Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Cognitive control persistently enhances hippocampal information processing

Nature volume 600, pages 484–488 (2021)Cite this article

Subjects

Abstract

Could learning that uses cognitive control to judiciously use relevant information while ignoring distractions generally improve brain function, beyond forming explicit memories? According to a neuroplasticity hypothesis for how some cognitive behavioural therapies are effective, cognitive control training (CCT) changes neural circuit information processing1,2,3. Here we investigated whether CCT persistently alters hippocampal neural circuit function. We show that mice learned and remembered a conditioned place avoidance during CCT that required ignoring irrelevant locations of shock. CCT facilitated learning new tasks in novel environments for several weeks, relative to unconditioned controls and control mice that avoided the same place during reduced distraction. CCT rapidly changes entorhinal cortex-to-dentate gyrus synaptic circuit function, resulting in an excitatory–inhibitory subcircuit change that persists for months. CCT increases inhibition that attenuates the dentate response to medial entorhinal cortical input, and through disinhibition, potentiates the response to strong inputs, pointing to overall signal-to-noise enhancement. These neurobiological findings support the neuroplasticity hypothesis that, as well as storing item–event associations, CCT persistently optimizes neural circuit information processing.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Cognitive control of spatial information processing in CA1.
Fig. 2: CCT facilitates subsequent learning.
Fig. 3: Cognitive control training persistently changes entorhinal-dentate  circuit function.
Fig. 4: Cognitive control training persistently changes inhibitory entorhinal-dentate circuit function.

Similar content being viewed by others

Data availability

The datasets are available from the corresponding author upon reasonable request. Source data are available at https://github.com/aafenton/Chung_etal_LearningToLearnData.git. Source data are provided with this paper.

Code availability

Code used for data processing and analysis is available from the corresponding author upon reasonable request at https://github.com/aafenton/Chung_etal_LearningToLearnCode.git.

References

  1. Subramaniam, K. et al. Computerized cognitive training restores neural activity within the reality monitoring network in schizophrenia. Neuron 73, 842–853 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Anguera, J. A. et al. Video game training enhances cognitive control in older adults. Nature 501, 97–101 (2013).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  3. Mishra, J., de Villers-Sidani, E., Merzenich, M. & Gazzaley, A. Adaptive training diminishes distractibility in aging across species. Neuron 84, 1091–1103 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Thomas, C. & Baker, C. I. Teaching an adult brain new tricks: a critical review of evidence for training-dependent structural plasticity in humans. Neuroimage 73, 225–236 (2013).

    Article  PubMed  Google Scholar 

  5. Alonso, A., van der Meij, J., Tse, D. & Genzel, L. Naïve to expert: Considering the role of previous knowledge in memory. Brain. Neurosci. Adv. 4, 2398212820948686 (2020).

    Article  PubMed  PubMed Central  Google Scholar 

  6. Yeager, D. S. et al. A national experiment reveals where a growth mindset improves achievement. Nature 573, 364–369 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  7. Bavelier, D., Green, C. S., Pouget, A. & Schrater, P. Brain plasticity through the life span: learning to learn and action video games. Annu. Rev. Neurosci. 35, 391–416 (2012).

    Article  CAS  PubMed  Google Scholar 

  8. Harlow, H. F. The formation of learning sets. Psychol. Rev. 56, 51–65 (1949).

    Article  CAS  PubMed  Google Scholar 

  9. Brewin, C. R. Theoretical foundations of cognitive-behavior therapy for anxiety and depression. Annu. Rev. Psychol. 47, 33–57 (1996).

    Article  CAS  PubMed  Google Scholar 

  10. de Villers-Sidani, E. et al. Recovery of functional and structural age-related changes in the rat primary auditory cortex with operant training. Proc. Natl Acad. Sci. USA 107, 13900–13905 (2010).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  11. Pavlowsky, A., Wallace, E., Fenton, A. A. & Alarcon, J. M. Persistent modifications of hippocampal synaptic function during remote spatial memory. Neurobiol. Learn. Mem. 138, 182–197 (2017).

    Article  PubMed  Google Scholar 

  12. Cimadevilla, J. M., Wesierska, M., Fenton, A. A. & Bures, J. Inactivating one hippocampus impairs avoidance of a stable room-defined place during dissociation of arena cues from room cues by rotation of the arena. Proc. Natl Acad. Sci. USA 98, 3531–3536 (2001).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  13. Hsieh, C. et al. Persistent increases of PKMzeta in memory-activated neurons trace LTP maintenance during spatial long-term memory storage. Eur. J. Neurosci. https://doi.org/10.1111/ejn.15137 (2021).

  14. Pastalkova, E. et al. Storage of spatial information by the maintenance mechanism of LTP. Science 313, 1141–1144 (2006).

    Article  ADS  CAS  PubMed  Google Scholar 

  15. Tsokas, P. et al. Compensation for PKMzeta in long-term potentiation and spatial long-term memory in mutant mice. eLife 5, e14846, (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  16. Talbot, Z. N. et al. Normal CA1 place fields but discoordinated network discharge in a Fmr1-null mouse model of fragile X syndrome. Neuron 97, 684–697 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. van Dijk, M. T. & Fenton, A. A. On how the dentate gyrus contributes to memory discrimination. Neuron 98, 832–845 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  18. Wesierska, M., Dockery, C. & Fenton, A. A. Beyond memory, navigation, and inhibition: behavioral evidence for hippocampus-dependent cognitive coordination in the rat. J. Neurosci. 25, 2413–2419 (2005).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Dvorak, D., Chung, A., Park, E. H. & Fenton, A. A. Dentate spikes and external control of hippocampal function. Cell Rep. 36, 109497 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Whishaw, I. Q. & Gorny, B. Path integration absent in scent-tracking fimbria-fornix rats: evidence for hippocampal involvement in “sense of direction” and “sense of distance” using self-movement cues. J. Neurosci. 19, 4662–4673 (1999).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Bures, J. et al. Dissociation of exteroceptive and idiothetic orientation cues: effect on hippocampal place cells and place navigation. Philos. Trans. R. Soc. Lond. B 352, 1515–1524 (1997).

    Article  ADS  CAS  Google Scholar 

  22. Jezek, K. et al. Stress-induced out-of-context activation of memory. PLoS Biol. 8, e1000570 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Garner, A. R. et al. Generation of a synthetic memory trace. Science 335, 1513–1516 (2012).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  24. Witter, M. P. The perforant path: projections from the entorhinal cortex to the dentate gyrus. Prog. Brain Res. 163, 43–61 (2007).

    Article  PubMed  Google Scholar 

  25. O’Reilly, K. C., Perica, M. I. & Fenton, A. A. Synaptic plasticity/dysplasticity, process memory and item memory in rodent models of mental dysfunction. Schizophr. Res. 207, 22–36 (2019).

    Article  PubMed  Google Scholar 

  26. Nitz, D. & McNaughton, B. Differential modulation of CA1 and dentate gyrus interneurons during exploration of novel environments. J. Neurophysiol. 91, 863–872 (2004).

    Article  PubMed  Google Scholar 

  27. Madar, A. D., Ewell, L. A. & Jones, M. V. Temporal pattern separation in hippocampal neurons through multiplexed neural codes. PLoS Comput. Biol. 15, e1006932 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  28. Mercer, L. F., Jr., Remley, N. R. & Gilman, D. P. Effects of urethane on hippocampal unit activity in the rat. Brain Res. Bull. 3, 567–570 (1978).

    Article  CAS  PubMed  Google Scholar 

  29. Ferrante, M., Migliore, M. & Ascoli, G. A. Feed-forward inhibition as a buffer of the neuronal input-output relation. Proc. Natl Acad. Sci. USA 106, 18004–18009 (2009).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  30. Vasuta, C. et al. Metaplastic regulation of CA1 Schaffer collateral pathway plasticity by Hebbian MGluR1a-mediated plasticity at excitatory synapses onto somatostatin-expressing interneurons. eNeuro 2, ENEURO.0051-15.2015 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  31. Donato, F., Rompani, S. B. & Caroni, P. Parvalbumin-expressing basket-cell network plasticity induced by experience regulates adult learning. Nature 504, 272–276 (2013).

    Article  ADS  CAS  PubMed  Google Scholar 

  32. Letzkus, J. J., Wolff, S. B. & Luthi, A. Disinhibition, a circuit mechanism for associative learning and memory. Neuron 88, 264–276 (2015).

    Article  CAS  PubMed  Google Scholar 

  33. Ling, D. S. et al. Protein kinase Mzeta is necessary and sufficient for LTP maintenance. Nat. Neurosci. 5, 295–296 (2002).

    Article  CAS  PubMed  Google Scholar 

  34. Takeuchi, T., Duszkiewicz, A. J. & Morris, R. G. The synaptic plasticity and memory hypothesis: encoding, storage and persistence. Philos. Trans. R. Soc. Lond. B 369, 20130288 (2013).

    Article  Google Scholar 

  35. Lee, J. W. & Jung, M. W. Separation or binding? Role of the dentate gyrus in hippocampal mnemonic processing. Neurosci. Biobehav. Rev. 75, 183–194 (2017).

    Article  PubMed  Google Scholar 

  36. Lee, J. W., Kim, W. R., Sun, W. & Jung, M. W. Role of dentate gyrus in aligning internal spatial map to external landmark. Learn. Mem. 16, 530–536 (2009).

    Article  PubMed  Google Scholar 

  37. Luna, V. M. et al. Adult-born hippocampal neurons bidirectionally modulate entorhinal inputs into the dentate gyrus. Science 364, 578–583 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  38. Lee, H. et al. Early cognitive experience prevents adult deficits in a neurodevelopmental schizophrenia model. Neuron 75, 714–724 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Aharoni, D. & Hoogland, T. M. Circuit investigations with open-source miniaturized microscopes: past, present and future. Front. Cell Neurosci. 13, 141 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  40. 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).

    Article  CAS  Google Scholar 

  41. Pnevmatikakis, E. A. et al. Simultaneous denoising, deconvolution, and demixing of calcium imaging data. Neuron 89, 285–299 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Zhou, P. et al. Efficient and accurate extraction of in vivo calcium signals from microendoscopic video data. eLife 7, e28728 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  43. Huang, L. et al. Relationship between simultaneously recorded spiking activity and fluorescence signal in GCaMP6 transgenic mice. eLife 10, e51675 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Olypher, A. V., Lansky, P., Muller, R. U. & Fenton, A. A. Quantifying location-specific information in the discharge of rat hippocampal place cells. J. Neurosci. Methods 127, 123–135 (2003).

    Article  CAS  PubMed  Google Scholar 

  45. Franklin, K. B. J. & Paxinos, G. The Mouse Brain in Stereotaxic Coordinates 2nd edn (Academic Press, 2001).

  46. Harrison, R. R. et al. A Low-power integrated circuit for a wireless 100-electrode neural recording system. IEEE J. Solid State Circuits 42, 123–133 (2007).

    Article  ADS  Google Scholar 

  47. Hrabetova, S. Extracellular diffusion is fast and isotropic in the stratum radiatum of hippocampal CA1 region in rat brain slices. Hippocampus 15, 441–450 (2005).

    Article  PubMed  Google Scholar 

  48. Holsheimer, J. Electrical conductivity of the hippocampal CA1 layers and application to current-source-density analysis. Exp. Brain Res. 67, 402–410 (1987).

    Article  CAS  PubMed  Google Scholar 

  49. Pettersen, K. H., Devor, A., Ulbert, I., Dale, A. M. & Einevoll, G. T. Current-source density estimation based on inversion of electrostatic forward solution: effects of finite extent of neuronal activity and conductivity discontinuities. J. Neurosci. Methods 154, 116–133 (2006).

    Article  PubMed  Google Scholar 

  50. Penttonen, M., Kamondi, A., Sik, A., Acsady, L. & Buzsaki, G. Feed-forward and feed-back activation of the dentate gyrus in vivo during dentate spikes and sharp wave bursts. Hippocampus 7, 437–450 (1997).

    Article  CAS  PubMed  Google Scholar 

  51. Bragin, A., Jando, G., Nadasdy, Z., van Landeghem, M. & Buzsaki, G. Dentate EEG spikes and associated interneuronal population bursts in the hippocampal hilar region of the rat. J. Neurophysiol. 73, 1691–1705 (1995).

    Article  CAS  PubMed  Google Scholar 

  52. Kanter, B. R. et al. A novel mechanism for the grid-to-place cell transformation revealed by transgenic depolarization of medial entorhinal cortex layer II. Neuron 93, 1480–1492.e1486 (2017).

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

Supported by NIH grants R01MH115304, R01NS105472, and R01AG043688. We thank T. Sacktor for valuable discussions.

Author information

Authors and Affiliations

  1. Center for Neural Science, New York University, New York City, USA

    Ain Chung, Alejandro Grau-Perales, Eliott R. J. Levy, Dino Dvorak, Nida Hussain & André A. Fenton

  2. Department of Psychology, Hunter College, City University of New York, New York City, USA

    Claudia Jou

  3. Neuroscience Institute at the NYU Langone Medical Center, New York City, USA

    André A. Fenton

Authors
  1. Ain Chung
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Claudia Jou
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Alejandro Grau-Perales
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Eliott R. J. Levy
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Dino Dvorak
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Nida Hussain
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. André A. Fenton
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

A.A.F. and A.C. designed experiments. E.L. and A.G.-P. performed the calcium imaging experiments and analyses. A.C., C.J., D.D. and A.G.-P. performed behavioural experiments and analyses. A.C. and D.D. performed in vivo physiology experiments and analyses; A.C. and C.J. performed ex vivo physiology experiments and analyses; A.C. and A.G.-P. performed immunohistochemistry experiments. N.H. contributed to behaviour and preparation for in vivo physiology experiments. A.A.F., A.C., C.J., E.L. and D.D. analysed and interpreted the results. A.A.F. and A.C. wrote the manuscript with contributions from C.J., D.D., and E.L. A.A.F. directed and obtained research funding.

Corresponding author

Correspondence to André A. Fenton.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks the anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Active place avoidance and control task behaviors.

a) Task variants: The apparatus and environment were essentially the same for the three task variants. More than 99% of the time in the apparatus, when there was no shock, the visual environment was identical across the tasks; only the place learning environment was different in that the arena surface was covered in shallow water. During the 500-ms shock the cognitive control training and place learning groups experienced an unpleasant foot shock (depicted as red shaded sectors), whereas the spatial exploration group did not. b) Example paths during select trials of the same representative mice; dots indicate the mouse’s location when it was shocked (red), and when it would have been shocked if the shock was on (black). c) The CCT (n=12), PL (n=11), and SE (N=9) groups differ in the initial but not the subsequent pretraining trials with no shock, because mice walk less in the shallow water (Group x Trial two-way RM ANOVA; group: F2,27 = 27.7, p = 10−7; η2 = 0.49; Trial: F1,50.5 = 2.44, p = 0.12; interaction: F2,50.5 = 5.88, p = 0.005, η2 = 0.19; PL during initial < all other group x trials measures). This resulted in group differences in the number of times the mice enter the location of the future shock zone (Group x Trial two-way RM ANOVA, Group: F2,31 = 10.51, p = 10−4, η2 = 0.22; Trial: F1,60.7 = 0.042, p = 0.8; Group X Trial: F2,60.7 = 2.93, p = 0.06; Tukey-HSD: ***CCT = SE > PL Consequently, performance in the initial and subsequent training sessions is analyzed separately. The group average performance measured during 30-min by entrances into the shock zone or equivalent area for the SE mice. Statistical analysis of subsequent training (see Supplemental Information for analysis of initial training; *p<0.05): Group X Trials1-3 two-way RM ANOVA, Group: F2,30 = 0.04; p = 0.97; Trials: F2, 56.8 = 12.3, p = 10−5, η2 = 0.19, Tukey-HSD: Ts1 > Ts2 = Ts3; Interaction; F4,63.2 = 0.93, p = 0.46). Place avoidance learning measured as decreasing entrances into the shock zone was robust with significant effects of Group and the Group X Trial interaction (Group: F2,45.6 = 72.3, p = 10−11, η2 = 0.23; Trial: F2,34.7 = 4.63, p = 0.02, η2 = 0.006; Interaction: F4,39.3 = 8.31, p = 10−5, η2 = 0.01; ***Tukey-HSD: SE > CCT = PL). The two conditioned groups showed similar learning (Group: F1,59.2 = 0.93, p = 0.34, η2 = 0.004; Trial: F2,33.5 = 39.5, p = 10−9, η2 = 0.20, Tukey-HSD Trial1 > Trial 2 > Trial 3; Interaction: F2,33.5 = 1.57, p = 0.2; η2 = 0.01). The groups differed on the 1-week retention test (F2,28 = 3.57, p = 0.04, η2 = 0.21). Tukey-HSD comparisons of groups: *p < 0.05. d) Total path walked measures activity: During initial training the conditioned mice walk less and restrict themselves to the arena periphery compared to the SE mice; the conditioned mice did not differ in the distance they walked (Group x Trial two-way RM ANOVA, Group: F2,41.2 = 11.51, p = 10−4, η2 = 0.14; Trial: F2,28.5 = 1.16, p = 0.33, η2 = 0.002; Group X Trial: F4,32.4 = 5.83, p = 0.001, η2 = 0.03; SE > CCT = PL). e) Time to 1st Entrance estimates between-session memory. One-week memory retention was robust with an obvious effect of Group measured by entrances into the shock zone (see Fig. 2b; F2,31 = 101.3, p = 10−14, η2 = 0.73, ***Tukey-HSD: SE > CCT = PL) as well as the time to first enter the shock zone in the two conditioned groups with PL better than CCT (Group: F1,31 = 3.25, *p = 0.05, η2 = 0.21). Line plot: mean±s.e.m.

Source data

Extended Data Fig. 2 Active place avoidance in a novel environment is improved by prior CCT training.

a left) Times to enter the shock zone for the first five times during subsequent training in entirely novel physical conditions (Ts1) are prolonged after initial training (Ti1) in the CCT group (n=12) comparison of slopes: F1,116 = 15.91, p = 0.0001, η2 = 0.12). a middle) CCT mice also prolonged entering the shock zone during subsequent training more than the other groups (comparison of slopes: F2,154 = 5.94, p = 0.003, η2 = 0.07). a right) PL (n=11) and SE (n=9) groups performed indistinguishably during their Ts1 from the CCT Ti1 (comparison of slopes: F2,154 = 2.60, p = 0.07). Solid lines depict linear regressions, and dotted lines depict 95% confidence interval bounds. b) The number of shock zone entrances during the first 5 min of CCT training. CCT Ti1 vs Ts1: paired t11 = 2.58, 1-tailed *p = 0.01, d = 0.75. CCT Ts1 vs PL Ts1: t21 = 2.48, 1-tailed *p = 0.01, d = 1.04; CCT Ts1 vs SE: t19 = 2.46, 1-tailed *p = 0.01, d = 1.09), which are all significant after Bonferroni corrections for the multiple (3) tests. CCT (n=12), PL (n=11), and SE (N=9). Box plot: 25-75%, median, whiskers: min/max. c) Schematic training protocol. d) Learned place avoidance during initial training predicts subsequent place avoidance in a novel environment: If place avoidance during initial training facilitates learning during subsequent training, then mice that express better learned place avoidance in initial training (measured as performance in RET) are predicted to show better performance during the early stages of subsequent training (Ts1). One-week retention of the conditioned place avoidance significantly predicts place avoidance in a novel environment amongst the CCT (left, n=12) but not the PL (right, n=11) mice. e) Similarly, if initial CCT facilitates subsequent learning in the novel environment, this predicts that if mice perform well during the early stages of subsequent training (Ts1) then they would have learned well as estimated by performance at the end of initial training (Ti3). Correlation between initial training trials (Ti1-3) and during subsequent (Ts1) training increased from Ti1 to Ti3 in the CCT group, but not within the PL group. Drawing Source: Ain Chung, André Fenton.

Source data

Extended Data Fig. 3 Targeting electrodes to investigate pathway-specific hippocampal responses to medial perforant path simulation.

a–c) Muscimol inactivation demonstrates DG is crucial for expression of active place avoidance several days after CCT training: a) Experimental design (1) and the tracks of an exemplar muscimol-injected mouse during each behavioral session (2). b) Schematic from Franklin and Paxinos45 illustrating the injection and guide cannulae target, a histology section of the cannulation track, as well as a histological section after injecting fluorogold (FG) to estimate the infusion spread (scale bar: 500μm). c) A measure of memory expression illustrates that targeting muscimol at DG reversibly impairs established active place avoidance memory (n = 5/group, Group X Retention session RM ANOVA Group: F1,8 = 2.46, p = 0.16; Retention session: F2,16 = 5.84, p = 0.01, η2 = 0.42; interaction: F2,16 = 4.01, p = 0.03, η2 = 0.33, two-sided Šidak’s multiple comparisons RET2 < RET1 = RET3 in Musimol injected group, Musimol vs PBS in RET2, ** p <0.01. Box plot: 25-75%, median, whiskers: min/max. d) Targeting of the recording and stimulating electrodes: Schematic and Nissl-stained histological sections from an exemplar mouse illustrating the left) recording and right) stimulation electrode sites. (Scale bars: 1mm) Schematics from Franklin and Paxinos45. e) Membrane potential vs. current source density (CSD) plots to illustrate signal localization: comparison of the raw field potential (mV) plots and corresponding CSD plots along corresponding channels of the linear silicon probe. Passive volume conduction strongly influences the signal recorded at adjacent electrodes in the field potential plots. Some examples are indicated by the red/blue boxes. 1: signal artifacts in CA1; 2: signal artifact at the hippocampal fissure; 3: volume conduction from MPP responses that occlude LPP responses; 4: volume conduction from the granule cell layer response occluding hilar responses. f) Representative sink and source signals and their locations from the CSD plots. g–j) Confirmation that electrical stimulation targeted the MPP. g) Nissl section and anti-human antigen (HA) immunostaining of transgenic mice (TRE-hM4Di-tetO+/- crossed with Ent-tTA+/-) that express the inhibitory DREADD under control of doxycycline withdrawal. The inhibitory DREADD immunoreactivity localizes to hippocampus-projecting MECII neurons. Immunofluorescence shows expression in MECII cell bodies, in the perforant path axons and in their termination zones in the molecular layer of the dentate gyrus and CA3. Expression is faint in CA1 str. lacunosum moleculare. Expression is in ~20% of MECII stellate cells and 5% in MECIII cells with less in adjacent regions (Kanter et al. 2017). h) Experimental design to evaluate effective stimulation of the MPP. i) Average evoked responses recorded with a 16-site linear silicon probe spanning the somato-dendritic axis of dorsal hippocampus. j) Input-output curves of fEPSP slope responses recorded from the DG molecular layer and population spike responses recorded from the DG hilus. These data quantify that responses to electrical stimulation targeted to the angular bundle is dose-dependently and reversibly suppressed by CNO-mediated activation of hM4Di in the MPP (Area X Dose RM ANOVA, Area: F1,8 = 7.55, p = 0.02, η2 = 0.49; Dose: F4,32 = 168.6 p = 10−21, η2 = 0.95; interaction: F4,32 = 5.78, p = 0.001, η2 = 0.42, two-sided Šidak-corrected within-area comparisons ***p<0.001). Box plot: 25-75%, median, whiskers: min/max. Line plot: mean±s.e.m. Drawing Source: Ain Chung, André Fenton.

Source data

Extended Data Fig. 4 Additional active place avoidance training does not incrementally change the response to medial perforant path stimulation.

a) Experimental design, b) Learning curves comparing initial CCT training, and subsequent CCT training to either a conflicting location of shock (Conflict CCT), or to a novel location of shock in a novel room (subsequent CCT). (Number of Entrances: Group X Trial RM ANOVA Group: F2,34 = 2.95; p = 0.07; Trial: F3,99 = 22.56, p = 10−5, η2 = 0.41; Interaction; F6,99 = 5.98, p = 10−5, η2 = 0.27, Time to 1st Entrance: Group X Trial RM ANOVA Group: : F2,34 = 4.42, p = 0.02, η2 = 0.21 Trial: F3,99 = 13.57; p = 10−5, η2 = 0.29; Interaction; F6,99 = 1.581, p = 0.16). (initial and subsequent CCT n=12, conflict group n=11). Line plot: mean±SEM. c) CCT (n=7) caused changes at the corresponding middle molecular layers of the supraDG but not the infraDG response to MPP stimulation (Area X Training RM ANOVA, area: F1,12 = 0.49, p = 0.50; training: F1,12 = 11.08, p = 0.006, η2 = 0.48; interaction: F1,12 = 6.69, p = 0.02, η2 = 0.36; Šidak’s multiple comparisons: only SupraDG changes circuit function **p = 0.003; supraDG was significant paired t6 = 5.09, **p = 0.002, d = 1.92) but not at infraDG sites (paired t6 = 0.46, p = 0.66). Scale bar: 50 μA/mm3, 5 ms. d) No further changes are observed in the synaptic responses after either additional CCT in a third, novel environment (n=5, supraDG: paired t4 = 0.45, p = 0.67; infraDG paired t4 = 0.12, p = 0.90) or in the same environment with a 180°-relocated shock zone that conflicts with the prior location of shock (n=5, supraDG paired t4 = 0.02, p = 0.98; infraDG paired t4 = 1.13, p = 0.32) The 4 mice/group that received both types of subsequent training, received conflict or novel environment training, in counterbalanced order. Box plot: 25-75%, median, whiskers: min/max. Line plot: mean±s.e.m. Drawing Source: Ain Chung, André Fenton.

Source data

Extended Data Fig. 5 Stability of electrode properties and placement, and time course of the learning-induced changes in the response to medial perforant path stimulation.

a–c) Stability of stimulation and recording electrode properties across the duration of the experiment. a) Average power of spontaneous theta oscillations measured at the molecular-hilar region of supraDG and b) stimulation electrode impedance, both measured during wheel running before active place avoidance training (see schematic at top). Neither theta power (n=8, F2.4, 21.6 = 0.65, p = 0.55) nor electrode impedance in 32 channels (F4,128= 0.294, p = 0.9) changed across the experiment, making it unlikely that changed electrode properties account for the CCT-induced differences in evoked responses. Box plot: 25-75%, median, whiskers: min/max. c) Session-averaged current source density (CSD) of LPP-triggered DSL (type 1) and MPP-triggered DSM (type2) spontaneous events also illustrate recording mechanical stability across the experiment. d) Time course of CCT-induced changes in MPP-DG synaptic responses. fEPSP slope was normalized to the baseline slope prior to exposure to the behavioral arena. The decrease of the fEPSP response is observed within 2 h of the first CCT trial. * indicates a significant difference from before to after a CCT trial (n = 7, Paired t-test: t6 = 3.40, *p = 0.01, d = 1.28) Box plot: 25-75%, median, whiskers: min/max. e) CCT causes place avoidance and reduced population synaptic responses to MPP stimulation that persist at least 60 days. Left) Conditioned place avoidance (n = 3, F1.109,2.038 = 31.07, p = 0.03, η2 = 0.94, pretraining > 1-wk retention = 60-d retention) and right) population synaptic responses (F1.015,2.030 = 19.30, p = 0.05, η2 = 0.90, baseline > 1-wk = 60-d, Tukey-HSD *p<0.05). Drawing Source: Ain Chung, André Fenton.

Source data

Extended Data Fig. 6 Naturally occurring, spontaneous dentate spikes originating from MPP activation confirm the findings using evoked responses to MPP stimulation.

a) A dentate spike (DS, stars) is the DG response to synchronous input from the medial and lateral perforant path, making it a physiological estimate of the MPP-DG connection that is independent of artificial stimulation. DS were identified in the DG hilus as sharp, positive waves of the local field potential with prominence (distance between peak and closest preceding or following trough) greater than 2 S.D. of all detected positive peaks, as well as width between 2.5 and 12.5 ms measured at 50% of the peak’s prominence. b) Two types of dentate spikes – DSL (type 1) and DSM (type 2) were identified using a CSD fingerprinting method, where all DS that exhibited a symmetric pair of current sinks in the distal molecular layers of DG were identified as DSL and DS that exhibited a symmetric pair of current sinks in the medial/proximal molecular layers of DG were identified as DSM. The average CSD profiles of DSL (left) and DSM (right) computed from all classified DS events with clearly distinct pairs of current sinks in the distal and medial/proximal molecular layers of DG respectively. c) Average CSD profiles of DS-associated current sinks in molecular layers of DG in supraDG (top row) and infraDG (bottom row) triggered by DSL (left column) and DSM (right column). Black and red colors represent before and after CCT training, respectively. c) Summary comparisons of maximum gradient of the DS sink (i.e. negative slope) measured before and after CCT. CCT only changed DSM (type 2) at the supraDG site but not at the infraDG site, confirming findings assessed by stimulus evoked responses (n=5, Paired t-test: t4 = 3.04, *p = 0.04, d = 1.36). Box plot: 25-75%, median, whiskers: min/max.

Source data

Extended Data Fig. 7 Characterization, hypotheses, and validation of experiments to assess the mechanism of the attenuated response to MPP stimulation after CCT.

a) Gad2-expressing interneurons are more abundant in supraDG than infraDG. Left) eYFP expressing interneurons in Gad2Cre-eYFP mice (scale bar 0.2 mm); Right) cell counts per section in mice after CCT (n=6) or SE (n=4) experience document enrichment of inhibitory interneurons in supraDG relative to infraDG (F1.215,10.93 = 21.40, p < 0.001, η2 = 0.69, SupraDG > InfraDG = Hilus, Tukey-HSD: ***p = 0.0008, **p = 0.003). Orange and green arrows indicate eYFP+ cells in the supraDG and infraDG, respectively. Box plot: 25-75%, median, whiskers: min/max. b) Two hypotheses to account for the CCT-induced suppression of DG responses to MPP stimulation: Schematic of MPP→granule cell (GC) depression and potentiation of GC inhibition hypotheses to explain the CCT-induced reduction of MPP response. c–i) The in vivo experiment under urethane anesthesia to investigate the role of inhibition in the CCT-induced effects. c) Experimental design, d) schematic of the preparation, e) photo illustrating the recording electrodes and optical stimulation, and f) The schematic is superimposed on a histological section of a Gad2-Cre-eYFP mouse’s dorsal hippocampus with eYFP (green) and immunofluorescence for DAPI (blue), to indicate the circuit that was targeted (scale bar 0.02mm). e) Atlas from Franklin and Paxinos45 and g) corresponding histological sections illustrating electrode and optical fiber placements (scale bar:1mm). Sixteen-channel linear electrode array recordings of LFPs across the somatodendritic axis of dorsal hippocampus h) under urethane anesthesia and i) in the freely-behaving mouse. Under urethane, LFPs reflecting rhythmic ongoing synaptic activity is much attenuated compared to in the behaving mouse. Voltage scale (red bar) = 1 mV.

Source data

Extended Data Fig. 8 validation of optogenetic activation of inhibition to evaluate whether altered inhibition contributes to the CCT-induced changes of the response to MPP stimulation.

a) Population spike responses are blocked by activating ChR2 in Gad2-expressing cells 5 ms before MPP stimulation in vivo under urethane anesthesia, regardless of prior training. Input-output relationships of the response to MPP stimulation with and without prior light activation of inhibitory-neuron ChR2 in the (upper) supraDG and (lower) infraDG. This demonstrates light-stimulated inhibition is effective in all groups. home cage (n = 5), SE (n = 6), PL (n = 4), and CCT mice (n = 7) b) fEPSP evoked responses at supraDG are reduced by activating Gad2-expressing cells at the same time as MPP stimulation in vivo, regardless of prior training. Upper) Input-output relationships of the response to MPP stimulation with and without concurrent light activation of inhibitory-neuron ChR2. Lower) The corresponding summaries represented as the area-under-the-curve (AUC) for statistical analysis of the group and light stimulation effects. Group X Stimulation ANOVA, Group: F3,13 = 0.80, p = 0.51; Stimulation: F1,13 = 16.05, p = 0.0015, η2 = 0.55; interaction: F2,10 = 0.4921, p = 0.70). home cage (n = 5), SE (n = 4), PL (n = 4), and CCT mice (n = 4) Box plot: 25-75%, median, whiskers: min/max. Line plot: mean±s.e.m.

Source data

Extended Data Fig. 9 in vivo paired-pulse inhibition is weaker after CCT.

a) CSD examples of single responses to paired-pulse inhibition in CCT, homecage, SE and PL mice (Scale bars: 100 μA/mm3, 5 ms); MPP-stimulated inhibition is attenuated at supraDG of CCT mice. b) Summary PPI curves and indices (supraDG: F3,18 = 3.22, *p = 0.04, η2 = 0.15; infraDG: F3,18 = 0.95, p = 0.43). b) Responses to MPP stimulation were recorded in urethane-anesthetized mice at the supra and infra granule cell layers of the DG. Significant changes were not detected although there is a hint of enhanced facilitation in the CCT group. (supraDG granule cell Group: F3,18 = 0.47, p = 0.70; InfraDG granule cell group: F3,18 = 0.19, p = 0.90). home cage (n = 5), SE (n = 6), PL (n = 4), and CCT mice (n = 7) Box plot: 25-75%, median, whiskers: min/max. Line plot: mean±s.e.m.

Source data

Extended Data Fig. 10 Validation of MPP-stimulation related optogenetic terminal activation of inhibition, and experience-related changes of MPP-stimulated responses in spontaneously-inactive ex vivo hippocampus slices.

a) Experimental design and photomicrograph of ex vivo hippocampus slice prepared 1-week posttraining (scale bar: 200μm). fEPSP responses to MPP terminal stimulation recorded with light-activated ChR2 in Gad2+ cells. b) Response recorded in infraDG to MPP stimulation with ChR2 activation of inhibitory neurons at different time offsets. left: infraDG (2-way RM ANOVA Group: F2,17 = 0.50, p = 0.62, Light-Stim ISI F2.06,34.95 = 2.82, p = 0.73; interaction: F10,85 = 1.15, p = 0.34) right: Light 2ms before (F2,17 = 1.14, p = 0.34). SE (n = 6), PL (n = 8), and CCT mice (n = 6) c) Input-output relationship of the response to MPP stimulation in ex vivo hippocampal slices. I-O curves (left) and their summary as the area-under-the-curve (AUC, right) upon which statistical group comparisons were made. (Upper) SupraDG: IO-AUC F3,32 = 3.80, p = 0.02, η2 = 0.26, Šidak’s multiple comparisons: HC vs PL *p = 0.021. (Lower) InfraDG: IO-AUC F3,31 = 0.93, p = 0.44 homecage (n = 6) SE (n = 10), PL (SupraDG n = 8, infraDG n = 9), and CCT mice (n = 10) Box plot: 25-75%, median, whiskers: min/max. Line plot: mean±s.e.m.

Source data

Supplementary information

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chung, A., Jou, C., Grau-Perales, A. et al. Cognitive control persistently enhances hippocampal information processing. Nature 600, 484–488 (2021). https://doi.org/10.1038/s41586-021-04070-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41586-021-04070-5

This article is cited by

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing