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:

Prefrontal and lateral entorhinal neurons co-dependently learn item–outcome rules

Nature (2024)Cite this article

Subjects

Abstract

The ability to learn novel items depends on brain functions that store information about items classified by their associated meanings and outcomes1,2,3,4, but the underlying neural circuit mechanisms of this process remain poorly understood. Here we show that deep layers of the lateral entorhinal cortex (LEC) contain two groups of ‘item–outcome neurons’: one developing activity for rewarded items during learning, and another for punished items. As mice learned an olfactory item–outcome association, we found that the neuronal population of LEC layers 5/6 (LECL5/6) formed an internal map of pre-learned and novel items, classified into dichotomic rewarded versus punished groups. Neurons in the medial prefrontal cortex (mPFC), which form a bidirectional loop circuit with LECL5/6, developed an equivalent item–outcome rule map during learning. When LECL5/6 neurons were optogenetically inhibited, tangled mPFC representations of novel items failed to split into rewarded versus punished groups, impairing new learning by mice. Conversely, when mPFC neurons were inhibited, LECL5/6 representations of individual items were held completely separate, disrupting both learning and retrieval of associations. These results suggest that LECL5/6 neurons and mPFC neurons co-dependently encode item memory as a map of associated outcome rules.

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

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to full article PDF
Buy now

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

Fig. 1: LECL5/6 cells encode outcome rules during associative learning.
Fig. 2: mPFC cells encode outcome rules concurrently with LECL5/6 cells during associative learning.
Fig. 3: LECL5/6 inhibition impairs mPFC outcome rules and new learning by mice.
Fig. 4: mPFC inhibition disrupts LECL5/6 outcome rules, memory acquisition and retrieval.

Similar content being viewed by others

Data availability

The neurophysiological data generated in this study are available on request.

Code availability

The neurophysiological data and analytical codes are available on request and will be deposited with a subsequent protocol paper.

References

  1. Suzuki, W. A. Associative learning signals in the brain. Prog. Brain Res. 169, 305–320 (2008).

    Article  PubMed  Google Scholar 

  2. Osada, T., Adachi, Y., Kimura, H. M. & Miyashita, Y. Towards understanding of the cortical network underlying associative memory. Phil. Trans. R. Soc. B 363, 2187–2199 (2008).

    Article  PubMed  PubMed Central  Google Scholar 

  3. Ozawa, T. & Johansen, J. P. Learning rules for aversive associative memory formation. Curr. Opin. Neurobiol. 49, 148–157 (2018).

    Article  CAS  PubMed  Google Scholar 

  4. Igarashi, K. M., Lee, J. Y. & Jun, H. Reconciling neuronal representations of schema, abstract task structure, and categorization under cognitive maps in the entorhinal–hippocampal–frontal circuits. Curr. Opin. Neurobiol. 77, 102641 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Squire, L. R. Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans. Psychol. Rev. 99, 195–231 (1992).

    Article  CAS  PubMed  Google Scholar 

  6. Buzsaki, G. & Moser, E. I. Memory, navigation and theta rhythm in the hippocampal–entorhinal system. Nat. Neurosci. 16, 130–138 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Morris, R. G. in The Hippocampus Book (ed. P. Andersen, P.) 581–714 (Oxford Univ. Press, 2007).

  8. Eichenbaum, H. On the integration of space, time, and memory. Neuron 95, 1007–1018 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. O’Keefe, J. & Nadel, L. The Hippocampus as a Cognitive Map (Oxford Univ. Press, 1978).

  10. Moser, E. I., Moser, M. B. & Roudi, Y. Network mechanisms of grid cells. Phil. Trans. R. Soc. B 369, 20120511 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  11. Price, J. L. An autoradiographic study of complementary laminar patterns of termination of afferent fibers to the olfactory cortex. J. Comp. Neurol. 150, 87–108 (1973).

    Article  CAS  PubMed  Google Scholar 

  12. Burwell, R. D. The parahippocampal region: corticocortical connectivity. Ann. N. Y. Acad. Sci. 911, 25–42 (2000).

    Article  ADS  CAS  PubMed  Google Scholar 

  13. Igarashi, K. M. et al. Parallel mitral and tufted cell pathways route distinct odor information to different targets in the olfactory cortex. J. Neurosci. 32, 7970–7985 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Young, B. J., Otto, T., Fox, G. D. & Eichenbaum, H. Memory representation within the parahippocampal region. J. Neurosci. 17, 5183–5195 (1997).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Deshmukh, S. S. & Knierim, J. J. Representation of non-spatial and spatial information in the lateral entorhinal cortex. Front. Behav. Neurosci. 5, 69 (2011).

    Article  PubMed  PubMed Central  Google Scholar 

  16. Tsao, A., Moser, M. B. & Moser, E. I. Traces of experience in the lateral entorhinal cortex. Curr. Biol. 23, 399–405 (2013).

    Article  CAS  PubMed  Google Scholar 

  17. Igarashi, K. M., Lu, L., Colgin, L. L., Moser, M. B. & Moser, E. I. Coordination of entorhinal–hippocampal ensemble activity during associative learning. Nature 510, 143–147 (2014).

    Article  ADS  CAS  PubMed  Google Scholar 

  18. Lee, J. Y. et al. Dopamine facilitates associative memory encoding in the entorhinal cortex. Nature 598, 321–326 (2021).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  19. Martin, C., Beshel, J. & Kay, L. M. An olfacto-hippocampal network is dynamically involved in odor-discrimination learning. J. Neurophysiol. 98, 2196–2205 (2007).

    Article  PubMed  Google Scholar 

  20. Cohen, J. Y., Haesler, S., Vong, L., Lowell, B. B. & Uchida, N. Neuron-type-specific signals for reward and punishment in the ventral tegmental area. Nature 482, 85–88 (2012).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  21. Luo, W. et al. Acquiring new memories in neocortex of hippocampal-lesioned mice. Nat. Commun. 13, 1601 (2022).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  22. Insausti, R., Herrero, M. T. & Witter, M. P. Entorhinal cortex of the rat: cytoarchitectonic subdivisions and the origin and distribution of cortical efferents. Hippocampus 7, 146–183 (1997).

    Article  CAS  PubMed  Google Scholar 

  23. Witter, M. P. & Amaral, D. G. in The Rat Nervous System 3rd edn (ed. Paxinos, G.) 635–704 (Elsevier, 2004).

  24. Zingg, B. et al. Neural networks of the mouse neocortex. Cell 156, 1096–1111 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Jun, H. et al. Disrupted place cell remapping and impaired grid cells in a knockin model of Alzheimer’s disease. Neuron 107, 1095–1112.e6 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Issa, J. B., Radvansky, B. A., Xuan, F. & Dombeck, D. A. Lateral entorhinal cortex subpopulations represent experiential epochs surrounding reward. Nat. Neurosci. 27, 536–546 (2024).

    Article  CAS  PubMed  Google Scholar 

  27. Mulder, A. B., Nordquist, R., Orgut, O. & Pennartz, C. M. Plasticity of neuronal firing in deep layers of the medial prefrontal cortex in rats engaged in operant conditioning. Prog. Brain Res. 126, 287–301 (2000).

    Article  CAS  PubMed  Google Scholar 

  28. Rushworth, M. F., Noonan, M. P., Boorman, E. D., Walton, M. E. & Behrens, T. E. Frontal cortex and reward-guided learning and decision-making. Neuron 70, 1054–1069 (2011).

    Article  CAS  PubMed  Google Scholar 

  29. Euston, D. R., Gruber, A. J. & McNaughton, B. L. The role of medial prefrontal cortex in memory and decision making. Neuron 76, 1057–1070 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Anastasiades, P. G. & Carter, A. G. Circuit organization of the rodent medial prefrontal cortex. Trends Neurosci. 44, 550–563 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Lillicrap, T. P., Santoro, A., Marris, L., Akerman, C. J. & Hinton, G. Backpropagation and the brain. Nat. Rev. Neurosci. 21, 335–346 (2020).

    Article  CAS  PubMed  Google Scholar 

  32. Konishi, M. I., Igarashi, K. M. & Miura, K. Biologically plausible local synaptic learning rules robustly implement deep supervised learning. Front. Neurosci. 17, 1160899 (2023).

    Article  PubMed  PubMed Central  Google Scholar 

  33. Hasegawa, I., Fukushima, T., Ihara, T. & Miyashita, Y. Callosal window between prefrontal cortices: cognitive interaction to retrieve long-term memory. Science 281, 814–818 (1998).

    Article  ADS  CAS  PubMed  Google Scholar 

  34. Tomita, H., Ohbayashi, M., Nakahara, K., Hasegawa, I. & Miyashita, Y. Top-down signal from prefrontal cortex in executive control of memory retrieval. Nature 401, 699–703 (1999).

    Article  ADS  CAS  PubMed  Google Scholar 

  35. Frankland, P. W. & Bontempi, B. The organization of recent and remote memories. Nat. Rev. Neurosci. 6, 119–130 (2005).

    Article  CAS  PubMed  Google Scholar 

  36. Tse, D. et al. Schema-dependent gene activation and memory encoding in neocortex. Science 333, 891–895 (2011).

    Article  ADS  CAS  PubMed  Google Scholar 

  37. Eichenbaum, H. Prefrontal–hippocampal interactions in episodic memory. Nat. Rev. Neurosci. 18, 547–558 (2017).

    Article  CAS  PubMed  Google Scholar 

  38. Kitamura, T. et al. Engrams and circuits crucial for systems consolidation of a memory. Science 356, 73–78 (2017).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  39. Tse, D. et al. Schemas and memory consolidation. Science 316, 76–82 (2007).

    Article  ADS  CAS  PubMed  Google Scholar 

  40. Baraduc, P., Duhamel, J. R. & Wirth, S. Schema cells in the macaque hippocampus. Science 363, 635–639 (2019).

    Article  ADS  CAS  PubMed  Google Scholar 

  41. Simons, J. S. & Spiers, H. J. Prefrontal and medial temporal lobe interactions in long-term memory. Nat. Rev. Neurosci. 4, 637–648 (2003).

    Article  CAS  PubMed  Google Scholar 

  42. Ito, H. T. Prefrontal–hippocampal interactions for spatial navigation. Neurosci. Res. 129, 2–7 (2018).

    Article  PubMed  Google Scholar 

  43. Spellman, T. et al. Hippocampal–prefrontal input supports spatial encoding in working memory. Nature 522, 309–314 (2015).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  44. Place, R., Farovik, A., Brockmann, M. & Eichenbaum, H. Bidirectional prefrontal–hippocampal interactions support context-guided memory. Nat. Neurosci. 19, 992–994 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Ito, H. T., Zhang, S. J., Witter, M. P., Moser, E. I. & Moser, M. B. A prefrontal–thalamo–hippocampal circuit for goal-directed spatial coding. Nature https://doi.org/10.1038/nature14396 (2015).

  46. Feierstein, C. E., Quirk, M. C., Uchida, N., Sosulski, D. L. & Mainen, Z. F. Representation of spatial goals in rat orbitofrontal cortex. Neuron 51, 495–507 (2006).

    Article  CAS  PubMed  Google Scholar 

  47. Wang, P. Y. et al. Transient and persistent representations of odor value in prefrontal cortex. Neuron 108, 209–224.e6 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  48. Basu, R. et al. The orbitofrontal cortex maps future navigational goals. Nature 599, 449–452 (2021).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  49. Jun, H., Chavez, J., Bramian, A. & Igarashi, K. M. Protocol for remapping of place cells in disease mouse models. STAR Protoc. 2, 100759 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  50. Kvitsiani, D. et al. Distinct behavioural and network correlates of two interneuron types in prefrontal cortex. Nature 20, 363–366 (2013).

    Article  ADS  Google Scholar 

Download references

Acknowledgements

We thank M. Witter and A. Treves, and members in the Igarashi laboratory for providing valuable comments on this work. The work was supported by the US National Institutes of Health (NIH) R01 grants (R01MH121736, R01AG063864, R01AG066806 and R01AG086441) and a BrightFocus Foundation Research grant (A2019380S) to K.M.I. H.J. was supported by the University of California, Irvine Medical Scientist Training Program (MSTP; T32GM008620) and the NIH F31 grant (F31AG069500). J.Y.L. was supported by the NIH F31 grant (F31AG074650).

Author information

Author notes

  1. These authors contributed equally: Heechul Jun, Jason Y. Lee

Authors and Affiliations

  1. Department of Anatomy and Neurobiology, School of Medicine, University of California Irvine, Irvine, CA, USA

    Heechul Jun, Jason Y. Lee, Nicholas R. Bleza, Ayana Ichii, Jordan D. Donohue & Kei M. Igarashi

  2. Department of Biomedical Engineering, Samueli School of Engineering, University of California Irvine, Irvine, CA, USA

    Kei M. Igarashi

  3. Center for Neural Circuit Mapping, School of Medicine, University of California Irvine, Irvine, CA, USA

    Kei M. Igarashi

  4. Center for the Neurobiology of Learning and Memory, University of California Irvine, Irvine, CA, USA

    Kei M. Igarashi

  5. Institute for Memory Impairments and Neurological Disorders, University of California Irvine, Irvine, CA, USA

    Kei M. Igarashi

Authors
  1. Heechul Jun
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jason Y. Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Nicholas R. Bleza
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Ayana Ichii
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Jordan D. Donohue
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Kei M. Igarashi
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

H.J., J.Y.L. and K.M.I. conceived the project and designed the experiments. H.J., J.Y.L., N.R.B., A.I. and J.D.D. performed the experiments. H.J., J.Y.L. and K.M.I. performed the analyses. H.J., J.Y.L. and K.M.I. wrote the paper with input from all authors.

Corresponding author

Correspondence to Kei M. Igarashi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature thanks Thomas McHugh and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

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 Histological validation of implanted sites.

(a) Recording positions in LEC of Rbp4-Cre mice for opt-tagging experiment (Figs. 1 and 4). Positions were marked with electrolytic lesioning. D, dorsal, V, ventral, M, medial, L, lateral. (b) Optic fiber positions in LEC of Rbp4-Cre mice injected with AAV-flex-Jaws-GFP into LEC for inhibition experiments (Fig. 3). Arrowhead, estimated tip of optic fibers. (c) Optic fiber positions in LEC of Rbp4-Cre mice injected with retroAAV-flex-Jaws-GFP into mPFC for inhibition of mPFC-projecting LEC cells (Extended Data Fig. 14). (d) Optic fiber positions in LEC of WT mice injected with AAV-hSyn-Jaws-GFP into mPFC for inhibition of mPFC terminals in the LEC (Fig. 4 and Extended Data Fig. 17). (e) Recording positions in mPFC of WT mice (Fig. 2). Positions were marked with electrolytic lesioning. (f) Optic fiber positions in mPFC of WT mice injected with AAV-hSyn-Jaws-GFP into mPFC for inhibition experiments (Fig. 4). *Mice used for recording LECL5/6 cells + mPFC terminal inhibition; **Mice used for recording mPFC cells + LECL5/6 inhibition.

Extended Data Fig. 2 Verification of opttagging procedure.

(a) Spike raster plots of a typical opttagged neuron which is seen to follow 1, 2, 5, and 10 Hz pulse trains (5 ms pulse, 10 pulses per train). (b) “Opto-PSTH” of 6 example opttagged neurons. Neurons i-v) feature optically evoked spikes with low latency and jitter (Bottom right) An excluded opttagged cell (vi), with a long latency of 9.1 ms, suggesting a synaptically tagged cell. (c) Latencies of all opttagged neurons plotted against the jitter of first spike after stimulation. Opttagged neurons had a wide variety of latencies (2-8.5 ms), making it difficult to delineate real opttagged cells and potential synaptically tagged cells. (d) PCA trajectories, Euclidian distances, and SI plot of LECL5/6 population as in Fig. 1i, excluding opttagged cells with latency greater than 8.5 ms (n = 213 cells). (e) PCA trajectories, Euclidian distances, and SI plot of LECL5/6 population as in Fig. 1i, excluding opttagged cells with latency greater than 5 ms (n = 153 cells). These results show a similar Odor-A and -a representation, and similar Odor-B and -b representation at T5. Thus, although the total population analysis may include some synaptically opttagged neurons, it is likely that our conclusion is not affected by their marginal numbers.

Extended Data Fig. 3 Cue-outcome profile of LECL5/6 neurons along the dorsoventral axis.

Recorded positions along dorsoventral axis of LEC. Distances were calculated from rhinal sulcus border to approximate location of amygdalopiriform transition area and normalized from 0 (dorsal) to 1 (ventral). The result demonstrated that while Aa cells were observed along the D-V axis, Bb cells were found only in the ventral part of the LEC.

Extended Data Fig. 4 Spike properties of LECL5/6 cells during correct associative learning.

(a-d) Spike properties of LECL5/6 cells. LECL5/6 cells were recorded in a session with Odor-A and Odor-B (AB-only session). After ~20 trials in AB session, associative learning (ABab) session was tested (T1-T5). (a) Spike firing of 228 LECL5/6 cells. Mean spike activity was averaged in 50 ms bins and shown in z-score compared with −1 – 0 s before odor onset. In this panel, cells were sorted using a cluster analysis of firing property in T5. (b) Mean firing rate of 228 LECL5/6 cells shown in z-score. (c) Percent responsive cells for each cue combination. Neurons with significant firing during 0.5-3.0 s after odor onset (odor + delay period) were counted (Wilcoxon signed-rank test, p < 0.05). (d) Percent responsive cells in periods of 0.5-1.0 s (odor), 2-3 s (delay) and 3-4 s (choice) after odor onset. Neurons with significant firing (p < 0.05) during each period were counted (two-sided signed-rank test). (e) Trajectories of neural firing of LECL5/6 cell population as in Fig. 1i, but presented throughout timepoints T1 – T5. (f) Euclidian distance between odor trial types. (g) Mean Euclidian distance during 0.5-3.0 s after odor onset (odor + delay period). Ninety-fifth percentile distance obtained from shuffled data denotes significant distance (red line). (h) The change of Similarity Index (SI) during associative learning was compared using the bootstrapping method (see Methods). PCA was performed from a resampled neuronal population, and this procedure was repeated 1000 times to make 1000 bootstraps. SI was calculated for each bootstrap, then SIs in T2–T5 were subtracted by that in T1, to test if there was a significant distribution above or below zero. SI for Odors A-a increased at T3-5 (p = 0.005, 0.012, 0.0005 respectively, bootstrapping test), as well as SI for Odors B-b at T3-5 (p = 0.0023, 0.0021, 0.003 respectively), confirming increasing similarity of representations between Odor-A and -a, and between B-b during leaning. *p < 0.05, **p < 0.01, ***p < 0.001, two-sided bootstrapping test. Data are presented as mean values +/- SEM.

Extended Data Fig. 5 Selectivity profile of mPFC and LECL5/6 single neurons during learning.

(a) Examples of neurons based on their selectivity profile between T1 vs T5 period. Type 1: Static cells with same selectivity to an odor cue; Type 2: Static cells without selectivity to any cue; Type 3: Plastic cells with selectivity in T1 to an odor cue but no selectivity in T5; Type 4: Plastic cells with no selectivity in T1 but gained selectivity to an odor cue in T5; Type 5: Plastic cells with selectivity to an odor cue in T1 but gained distinct selectivity in T5. Significant selectivity (p < 0.05) was assessed using spike response during 0.5 – 3 s after cue onset, compared to 1 s pre-cue period (two-sided rank sum test). Data are presented as mean values +/- SEM. (b) (Top) Percentage of static (54%) or plastic (46%) mPFC cells. Static cells were further subdivided into type1 (4%) and type 2 (50%). Plastic cells were further subdivided into type 3 (18%), type 4 (14%), and type 5 (14%). (Bottom) Percentage of static (50%) or plastic (50%) LEC5/6 cells. Static cells were further subdivided into type1 (4%) and type 2 (46%). Plastic cells were further subdivided into type 3 (19%), type 4 (13%), and type 5 (18%).

Extended Data Fig. 6 Additional principal component analyses for LECL5/6 cell recordings.

(a) Error-exclusion analysis. Trajectories of neural firing of LECL5/6 cell population using only correct (hit) trials for Odors-A and -a and correct rejection (CR) trials for Odors-B and -b. The pattern of A-a vs. B-b dichotomic classification was again observed when the incorrect trials were removed from the PC analysis, excluding the possibility that the representations in Fig. 1i emerged from the increasing rate of correct trials. (b) Movement analysis. Trajectories of neural firing of LECL5/6 cell population using only lick (hit) trials for Odors-A and -a, and error lick (false alarm, FA) trials for Odors-B and -b. Although all of them are trials in which mice made lick responses, the A-a vs. B-b dichotomic classification similar to Fig. 1i was observed, suggesting that LECL5/6 cells do not simply represent lick-related motor information.

Extended Data Fig. 7 Effect of individual Aa and Bb cells on population activity.

PCA trajectories, Euclidian distances, and SI plot of LECL5/6 population, excluding individual Aa and Bb cells (n = 137 cells). The result showed unstable grouping between Odor-A and Odor-a (shown by unstable SIA-b in the SI plot), and no development of grouping between Odor-B and Odor-b (shown by low SIB-b throughout the session). This result suggests that Aa cells and Bb cells contribute to the generation of outcome rule representation.

Extended Data Fig. 8 Characteristics of Rbp4+ LECL5/6 cells.

(a) Electrophysiological features for classifying putative principal neurons. Peak to valley time of spike waveform was used to distinguish putative interneurons from principal neurons in Rbp4+ LECL5/6 cells. (Left) Dashed line (230µs) represents cut off for wide spike (WS) excitatory neurons (n = 176; 77%) to narrow spike (NS) interneurons (n = 52; 23%) classification (Bartho et al., 2004). (Right) Independent trajectories of neural firing of LECL5/6 WS neurons and NS neurons both demonstrated the grouping of Odors-A and -a and grouping of Odors-B and -b. (b) LECL5/6 labeling with excitatory cell marker. (Left) AAV-flex-tdTomato (red) and AAV-flex-CaMKIIa-GFP (green) were injected into LEC5/6 of Rbp4-Cre mice. (Middle) Coronal sections of LEC layer 5/6. CaMKIIa (green) GFP signal reveals excitatory neurons. TdTomato (red) signal reveals Rbp4+ LECL5/6 cells. Bottom panel demonstrates magnified windows from top panel. Yellow arrow points to example LECL5/6 excitatory cell expressing both CaMKIIa GFP and tdTomato. White arrow points to example LECL5/6 non-excitatory cell expressing only tdTomato. (Right) Percentage of double-positive neurons among tdTomato+ neurons (80.08 ± 2.75 % from 8 sections obtained from n = 2 mice). Data are presented as mean values +/- SEM.

Extended Data Fig. 9 Spike properties of LECL5/6 cells during spontaneous error sessions.

(a) Behavioral performance of LEC Rbp4-Cre recording mice in correct sessions (Fig. 1) and error sessions where mice spontaneously could not reach >80% performance criteria. Data are presented as mean values +/- SEM. (a1) Top, learning curves during correct and error sessions. Familiar cue performance (A/B) in blue, novel cue performance (a/b) in red. Bottom, percent correct trials in T5 for correct vs. error sessions (p = 8.8e-8, two-way ANOVA; A/B correct vs. a/b error, *p = 3.8e-9 or less, post-hoc Tukey test; n = 8 mice). (a2) (Left) Percentage of correct trials plotted in each odor trial type. (Right) Performance of mice in the last 10 trials (p = 4.8e-28, two-way ANOVA; ***p = 6e-8 between Odor-a correct vs. odor-a error, post-hoc Tukey test; n = 8 mice). (b-e) Same as in Extended Data Fig. 4a–d, but for error sessions (n = 72 cells). (f) Proportion of Aa cells (left) and Bb cells (right) in correct vs. error sessions. Both response types were missing in error sessions. (g) Same as in in Extended Data Fig. 4e–g, but for error sessions. SI plot shows that SIA-a did not increase to positive, indicating Odor-a was not classified together with Odor-A. (h) Bootstrapping analysis. (Top) Same plot as in Extended Data Fig. 4h, but for error sessions. LECL5/6 cells did not develop similar representations between Odors-A and -a, and between Odors-B and -b observed in correct sessions. (Bottom) Direct comparison between correct and error sessions confirmed the disappearance of similar representations between Odors-A and -a at T3-T5 of error sessions (p = 0.27, 0.14, 0.13 respectively). The comparison also indicates more similar representations between Odors-a and -b throughout error sessions T1-T5 (p = 0.043, 0.013, 0.0056, 0.026, 0.0054 respectively). *p < 0.05, **p < 0.01, ***p < 0.001, two-sided bootstrapping test. Data are presented as mean values +/- SEM.

Extended Data Fig. 10 Spike properties of mPFC cells during correct associative learning sessions.

(a) Spike firing of 779 mPFC cells. Mean spike activity was averaged in 50 ms bins and shown in z-score compared with −1 – 0 s before odor onset. In this panel, cells were sorted using a cluster analysis of firing property in T5. (b) Mean firing rate of 779 mPFC cells shown in z-score. (c) Percent responsive cells in correct T5 (top) and error T5 (bottom). Neurons with significant firing (p < 0.05) during 0.5-3.0 s after odor onset (odor + delay period) were counted (two-sided signed-rank test). (d) Percent responsive cells in periods of 0.5-1.0 s (odor), 2-3 s (delay) and 3-4 s (choice) after odor onset. Neurons with significant firing (p < 0.05) during each period were counted (two-sided signed-rank test). (e) Trajectories of neural firing of mPFC cell population as shown in Fig. 2h, but for timepoints T1 – T5. (f) Euclidian distance between odor trial types. (g) Mean Euclidian distance during 0.5-3.0 s after odor onset (odor + delay period) for timepoints T1 – T5 of correct sessions. Ninety-fifth percentile distance obtained from shuffled data denotes significant distance (red line). (h) Bootstrapping analysis for Fig. 2j. SI for Odors A-a and B-b showed significant increases in T5 compared to T1 (p = 0.0039 and p = 1.7e-8, respectively, bootstrapping test), while SI for Odors A-B and a-b showed significant decreases (p = 3.1e-5 and p = 0.029, respectively), confirming increasing similarity between odors of the same outcome, and decreasing similarity between odors of the different outcome. *p < 0.05, **p < 0.01, ***p < 0.001, two-sided test, n = 1000 bootstraps each. Data are presented as mean values +/- SEM.

Extended Data Fig. 11 Early emergence of grouping of Odor-A and Odor-a rewarded cues.

(a) Trajectories of neural firing of LECL5/6 cell population as in Fig. 1i, but presented throughout timepoints T1a (first five trials of T1) and T1b (second five trials of T1). (b) Trajectories of neural firing of mPFC cell population as above.

Extended Data Fig. 12 Additional principal component analyses for mPFC cell recordings.

(a) Error-exclusion analysis. Trajectories of neural firing of mPFC cell population using only correct (hit) trials for Odors-A and -a, and correct rejection (CR) trials for Odors-B and -b. The pattern of A-a vs. B-b dichotomic classification was again observed when the incorrect trials were removed from the PC analysis, excluding the possibility that the representations in Fig. 2h emerged from the increasing rate of correct trials. (b) Movement analysis. Trajectories of neural firing of mPFC cell population using only lick (hit) trials for Odors-A and -a and error lick (false alarm, FA) trials for Odors-B and -b. Although all of them are trials in which mice made lick responses, the A-a vs. B-b dichotomic classification similar to Fig. 2h was observed, suggesting that mPFC cells do not simply represent lick-related motor information.

Extended Data Fig. 13 Spike properties of mPFC cells during spontaneous error sessions.

(a) Behavioral performance of mPFC recording mice in correct sessions (Fig. 2) and error sessions where mice spontaneously could not reach >80% performance criteria. Data are presented as mean values +/- SEM. (a1) Behavioral performance of mPFC recording mice in correct vs. error sessions. Top, learning curves during correct and error sessions. Familiar cue performance (A/B) in blue, novel cue performance (a/b) in red. Bottom, percentage of correct trials in T5 for correct vs. error sessions (p = 2.4e-7, ANOVA; p = 1.9e-8 or less, post-hoc Tukey test; n = 6 mice). (a2) mPFC recording mice data plotted for percent correct trials in each odor trial type as a function of trial number for each odor type. (Right) Performance of mice in the last 10 trials (p = 1.99e-11, ANOVA; ***p = 6e-8 between Odor-a correct vs. odor-a error, post-hoc Tukey test; n = 6 mice). (b-e) Same as in Extended Data Fig. 10a–d, but for error sessions (n = 116 cells). (f) Proportion of mPFC Aa cells (left) and Bb cells (right) in correct vs. error sessions. Aa: p = 1.8e-7, Bb: p = 2.2e-13, two-sided binomial test. (g) Same as in Extended Data Fig. 10e–g, but for error sessions. (h) Bootstrapping analysis. (Top) Same plot as in Extended Data Fig. 10h, but for error sessions. During the error sessions, mPFC cells did not develop similar representations between Odors-A and -a (T2-T5: p = 0.24, 0.39, 0.36, 0.36, respectively) and between Odors-B and -b (T2-T5: p = 0.36, 0.034, 0.089, 0.43, respectively) observed in correct sessions. (Bottom) Direct comparison between correct and error sessions confirmed the disappearance of similar representations between Odors-A and -a (T3-T5: p = 0.006, 0.0005, 6.3e-8, respectively), and similar representations between Odors-B and -b (T5, p = 0.0002). The comparison also indicates more similar representations between Odors-A and -B (T1-T5: p = 3e-6, 1.8e-6, 6.2e-8. 4e-7, 1.2e-7, respectively), and between Odors-a and -b throughout the error sessions (T1-T5: p = 7e-6, 1e-16, 3e-15, 1e-16, 1e-16, respectively). *p < 0.05, **p < 0.01, ***p < 0.001, bootstrapping test, n = 1000 bootstraps each. Data are presented as mean values +/- SEM.

Extended Data Fig. 14 Behavioral performance in LECL5/6 inhibition experiments.

(a) Behavioral performance of LEC Rbp4-Cre mice in Fig. 3a,b, but plotted for each odor trial type. Learning curves during control (left) and LECL5/6 inhibition (middle) sessions. (Right) Performance of mice in the last 10 trials (p = 3.8e-13, two-way ANOVA; p = 6e-8 between Odor-a in control vs. Odor-a inhibition sessions, post-hoc Tukey test; n = 10 mice). Data are presented as mean values +/- SEM. (b-d) mPFC-projecting LECL5/6 cell inhibition experiments. Data are presented as mean values +/- SEM. (c) (Left) Learning curves during control and inhibition sessions. (Middle) Percent correct trials in T5 for control vs. inhibition sessions (p = 3.4e-6, two-way ANOVA; A/B control vs. a/b inhibition, p = 3.8e-9; a/b correct vs. a/b error, p = 7.2e-9; A/B error vs. a/b error, p = 3.9e-9, post-hoc Tukey test; n = 10 mice). (Right) Proportion of correct sessions between control and inhibition conditions (p = 0.0064, two-sided binomial test; n = 30 control and 30 inhibition sessions). (d) Percent correct trials in each odor trial type as a function of trial number for each odor type. (Right) Performance of mice in the last 10 trials (p = 1.5e-16, two-way ANOVA; p = 6e-8 for Odor-a in correct sessions vs. odor-a in error sessions, post-hoc Tukey test; n = 10 mice).

Extended Data Fig. 15 Modulation of neuronal activity with Jaws inhibition.

(a) (Top) Recording of mPFC cells during inhibition. Jaws were expressed in mPFC cells using synapsin promoter driven AAVs. (Middle) Correlation between change in firing rate and pre-stimulation firing rate divided into principal neurons and interneurons based on their spike waveform peak-to-trough width threshold of 230 μs. Each dot represents a single cell. Cells were classified as inhibited, unaffected, or disinhibited using two-sided rank sum test (p < 0.05 threshold) between pre-stimulation firing rate and firing rate during inhibition. (Bottom) Percentage of inhibited (21.9%), unaffected (60.3%), or disinhibited (17.8%) mPFC cells during inhibition. (b) (Top) Opt-tag recording of LECL5/6 cells during inhibition. ChR2 and Jaws were expressed in LECL5/6 of Rbp4-Cre mice. (Middle) Correlation between change in firing rate and pre-stimulation firing rate. Each dot represents a single cell. Cells were classified as inhibited, unaffected, or disinhibited using two-sided rank sum test (p < 0.05 threshold) between pre-stimulation firing rate and firing rate during inhibition. (Bottom) Percentage of inhibited (12.5%), unaffected (75%), or disinhibited (12.5%) LEC5/6 cells during inhibition.

Extended Data Fig. 16 Detailed analyses from recording of mPFC cells with LECL5/6 inhibition.

(a-b) Behavioral performance only from n = 4 mice used in mPFC cell recording with simultaneous inhibition of LECL5/6 cells. (b1) (Top) Learning curves during control and inhibition sessions. (Bottom left) Percentage of correct trials in T5 for control vs. inhibition sessions (p = 7.5e-9, two-way ANOVA; p = 5.6e-9 or less, post-hoc Tukey test; n = 4 mice). (Bottom right) Proportion of correct sessions between control and inhibition conditions (p = 1.6e-10, two-sided binomial test; n = 24 control and 24 inhibition sessions). Data are presented as mean values +/- SEM. (b2) Percent correct trials in each odor trial type as a function of trial number for each odor type. (Right) Performance of mice in the last 10 trials (p = 2.8e-18, two-way ANOVA; p = 6e-6 between Odor-a in correct sessions vs. Odor-a in error sessions, post-hoc Tukey test; n = 4 mice). Data are presented as mean values +/- SEM. (c) (Top) Trajectories of neural firing of n = 497 mPFC cell population in no-laser control sessions. (Middle) Euclidian distance between odor trial types. (Bottom) Mean Euclidian distance during 0.5-3.0 s after odor onset (odor + delay period) for timepoints T1 – T5 of correct sessions. Ninety-fifth percentile distance obtained from shuffled data denotes significant distance (red line). (Right) Similarity index. (d) Same as (c), but for inhibition sessions (laser on). (e) Bootstrapping analysis. (Top) In control sessions, SI between Odors A-a and between Odors B-b showed significant increase from T1 to T5 (p = 0.014, p = 2.6e-6, respectively). SI for Odors A-B and a-b showed significant decrease from T1 to T5 (p = 5.4e-6, p = 0.0027, respectively, bootstrapping test). (Middle) In inhibition sessions, SI for all odor pairs did not show significant differences from T1 to T5 (Aa: p = 0.15, Bb: p = 0.08, AB: p = 0.45, ab: p = 0.34), confirming the impairment of outcome classification. (Bottom) Direct comparison between control and inhibition sessions confirmed the disappearance of similar representations between Odors-A and -a (T5: p = 7e-6). The comparison also indicates disappearance of separate representations between Odors-A and -B and between Odors-a and -b in inhibition sessions (p = 2e-4, 8e-13, respectively). n = 1000 bootstraps; *p < 0.05, **p < 0.01, ***p < 0.001, bootstrapping test. Data are presented as mean values +/- SEM.

Extended Data Fig. 17 Behavioral performance for mPFC inhibitions.

(a) Behavioral performance during mPFC inhibition in Fig. 4a,b, but plotted for each odor trial type. Learning curves during control (left) and mPFC inhibition (middle) sessions. (Right) Performance of mice in the last 10 trials (p = 1e-32, two-way ANOVA; p = 6e-8 between Odor-A in control sessions vs. Odor-A in inhibition sessions; p = 6e-8 between Odor-a in control sessions vs. Odor-a in inhibition sessions, post-hoc Tukey test; n = 13 mice). Data are presented as mean values +/- SEM. (b-d) Behavior performance during mPFC terminal inhibition. Data are presented as mean values +/- SEM. (c) (Left) Learning curves during control and inhibition sessions. (Middle) Percent correct trials in T5 for control vs. inhibition sessions (p = 0.029, two-way ANOVA; p = 0.0088 or less, post-hoc Tukey test; n = 7 mice). (Right) Proportion of correct sessions between control and inhibition conditions (p = 0.0001, two-way binomial test; n = 31 control and 21 inhibition sessions). (d) Percent correct trials in each odor trial type as a function of trial number for each odor type. (Right) Performance of mice in the last 10 trials (p = 0.0037, two-way ANOVA; p = 0.0003 between Odor-a in correct sessions vs. Odor-a in error sessions, post-hoc Tukey test; n = 7 mice). Although not significant, we observed a trend of decrease in the performance for Odor-A during inhibition (p = 0.63, Odor-A in correct sessions vs. Odor-A in error sessions, post-hoc Tukey test).

Extended Data Fig. 18 Detailed analyses from recording of LECL5/6 cells with mPFC terminal inhibition.

(a) Opt-tag recording of LECL5/6 cells with simultaneous inhibition of mPFC inputs in LEC. (b) (Top) Trajectories of neural firing of n = 181 LECL5/6 cell population in no-laser control sessions. (Middle) Euclidian distance between odor trial types. (Bottom) Mean Euclidian distance during 0.5-3.0 s after odor onset (odor + delay period) for timepoints T1 – T5 of correct sessions. Ninety-fifth percentile distance obtained from shuffled data denotes significant distance (red line). (Right) Similarity index. (c) Same as (b), but for n = 249 LECL5/6 cell population during mPFC terminal inhibition sessions (laser on). (d) Bootstrapping analysis. (Top) In control sessions, SI between Odors A-a and between Odors B-b showed significant increase from T1 to T5 (p = 0.012, p = 0.0095, respectively). (Middle) In inhibition sessions, SI for all odor pairs did not show significant differences from T1 to T5 (Aa: p = 0.41, Bb: p = 0.07, AB: p = 0.15, ab: p = 0.23), confirming the impairment of outcome classification. (Bottom) Direct comparison between control and inhibition sessions confirmed the disappearance of similar representations between Odors-A and -a (p = 0.0014 at T5). *p < 0.05, **p < 0.01, ***p < 0.001, two-sided bootstrapping test, n = 1000 bootstraps each. Data are presented as mean values +/- SEM.

Extended Data Fig. 19 Fast development of odor outcome representation in mPFC population.

(a) Development of task performance, Similarity Index (SI) between Odors-A and a (SIA-a) for LECL5/6 cell population, and SI between Odors B and b (SI B-b) for LECL5/6 cell population. Variables are normalized onto a scale from 0 (T1a) to 1 (T5). (b) Development of task performance, SIA-a for mPFC population and SIB-b for mPFC population as above.

Extended Data Fig. 20 Decoding analyses of LECL5/6 and mPFC population activity.

(a-b) A support vector machine (SVM) decoder was trained to discriminate novel Odor-a vs. Odor-b, using neural responses to familiar Odor-A vs. Odor-B in the same timepoint (T) as training data. For each 100 ms time bin, 100 SVMs were trained and their performances averaged to achieve the final % correct value. Decoding performance increases quickly in mPFC between T1-T2 (~90% at Trials 6-10 (T1b)), while performance only gradually reached ~90% at T3 in LECL5/6, indicating a quicker development of outcome representation in mPFC. T1A = trials 1-5; T1B = trials 6-10. a) Decoder performance for Correct sessions. (Top row) LECL5/6 population. (Bottom row) mPFC population. Right: Average performance from the 0.5-3 s window of each timepoint. b) As above, but for Error sessions. (c-d) The decoder was trained to discriminate within outcome categories (Odor-A vs. Odor-a OR Odor-B vs. Odor-b), using half the trials of one timepoint to train and the other half to test. Although the discriminatory power between Odor-A and Odor-a gradually decreased from T1 to T5, it remained at ~75% at T5. This result suggests that LEC neural activities still have information distinct enough for the decoder to discriminate between Odor-A and -a at T5. This was also the case for Odor-B vs Odor-b (78% at T5), as well as mPFC (82% at T5 for A vs a, 98% at T5 for B vs b). (c) Decoder performance for LECL5/6 population during Correct sessions. (d) Decoder performance for mPFC population during Correct sessions.

Extended Data Fig. 21 Characterizing monosynaptic input to Rbp4+ LECL5/6 cells using retrograde rabies tracing.

Monosynaptic retrograde tracing of Rbp4+ LECL5/6 neurons. (Left) Cre-depndent glycoprotein and TVA-GFP virus (green) and G-deleted rabies virus expressing mCherry (red) were injected into the LECL5/6 of Rbp4-Cre mice. (Right) Top panel shows coronal sections revealing starter Rbp4+ cells (indicated by white arrows) that are double positive (red and green) near the injection site. The remaining panels demonstrate coronal sections with examples of presynaptically labelled neurons (red only) from local LEC, posterior piriform cortex, amygdala and hippocampal CA1 of the intermediate-ventral tiers (indicated by white arrows). (BLA = Basolateral amygdala, BMP = Basomedial posterior amygdala, BLV = Basolateral ventral amygdala, LaVL = Lateral amygdala ventrolateral, LaVM = Lateral amygdala ventromedial and LaDL = Lateral amygdala dorsolateral).

Supplementary information

Supplementary Information

Supplementary Results and Discussion which report and discuss additional results that are not logically connected to the main text.

Reporting Summary

Supplementary Tables

Supplementary Tables 1–5

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Jun, H., Lee, J.Y., Bleza, N.R. et al. Prefrontal and lateral entorhinal neurons co-dependently learn item–outcome rules. Nature (2024). https://doi.org/10.1038/s41586-024-07868-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1038/s41586-024-07868-1

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