Adaptive disinhibitory gating by VIP interneurons permits associative learning

Article metrics

Abstract

Learning drives behavioral adaptations necessary for survival. While plasticity of excitatory projection neurons during associative learning has been extensively studied, little is known about the contributions of local interneurons. Using fear conditioning as a model for associative learning, we found that behaviorally relevant, salient stimuli cause learning by tapping into a local microcircuit consisting of precisely connected subtypes of inhibitory interneurons. By employing deep-brain calcium imaging and optogenetics, we demonstrate that vasoactive intestinal peptide (VIP)-expressing interneurons in the basolateral amygdala are activated by aversive events and provide a mandatory disinhibitory signal for associative learning. Notably, VIP interneuron responses during learning are strongly modulated by expectations. Our findings indicate that VIP interneurons are a central component of a dynamic circuit motif that mediates adaptive disinhibitory gating to specifically learn about unexpected, salient events, thereby ensuring appropriate behavioral adaptations.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Aversive foot shocks activate VIP BLA interneurons during fear learning.
Fig. 2: Long-range connectivity of distinct BLA interneuron subtypes.
Fig. 3: VIP BLA interneurons preferentially target other interneuron subtypes over projection neurons.
Fig. 4: VIP BLA interneuron activation disinhibits projection neurons.
Fig. 5: VIP BLA interneurons control projection neuron US activity and CS plasticity.
Fig. 6: VIP BLA interneuron activation during the aversive US is necessary for learning.
Fig. 7: VIP BLA interneuron US activity decreases during learning.
Fig. 8: VIP BLA interneuron activity is modulated by expectation.

Data availability

Data from this study as well as material from custom products are available from the corresponding authors upon request.

Code availability

Custom-written codes used to analyze data from this study are available from the corresponding authors upon request.

References

  1. 1.

    Froemke, R. C. Plasticity of cortical excitatory–inhibitory balance. Annu. Rev. Neurosci. 38, 195–219 (2015).

  2. 2.

    Kepecs, A. & Fishell, G. Interneuron cell types are fit to function. Nature 505, 318–326 (2014).

  3. 3.

    Tovote, P., Fadok, J. P. & Lüthi, A. Neuronal circuits for fear and anxiety. Nat. Rev. Neurosci. 16, 317–331 (2015).

  4. 4.

    LeDoux, J. E. Emotion circuits in the brain. Annu. Rev. Neurosci. 23, 155–184 (2000).

  5. 5.

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

  6. 6.

    McNally, G. P., Johansen, J. P. & Blair, H. T. Placing prediction into the fear circuit. Trends Neurosci. 34, 283–292 (2011).

  7. 7.

    Krabbe, S., Gründemann, J. & Lüthi, A. Amygdala inhibitory circuits regulate associative fear conditioning. Biol. Psychiatry 83, 800–809 (2018).

  8. 8.

    Wolff, S. B. E. et al. Amygdala interneuron subtypes control fear learning through disinhibition. Nature 509, 453–458 (2014).

  9. 9.

    Johansen, J. P. et al. Optical activation of lateral amygdala pyramidal cells instructs associative fear learning. Proc. Natl Acad. Sci. USA 107, 12692–12697 (2010).

  10. 10.

    Pfeffer, C. K., Xue, M., He, M., Huang, Z. J. & Scanziani, M. Inhibition of inhibition in visual cortex: the logic of connections between molecularly distinct interneurons. Nat. Neurosci. 16, 1068–1076 (2013).

  11. 11.

    Pi, H. J. et al. Cortical interneurons that specialize in disinhibitory control. Nature 503, 521–524 (2013).

  12. 12.

    Muller, J. F., Mascagni, F. & McDonald, A. J. Synaptic connections of distinct interneuronal subpopulations in the rat basolateral amygdalar nucleus. J. Comp. Neurol. 456, 217–236 (2003).

  13. 13.

    Rhomberg, T. et al. VIP-immunoreactive interneurons within circuits of the mouse basolateral amygdala. J. Neurosci. 38, 6983–7003 (2018).

  14. 14.

    Grewe, B. F. et al. Neural ensemble dynamics underlying a long-term associative memory. Nature 543, 670–675 (2017).

  15. 15.

    Gründemann, J. et al. Amygdala ensembles encode behavioral states. Science 364, eaav8736 (2019).

  16. 16.

    Bienvenu, T. C. M., Busti, D., Magill, P. J., Ferraguti, F. & Capogna, M. Cell-type-specific recruitment of amygdala interneurons to hippocampal theta rhythm and noxious stimuli in vivo. Neuron 74, 1059–1074 (2012).

  17. 17.

    Lanuza, E., Nader, K. & LeDoux, J. E. Unconditioned stimulus pathways to the amygdala: effects of posterior thalamic and cortical lesions on fear conditioning. Neuroscience 125, 305–315 (2004).

  18. 18.

    Berret, E. et al. Insular cortex processes aversive somatosensory information and is crucial for threat learning. Science 364, eaaw0474 (2019).

  19. 19.

    Jiang, X. et al. Principles of connectivity among morphologically defined cell types in adult neocortex. Science 350, aac9462 (2015).

  20. 20.

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

  21. 21.

    Sengupta, A. et al. Basolateral amygdala neurons maintain aversive emotional salience. J. Neurosci. 38, 3001–3012 (2018).

  22. 22.

    Francavilla, R. et al. Connectivity and network state-dependent recruitment of long-range VIP-GABAergic neurons in the mouse hippocampus. Nat. Commun. 9, 5043 (2018).

  23. 23.

    Senn, V. et al. Long-range connectivity defines behavioral specificity of amygdala neurons. Neuron 81, 428–437 (2014).

  24. 24.

    Vogel, E., Krabbe, S., Gründemann, J., Wamsteeker Cusulin, J. I. & Lüthi, A. Projection-specific dynamic regulation of inhibition in amygdala micro-circuits. Neuron 91, 644–651 (2016).

  25. 25.

    Varga, C., Lee, S. Y. & Soltesz, I. Target-selective GABAergic control of entorhinal cortex output. Nat. Neurosci. 13, 822–824 (2010).

  26. 26.

    Lu, J. et al. Selective inhibitory control of pyramidal neuron ensembles and cortical subnetworks by chandelier cells. Nat. Neurosci. 20, 1377–1383 (2017).

  27. 27.

    Herry, C. et al. Switching on and off fear by distinct neuronal circuits. Nature 454, 600–606 (2008).

  28. 28.

    Quirk, G. J., Repa, C. & LeDoux, J. E. Fear conditioning enhances short-latency auditory responses of lateral amygdala neurons: parallel recordings in the freely behaving rat. Neuron 15, 1029–1039 (1995).

  29. 29.

    Gambino, F. et al. Sensory-evoked LTP driven by dendritic plateau potentials in vivo. Nature 515, 116–119 (2014).

  30. 30.

    Johansen, J. P. et al. Hebbian and neuromodulatory mechanisms interact to trigger associative memory formation. Proc. Natl Acad. Sci. USA 111, E5584–E5592 (2014).

  31. 31.

    Humeau, Y. et al. Dendritic spine heterogeneity determines afferent-specific Hebbian plasticity in the amygdala. Neuron 45, 119–131 (2005).

  32. 32.

    Shin, R.-M., Tsvetkov, E. & Bolshakov, V. Y. Spatiotemporal asymmetry of associative synaptic plasticity in fear conditioning pathways. Neuron 52, 883–896 (2006).

  33. 33.

    Shaban, H. et al. Generalization of amygdala LTP and conditioned fear in the absence of presynaptic inhibition. Nat. Neurosci. 9, 1028–1035 (2006).

  34. 34.

    Morrison, D. J. et al. Parvalbumin interneurons constrain the size of the lateral amygdala engram. Neurobiol. Learn. Mem. 135, 91–99 (2016).

  35. 35.

    Hangya, B., Ranade, S. P., Lorenc, M. & Kepecs, A. Central cholinergic neurons are rapidly recruited by reinforcement feedback. Cell 162, 1155–1168 (2015).

  36. 36.

    Fu, Y. et al. A cortical circuit for gain control by behavioral state. Cell 156, 1139–1152 (2014).

  37. 37.

    Letzkus, J. J. et al. A disinhibitory microcircuit for associative fear learning in the auditory cortex. Nature 480, 331–335 (2011).

  38. 38.

    Bromberg-Martin, E. S., Matsumoto, M. & Hikosaka, O. Dopamine in motivational control: rewarding, aversive, and alerting. Neuron 68, 815–834 (2010).

  39. 39.

    Lin, S.-C. & Nicolelis, M. A. L. Neuronal ensemble bursting in the basal forebrain encodes salience irrespective of valence. Neuron 59, 138–149 (2008).

  40. 40.

    Tye, K. M., Stuber, G. D., de Ridder, B., Bonci, A. & Janak, P. H. Rapid strengthening of thalamo-amygdala synapses mediates cue-reward learning. Nature 453, 1253–1257 (2008).

  41. 41.

    Shabel, S. J. & Janak, P. H. Substantial similarity in amygdala neuronal activity during conditioned appetitive and aversive emotional arousal. Proc. Natl Acad. Sci. USA 106, 15031–15036 (2009).

  42. 42.

    Belova, M. A., Paton, J. J., Morrison, S. E. & Salzman, C. D. Expectation modulates neural responses to pleasant and aversive stimuli in primate amygdala. Neuron 55, 970–984 (2007).

  43. 43.

    Beyeler, A. et al. Divergent routing of positive and negative information from the amygdala during memory retrieval. Neuron 90, 348–361 (2016).

  44. 44.

    Kim, J., Pignatelli, M., Xu, S., Itohara, S. & Tonegawa, S. Antagonistic negative and positive neurons of the basolateral amygdala. Nat. Neurosci. 19, 1636–1646 (2016).

  45. 45.

    Johansen, J. P., Tarpley, J. W., LeDoux, J. E. & Blair, H. T. Neural substrates for expectation-modulated fear learning in the amygdala and periaqueductal gray. Nat. Neurosci. 13, 979–986 (2010).

  46. 46.

    Pakan, J. M. et al. Behavioral-state modulation of inhibition is context-dependent and cell type specific in mouse visual cortex. eLife 5, 79 (2016).

  47. 47.

    Dipoppa, M. et al. Vision and locomotion shape the interactions between neuron types in mouse visual cortex. Neuron 98, 602–615.e8 (2018).

  48. 48.

    Kuchibhotla, K. V. et al. Parallel processing by cortical inhibition enables context-dependent behavior. Nat. Neurosci. 20, 62–71 (2017).

  49. 49.

    Burgess, C. R. et al. Hunger-dependent enhancement of food cue responses in mouse postrhinal cortex and lateral amygdala. Neuron 91, 1154–1169 (2016).

  50. 50.

    Calhoon, G. G. et al. Acute food deprivation rapidly modifies valence-coding microcircuits in the amygdala. Preprint at bioRxiv https://doi.org/10.1101/285189 (2018).

  51. 51.

    Taniguchi, H. et al. A resource of Cre driver lines for genetic targeting of GABAergic neurons in cerebral cortex. Neuron 71, 995–1013 (2011).

  52. 52.

    Hippenmeyer, S. et al. A developmental switch in the response of DRG neurons to ETS transcription factor signaling. PLoS Biol. 3, e159 (2005).

  53. 53.

    Madisen, L. et al. Transgenic mice for intersectional targeting of neural sensors and effectors with high specificity and performance. Neuron 85, 942–958 (2015).

  54. 54.

    He, M. et al. Strategies and tools for combinatorial targeting of GABAergic neurons in mouse cerebral cortex. Neuron 91, 1228–1243 (2016).

  55. 55.

    Chen, T.-W. et al. Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature 499, 295–300 (2013).

  56. 56.

    Xu, C. et al. Distinct hippocampal pathways mediate dissociable roles of context in memory retrieval. Cell 167, 961–972.e16 (2016).

  57. 57.

    Fenno, L. E. et al. Targeting cells with single vectors using multiple-feature Boolean logic. Nat. Methods 11, 763–772 (2014).

  58. 58.

    Ghosh, K. K. et al. Miniaturized integration of a fluorescence microscope. Nat. Methods 8, 871–878 (2011).

  59. 59.

    Paxinos, G. & Franklin, K. The Mouse Brain in Stereotaxic Coordinates (Academic Press, 2001).

  60. 60.

    Schneider, C. A., Rasband, W. S. & Eliceiri, K. W. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675 (2012).

  61. 61.

    Thevenaz, P., Ruttimann, U. E. & Unser, M. A pyramid approach to subpixel registration based on intensity. IEEE Trans. Image Process 7, 27–41 (1998).

  62. 62.

    Mukamel, E. A., Nimmerjahn, A. & Schnitzer, M. J. Automated analysis of cellular signals from large-scale calcium imaging data. Neuron 63, 747–760 (2009).

  63. 63.

    Callaway, E. M. & Luo, L. Monosynaptic circuit tracing with glycoprotein-deleted rabies viruses. J. Neurosci. 35, 8979–8985 (2015).

  64. 64.

    Sommer, C., Straehle, C., Koethe, U. & Hamprecht, F. A. Ilastik: interactive learning and segmentation toolkit. In Proc. IEEE International Symposium on Biomedical Imaging: From Nano to Macro (IEEE, 2011).

  65. 65.

    Rothman, J. S. & Silver, R. A. NeuroMatic: an integrated open-source software toolkit for acquisition, analysis and simulation of electrophysiological data. Front. Neuroinform. 12, 14 (2018).

  66. 66.

    Courtin, J. et al. Prefrontal parvalbumin interneurons shape neuronal activity to drive fear expression. Nature 505, 92–96 (2014).

  67. 67.

    Douglass, A. M. et al. Central amygdala circuits modulate food consumption through a positive-valence mechanism. Nat. Neurosci. 20, 1384–1394 (2017).

  68. 68.

    Betley, J. N. et al. Neurons for hunger and thirst transmit a negative-valence teaching signal. Nature 521, 180–185 (2015).

  69. 69.

    Owen, S. F., Berke, J. D. & Kreitzer, A. C. Fast-spiking interneurons supply feedforward control of bursting, calcium, and plasticity for efficient learning. Cell 172, 683–695.e15 (2018).

  70. 70.

    Ährlund-Richter, S. et al. A whole-brain atlas of monosynaptic input targeting four different cell types in the medial prefrontal cortex of the mouse. Nat. Neurosci. 22, 657–668 (2019).

  71. 71.

    Yetman, M. J. et al. Intersectional monosynaptic tracing for dissecting subtype-specific organization of GABAergic interneuron inputs. Nat. Neurosci. 22, 492–502 (2019).

  72. 72.

    Fadok, J. P. et al. A competitive inhibitory circuit for selection of active and passive fear responses. Nature 542, 96–100 (2017).

  73. 73.

    Botta, P. et al. Regulating anxiety with extrasynaptic inhibition. Nat. Neurosci. 18, 1493–1500 (2015).

  74. 74.

    Klavir, O., Prigge, M., Sarel, A., Paz, R. & Yizhar, O. Manipulating fear associations via optogenetic modulation of amygdala inputs to prefrontal cortex. Nat. Neurosci. 20, 836–844 (2017).

Download references

Acknowledgements

The authors thank all members of the Lüthi and Ferraguti labs for helpful discussions and comments. They thank P. Argast, P. Buchmann, A. Kovacevic, T. Lu and all staff of the FMI Animal Facility for excellent technical assistance. They further thank the Facility for Imaging and Microscopy at the FMI, in particular S. Bourke and R. Thierry, and the FMI IT department, in particular D. Flanders, S. Grzybek, R. Milani and S. van Eden, for their support with data acquisition and analyses, as well as M. Stadler for statistical advice. They are grateful to the GENIE Program at Janelia Research Campus of the Howard Hughes Medical Institute for making GCaMP6 material available, C. Ramakrishnan and K. Deisseroth for viral constructs, and S. Arber and Z. J. Huang for sharing mouse lines. The authors further thank Inscopix for providing access to the nVoke integrated imaging and optogenetics system. This work was supported by the following grants to the following individuals: the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement no. 669582), the National Center of Competences in Research: “SYNAPSY—The Synaptic Bases of Mental Diseases” (financed by the Swiss National Science Foundation, SNSF, 51NF40-158776) and an SNSF core grant (310030B_170268) all to A.L.; the Austrian Science Fund (Fonds zur Förderung der Wissenschaftlichen Forschung), Sonderforschungsbereich grants F44-17-B23 and W012060-10 to F.F.; a Young Investigator Grant from the Brain & Behavior Research Foundation (23593) and a Career Development Award from the Synapsis Foundation—Alzheimer Research Switzerland ARS (2018-CDA02) to S.K.; a NENS exchange grant to E.P.; an EMBO Long-Term Fellowship (1579-2010) to C.X.; a SNSF Ambizione grant (PZ00P3_180057) to J.C.; a SNSF Ambizione grant (PZ00P3_154765) to J.G.; and an SNSF Professorship (PP00P3_170672) and ERC Starting Grant (803870) to J.G.; as well as by the Novartis Research Foundation.

Author information

S.K., E.P., J.G., F.F. and A.L. designed the project. S.K., J.G. and Y.B. performed and analyzed the deep-brain imaging experiments. E.P. and S.D. performed and analyzed the rabies tracings. S.K. performed and analyzed the in vitro electrophysiology experiments. E.P. performed and analyzed the optogenetic manipulation experiments. J.C. performed and analyzed the in vivo electrophysiology experiments (with the help of S.K. and C.M). S.K., E.P., S.D., C.M. and T.E. performed and analyzed immunohistochemistry. C.X., K.Y. and M.M. designed, generated and validated viral constructs. S.K., E.P., J.G., F.F. and A.L. wrote the manuscript. All authors contributed to the experimental design, interpretation of the data and commented on the manuscript.

Correspondence to Jan Gründemann or Francesco Ferraguti or Andreas Lüthi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Neuroscience thanks Balazs Hangya and Ekaterina Likhtik 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.

Integrated supplementary information

Supplementary Figure 1 Heterogeneity of CS and US responses in BLA interneuron subtypes.

a, Representative example image of GCaMP6 expression in VIP interneurons in the BLA of VIP-cre mice. Scale bar, 20 µm. b, Quantification of co-localization of viral GCaMP6 expression with VIP detected by immunohistochemistry (N=3 mice). c, Schematic illustrating reconstructed implant sites of GRIN lenses (blue lines) within the BLA of VIP-cre mice for deep brain imaging experiments presented in Fig. 1 matched to a mouse brain atlas (N=7 mice). LA, lateral amygdala; BA, basal amygdala; CEA, central amygdala. d, Freezing levels before and after fear conditioning in GRIN lens-implanted VIP-cre mice (N=7). e, Representative example image of GCaMP6 expression in PV interneurons in the BLA of PV-cre mice. Scale bar, 20 µm. f, Left to right, GRIN lens implant sites in PV-cre mice (N=4), example Ca2+ responses of PV BLA interneurons to CS+ and US presentations during fear conditioning and percentage of cells with significantly increased or decreased Ca2+ responses during stimulus presentations (n=46 cells from N=4 mice). g, Example image of GCaMP6 expression in SOM interneurons in the BLA of SOM-cre mice. Scale bar, 20 µm. h, Left to right, GRIN lens implant sites in SOM-cre mice (N=5), example Ca2+ responses to CS+ and US presentations during fear conditioning and percentage of SOM interneurons with significantly increased or decreased Ca2+ responses during stimulus presentations (n=152 cells from N=5 mice). i, Fraction of US responsive VIP, PV and SOM BLA interneurons averaged across mice US excited: Kruskal-Wallis test, H=9.759, P=0.0018; Dunn’s multiple comparisons test, VIP vs SOM, P=0.0156, PV vs SOM, P=0.0261; US inhibited: Kruskal-Wallis test, H=8.639, P=0.0054; Dunn’s multiple comparisons test, VIP vs SOM, P=0.0109). j, Fraction of CS+ responsive VIP, PV and SOM BLA interneurons averaged across mice (CS+ inhibited: Kruskal-Wallis test, H=7.815, P=0.0112; Dunn’s multiple comparisons test, VIP vs SOM, P=0.0156). k, Fraction of CS- responsive VIP, PV and SOM BLA interneurons averaged across mice. For panels i-k: VIP, N=7 mice; PV, N=4; SOM, N=5. Example images and traces are representative for N=7 mice (a), N=4 (e, f), N=5 (g,h). Box-and-whisker plots show median values and 25th/75th percentiles with 10th to 90th percentile whiskers, dots additionally indicate the mean. Bar graphs are mean and s.e.m., circles individual data points. * P<0.05. Details of statistical analysis are listed in Supplementary Table 3.

Supplementary Figure 2 Monosynaptic rabies tracing from VIP, PV and SOM interneurons in the BLA.

a, Representative example image of 2A peptide and rabies-GFP (RV-GFP) expression in PV interneurons in the BLA of PV-cre mice. Yellow arrowheads point to identified starter cells expressing both TVA950-2A-CVS11G construct and RV-GFP. LA, lateral amygdala; BA, basal amygdala. Scale bar, 200 µm. High magnification image depicts an example starter cell. Scale bar, 20 µm. b, Co-expression of 2A peptide and RV-GFP in SOM interneurons in the BLA of SOM-cre mice. Yellow arrowheads point to identified starter cells expressing both TVA950-2A-CVS11G and RV-GFP. Scale bar, 200 µm and 20 µm (high magnification). c-e, Specificity of TVA950-2A-CVS11G expression. Example images show co-expression of TVA950-2A-CVS11G and c, VIP, d, PV and e, SOM in VIP-cre, PV-cre and SOM-cre mice, respectively. Scale bars, 20 µm. f, Quantification of co-localization of 2A peptide with interneuron markers detected by immunohistochemistry (VIP, N=1 mouse; PV, N=2; SOM, N=2). Data is presented as mean and s.e.m., circles are individual data points. g, Representative image illustrating absence of RV-GFP expression in the BLA without preceding TVA950-2A-CVS11G injection. Scale bar, 200 µm. Confocal images are representative for N=6 mice (a, d), N=9 (b, e), N=7 (c), N=3 (g).

Supplementary Figure 3 Monosynaptic inputs to VIP, PV and SOM interneurons in the BLA.

a, Serial reconstruction of representative example mouse brains depicting monosynaptic inputs to VIP (top), PV (middle) and SOM (bottom) BLA interneurons. Corresponding injection sites are shown in Fig. 2 (VIP, mouse #8 in Fig. 2 heat map, 49% LA starter cells) and Supplementary Fig. 2 (PV, mouse #6, 58% LA; SOM, mouse #8, 38% LA). Top row displays matching mouse brain atlas planes. MO, medial orbital cortex; Ai, agranular insular cortex; BF, basal forebrain; Pir, piriform cortex; dMT, dorsal midline thalamic nuclei; LOT, nucleus of the lateral olfactory tract; CxA, cortex-amygdala transition zone; VMH, ventromedial hypothalamus; PLCo, posterolateral cortical amygdaloid nucleus; BLA, basolateral amygdala; AuT, auditory thalamus; PIL, posterior intralaminar thalamus; AuC, auditory cortices; RhC, rhinal cortices; vHC, ventral hippocampus; DR, dorsal raphe nucleus. Scale bar, 1 mm. b, Fraction of inputs over total input numbers for each identified brain area projecting to VIP, PV and SOM BLA interneurons (VIP, N=8 mice; PV, N=6; SOM, N=9). c, A subset of basal forebrain presynaptic inputs to VIP BLA interneurons expresses choline acetyltransferase (ChAT, yellow arrowheads; 19.7±4.6%, N=3). LPO, lateral preoptic area; VP, ventral pallidum; SI, substantia innominata, basal part; HDB, nucleus of the horizontal limb of the diagonal band; MCPO, magnocellular preoptic nucleus. Scale bars, 200 µm and 20 µm (high magnification). Confocal images are representative for N=7 mice (a, VIP), N=6 (a, PV), N=9 (a, SOM), N=3 (c). Box-and-whisker plots show median values and 25th/75th percentiles with 10th to 90th percentile whiskers, dots additionally indicate the mean. Details of rabies tracing analysis are specified in Supplementary Table 1.

Supplementary Figure 4 Interconnectivity of BLA interneuron subtypes.

a, ChR2-EYFP in VIP BLA interneurons. Top left, example recording of action potential generation by ChR2 activation with blue light in cell-attached mode. Scale bars, 20 pA, 2 ms. Top right, average latency to light-evoked action potentials by ChR2 activation (n=7 cells). Bottom, confocal image of the example cell expressing ChR2 filled with biocytin during whole-cell recordings to confirm VIP expression. Scale bar, 20 µm. b-d, Example traces of IPSCs evoked by VIP BLA network photostimulation before and after application of the GABAA-receptor antagonist picrotoxin in b, PV interneurons, c, SOM interneurons and d, PNs of the BLA. Scale bars, 100 pA, 10 ms. e, The amplitude of light-evoked IPSCs is significantly reduced by picrotoxin in PV (ratio paired t-test, two-sided, t(3)=10.31, P=0.0019, n=4) and SOM (ratio paired t-test, two-sided, t(3)=15.15, P=0.0006, n=4) interneurons as well as PNs (ratio paired t-test, two-sided, t(3)= 6.988, P=0.0060, n=4). f, Position of recorded and reconstructed cells matched to a mouse brain atlas, symbol size refers to input strength upon VIP activation (PV, n=60; SOM, n=46, PN, n=70). g, Top, example recording from a VIP BLA interneuron receiving short-latency inhibitory inputs upon PV BLA interneuron network activation with ChR2 (green bar). Scale bars, 100 pA, 10 ms. Bottom, corresponding confocal image confirming VIP expression in the biocytin-filled cell. Scale bar, 20 µm. h, Same for a SOM BLA interneuron. i, High connectivity from PV BLA interneurons to VIP and SOM interneurons (VIP, 97.6%, 40 of 41 cells from N=3 mice; SOM, 95.3%, 41 of 43 cells from N=3 mice). j, IPSC amplitudes are higher in VIP BLA interneurons compared to SOM interneurons (Mann-Whitney U test, two-sided, P<0.0001; VIP, n=40; SOM, n=41). k, Charge transfer in VIP BLA interneurons is larger compared to SOM interneurons (Mann-Whitney U test, two-sided, P=0.0002; VIP, n=40; SOM, n=41). l, Position of recorded and reconstructed cells matched to a mouse brain atlas, symbol size refers to input strength upon PV activation (see panel f for legend; VIP, n=41; SOM, n=43, PN, n=35). m, Differential inputs from PV interneurons in the lateral (LA) and basal (BA) subdivisions of the BLA. Connectivity is not significantly different within LA and BA subpopulations, but amplitudes differ with overall larger inputs in the BA (LA: Kruskal-Wallis test, H=26.20, P<0.0001; Dunn’s multiple comparisons test VIP vs SOM, P=0.0003, SOM vs PN, P<0.0001; VIP, n=20, SOM, n=20, PN, n=14; BA: Kruskal-Wallis test, H=28.60, P<0.0001; Dunn’s multiple comparisons test VIP vs PN, P=0.0083, SOM vs PN, P<0.0001; VIP, n=20 cells, SOM, n=21, PN, n=20). n, Top, example traces from a VIP BLA interneuron receiving short-latency inhibitory inputs by brief SOM BLA interneuron network activation with ChR2 (red bar). Scale bars, 100 pA, 10 ms. Bottom, corresponding confocal image confirming VIP expression in the recorded cell. Scale bar, 20 µm. o, Same for a PV BLA interneuron. p, High connectivity from SOM BLA interneurons to VIP and SOM interneurons (VIP, 85.7%, 42 of 49 cells from N=4 mice; PV, 88.1%, 37 of 42 cells from N=3 mice). q-r, Neither IPSC q, amplitude nor r, charge transfer upon SOM BLA network photostimulation are different between VIP and PV interneurons (VIP, n=42; PV, n=37). s, Position of recorded and reconstructed cells matched to a mouse brain atlas, symbol size refers to input strength upon SOM activation (see panel f for legend; VIP, n=49; PV, n=42, PN, n=33). t, Differential inputs from SOM interneuron in the lateral (LA) and basal (BA) subdivisions of the BLA. Connectivity is not significantly different within LA and BA subpopulations, but amplitudes differ (LA: Kruskal-Wallis test, H=10.88, P=0.0043; Dunn’s multiple comparisons test VIP vs PN, P=0.0032; VIP, n=22, PV, n=9, PN, n=13; BA: Kruskal-Wallis test, H=27.95, P<0.0001; Dunn’s multiple comparisons test VIP vs PN, P=0.0020, PV vs PN, P<0.0001; VIP, n=20 cells, PV, n=28, PN, n=20). Confocal images and example traces are representative for n=7 cells (a), n=41 (g), n=43 (h), n=49 (n), n=42 (o). Traces in b-d, f-g, k-l are 20 individual trials (gray) with corresponding average IPSC (black) from one cell. Dots in e represent individual data points, horizontal lines additionally indicate the mean. Circles/triangles in m and t represent mean amplitude with s.e.m. Box-and-whisker plots show median values and 25th/75th percentiles with 10th to 90th percentile whiskers, dots additionally indicate the mean. ** P<0.01, *** P<0.001. Further details of slice electrophysiology analysis are summarized in Supplementary Table 2. Additional details of statistical analysis are listed in Supplementary Table 3.

Supplementary Figure 5 Postsynaptic effects of ChR2-dependent VIP BLA interneuron activation.

a, Fractions of BLA cells inhibited by optogenetic VIP interneuron network stimulation in vitro significantly differs in PNs compared to PV and SOM interneurons (PV, 79%, 11 of 14 cells; SOM, 92%, 11 of 12 cells; PN, 8%, 1 of 12 cells; Pearson’s χ2 test, P<0.0001, Fisher’s exact test, two-sided, PV vs PN, P=0.0005, SOM vs PN, P=0.0001). b, Location of electrode tips and optical fiber placement for in vivo single-unit recordings. Blue shades indicate maximum virus spread (N=6 mice). LA, lateral amygdala; BA, basal amygdala; CEA, central amygdala. c, Electrophysiological identification of putative interneurons and projection neurons in vivo. Recorded BLA neurons (n=68 cells from N=6 mice) were classified as putative interneurons (n=6, orange circles) or putative projection neurons (n=62, gray circles) based on three extracellular electrophysiological properties: firing frequency, spike half width and afterhyperpolarization (AHP, see methods). Insets show corresponding normalized action potential waveforms. d, Examples of putative BLA interneurons inhibited by VIP interneuron photostimulation (2 s). Dots indicate individual action potentials in repeated trials, bottom PSTH shows average frequency across 100 trials. e, Fraction of putative BLA interneurons significantly inhibited or disinhibited by VIP interneuron photostimulation (n=6, left) and corresponding average z-scored activity profiles (right; inhibited, n=3, disinhibited, n=1, not responsive, n=2). f, Mean z-scored activity of all recorded putative PNs (n=62) and interneurons (n=6) averaged independent of response profile. Traces in e and f are mean with s.e.m. or traces from individual cells if n<3. *** P<0.001. Details of statistical analysis are listed in Supplementary Table 3.

Supplementary Figure 6 ArchT expression in VIP BLA interneurons.

a, Representative example image of ArchT-GFP expression in VIP interneurons in the BLA of VIP-cre mice. Scale bar, 20 µm. b, Quantification of co-localization of ArchT-GFP expression with VIP detected by immunohistochemistry (N=3 mice). c, Representative patch-clamp recording of an ArchT-GFP expressing VIP BLA interneuron. Top, suppression of spontaneous action potential generation by 4.5 s yellow light. Scale bars, 20 mV, 500 ms. Middle, ArchT activation with yellow light diminishes action potentials evoked by depolarizing current steps (-50 pA, 0 pA, and +50 pA current injections while holding the cell at -60 mV). Scale bars, 20 mV, 200 ms. Bottom, confocal image of the same ArchT-GFP+ cell filled with biocytin during whole-cell recordings to confirm VIP expression. Scale bar, 10 µm. d, Spontaneous action potentials are reliably inhibited by application of yellow light (Wilcoxon matched-pairs signed rank test, two-sided, P=0.001, n=11 cells). e, ArchT activation decreases excitability of VIP BLA interneurons. Spike rate was normalized to the maximum frequency in baseline condition for each cell. Sigmoidal curve fitting reveals a significant shift to the right of input-output curves with ArchT activation (15.7 pA shift; Ihalf baseline 99.9 pA, Ihalf light 115.4 pA; paired t-test, two-sided, t(11)=2.356, P=0.0381; n=12 cells) and decreased gain (slope baseline 43.5%/pA, light 36.6%/pA; paired t-test, two-sided, t(11)=2.740, P=0.0192; n=12 cells) without affecting maximum output. f, Representative example traces from a VIP BLA interneuron expressing ArchT-tdTomato, demonstrating reliable spike suppression with yellow, but not blue light. Further, only yellow but not blue light activates ArchT at a holding potential of -60 mV, leading to membrane potential hyperpolarization. Scale bars, 20 mV, 500 ms. g, Yellow light significantly decreases spike probability in VIP BLA interneurons, while blue light has no effect (Friedman test, F=40.71, P<0.0001; Dunn’s multiple comparisons test, C yellow vs C yellow +light, P<0.0001, C yellow +light vs C blue +light, P=0.0002; n=15 cells). Note that blue light used for nVoke in vivo imaging experiments was further of shorter wavelength and lower intensity (448 nm, 0.4-0.7 mW) compared to slice electrophysiology to exclude unwanted cross-excitation of ArchT. h, Yellow light (yellow line; 590 nm, 12 mW at objective, 20 s) does not affect Ca2+ fluorescence in VIP BLA interneurons expressing GCaMP6 in vivo (n=95 cells from N=3 mice, trace represents mean and s.e.m.). Scale bars, 0.5% ∆F/F, 10 s. i, Yellow light induces a decrease in Ca2+ fluorescence in VIP BLA co-expressing GCaMP6 and ArchT-tdTomato (n=80 from N=3 mice). Scale bars, 0.5% ∆F/F, 10 s. j, Average amplitude during yellow light application (20 s) is significantly different between GCaMP6-only controls and VIP interneurons expressing GCaMP6 with ArchT (Mann-Whitney U test, two-sided, P<0.0001; control, n=95; ArchT, n=80). k, Schematic illustrating reconstructed implant sites of GRIN lenses within the BLA for VIP nVoke experiments shown in Fig. 5 matched to a mouse brain atlas (gray lines, GCaMP6 in VIP, N=4 mice; yellow lines, GCaMP6 and ArchT in VIP, N=3). LA, lateral amygdala; BA, basal amygdala; CEA, central amygdala. Confocal images and example traces are representative for N=3 mice (a), n=12 cells (c), n=15 (f). Circles in b show individual data points, connected circles in d and g represent individual paired data points, horizontal lines additionally indicate the mean. Box-and-whisker plots show median values and 25th/75th percentiles with 10th to 90th percentile whiskers, dots additionally indicate the mean. All other data is presented as mean and s.e.m. * P<0.05, *** P<0.001. All details of statistical analysis are listed in Supplementary Table 3.

Supplementary Figure 7 Combined deep brain calcium imaging and optogenetic manipulation.

a, Representative example image showing concomitant expression of CaMKII-GCaMP6 and cre-dependent ArchT-tdTomato in the BLA of VIP-cre mice. Immunohistochemical counterstaining against CaMKII and VIP confirms specificity of viral constructs. Scale bar, 20 µm. b, Implant sites of GRIN lenses (lines) and maximum tdTomato/ArchT-tdTomato in VIP virus spread (shades) for CaMKII-CGaMP6 nVoke imaging experiments shown in Fig. 5 (left: control, tdTomato in VIP, N=6 mice; right: ArchT-tdTomato in VIP, N=4). LA, lateral amygdala; BA, basal amygdala; CEA, central amygdala. c, Average CS and US responses for all pairings for cells clustered based on their CS activity pattern during the last three trials illustrating CS responsive PNs (CS-up pattern, Cluster 1, n=132 cells from N=10 mice) or CS non-responsive PNs (Cluster 2: active during both baseline and CS in trials 4-6, n=184; Cluster 3: showing no activity during baseline and CS in trials 4-6, n=229) from both control and ArchT mice. Scale bar, 0.2% ∆F/F. d, Inhibition of VIP interneurons during the US with ArchT significantly changes CS activity patterns in BLA PNs (Pearson’s χ2 test, P=0.0001; control, n=349; ArchT, n=196). e, Difference in US responses between light and no-light trials for CS-up PNs (Cluster 1; control, n=104; ArchT, n=28). Confocal images are representative for N=4 mice (a). Box-and-whisker plots show median values and 25th/75th percentiles with 10th to 90th percentile whiskers, dots additionally indicate the mean. *** P<0.001. Details of statistical analysis are listed in Supplementary Table 3.

Supplementary Figure 8 Optogenetic inhibition of VIP BLA interneurons during auditory fear conditioning.

a, Schematic illustrating the entire 5-day behavioral paradigm used for optogenetic loss-of-function experiments, including details about CS, US and yellow light pattern applied. b-c, Representative example images of bilateral expression of b, ArchT-GFP and c, GFP in VIP interneurons in the BLA of VIP-cre mice with corresponding optical fiber placement (dashed lines). Scale bar, 200 µm. d, Position of optical fiber tips (symbols) and maximum virus spread (shades) in all mice included in optogenetic experiments matched to a mouse brain atlas. LA, lateral amygdala; BA, basal amygdala; CEA, central amygdala. e, CS presentations on habituation day do not induce freezing in naïve mice. f, Optogenetic inhibition of VIP BLA interneurons has no effect on freezing during or after yellow light stimulus presentation in naïve mice (ISI, inter-stimulus interval). g, Similarly, light stimulation during the habituation session does not affect running speed in either of the light-treated groups. h, Maximum acceleration during the aversive US during fear conditioning. i, Left to right, distance travelled, maximum speed and maximum acceleration during the aversive US during reconditioning. j, Post-shock freezing during reconditioning (Kruskal-Wallis test, H=6.437, P=0.04; Dunn’s multiple comparisons test, ArchT vs GFP, P=0.0398). k, Optogenetic inhibition of VIP BLA interneurons for 4.5 s or 10 s at the end of the retrieval 2 session does not affect freezing behavior. For panels d-k: ArchT, N=14 mice; GFP, N=11; ArchT no light ctrl, N=12. Confocal images are representative for N=26 mice (b), N=11 (c). Box-and-whisker plots show median values and 25th/75th percentiles with 10th to 90th percentile whiskers, dots additionally indicate the mean. * P<0.05. Details of statistical analysis are listed in Supplementary Table 3.

Supplementary Figure 9 Optogenetic activation of VIP BLA interneurons does not induce associative fear learning.

a, Schematic illustrating the entire behavioral paradigm used for optogenetic gain-of-function experiments, including details about CS, US and blue light pattern applied. b-c, Representative example images of bilateral expression of b, ChR2-EYFP and c, GFP in VIP interneurons in the BLA of VIP-cre mice with corresponding optical fiber placement (dashed lines). Scale bar, 200 µm. d, Position of optical fiber tips (squares) and maximum virus spread (shades) in all mice included in optogenetic experiments matched to a mouse brain atlas. LA, lateral amygdala; BA, basal amygdala; CEA, central amygdala. e, CS presentations on habituation day do not induce freezing in naïve mice. f-g, Optogenetic activation of VIP BLA interneurons has no effect on f, freezing or g, running speed during or after blue light stimulus presentation in naïve mice (ISI, inter-stimulus interval). h, Optogenetic activation of VIP BLA coupled to CS presentations does not induce associative fear learning as measured by freezing responses to the CS on retrieval day. i, Similarly, no difference is detected in the average speed during the CS between groups. j, Optogenetic activation of VIP BLA interneurons for 2 s or 10 s on a separate laser test day does not induce freezing behavior. For panels d-j: ChR2, N=8 mice; GFP, N=9. Confocal images are representative for N=8 mice (b), N=9 (c). Box-and-whisker plots show median values and 25th/75th percentiles with 10th to 90th percentile whiskers, dots additionally indicate the mean. Details of statistical analysis are listed in Supplementary Table 3.

Supplementary Figure 10 Plasticity of CS and US responses in VIP BLA interneurons during learning.

a, Full repeated fear conditioning paradigm. b, Implant sites of GRIN lenses (blue lines, N=7) within the BLA of VIP-cre mice for repeated fear conditioning experiments. c, Average mean CS responses increase across trials of repeated fear conditioning (Friedman test, F=85.63, P<0.0001; Dunn’s multiple comparisons test comparing CS+ responses to preceding trial, D1 CS2 vs CS3, P<0.0001, D2 CS1 vs CS2, P=0.0013, D2 CS1 vs CS2, P=0.0112, D2 CS2 vs CS3, P=0.0133; n=201 cells from N=7 mice). d, Average mean US responses decrease across trials of repeated fear conditioning (Friedman test, F=185.4, P<0.0001; Dunn’s multiple comparisons test comparing US responses to preceding trial, D1 CS2 vs CS3, P=0.0004; n=201). e, Full conditioning paradigm applying unpredicted weak and predicted strong foot shocks after association of weak foot shocks with a short auditory CS. f, Implant sites of GRIN lenses (blue lines, N=5). g, Example Ca2+ traces from VIP BLA interneurons during conditioning. Arrowheads indicate time points of CS presentations (gray) with weak US (red). Scale bars, 10% ∆F/F, 60 s. h, Mean CS/US response from all recorded VIP BLA interneurons (n=122 cells from N=5 mice) averaged across six trials during conditioning. Scale bars, 0.5% ∆F/F, 5 s. i, Corresponding average US responses for individual trials (n=122). Scale bars, 0.5% ∆F/F, 2 s. j, Decrease of mean US amplitudes across trials (Friedman test, F=33.71, P<0.0001; Dunn’s multiple comparisons test comparing US responses to preceding trial, US5 vs US6, P=0.0058, n=122). k, Fraction of VIP BLA interneurons significantly modulated by the US in trial 1 and trial 3 (Pearson’s χ2 test, P=0.0261, n=122) during conditioning. l, Mean CS responses in VIP BLA neurons (n=111, N=5) averaged for trials with subsequent US (black) and CS only trials (blue). Scale bars, 0.5% ∆F/F, 5 s. m, Fraction of VIP BLA interneurons significantly modulated by the US in trial 1 and first unexpected trial 3 (n=111). n, Mean CS/US responses in VIP BLA neurons (n=108, N=5) averaged for trials with subsequent weak (black) or strong US (red). Scale bars, 0.5% ∆F/F, 5 s. o, Fraction of VIP BLA interneurons significantly modulated by the US in trial 1 and first stronger trial 3 (Pearson’s χ2 test, P=0.005; Fisher’s exact test US excited, two-sided, P=0.0312, n=108). Data is shown as mean and s.e.m. * P<0.05, ** P<0.01, *** P<0.001. Details of statistical analysis are listed in Supplementary Table 3.

Supplementary information

Supplementary Information

Supplementary Figs. 1–10 and Supplementary Tables 1–4.

Reporting Summary

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Krabbe, S., Paradiso, E., d’Aquin, S. et al. Adaptive disinhibitory gating by VIP interneurons permits associative learning. Nat Neurosci 22, 1834–1843 (2019) doi:10.1038/s41593-019-0508-y

Download citation