The formation and retrieval of conditioned fear memories critically depend on the amygdala. Here we identify an inhibitory projection from somatostatin-positive neurons in the central amygdala to parvalbumin-positive neurons in the zona incerta that is required for both recent and remote fear memories. Thus, the amygdala inhibitory input to parvalbumin-positive neurons in the zona incerta, a nucleus not previously implicated in fear memory, is an essential component of the fear memory circuitry.
Subscribe to Journal
Get full journal access for 1 year
only $4.92 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
The DNA sequence of the DIO-SynaptoTag plasmid has been deposited in GenBank under accession code MH883617. The data that support the findings of this study are available from the corresponding authors upon reasonable request.
Maren, S. Neurobiology of Pavlovian fear conditioning. Annu. Rev. Neurosci. 24, 897–931 (2001).
Fanselow, M. S. & Poulos, A. M. The neuroscience of mammalian associative learning. Annu. Rev. Psychol. 56, 207–234 (2005).
Xu, W. & Südhof, T. C. A neural circuit for memory specificity and generalization. Science 339, 1290–1295 (2013).
Do-Monte, F. H., Quiñones-Laracuente, K. & Quirk, G. J. A temporal shift in the circuits mediating retrieval of fear memory. Nature 519, 460–463 (2015).
Penzo, M. A. et al. The paraventricular thalamus controls a central amygdala fear circuit. Nature 519, 455–459 (2015).
Vetere, G. et al. Chemogenetic interrogation of a brain-wide fear memory network in mice. Neuron 94, 363–374.e4 (2017).
Frankland, P. W. & Bontempi, B. The organization of recent and remote memories. Nat. Rev. Neurosci. 6, 119–130 (2005).
Mitrofanis, J. Some certainty for the “zone of uncertainty”? Exploring the function of the zona incerta. Neuroscience 130, 1–15 (2005).
Liu, K. et al. Lhx6-positive GABA-releasing neurons of the zona incerta promote sleep. Nature 548, 582–587 (2017).
Zhang, X. & van den Pol, A. N. Rapid binge-like eating and body weight gain driven by zona incerta GABA neuron activation. Science 356, 853–859 (2017).
Chou, X. L. et al. Inhibitory gain modulation of defense behaviors by zona incerta. Nat. Commun. 9, 1151 (2018).
Xu, W. et al. Distinct neuronal coding schemes in memory revealed by selective erasure of fast synchronous synaptic transmission. Neuron 73, 990–1001 (2012).
Wheeler, A. L. et al. Identification of a functional connectome for long-term fear memory in mice. PLoS Comput. Biol. 9, e1002853 (2013).
Biseul, I. et al. Fear recognition is impaired by subthalamic nucleus stimulation in Parkinson’s disease. Neuropsychologia 43, 1054–1059 (2005).
Okun, M. S. et al. Pseudobulbar crying induced by stimulation in the region of the subthalamic nucleus. J. Neurol. Neurosurg. Psychiatry 75, 921–923 (2004).
Essig, J. & Felsen, G. Warning! Dopaminergic modulation of the superior colliculus. Trends Neurosci. 39, 2–4 (2016).
Chen, J. & Kriegstein, A. R. A GABAergic projection from the zona incerta to cortex promotes cortical neuron development. Science 350, 554–558 (2015).
Li, H. et al. Experience-dependent modification of a central amygdala fear circuit. Nat. Neurosci. 16, 332–339 (2013).
Haubensak, W. et al. Genetic dissection of an amygdala microcircuit that gates conditioned fear. Nature 468, 270–276 (2010).
Wilensky, A. E. et al. Rethinking the fear circuit: the central nucleus of the amygdala is required for the acquisition, consolidation, and expression of Pavlovian fear conditioning. J. Neurosci. 26, 12387–12396 (2006).
Zolotukhin, S. et al. Recombinant adeno-associated virus purification using novel methods improves infectious titer and yield. Gene Ther. 6, 973–985 (1999).
Beier, K. T. et al. Circuit architecture of VTA dopamine neurons revealed by systematic input-output mapping. Cell 162, 622–634 (2015).
De Franceschi, G., Vivattanasarn, T., Saleem, A. B. & Solomon, S. G. Vision guides selection of freeze or flight defense strategies in mice. Curr. Biol. 26, 2150–2154 (2016).
Otsuka, S. et al. Roles of Cbln1 in non-motor functions of mice. J. Neurosci. 36, 11801–11816 (2016).
Stephenson-Jones, M. et al. A basal ganglia circuit for evaluating action outcomes. Nature 539, 289–293 (2016).
We thank K. Beier for help with the rabies tracing experiments; F. Luo for help with the visually induced freezing behavior; J. Liang, S. Lu, and R. Land for assistance; and L. Chen for feedback on the manuscript. This study was supported by a National Institutes of Health (NIH)/National Institutes of Mental Health grant (no. MH052804 to T.C.S.) and a NIH/National Institute of Neurological Disorders and Stroke grant (no. 1K01NS105155 to M.Z.).
The authors declare no competing interests.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Integrated supplementary information
(a) Top, illustration of bilateral injection of AAV CAG-eGFP-2A-TetTox into ZI. Bottom, representative images of eGFP expression in different brain sections, arranged from rostral to caudal. (b) Sample images of eGFP expression in different brain slices after injection of AAVs expressing double-loxP-flanked TetTox and eGFP (CAG-DIO-eGFP-2A-TetTox) into the ZI of PV-Cre mice. Experiments in a and b were independently replicated with similar results more than ten times.
Supplementary Figure 2 Behavioral characterization of mice after inactivation of all synaptic outputs from the medial ZI by expression of TetTox.
(a) Illustration of bilateral injections of AAVs expressing TetTox and eGFP (CAG-eGFP-2A-TetTox; test) or only eGFP (CAG-eGFP; control) into the ZI. (b) Body weight of test and control mice at P50 and P100. (c–k) Performance of test and control mice in a series of behavioral assays (c, open field; d, spontaneous alternating Y-maze; e, elevated plus maze; f, novel object recognition; g, three chamber social interaction paradigm; h, acoustic startle response; i, pre-pulse inhibition; j, rotarod; and k, water T maze. In g, “S” stands for “social”, “C” stands for “center” and “NS” stands for “non-social”. Quantitative data are means ± s.e.m. Statistics were based on two-sided unpaired t tests or Mann-Whitney U tests (for datasets that were not normally distributed) for two-group comparisons. P < 0.05 was considered significant (*P < 0.05, **P < 0.01, ***P < 0.001). In c, P = 0.13, 0.03. In j, P = 0.11, 0.11, 0.034, 0.013, 5 × 10−4, 0.004, 0.009, 0.019, 2 × 10−4. n = 7, 8 for control and TetTox injected mice respectively.
Supplementary Figure 3 Fear memory deficits of ZI TetTox-injected mice could not be explained by reduced foot shock responses or freezing expression.
(a) Running speed of test and control mice (the same cohort analyzed in Fig. 1f) before, during and after delivery of foot shock. P = 0.01 (during second foot shock), 0.02 (during third foot shock). (b) Running speed of test and control mice (the same cohort analyzed in Fig. 1r) before, during and after delivery of foot shock. P = 0.01 (post second foot shock), 0.01 (pre third foot shock), 0.01 (post third foot shock). (c) Freezing percentage of test (DIO-TetTox bilaterally injected in ZI of PV-Cre mice) and control mice (DIO-eGFP bilaterally injected in ZI of PV-Cre mice) before and after being exposed to sweeping visual stimuli. (d) Freezing percentage of test and control mice (the same cohort used in c) before and after being exposed to TMT. Quantitative data are means ± s.e.m. Statistics were based on two-sided unpaired t tests or Mann-Whitney tests (for datasets that were not normally distributed) for two-group comparisons. P < 0.05 was considered significant (*P < 0.05, **P < 0.01, ***P < 0.001). All numbers of mice used are presented in right bars.
Supplementary Figure 4 Behavioral characterization of mice after inactivation of synaptic outputs from only PV+ neurons in the ZI.
The panels are arranged for the same experiments as in Supplementary Figure 2 in the same order, but experiments were performed with mice in which TetTox and eGFP or only eGFP were expressed in PV+-neurons of the ZI. This was achieved by infecting ZI neurons in PV-Cre mice with AAVs containing double-loxP-flanked coding regions for TetTox and eGFP or eGFP alone. Note that for the behaviors assayed, the inactivation of all neurons in the ZI or of only PV+-neurons produced similar impairment patterns. Quantitative data are means ± s.e.m. Statistics were based on two-sided unpaired t tests or Mann-Whitney tests (for datasets that were not normally distributed) for two-group comparisons. P < 0.05 was considered significant (*P < 0.05, **P < 0.01, ***P < 0.001). In c, P = 0.36, 0.008. In e, P = 0.006. In h, P = 0.3, 0.2, 0.5, 0.13, 0.005. In i, P = 0.013, 0.024, 0.019, 0.041. In j, P = 0.36, 0.17, 0.009, 0.03, 0.004, 0.003, 7 × 10−4, 1 × 10−4, 8 × 10−4. n = 9, 14 for control and TetTox injected mice respectively.
Supplementary Figure 5 Examples of additional brain regions that provide presynaptic inputs to PV+ neurons in the ZI besides those shown in Fig. 2c.
Brain regions shown in the panels are labeled by abbreviations that are explained on the top right. Dashed lines indicate approximate boundaries between various brain nuclei. Experiments were independently replicated with similar results five times.
Supplementary Figure 6 Retrograde tracing using Alexa488-conjugated cholera toxin subunit B (CTB-488) reproduces the presynaptic inputs identified by retrograde rabies tracing in Fig. 2c,d and Supplementary Fig. 5, and defines additional inputs that may be targeting non-PV+ neurons in the ZI.
(a) Illustration of the CTB-488 injection into the left ZI (left), and example image showing the injection site (right). (b–j) Representative images arranged in a rostral to caudal sequence to display CTB-488 labelled neurons in different brain areas after ZI injections of CTB-488. Please note that the exposures were increased compared to a to reveal presynaptic neurons. As a result, there are strong false positive signals around injection site in g-i. Abbreviations used in the various panels are explained on the top right. (k–n) Representative images illustrating the co-localization of CTB retrogradely labelled CeA neurons with SST+- and PRKCD+-neurons. The same injection scheme was used as in a, except that SST-Ai14 mice were used. PRKCD+-neurons were recognized by antibody staining. (o–r) Enlarged images of the boxed area shown in k-n. Experiments in a–j and k-r were independently repeated with similar results four and three times, respectively.
(a) Experimental strategy with channelrhodopsin (ChR2-eYFP) and TetTox (TetTox-eGFP) co-expression in the CeA of PV-Cre/Ai14 mice. Postsynaptic currents recorded from PV+-neurons were evoked by optogenetic stimulation in the presence of TTX and 4-AP. (b) Synaptic response amplitude plot. n = 21 recorded PV+-neurons. n.d.: not detected.
Supplementary Figure 8 Verification of injection sites for AAVs expressing WGA-Cre in the ZI and TetTox-eGFP in the CeA.
(a) Experimental strategy. (b) Representative image showing the expression of eGFP in CeA. (c) Top, representative image showing the expression of mCherry and WGA-Cre in the ZI and the green axons from CeA neurons that innervate the ZI. Bottom, enlarged confocal image showing the boxed area. The ZI area was delineated by white dashed lines. LHb, lateral habenula; STN, subthalamic nucleus; PSTN, parasubthalamic nucleus. (d,e) Different sections illustrating the collateral projections of CeA-ZI pathway. PF, parafascicular nucleus; SNc, substantia nigra compact part. Experiments were independently repeated with similar results more than ten times. Note that WGA-Cre ‘starter cells’ (that is, the initially infected cells) are red owing to co-expressed mCherry; WGA-Cre mediates retrograde trans-synaptic transfer of Cre to presynaptic inputs, but not of mCherry, and thus input cells are not red. Input cells in CeA, however, are green because the retrogradely transported Cre activates the double floxed TetTox-2A-eGFP expression, thereby silencing CeA-ZI projections. CeA-ZI pathway also have other collateral projections (labelled by arrows).
Supplementary Figure 9 SynaptoTag tracing of efferent targets of ZI PV+ neurons identifies neurons in thalamic regions, the superior colliculus, the periaqueductal gray and the ZI itself as postsynaptic targets of ZI PV+ neurons.
(a) Illustration of the injection of AAVs expressing double-loxP-flanked SynaptoTag (DIO-SynaptoTag) into the left ZI of PV-Cre mice. (b) Schematic of the DIO-SynaptoTag construct. (c) Representative images showing the AAV injection site in the ZI and presynaptic terminals formed by ZI PV+-neurons in the thalamus (left overview; right, enlarged area from the boxed region in the overview). Note that in the enlarged image the red fluorescence intensity was lowered to reveal the location of infected neurons in ventral ZI. Green fluorescence, representing presynaptic terminals, could be observed in ZI itself and in the ventral posteromedial thalamic nucleus (VPM) and the posterior complex of the thalamus (PO). (d) Representative images showing strong presynaptic SynaptoTag signals in the superior colliculus (SC) and the periaqueductal gray (PAG) of the midbrain (left overview; right, enlarged area from the boxed region in the overview). Note that the presynaptic terminals are located in the lateral and deeper layer of the SC, as well as in the lateral part of PAG. (e,f) High magnification images showing the red fluorescence in the cell body and axon fibers and the green fluorescence in presynaptic terminals in the ZI (e) and the VPM (f). Experiments were independently repeated with similar results three times.
Supplementary Figure 10 Silencing medial ZI–lateral SC pathway does not replicate the fear memory deficits observed in ZI TetTox-injected mice.
(a) Stereotaxic viral injection strategy. (b) Representative image showing the expression of eGFP in medial ZI. (c) Representative image showing the expression of mCherry and WGA-Cre in the lateral SC and the green axons from ZI neurons that innervate the lateral SC. Note that mCherry signal is in the lateral SC without spreading to PAG, and eGFP+-neurons are restricted to the ventral ZI, where PV+-neurons are supposed to be located. Imaging experiments were independently repeated with similar results five times. (d) Tests of memory acquisition, recent memory and remote memory after blocking medial ZI-lateral SC pathway. Quantitative data are means ± s.e.m. Statistics were based on two-sided unpaired t tests or Mann-Whitney tests (for datasets that were not normally distributed). P = 0.043 for the comparison of contextual recent memory. n = 10, 10 for control and TetTox injected mice respectively. Note that the same cohort of mice were used for recent and remote fear memory test.
Supplementary Figure 11 Silencing medial ZI-lateral PAG pathway does not replicate the fear memory deficits observed in ZI TetTox-injected mice.
(a–d) The panels are arranged for the same experiments as in Supplementary Figure 10 in the same order, but experiments were performed with WGA-Cre and mCherry expressed in the lateral PAG. Representative images in b-c were independently repeated with similar results five times. Quantitative data are means ± s.e.m. Statistics were based on two-sided unpaired t tests or Mann-Whitney tests (for datasets that were not normally distributed). P < 0.05 was considered significant. n = 7, 7 for control and TetTox injected mice respectively. Note that the same cohort of mice were used for recent and remote fear memory test.
Supplementary Figure 12 Optogenetically activating ZI PV+ neurons does not directly drive freezing behavior.
(a) Stereotaxic viral injection and optic fiber implantation strategy. (b) Representative image showing the expression of ChR2 in ZI PV+-neurons. Experiments were independently repeated with similar results three times. (c) Representative trace of membrane potential change in response to a 15 Hz train of blue laser stimulation. (d) Summary of animals’ freezing behavior in response to five repeats of 15 Hz light stimulation (see methods). Statistics were based on two-sided paired t tests or Mann-Whitney tests (for datasets that were not normally distributed). P < 0.05 was considered significant. n = 4 mice. Different stimulation frequencies were tested and none of them were found to drive freezing (data not shown).
About this article
Cite this article
Zhou, M., Liu, Z., Melin, M.D. et al. A central amygdala to zona incerta projection is required for acquisition and remote recall of conditioned fear memory. Nat Neurosci 21, 1515–1519 (2018). https://doi.org/10.1038/s41593-018-0248-4
Direct Septum-Hippocampus Cholinergic Circuit Attenuates Seizure Through Driving Somatostatin Inhibition
Biological Psychiatry (2020)
Molecular Psychiatry (2020)
Journal of Clinical Investigation (2020)