Abstract

Dopamine modulates medial prefrontal cortex (mPFC) activity to mediate diverse behavioural functions1,2; however, the precise circuit computations remain unknown. One potentially unifying model by which dopamine may underlie a diversity of functions is by modulating the signal-to-noise ratio in subpopulations of mPFC neurons3,4,5,6, where neural activity conveying sensory information (signal) is amplified relative to spontaneous firing (noise). Here we demonstrate that dopamine increases the signal-to-noise ratio of responses to aversive stimuli in mPFC neurons projecting to the dorsal periaqueductal grey (dPAG). Using an electrochemical approach, we reveal the precise time course of pinch-evoked dopamine release in the mPFC, and show that mPFC dopamine biases behavioural responses to aversive stimuli. Activation of mPFC–dPAG neurons is sufficient to drive place avoidance and defensive behaviours. mPFC–dPAG neurons display robust shock-induced excitations, as visualized by single-cell, projection-defined microendoscopic calcium imaging. Finally, photostimulation of dopamine terminals in the mPFC reveals an increase in the signal-to-noise ratio in mPFC–dPAG responses to aversive stimuli. Together, these data highlight how dopamine in the mPFC can selectively route sensory information to specific downstream circuits, representing a potential circuit mechanism for valence processing.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Data availability

Data, unprocessed and unpseudocoloured images, and custom MATLAB codes are available upon request.

Additional information

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

References

  1. 1.

    Arnsten, A. F. T. Stress signalling pathways that impair prefrontal cortex structure and function. Nat. Rev. Neurosci. 10, 410–422 (2009).

  2. 2.

    Miller, E. K. & Cohen, J. D. An integrative theory of prefrontal cortex function. Annu. Rev. Neurosci. 24, 167–202 (2001).

  3. 3.

    Cohen, J. D., Braver, T. S. & Brown, J. W. Computational perspectives on dopamine function in prefrontal cortex. Curr. Opin. Neurobiol. 12, 223–229 (2002).

  4. 4.

    Kroener, S., Chandler, L. J., Phillips, P. E. M. & Seamans, J. K. Dopamine modulates persistent synaptic activity and enhances the signal-to-noise ratio in the prefrontal cortex. PLoS ONE 4, e6507 (2009).

  5. 5.

    Rolls, E. T., Loh, M., Deco, G. & Winterer, G. Computational models of schizophrenia and dopamine modulation in the prefrontal cortex. Nat. Rev. Neurosci. 9, 696–709 (2008).

  6. 6.

    Winterer, G. & Weinberger, D. R. Genes, dopamine and cortical signal-to-noise ratio in schizophrenia. Trends Neurosci. 27, 683–690 (2004).

  7. 7.

    Popescu, A. T., Zhou, M. R. & Poo, M.-M. Phasic dopamine release in the medial prefrontal cortex enhances stimulus discrimination. Proc. Natl Acad. Sci. USA 113, E3169–E3176 (2016).

  8. 8.

    Noudoost, B. & Moore, T. Control of visual cortical signals by prefrontal dopamine. Nature 474, 372–375 (2011).

  9. 9.

    Williams, G. V. & Goldman-Rakic, P. S. Modulation of memory fields by dopamine D1 receptors in prefrontal cortex. Nature 376, 572–575 (1995).

  10. 10.

    Burgos-Robles, A. et al. Amygdala inputs to prefrontal cortex guide behavior amid conflicting cues of reward and punishment. Nat. Neurosci. 20, 824–835 (2017).

  11. 11.

    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).

  12. 12.

    Kim, C. K. et al. Simultaneous fast measurement of circuit dynamics at multiple sites across the mammalian brain. Nat. Methods 13, 325–328 (2016).

  13. 13.

    Lammel, S., Ion, D. I., Roeper, J. & Malenka, R. C. Projection-specific modulation of dopamine neuron synapses by aversive and rewarding stimuli. Neuron 70, 855–862 (2011).

  14. 14.

    Abercrombie, E. D., Keefe, K. A., DiFrischia, D. S. & Zigmond, M. J. Differential effect of stress on in vivo dopamine release in striatum, nucleus accumbens, and medial frontal cortex. J. Neurochem. 52, 1655–1658 (1989).

  15. 15.

    Thierry, A. M., Tassin, J. P., Blanc, G. & Glowinski, J. Selective activation of mesocortical DA system by stress. Nature 263, 242–244 (1976).

  16. 16.

    Mantz, J., Thierry, A. M. & Glowinski, J. Effect of noxious tail pinch on the discharge rate of mesocortical and mesolimbic dopamine neurons: selective activation of the mesocortical system. Brain Res. 476, 377–381 (1989).

  17. 17.

    Finlay, J. M., Zigmond, M. J. & Abercrombie, E. D. Increased dopamine and norepinephrine release in medial prefrontal cortex induced by acute and chronic stress: effects of diazepam. Neuroscience 64, 619–628 (1995).

  18. 18.

    Heien, M. L. A. V., Phillips, P. E. M., Stuber, G. D., Seipel, A. T. & Wightman, R. M. Overoxidation of carbon-fiber microelectrodes enhances dopamine adsorption and increases sensitivity. Analyst 128, 1413–1419 (2003).

  19. 19.

    Bandler, R. & Carrive, P. Integrated defence reaction elicited by excitatory amino acid microinjection in the midbrain periaqueductal grey region of the unrestrained cat. Brain Res. 439, 95–106 (1988).

  20. 20.

    Deng, H., Xiao, X. & Wang, Z. Periaqueductal gray neuronal activities underlie different aspects of defensive behaviors. J. Neurosci. 36, 7580–7588 (2016).

  21. 21.

    Tovote, P. et al. Midbrain circuits for defensive behaviour. Nature 534, 206–212 (2016).

  22. 22.

    Franklin, T. B. et al. Prefrontal cortical control of a brainstem social behavior circuit. Nat. Neurosci. 20, 260–270 (2017).

  23. 23.

    Murugan, M. et al. Combined social and spatial coding in a descending projection from the prefrontal cortex. Cell 171, 1663–1677 (2017).

  24. 24.

    Otis, J. M. et al. Prefrontal cortex output circuits guide reward seeking through divergent cue encoding. Nature 543, 103–107 (2017).

  25. 25.

    Britt, J. P. et al. Synaptic and behavioral profile of multiple glutamatergic inputs to the nucleus accumbens. Neuron 76, 790–803 (2012).

  26. 26.

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

  27. 27.

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

  28. 28.

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

  29. 29.

    Klapoetke, N. C. et al. Independent optical excitation of distinct neural populations. Nat. Methods 11, 338–346 (2014).

  30. 30.

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

  31. 31.

    Nagel, G. et al. Channelrhodopsin-2, a directly light-gated cation-selective membrane channel. Proc. Natl Acad. Sci. USA 100, 13940–13945 (2003).

  32. 32.

    Zhang, F., Wang, L.-P., Boyden, E. S. & Deisseroth, K. Channelrhodopsin-2 and optical control of excitable cells. Nat. Methods 3, 785–792 (2006).

  33. 33.

    Gradinaru, V., Thompson, K. R. & Deisseroth, K. eNpHR: a Natronomonas halorhodopsin enhanced for optogenetic applications. Brain Cell Biol. 36, 129–139 (2008).

  34. 34.

    Schobert, B. & Lanyi, J. K. Halorhodopsin is a light-driven chloride pump. J. Biol. Chem. 257, 10306–10313 (1982).

  35. 35.

    Akerboom, J. et al. Genetically encoded calcium indicators for multi-color neural activity imaging and combination with optogenetics. Front. Mol. Neurosci. 6, 2 (2013).

  36. 36.

    Kremer, E. J., Boutin, S., Chillon, M. & Danos, O. Canine adenovirus vectors: an alternative for adenovirus-mediated gene transfer. J. Virol. 74, 505–512 (2000).

  37. 37.

    Schindelin, J., Arganda-Carreras, I., & Frise, E. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).

  38. 38.

    Witten, I. B. et al. Recombinase-driver rat lines: tools, techniques, and optogenetic application to dopamine-mediated reinforcement. Neuron 72, 721–733 (2011).

  39. 39.

    Tsai, H.-C. et al. Phasic firing in dopaminergic neurons is sufficient for behavioral conditioning. Science 324, 1080–1084 (2009).

  40. 40.

    Nieh, E. H. et al. Inhibitory input from the lateral hypothalamus to the ventral tegmental area disinhibits dopamine neurons and promotes behavioral activation. Neuron 90, 1286–1298 (2016).

  41. 41.

    Sparta, D. R. et al. Construction of implantable optical fibers for long-term optogenetic manipulation of neural circuits. Nat. Protoc. 7, 12–23 (2011).

  42. 42.

    Robinson, D. L., Venton, B. J., Heien, M. L. A. V. & Wightman, R. M. Detecting subsecond dopamine release with fast-scan cyclic voltammetry in vivo. Clin. Chem. 49, 1763–1773 (2003).

  43. 43.

    Keithley, R. B. & Wightman, R. M. Assessing principal component regression prediction of neurochemicals detected with fast-scan cyclic voltammetry. ACS Chem. Neurosci. 2, 514–525 (2011).

  44. 44.

    Keithley, R. B., Heien, M. L. & Wightman, R. M. Multivariate concentration determination using principal component regression with residual analysis. Trends Anal. Chem. 28, 1127–1136 (2009).

  45. 45.

    Conte, W. L., Kamishina, H. & Reep, R. L. Multiple neuroanatomical tract-tracing using fluorescent Alexa Fluor conjugates of cholera toxin subunit B in rats. Nat. Protoc. 4, 1157–1166 (2009).

  46. 46.

    Zhuang, X., Masson, J., Gingrich, J. A., Rayport, S. & Hen, R. Targeted gene expression in dopamine and serotonin neurons of the mouse brain. J. Neurosci. Methods 143, 27–32 (2005).

  47. 47.

    Resendez, S. L. et al. Visualization of cortical, subcortical and deep brain neural circuit dynamics during naturalistic mammalian behavior with head-mounted microscopes and chronically implanted lenses. Nat. Protoc. 11, 566–597 (2016).

  48. 48.

    Jennings, J. H. et al. Visualizing hypothalamic network dynamics for appetitive and consummatory behaviors. Cell 160, 516–527 (2015).

  49. 49.

    Ziv, Y. et al. Long-term dynamics of CA1 hippocampal place codes. Nat. Neurosci. 16, 264–266 (2013).

  50. 50.

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

  51. 51.

    Novák, P. & Zahradník, I. Q-method for high-resolution, whole-cell patch-clamp impedance measurements using square wave stimulation. Ann. Biomed. Eng. 34, 1201–1212 (2006).

  52. 52.

    Burgos-Robles, A., Vidal-Gonzalez, I., Santini, E. & Quirk, G. J. Consolidation of fear extinction requires NMDA receptor-dependent bursting in the ventromedial prefrontal cortex. Neuron 53, 871–880 (2007).

  53. 53.

    Paxinos, G. & Watson, C. The Rat Brain in Stereotaxic Coordinates: Hard Cover Edition (Academic, Cambridge, MA, 2006).

  54. 54.

    Paxinos, G. & Franklin, K. B. The Mouse Brain in Stereotaxic Coordinates (Gulf, Houston, 2004).

  55. 55.

    Kupferschmidt, D. A., Juczewski, K., Johnson, K. A. & Lovinger, D. M. Parallel, but dissociable, processing in discrete corticostriatal inputs encodes skill learning. Neuron 96, 476–489 (2017).

Download references

Acknowledgements

We thank I. Witten, C. Cameron, N. Parker, M. Murugan, P. Zhou and L. Paninski for advice and code for CNMF-E analysis; M. Schnitzer and D. Cai for advice regarding endoscopic imaging; Y.-N. Leow, A. Shea and N. Golan for histological assistance; N. Imamura and C. Leppla for technical training. We recognize the generosity of the Genetically-Encoded Neuronal Indicator and Effector (GENIE) program, the Janelia Farm Research Campus, V. Jayaraman, R. A. Kerr, D. S. Kim, L. L. Looger and K. Svoboda for providing GCaMP6m. We acknowledge Inscopix for a scientific collaboration and providing early access to nVoke and L. Cardy and A. Stamatakis of Inscopix for technical assistance. We thank E. J. Kremer for providing CAV2-Cre vector; UNC vector core for ChR2, NpHR and ChrimsonR vectors; University of Pennsylvania vector core for GCaMP6m packaging; R. Neve (formerly at the Gene Transfer Core Facility at MIT, now at Massachusetts General Hospital) for packaging the AAV-DIO-synaptophysin-mCherry construct; J. Crittenden for D1-TdTomato/D2-GFP mice and T. Okuyama for Drd1a-Cre and Drd2-Cre mice. K.M.T. is a New York Stem Cell Foundation–Robertson Investigator and a McKnight Scholar, and this work was supported by funding from the JPB Foundation, PIIF, PNDRF, JFDP, Klingenstein Foundation, NARSAD Young Investigator Award, New York Stem Cell Foundation, NIH R01-MH102441-01 (NIMH), NIH Director’s New Innovator Award DP2-DK-102256-01 (NIDDK), and Pioneer Award DP1-AT009925 (NCCIH). C.M.V.W. and E.H.N. were supported by the NSF Graduate Research Fellowship and Integrative Neuronal Systems Training Fellowship (T32 GM007484). C.A.S. is supported by NIH grants F32 MH111216 (NIMH) and K99 DA045103 (NIDA). G.A.M. was supported by the Charles A. King Trust Postdoctoral Research Fellowship Program, Bank of America, N.A., Co-Trustees. R.W. and N.P.-C. acknowledge funding from the Simons Center Postdoctoral Fellowship. R.W. also recognizes funding from the Netherlands Organization for Scientific Research (NWO) RUBICON. C.A.S., A.B., A.B.-R. and R.W. recognize support from the NARSAD Young Investigator Award.

Reviewer information

Nature thanks P. Phillips and the anonymous reviewer(s) for their contribution to the peer review of this work.

Author information

Author notes

    • Romy Wichmann
    •  & Kay M. Tye

    Present address: Salk Institute for Biological Sciences, La Jolla, CA, USA

  1. These authors contributed equally: Caitlin M. Vander Weele, Cody A. Siciliano, Gillian A. Matthews

Affiliations

  1. The Picower Institute for Learning and Memory, Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA

    • Caitlin M. Vander Weele
    • , Cody A. Siciliano
    • , Gillian A. Matthews
    • , Praneeth Namburi
    • , Ehsan M. Izadmehr
    • , Isabella C. Espinel
    • , Edward H. Nieh
    • , Evelien H. S. Schut
    • , Nancy Padilla-Coreano
    • , Anthony Burgos-Robles
    • , Chia-Jung Chang
    • , Eyal Y. Kimchi
    • , Anna Beyeler
    • , Romy Wichmann
    • , Craig P. Wildes
    •  & Kay M. Tye
  2. Department of Cognitive Neuroscience, Radboudumc Nijmegen, Nijmegen, The Netherlands

    • Evelien H. S. Schut

Authors

  1. Search for Caitlin M. Vander Weele in:

  2. Search for Cody A. Siciliano in:

  3. Search for Gillian A. Matthews in:

  4. Search for Praneeth Namburi in:

  5. Search for Ehsan M. Izadmehr in:

  6. Search for Isabella C. Espinel in:

  7. Search for Edward H. Nieh in:

  8. Search for Evelien H. S. Schut in:

  9. Search for Nancy Padilla-Coreano in:

  10. Search for Anthony Burgos-Robles in:

  11. Search for Chia-Jung Chang in:

  12. Search for Eyal Y. Kimchi in:

  13. Search for Anna Beyeler in:

  14. Search for Romy Wichmann in:

  15. Search for Craig P. Wildes in:

  16. Search for Kay M. Tye in:

Contributions

C.M.V.W. and K.M.T. conceived the project. C.M.V.W., C.A.S., G.A.M., E.M.I., I.C.E., E.H.N., E.H.S.S. and N.P.-C. collected data. C.M.V.W., E.H.N., G.A.M., C.A.S., I.C.E., C.-J.C., P.N. and K.M.T. analysed data. E.H.N., P.N., C.-J.C. and E.Y.K. provided MATLAB scripts and advice for data analysis. R.W., A.B., C.P.W. and A.B.-R. provided technical training. C.M.V.W., C.A.S., G.A.M., E.H.N. and K.M.T. contributed to experimental design. C.M.V.W. and K.M.T. wrote the paper. All authors contributed to the editing of the manuscript.

Competing interests

The authors declare no competing interests.

Corresponding author

Correspondence to Kay M. Tye.

Extended data figures and tables

  1. Extended Data Fig. 1 Investigation of catecholamine terminal density and dopamine release dynamics in the mPFC.

    a, b, Injection of viral constructs (a) enabling Cre-dependent expression into the LC and VTA of TH::Cre mice resulted in (b) fluorescent labelling of TH positive (TH+) noradrenergic (NE) neurons in the LC and dopamine neurons in the VTA. c, Examination of VTADA and LCNA fluorescent terminal labelling in the mPFC revealed different patterns of innervation by VTADA and LCNA neurons across cortical layers in the prelimbic subregion of the mPFC (n = 3 mice). d, VTADA terminals were densest in the deep (5 and 6) layers of the mPFC, whereas LCNA terminals were denser in superficial (1 and 2/3) layers. e, Schematic of strategy for differentiating dopamine and noradrenaline neurotransmission in the mPFC using FSCV. VTADA neurons were selectively transduced with ChR2 in TH::Cre rats. After incubation, rats were prepared for anaesthetized FSCV recordings, in which an optical fibre was implanted over the VTA and a guide cannula was positioned over the LC for TTX-mediated pharmacological inhibition. f, A glass-encased carbon-fibre recording electrode was lowered into the mPFC for FSCV neurochemical measurements. Schematic representation of all recording electrode locations for ChR2 FSCV experiments. g, Representative image of guide cannula track positioned over LCNA cell bodies. Yellow, TH. hj, When VTADA and LCNA neurons were intact, tail pinch (10 s in duration) rapidly increased extracellular catecholamine concentration (CAT), as shown in a representative false colour plot (h), average CAT trace (i), and concentration quantification (j). n = 5 rats; two-tailed paired t-test, t4 = 3.402, *P = 0.027. Colour plot insets, representative cyclic voltammograms. k, TTX–Fast Green injection locations. ln, After LC inactivation via intra-LC infusion of TTX, tail-pinch-evoked responses were maintained (two-tailed paired t-test, t4 = 5.249, **P = 0.006. o, Pharmacological inactivation of the LC with TTX did not significantly alter tail-pinch-evoked catecholamine release in the mPFC. Two-way repeated measures ANOVA, F5,40 = 0.061, P = 0.997. p, Representative image of FSCV electrode track in the mPFC. q, Representative confocal image of ChR2–mCherry expression (red) in VTADA cell bodies. Yellow, TH immunostaining. r, Schematic of strategy to verify dependence of pinch-evoked increases in CAT neurotransmission on VTADA neurons. s, Histologically verified FSCV recording electrode locations for NpHR experiments. t, Electrical stimulation (60 Hz, 60 pulses, 200 μA) of the dorsal VTA evoked distinct patterns of dopamine release in the NAc and mPFC (n = 5 rats). u, Optical inhibition (20 s constant 593 nm, 5 mW) of NpHR-expressing VTADA neurons attenuated tail-pinch-evoked CAT release in the mPFC. Two-way repeated measures ANOVA, F5,40 = 2.857, P = 0.027; Bonferroni post-hoc tests, **P < 0.01. v, Schematic of viral strategy to optically manipulate ChR2-expressing VTADA terminals in the mPFC and record from dPAG- and NAc-projectors retrogradely labelled with CTB with ex vivo electrophysiology. w, No evidence of co-release of fast-synaptic neurotransmitters (that is, glutamate and GABA (γ-aminobutyric acid)) from VTADA terminals onto either mPFC–dPAG (teal) or mPFC–NAc (pink) populations following optical stimulation in voltage-clamp (left) and current-clamp (right). Error bars and shading represent s.e.m. A.U., arbitrary fluorescence units. The rat brain in this figure was reproduced with permission from Paxinos and Watson, 200653.

  2. Extended Data Fig. 2 Activation of VTADA terminals in the mPFC does not support real-time or conditioned place preference.

    a, Schematic of strategy for manipulating dopamine release in the mPFC. VTADA neurons were selectively transduced with ChR2 in TH::Cre rats and guide cannulae were implanted over the mPFC for the insertion of an optical fibre for light delivery. b, Representative confocal image of ChR2–eYFP expression in VTADA–mPFC underneath a guide cannula (left) and expression in the VTA (right). c, Histological verification of guide cannulae placements in the mPFC for ChR2 subjects (left) and eYFP controls (right). d, Schematic of experimental design for RTPP/A. When rats entered the ON zone, laser light stimulation was activated for the duration of the time spent in the ON zone (20 Hz, 60 pulses, every 30 s, 20 mW of 473 nm). When rats entered the OFF zone, light stimulation was terminated for the duration of time spent in the OFF zone. e, Optogenetic stimulation of VTADA terminals did not evoke real-time place avoidance or preference in VTADA–mPFC::ChR2 rats (n = 5), compared to VTADA–mPFC::eYFP controls (n = 5), measured by difference score (minutes spent in the ON zone −  minutes spent in the OFF zone). Two-tailed unpaired t-test, t8 = 0.9337, P = 0.3778. f, Schematic of experimental design for CPP/A. Day 1 consisted of a habituation period in which time spent on each compartment of the arena was recorded. On days 2 and 3, a divider was placed in the middle of the chamber to separate the two compartments and rats received either no stimulation (OFF) or stimulation (ON) (20 Hz, 60 pulses, every 30 s, 20 mW), counterbalanced across days. On day 4, the divider was removed and time spent in each compartment was recorded in the absence of stimulation (that is, test day). g, Optogenetic stimulation of VTADA terminals did not support conditioned place aversion or preference in VTADA–mPFC::ChR2 animals (n = 6), compared to VTADA–mPFC::eYFP controls (n = 5), measured by difference score. Two-tailed unpaired t-test, t9 = 0.3192, P = 0.7569. h, Schematic of task used to examine dopamine modulation of reward and fear-motivated behaviours during competition. i, During sucrose training, a conditioned stimulus (CS) (light or tone, counterbalanced) predicted sucrose delivery (CSsuc). Sucrose was removed from the delivery port by vacuum if not collected. j, VTADA–mPFC::ChR2 rats (n = 7) and VTADA–mPFC::eYFP controls (n = 6) acquired sucrose conditioning similarly. Two-way repeated measures ANOVA, F2,22 = 0.7, P = 0.5090. k, During discrimination, the alternative CS (light or tone, counterbalanced) was introduced and predicted foot shock (CSshk). l, Average traces showing time spent in the sucrose port before, during, and after each CS presentation (grouped, n = 13 rats). m, Time spent in the sucrose port did not differ between VTADA–mPFC::ChR2 rats (n = 7) and VTADA–mPFC::eYFP controls (n = 6) during CSsuc or CSshk presentation. Repeated measures two-way ANOVA, F1,11 = 0.54, P = 0.4789. n, Average traces showing time spent freezing before, during, and after each CS presentation (grouped, n = 13 rats). o, Time spent freezing did not differ between VTADA–mPFC::ChR2 rats and VTADA–mPFC::eYFP controls during CSsuc or CSshk presentation. Repeated measures two-way ANOVA, F1,11 = 0.01, P = 0.9281. p, During competition sessions, the average time spent in the reward port for CSsuc trials during ON sessions and CSsuc trials during OFF sessions did not differ between ChR2 rats (n = 7, closed) and eYFP controls (n = 6, open). Repeated measures two-way ANOVA, F1,11 = 0.82, P = 0.3845. Note that during ON sessions, stimulation was only delivered during the CScomp trials. q, Average time spent freezing for CSsuc trials during ON sessions and CSsuc trials during OFF sessions did not differ between ChR2 rats (closed) and eYFP controls (open). Repeated measures two-way ANOVA, F1,11 =  1.35, P = 0.2705. r, During competition sessions, the average time spent in the reward port for CSshk trials during ON sessions and CSshk trials during OFF sessions was not different between ChR2 (closed) and eYFP controls (open). Repeated measures two-way ANOVA, F1,11 = 0.94, P = 0.354. s, During competition sessions, the average time spent freezing for CSshk trials during ON sessions and CSshk trials during OFF sessions was not different between ChR2 rats (closed) and eYFP controls (open). Repeated measures two-way ANOVA, F1,11 = 0.16, P = 0.6998. Error bars and shading represent s.e.m.

  3. Extended Data Fig. 3 Putative connection strength of mPFC projections to downstream targets, layer localization of projectors, and collateralization.

    a, Schematic of strategy in which anterogradely travelling virus was injected into the prelimbic and infralimbic subregions of the mPFC and fluorescence was quantified in several downstream brain regions. b, Orange boxes represent approximate locations of fluorescence quantification, as a proxy for connection strength. n = 3 rats. c, d, Representative images (c) and quantification of fluorescence (d) in the mPFC and downstream targets in the rat. e, Microinjections of CTB conjugated to fluorescent proteins (Alexa Fluor 488, Alexa Fluor 555 or Alexa Fluor 647, counterbalanced) were placed in the dPAG and NAc to retrogradely label the cell bodies of projection neurons in the rat mPFC (n = 3 rats). f, Representative confocal images of CTB injections in the NAc and dPAG of the rat. g, Representative confocal image of retrogradely labelled neurons in the rat mPFC. h, As a population, only 11 out of 1,679 CTB+ neurons in the mPFC were dual-labelled. i, Fluorescence quantification of retrogradely labelled mPFC–dPAG and mPFC–NAc neurons revealed differences in cell-body location across cortical layers in the rat mPFC. j, In the rat, dPAG projectors predominantly originate from deep layer 5, whereas NAc projectors are located in both superficial layers 2/3 and deep layer 5. k, Microinjections of CTB conjugated to fluorescent proteins were placed in the dPAG and NAc to retrogradely label the cell bodies of projection neurons in the mouse mPFC. l, Representative confocal images of CTB injections in the NAc and dPAG of the mouse (n = 3 mice). m, Representative confocal image of retrogradely labelled neurons in the mouse mPFC. n, As a population, only 17 out of 458 CTB+ neurons in the mPFC were dual-labelled. o, Fluorescence quantification of retrogradely labelled mPFC–dPAG and mPFC–NAc neurons revealed differences in cell-body location across cortical layers in the mouse mPFC. p, In the mouse, dPAG projectors predominantly originate from deep layer 5, whereas NAc projectors are located in both superficial layers 2/3 and deep layer 5. q, Schematic of viral strategy to explore downstream fluorescence from mPFC–NAc::eYFP (n = 3 rats) and mPFC–dPAG::eYFP (n = 3 rats) projectors. r, Quantification of fluorescence in the mPFC and downstream brain regions originating from mPFC–dPAG::eYFP and mPFC–NAc::eYFP neurons. s, t, Representative confocal images from a mPFC–dPAG::eYFP subject (s) and a mPFC–NAc::eYFP subject (t). u, Schematic of viral strategy to explore downstream terminals from mPFC–NAc::synaptophysin (n = 3 mice) and mPFC–dPAG::synaptophysin (n = 3 mice) projectors. v, Quantification of fluorescence in the mPFC and downstream brain regions originating from mPFC–dPAG::synaptophysin and mPFC–NAc::synaptophysin neurons. w, x, Representative confocal images from a mPFC–dPAG::synaptophysin subject (w) and a mPFC–NAc::synaptophysin subject (x). BLA, basolateral amygdala; agIN, agranular insula; cl, claustrum; dStr, dorsal striatum (medial); LH, lateral hypothalamus; LS, lateral septum; PHA, posterior hypothalamic area; PVT, paraventricular nucleus of the thalamus; vPAG, ventral periaqueductal grey. Rat and mouse brains in this figure have been reproduced with permission from Paxinos and Watson, 200653, and Paxinos and Franklin, 200454, respectively. Error bars and dashed lines represent s.e.m. Scale bars, 50 μm.

  4. Extended Data Fig. 4 mPFC–NAc photostimulation does not support place preference or aversion.

    a, Schematic of viral transduction strategy to achieve optogenetic control of rat mPFC neurons projecting to the NAc. b, Representative image of NAc-projecting mPFC neurons expressing ChR2 (left) and ChR2+ terminals in the NAc (right). c, Histological verification of bilateral optical-fibre implant locations above the mPFC and virus injection locations in the NAc. d, Representative locomotor heat maps of mPFC–NAc::ChR2 (top) and mPFC–NAc::eYFP (bottom) subjects in the RTPP/A assay. e, Optogenetic stimulation of mPFC–NAc neurons did not evoke real-time place avoidance or preference in mPFC–NAc::ChR2 animals (n = 7 rats), compared to mPFC–NAc::eYFP controls (n = 6 rats), measured by difference score (minutes spent in the ON zone – minutes spent in OFF zone). Two-tailed unpaired t-test, t11 = 0.5549, P = 0.5901. f, Representative locomotor heat map of mPFC–NAc::ChR2 subject in CPP/A assay. g, Optogenetic stimulation of mPFC–NAc neurons did not evoke real-time place aversion or preference in mPFC–NAc::ChR2 animals (n = 6 rats), compared to mPFC–NAc::eYFP controls (n = 6 rats). Two-tailed unpaired t-test, t10 = 0.2143, P = 0.8346. h, Representative locomotor heat maps of a mPFC–NAc::ChR2 subject during 3 min OFF–ON–OFF epochs during the open-field test. i, j, Optical activation of mPFC–NAc::ChR2 (n = 6 rats) did not change time spent in the centre region compared to eYFP controls (n = 5 rats) (i; two-way repeated measures ANOVA, F2,18 = 0.74, P = 0.4913), or general locomotor activity (j; two-way repeated measures ANOVA, F2,18 = 0.61, P = 0.5532). Data are mean ± s.e.m. The rat brains in this figure were reproduced with permission from Paxinos and Watson, 200653.

  5. Extended Data Fig. 5 Activation of mPFC terminals in the dPAG increases marble burying and activation of mPFC–dPAG cell bodies does not affect anxiety-like behaviour.

    a, Schematic of viral strategy to achieve optogenetic control of ChR2-expressing mPFC terminals in the dPAG. b, c, Representative image of ChR2+ neurons in the mPFC (b) and ChR2+ terminals in the dPAG (c) (optic fibre lesions indicated by dashed lines). d, Histological verification of bilateral virus injection locations in the mPFC and bilateral optic fibre implant locations above the dPAG. e, Representative locomotor heat maps of mPFC–dPAG::ChR2 (top) and mPFC–dPAG::eYFP control subject (bottom) in the RTPP/A assay. f, Percent of time spent in the ON and OFF zones of the arena in mPFC–dPAG::ChR2 and mPFC–dPAG::eYFP subjects. g, Optogenetic stimulation of mPFC terminals in the dPAG resulted in a trend towards avoidance in the RTPA assay in mPFC–dPAG::ChR2 animals (n = 5 rats), compared with mPFC–dPAG::eYFP controls (n = 8 rats). Two-tailed unpaired t-test, t11 = 1.830, #P = 0.0944). h, Representative arena of mPFC–dPAG::ChR2 animal after marble-burying assay when optical stimulation was OFF (top) and ON (bottom). i, Number of marbles buried in mPFC–dPAG::ChR2 (n = 5 rats) and mPFC–dPAG::eYFP (n = 6 rats) during OFF and ON sessions. j, k, Optical stimulation of mPFC–dPAG neurons resulted in more marbles buried by mPFC–dPAG::ChR2 animals, compared with mPFC–dPAG::eYFP controls (j; two-tailed unpaired t-test, t9 = 2.839, *P = 0.0194) and more time digging (k; one-tailed unpaired t-test, t9 = 2.775, *P = 0.0108). l, Functional ChR2 expression in mPFC–dPAG neurons was verified by targeted ex vivo whole-cell patch-clamp electrophysiology. Recording from a ChR2-expressing mPFC–dPAG neuron in voltage-clamp mode showing sustained inward current elicited by a 1-s pulse of 470-nm light. m, n, In current-clamp mode, action potentials were elicited by 1-Hz (m) and 20-Hz light trains (n). 470 nm, 5-ms pulse duration. o, Representative confocal images of mPFC–dPAG::ChR2 (top) and mPFC–dPAG::eYFP expressing neurons showing immediate early gene (c-Fos) expression following 5 min blue (473 nm) light exposure (20 Hz, 5-ms pulse duration, 15 mW). p, Laser light stimulation (473 nm) enhanced the number of c-Fos-positive ChR2-expressing mPFC–dPAG neurons compared with control mPFC–dPAG::eYFP neurons. mPFC–dPAG::ChR2, n = 4 rats; mPFC–dPAG::eYFP, n = 3 rats; two-tailed unpaired t-test, t5 = 3.707, *P = 0.014. q, Histological verification of bilateral optical-fibre implant locations above the mPFC and virus injection locations in the dPAG for mPFC–dPAG::ChR2/eYFP-expressing rats. r, Representative locomotor heat maps of a mPFC–dPAG::ChR2 subject during 3 min OFF–ON–OFF epochs in the open-field test. s, t, Optical activation of mPFC–dPAG::ChR2 (n = 15 rats) did not change time spent in the centre region compared to eYFP controls (s; n = 18 rats, two-way repeated measures ANOVA, group × epoch, F2,62 = 0.37, P = 0.69), or general locomotor activity (t; distance travelled, two-way repeated measures ANOVA, group × epoch interaction, F2,62 = 0.9412, P = 0.3957). u, Quantification of behaviours (percentage of time engaging) during marble-burying assay. v, Representative confocal image of viral spread in the PAG, visualized by co-injection of AAV5-hSyn-mCherry (hSyn, synapsin, red) with CAV2-Cre in a subset of mPFC–dPAG::ChR2/eYFP expressing rats. w, Illustration of reconstructed injection locations and spread in co-injected subjects. n = 14 total, 7 ChR2, 7 eYFP. Error bars indicate s.e.m. The rat brains in this figure were reproduced with permission from Paxinos and Watson, 200653.

  6. Extended Data Fig. 6 Analysis and additional data from epifluorescent calcium imaging experiments during sucrose and shock delivery.

    a, Schematic of strategy for monitoring neuronal activity in mPFC–dPAG and mPFC–NAc neurons using in vivo calcium imaging. b, Representative confocal images of mPFC–NAc::GCaMP6m (left) and mPFC–dPAG::GCaMP6m neurons (right) underneath GRIN lenses (dashed lines). c, Dynamic calcium fluctuations were monitored during a 15-min recording session in which mice were allowed to self-administer sucrose via a sucrose lickometer or had random, unsignalled foot shocks delivered. d, As a population, mPFC–dPAG::GCaMP6m (n = 6 mice) were activated to foot shock (green, two-tailed paired t-test, t5 = 2.616, *P = 0.0473) or inhibited by the initiation of a sucrose bout (purple, two-tailed paired t-test, t5 = 6.982, ***P = 0.0009) as measured by the bulk fluorescence across the entire FOV (−3 to 0 s, pre-shock/sucrose; 0–3 s, shock/sucrose). e, As a population, mPFC–NAc::GCaMP6m (n = 5 mice) were not responsive to foot shock (green, two-tailed paired t-test, t4 = 0.1520, P = 0.8866) or the initiation of a sucrose bout (purple, two-tailed paired t-test, t4 = 0.2678, P = 0.8021) (−3 to 0 s, pre-shock/sucrose; 0–3 s, shock/sucrose). f, mPFC–dPAG::GCaMP6m and mPFC–NAc::GCaMP6m mice did not differ in the number of lick bouts initiated during the sucrose session. Two-tailed unpaired t-test, t9 = 0.1666, P = 0.8714. g, Peak-to-noise heat map generated from a representative FOV with seed pixels overlaid (black X). h, mPFC–dPAG::GCaMP6m neurons (n = 118 ROIs) had more frequent calcium transients than mPFC–NAc::GCaMP6m neurons (n = 169 ROIs) during the shock session. Number of events difference score (shock − sucrose): dPAG Mdn, 51.5; NAc Mdn, −6. Two-tailed Mann–Whitney test, U = 5,840, ***P < 0.0001. i, mPFC–dPAG::GCaMP6m neurons had higher amplitude transients than mPFC–NAc::GCaMP6m neurons during the shock session. Amplitude of events difference score (shock − sucrose): dPAG Mdn, 0.9031; NAc Mdn, −0.3549. Mann–Whitney test, U = 6,672, ***P < 0.0001. j, Dendrogram of agglomerative hierarchical clustering. Different colours represent clusters based on average responses per ROI to footshock and sucrose. k, Histologically verified locations of GRIN lens implants. lac, In addition to using CNMF-E, imaging data were analysed using two other approaches: 1) a modified constrained CNMF-E algorithm considering calcium fluctuations can have negative transients, associated with a decrease in firing24,55 (for the approach, we did not constrain temporal components to >0) and 2) a ROI-based method (that is, ‘non-ROI’, rac). l, m, Calcium signals were extracted from individual ROIs and the average calcium traces per ROI were aligned to shock and sucrose bout onset for mPFC–NAc::GCaMP6m (l) and mPFC–dPAG::GCaMP6m recordings (m). n, The distribution of shock- and sucrose-excited cells for mPFC–dPAG::GCaMP6m neurons was different from mPFC–NAc::GCaMP6m neurons. χ2 = 10.95, **P = 0.0042. o, Representative calcium traces from a mPFC–dPAG::GCaMP6m neuron during shock (top) and sucrose (bottom) recording sessions. Individual calcium transients (yellow dots) were identified and quantified. p, mPFC–dPAG::GCaMP6m neurons (n = 118 ROIs) had more frequent calcium transients than mPFC–NAc::GCaMP6m neurons (n = 169 ROIs) during the shock session. Number of events difference score (shock − sucrose): dPAG Mdn, 43; NAc Mdn, −3. Two-tailed Mann–Whitney test, U = 4,373, ***P < 0.0001. q, mPFC–dPAG::GCaMP6m neurons had higher amplitude calcium transients compared to mPFC–NAc::GCaMP6m neurons during the shock session. Amplitude of events difference score (shock − sucrose): dPAG Mdn, 1.329; NAc Mdn, −0.2459. Two-tailed Mann–Whitney test, U = 7,164, ***P < 0.0001. r, Mean t-projection image of the entire FOV through the relay lens after image pre-processing. Recordings were converted to changes in fluorescence compared to background fluorescence (F − F0)/F0 using the mean t-projection image as reference (F0). s, Calcium signals arising from ROIs were identified using independent and principal component analyses (PCA/ICA). t, Identified PCA/ICA filters were thresholded at their half-maximum values to define possible ROIs and were screened for neuronal morphology. ROIs were only accepted if the threshold filters included only on contiguous region with an eccentricity of <0.85 and an area between 30–350 pixels. In this example, 61 ROIs (of the original 150 independent components (ICs)) met these criteria. u, Accepted ROI filters were then merged if their areas overlapped by more than 60%. In this example, 24 ROIs were merged for a remaining total of 37 valid ROIs. v, To acquire the non-ROI thresholded image for background subtraction, max z projections of individual recordings were created and thresholded to separate ROIs and their processes from the rest of the FOV. Average signal from the remaining pixels was used as a proxy for the whole-field changes in fluorescence, and regressed from the signal extracted from each ROI. w, x, Calcium transients (yellow dots) within individual mPFC–dPAG::GCaMP6m neurons (w) and mPFC–NAc::GCaMP6m neurons (x) were quantified (representative traces). y, mPFC–dPAG::GCaMP6m neurons (n = 113 of 118 ROIs) had more frequent calcium transients than mPFC–NAc::GCaMP6m neurons (n = 157 ROIs) during the shock session. Difference score (shock − sucrose): dPAG Mdn, 30; NAc Mdn, 6. Two-tailed Mann–Whitney test, U = 6,392, ***P < 0.0001. z, mPFC–dPAG::GCaMP6m neurons had calcium transients of larger amplitude than mPFC–NAc::GCaMP6m neurons during the shock session. Difference score (shock − sucrose): dPAG Mdn, 0.5158; NAc Mdn, −0.0615. Two-tailed Mann–Whitney test, U = 7,065, **P = 0.0044. aa, ab, Average calcium traces per cell for mPFC–dPAG::GCaMP6m neurons (aa) and mPFC–NAc::GCaMP6m neurons (ab) were aligned to shock (left) and sucrose bout (right). ac, The distribution of shock- and sucrose-excited cells for mPFC–dPAG::GCaMP6m (n = 118 ROIs) neurons was different from that for mPFC–NAc::GCaMP6m neurons (n = 157 ROIs). χ2 = 32.33, ***P < 0.0001. Error bars and ‘+’ indicate s.e.m. Scale bar, 100 μm. The mouse brains in this figure were reproduced with permission from Paxinos and Franklin, 200454.

  7. Extended Data Fig. 7 VTADA effects on mPFC projectors across time and their properties.

    a, Representative confocal image of mPFC–dPAG labelled neurons. b, Representative examples of reconstructed mPFC–NAc and mPFC–PAG neurons. c, Sholl analysis of mPFC–NAc (n = 4 cells) and mPFC–dPAG (n = 4 cells) subpopulations. d, Schematic of viral strategy to optically manipulate ChR2-expressing VTADA terminals in the mPFC and record from dPAG- and NAc-projectors retrogradely labelled with CTB using ex vivo electrophysiology. e, Representative images of a recorded mPFC–dPAG neuron (neurobiotin+ and CTB+) surrounded by ChR2–eYFP+ VTADA terminals. f, Representative traces of a mPFC–dPAG and mPFC–NAc neuron during a current step without (top) and with (bottom) optogenetic activation of VTADA terminals in the presence of type D2-type dopamine receptor blockade by bath-applied raclopride. g, The change in spike number with optical stimulation (ON–OFF) recorded from mPFC–dPAG (n = 5 cells) and mPFC–NAc neurons (n = 14 cells) in the presence of D2-receptor antagonism. h, The change in spike number with optical stimulation (ON–OFF) was different between mPFC–dPAG (n = 17 cells) and mPFC–NAc neurons (n = 24 cells) and was blocked by D2-receptor antagonism. One-way ANOVA, F3,56 = 5.343, P = 0.0026; Bonferroni multiple comparisons tests: dPAG vs NAc, *P = 0.0040; NAc vs NAc + raclopride, *P = 0.0034. i, Change in the number of spikes per step with optical stimulation (ON–OFF) for individual sweeps. mPFC–NAc neurons exhibited a more robust decrease in spike number during VTADA terminal stimulation during the first few sweeps, an effect that diminished in later sweeps. j, Representative traces showing firing elicited in mPFC–dPAG and mPFC–NAc neurons in response to current ramp with and without VTADA terminal stimulation (grey dashed line indicates time of first action potential without optical stimulation). Scale bars, 50 mV, 500 ms. k, Optical stimulation of VTADA terminals increased the current required to elicit an action potential (rheobase) in NAc projectors. The change in rheobase with optical stimulation (ON–OFF) was different between dPAG projectors (n = 5 cells) and NAc projectors (n = 9 cells). Two-tailed unpaired t-test, t12 = 2.669, P = 0.0205. l, m, Neither resting membrane potential (mPFC–dPAG, n = 16 cells; mPFC–NAc, n = 13 cells) (l) nor capacitance (m) differed between dPAG-projectors (n = 50 cells) and NAc-projectors (n = 27 cells). Resting membrane potential: two-tailed unpaired t-test, t27 = 0.6265, P = 0.5363; capacitance: two-tailed unpaired t-test, t75 = 0.8643, P = 0.3902. n, The current-voltage (IV) relationship of mPFC–dPAG (n = 16 cells) and mPFC–NAc (n = 13 cells) neurons obtained by applying a series of current steps in voltage-clamp mode. Two-way ANOVA, F12,324 = 10.16, P < 0.0001. o, The membrane resistance was significantly greater in NAc projectors (n = 27 cells) compared to dPAG projectors (n = 50 cells). Two-tailed unpaired t-test, t75 = 7.030, ***P < 0.0001. p, Representative traces showing action potential firing in mPFC–dPAG and mPFC–NAc neurons in response to a depolarizing current step. Scale bars, 50 mV, 500 ms. q, r, Instantaneous (I) (q) and steady-state (SS) (r) firing frequency in dPAG and NAc projectors in response to increasing current steps. s, Schematic of strategy for identifying dopamine type 1 receptor (D1) and dopamine type 2 receptor (D2) on mPFC-projector populations using transgenic mice (Drd1a-Cre (n = 3 mice) and Drd2-Cre (n = 3 mice)), retrograde labelling, and Cre-dependent eYFP recombination. t, u, Representative confocal images of NAc CTB injections sites (upper left), mPFC terminal fluorescence (lower left), and mPFC–NAc cell bodies (right) in a Drd1a-Cre::eYFP mouse (t) and Drd2-Cre::eYFP mouse (u). v, w, Representative confocal images of dPAG CTB injections sites (upper left), mPFC terminal fluorescence (lower left), and mPFC–dPAG cell bodies (right) in a Drd1a-Cre::eYFP mouse (v) and a Drd2-Cre::eYFP mouse (w). x, 5% of mPFC–dPAG CTB+ neurons were Drd1a+ (19/378), whereas 31.5% of mPFC–NAc CTB+ neurons were co-labelled as Drd1a+ (151/479) (D1 χ2 = 93.29, ***P < 0.0001). y, 27.6% of mPFC–dPAG CTB+ neurons were Drd2+ (74/342), whereas 86.3% of mPFC–NAc CTB+ neurons were co-labelled as Drd2+ (414/480) (D2 χ2 = 345.6, ***P < 0.0001). Error bars, shading, and ‘+’ represent s.e.m.

  8. Extended Data Fig. 8 Investigation of VTA projections to the dPAG for simultaneous epifluorescent imaging in mPFC–dPAG neurons and excitation of VTADA terminals.

    a, To verify that VTA neurons do not project to the dPAG (to allow for CAV2-Cre mediated GCaMP6m expression in dPAG neurons and simultaneous expression of the excitatory opsin Chrimson in VTADA neurons in DAT::Cre mice), VTA slices were immunostained for tyrosine hydroxylase (TH) in rats injected with the retrograde tracer CTB in the dPAG. b, Of 1,400 DAPI+ cells counted in the VTA, 792 (56%) were TH+, 16 (1.1%) were CTB+, and 0 were TH+ and CTB+. The lack of CTB+ cells suggests that VTA does not make a prominent projection to the dPAG. c, Schematic of strategy to simultaneously image fluorescent calcium activity in mPFC–dPAG::GCaMP6m neurons and activate VTADA–mPFC. d, Histological verification of GRIN lens locations in the mPFC in mPFC–dPAG::GCaMP6m × VTADA::Chrimson subjects and control mPFC–dPAG::GCaMP6m × VTADA::mCherry subjects. e, Representative confocal images of mPFC–dPAG::GCaMP6m and VTADA::Chrimson expression in the mPFC. f, Schematic of experimental design. During the ON epoch, a 590-nm LED stimulated Chrimson expressing VTADA–mPFC (20 Hz, 60 pulses of 5 ms, every 30 s). g, Individual ROI transient frequency analysed with CNMF-E. h, Individual ROI transient amplitude analysed with CNMF-E. ik, Data analysed using a non-ROI thresholded subtraction method (Chrimson: n = 4 mice, 44 ROIs; mCherry: n = 5 mice, 50 ROIs). i, Representative traces from a mPFC–dPAG::GCaMP6m neuron during the OFF–ON–OFF recording epochs. Calcium transients (yellow dots) for each neuron were identified and quantified. j, VTADA–mPFC stimulation decreased the average calcium event frequency per neuron, during both the ON and second OFF epochs. Data normalized to first OFF epoch; two-way repeated measure ANOVA, F2,184 = 9.209, P = 0.0002; Bonferroni multiple comparisons tests, P < 0.05. k, VTADA–mPFC stimulation increased the average calcium event amplitude per cell during the ON epoch, an effect that recovered in the second OFF epoch. Data normalized to first OFF epoch; two-way repeated measure ANOVA, F2,184 = 3.756, P = 0.0252; Bonferroni multiple comparisons tests: P < 0.05. ln, Data analysed using CNMF-E with removal of non-negative temporal constraints. Chrimson: n = 4 mice, 44 ROIs; mCherry: n = 5 mice; 50 ROIs. l, Representative traces from a mPFC–dPAG::GCaMP6m neuron during each 10 min OFF–ON–OFF recording epoch. Calcium transients (yellow dots) for each neuron were identified and quantified. m, VTADA–mPFC stimulation decreased the average calcium event frequency per neuron, during both the ON and second OFF epochs. Data normalized to first OFF epoch; two-way repeated measure ANOVA, F2,184 = 43.62, P < 0.0001; Bonferroni multiple comparisons tests: P < 0.05. n, VTADA–mPFC stimulation increased the average calcium event amplitude per cell during the ON epoch, an effect that recovered in the second OFF epoch. Data normalized to first OFF epoch; two-way repeated measure ANOVA, F2,184 = 3.50, P = 0.0322; Bonferroni multiple comparisons tests: P < 0.05. o, Schematic of viral strategy to optically manipulate ChR2-expressing VTADA terminals in the mPFC and record from mPFC–dPAG::ChR2 and non-expressing neighbouring neurons with ex vivo electrophysiology. p, Number of non-expressing cells with different responses to 1 Hz, 5-ms blue light delivery. q, Latency to action-potential peak for all ChR2-expressing cells plotted against light power density. Error bars represent s.e.m. The mouse brains in this figure were reproduced with permission from Paxinos and Franklin, 200454.

  9. Extended Data Fig. 9 Additional data for head-fixed electrophysiological recordings.

    a, Schematic of strategy to manipulate VTADA terminals in the mPFC and optically identify mPFC–dPAG::ChR2 neurons using in vivo head-fixed electrophysiology. b, Representative image of recording track in the mPFC (Rb, red retrobeads) and ChR2–eYFP-expressing mPFC–dPAG neurons. Representative image of ChR2–eYFP-expressing terminals surrounding the PAG. c, Histologically verified locations of recording tracks for in vivo head-fixed electrophysiology experiments. d, Population z-score of all phototagged units aligned to 1 Hz, 5-ms pulse of 473 nm. e, Photoresponse latencies showing <8 ms response latency from all 32 mPFC–dPAG::ChR2 units. f, g, PSTH from representative phototagged unit (f) and population z-score showing no response (g) to 20 Hz, 60 pulses of 593-nm laser light used for VTADA::Chrimson terminal activation. h, Representative PSTH of the firing rate in response to the onset of 5-ms pulse of 473-nm laser light used for phototagging. i, 204 mPFC units were recorded (n = 3 mice, 5 recording sessions) and 32 phototagged units were identified as mPFC–dPAG projectors (blue), 73 were photoinhibited (red), and 99 remained unidentified (grey). j, k, Neural response magnitudes to airpuff (x axis) and sucrose (y axis) in phototagged (j; blue) and unidentified (k; black) populations. l, m, In a subset of mice, both 405 and 473-nm laser light were used for phototagging. l, Representative phototagged unit showing faithful responses to 1 Hz, 5-ms pulses of both 473 and 405-nm light. m, Representative phototagged unit showing blunted response to 1 s of 405-nm, compared to 473-nm light. n, Representative PSTHs of photoidentified mPFC–dPAG units aligned to airpuff (green) and sucrose (purple). Histograms show neural responses in the OFF–ON–OFF epochs. o, p, Individual neural responses (AUC (0–500 ms post-stimulus presentation)) of every phototagged unit (n = 32 units) to airpuff (o) and sucrose (p) in each of the three recording epochs (OFF–ON–OFF). q, r, VTA dopamine terminal stimulation in the mPFC did not change the baseline firing rate (FR) in the 3-s pre-stimulus windows in the phototagged (q; Friedman test, χ2 = 2.472, P = 0.2905) or unidentified (r; Friedman test, χ2 = 0.4242, P = 0.8089) populations. s, VTADA terminal activation increased the frequency within a burst in the phototagged population. t, VTADA terminal activation did not affect burst characteristics in the unidentified population. Error bars indicate s.e.m. The mouse brains in this figure were reproduced with permission from Paxinos and Franklin, 200454.

  10. Extended Data Fig. 10 Dopamine-attenuates responses to airpuff in photoinhibited mPFC neurons.

    a, Schematic of strategy to manipulate VTADA terminals in the mPFC and optically identify mPFC–dPAG::ChR2 neurons using in vivo head-fixed electrophysiology. n = 3 mice, 5 recording sessions. b, 35.8% of recorded units (73/204) were photoinhibited. c, d, Representative PSTHs of a photoinhibited unit in response to 1 Hz, 5 ms (c) and 1 s (d) of 473-nm light. e, Neural response magnitudes to airpuff (x axis) and sucrose (y axis) in photoinhibited (red) population. f, VTADA terminal stimulation in the mPFC increased the baseline firing rate in the 3-s pre-stimulus windows in the photoinhibited population (n = 73 units) during the ON and second OFF epochs (Friedman test, χ2 = 16.22; P = 0.0003; Dunn’s multiple comparisons tests, P < 0.05). g, Population z-score of photoinhibited units aligned to airpuff in each of the recording epochs. h, In photoinhibited neurons, VTADA terminal stimulation attenuated neural responses to airpuff (Friedman test, χ2 = 8.329, P = 0.0155; Dunn’s multiple comparisons tests, P < 0.05). i, Population z-score of photoinhibited units aligned to sucrose in each of the recording epochs. j, In photoinhibited neurons, VTADA terminal stimulation did not affect neural responses to sucrose (Friedman test, χ2 = 0.4492, P = 0.7988; Dunn’s multiple comparisons tests, P > 0.05. k, VTADA terminal activation did not affect burst characteristics in the photoinhibited population. Error bars and shading represent s.e.m.

Supplementary information

  1. Supplementary Information

    This file contains a Supplementary Discussion

  2. Reporting Summary

  3. Video 1: Marble burying in mPFC-dPAG::ChR2. Representative video of marble burying behavior in an mPFC-dPAG::ChR2 subject during light off (left) or during stimulation of mPFC-dPAG projectors (right) (473 nm light, 20 Hz, 5 ms pulse width, continuous throughout 12 min session, n = 13 rats).

  4. Video 2: Calcium imaging analysis of mPFC-dPAG with CNMF-E. Example of constrained non-negative matrix factorization for microendoscopic data (CNMF-E) of GCaMP6m signals recorded with a miniature head-mounted microscope through a GRIN lens in mPFC. Each panel represents a component of the analysis process, played in synchrony. Raw data shows data as collected with no modifications. Background shows the estimation of background fluorescence (i.e. noise) by the algorithm. Raw-BG shows signals after removal of the background noise. Denoised and demixed show ‘cleaned’ signals for each neuron extracted from the video, and residuals show the remaining noise that was removed. Video shows a representative animal for mPFC-dPAG projectors, played at 30× real time.

  5. Video 3: Calcium imaging analysis of mPFC-NAc with CNMF-E. Example of constrained non-negative matrix factorization for microendoscopic data (CNMF-E) of GCaMP6m signals recorded with a miniature head-mounted microscope through a GRIN lens in mPFC. Each panel represents a component of the analysis process, played in synchrony. Raw data shows data as collected with no modifications. Background shows the estimation of background fluorescence (i.e. noise) by the algorithm. Raw-BG shows signals after removal of the background noise. Denoised and demixed show ‘cleaned’ signals for each neuron extracted from the video, and residuals shows the remaining noise that was removed. Video shows a representative animal for mPFC-NAc projectors, played at 30× real time.

  6. Video 4: Calcium imaging recordings in mPFC-dPAG and mPFC-NAc in response to foot shock. Representative video of mPFC-dPAG (left) and mPFC-NAc (right) projectors during exposure to foot shock. Videos were preprocessed (motion corrected and cropped), and converted to changes in fluorescence compared to background fluorescence using the equation (− F0)/F0, where F0 is equal to the mean fluorescence in each pixel over the entire video. Video is played in real time, and foot shock onset is marked by the white square in the center of the screen.

  7. Video 5: Calcium imaging recordings in mPFC-dPAG and mPFC-NAc in response to sucrose. Representative video of mPFC-dPAG (left) and mPFC-NAc (right) projectors during sucrose self-administration. Videos were preprocessed (motion corrected and cropped), and converted to changes in fluorescence compared to background fluorescence using the equation (− F0)/F0, where F0 is equal to the mean fluorescence in each pixel over the entire video. Video is played in real time, and sucrose bout initiation is marked by the white square in the center of the screen.

About this article

Publication history

Received

Accepted

Published

DOI

https://doi.org/10.1038/s41586-018-0682-1

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.