A whole-brain map of long-range inputs to GABAergic interneurons in the mouse medial prefrontal cortex

Abstract

The medial prefrontal cortex (mPFC) contains populations of GABAergic interneurons that play different roles in cognition and emotion. Their local and long-range inputs are incompletely understood. We used monosynaptic rabies viral tracers in combination with fluorescence micro-optical sectioning tomography to generate a whole-brain atlas of direct long-range inputs to GABAergic interneurons in the mPFC of male mice. We discovered that three subtypes of GABAergic interneurons in two areas of the mPFC are innervated by same upstream areas. Input from subcortical upstream areas includes cholinergic neurons from the basal forebrain and serotonergic neurons (which co-release glutamate) from the raphe nuclei. Reconstruction of single-neuron morphology revealed novel substantia innominata–anteromedial thalamic nucleus–mPFC and striatum–anteromedial thalamic nucleus–mPFC circuits. Based on the projection logic of individual neurons, we classified cortical and hippocampal input neurons into several types. This atlas provides the anatomical foundation for understanding the functional organization of the mPFC.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Characterization of monosynaptic inputs to three types of interneurons in different subregions of the mPFC.
Fig. 2: Visualization of the whole-brain input neurons to three types of interneurons in two subregions of the mPFC and quantification of the input neurons in individual brain regions.
Fig. 3: Modulation of GABAergic neurons in the mPFC by multiple neurotransmitters.
Fig. 4: Characterization of the functional inputs from cholinergic neurons in basal forebrain and serotonin neurons in raphe nuclei to GABAergic neurons in the mPFC.
Fig. 5: A neural circuit among mPFC, SI and AM.
Fig. 6: Distribution of input neurons in the neocortex to three types of GABAergic neurons in the mPFC.
Fig. 7: Characterization of the morphological properties of the cortical neurons that directly input to GABAergic neurons in the mPFC.
Fig. 8: Characterization of the morphological properties of the hippocampal pyramidal neurons that directly input to GABAergic neurons in the mPFC.

Data availability

The data that support the findings of this study are available in the Supplementary Tables and from the corresponding author upon reasonable request. Data including the whole-brain distribution of the input neurons and the reconstructions of the neural morphology in the neocortex and hippocampus can be accessed at http://atlas.brainsmatics.org/a/sun1903.

References

  1. 1.

    Pinto, L. & Dan, Y. Cell-type-specific activity in prefrontal cortex during goal-directed behavior. Neuron 87, 437–450 (2015).

  2. 2.

    Kvitsiani, D. et al. Distinct behavioural and network correlates of two interneuron types in prefrontal cortex. Nature 498, 363–366 (2013).

  3. 3.

    Croarkin, P. E., Levinson, A. J. & Daskalakis, Z. J. Evidence for GABAergic inhibitory deficits in major depressive disorder. Neurosci. Biobehav. Rev. 35, 818–825 (2011).

  4. 4.

    Lewis, D. A., Hashimoto, T. & Volk, D. W. Cortical inhibitory neurons and schizophrenia. Nat. Rev. Neurosci. 6, 312–324 (2005).

  5. 5.

    Khoshkhoo, S., Vogt, D. & Sohal, V. S. Dynamic, cell-type-specific roles for GABAergic interneurons in a mouse model of optogenetically inducible seizures. Neuron 93, 291–298 (2017).

  6. 6.

    Tremblay, R., Lee, S. & Rudy, B. GABAergic interneurons in the Neocortex: from cellular properties to circuits. Neuron 91, 260–292 (2016).

  7. 7.

    Hoover, W. B. & Vertes, R. P. Anatomical analysis of afferent projections to the medial prefrontal cortex in the rat. Brain Struct. Funct. 212, 149–179 (2007).

  8. 8.

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

  9. 9.

    Oh, S. W. et al. A mesoscale connectome of the mouse brain. Nature 508, 207–214 (2014).

  10. 10.

    Kim, T. et al. Cortically projecting basal forebrain parvalbumin neurons regulate cortical gamma band oscillations. Proc. Natl Acad. Sci. USA 112, 3535–3540 (2015).

  11. 11.

    Kabanova, A. et al. Function and developmental origin of a mesocortical inhibitory circuit. Nat. Neurosci. 18, 872–882 (2015).

  12. 12.

    Delevich, K., Tucciarone, J., Huang, Z. J. & Li, B. The mediodorsal thalamus drives feedforward inhibition in the anterior cingulate cortex via parvalbumin interneurons. J. Neurosci. 35, 5743–5753 (2015).

  13. 13.

    Wickersham, I. R. et al. Monosynaptic restriction of transsynaptic tracing from single, genetically targeted neurons. Neuron 53, 639–647 (2007).

  14. 14.

    Wall, N. R., Wickersham, I. R., Cetin, A., De La Parra, M. & Callaway, E. M. Monosynaptic circuit tracing in vivo through Cre-dependent targeting and complementation of modified rabies virus. Proc. Natl Acad. Sci. USA 107, 21848–21853 (2010).

  15. 15.

    Gong, H. et al. High-throughput dual-colour precision imaging for brain-wide connectome with cytoarchitectonic landmarks at the cellular level. Nat. Commun. 7, 12142 (2016).

  16. 16.

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

  17. 17.

    Watabe-Uchida, M., Zhu, L. S., Ogawa, S. K., Vamanrao, A. & Uchida, N. Whole-brain mapping of direct inputs to midbrain dopamine neurons. Neuron 74, 858–873 (2012).

  18. 18.

    Miyamichi, K. et al. Dissecting local circuits: parvalbumin interneurons underlie broad feedback control of olfactory bulb output. Neuron 80, 1232–1245 (2013).

  19. 19.

    Ohara, S. et al. Dual transneuronal tracing in the rat entorhinal-hippocampal circuit by intracerebral injection of recombinant rabies virus vectors. Front. Neuroanat. 3, 1 (2009).

  20. 20.

    DeNardo, L. A., Berns, D. S., DeLoach, K. & Luo, L. Q. Connectivity of mouse somatosensory and prefrontal cortex examined with trans-synaptic tracing. Nat. Neurosci. 18, 1687–1697 (2015).

  21. 21.

    Beier, K. T. et al. Circuit architecture of VTA dopamine neurons revealed by systematic input-output mapping. Cell 162, 622–634 (2015).

  22. 22.

    Beier, K. T. et al. Topological organization of ventral tegmental area connectivity revealed by viral-genetic dissection of input−output relations. Cell Rep. 26, 159–167 (2019).

  23. 23.

    Picciotto, M. R., Higley, M. J. & Mineur, Y. S. Acetylcholine as a neuromodulator: cholinergic signaling shapes nervous system function and behavior. Neuron 76, 116–129 (2012).

  24. 24.

    Lesch, K. P. & Waider, J. Serotonin in the modulation of neural plasticity and networks: implications for neurodevelopmental disorders. Neuron 76, 175–191 (2012).

  25. 25.

    Zaborszky, L. et al. Neurons in the basal forebrain project to the cortex in a complex topographic organization that reflects corticocortical connectivity patterns: an experimental study based on retrograde tracing and 3D reconstruction. Cereb. Cortex 25, 118–137 (2015).

  26. 26.

    Conti, F., DeBiasi, S., Minelli, A., Rothstein, J. D. & Melone, M. EAAC1, a high-affinity glutamate transporter, is localized to astrocytes and gabaergic neurons besides pyramidal cells in the rat cerebral cortex. Cereb. Cortex 8, 108–116 (1998).

  27. 27.

    Ren, J. et al. Anatomically defined and functionally distinct dorsal raphe serotonin sub-systems. Cell 175, 472–487 (2018).

  28. 28.

    Saunders, A., Granger, A. J. & Sabatini, B. L. Corelease of acetylcholine and GABA from cholinergic forebrain neurons. eLife 4, e06412 (2015).

  29. 29.

    Liu, Z. X. et al. Dorsal raphe neurons signal reward through 5-HT and glutamate. Neuron 81, 1360–1374 (2014).

  30. 30.

    Tamamaki, N. et al. Green fluorescent protein expression and colocalization with calretinin, parvalbumin, and somatostatin in the GAD67-GFP knock-in mouse. J. Comp. Neurol. 467, 60–79 (2003).

  31. 31.

    Schwarz, L. A. et al. Viral-genetic tracing of the input-output organization of a central noradrenaline circuit. Nature 524, 88–92 (2015).

  32. 32.

    Larsen, D. D., Wickersham, I. R. & Callaway, E. M. Retrograde tracing with recombinant rabies virus reveals correlations between projection targets and dendritic architecture in layer 5 of mouse barrel cortex. Front. Neural Circuits 1, 5 (2008).

  33. 33.

    Rock, C., Zurita, H., Lebby, S., Wilson, C. J. & Apicella, A. J. Cortical circuits of callosal GABAergic neurons. Cereb. Cortex 28, 1154–1167 (2018).

  34. 34.

    Economo, M. N. et al. Distinct descending motor cortex pathways and their roles in movement. Nature 563, 79–84 (2018).

  35. 35.

    van Aerde, K. I. & Feldmeyer, D. Morphological and physiological characterization of pyramidal neuron subtypes in rat medial prefrontal cortex. Cereb. Cortex 25, 788–805 (2015).

  36. 36.

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

  37. 37.

    Padilla-Coreano, N. et al. Direct ventral hippocampal-prefrontal input is required for anxiety-related neural activity and behavior. Neuron 89, 857–866 (2016).

  38. 38.

    Soltesz, I. & Losonczy, A. CA1 pyramidal cell diversity enabling parallel information processing in the hippocampus. Nat. Neurosci. 21, 484–493 (2018).

  39. 39.

    Fanselow, M. S. & Dong, H. W. Are the dorsal and ventral hippocampus functionally distinct structures? Neuron 65, 7–19 (2010).

  40. 40.

    Abbas, A. I. et al. Somatostatin interneurons facilitate hippocampal-prefrontal synchrony and prefrontal spatial encoding. Neuron 100, 926–939 (2018).

  41. 41.

    Hoover, W. B. & Vertes, R. P. Projections of the medial orbital and ventral orbital cortex in the rat. J. Comp. Neurol. 519, 3766–3801 (2011).

  42. 42.

    Ballinger, E. C., Ananth, M., Talmage, D. A. & Role, L. W. Basal forebrain cholinergic circuits and signaling in cognition and cognitive decline. Neuron 91, 1199–1218 (2016).

  43. 43.

    Szonyi, A. et al. The ascending median raphe projections are mainly glutamatergic in the mouse forebrain. Brain Struct. Funct. 221, 735–751 (2016).

  44. 44.

    Allen, T. G. J., Abogadie, F. C. & Brown, D. A. Simultaneous release of glutamate and acetylcholine from single magnocellular ‘cholinergic’ basal forebrain neurons. J. Neurosci. 26, 1588–1595 (2006).

  45. 45.

    Kim, J. H. et al. Selectivity of neuromodulatory projections from the basal forebrain and locus ceruleus to primary sensory cortices. J. Neurosci. 36, 5314–5327 (2016).

  46. 46.

    Li, X. N. et al. Generation of a whole-brain atlas for the cholinergic system and mesoscopic projectome analysis of basal forebrain cholinergic neurons. Proc. Natl Acad. Sci. USA 115, 415–420 (2018).

  47. 47.

    Wei, W. & Wang, X. J. Inhibitory control in the cortico-basal ganglia-thalamocortical loop: complex regulation and interplay with memory and decision processes. Neuron 92, 1093–1105 (2016).

  48. 48.

    Zhang, S. Y. et al. Organization of long-range inputs and outputs of frontal cortex for top-down control. Nat. Neurosci. 19, 1733–1742 (2016).

  49. 49.

    Ahrlund-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).

  50. 50.

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

  51. 51.

    Madisen, L. et al. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat. Neurosci. 13, 133–140 (2010).

  52. 52.

    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).

  53. 53.

    Zhang, Z. et al. Whole-brain mapping of the inputs and outputs of the medial part of the olfactory tubercle. Front Neural Circuits 11, 52 (2017).

  54. 54.

    Ragan, T. et al. Serial two-photon tomography for automated ex vivo mouse brain imaging. Nat. Methods 9, 255–258 (2012).

  55. 55.

    Zhang, J. et al. Presynaptic excitation via GABAB receptors in habenula cholinergic neurons regulates fear memory expression. Cell 166, 716–728 (2016).

  56. 56.

    Peng, J. et al. A quantitative analysis of the distribution of CRH neurons in whole mouse brain. Front Neuroanat. 11, 63 (2017).

  57. 57.

    Ni H., et al. A robust image registration interface for large volume brain atlas. Preprint at bioRxiv https://doi.org/10.1101/377044 (2018).

  58. 58.

    Kuan, L. et al. Neuroinformatics of the allen mouse brain connectivity atlas. Methods 73, 4–17 (2015).

Download references

Acknowledgements

We thank C. Zhou, T. Luo, X. Peng, C. Tan and Z. Duan for help with experiments and data analysis. We thank the Optical Bioimaging Core Facility of HUST for support with data acquisition, as well as the Analytical and Testing Center of HUST for spectral measurements. This work was financially supported by NSFC projects (nos. 61721092, 91632302, 91749209 and 31871088) and the Director Fund of WNLO.

Author information

Q.L. and H.G. conceived and designed the study. Q.S., X.L. and M.R. performed the tracing experiments. Y.R. and M.L. performed the physiological experiments. Q.Z., X.Z., C.Z. and J.Y. performed the whole-brain data acquisition. M.Z., H.N. and A.L. performed the imaging processing. Q.S., M.R., P.L. and X.L. performed the neuron reconstruction. Q.S., X.L., H.G. and Q.L. wrote the paper.

Correspondence to Qingming Luo.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information: Nature Neuroscience thanks Ian Wickersham and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Integrated supplementary information

Supplementary Figure 1 Validation of the specificity of virus labeling.

a, b. immunostaining against parvalbumin and somatostatin at the injection sites. Parvalbumin is expressed in the TVA-mCherry and EGFP positive neurons in PV-Cre mice and somatostatin is expressed in the TVA-mCherry and EGFP positive neurons in SST-Cre mice (PV+, n=6 mice, SST+, n=6 mice). c-n. Control experiments of RV monosynaptic tracing. c-e. No neuron was labeled when RV-EnvA-GFP was injected into mouse brain alone (n=3 mice). f-j. Only the specific neurons at the injection site were labeled when AAV-DIO-RG was absent (n=4 mice). k. Quantification of AAV/RV labeled neurons that are PV+ (n=4 mice, data shown as mean±s.e.m). l-n. A few neurons were labeled at the injection site when AAV helper virus and RV were injected into the wild type mouse brain but no neuron was labeled outside the injection stite (n=3 mice). Scale bar a,b 100 µm, c-e, 200 µm, f.1mm, g-j. 50 µm, l-n. 1 mm

Supplementary Figure 2 Layer distribution of three types of interneurons in mPFC.

a. Representatitve coronal sections showing distribution of PV, SST and VIP positive neurons. The PV, SST and VIP neurons were labeled by crossing PV, SST and VIP-Cre driver lines with Ai14 reporter line. b. Quantification of the layer distribution of PV, SST and VIP positive neurons in mPFC. (SST+, n=3 mice; PV+, n=3 mice; VIP+, n=3 mice, data shown as mean±s.e.m) Scale bar a. 1 mm.

Supplementary Figure 3 Whole brain inputs to SST, PV and VIP positive neurons in mPFC.

a. Representatitve coronal sections showing EGFP labeled cells input to SST, PV and VIP positive neurons in different brain regions with PL targeting. The projections were 50µm. b. A sample of 210-serial section raw data set of RV labeled mouse brain imaged by fMOST system(SST+, n=18 mice; PV+, n=14 mice; VIP+, n=10 mice). The projections were 50µm. Scale bar, a.1 mm.

Supplementary Figure 4 Direct comparison of the long range inputs to SST positive neurons in two subregions of mPFC within individual brain sample.

a, Representatitve coronal sections showing EGFP labeled cells and DsRed labeled cells in different brain areas. The sections were registered to the Allen CCF3.0. b, The distribution of EGFP labeled cells and DsRed labeled cells in different brain regions (For dual-color RV labeling, SST+, n=2 mice, PV+, n=2 mice, VIP+, n=2 mice). Scale bars are 1 mm in a and b.

Supplementary Figure 5 Visualization of PL and ILA inputs with dual-color CTb labeling.

a, b. Conjugated CTb with different fluorescence was injected into PL and ILA, respectively. c-j. The CTb labeled neurons in major brain areas that innervate PL and ILA were shown in c-j. The number of PL projecting, ILA projecting and double projecting neurons were also quantified (one-way ANOVA followed by Tukey’s post hoc tests, **p < 0.01, ***p<0.001, ****p<0.0001, for detailed p value, see supplementary table 3, n=3 mice, data shown as mean±s.e.m). Scale bars in a and b are 1mm, those in c-j are 100µm.

Supplementary Figure 6 Quantitative analysis of whole-brain monosynaptic inputs to different GABAergic neurons in the medial prefrontal cortex at grouped brain area level, related to Fig 2.

(one-way ANOVA followed by Tukey’s post hoc tests, *p < 0.05, ***p < 0.001, ****p<0.0001,for detailed p value, see main article text, SST+, n=18 mice; PV+, n=14 mice; VIP+, n=10 mice, data shown as mean±s.e.m).

Supplementary Figure 7 Representative images showing the brain areas that differently innervated SST+, PV+ and VIP+ neurons in mPFC, related to Fig. 2.

PV+ and VIP+ neurons receive more inputs from some cortical areas (a, c, f) while SST+ neurons receive more inputs from some subcortical nuclei (b, d, e, g, h)(SST+, n=18 mice; PV+, n=14 mice; VIP+, n=10 mice). Scale bar 100 µm.

Supplementary Figure 8 Quantitative analysis of the differences between the inputs of prelimbic area and infralimbic area.

a. Brain areas that differently innervated SST+ neurons in PL and IL. b. Brain areas that differently innervated VIP+ neurons in PL and IL.(one-way ANOVA followed by Tukey’s post hoc tests, *p < 0.05, **p < 0.01, for detailed p value, see main article text, SST+, n=18 mice; PV+, n=14 mice; VIP+, n=10 mice, data shown as mean±s.e.m).

Supplementary Figure 9 Analysis of the relationship between spatial distribution of starter cells and input patterns.

a. A 3-dimensional view of the spatial distribution of starter cells in PL and ILA. The starter cells of PL targeting sample were shown in red and the starter cells of ILA targeting sample were shown in green. b. The mean vertical distance to midline of the starter cells in SST, PV and VIP-Cre mice (SST, n=12 mice, PV, n=8 mice, VIP, n=8 mice; one-way ANOVA followed by Tukey’s post hoc tests, SST+ versus VIP+, p=0.045, PV+ vs VIP+, p=0.009, data shown as mean±s.e.m). c. The mean vertical distance to brain surface of the starter cells in PL targeting samples and ILA targeting samples (SST, n=12 mice, PV, n=8 mice, VIP, n=8 mice, two-tailed unpaired t test, p<0.0001, data shown as mean±s.e.m). d. The mean vertical distance to the coronal plane under bregma point of the starter cells in PL targeting samples and ILA targeting samples (SST, n=12 mice, PV, n=8 mice, VIP, n=8 mice, two-tailed unpaired t test, p<0.0001, data shown as mean±s.e.m) (Bregma point is defined as the point on the top of the skull where the coronal and sagittal sutures meet). e. Linear regression analysis with mean vertical distance to midline, brain surface and the coronal plane under bregma point. f. Linear regression analysis with COM, Cre line and COM Cre line combination. (one-way ANOVA followed by Tukey’s post hoc tests, *p < 0.05, **p < 0.01, ****p<0.0001, SST, n=12 mice, PV, n=8 mice, VIP, n=8 mice) COM, starter cell center of mass.

Supplementary Figure 10 Characterization of the neurochemical properties of input neurons by double immunochemical staining, related to Fig. 3.

a. Input neurons in HDB express choline acetyltransferase but not EAAC1 (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice). b. Input neurons in HDB express EAAC1 but not choline acetyltransferase (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice). c. Input neurons in dorsal raphe nucleus express TPH2 but not EAAC1 (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice).d. Input neurons in dorsal raphe nucleus express EAAC1 but not TPH2 (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice). e, f. Input neurons in VTA express EAAC1 and TH, respectively (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice). g. Quantification of EGFP labelled neurons in VTA that are Dopaminergic and EAAC1 positive (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice, data shown as mean±s.e.m). h. Colocalization rate of chat and EAAC1 expression of neurons in basal forebrain that project to mPFC (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice, data shown as mean±s.e.m). i. Colocalization rate of TPH2 and EAAC1 expression of neurons in raphe nuclei that project to mPFC. (SST-Cre, n=6 mice; PV-Cre, n=6 mice; VIP-Cre, n=8 mice, data shown as mean±s.e.m) Scale bar=50 µm.

Supplementary Figure 11 Neuron in anteromedial thalamic nucleus that project to mPFC receive inputs from multiple brain areas, related to Fig. 5.

a.Brain areas that innervated AM and mPFC. b. Brain areas that innervated AM alone. n=6 mice. Scale bar=100 µm.

Supplementary Figure 12 Characterization and quantification of the neurons that directly project to mPFC and indirectly connect with mPFC via mPFC projecting neurons in AM.

a, b. Both cholinergic and non-cholinergic neurons in SI can project to mPFC and form synaptic connections with neurons in AM that project to mPFC (n=4 mice). c. Quantification of mPFC projecting neurons, AM projecting neurons and double projecting neurons in SI. d. Quantification of double projecting neurons in SI that are cholinergic. e-h. Quantification of mPFC projecting neurons, AM projecting neurons and double projecting neurons in MOs, LHA, HDB and CA1. (For quantification of neurons in SI, MOs, HDB, n=4 mice; for quantification of neurons in LHA, CA1, n=3 mice, data shown as mean±s.e.m)Scale bars in a and b are 50 µm.

Supplementary Figure 13

Seven different types of cortical neurons innervated GABAergic neurons in mPFC.

Supplementary Figure 14 Quantitative comparison of the fine morphology of four types of different pyramidal neurons in layer V that target GABAergic neurons in mPFC, related to fig. 7.

Histograms of the dendrite length and the apical dendrite length are shown in a, b, respectively.(a, one-way ANOVA followed by Tukey’s post hoc tests, *p < 0.05, **p < 0.01, ***p<0.001, ****p<0.0001, callosal neurons, n=19, corticofugal neurons, n=17, corticospinal neurons, n=5, associative neurons, n=16, callosal neurons vs corticofugal neurons, p=0.0084, callosal neurons vs corticospinal neurons, p<0.0001, corticofugal neurons vs corticospinal neurons, p=0.033, corticospinal neurons vs associative neurons, p<0.0001, data shown as mean±s.e.m; b, two-tailed unpaired t test, p=0.0006, data shown as mean±s.e.m).

Supplementary information

Supplementary Figs. 1–14 and Supplementary Tables 1–15.

Reporting Summary

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Further reading