Article | Published:

Active dendritic integration and mixed neocortical network representations during an adaptive sensing behavior

Nature Neurosciencevolume 21pages15831590 (2018) | Download Citation

Abstract

Animals strategically scan the environment to form an accurate perception of their surroundings. Here we investigated the neuronal representations that mediate this behavior. Ca2+ imaging and selective optogenetic manipulation during an active sensing task reveals that layer 5 pyramidal neurons in the vibrissae cortex produce a diverse and distributed representation that is required for mice to adapt their whisking motor strategy to changing sensory cues. The optogenetic perturbation degraded single-neuron selectivity and network population encoding through a selective inhibition of active dendritic integration. Together the data indicate that active dendritic integration in pyramidal neurons produces a nonlinearly mixed network representation of joint sensorimotor parameters that is used to transform sensory information into motor commands during adaptive behavior. The prevalence of the layer 5 cortical circuit motif suggests that this is a general circuit computation.

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

The data that support the findings of this study are stored on Janelia Research Campus servers and are available from the corresponding author upon reasonable request.

Additional information

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

References

  1. 1.

    Yarbus, A. L. in Eye Movements and Vision 171–211 (Springer, New York, NY, USA, 1967).

  2. 2.

    Kleinfeld, D., Ahissar, E. & Diamond, M. E. Active sensation: insights from the rodent vibrissa sensorimotor system. Curr. Opin. Neurobiol. 16, 435–444 (2006).

  3. 3.

    Cullen, K. E. Sensory signals during active versus passive movement. Curr. Opin. Neurobiol. 14, 698–706 (2004).

  4. 4.

    Mitchinson, B., Martin, C. J., Grant, R. A. & Prescott, T. J. Feedback control in active sensing: rat exploratory whisking is modulated by environmental contact. Proc. Biol. Sci. 274, 1035–1041 (2007).

  5. 5.

    Pouget, A. & Snyder, L. H. Computational approaches to sensorimotor transformations. Nat. Neurosci. 3(Suppl), 1192–1198 (2000).

  6. 6.

    Denève, S. & Pouget, A. Basis functions for object-centered representations. Neuron 37, 347–359 (2003).

  7. 7.

    Rigotti, M. et al. The importance of mixed selectivity in complex cognitive tasks. Nature 497, 585–590 (2013).

  8. 8.

    Pouget, A. & Sejnowski, T. J. Spatial transformations in the parietal cortex using basis functions. J. Cogn. Neurosci. 9, 222–237 (1997).

  9. 9.

    Andersen, R. A., Snyder, L. H., Bradley, D. C. & Xing, J. Multimodal representation of space in the posterior parietal cortex and its use in planning movements. Annu. Rev. Neurosci. 20, 303–330 (1997).

  10. 10.

    Asaad, W. F., Rainer, G. & Miller, E. K. Neural activity in the primate prefrontal cortex during associative learning. Neuron 21, 1399–1407 (1998).

  11. 11.

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

  12. 12.

    Alexander, A. S. & Nitz, D. A. Retrosplenial cortex maps the conjunction of internal and external spaces. Nat. Neurosci. 18, 1143–1151 (2015).

  13. 13.

    Barak, O., Rigotti, M. & Fusi, S. The sparseness of mixed selectivity neurons controls the generalization-discrimination trade-off. J. Neurosci. 33, 3844–3856 (2013).

  14. 14.

    Miller, K. D. Canonical computations of cerebral cortex. Curr. Opin. Neurobiol. 37, 75–84 (2016).

  15. 15.

    Mel, B. W. & Koch, C. Sigma-Pi learning: on radial basis functions and cortical associative learning. Adv. Neural Inf. Process. Syst. 2, 474–481 (1989).

  16. 16.

    Koch, C. & Poggio, T. in Single Neuron Computation (eds. McKenna, T.M., Davis, J.L. & Zornetzer, S.F.) 315–345 (Academic Press, San Diego, CA, USA, 1992).

  17. 17.

    Silver, R. A. Neuronal arithmetic. Nat. Rev. Neurosci. 11, 474–489 (2010).

  18. 18.

    Harnett, M. T., Xu, N.-L., Magee, J. C. & Williams, S. R. Potassium channels control the interaction between active dendritic integration compartments in layer 5 cortical pyramidal neurons. Neuron 79, 516–529 (2013).

  19. 19.

    Harnett, M. T., Magee, J. C. & Williams, S. R. Distribution and function of HCN channels in the apical dendritic tuft of neocortical pyramidal neurons. J. Neurosci. 35, 1024–1037 (2015).

  20. 20.

    Mao, T. et al. Long-range neuronal circuits underlying the interaction between sensory and motor cortex. Neuron 72, 111–123 (2011).

  21. 21.

    Petreanu, L. et al. Activity in motor-sensory projections reveals distributed coding in somatosensation. Nature 489, 299–303 (2012).

  22. 22.

    Hill, D. N., Curtis, J. C., Moore, J. D. & Kleinfeld, D. Primary motor cortex reports efferent control of vibrissa motion on multiple timescales. Neuron 72, 344–356 (2011).

  23. 23.

    Kleinfeld, D. & Deschênes, M. Neuronal basis for object location in the vibrissa scanning sensorimotor system. Neuron 72, 455–468 (2011).

  24. 24.

    Xu, N.-L. et al. Nonlinear dendritic integration of sensory and motor input during an active sensing task. Nature 492, 247–251 (2012).

  25. 25.

    O’Connor, D. H. et al. Vibrissa-based object localization in head-fixed mice. J. Neurosci. 30, 1947–1967 (2010).

  26. 26.

    Pammer, L. et al. The mechanical variables underlying object localization along the axis of the whisker. J. Neurosci. 33, 6726–6741 (2013).

  27. 27.

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

  28. 28.

    Yu, C., Derdikman, D., Haidarliu, S. & Ahissar, E. Parallel thalamic pathways for whisking and touch signals in the rat. PLoS Biol. 4, e124 (2006).

  29. 29.

    Moore, J. D., Mercer Lindsay, N., Deschênes, M. & Kleinfeld, D. Vibrissa self-motion and touch are reliably encoded along the same somatosensory pathway from brainstem through thalamus. PLoS Biol. 13, e1002253 (2015).

  30. 30.

    Curtis, J. C. & Kleinfeld, D. Phase-to-rate transformations encode touch in cortical neurons of a scanning sensorimotor system. Nat. Neurosci. 12, 492–501 (2009).

  31. 31.

    Crochet, S., Poulet, J. F., Kremer, Y. & Petersen, C. C. Synaptic mechanisms underlying sparse coding of active touch. Neuron 69, 1160–1175 (2011).

  32. 32.

    Han, X. et al. A high-light sensitivity optical neural silencer: development and application to optogenetic control of non-human primate cortex. Front. Syst. Neurosci. 5, 18 (2011).

  33. 33.

    Salinas, E. & Abbott, L. F. Coordinate transformations in the visual system: how to generate gain fields and what to compute with them. Prog. Brain Res. 130, 175–190 (2001).

  34. 34.

    Salinas, E. & Abbott, L. F. A model of multiplicative neural responses in parietal cortex. Proc. Natl. Acad. Sci. USA 93, 11956–11961 (1996).

  35. 35.

    Zipser, D. & Andersen, R. A. A back-propagation programmed network that simulates response properties of a subset of posterior parietal neurons. Nature 331, 679–684 (1988).

  36. 36.

    Buonomano, D. V. & Maass, W. State-dependent computations: spatiotemporal processing in cortical networks. Nat. Rev. Neurosci. 10, 113–125 (2009).

  37. 37.

    Jaeger, H. & Haas, H. Harnessing nonlinearity: predicting chaotic systems and saving energy in wireless communication. Science 304, 78–80 (2004).

  38. 38.

    Sussillo, D. & Abbott, L. F. Generating coherent patterns of activity from chaotic neural networks. Neuron 63, 544–557 (2009).

  39. 39.

    Spruston, N. Pyramidal neurons: dendritic structure and synaptic integration. Nat. Rev. Neurosci. 9, 206–221 (2008).

  40. 40.

    Larkum, M. E., Zhu, J. J. & Sakmann, B. A new cellular mechanism for coupling inputs arriving at different cortical layers. Nature 398, 338–341 (1999).

  41. 41.

    Takahashi, N., Oertner, T. G., Hegemann, P. & Larkum, M. E. Active cortical dendrites modulate perception. Science 354, 1587–1590 (2016).

  42. 42.

    Bittner, K. C. et al. Conjunctive input processing drives feature selectivity in hippocampal CA1 neurons. Nat. Neurosci. 18, 1133–1142 (2015).

  43. 43.

    Manita, S. et al. A top-down cortical circuit for accurate sensory perception. Neuron 86, 1304–1316 (2015).

  44. 44.

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

  45. 45.

    Zhang, S. et al. Selective attention. Long-range and local circuits for top-down modulation of visual cortex processing. Science 345, 660–665 (2014).

  46. 46.

    Chen, S. X., Kim, A. N., Peters, A. J. & Komiyama, T. Subtype-specific plasticity of inhibitory circuits in motor cortex during motor learning. Nat. Neurosci. 18, 1109–1115 (2015).

  47. 47.

    Makino, H. & Komiyama, T. Learning enhances the relative impact of top-down processing in the visual cortex. Nat. Neurosci. 18, 1116–1122 (2015).

  48. 48.

    Xu, T. et al. Rapid formation and selective stabilization of synapses for enduring motor memories. Nature 462, 915–919 (2009).

  49. 49.

    Bittner, K. C., Milstein, A. D., Grienberger, C., Romani, S. & Magee, J. C. Behavioral time scale synaptic plasticity underlies CA1 place fields. Science 357, 1033–1036 (2017).

  50. 50.

    Khan, A. G. et al. Distinct learning-induced changes in stimulus selectivity and interactions of GABAergic interneuron classes in visual cortex. Nat. Neurosci. 21, 851–859 (2018).

  51. 51.

    Greenberg, D. S. & Kerr, J. N. Automated correction of fast motion artifacts for two-photon imaging of awake animals. J. Neurosci. Methods 176, 1–15 (2009).

  52. 52.

    Guizar-Sicairos, M., Thurman, S. T. & Fienup, J. R. Efficient subpixel image registration algorithms. Opt. Lett. 33, 156–158 (2008).

  53. 53.

    Bingham, E. & Hyvärinen, A. A fast fixed-point algorithm for independent component analysis of complex valued signals. Int. J. Neural Syst. 10, 1–8 (2000).

  54. 54.

    Hyvärinen, A. & Oja, E. Independent component analysis: algorithms and applications. Neural Netw. 13, 411–430 (2000).

  55. 55.

    O’Connor, D. H., Peron, S. P., Huber, D. & Svoboda, K. Neural activity in barrel cortex underlying vibrissa-based object localization in mice. Neuron 67, 1048–1061 (2010).

  56. 56.

    Clack, N. G. et al. Automated tracking of whiskers in videos of head fixed rodents. PLoS Comput. Biol. 8, e1002591 (2012).

  57. 57.

    O’Connor, D. H. et al. Neural coding during active somatosensation revealed using illusory touch. Nat. Neurosci. 16, 958–965 (2013).

  58. 58.

    Cajigas, I., Malik, W. Q. & Brown, E. N. nSTAT: open-source neural spike train analysis toolbox for Matlab. J. Neurosci. Methods 211, 245–264 (2012).

  59. 59.

    Dimatteo, I., Genovese, C. R. & Kass, R. E. Bayesian curve‐fitting with free‐knot splines. Biometrika 88, 1055–1071 (2001).

Download references

Acknowledgements

We thank M. Tadross and K. Bittner for comments on the manuscript; R. Behnam, A. Karpova, and the Molecular Biology & Virus services at JRC for GCaMP6 and ArchT constructs; the Svoboda lab for help with software and analysis; A. Hu and M. Copeland for histology; S. Lindo for in utero surgeries; and J.J. Macklin for light probes. This work was supported by the Howard Hughes Medical Institute.

Author information

Author notes

  1. These authors contributed equally: Gayathri N. Ranganathan, Pierre F. Apostolides.

Affiliations

  1. Howard Hughes Medical Institute, Janelia Research Campus, Ashburn, VA, USA

    • Gayathri N. Ranganathan
    • , Pierre F. Apostolides
    • , Shaul Druckmann
    •  & Jeffrey C. Magee
  2. Kresge Hearing Research Institute Department of Otolaryngology, University of Michigan , Ann Arbor, MI, USA

    • Pierre F. Apostolides
  3. McGovern Institute for Brain Research, Massachusetts Institute of Technology, Cambridge, MA, USA

    • Mark T. Harnett
  4. Institute of Neuroscience, State Key Laboratory of Neuroscience, CAS Center for Excellence in Brain Science and Intelligence Technology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China

    • Ning-Long Xu
  5. Howard Hughes Medical Institute, Baylor College of Medicine, Houston, TX, USA

    • Jeffrey C. Magee

Authors

  1. Search for Gayathri N. Ranganathan in:

  2. Search for Pierre F. Apostolides in:

  3. Search for Mark T. Harnett in:

  4. Search for Ning-Long Xu in:

  5. Search for Shaul Druckmann in:

  6. Search for Jeffrey C. Magee in:

Contributions

G.N.R., N.-L.X., and J.C.M. conceived the project and designed the experiments. G.N.R. performed behavioral and chronic imaging experiments and analysis. P.F.A. and G.N.R. performed juxtacellular electrophysiology experiments and analysis. P.F.A. performed slice experiments and analysis. M.T.H. performed whole-cell voltage recordings and analysis. G.N.R. and S.D. performed population-encoding analyses. G.N.R. and J.C.M. wrote the paper with input from all authors.

Competing interests

The authors declare no competing interests.

Corresponding author

Correspondence to Jeffrey C. Magee.

Integrated supplementary information

  1. Supplementary Figure 1 GCaMP6f signals in L5 proximal dendrites are linear over a wide range of spike frequencies.

    a, Whole-cell recordings were made from layer 5 neurons in somatosensory cortex. GCaMP6f was expressed via in utero injections of a rAAV2/1-syn-flex-GCaMP6f virus in RBP4-cre mice. Micrograph shows a 2-photon z-stack of a GCaMP6f-expressing neuron recorded in a brain slice. Linescan location for panels b and c is shown by the red dashed line. b, Left: Example traces of a single trial spike train (top traces) and GCaMP6f fluorescence changes at the linescan location denoted in a. Fluorescence traces are averages of 4,3,3,3. Right: Summary data (n = 8 cells from 3 mice) showing the fluorescence increase for trains of 5 spikes at frequencies ranging from 10-100 Hz. Gray dashed lines are single experiments, red is mean ± sem. c) same as panel b, but for trains of 10 spikes (n = 8 cells from 3 mice). Traces are averages of 4,4,3,3 trials respectively.

  2. Supplementary Figure 2 Touch location tuning in L5 network and somatic responses.

    a, Raster plots of Ca2+ signals sorted by object location. Highlighted locations are those shown in Fig. 2. b, Normalized mean Ca2+ signal amplitudes from all cells plotted as a function of object location with the preferred location centered at 0. Gray lines are individual cells. Black symbols are mean±sem where n=14, 20, 32, 39, 88, 59, 53, 41, 34 for locations -4 through +4. c, Plot of average touch magnitudes sorted by location of touch. Plotted is mean±sem where n=258, 263, 592, 541, 526, 321 for locations 1 through 6. d, Normalized location tuning from slopes of Ca2+ response dependence on touch magnitude from all cells. Preferred location is centered at 0. Gray lines are individual cells. Black symbols are mean±sem where n=12, 25, 46, 57, 88, 70, 62, 51, 35 for positions -4 through +4. e, Histogram of selectivity indices from all touch responsive neurons recorded. Mean±sem shown. f, Normalized location tuning from mean Ca2+ responses from signals recorded from all neurons.

  3. Supplementary Figure 3 Determination of self-motion signal.

    a, Plots showing whisker angle (□touch, blue) and whisker phase (□touch, green) at touch sorted by object location (1-6). Correlation measures (Pearson’s correlation coefficient) are reported for 3315 protraction touches. b, Results of a linear classifier trained to predict object location using □touch (blue) and □touch (green). c, Average accuracy in prediction of whisker angle (□touch) and whisker phase (□touch) at touch from L5 Ca2+ signals (20 runs from 88 neurons). d, Histogram of pairwise cross-correlation coefficients from the selectivity of all L5 neurons recorded for □touch (blue) and □touch (green). e, Histogram of area under ROC curve for all neurons collected (n=88), analyzed for their discriminability of □touch (blue) and □touch (green). f, Selectivity indices of neurons (n=88) for □touch plotted against the corresponding selectivity index for □touch. R value shown is Pearson’s correlation coefficient.

  4. Supplementary Figure 4 Simultaneous silencing and imaging activity.

    a, Co-expression of ArchT (red fluorophore) and GCaMP6f (green fluorophore) in the somata and apical dendrites of L5 pyramidal neurons. Similar results were found in all mice tested (n=3). b, Intensity profile of dispersive light. Similar results were found in all mice tested (n=3). c, Tuft Ca2+ signal amplitudes from multiple branches of the same neuron. Classification of events as global Ca2+ spikes was based on the presence of a threshold crossing Ca2+ event (> 30%∆F/F, dotted line) in all the recorded branches. d, Global events showed higher cross-correlation values. e, Effect of light activated selective tuft silencing on the probability of Ca2+ events recorded in the output compartment of L5 neurons (n=200 cells; two-sided, paired Student’s t-test)(gray lines are individual cells, open circles are mean±sem). f, Relative probability of output Ca2+ signals as a function of Ca2+ event amplitude, control (black) and light manipulation (red) trials. g, Fractional reduction in Ca2+ event probability vs. amplitude of Ca2+ events (10% more reduction in event probability for every 200% increase in event amplitude). h, Whole-cell recordings from L5 neurons under anesthetized conditions. F-I curves from all neurons showing small rightward shift from light activation (symbols are mean±sem, n=6 cells). i, Summary data showing resulting change in input – output relationship from the effects of light induced shift in H (two-sided, paired Student’s t-test). j, Expected fractional change in event probability for a simulated subtractive light effect using control data in f

  5. Supplementary Figure 5 ArchT expression in L5 neurons is similar using virus or transgenic mice.

    a, Whole-cell recordings were obtained from layer 5 neurons in acute brain slices. ArchT was expressed by injecting Rbp4-cre mice with an ArchT virus (see Methods) or by breeding Rbp4-cre and Ai40D mice. Micrograph shows a 2-photon maximal intensity z-stack of a virally infected layer 5 neuron filled with Alexa594, along with neighboring infected cells (green). Scale bar = 100 µm. A fluorescence ratio of recorded to neighboring cells in the same field of view was 2.9 ± 0.5 (mean and sem), suggesting that our visual targeting approach may have biased our recordings to neurons exhibiting a relatively high expression level. b, Top: Example traces showing that 590 nm light pulses (300 ms) via full-field illumination hyperpolarizes the membrane potential of somatosensory cortex layer 5 neurons expressing ArchT via transgenic or viral methods (black and red traces, respectively). c, Summary data showing the effect of light intensity on membrane potential hyperpolarization for n=9 and n=10 neurons in transgenic and virus groups, respectively. Solid lines represent individual experiments and circles are mean ± SEM. A two-way repeated measures ANOVA reveals a main effect of light intensity (F(5,85)=96.17, p = 0.000073), but the membrane potential hyperpolarization was not significantly different between transgenic and virus groups at any light intensity (p=0.239 for each light intensity after correcting for multiple comparisons, Bonferonni post-hoc test).

  6. Supplementary Figure 6 Simulated impact of tuft ArchT activation on dendritic and neuronal output.

    Simulation uses the scheme shown in Fig. 5m, n. a, Dendritic hyperpolarization via ArchT activation effectively shifts the dendritic activation function to higher input values. This dramatically decreases the slope of the activation function engaged by the dendritic input (0-6 units under both control and light conditions). Thus for the same range of dendritic (motor) inputs, the dendritic output (current leaving the tuft regions) is highly reduced by ArchT activation. To simulate the interaction of the proximal and distal active compartments the slope of the dendritic activation function is itself a function of sensory input. This mimics the effect of higher bAP frequency on the rate of dendritic Ca2+ spikes (22,44). b, Dendritic output generated by running a range of dendritic input (0-6 units) through either the control (x-axis) or light (y-axis) dendritic activation function for three different levels of sensory input (5, 8, 11 units). c, Subtraction of light output values from the control values plotted against control values showing the very large divisive effect of shifting the dendritic functions by light. B and C reproduce data shown in Fig. 5c, d. d, Soma/axon activation function where output is AP frequency and input is combined dendritic output (as above) and sensory input (5-11 units). Total input range is greatly curtailed by simulated ArchT activation (light) due to reduced dendritic output. e&f, as in b&c but using the soma/axonal activation function to determine neuronal output. This simulates divisive impact on output, as observed in Fig. 4g, h. g, Simulation reproducing location dependent (motor input, 0-6 units) impact on the slope of the response (output frequency)-touch magnitude (sensory input, 5-11 units) relationship. h, Same for light conditions. These plots reproduce data shown in Fig. 5.

  7. Supplementary Figure 7 Selective dendritic silencing disrupts mixed selectivity in L5 neurons and the L5 network representation of object location.

    a, Touch evoked response amplitudes from an example L5 neuron under control (black) and tuft silenced (red) conditions, sorted by location of whisker touch. b, Reduction in mean Ca2+ signal amplitude from selectively diminishing global Ca2+ spikes in L5 neurons (n=200 cells; two-sided, paired Student’s t-test; gray lines are individual cells and black are mean±sem). c, Touch magnitude is not changed by light (n=1968 for control and 2223 touches for light; two-sided, paired Student’s t-test). The ends of the box are upper and lower quartiles, vertical line is median and whiskers are the minimum and maximum data values. d, Reduction in location tuning in normalized mean response amplitudes from diminishing Ca2+ spikes, control (black) and tuft-silenced (red) conditions. Symbols are mean±sem where n=18, 66, 105, 129, 200, 146, 108, 69, 19 for locations -4 to +4 in control (left) and 16, 64, 103, 119, 200, 146, 108, 69, 19 for locations -4 to +4 in light (right). e, Nonlinear reduction in location tuning from slopes of responses amplitude as a function of touch magnitude from diminishing Ca2+ spikes. Symbols are mean±sem where n=10, 35, 80, 149, 135, 112, 65 for locations -3 to +3 in control (left) and 10, 35, 80, 143, 126, 112, 65 for locations -3 to +3 in light (right). In d&e light colored lines are individual cells.

  8. Supplementary Figure 8 Selective silencing of Ca2+ spikes degrades selectivity to whisker phase at touch and whisker angle at touch.

    a, Touch evoked response amplitudes from an example L5 neuron under control (black) and tuft silenced (red) conditions, sorted by whisker angle at touch (□touch). Values shown are mean±sem for n=42, 66, 51, 43 trials, from left to right. b, Touch evoked response amplitudes of the same neuron sorted by whisker phase at touch (□touch). Values shown are for n=21, 105, 43, 33 trials, from left to right. c, Reduction in □touch tuning in normalized mean response amplitudes from diminishing Ca2+ spikes, control (black) and tuft-silenced (red) conditions. Black symbols are mean±sem where n=72, 88, 113, 152, 94, 65, 40 for angles ±19.75°. Red symbols are mean±sem with samples sizes of 72, 88, 113, 152, 94, 65, 40 for angles ±19.75°. d, Reduction in □touch tuning in normalized mean response amplitudes. Black symbols are mean±sem with samples sizes of 36, 90, 144, 170, 200, 164, 110, 56, 30 for phases ±1.6π. Red symbols are mean±sem where n=36, 89, 144, 168, 197, 160, 110, 56, 30 for phases ±1.6π. In c&d light colored lines are individual cells. e, Histogram of pairwise correlation coefficients of neurons’ selectivity to □touch. f, Histogram of pairwise correlation coefficients of neurons’ selectivity to □touch. g, Scatter plot of selectivity indices of individual L5 neurons calculated for □touch and □touch, showing joint encoding of both parameters.

  9. Supplementary Figure 9 Selective silencing of Ca2+ spikes affects sensorimotor behavior (part 1).

    a, Whisker occupancy histograms over baseline (1&2) and switch (3-8) sessions where the mean object location (vertical red line) was moved anterior, in an example control mouse. b, Relative change in mean whisker occupancy over the eight sessions (above) and relative change in whisker angle (below) in the same example mouse. c, Whisker occupancy histograms from baseline and switch sessions similar to a, in a mouse exposed to selective silencing of global Ca2+ spikes in L5 pyramidal neurons in the vibrissal cortex. d, Relative change in mean whisker occupancy (above) and mean whisker angle (below) across 8 sessions in the same perturbed mouse. e, Summary plot of relative change in whisker occupancy in all control mice (gray, left) and all mice with diminished global L5 Ca2+ spikes (red, right). f, Summary plot of relative change in whisker angle for control (left) and perturbed (right) mice as in e

  10. Supplementary Figure 10 Selective silencing of Ca2+ spikes affects sensorimotor behavior (part 2).

    a, Mean change in whisker angle across all mice (control (black) and perturbed (red). Asterisks indicate statistical significance (two-sample Kolmogorov-Smirnov test) between the groups where p=0.0074 at session 3 and 0.013 at session 5. b, Fraction of false alarms during baseline and switch sessions in the control (black) and active L5 integration perturbed (red) mice. c, Behavioral performance (d’) during baseline and switch sessions in control (black) and perturbed (red) mice showing no drop in detection or choice performance. d, Fraction of reward trials in control and tuft silenced mice. e, Mean fraction of whisker contact trials in control (black) and tuft silenced (red) mice). f, Behavioral performance in a subset of mice (n=4) and the impact of trimming the single (C2) whisker after 6 sessions. Significant drop in performance to chance levels from trimming the C2 whisker shows that task performance was dependent on active C2 whisker contacts alone. In a-e, all symbols are mean±sem.

  11. Supplementary Figure 11 Proposed scheme for active dendritic integration contributing to mixed network representations and sensorimotor behavior.

    a, Illustration of nonlinear mixing of spatially segregated multi-modal inputs in L5 pyramidal neurons through active dendritic integration. Above, inputs carrying whisker location information are shown as received by the tuft regions of L5 pyramidal neurons through long-range synaptic inputs from sources like motor cortex (M1) or posteriomedial nuclei in the thalamus (PoM) (schematic shows an example input with location tuning of ~10°). Below, inputs carrying touch information, including the identity of whisker that contacted the object, received predominantly at the proximal dendritic regions of the L5 neuron from sources like the ventral posteriomedial thalamic nucleus (VPM) and from neuronal populations in layers 4 and 2/3 of the principle column corresponding to that whisker (schematic shows a C3 whisker representation for example). Dashed line represents weak long-range synaptic inputs from sources like motor cortex (M1) or posteriomedial nuclei in the thalamus (PoM) to proximal regions. Coincident tuned inputs to both regions (distal and proximal) during active touch by the preferred whisker (C3) at the preferred location (~10°) enables the otherwise restricted back-propagating action potential (bAP) to evoke a long-duration, Ca2+ - plateau potential in the distal dendrites. This results in a nonlinear increase in high-frequency bursts of action potentials in somatic output. Thus, the neuron generates a nonlinearly mixed selectivity to the two input modalities, whisker location (10°) and whisker touch (C3 touch). This manifests as a somatotopic map gain modulated by whisker location. b, Network impact, active dendritic integration in a population of L5 neurons results in nonlinearly mixed selectivity to various combinations of whisker angle and whisker touch. Mixed tuning to whisker identity (somatotopy) and whisker location is shown by 3-D surface plots pseudocolored to reflect neuronal output intensity (increasing from blue to red). Such a distributed population forms a high-dimensional basis network that effectively represents object location suitable for a linear read-out by various downstream areas. The resulting activity pattern (e.g. mapping to motor co-ordinates) from an example linear sum is depicted in a potential downstream neuron5. c, Illustration of the behavioral result of the neuronal and network level consequences of active dendritic integration. The sensory-motor transformation produces motor related activity in downstream regions that drives whisking patterns that are dependent on the sensory environment.

Supplementary information

About this article

Publication history

Received

Accepted

Published

DOI

https://doi.org/10.1038/s41593-018-0254-6