Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Distinct descending motor cortex pathways and their roles in movement


Activity in the motor cortex predicts movements, seconds before they are initiated. This preparatory activity has been observed across cortical layers, including in descending pyramidal tract neurons in layer 5. A key question is how preparatory activity is maintained without causing movement, and is ultimately converted to a motor command to trigger appropriate movements. Here, using single-cell transcriptional profiling and axonal reconstructions, we identify two types of pyramidal tract neuron. Both types project to several targets in the basal ganglia and brainstem. One type projects to thalamic regions that connect back to motor cortex; populations of these neurons produced early preparatory activity that persisted until the movement was initiated. The second type projects to motor centres in the medulla and mainly produced late preparatory activity and motor commands. These results indicate that two types of motor cortex output neurons have specialized roles in motor control.

This is a preview of subscription content, access via your institution

Access options

Buy this article

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Taxonomy of motor cortex glutamatergic neurons based on scRNA-seq.
Fig. 2: Two types of PT neuron in the motor cortex.
Fig. 3: PT neuron type markers.
Fig. 4: Cell type-specific extracellular neurophysiology.
Fig. 5: Persistent preparatory activity in PTupper neurons.
Fig. 6: Movement commands in PTlower neurons.

Similar content being viewed by others

Data availability

The full single-cell RNA-seq dataset has been described previously21. Bulk RNA-seq data are available at the Gene Expression Omnibus (GEO) under accession GSE119182. Electrophysiology data are available at and Figshare (doi: 10.25378/janelia.7007846).


  1. Svoboda, K. & Li, N. Neural mechanisms of movement planning: motor cortex and beyond. Curr. Opin. Neurobiol. 49, 33–41 (2018).

    Article  CAS  Google Scholar 

  2. Shenoy, K. V., Sahani, M. & Churchland, M. M. Cortical control of arm movements: a dynamical systems perspective. Annu. Rev. Neurosci. 36, 337–359 (2013).

    Article  CAS  Google Scholar 

  3. Evarts, E. V. Pyramidal tract activity associated with a conditioned hand movement in the monkey. J. Neurophysiol. 29, 1011–1027 (1966).

    Article  CAS  Google Scholar 

  4. Kaufman, M. T. et al. The largest response component in the motor cortex reflects movement timing but not movement type. eNeuro 3, (2016).

    Article  Google Scholar 

  5. Kiritani, T., Wickersham, I. R., Seung, H. S. & Shepherd, G. M. G. Hierarchical connectivity and connection-specific dynamics in the corticospinal-corticostriatal microcircuit in mouse motor cortex. J. Neurosci. 32, 4992–5001 (2012).

    Article  CAS  Google Scholar 

  6. Kawaguchi, Y. Pyramidal cell subtypes and their synaptic connections in layer 5 of rat frontal cortex. Cereb. Cortex 27, 5755–5771 (2017).

    Article  Google Scholar 

  7. Hooks, B. M. et al. Organization of cortical and thalamic input to pyramidal neurons in mouse motor cortex. J. Neurosci. 33, 748–760 (2013).

    Article  CAS  Google Scholar 

  8. von Economo, C. & Parker, S. The Cytoarchitectonics of the Human Cerebral Cortex (Oxford University Press, London 1929).

    Google Scholar 

  9. Kuypers, H. G. J. M. in Comprehensive Physiology (ed. Terjung, R.) (John Wiley & Sons, Inc., New Jersey, 1981).

  10. Lemon, R. N. Descending pathways in motor control. Annu. Rev. Neurosci. 31, 195–218 (2008).

    Article  CAS  Google Scholar 

  11. Cheney, P. D. & Fetz, E. E. Functional classes of primate corticomotoneuronal cells and their relation to active force. J. Neurophysiol. 44, 773–791 (1980).

    Article  CAS  Google Scholar 

  12. Lawrence, D. G. & Kuypers, H. G. The functional organization of the motor system in the monkey. I. The effects of bilateral pyramidal lesions. Brain J. Neurol. 91, 1–14 (1968).

    Article  CAS  Google Scholar 

  13. Deschênes, M., Bourassa, J. & Pinault, D. Corticothalamic projections from layer V cells in rat are collaterals of long-range corticofugal axons. Brain Res. 664, 215–219 (1994).

    Article  Google Scholar 

  14. Kita, T. & Kita, H. The subthalamic nucleus is one of multiple innervation sites for long-range corticofugal axons: a single-axon tracing study in the rat. J. Neurosci. Off. J. Soc. Neurosci. 32, 5990–5999 (2012).

    Article  CAS  Google Scholar 

  15. Guo, Z. V. et al. Maintenance of persistent activity in a frontal thalamocortical loop. Nature 545, 181–186 (2017).

    Article  ADS  CAS  Google Scholar 

  16. Jones, E. G. & Wise, S. P. Size, laminar and columnar distribution of efferent cells in the sensory-motor cortex of monkeys. J. Comp. Neurol. 175, 391–437 (1977).

    Article  CAS  Google Scholar 

  17. Wang, X. et al. Deconstruction of corticospinal circuits for goal-directed motor Skills. Cell 171, 440–455.e14 (2017).

    Article  CAS  Google Scholar 

  18. Li, N., Chen, T.-W., Guo, Z. V., Gerfen, C. R. & Svoboda, K. A motor cortex circuit for motor planning and movement. Nature 519, 51–56 (2015).

    Article  ADS  CAS  Google Scholar 

  19. Tanji, J. & Evarts, E. V. Anticipatory activity of motor cortex neurons in relation to direction of an intended movement. J. Neurophysiol. 39, 1062–1068 (1976).

    Article  CAS  Google Scholar 

  20. Turner, R. S. & DeLong, M. R. Corticostriatal activity in primary motor cortex of the macaque. J. Neurosci. 20, 7096–7108 (2000).

    Article  CAS  Google Scholar 

  21. Tasic, B. et al. Shared and distinct transcriptomic cell types across neocortical areas. Nature (2018).

    Article  ADS  CAS  Google Scholar 

  22. Economo, M. N. et al. A platform for brain-wide imaging and reconstruction of individual neurons. eLife 5, e10566 (2016).

    Article  Google Scholar 

  23. Stanek, E., IV, Cheng, S., Takatoh, J., Han, B.-X. & Wang, F. Monosynaptic premotor circuit tracing reveals neural substrates for oro-motor coordination. eLife 3, e02511 (2014).

    Article  Google Scholar 

  24. Tervo, D. G. et al. A designer AAV variant permits efficient retrograde access to projection neurons. Neuron 92, 372–382 (2016).

    Article  CAS  Google Scholar 

  25. Guo, Z. V. et al. Flow of cortical activity underlying a tactile decision in mice. Neuron 81, 179–194 (2014).

    Article  CAS  Google Scholar 

  26. Chen, T.-W., Li, N., Daie, K. & Svoboda, K. A map of anticipatory activity in mouse motor cortex. Neuron 94, 866–879 (2017).

    Article  CAS  Google Scholar 

  27. Li, N., Daie, K., Svoboda, K. & Druckmann, S. Robust neuronal dynamics in premotor cortex during motor planning. Nature 532, 459–464 (2016).

    Article  ADS  CAS  Google Scholar 

  28. Cunningham, J. P. & Yu, B. M. Dimensionality reduction for large-scale neural recordings. Nat. Neurosci. 17, 1500–1509 (2014).

    Article  CAS  Google Scholar 

  29. Komiyama, T. et al. Learning-related fine-scale specificity imaged in motor cortex circuits of behaving mice. Nature 464, 1182–1186 (2010).

    Article  ADS  CAS  Google Scholar 

  30. Travers, J. B., Dinardo, L. A. & Karimnamazi, H. Motor and premotor mechanisms of licking. Neurosci. Biobehav. Rev. 21, 631–647 (1997).

    Article  CAS  Google Scholar 

  31. Druckmann, S. & Chklovskii, D. B. Neuronal circuits underlying persistent representations despite time varying activity. Curr. Biol. 22, 2095–2103 (2012).

    Article  CAS  Google Scholar 

  32. Kaufman, M. T., Churchland, M. M., Ryu, S. I. & Shenoy, K. V. Cortical activity in the null space: permitting preparation without movement. Nat. Neurosci. 17, 440–448 (2014).

    Article  CAS  Google Scholar 

  33. Hattox, A. M. & Nelson, S. B. Layer V neurons in mouse cortex projecting to different targets have distinct physiological properties. J. Neurophysiol. 98, 3330–3340 (2007).

    Article  Google Scholar 

  34. Catsman-Berrevoets, C. E. & Kuypers, H. G. A search for corticospinal collaterals to thalamus and mesencephalon by means of multiple retrograde fluorescent tracers in cat and rat. Brain Res. 218, 15–33 (1981).

    Article  CAS  Google Scholar 

  35. Steriade, M. & Yossif, G. Afferent and recurrent collateral influences on cortical somatosensory neurons. Exp. Neurol. 56, 334–360 (1977).

    Article  CAS  Google Scholar 

  36. Sherman, S. M. Thalamus plays a central role in ongoing cortical functioning. Nat. Neurosci. 19, 533–541 (2016).

    Article  CAS  Google Scholar 

  37. Yamawaki, N. & Shepherd, G. M. G. Synaptic circuit organization of motor corticothalamic neurons. J. Neurosci. 35, 2293–2307 (2015).

    Article  CAS  Google Scholar 

  38. Sherman, S. M. & Guillery, R. W. On the actions that one nerve cell can have on another: distinguishing “drivers” from “modulators”. Proc. Natl Acad. Sci. USA 95, 7121–7126 (1998).

    Article  ADS  CAS  Google Scholar 

  39. Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9, 559 (2008).

    Article  Google Scholar 

  40. Tasic, B. et al. Adult mouse cortical cell taxonomy revealed by single cell transcriptomics. Nat. Neurosci. 19, 335–346 (2016).

    Article  CAS  Google Scholar 

  41. Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).

    Article  Google Scholar 

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

    Article  ADS  CAS  Google Scholar 

  43. Paletzki, R. & Gerfen, C. R. Whole mouse brain image reconstruction from serial coronal sections using FIJI (ImageJ). Curr. Protoc. Neurosci. 73, 1.25.1–1.25.21 (2015).

    Article  Google Scholar 

  44. Sugino, K. et al. Molecular taxonomy of major neuronal classes in the adult mouse forebrain. Nat. Neurosci. 9, 99–107 (2006).

    Article  CAS  Google Scholar 

  45. Sugino, K. et al. Cell-type-specific repression by methyl-CpG-binding protein 2 is biased toward long genes. J. Neurosci. 34, 12877–12883 (2014).

    Article  CAS  Google Scholar 

  46. Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120 (2014).

    Article  CAS  Google Scholar 

  47. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).

    Article  CAS  Google Scholar 

  48. McCarthy, D. J., Chen, Y. & Smyth, G. K. Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation. Nucleic Acids Res. 40, 4288–4297 (2012).

    Article  CAS  Google Scholar 

  49. Guo, Z. V. et al. Procedures for behavioral experiments in head-fixed mice. PLoS One 9, e88678 (2014).

    Article  ADS  Google Scholar 

  50. Jun, J. J. et al. Real-time spike sorting platform for high-density extracellular probes with ground-truth validation and drift correction. Preprint at (2017).

  51. Towe, A. L., Patton, H. D. & Kennedy, T. T. Response properties of neurons in the pericruciate cortex of the cat following electrical stimulation of the appendages. Exp. Neurol. 10, 325–344 (1964).

    Article  CAS  Google Scholar 

  52. Aarts, E., Verhage, M., Veenvliet, J. V., Dolan, C. V. & van der Sluis, S. A solution to dependency: using multilevel analysis to accommodate nested data. Nat. Neurosci. 17, 491–496 (2014).

    Article  CAS  Google Scholar 

  53. van der Leeden, R. in Handbook of Multilevel Analysis (eds de Leeuw, J. & Meijer, E.) 401-433 (Springer, New York, 2008).

  54. Lein, E. S. et al. Genome-wide atlas of gene expression in the adult mouse brain. Nature 445, 168–176 (2007).

    Article  ADS  CAS  Google Scholar 

Download references


We thank A. Lemire and K. Aswath for single-cell sorting and bulk RNA-seq, S. Lindo for stereotaxic surgeries, D. Kao and T. Wang for help with RNA-seq analysis, D. Alcor for imaging assistance, and N. Li and H. Inagaki for help with electrophysiological recordings and for discussion. We thank M. Cembrowski, E. Bloss and F. Henry for discussion. H. Inagaki, M. Sherman, S. Romani, L. Luo, G. Shepherd and T. Wang provided comments on the manuscript. Imaging and reconstructions of neuronal morphology was performed by the Janelia MouseLight project ( This work was funded by the Howard Hughes Medical Institute, the Allen Institute for Brain Science, the NIH Brain Initiative (U01MH105982 to H.Z.), and the Intramural Research Program of the NIMH (ZIA-MH002497-29 to C.R.G).

Reviewer information

Nature thanks P. Carninci, M. Fee, T. Mrsic-Flogel and the anonymous reviewer(s) for their contribution to the peer review of this work.

Author information

Authors and Affiliations



M.N.E., K.S., S.V., L.L.L., B.T. and H.Z.: conception and design of the experiments. B.T., L.T.G., T.N.N., K.A.S., Z.Y. and H.Z.: scRNA-seq experiments, with material provided by M.N.E., S.V. and L.W. M.N.E., E.B., J.W. and J.C.: single neuron reconstructions. S.V., C.R.G. and M.N.E: in situ experiments and analysis. M.N.E.: electrophysiology experiments. M.N.E., S.V. and C.R.G.: anatomical experiments. M.N.E., S.V., V.M., L.T.G., Z.Y., K.S. and L.L.L.: data analysis. M.N.E., K.S. and L.L.L.: wrote the paper with input from all authors.

Corresponding author

Correspondence to Karel Svoboda.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

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

Extended data figures and tables

Extended Data Fig. 1 Differentially expressed genes in single-cell and bulk RNA-seq.

a, b, Heat map of expression (from scRNA-seq) of differentially expressed genes identified from scRNA-seq (a) and bulk RNA-seq (b). Columns represent individual cells, grouped by transcriptomic cluster (indicated below each colour map; Slco2a1, n = 203; Npsr1, n = 43; Hpgd, n = 122) and retrograde labelling site (indicated above). Grey indicates cells isolated from transgenic lines or other targets. Rows represent genes differentially expressed between the Slco2a1 and Hpgd–Npsr1 clusters. Colour bar shows transcript expression in counts per million mapped reads (CPM + 1) on a log-scale. c, d, Heat map of expression (based on bulk RNA-seq) of differentially expressed genes identified from scRNA-seq (c) and bulk RNA-seq (d) datasets. Rows represent genes, coloured by differential bulk RNA-seq expression between thalamus-labelled and medulla-labelled PT neurons. Columns in the heat map represent individual replicates (six each for PTupper and PTlower). Colours show log-transformed transcript intensity (CPM + 1) in z-scored units. Blue shows replicates with low expression (z-score = −2); red shows replicates with high expression (z-score = +2).

Extended Data Fig. 2 Distribution of PT axon collaterals.

a, Axonal lengths of single-neuron reconstructions within targets in thalamus and medulla (thalamus-targeting PT neurons, n = 8, green; medulla-targeting PT neurons, n = 4, magenta). b, Axonal lengths within other selected PT targets. c, Axonal termini in single-neuron reconstructions within thalamic and medullary targets. d, Axonal termini within other selected PT targets. e, Bulk projections of PTupper and PTlower populations. Groups of cortical neurons were labelled from the thalamus (PTupper; mouse #1), the medulla (PTlower; mouse #2), or both targets (mouse #3) using rAAV2-retro expressing spectrally distinct fluorescent proteins. Top, schematics of the labelling procedures. Left, rostro-caudal level (relative to Bregma). Right, annotated coronal sections taken from the Allen Mouse Brain Atlas54 with imaged area indicated. Both cell types extended axon collaterals to motor-related superior colliculus, but to different parts. Axons from PTupper cells were apparent throughout all superior colliculus layers, with a dense projection to the ventrolateral aspect; PTlower neurons were restricted to the ventral superior colliculus and were concentrated more caudally. Both groups project to the pons, particularly the pontine grey, but with terminations in largely non-overlapping zones. PTupper cells projected to the globus pallidus external segment and broadly targeted the dorsal, lateral and ventral striatum. PTlower cells projected sparsely to the lateral striatum. PTlower neurons projected to the central amygdala and parasubthalamic nucleus. PTlower neurons also made up most of the projection to the red nucleus, parabrachial nucleus, substantia nigra pars compacta, motor and sensory trigeminal nuclei in the hindbrain, and through the medullary pyramids. Both cell types extended axon collaterals locally within the same sublamina as their somata, the subthalamic nucleus, zona incerta, and the midbrain reticular nucleus. PTupper cells project more broadly to layer 1 in motor cortex. Mice 1 and 2 were used for electrophysiological recordings. For this reason, projections were labelled with ChR2(H134R)–YFP.

Extended Data Fig. 3 Spatial distribution of thalamus- and medulla-projecting PT neurons.

a, The nuclei of PT neurons were retrogradely labelled from the thalamus (green cells) and medulla (magenta cells) using rAAV2-retro. Thalamus-projecting PT neurons are in upper L5b throughout motor cortex, whereas medulla-projecting PT neurons are in deep L5b. Schematics to the left of each image set are annotated coronal sections (Allen Mouse Brain Atlas54). b, rAAV2-retro injection sites in the thalamus (left) and medulla (right). Three biological replicates of this experiment yielded similar results.

Extended Data Fig. 4 Correspondence between scRNA-seq transcriptomic clusters, projection targets, and bulk RNA-seq data.

a, Pairwise correlations (Pearson’s r) in gene expression between all cells in the Slco2a1 cluster and within subsets of Slco2a1 cells identified by projection target. Expression patterns of cells identified from a common target were not more similar than randomly chosen cells. b, As in a for the Hpgd expression cluster. c, Pairwise correlation in gene expression of cells retrogradely labelled from the superior colliculus within the Slco2a1 cluster, within the Hpgd cluster, and between cells from different clusters. Correlations between cells from different clusters were significantly lower than within-cluster correlations (between Slco2a1, superior colliculus and Hpgd, superior colliculus (n = 1,677 pairs) versus within Slco2a1, superior colliculus (n = 903 pairs): P < 1 × 10−10; between Slco2a1, superior colliculus and Hpgd, superior colliculus (n = 1,677 pairs) versus within Hpgd, superior colliculus (n = 741 pairs): P < 1 × 10−10; two-sided Wilcoxon signed rank test). Correlation analysis was not performed for the Npsr1 cluster as the number of cells was relatively small (Fig. 2e). d, Number of differentially expressed genes in bulk RNA-seq data between groups of cells labelled from different projection targets. Notably, only a single gene was identified as differentially expressed between the thalamus/medulla group and the superior colliculus/pons group, indicating that the set of PT neurons projecting to either the thalamus or medulla probably represents a superset of PT neurons. There is unlikely to be a transcriptomically distinct group of PT neurons with projections to the superior colliculus or pons that lacks projections to the medulla and thalamus. Although we cannot rule out a transcriptomically distinct subset that lacks projections to all of the thalamus, medulla, pons and superior colliculus, such neurons were not detected in single-cell reconstructions. e, Total genes detected by bulk RNA-seq and scRNA-seq in PTupper neurons. Both: genes detected in bulk RNA-seq and scRNA-seq; single-cell: additional genes detected only in scRNA-seq; bulk: additional genes detected only in bulk RNA-seq. f, As in e for PTlower neurons. g, Total number of genes detected across all experiments. In eg, single cell reads were downsampled such that total read depth was the same for scRNA-seq and bulk RNA-seq. h, Number of genes detected in scRNA-seq for each PTupper and PTlower neuron (‘X’, median; PTupper, 9,936 genes; PTlower, 9,865 genes). i, Mean fold change in expression (measured as log2(CPM + 1)) of all genes detected by both methods between neurons in the Npsr1–Hpgd clusters (PTupper) and neurons in the Slco2a1 cluster (PTlower) as determined by scRNA-seq (x axis) and bulk RNA-seq (y axis). Colour represents classification accuracy between the Npsr1–Hpgd and Slco2a1 clusters using a binary (detected/not detected) version of the scRNA-seq data. The 100 most discriminative neurons are coloured for each type. The correlation coefficient (Pearson’s r) in fold change expression was 0.87 for this set of 200 differentially expressed genes.

Extended Data Fig. 5 Electrophysiology and trial-averaged spike rates for identified PT neurons.

a, PTupper neurons (n = 61). Left, inter-spike interval histograms; right: trial-averaged activity on lick right (blue) and lick left trials (red). Grey shaded area in inter-spike interval histograms represents the interval of −2.5 ms to 2.5 ms. b, PTlower neurons (n = 69). Boxed region indicates neurons recorded in the right ALM (ipsilateral to injection site in medulla). All other neurons were recorded in the left ALM (contralateral to injection site). c, Bursting cells (cells in which greater than 10% of inter-spike intervals were less than 5 ms) were rare in the PTupper population (3.3%) and more common in the PTlower population (18.8%; P = 0.006, Fisher’s exact test).

Extended Data Fig. 6 Variance in trial-averaged activity explained by PT cell class.

a, Green line: variance of trial-averaged activity explained by increasing numbers of principal components across the population of PTupper neurons (n = 61). Dark green line and 95% confidence interval: expected variance explained by the same number of components for size-matched samples (n = 1,000 repetitions) of simultaneously recorded, unidentified neurons. b, Ratio of variance explained between PTupper neurons and bootstrapped distribution of simultaneously recorded unidentified neurons. Error bars represent s.d. of distribution. c, d, As in a and b, but for PTlower neurons (n = 69). Asterisks denote points significantly greater than unity (P < 0.05, bootstrap).

Extended Data Fig. 7 Late delay epoch coding direction and similarity with other trial epochs.

a, Time course of the linear combination of neuronal activity that best differentiates trial types in the 400 ms immediately before the go cue (late coding direction; CDlate) on lick right (blue) and lick left (red) trials for PTupper (top; n = 61) and PTlower (bottom; n = 69) neurons. b, Difference in CDlate projections on lick right and lick left trials (selectivity) in each population. Selectivity along CDlate is present in both populations, and persists after the go cue, but is not strongly modulated during movement initiation. Shaded regions represent the s.d. of the distribution produced by hierarchical bootstrapping (n = 1,000 iterations) in a and 5–95% confidence intervals in b (denoting region significantly greater than zero, P < 0.05 one-sided test, bootstrap). c, A coding direction (CD) was calculated at all individual time points. Heat maps represent the correlation (inner product) of the CD between pairs of time points. In PTupper neurons (top), the coding direction remained similar across the sample and delay epochs. In PTlower neurons (bottom), coding directions in the delay epoch were largely orthogonal to coding directions calculated in the sample epoch. The upcoming movement direction is encoded in a persistent manner in the PTupper population, but not the PTlower population. Right, expanded view of the change in coding direction around the time of the go cue. An abrupt change in the coding direction occurs immediately after the go cue onset in the PTlower population. A change also occurs in the PTupper population, but more slowly (over several hundred milliseconds), largely after initiation of movement. d, Correlation (Pearson’s r) between CDearly vector weights and CDlate vector weights for PTupper neurons (top) and PTlower neurons (bottom). e, Correlation between the CD and CDlate normalized to the mean correlation in the 400 ms preceding the go cue. A rapid change in the CD occurs in the PTlower population following the go cue. Shaded areas represent the s.d. of the bootstrapped distribution (n = 1,000 iterations).

Extended Data Fig. 8 Comparison between identified PT populations and simultaneously recorded untagged neurons.

Time course of selectivity along CDearly (as in Fig. 5b), CDgo (as in Fig. 6b), and CDgo aligned to the last lick in each trial (as in Extended Data Fig. 10b) calculated as follows: (1) from PTupper (n = 61) and PTlower (n = 69) neurons (top row), (2) from simultaneously recorded but unidentified neurons (second row; n = 495 from PTupper cohort; n = 511 simultaneously recorded with PTlower neurons), (3) after removing unidentified neurons from the PTlower experiments recorded from the contralateral hemisphere (third row; n = 276 remaining), which were recorded at a lower average depth in a different recording configuration, and (4) from PTupper and PTlower neurons with firing rates z-scored based on their firing rates in the epoch preceding the stimulus (bottom row). CD projections in the populations of unidentified neurons were similar and nearly indistinguishable after they were approximately depth-matched. Shaded regions represent 5–95% confidence intervals in bootstrapped distribution as in all other figures (n = 1,000 iterations; denotes region significantly greater than zero, P < 0.05 one-sided test, bootstrap). b, Correlation of CD with CDlate (inner product; as in Extended Data Fig. 7e) around the time of the go cue in each PT population (left), in simultaneously recorded unidentified populations (middle) and approximately depth matched populations (right). The change in correlations around the time of the go cue in the unidentified populations were similar and intermediate between that observed in the PTupper population and PTlower populations. c, Early trial type decoding (as in Extended Data Fig. 9) in each PT population (left) and the simultaneously recorded populations of unidentified neurons (right).

Extended Data Fig. 9 Decoding of trial type in PT neuron types.

a, Accuracy of trial type classification by single neurons in the 400 ms immediately after stimulus onset. 24.6% (15 out of 61) of PTupper neurons predicted trial type with at least 70% accuracy, whereas only 4.4% (3 out of 69) of PTlower neurons did so. Mean accuracy was also significantly higher in PTupper neurons (PTupper: 64.4 ± 1.0%; PTlower: 58.9 ± 0.6%, mean ± s.e.m.; P = 1 × 10−4, two-sided Mann–Whitney test). The ten most discriminative neurons all belonged to the PTupper population. b, Cumulative distribution function of the data in a. c, d, As in a and b but decoding only based on spike rates rectified at baseline. Trial-type selectivity during the sample epoch in PTlower neurons was predominantly characterized by a modest suppression of spiking on one trial type, probably reflecting widespread lateral inhibition. Disregarding spike rate changes below baseline, no PTlower neurons predicted trial type with at least 70% accuracy, whereas the same 24.6% of PTupper neurons continued to do so and accounted for 20 out of 21 of the most predictive neurons (PTupper: 62.7 ± 1.1%; PTlower: 56.7 ± 0.4%, mean ± s.e.m.; P = 9 × 10−7, two-sided Mann–Whitney test). As soon as the trial type is cued by the stimulus, upcoming movement direction is encoded robustly in a subset of PTupper cells and only minimally in PTlower cells.

Extended Data Fig. 10 Movement termination signals in PTlower neurons.

a, Selectivity along CDlate (as in Extended Data Fig. 7b) for PTupper (green, left; n = 61) and PTlower neurons (magenta, right; n = 69) aligned to the last lick in the response epoch. b, Selectivity along CDgo (same as Fig. 6b) aligned to the last lick for each PT type. Shaded regions in a and b represent 5–95% confidence intervals around the mean using hierarchical bootstrapping (n = 1,000 iterations; denoting region significantly greater than zero, P < 0.05 one-sided test, bootstrap). c, Correlation of coding direction weights at all pairs of time points after the go cue for PTupper neurons (left) and PTlower neurons (right) using last-lick aligned spike rates. An additional transition in the population dynamics accompanies the termination of movement in PTlower neurons, whereas there is no correlate of movement termination in PTupper neurons. The change in dynamics at the offset of movement was less abrupt than at movement onset, probably a result of aligning data to the last lick-port contact, which does not precisely mark the cessation of movement. d, Spike raster plots (top) and trial-averaged activity (bottom) for four example PTlower neurons aligned to the go cue (lick right: blue; lick left: red) and the last lick-port contact (lick right: dark blue; lick left: dark red).

Supplementary information

Supplementary Tables

This file contains Supplementary tables 1-4.

Reporting Summary

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

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

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


This article is cited by


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.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing