Transcription factor expression defines subclasses of developing projection neurons highly similar to single-cell RNA-seq subtypes Whitney E. Heavnera,b, Shaoyi Jib, James H. Notwellc, Ethan S. Dyerd,e,f, Alex M. Tsengc, Johannes Birgmeierc, Boyoung Yooc, Gill Bejeranoc,g,h,i, and Susan K. McConnellb,1 aCenter for Integrative Brain Research, Seattle Children’s Research Institute, Seattle, WA 98101; bDepartment of Biology, Stanford University, Stanford, CA 94305; cDepartment of Computer Science, Stanford University, Stanford, CA 94305; dStanford Institute for Theoretical Physics, Stanford University, Stanford, CA 94305; eDepartment of Physics and Astronomy, The Johns Hopkins University, Baltimore, MD 21218; fGoogle, Mountain View, CA 94043; gDepartment of Developmental Biology, Stanford University, Stanford, CA 94305; hDepartment of Pediatrics, Stanford University School of Medicine, Stanford University, Stanford, CA 94305; and iDepartment of Biomedical Data Science, Stanford University, Stanford, CA 94305 Contributed by Susan K. McConnell, August 12, 2020 (sent for review April 27, 2020; reviewed by Anthony Paul Barnes and Seth Blackshaw) We are only just beginning to catalog the vast diversity of cell clustering and visualization tools, such as t-distributed stochastic types in the cerebral cortex. Such categorization is a first step neighbor embedding (t-SNE) (8) and uniform manifold ap- toward understanding how diversification relates to function. All proximation and projection (UMAP) (9). Default t-SNE, how- cortical projection neurons arise from a uniform pool of progenitor ever, does not preserve global structure and therefore may miss cells that lines the ventricles of the forebrain. It is still unclear how biologically meaningful hierarchies in single-cell data (10). these progenitor cells generate the more than 50 unique types of UMAP, on the other hand, was shown to preserve more local mature cortical projection neurons defined by their distinct gene- structure and continuity between cell subtypes than t-SNE, sug- expression profiles. Moreover, exactly how and when neurons di- gesting that UMAP improves on the utility of t-SNE for yielding versify their function during development is unknown. Here we insights into cell fate trajectories (9). A single cortical progenitor relate gene expression and chromatin accessibility of two sub- cell, for example, generates many daughter cells, each of which classes of projection neurons with divergent morphological and follows a specific trajectory to become a highly specialized cell functional features as they develop in the mouse brain between type among a larger population with similar features. The cere- embryonic day 13 and postnatal day 5 in order to identify tran- bral cortex can be divided into two such large populations: scriptional networks that diversify neuron cell fate. We compare Projection neurons of the deep layers (DL), layer 6 (L6) and L5, these gene-expression profiles with published profiles of single cells isolated from similar populations and establish that layer- which are populated first, followed in order by projection neu- defined cell classes encompass cell subtypes and developmental rons of the upper layers (UL), L4, then L2/3 (11). Although DL trajectories identified using single-cell sequencing. Given the and UL subclasses are heterogeneous, cells within each subclass depth of our sequencing, we identify groups of transcription fac- share common morphological and electrophysiological proper- tors with particularly dense subclass-specific regulation and ties. The majority of DL neurons, for example, are corticofugal subclass-enriched transcription factor binding motifs. We also de- projection neurons (CFPNs), which send their axons to areas scribe transcription factor-adjacent long noncoding RNAs that de- outside the cortex, including the thalamus and the spinal cord. fine each subclass and validate the function of Myt1l in balancing The majority of UL neurons are cortico–cortical projection the ratio of the two subclasses in vitro. Our multidimensional ap- proach supports an evolving model of progressive restriction of Significance cell fate competence through inherited transcriptional identities. Neurons of the cerebral cortex arise from a common progenitor transcription | gene regulation | cortical development | pool that progressively generates different cell subtypes with next-generation sequencing distinct features, including deep layer (DL) and upper layer (UL) projection neurons. Several transcription factors (TFs) are he cerebral cortex is the region of the human brain respon- known to specify whether a progenitor cell will become UL or Tsible for perception, language, complex thinking, and motor DL; however, it is still unknown which upstream and down- control. Neurons in the cortex can be subdivided into two broad stream factors contribute to this fate decision. Here, we deeply classes: Excitatory glutamatergic projection neurons and inhibi- characterize differential gene expression and chromatin ac- tory GABAergic interneurons. Subclasses of cells exist within cessibility of UL and DL neurons. We identify TFs that are highly each of these broad classes. Recent efforts have sought to differentially expressed and densely regulated and find that UL identify the complete catalog of cell types in the cortex using neurons retain the transcriptomic signature of their mother transcriptional profiling (1–4). At least 13 unique types of ex- cells. We also discover a previously unknown role for the TF citatory neurons have been identified in the developing mouse Myt1l in cell fate specification. cortex (2), while the adult mouse cortex contains at least 52, a Author contributions: W.E.H., S.J., J.H.N., G.B., and S.K.M. designed research; W.E.H. and number that may rise with increased computational power to S.J. performed research; W.E.H., S.J., J.H.N., E.S.D., A.M.T., J.B., B.Y., G.B., and S.K.M. an- delineate cell types among single cells (1, 3). alyzed data; and W.E.H., S.J., J.H.N., E.S.D., A.M.T., J.B., G.B., and S.K.M. wrote the paper. While our ability to identify neuron subtypes contributes to Reviewers: A.P.B., Oregon Health & Science University; and S.B., The Johns Hopkins Uni- our understanding of functional divisions within cortical circuits versity School of Medicine. (5), how these circuits are specified during development remains The authors declare no competing interest. unclear. Several recent studies have used single-cell RNA se- Published under the PNAS license. quencing (scRNA-seq) to identify how differential gene expres- 1To whom correspondence may be addressed. Email: [email protected]. sion over time establishes all of the identifiable cell subtypes in This article contains supporting information online at https://www.pnas.org/lookup/suppl/ the developing brain (2, 6, 7). Delineating cell types using doi:10.1073/pnas.2008013117/-/DCSupplemental. scRNA-seq profiles relies on dimensionality reduction with First published September 18, 2020. 25074–25084 | PNAS | October 6, 2020 | vol. 117 | no. 40 www.pnas.org/cgi/doi/10.1073/pnas.2008013117 Downloaded by guest on September 26, 2021 neurons (CPNs), many of which send their axons across the midline along the corpus callosum. While CPNs overwhelmingly occupy the ULs, they can also be found in the DLs. A network of key transcription factors (TFs) expressed in early projection neurons determines whether a cell will acquire CFPN or CPN fate (12–14). Previous studies in the mouse revealed that TBR1 and FEZF2 regulate L6 and L5 CFPN development, re- spectively (15–22), while SATB2 is required for CPN fate (23–25). The upstream and downstream transcription cascade governing layer-specific maturation over time, however, is still unknown. No single study has directly assessed chromatin accessibility—an important indicator of gene regulation—and gene expression in CPNs or CFPNs during a specific stage of embryonic or early postnatal development. Moreover, it is still unknown how such transcriptional profiles differ between CPNs and CFPNs at comparable stages of development. One previous screen used bulk RNA-seq to identify genes that are differen- tially expressed between projection neuron subtypes on the same day of development in the mouse, when DL neurons are more developmentally mature than UL neurons (26). Here, in order to identify transcriptional states that determine DL versus UL fate, we chose sequencing depth over sequencing breadth and analyzed cell type-specific gene expression plus differential chromatin accessibility over three defined stages of development in two sets of genetically defined neurons. We compared subclass- and stage-defining gene sets with subtype- defining gene sets identified using scRNA-seq and found broad agreement between our bulk-sequenced neuron populations and populations identified by scRNA-seq. Given the robustness of NEUROSCIENCE bulk sequencing, we also joined open chromatin with gene ex- pression within these subclasses and identified densely regulated TFs with previously uncharacterized roles in specifying UL ver- Fig. 1. Excitatory projection neurons of the mouse neocortex can be sub- sus DL fate. In addition, we identified long-noncoding RNAs divided into two broad classes, here called upper layer (UL) and deep layer adjacent to important TF genes (TF-lncRNAs) with highly spe- (DL), each identified by a specific combination of TFs and labeled by an exclusive fluorescent reporter. (A) Schematic representation of neocortical cific spatial