Single-Cell Transcriptomics of the Developing Lateral PNAS PLUS Geniculate Nucleus Reveals Insights Into Circuit Assembly and Refinement
Total Page：16
File Type：pdf, Size：1020Kb
Single-cell transcriptomics of the developing lateral PNAS PLUS geniculate nucleus reveals insights into circuit assembly and refinement Brian T. Kalisha,b,1, Lucas Cheadlea,1, Sinisa Hrvatina, M. Aurel Nagya,c, Samuel Riveraa, Megan Crowd, Jesse Gillisd, Rory Kirchnere, and Michael E. Greenberga,2 aDepartment of Neurobiology, Harvard Medical School, Boston, MA 02115; bDivision of Newborn Medicine, Department of Medicine, Boston Children’s Hospital, Boston, MA 02115; cProgram in Neuroscience, Harvard Medical School, Boston, MA 02115; dCold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724; and eBioinformatics Core, Department of Biostatistics, Harvard T.H. Chan School of Public Health, Boston, MA 02115 Contributed by Michael E. Greenberg, December 18, 2017 (sent for review October 23, 2017; reviewed by Matthew B. Dalva and Sacha B. Nelson) Coordinated changes in gene expression underlie the early pattern- the cleft to organize synaptic architecture and composition (15– ing and cell-type specification of the central nervous system. 17). The process of synaptogenesis is accompanied by myelination, However, much less is known about how such changes contribute which accrues as the circuit matures, ensuring the efficiency of to later stages of circuit assembly and refinement. In this study, we action potential propagation within both local and long-range employ single-cell RNA sequencing to develop a detailed, whole- circuits (18, 19). Finally, synapses are strengthened or eliminated transcriptome resource of gene expression across four time points in based upon a process of activity-dependent synaptic refinement the developing dorsal lateral geniculate nucleus (LGN), a visual that is in part mediated through the tagging of synapses with structure in the brain that undergoes a well-characterized program complement proteins such as C1q and C3 (20–22). Thus far, the of postnatal circuit development. This approach identifies markers molecular mechanisms of circuit assembly and refinement have defining the major LGN cell types, including excitatory relay neurons, primarily been examined on a gene-by-gene basis and have for the oligodendrocytes, astrocytes, microglia, and endothelial cells. Most most part not yet been studied at a genome-wide level. In addition, cell types exhibit significant transcriptional changes across develop- how nonneuronal cell types in the developing brain contribute ment, dynamically expressing genes involved in distinct processes to the development and plasticity of synapses requires further NEUROSCIENCE including retinotopic mapping, synaptogenesis, myelination, and investigation. synaptic refinement. Our data suggest that genes associated with The temporal, spatial, and cell-type specificities of the mo- synapse and circuit development are expressed in a larger proportion lecular code that defines synapse formation and refinement are of nonneuronal cell types than previously appreciated. Furthermore, dependent upon the active regulation of gene transcription (22). we used this single-cell expression atlas to identify the Prkcd-Cre To date, two types of experiments have primarily contributed to mouse line as a tool for selective manipulation of relay neurons our understanding of the transcriptional dynamics underly- during a late stage of sensory-driven synaptic refinement. This ing postnatal brain development: whole-tissue sequencing, and transcriptomic resource provides a cellular map of gene expression across several cell types of the LGN, and offers insight into the Significance molecular mechanisms of circuit development in the postnatal brain. Neurons and nonneuronal cells in the developing brain dy- lateral geniculate nucleus | LGN | retinogeniculate | transcriptomics | namically regulate gene expression as neural connectivity is refinement established. However, the specific gene programs activated in distinct cell populations during the assembly and refinement of he transcriptome of the developing brain is highly dynamic. many intact neuronal circuits have not been thoroughly char- TEarly embryonic stages of neural tissue induction, brain re- acterized. In this study, we take advantage of recent advances gion specification, and neuronal differentiation are all regulated, in transcriptomic profiling techniques to characterize gene ex- in part, by the expression of specific transcription factors that pression in the postnatal developing lateral geniculate nucleus coordinate unique gene programs within each cell type and tissue (LGN) at single-cell resolution. Our data reveal that genes in- (1–3). However, the cellular heterogeneity and anatomical com- volved in brain development are dynamically regulated in all plexity of the developing brain continue to present a major chal- major cell types of the LGN, suggesting that the establishment lenge to the study of how transcriptional changes contribute to the of neural connectivity depends upon functional collaboration assembly and refinement of neural circuits. Recent advances in between multiple neuronal and nonneuronal cell types in this single-cell RNA sequencing have enabled the profiling of gene brain region. expression patterns across thousands of individual cells in the mature brain, revealing a remarkable degree of previously un- Author contributions: B.T.K., L.C., and M.E.G. designed research; B.T.K., L.C., and S.R. derappreciated taxonomic diversity (4–11). However, the contri- performed research; S.H., M.A.N., M.C., J.G., and R.K. contributed new reagents/analytic tools; B.T.K., L.C., S.H., M.A.N., M.C., and R.K. analyzed data; and B.T.K., L.C., and M.E.G. bution of coordinated, cell type-specific gene expression programs wrote the paper. to postnatal brain development remains poorly understood. Reviewers: M.B.D., Thomas Jefferson University; and S.B.N., Brandeis University. The establishment and refinement of neural connectivity are The authors declare no conflict of interest. fundamental aspects of neurodevelopment. Circuit assembly is This open access article is distributed under Creative Commons Attribution-NonCommercial- initially specified by a molecular code of complementary guidance NoDerivatives License 4.0 (CC BY-NC-ND). and adhesion molecules. During this multistage process, potential Data deposition: The data reported in this paper have been deposited in the Gene Expres- synaptic partners make direct contact through molecular in- sion Omnibus (GEO) database, https://www.ncbi.nlm.nih.gov/geo (accession no. GSE108761). teractions between axon guidance cues, such as Ephrins and 1B.T.K. and L.C. contributed equally to this work. – Eph receptors (12 14). Synapses are then specified and further 2To whom correspondence should be addressed. Email: [email protected]. differentiated through physical contact between synaptogenic edu. adhesion molecules, such as Neurexins and Neuroligins, that are This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. embedded in pre- and postsynaptic membranes and bind across 1073/pnas.1717871115/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1717871115 PNAS Early Edition | 1of10 Downloaded by guest on September 28, 2021 sequencing of specific cell populations isolated through the use a quality control (QC) assessment to remove cells that may of mouse genetics or by mechanical techniques such as FACS represent doublets instead of individual cells. Specifically, we and immunopanning. The application of whole-tissue sequencing removed from the dataset cells that expressed aberrantly high or to heterogeneous brain regions such as the prefrontal cortex has low levels of genes, or cells expressing a particularly high per- shown that the transcriptome of the embryonic brain is highly centage of mitochondrial genes (SI Experimental Procedures). dynamic and becomes more stable as the brain matures (23–26). After removing these, we proceeded to analyze the remaining However, this approach has been technically limited by an in- 35,326 cells that passed QC (7,499 cells from P5, 7,596 cells from ability to disentangle cell type-specific gene expression from P10, 13,091 cells from P16, and 7,140 cells from P21). population-level changes. Furthermore, most studies have fo- Using the Seurat software package for R, we next identified cused on transcriptional changes between embryonic develop- highly variable genes by calculating the average expression and ment and adulthood, with few studies examining the critical first distribution of each gene across all cells (31). Genes with high cell- few weeks of postnatal brain development during which neuronal to-cell variability were used to perform principal component connectivity is established. analysis (PCA) and linear dimension reduction. We conducted a On the other hand, strategies of transcriptional analysis fol- semisupervised clustering analysis that included an unsupervised lowing the mechanical isolation of tissues are limited to approach followed by manual filtering. Given the multidimen- investigator-selected cell types and may not allow for the pro- sional nature of the data, we used t-distributed stochastic neighbor filing of a large number of cell types within a given sample, which embedding (t-SNE) to visualize cell clustering. Through analysis of can be useful for providing biological context and for comparing all cells across all time points, we identified 29 cell clusters with levels of a gene of interest across multiple cell types (27–29). distinct gene expression signatures with greater than 100 cells