Changes in genome architecture and transcriptional dynamics progress independently of sensory experience during post-natal brain development

•High-resolution single-cell transcriptome and 3D genome atlas of developing mouse brain•3D “structure types” behind transcriptional cell types•Major transformation after birth, unaffected by experience•Allele-specific structure imprinted genes, including one that spans a chromosome Both transcription three-dimensional (3D) architecture the mammalian play critical roles in neurodevelopment its disorders. However, structures single brain cells have not been solved; little is known about dynamics birth. Here, we generated (3,517 cells) (3,646 cortex hippocampus using our high-resolution multiple annealing looping-based amplification cycles for digital transcriptomics (MALBAC-DT) diploid chromatin conformation capture (Dip-C) methods multi-omic analysis pipelines. In adults, delineate all major types, with high correlation between A/B compartments gene expression. During development, both are extensively transformed first post-natal month. neurons, rewired across scales, correlated expression modules, independent sensory experience. Finally, examine allele-specific revealing local (chr)-wide differences. These findings uncover an unknown dimension neurodevelopment. Two intimately related dimensions genome—gene architecture—are crucial development nervous systems. The dynamic interplay cell-type-specific expression, structure, experience, other factors (e.g., epigenetic marks) underlies brain’s immense plasticity functions. Dysregulation leads to debilitating neurodevelopmental disorders, such as autism (Satterstrom et al., 2020Satterstrom F.K. Kosmicki J.A. Wang J. Breen M.S. De Rubeis S. An J.Y. Peng M. Collins R. Grove Klei L. al.Autism Sequencing ConsortiumiPSYCH-Broad ConsortiumLarge-Scale Exome Study Implicates Developmental Functional Changes Neurobiology Autism.Cell. 2020; 180: 568-584.e23Abstract Full Text PDF PubMed Scopus (424) Google Scholar) schizophrenia (Howrigan 2020Howrigan D.P. Rose S.A. Samocha K.E. Fromer Cerrato F. Chen W.J. Churchhouse C. Chambert K. Chandler S.D. Daly M.J. al.Exome sequencing schizophrenia-affected parent-offspring trios reveals risk conferred protein-coding de novo mutations.Nat. Neurosci. 23: 185-193Crossref (39) Scholar). Genome topology also determines distribution somatic DNA damage mutations human neurons (Zhu 2019Zhu Q. Niu Y. Gundry Sheng Zong Topologically Dependent Abundance Spontaneous Damage Single Human Cells.bioRxiv. 2019; https://doi.org/10.1101/859686Crossref (0) Recent progress has made characterize brain; however, existing studies 3 limitations. First, resolved. For example, bulk (3C/Hi-C) assays can only measure ensemble-averaged maps genome. They lump together diverse sub-types (and/or glia), masking features (Lu 2020Lu Liu X. Huang W.K. Giusti-Rodríguez P. Cui Zhang Xu W. Wen Z. Ma Rosen J.D. al.Robust Hi-C Maps Enhancer-Promoter Interactions Reveal Function Non-coding Neural Development Diseases.Mol. Cell. 79: 521-534.e15Abstract (28) Scholar; Won 2016Won H. la Torre-Ubieta Stein J.L. Parikshak N.N. Opland C.K. Gandal Sutton G.J. Hormozdiari Lu D. al.Chromosome elucidates regulatory relationships brain.Nature. 2016; 538: 523-527Crossref (252) So far, revealed refolding either differentiated vitro (Bonev 2017Bonev B. Mendelson Cohen N. Szabo Fritsch Papadopoulos G.L. Lubling Lv Hugnot J.P. Tanay A. Cavalli G. Multiscale Rewiring during Mouse Development.Cell. 2017; 171: 557-572.e24Abstract (462) Rajarajan 2018Rajarajan Borrman T. Liao Schrode Flaherty E. Casiño Powell Yashaswini LaMarca E.A. Kassim al.Neuron-specific signatures chromosomal connectome associated risk.Science. 2018; 362: eaat4311Crossref (76) or isolated at time point from embryonic Single-cell 3C/Hi-C achieved adult brains (Lee 2019Lee D.S. Luo Zhou Chandran Rivkin Bartlett Nery J.R. Fitzpatrick O’Connor Dixon Ecker Simultaneous profiling methylation cells.Nat. Methods. 16: 999-1006Crossref (53) but had low resolution, yielded no structures, could distinguish different neuronal based on structural information alone. Second, comprehensively studied brain. Although diversity characterized (Di Bella 2020Di D.J. Habibi Yang S.-M. Stickels R.R. Brown Yadollahpour Macosko E.Z. Regev Arlotta Molecular Logic Cellular Diversification Mammalian Cerebral Cortex.bioRxiv. https://doi.org/10.1101/2020.07.02.185439Crossref La Manno 2020La Siletti Furlan Gyllborg Vinsland Langseth C.M. Khven I. Johnsson Nilsson Lönnerberg al.Molecular brain.bioRxiv. https://doi.org/10.1101/2020.07.02.184051Crossref Lake 2016Lake B.B. Ai Kaeser G.E. Salathia N.S. Yung Y.C. Wildberg Gao Fung H.L. al.Neuronal subtypes single-nucleus RNA brain.Science. 352: 1586-1590Crossref (403) Tasic 2016Tasic Menon V. Nguyen T.N. Kim T.K. Jarsky Yao Levi Gray L.T. Sorensen Dolbeare al.Adult cortical taxonomy transcriptomics.Nat. 19: 335-346Crossref (763) Scholar, 2018Tasic Graybuck Smith K.A. Bertagnolli Goldy Garren Economo M.N. Viswanathan al.Shared distinct transcriptomic types neocortical areas.Nature. 563: 72-78Crossref (442) Zeisel 2018Zeisel Hochgerner Memic van der Zwan Häring Braun Borm L.E. Architecture Nervous System.Cell. 174: 999-1014.e22Abstract Scholar), few probed dynamics, each limited set points sampled (Carter 2018Carter R.A. Bihannic Rosencrance Hadley Tong Phoenix Natarajan Easton Northcott P.A. Gawad A Single-Cell Transcriptional Atlas Developing Murine Cerebellum.Curr. Biol. 28: 2910-2920.e2Abstract Kalish 2020Kalish B.T. Barkat T.R. Diel E.E. E.J. Greenberg M.E. Hensch Single-nucleus auditory period triggers brakes.Proc. Natl. Acad. Sci. USA. 117: 11744-11752Crossref (14) 2020Kim D.W. Washington P.W. Z.Q. Lin S.H. Sun Ismail Jiang Blackshaw cellular molecular landscape hypothalamic patterning differentiation late postnatal development.Nat. Commun. 11: 4360Crossref (18) Li 2018Li Santpere Imamura Kawasawa Evgrafov O.V. Gulden F.O. Pochareddy Sunkin S.M. Shin Zhu al.BrainSpan ConsortiumPsychENCODE SubgroupIntegrative functional genomic neuropsychiatric risks.Science. 362Crossref (206) Rosenberg 2018Rosenberg A.B. Roco Muscat Kuchina Sample Peeler Mukherjee al.Single-cell spinal cord split-pool barcoding.Science. 360: 176-182Crossref (400) Tiklová 2019Tiklová Björklund A.K. Lahti Fiorenzano Nolbrant Gillberg Volakakis Yokota Hilscher M.M. Hauling midbrain dopamine neuron emerging 10: 581Crossref (70) Furthermore, most these used low-sensitivity methods, detecting transcripts per cell. Third, offer in-depth, integrative data conjunction omic datasets. above-mentioned reasons, neither address true relationship their underlying nor trace vivo. fill this long-standing technological knowledge gap. Using highly sensitive accurate method (Chapman 2020Chapman A.R. Lee D.F. Cai Xie X.S. Correlated Gene Modules Uncovered Transcriptomics High Detectability Accuracy.bioRxiv. https://doi.org/10.1101/2019.12.31.892190Crossref created forebrain, encompassing ?3,500 2 regions 7 points. streamlined version (Tan 2018Tan Xing Chang C.H. Three-dimensional cells.Science. 361: 924-928Crossref (148) Tan 2019Tan Daley visual olfactory systems.Nat. Struct. Mol. 26: 297-307Crossref (33) atlas, ?2,000 contact ?800 6 We further analyzed sensory-deprived mice, ?1,700 5 developed visualization methods. directly addressed several fundamental questions developmental biology. On side, observed around day (P) 14 main lineages: astrocytes, oligodendrocytes. primary source variation was developmentally regulated, modules. found cell-type identity encoded wiring genome, because alone separate into 13 types.” Between differential compartmentalization well P7 P28 lineages. emerged transition; neonatal contrast, adopted more primordial state resembled cultured/differentiated vitro. gained mechanistic insights transformation. large-scale, neuron-specific radial reconfiguration, which be linked global non-CpG (Lister 2013Lister Mukamel Urich Puddifoot C.A. Johnson N.D. Lucero Dwork A.J. Schultz M.D. al.Global epigenomic reconfiguration development.Science. 2013; 341: 1237905Crossref (1158) shared striking similarity remodeling peripheral system (Bashkirova 2020Bashkirova Monahan Campbell C.E. Osinski J.M. Schieren Barnea Gronostajski R.M. Lomvardas Homeotic Regulation Olfactory Receptor Choice via NFI-dependent Heterochromatic Silencing Genomic Compartmentalization.bioRxiv. https://doi.org/10.1101/2020.08.30.274035Crossref Clowney 2012Clowney LeGros M.A. Mosley C.P. F.G. Markenskoff-Papadimitriou E.C. Myllys Larabell Nuclear aggregation receptor genes governs monogenic expression.Cell. 2012; 151: 724-737Abstract (211) 2019Monahan Horta LHX2- LDB1-mediated trans interactions regulate choice.Nature. 565: 448-453Crossref (105) long-range rewiring and/or disease-implicated enhancers. (scA/B) regulated although many displayed discordant temporally shifted changes structure. deprivation effect cortex, suggesting genetically predetermined. report genome-wide, survey genes. parent-of-origin-specific least 29 loci (Perez 2015Perez Rubinstein Fernandez D.E. Santoro S.W. Needleman L.A. Ho-Shing O. Choi J.J. Zirlinger S.K. J.S. Dulac Quantitative interrogation parent-of-origin allelic biases brain.eLife. 2015; 4: e07860Crossref (41) extreme case, difference seen extending tens Mbs Prader-Willi/Angelman syndrome (PWS/AS) locus. performed MALBAC-DT 3,517 ages: P1, P7, P14, P21, P28, P56, P180 (Figures 1A 1B , top). Dense sampling ensured weekly resolution month—a cognitive plasticity. To minimize stress (Lacar 2016Lacar Linker S.B. Jaeger B.N. Krishnaswami S.R. Barron Kelder M.J.E. Parylak S.L. Paquola A.C.M. Venepally Novotny al.Nuclear RNA-seq activation.Nat. 7: 11022Crossref (179) were rapidly dissected tissues individual nuclei (Krishnaswami 2016Krishnaswami Grindberg R.V. Lacar Bhutani Pham Erwin Miller al.Using postmortem neurons.Nat. Protoc. 499-524Crossref (158) (STAR methods). Note still referred them “cells.” allowed us sensitivity accuracy detected average 12.8 k mRNA molecules (also unique molecule identifiers [UMIs]; SD = 8.4 k, minimum 1.7 maximum 49.8 k) 3.4 (SD 1.3 1.0 7.8 This order magnitude higher than previous mice (Kalish visualized uniform manifold approximation projection (UMAP) (Hafemeister Satija, 2019Hafemeister Satija Normalization variance stabilization regularized negative binomial regression.Genome 20: 296Crossref (422) Stuart 2019Stuart Butler Hoffman Hafemeister Papalexi Mauck 3rd, W.M. Hao Stoeckius Smibert Comprehensive Integration Data.Cell. 177: 1888-1902.e1821Abstract (2364) Cells clearly separated clusters, age tissue patterns (Figure 1C). clusters “transcriptome later sections. Louvain clustering identified 4 types: oligodendrocytes, (“microglia, etc.”: microglia small number vascular 1C, left; Figures 1D 1E, top), expressing specific marker S1; Table S1). question biology what extent continue mature profile. especially important brain—most become postmitotic point, must last lifetime. UMAP plot, within type, segregated primarily age, clearest distinction P14 middle). suggested examined oligodendrocytes astrocytes. Oligodendrocytes sub-types: oligodendrocyte progenitors (Pdgfra+), newly formed (Tns3+), (Mal+) left), consistent (Cahoy 2008Cahoy Emery Kaushal Foo L.C. Zamanian Christopherson K.S. Lubischer Krieg Krupenko al.A database oligodendrocytes: new resource understanding function.J. 2008; 264-278Crossref (2078) Starting type (oligodendrocyte progenitors) led emergence over time: P21 bottom). Similarly, astrocytes (“adult astrocytes”; Gjb6+) addition (Tnc+) middle), study now focus neurons. Different excitatory inhibitory carry out computations. Neurons regions, stages, exhibit properties. investigate basis diversity, separately 2,277 complex according ages, functions 2A). sub-types, prominent gradient seen—usually separation 2A, second right). classified 20 Figure 2B). merged stages same sub-type when possible, leading 16 immature types. Among 8 excitatory: pyramidal layers (L2/3: Camk2d+; L4: Rorb+; L5: Tshz2+; L6: Tle4+, among sub-cluster Ctgf+ L6b visible), hippocampal cornu ammonis (CA) region (CA1: Gm2115+; CA3: Grik4+), granule dentate gyrus (DG) (Prox1+), Cajal-Retzius (CR) (Nhlh2+) 1B; inhibitory: interneuron (SST: Sst+; PV: Pvalb+; VIP: Vip+; LAMP5: Lamp5+ [Tasic Scholar], some neuron-derived neurotrophic factor [NDNF] interneurons Scholar]; MEIS2: Meis2 [Di Jin 2020Jin Simmons Guo Shetty A.S. Ko Jokhi Robinson Oyler Curry al.In vivo Perturb-Seq glial abnormalities genes.Science. 370: eaaz6063Crossref (22) Scholar]), medium spiny (MSNs: Tac1+, Meis2+ [Jin S1, remaining corresponded L2–L5 (mostly P1–P7), intermediate L2–L4 (P14), (P1–P7), SST/PV (P1–P7) principal-component (PCA) dissect sources cell-to-cell heterogeneity 2C). Despite neurons’ well-known specialization principal component (PC)—explaining 3.6% total variance—corresponded progression ages. By comparison, next PCs, distinguished (granule cells), explained 2.1% 1.7%, respectively. positioned along PC 2D). harbored (neonatal) partly neurogenesis DG Genes maturation 1 loadings 2E; Tables S2 S3) 2E). As confirmation, modules (Langfelder Horvath, 2008Langfelder Horvath WGCNA: R package weighted network analysis.BMC Bioinformatics. 9: 559Crossref (7397) largest modules—termed “neonatal” “adult”—corresponded loadings. 3,000 PCA, 75 module ranked 1st 123rd (mean rank 44), while 72 148th positive 53) 2E, S2). less stringent list top 10,000 (rather 3,000) variable 122 151 (Table S3). eigengenes eigengene expressed P1 sharply declined remained onward 2F, steadily increased until plateauing P56 therefore represented modes Ontology (GO) (Thomas 2003Thomas P.D. Kejariwal Mi Karlak Daverman Diemer Muruganujan Narechania PANTHER: library protein families subfamilies indexed function.Genome Res. 2003; 13: 2129-2141Crossref (1988) enriched (36 genes; false discovery rate [FDR] × 10?17), process (43; 10?13), (22; axon (17; 10?12), constituent cytoskeleton (5; 10?3), polymeric cytoskeletal fiber (12; 10?4) (Tables Neonatal included Tuba1a/b2a/b2b/b3/b5, Stmn1/2, Syne2), guidance Gap43, Nrp1, Dpysl3/5, Plxnb1, Sema3c), Lhx2, Sox4/11, Tcf4, Nfib, Nrep, Rcor2, Ctnnb1). Some enrichments microarray (Lyckman 2008Lyckman A.W. Horng Leamey Tropea Watakabe Van Wart McCurry Yamamori Sur period: synaptic reversal deprivation.Proc.