The full neutrophil heterogeneity and differentiation landscape remains incompletely characterized. Here, we profiled >25,000 differentiating and mature mouse neutrophils using single-cell RNA sequencing to provide a comprehensive transcriptional landscape of neutrophil maturation, function and fate decision in their steady state and during bacterial infection. Eight neutrophil populations were defined by distinct molecular signatures. The three mature peripheral blood neutrophil subsets arise from distinct maturing bone marrow neutrophil subsets. Driven by both known and uncharacterized transcription factors, neutrophils gradually acquire microbicidal capability as they traverse the transcriptional landscape, representing an evolved mechanism for fine-tuned regulation of an effective but balanced neutrophil response. Bacterial infection reprograms the genetic architecture of neutrophil populations, alters dynamic transitions between subpopulations and primes neutrophils for augmented functionality without affecting overall heterogeneity. In summary, these data establish a reference model and general framework for studying neutrophil-related disease mechanisms, biomarkers and therapeutic targets at single-cell resolution.
Subscribe to Journal
Get full journal access for 1 year
only $17.42 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
The data that support the findings of this study are available from the corresponding author upon request. All sequencing data generated in this study have been deposited at NCBI’s Gene Expression Omnibus (GEO) repository and are accessible through GEO Series accession number GSE137540. The published data used in this study were retrieved from the GEO (accession numbers GSE70245 (ref. 25), GSE109467 (ref. 26), GSE117129 (ref. 27), GSE92575 (ref. 13) and GSE120409 (ref. 29)). To align the differentiating neutrophil clusters characterized in this study to the proNeu population, we also utilized the most recent data from ref. 30 (kindly provided by L. G. Ng).
Nicolas-Avila, J. A., Adrover, J. M. & Hidalgo, A. Neutrophils in homeostasis, immunity, and cancer. Immunity 46, 15–28 (2017).
Nauseef, W. M. & Borregaard, N. Neutrophils at work. Nat. Immunol. 15, 602–611 (2014).
Silvestre-Roig, C., Hidalgo, A. & Soehnlein, O. Neutrophil heterogeneity: implications for homeostasis and pathogenesis. Blood 127, 2173–2181 (2016).
Ley, K. et al. Neutrophils: new insights and open questions. Sci. Immunol. 3, eaat4579 (2018).
Scapini, P., Marini, O., Tecchio, C. & Cassatella, M. A. Human neutrophils in the saga of cellular heterogeneity: insights and open questions. Immunol. Rev. 273, 48–60 (2016).
Ng, L. G., Ostuni, R. & Hidalgo, A. Heterogeneity of neutrophils. Nat. Rev. Immunol. 19, 255–265 (2019).
Adrover, J. M., Nicolas-Avila, J. A. & Hidalgo, A. Aging: a temporal dimension for neutrophils. Trends Immunol. 37, 334–345 (2016).
Yvan-Charvet, L. & Ng, L. G. Granulopoiesis and neutrophil homeostasis: a metabolic, daily balancing act. Trends Immunol. 40, 598–612 (2019).
Doerschuk, C. M. Leukocyte trafficking in alveoli and airway passages. Respir. Res. 1, 136–140 (2000).
Wang, Q. & Doerschuk, C. M. The signaling pathways induced by neutrophil–endothelial cell adhesion. Antioxid. Redox Signal. 4, 39–47 (2002).
Adlung, L. & Amit, I. From the Human Cell Atlas to dynamic immune maps in human disease. Nat. Rev. Immunol. 18, 597–598 (2018).
Stubbington, M. J. T., Rozenblatt-Rosen, O., Regev, A. & Teichmann, S. A. Single-cell transcriptomics to explore the immune system in health and disease. Science 358, 58–63 (2017).
Giladi, A. et al. Single-cell characterization of haematopoietic progenitors and their trajectories in homeostasis and perturbed haematopoiesis. Nat. Cell Biol. 20, 836–846 (2018).
Nestorowa, S. et al. A single-cell resolution map of mouse hematopoietic stem and progenitor cell differentiation. Blood 128, e20–e31 (2016).
Velten, L. et al. Human haematopoietic stem cell lineage commitment is a continuous process. Nat. Cell Biol. 19, 271–281 (2017).
Paul, F. et al. Transcriptional heterogeneity and lineage commitment in myeloid progenitors. Cell 163, 1663–1677 (2015).
Karamitros, D. et al. Single-cell analysis reveals the continuum of human lympho-myeloid progenitor cells. Nat. Immunol. 19, 85–97 (2018).
Qiu, X. et al. Reversed graph embedding resolves complex single-cell trajectories. Nat. Methods 14, 979–982 (2017).
Eash, K. J., Greenbaum, A. M., Gopalan, P. K. & Link, D. C. CXCR2 and CXCR4 antagonistically regulate neutrophil trafficking from murine bone marrow. J. Clin. Invest. 120, 2423–2431 (2010).
Cowland, J. B. & Borregaard, N. Granulopoiesis and granules of human neutrophils. Immunol. Rev. 273, 11–28 (2016).
Borregaard, N., Sørensen, O. E. & Theilgaard-Mönch, K. Neutrophil granules: a library of innate immunity proteins. Trends Immunol. 28, 340–345 (2007).
Sørensen, O., Arnljots, K., Cowland, J. B., Bainton, D. F. & Borregaard, N. The human antibacterial cathelicidin, hCAP-18, is synthesized in myelocytes and metamyelocytes and localized to specific granules in neutrophils. Blood 90, 2796–2803 (1997).
Hoogendijk, A. J. et al. Dynamic transcriptome–proteome correlation networks reveal human myeloid differentiation and neutrophil-specific programming. Cell Rep. 29, 2505–2519.e4 (2019).
Satake, S. et al. C/EBPβ is involved in the amplification of early granulocyte precursors during candidemia-induced “emergency” granulopoiesis. J. Immunol. 189, 4546–4555 (2012).
Olsson, A. et al. Single-cell analysis of mixed-lineage states leading to a binary cell fate choice. Nature 537, 698–702 (2016).
Evrard, M. et al. Developmental analysis of bone marrow neutrophils reveals populations specialized in expansion, trafficking, and effector functions. Immunity 48, 364–379.e8 (2018).
Zhu, Y. P. et al. Identification of an early unipotent neutrophil progenitor with pro-tumoral activity in mouse and human bone marrow. Cell Rep. 24, 2329–2341.e8 (2018).
Kim, M.-H. et al. A late-lineage murine neutrophil precursor population exhibits dynamic changes during demand-adapted granulopoiesis. Sci. Rep. 7, 39804 (2017).
Muench, D. E. et al. Mouse models of neutropenia reveal progenitor-stage-specific defects. Nature 2, 109–114 (2020).
Kwok, I. et al. Combinatorial single-cell analyses of granulocyte–monocyte progenitor heterogeneity reveals an early uni-potent neutrophil progenitor. Immunity https://doi.org/10.1016/j.immuni.2020.06.005 (2020).
Borregaard, N. & Herlin, T. Energy metabolism of human neutrophils during phagocytosis. J. Clin. Invest. 70, 550–557 (1982).
Chervenick, P. A., Boggs, D. R., Marsh, J. C., Cartwright, G. E. & Wintrobe, M. M. Quantitative studies of blood and bone marrow neutrophils in normal mice. Am. J. Physiol. 215, 353–360 (1968).
Colvin, G. A. et al. Murine marrow cellularity and the concept of stem cell competition: geographic and quantitative determinants in stem cell biology. Leukemia 18, 575–583 (2004).
La Manno, G. et al. RNA velocity of single cells. Nature 560, 494–498 (2018).
Broxmeyer, H. E. Chemokines in hematopoiesis. Curr. Opin. Hematol. 15, 49–58 (2008).
Furze, R. C. & Rankin, S. M. Neutrophil mobilization and clearance in the bone marrow. Immunology 125, 281–288 (2008).
Suratt, B. T. et al. Neutrophil maturation and activation determine anatomic site of clearance from circulation. Am. J. Physiol. Lung Cell Mol. Physiol. 281, L913–L921 (2001).
Theilgaard-Monch, K. et al. The transcriptional program of terminal granulocytic differentiation. Blood 105, 1785–1796 (2005).
Monticelli, S. & Natoli, G. Transcriptional determination and functional specificity of myeloid cells: making sense of diversity. Nat. Rev. Immunol. 17, 595–607 (2017).
Aibar, S. et al. SCENIC: single-cell regulatory network inference and clustering. Nat. Methods 14, 1083–1086 (2017).
Lambrechts, D. et al. Phenotype molding of stromal cells in the lung tumor microenvironment. Nat. Med. 24, 1277–1289 (2018).
Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902.e21 (2019).
Kwak, H. J. et al. Myeloid cell-derived reactive oxygen species externally regulate the proliferation of myeloid progenitors in emergency granulopoiesis. Immunity 42, 159–171 (2015).
Manz, M. G. & Boettcher, S. Emergency granulopoiesis. Nat. Rev. Immunol. 14, 302–314 (2014).
Tirosh, I. et al. Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science 352, 189–196 (2016).
Uhl, B. et al. Aged neutrophils contribute to the first line of defense in the acute inflammatory response. Blood 128, 2327–2337 (2016).
Kolaczkowska, E. The older the faster: aged neutrophils in inflammation. Blood 128, 2280–2282 (2016).
Luo, H. R. & Loison, F. Constitutive neutrophil apoptosis: mechanisms and regulation. Am. J. Hematol. 83, 288–295 (2008).
Schneider, W. M., Chevillotte, M. D. & Rice, C. M. Interferon-stimulated genes: a complex web of host defenses. Annu. Rev. Immunol. 32, 513–545 (2014).
Zilionis, R. et al. Single-cell transcriptomics of human and mouse lung cancers reveals conserved myeloid populations across individuals and species. Immunity 50, 1317–1334.e10 (2019).
Sakai, J. et al. Reactive oxygen species-induced actin glutathionylation controls actin dynamics in neutrophils. Immunity 37, 1037–1049 (2012).
Adrover, J. M. et al. A neutrophil timer coordinates immune defense and vascular protection. Immunity 50, 390–402.e10 (2019).
Hou, Q. et al. Inhibition of IP6K1 suppresses neutrophil-mediated pulmonary damage in bacterial pneumonia. Sci. Transl. Med. 10, eaal4045 (2018).
Loison, F. et al. Proteinase 3-dependent caspase-3 cleavage modulates neutrophil death and inflammation. J. Clin. Invest. 124, 4445–4458 (2014).
Karatepe, K. et al. Proteinase 3 limits the number of hematopoietic stem and progenitor cells in murine bone marrow. Stem Cell Rep. 11, 1092–1105 (2018).
Butler, A., Hoffman, P., Smibert, P., Papalexi, E. & Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat. Biotechnol. 36, 411–420 (2018).
Becht, E. et al. Dimensionality reduction for visualizing single-cell data using UMAP. Nat. Biotechnol. 37, 38–44 (2018).
Kowalczyk, M. S. et al. Single-cell RNA-seq reveals changes in cell cycle and differentiation programs upon aging of hematopoietic stem cells. Genome Res. 25, 1860–1872 (2015).
Tirosh, I. et al. Single-cell RNA-seq supports a developmental hierarchy in human oligodendroglioma. Nature 539, 309–313 (2016).
Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16, 284–287 (2012).
Consortium, G. O. The Gene Ontology resource: 20 years and still GOing strong. Nucleic Acids Res. 47, D330–D338 (2018).
Benaglia, T., Chauveau, D., Hunter, D. & Young, D. mixtools: An R package for analyzing finite mixture models. J. Stat. Softw. https://doi.org/10.18637/jss.v032.i06 (2009).
Zhang, D. et al. Neutrophil ageing is regulated by the microbiome. Nature 525, 528–532 (2015).
Suratt, B. T. et al. Role of the CXCR4/SDF-1 chemokine axis in circulating neutrophil homeostasis. Blood 104, 565–571 (2004).
Martin, C. et al. Chemokines acting via CXCR2 and CXCR4 control the release of neutrophils from the bone marrow and their return following senescence. Immunity 19, 583–593 (2003).
Newman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods 12, 453–457 (2015).
Mullen, K. M. & van Stokkum, I. H. M. nnls: The Lawson–Hanson algorithm for non-negative least squares (NNLS) v1.4 (2012); https://cran.r-project.org/web/packages/nnls/index.html
Kjeldsen, L., Bainton, D. F., Sengeløv, H. & Borregaard, N. Structural and functional heterogeneity among peroxidase-negative granules in human neutrophils: identification of a distinct gelatinase-containing granule subset by combined immunocytochemistry and subcellular fractionation. Blood 82, 3183–3191 (1993).
Gu, Z., Eils, R. & Schlesner, M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32, 2847–2849 (2016).
The authors thank A. Hidalgo, L. G. Ng, J. Manis and L. Chai for helpful discussions and suggestions. F.M., Y.X. and T.C. are supported by grants from the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (2017-I2M-1-015, 2016-I2M-1-017 and 2016-12M-1-003), the Non-profit Central Research Institute Fund of the Chinese Academy of Medical Sciences (2018RC31002, 2018PT32034 and 2017PT31033), the National Natural Science Foundation of China (81970107 and 81421002) and the Natural Science Foundation of Tianjin City (18JCYBJC25700). Q.S. and C.L. were supported by the National Natural Science Foundation of China (31871266), Chinese National Key Projects of Research and Development (2016YFA0100103) and NSFC Key Research Grant 71532001. H.R.L. is supported by National Institutes of Health grants (1 R01 AI142642, 1 R01 AI145274, 1 R01 AI141386, R01HL092020 and P01 HL095489) and a grant from FAMRI (CIA 123008). Part of the data analysis was performed on the High Performance Computing Platform of the Center for Life Sciences, Peking University.
The authors declare no competing interests.
Peer review information Jamie D. K. Wilson and Ioana Visan were the primary editors on this article and managed its editorial process and peer review in collaboration with the rest of the editorial team.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended Data Fig. 1 Sample preparation, quality controls, and related parameters and results related to scRNA-seq analysis.
a, Fluorescence-activated cell sorting (FACS) strategy for scRNA-seq sample preparation. b, Summary of sample information. c, Cell viability percentages immediately before cells were loaded into the 10X Chromium Controller. d, Representative GEM formation after the 10X Chromium Controller under the microscope. e, Violin plots of the number of genes, number of UMIs, mitochondria count percentage, and UMI per gene of all QC-passed cells in different organs. f, Uniform manifold approximation and projection (UMAP) of 19,582 cells from the bone marrow (BM), peripheral blood (PB), and spleen (SP) colored by sample origin and cell type, respectively. Expression of unique genes specifically distinguished each cluster and associated them with neutrophils (Neu) (S100a8 and S100a9), myeloid progenitors (MP) (Cd34, Kit, Mpo and Elane), hematopoietic stem progenitor cells (HSPC, not including MP) (Cd34, Kit, Mpo- and Elane-), monocytes (Mono) (S100a4 and Ccl9/MIP-1γ), B cells (Cd79a and Cd79b), T cells (Cd3d and Ccl5), and dendritic cells (DC) (Siglech), respectively. Cont: contaminated cells. g, Heatmap showing the five highest differentially expressed genes (DEGs) per cell type for all QC-passed cells. h, As in e but using only neutrophils in different organs. i, Comparison of Gr1+ BM neutrophil populations in our data with Ly6g+ BM neutrophil populations in Dr. Ido Amit’s data. Cluster labels are transferred from our data to Dr. Ido Amit’s data13(Methods). Left: UMAPs of 3591 Gr1+ neutrophils and 2304 Ly6g+ neutrophils colored by data set or cluster identity. Right: Neutrophil compositions in our data and Dr. Ido Amit’s data. j, Violin plots of the number of genes and number of UMIs of our Gr1+ neutrophils and Dr. Ido Amit’s Ly6g+ neutrophils.
Extended Data Fig. 2 scRNA-seq defined neutrophil populations correlated with previously reported neutrophil subpopulations.
a, Correlation of scRNA-seq defined neutrophil populations with the indicated four samples characterized by Olsson et al.25. Shown are the fraction of each indicated scRNA-seq defined cluster in the four samples. The cluster identity of each cell was inferred based on the transcriptomic similarity between this cell and the reference clusters (G0-G5) defined in current study. b, Correlation of scRNA-seq defined neutrophil populations with the neutrophil subtypes reported by Evrard et al.26. Coefficient matrix showing deconvolution results of bulk profiles of indicated neutrophil subpopulations. The 20 highest DEGs per single-cell group (G0-G5) were selected as signatures for deconvolution. Each column is normalized by column sums. (c, d), Correlation of scRNA-seq defined neutrophil populations with the C1 and C2 neutrophil clusters reported by Zhu et al.27. c, Top: t-distributed stochastic neighbor embedding (t-SNE) plot of the C1 and C2 cells characterized by Zhu et al. The raw data was retrieved from GEO website and reanalyzed. Bottom: t-SNE plot of C1 and C2 cells colored based on scRNA-seq defined clusters (G0-G5). The cluster identity of each cell was determined as described in (A). d, Heatmap showing row-scaled expression of the 10 highest DEGs of C1 and C2 in indicated scRNA-seq defined clusters. Signature genes Ly6g, Cebpa and Cebpe were also included in the map. (e-f), Correlation of scRNA-seq defined neutrophil populations with the Stage I and Stage II neutrophils defined by Giladi et al.13. e, The Stage I (x-axis) and Stage II (y-axis) score of each single cell in the reference sample of current study (Fig.1b). The scRNA-seq defined neutrophil identity of each cell is indicated. f, Violin plots of Stage I and Stage II score for each scRNA-seq defined neutrophil population. g, Correlation between scRNA-seq-defined neutrophil populations and the neutrophil subpopulations reported by Muench et al.29 Left: The fraction of indicated scRNA-seq defined clusters (G0-G5) in samples characterized by Muench et al. Right: The fraction of indicated scRNA-seq defined clusters (G0-G5) in each neutrophil subtype. Each row is normalized by row sums. The cluster identity of each cell was determined as described in (a). h, Correlation between scRNA-seq-defined neutrophil populations and the neutrophil subpopulations reported by Kwok et al.30 Shown are the fraction of indicated scRNA-seq defined clusters in each neutrophil subtype. Each row is normalized by row sums.
Extended Data Fig. 3 Three major neutrophil subpopulations, including an ISG-expressing G5b population, were identified in the PB and spleen.
a, Heatmap showing row-scaled expression of 47 interferon-stimulated genes (ISGs) for each averaged cluster. b, Monocle trajectories of neutrophil population G5a, G5b, and G5c. Each dot represents a single cell. c, Violin plots of the number of genes, number of UMIs, mitochondria count percentage, and UMI per gene of neutrophils in each cluster.
a-d, Violin plot of phagocytosis score (GO:0006911), chemotaxis score (GO:0030593), neutrophil activation score (GO:0042119), and NADPH oxidase score for each cluster. e, Heatmap showing relative expression of seven genes of the NADPH oxidase complex for all neutrophils. f, As in (a-d) but displaying mitochondria-mediated ROS production score (reactive oxygen species biosynthetic process, GO:1903409) for each cluster. g, Violin plots of metabolic scores for each cluster. Glycolysis (Reactome Pathway Database #R-MMU-70171); Oxidative phosphorylation (GO:000619); Electron transport chain (GO:0022900); Tricarboxylic acid cycle (GO:0006099). h-i, Heatmaps showing relative expression of glycolysis-related genes and glucose transport-related genes.
a, Heatmap showing row-scaled expression of the ten highest DEGs per organ for each averaged organ profile. b, As in a. but for each G5 subpopulation between PB and SP. KEGG analysis of DEGs for each G5 in these two organs. Left: selected KEGG terms with Benjamini-Hochberg-corrected P-values < 0.05 are shown. c, The percentages of each neutrophil subpopulation in the BM and PB calculated based on the scRNA-seq data (Fig.1c). d, The absolute numbers of each neutrophil subpopulation in the BM and PB calculated based on the percentage in (c) and predicted total BM neutrophil count32,33. (e-j), The transcriptome feature of the three G5 populations in the BM, PB, and SP. e, Heatmap showing row-scaled expression of DEGs across organs in each G5 cluster. (f-h), GO analysis of DEGs across organs. Selected GO terms with Benjamini-Hochberg-corrected P-values < 0.05 (one-sided Fisher’s exact test) are shown. i, Violin plots of maturation score and apoptosis score for each G5 neutrophil subpopulation across organs. j, Proportions of apoptotic cells in each G5 neutrophil subpopulation across organs.
a, Number of white blood cells and the proportion of neutrophils in mice before and after E. coli challenge evaluated by a hematology analyzer (Mindray BC-5000 Vet). Results are the mean ± SD of three independent experiments. b, Experimental scheme of the sample collection process after E. coli challenge. c, Summary of sample information. Organ distribution of neutrophils is shown on the right. d, Cell viability percentages immediately before cells were loaded into the 10X Chromium Controller. e, UMAPs of all 24,943 cells from BM, PB, and SP from E. coli-challenged mice colored by sample origin and cell type, respectively. f, Heatmap showing row-scaled expression of the five highest DEGs for all QC-passed cells colored by cell type. g, Comparisons of the number of genes, number of UMIs, mitochondria count percentage, and UMI per gene of all QC-passed cells in each organ before and after E. coli challenge. h, As in g but only of all neutrophils in each organ. i-j, Heatmaps showing expression of 7 genes of the NADPH oxidase complex (i) and neutrophil granule-related genes (j) for all neutrophils.
Extended Data Fig. 7 Differentially expressed genes in each neutrophil subpopulation in E. coli-challenged mice.
a, MA plots displaying genes that are up- (red) or downregulated (blue) after E. coli challenge for each cluster. Dashed lines denote fold change thresholds used when identifying DEGs. b, Gene ontology (GO) analysis of DEGs before and after E. coli challenge for each cluster. Selected GO terms with Benjamini-Hochberg-corrected P-values < 0.05 (one-sided Fisher’s exact test) are shown.
a, UMAP of the regulon activity matrix of 32,888 cells (11,992 normal neutrophils, 13,687 challenged neutrophils, and 7209 other cells under normal conditions) colored by Seurat cluster identity (top) or experimental condition (bottom, only neutrophils). b, Heatmap of the t-values of regulon activity derived from a generalized linear model for the difference between cells from one challenged neutrophil subpopulation and cells from the corresponding normal subpopulation. Only regulons with at least one absolute t-value greater than 18 are visualized. Regulons are hierarchically clustered based on challenge-response pattern (purple: upregulated, yellow: first up- then downregulated, green: downregulated) c, Heatmap showing activity change of regulons identified in (b) during normal group transitions.
a, Overview of study design and the gating strategy for isolating human PB neutrophils. b, UMAP plots of neutrophils from three healthy donors (D1, D2, or D3) colored by cluster identity. c, The combined UMAP plot of the three donors. d, Dot plot showing scaled expression of selected signature genes for each cluster colored by average expression of each gene in each cluster scaled across all clusters. Dot size represents the percentage of cells in each cluster with more than one read of the corresponding gene. The analysis was conducted using cells from all three human donors. e, Row-scaled expression of the ten highest differentially expressed genes (Bonferroni-corrected P values < 0.05, Student’s t-test) in each neutrophil cluster. D1+D2+D3, the analysis was conducted using cells from all three human donors. f, Row-scaled expression of 37 interferon-stimulated genes in each neutrophil cluster. The analysis was conducted using cells from all three human donors.
Extended Data Fig. 10 Single-cell transcriptome profiling reveals eight neutrophil subpopulations defined by distinct molecular signatures.
Summary of dynamic change of morphology, gene expression (Ly6g and c-kit), TF expression (Cebpe and Cebpb) and granules (Azurophil, Specific, Gelatinase granules and Secretory Vesicles) between each subpopulation. Comparison between our scRNA transcriptome profiles with other published neutrophil populations13,25,26,27,28,29,30.
About this article
Cite this article
Xie, X., Shi, Q., Wu, P. et al. Single-cell transcriptome profiling reveals neutrophil heterogeneity in homeostasis and infection. Nat Immunol 21, 1119–1133 (2020). https://doi.org/10.1038/s41590-020-0736-z