bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
An atlas of the aging lung mapped by single cell transcriptomics and deep tissue proteomics
Ilias Angelidis1*, Lukas M. Simon2*, Isis E. Fernandez1, Maximilian Strunz1, Christoph H. Mayr1, Flavia R. Greiffo1, George Tsitsiridis2, Elisabeth Graf3, Tim‐Matthias Strom3, Oliver Eickelberg4, Matthias Mann5, Fabian J. Theis2,6 #, and Herbert B. Schiller1, #
Aging promotes lung function decline and susceptibility to chronic lung diseases, which are the third leading cause of death worldwide. We used single cell transcriptomics and mass spectrometry to quantify changes in cellular activity states of 30 cell types and the tissue proteome from lungs of young and old mice. Aging led to increased transcriptional noise, indicating deregulated epigenetic control. We observed highly distinct effects of aging on cell type level, uncovering increased cholesterol biosynthesis in type‐2 pneumocytes and lipofibroblasts as a novel hallmark of lung aging. Proteomic profiling revealed extracellular matrix remodeling in old mice, including increased collagen IV andI XV and decreased Fraser syndrome complex proteins and Collagen XIV. Computational integration of the aging proteome and single cell transcriptomes predicted the cellular source of regulated proteins and created a first unbiased reference of the aging lung. The lung aging atlas can be accessed via an interactive user‐friendly webtool at: https://theislab.github.io/LungAgingAtlas ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ Keywords: aging, chronic lung disease, single cell transcriptomics, proteomics, extracellular matrix, lipid metabolism, lung cell atlas ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ 1. Helmholtz Zentrum München, Institute of Lung Biology and Disease, Member of the German Center for Lung Research (DZL), Munich, Germany 2. Helmholtz Zentrum München, Institute of Computational Biology, Munich, Germany 3. Helmholtz Zentrum München, Institute of Human Genetics, Munich, Germany 4. University of Colorado, Department of Medicine, Division of Respiratory Sciences and Critical Care Medicine, Denver, CO, USA 5. Max Planck Institute of Biochemistry, Department of Proteomics and Signal Transduction, Martinsried, Germany 6. Department of Mathematics, Technische Universität München, Munich, Germany
*… these authors contributed equally to this work # ... correspondence to Fabian J. Theis (fabian.theis@helmholtz‐muenchen.de) and Herbert B. Schiller (herbert.schiller@helmholtz‐muenchen.de) ------The intricate structure of the lung enables gas exchange frequency in mice, thereby decreasing mucociliary between inhaled air and circulating blood. As the organ clearance and partially explaining the predisposition of with the largest surface area (~70 m2 in humans), the lung the elderly to pneumonia5. Senescence of the immune is constantly exposed to a plethora of environmental system in the elderly has been linked to a phenomenon insults. A range of protection mechanisms are in place, called `inflammaging´, which refers to elevated levels of including a highly specialized set of lung resident innate tissue and circulating pro‐inflammatory cytokines in the and adaptive immune cells that fight off infection, as well absence of an immunological threat6. Several previous as several stem and progenitor cell populations that studies analyzing the effect of aging on pulmonary provide the lung with a remarkable regenerative capacity immunity point to age dependent changes of the immune upon injury1. These protection mechanisms seem to repertoire as well as activity and recruitment of immune deteriorate with advanced age, since aging is the main cells upon infection and injury4. Vulnerability to oxidative risk factor for developing chronic lung diseases, including stress, pathological nitric oxide signaling, and deficient chronic obstructive pulmonary disease (COPD), lung recruitment of endothelial stem cell precursors have been cancer, and interstitial lung disease (ILD)2, 3. Advanced age described for the aged pulmonary vasculature7. The ECM causes a progressive impairment of lung function even in of old lungs features changes in tensile strength and otherwise healthy individuals, featuring structural and elasticity, which were discussed to be a possible immunological alterations that affect gas exchange and consequence of fibroblast senescence8. Using atomic susceptibility to disease4. Aging decreases ciliary beat force microscopy, age‐related increases in stiffness of
1 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
parenchymal and vessel compartments were samples contributed as little as a single cell to this demonstrated recently9, however, the causal molecular megakaryocyte cluster, emphasizing the power and changes underlying these effects are unknown. accuracy of the computational workflow used here for data integration from multiple mice. Aging is a multifactorial process that leads to these molecular and cellular changes in a complicated series of We used differential gene expression analysis to events. The hallmarks of aging encompass cell‐intrinsic determine cell type specific marker genes with highly effects, such as genomic instability, telomere attrition, different levels between clusters (Fig. 1c, Table S1). The epigenetic alterations, loss of proteostasis, deregulated clusters were annotated with assumed cell type identities nutrient sensing, mitochondrial dysfunction and based on (1) known marker genes derived from expert senescence, as well as cell‐extrinsic effects, such as annotation in literature, and (2) enrichment analysis altered intercellular communication and extracellular 2, 3 using Fisher’s exact test of gene expression signatures of matrix remodeling . The lung contains at least 40 18 10 isolated cell types from databases including ImmGen distinct cell types , and specific effects of age on cell type and xCell19. Correlation analysis of marker gene signatures level have never been systematically analyzed. In this revealed that similar cell types clustered together, study, we build on rapid progress in single cell 11, 12 implying correct cell type annotation (Fig. 1c). We used transcriptomics , which recently enabled the the matchSCore tool to compare the cluster identities of generation of a first cell type resolved census of murine 13 our dataset with the lung data in the recently published lungs , serving as a starting point for investigating the Mouse Cell Atlas13, and found very good agreement in lung in distinct biological conditions as shown for lung cluster identities and annotations (Fig. S1c). aging in the present work. We computationally integrated single cell signatures of aging with state of the art mass 14 It was suggested that aging is the result of an increase in spectrometry driven proteomics to generate the first transcriptional instability rather than a coordinated multi‐omics whole organ resource of aging associated transcriptional program, and that an aging associated molecular and cellular alterations in the lung. increase in transcriptional noise can lead to `fate drifts´ 20, 21 and ambiguous cell type identities . Therefore, we Results quantified transcriptional noise by calculating the median euclidean distance between the gene expression profile A single cell atlas of mouse lung aging reveals common of all single cells from each mouse and the average gene deregulated transcriptional control expression profile of the respective cell type. A significant positive association between transcriptional noise and To generate a cell‐type resolved map of lung aging we age was observed after accounting for cell type identity as performed highly parallel genome‐wide expression 15 a covariate (Analysis of variance, P: <1e‐5). Thus, this profiling of individual cells using the Dropseq workflow , analysis confirmed an increase in transcriptional noise which uses both molecule and cell‐specific barcoding, with aging in most cell types (Fig. 1d) in line with previous enabling great cost efficiency and accurate quantification 21 20 16 reports in the human pancreas or mouse CD4+ T cells . of transcripts without amplification bias . Single cell suspensions of whole lungs were generated from 3 (n=8) Multi‐omics data integration of mRNA and protein aging and 24 (n=7) month old mice. After quality control, a total signatures of 14,813 cells were used for downstream analysis (Fig. 1a). Unsupervised clustering analysis revealed 36 distinct To capture age dependent alterations in both mRNA and clusters corresponding to 30 cell types, including all major protein content, a second independent cohort of young known epithelial, mesenchymal, and leukocyte lineages and old mice was analyzed with a state of the art shotgun (Fig. 1b and d). We observed very good overlap across proteomics workflow (Fig. S2 and Table S2). To compare mouse samples and most clusters were derived from proteome and transcriptome data we generated in silico >70% of the mice of both age groups (Fig. S1a). The bulk samples from the scRNA‐seq data by summing definition of cell types (clusters in tSNE map) was very expression counts from all cells for each mouse comparable between old and young mice, indicating that individually (Fig. 2a). Differential gene expression analysis thee cell typ identity was not strongly confounded by the of 21969 genes from in silico bulks revealed a total of aging effects (Fig. S1b). Two clusters exclusively contained 2362 differentially expressed genes (FDR < 10%) between cells from a single mouse and were removed from theo tw age groups (Fig. 2b, Table S3). From whole lung downstream analysis. Interestingly, we identified even tissue proteomes we quantified 5212 proteins across rare (<1%, 43 cells) cell types such as megakaryocytes, conditions and found 213 proteins to be significantly which were recently identified as a novel unexpected regulated with age, including 32 ECM proteins (FDR < tissue resident cell type in mouse lung17. Of note, some 10%, Fig. 2c, Supplementary Fig. S2 and Table S2).
2 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
a Cell type resolved analysis of lung aging whole lung 3 month (n=8) Highseq 4000 Data analysis single cell suspensions Single cell library 15k cells (Dropseq) 24 month (n=7)
Transcriptional noise Difference − 0.02 0.25 0.30 0.35 0.40 bdn = 15 (mice); k=14,813 (cells) 0.00 0.02 young old 18 23 30 Plasma_cells 13 Eosinophils P-value: Mesothelial_cells < 0.05 11 28 Lipofibroblast 29 Club_cells > 0.05 8 12 22 Type1_pneumocytes Interstitial macrophages Ciliated_cells 17 24 Ccl17+/Cd103−/Cd11b−_dend ritic_cells 25 Mki67+_proliferating_cells 14 1 Fn1+_macrophage 27 Vcam1+_endothelial_cells classical_monocyte_( Ly6c2+) 10 Goblet_cells tSNE 2 Natural_Killer_cells 9 19 20 Capillary_endothelial_cells Alveolar_macrophage 16 Cd103+/Cd11b−_dend ritic_cells 6 Type_2_pneumocytes 21 2 7 Cd4+_T_cells 3 B_cells Interstitial_Fibroblast 26 5 CD209+/Cd11b+_dend ritic_cells vascular_endothelial_cells 4 CD8+_T_cells Megakaryocytes 15 non−classical_monocyte_( Ly6c2−) Smooth_muscle_cells lymphatic_endothelial_cells tSNE 1 Neutrophils higher higher in young in old c 1 Alveolar_macrophage 2 Mki67+_proliferating_cells 3 Natural_Killer_cells 4 Plasma_cells 5 B_cells 6 CD4+_T_cells 7 CD8+_T_cells 8 Interstital macrophages pct.exp 9 non−classical_monocyte_(Ly6c2−) 0.00 10 classical_monocyte_(Ly6c2+) 0.25 0.50 11 CD103+/Cd11b−_dendritic_cells 0.75 12 CD209+/Cd11b+_dendritic_cells 1.00 13 Ccl17+/CD103−/CD11b−_dendritic_cells 14 Megakaryocytes 15 Neutrophils avg.exp.scale 16 Eosinophils 2 17 Fn1+_macrophage 18 Lymphatic_endothelial_cells 1 19 Vcam1+_endothelial_cells 0 20 Vascular_endothelial_cells 21 Capillary_endothelial_cells 22 Mesothelial_cells 23 Smooth_muscle_cells 24 Interstitial_Fibroblast 25 Lipofibroblast 26 Type 1_pneumocytes 27 Type 2_pneumocytes 28 Ciliated_cells 29 Club_cells 30 Goblet_cells Vwf Dcn Ngp rbc2 Inmt Ear2 Prg4 Msln Itgax Itgae Ppbp Sftpd Foxj1 C1qb T Mzb1 Itgam Plac8 Ccl17 Acta2 Cxcr2 Rtkn2 Ly6c2 Gzma Ednrb Bpifb1 Cd79b Cd8b1 Mmrn1 Col1a2 Cd209a Scgb1a1
Tmem100
Figure 1. A single cell atlas of mouse lung aging. (a) Rows represent hierarchically clustered cell types, Experimental design – whole lung single cell suspensions of demonstrating similarities of transcriptional profiles. (d) Boxplot young and old mice were analyzed using the Dropseq workflow. illustrates transcriptional noise by age and celltype. Barplot to (b) The tSNE visualization shows unsupervised transcriptome the right depicts the difference between the median clustering, revealing 30 distinct cellular identities. (c) The transcriptional noise values between young and old mice. Purple dotplot shows (1) the percentage of cells expressing the color indicates differences trending towards statistical respective selected marker gene using dot size and (2) the significant (P: <0.05). average expression level of that gene based on UMI counts.
3 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
Prediction of the upstream regulators22 of the observed inverted in old mice with more than twofold more ciliated expression changes in either the transcriptome or cells over Club cells (Fig. S3c). We validated this finding in proteome data gave very similar results (Fig. 2d). In both situ by quantifying airway club and ciliated cells using datasets from independent mouse cohorts, we immunostainings of Foxj1 (ciliated cell marker) and CC10 discovered a pro‐inflammatory signature, which included (club cell marker) (Fig. S3d). Club cell numbers were upregulation of Il6, Il1b, Tnf, and Ifng, as well as the decreased in old mice (Fig S3e), while ciliated cells were downregulation of Pparg and Il10 (Fig. 2d). Furthermore, increased (Fig. S3f), leading to a significantly altered ratio to reveal common or distinct regulation of gene of club to ciliated cells in aged mouse airways (Fig. S3g). annotation categories in the transcriptome or proteome we performed a two dimensional annotation enrichment Age dependent changes in composition and organization analysis23 (Table S4). Again, most gene categories of the pulmonary extracellular matrix regulated by age were showing the same direction in The extracellular matrix (ECM) can act as a solid phase‐ transcriptome and proteome so that the positive Pearson binding interface for hundreds of secreted proteins, correlation of the annotation enrichment scores was creating an information‐rich signaling template for cell highly significant (Fig. 2e). We observed several hallmarks function and differentiation24. Alterations in ECM of aging, including a decline in mitochondrial function and composition and possibly architecture in the aging lung upregulation of pro‐inflammatory pathways have been suggested25, however experimental evidence (`inflammaging´). Interestingly, we detected a strong using unbiased mass spectrometry is scarce. We increase in immunoglobulins in both datasets, as well as previously developed the quantitative detergent solubility higher levels of MHC class I, which is consistent with the profiling (QDSP) method to add an additional dimension observed increase in the interferon pathway (Fig. 2e). of protein solubility to tissue proteomes26‐28. In QDSP, Many extracellular matrix genes, such as Collagen III were proteins are extracted from tissue homogenates with downregulated on both the mRNA and protein level (Fig. increasing stringency of detergents, which typically leaves 2f), while the levels of all basement membrane‐ ECM proteins enriched in the insoluble last fraction. This associated Collagen IV genes were increased on the enables better coverage of ECM proteins and analysis of protein level, but decreased at the mRNA level (Fig. 2g). the strength of their associations with higher order ECM The differential regulation of Collagen IV transcripts and structures such as microfibrils or collagen networks. We proteins highlights the importance of combined RNA and applied this method to young and old mice and compared protein analysis. We validated the increased protein protein solubility profiles between the two groups (Fig. abundance of Collagen IV using immunofluorescence and 3a). Differential comparison of the solubility profiles found that interestingly the main increase in Collagen IV between young and old mice revealed 74 proteins, in old mice was found around airways and vessels (Fig. including eight ECM proteins, with altered solubility 2h). profiles (FDR <20%) (Table S5).
Using principal component analysis 2of 43 secreted The combination of tissue proteomics with single cell extracellular proteins we found that the protein solubility transcriptomics enabled us to predict the cellular source fractions separated in component 1, while the age groups of the regulated proteins, which can be explored in the separated in component 4 of the data (Fig. 3b). Thus, online webtool. Furthermore, single cell RNAseq can principal component analysis enabled the stratification of disentangle relative frequency changes of cell types from secreted proteins by their biochemical solubility and their real changes in gene expression within a given cell type. differential behavior upon aging (Fig. 4c). This analysis We analyzed age dependent alterations of relative also showed that neither the abundance nor the solubility frequencies of the 30 cell types represented in our of many ECM proteins, including Collagen I and basement dataset. Since the cell type frequencies are proportions membrane laminins, was altered (Fig. 3c). We uncovered the data is compositional. Therefore, it is impossible to several ECM proteins with greatly reduced abundance statistically discern if a relative change in cell type and/or significant changes in their solubility profiles. frequency is caused by the increase of a given cell type or While the most abundant basement membrane laminin the decrease of another. However, after performing chain (Lamc1) was unaltered in both abundance (Fig. 3d) dimension reduction using multidimensional scaling of and solubility (Fig. 3g), serving as a control for overall the cell type proportions we observed a significant integrity of the basement membrane and the quality of association between the first coordinate and age (Fig. our data, the basement membrane‐associated trimeric S3a, b; Wilcoxon test, P < 0.005), indicating that cell type Fraser Syndrome complex (consisting of Fras1, Frem1, frequencies differed between young and old mice. and Frem2) was downregulated (Fig. 3e) and more Interestingly, the ratio of club to ciliated cells was altered soluble (Fig.) 3h in old age. in old mice. While Dropseq data from young mice contained more club cells than ciliated cells, this ratio was
4 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
Figure 2. Multi‐omics analysis of lung aging. (a) Experimental ageing. (e) The scatter plot shows the result of a two design ‐ two independent cohorts of young and old mice were dimensional annotation enrichment analysis based on fold analyzed by RNA‐seq and mass spectrometry driven proteomics changes in the transcriptome (x‐axis) and proteome (y‐axis), respectively. (b, c) The volcano plots show significantly which resulted in a significant positive correlation of both regulated genes in the (b) transcriptome and (c) proteome. (d) datasets. Types of databases used for gene annotation are color Gene expression and protein abundance fold changes were used coded as depicted in the legend. (f, g) Normalized relative to predict upstream regulators that are known to drive gene abundance of the depicted (f) interstitial matrix and (g) Collagen expression responses similar to the ones experimentally IV genes in the transcriptome and proteome experiments observed. Upstream regulators could be cytokines or respectively. (h) Immunofluorescence image of collagen type IV transcription factors. The color coded activation z‐score using confocal microscopy at 25x magnification. Note, the illustrates the prediction of increased or decreased activity upon increased fluorescence intensity around airways in old mice.
Incorporation of the Fraser syndrome complex within the three proteins29, indicating that this assembly and/or the basement membrane (rendering it more insoluble) has expression of either one or all subunits of the complex is been shown to depend on extracellular assembly of all perturbed in old mice. Fraser syndrome is a skin‐blistering
5 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
disease which points to an important function of the In alveolar type 2 pneumocytes, 121 mRNAs were Fraser syndrome complex proteins in linking the epithelial significantly regulated (FDR<10%, Fig. 4c). We observed a basement membrane to the underlying mesenchyme29. In strong increase of the Major Histocompatibility Complex the lungs of adult mice, expression is restricted to the (MHC) class I proteins H2‐K1, H2‐Q7, H2‐D1, and B2m mesothelium; Fras1 ‐/‐ mice develop lung lobulation (Fig. 4c, d), which we validated using an independent flow defects30. Indeed, we were able to confirm the cytometry experiment (Fig. 4f). Elevated MHC class I mesothelium specific expression of Fras1 using our levels likely result in increased presentation of self‐ scRNAseq data (Fig. 3j, k). Another ECM protein with antigens to the immune system and are consistent with markedly reduced protein levels (Fig. 3f) and increased our observation of a prominent Interferon‐gamma solubility (Fig. 3l) was Collagen XIV, a collagen of the signature in old mice (Fig. 2d), which is known to activate FACIT family of collagens that is associated with the MHC class I expression39. Type‐2 pneumocytes of old mice surface of Collagen I fibrils and may function by featured a highly significant upregulation of the enzyme integrating collagen bundles31. Collagen XIV is a major Acyl‐CoA desaturase 1 (Scd1), which is the fatty acyl Δ9‐ ECM binding site for the proteoglycan Decorin32, which is desaturating enzyme that converts saturated fatty acids known to regulate TGF‐beta activity33, 34. Interestingly, our (SFA) into monounsaturated fatty acids (MUFA) (Fig. 4c, scRNAseq data localized Collagen XIV expression to e). The age dependent upregulation of Scd1 in type‐2 interstitial fibroblasts, which also specifically expressed pneumocytes is a novel observation, which may have Decorin and were distinct from the lipofibroblasts that important implications since Scd1 is thought to induce showed very little expression of this particular collagen adaptive stress signaling that maintains cellular (Fig. 3 j/k). persistence and fosters survival and cellular functionality under distinct pathological conditions40.
Cell type specific effects of aging To obtain a meta‐analysis of changes in previously characterized gene expression modules and pathways, we Cell type‐resolved differential gene expression testing used cell type resolved mRNA fold changes for gene between age groups in the single cell data sets identified annotation enrichment analysis (Fig. 5a and b, Table S7) 391 significantly regulated genes (FDR < 10%) (Fig. 4a; and upstream regulator analysis (Fig. 5c‐e). The analysis Table S6). Alveolar macrophages and type 2 revealed cell type specific alterations in gene expression pneumocytes, the two cell types with highest number of programs upon aging. For instance, comparing club cells cells in the dataset, are discussed as an example for the to type‐2 pneumocytes, showed that Nrf2 (Nfe2l2) type of insight that can be gained from our cell type mediated oxidative stress responses were higher in type‐ resolved resource. Both cell types showed a clearly 2 pneumocytes of old mice and lower in club cells (Fig. altered phenotype in aged mice. 5c). Aging is known to affect growth signaling via the In alveolar macrophages, we found 125 significantly evolutionary conserved Igf‐1/Akt/mTOR axis2. regulated mRNAs (FDR<10%, Fig. 4b), including the Interestingly, we found evidence for increased mTOR downregulation of the genes for Eosinophil cationic signaling in type‐2 and club cells, but not in ciliated and protein 1 & 2 (Ear1 & Ear2), which have ribonuclease goblet cells (Fig. 5c). Mesenchymal cells showed activity ande ar thought to have potent innate immune remarkable differences in their aging response (Fig. 5d). 35 functions as antiviral factors . The cell surface scavenger‐ For instance, we observed the pro‐inflammatory Il1b type receptor Marco, mediating ingestion of unopsonized signature in capillary endothelial cells, as well as in 36 environmental particles , was downregulated in old mice mesothelial and smooth muscle cells, but not in the other which may cause impaired particle clearance by alveolar mesenchymal cell types. In myeloid cell types we found macrophages. We observed higher levels of the C/EBP both differences and similarities in the aging response beta (Cebpb), which is an important transcription factor (Fig. 5e). While, an increased interferon gamma and regulating the expression of genes involved in immune reduced Il10 signature in old mice was consistently and inflammatory responses37, 38. Several genes that have observed, other effects were more specific, such as the been shown to be upregulated in lung injury, repair and increase in Stat1 target genes in classical monocytes 28 fibrosis , such as Spp1, Gpnmb, and Mfge8 were also (Ly6c2+), which was not observed in non‐classical induced in alveolar macrophages of old mice, which may monocytes (Ly6c2‐). be a consequence of the ongoing `inflammaging´.
6 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
a Quantitative detergent Detergent extraction Mass Protein solubility solubility profiling (QDSP) stringency spectrometry - extracellular matrix profiling (4 protein fractions) 3 month Total tissue homogenate MS-intensity
24 month FR1 FR2 FR3 INSOL FR1 FR2 FR3 INSOL
bcFR1 FR2 young FR3 INSOL old 432 secreted proteins (+ Matrisome) Protein solubility Col14a1 Frem1 2
15 Frem2 A1bg Col6a4 Nppa Col6a6 10 1 Serpinb6b Ltf Fmod Fras1 Lama2 Calu
5 Col1a2 Fgf2 Lamc1 0 Ace higher in young Col6a3 Col1a1 old in higher Lama5 0 Col4a3 C4b Vtn Col4a2
-1 Apoa1 Serpina1a C1qtnf7 Scgb1a1 Ighm -5 B2m Mfge8 Chil3 Fetub Col16a1 Component 4 [1e-1] -2 Component 4 (2.6%)
-10 C9 -15 -3 Serpina1e -40 -30 -20 -10 0 10 20 30 40 50 -1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Component 1 (60.7%) Component 1 [1e-1]
defLamc1 - protein abundance Frem1 - protein abundance Col14a1 - protein abundance 40 32 33 39 30 32 38 31 28 37 30 26 36 29 35 24 28 MS-intensity [log2] MS-intensity [log2] MS-intensity [log2] MS-intensity 34 22 27 young old young old young old
ghiLamc1 - QDSP Frem1 - QDSP 4 Col14a1 - QDSP 8 4 young young young 6 old old old 2 2 4 2 0 0 0 -2 -2 -2
Normalized MS-intensity -4 Normalized MS-intensity -4 Normalized MS-intensity FR1 FR2 FR3 INSOL FR1 FR2 FR3 INSOL -4 FR1 FR2 FR3 INSOL
jkSub-tSNE of mesenchymal cells (EC excluded) 7- Pericytes 20 avg.exp.scale 6 - Fibroblast subtype 2 2 0 1 0 5 - Fibroblast subtype 1 0 10 −1 1 1 7 2 4 - TCF21+ Lipofibroblasts pct.exp 3 3 0.00 0 4 3 - Smooth muscle cells 0.25 5 5
tSNE_2 0.50 6 2 - Interstitial matrix fibroblasts −10 7 1 - Mesothelial cells - state2 0.75 2 6 0 - Mesothelial cells - state1 −20 4 Number of cells ras1 Msln Tcf21 F Acta2 Fndc1 Cspg4
−10 0 10 20 Lama2 Col3a1 Col1a2 Col4a1 Col16a1 tSNE_1 Col14a1
Figure 3. Quantitative detergent solubility profiling of the symbol shape, in component 1 and the age groups, as indicated aging lung proteome. (a) Experimental design – extraction of by color, in component 4. (c) The loadings of the PCA are shown. proteins from whole lung homogenates with increasing (d‐f) Relative differences in MS‐intensity (abundance) of the detergent stringency results in four distinct protein fractions, indicated proteins. (g‐i) The normalized MS‐intensity across the which are analyzed by mass spectrometry. (b) The projections of four protein fractions from differential detergent extraction a principal component analysis (PCA) of 432 proteins with the highlights changes in protein solubility between young and old annotation `secreted´ in the Uniprot and/or Matrisome mice for the indicated proteins. (j) The tSNE map shows 483 database separate the four protein fractions, indicated by cells of mesenchymal origin (endothelial cells excluded),
7 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
revealing 8 distinct cellular states. (k) The dotplot shows (1) the gene based on UMI counts. Proteins that were found to be up‐ percentage of cells expressing the respective selected marker or downregulated are highlighted in red and blue respectively. gene using dot size and (2) the average expression level of that
Increased cholesterol biosynthesis in aged type‐2 multi‐omics systems biology tools for the analysis of lung pneumocytes and lipofibroblasts aging are needed25. In this work, we present the first single cell survey of mouse lung aging and Pulmonary surfactant homeostasis is a tightly regulated computationally integrate transcriptomics data with deep process that involves synthesis of lipids by type‐2 tissue proteomics data to build an atlas of the aging lung. pneumocytes and lipofibroblasts41. Lipid metabolism in The lung aging atlas and associated raw data can be alveolar type‐2 cells is regulated by sterol‐response accessed at https://theislab.github.io/LungAgingAtlas element‐binding proteins (SREBPs), such as Srebf2 and (Figure S4). It features five dimensions that can be their negative regulators Insig1 and Insig2. Deletion of navigated through gene and cell type specific queries: (1) Insig1/2 in mouse type‐2 pneumocytes activated SREBPs cell type specific expression of genes and marker and led to the accumulation of neutral lipids (cholesterol signatures for 30 cell types, (2) regulation of gene esters and trigylcerids) in type‐ 2 pneumocytes and expression by age on cell type level, (3) cell type resolved alveolar macrophages, accompanied by lipotoxicity‐ pathway and gene category enrichment analysis, (4) related lung inflammation and tissue remodeling42. regulation of protein abundance by age on tissue level, (5) Interestingly, we observed very similar gene expression regulation of protein solubility by age. changes in type‐2 pneumocytes of old mice as reported
for the Insig1/2 deletion. Consistently, the upstream The highly multiplexed nature of droplet based single cell regulator analysis predicted increased activity of Srebf2 RNA sequencing used in this study allows the direct and reduced activity of Insig1 specifically in type‐2 analysis of thousands of individual cells freshly isolated pneumocytes of old mice (Fig. 5c). The upstream from whole mouse lungs, providing unbiased regulator analysisd was base on 25 known targets of classification of cell types and cellular states. Two SREBP/Insig1, all of which were increased in aged type‐2 previous studies have analyzed aging effects using single pneumocytes (Fig. 6a). Using gene annotation enrichment 20, 21 cell transcriptomics and found increased analysis on Uniprot Keywords, GO terms and KEGG transcriptional variability between cells in human pathways (Table S7) we found increased cholesterol pancreas and T cells. In this study, we identify aging biosynthesis as the top hit in type‐II pneumocytes and associated increased transcriptional noise, which may lipofibroblasts and no other cell type (Fig. 6b). Indeed, result from deregulated epigenetic control, in most cell most of the Insig1/2 target genes are directly involved in types of the lung, indicating that this phenomenon is a cholesterol biosynthesis (Fig. 6c). general hallmark of aging that likely affects most cell To confirm the increased cholesterol biosynthesis and types in both mice and humans. This new concept is analyze the actual lipid content of the cells we used the supported by our study and it will be interesting when Nile red dye to stain neutral lipids in cells of a whole lung and how future investigations will shed light on the suspension after depletion of leukocytes (Fig. 6d). Using molecular mechanisms driving this phenomenon. flow cytometry we quantified the Nile red lipid staining and found a significant increase in both mean We have used two independent cohorts of young and old fluorescence intensity (Fig. 6d, e), and number of Nile red mice and uncovered remarkably well conserved aging positive cells (Fig. 6f, g) in the CD45 negative cells of old signatures in both mRNA and protein. Thus, the two mice. CD45+ cells were not significantly altered, indicating datasets validate each other and show that on gene that the increase in neutral lipid content is specific to category and pathway level, the analysis of protein and epithelial cells and fibroblasts. Thus, we have shown for mRNA content can lead to overall similar results with the first time that increased cholesterol biosynthesis and important differences. Hallmarks of aging, such as the neutral lipid content in type‐2 pneumocytes and downregulation of mitochondrial oxidative lipofibroblasts is a hallmark of lung aging. phosphorylation and the upregulation of pro‐ inflammatory signaling pathways were consistently
observed in both datasets. On the level of individual
genes/proteins, however, we often observed interesting Discussion differences, which indicates that for functional analysis of a particular gene/protein it remains essential to also Increasing both health and lifespan of humans is one of analyze the protein, which ultimately executes biological the prime goals of the modern society. In order to better functions. understand age‐related chronic lung diseases such as COPD, lung cancer or fibrosis, intense efforts in integrated
8 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
a -1 0 1 b 0 10 100 Differential gene expression Cell type resolved DGE analysis logFC - alveolar macrophages % FDR 391 significantly regulated genes (FDR < 10%) [old / young] 45 Ear2 Alveolar_macrophage
non-classical_monocyte_(Ly6c2-) 40 B_cells Cd63 Mesothelial_cells 35 Type_2_pneumocytes Lyz2 vascular_endothelial_cells Capillary_endothelial_cells 30 Wfdc17
Ciliated_cells Cebpb
Club_cells 25 Psap Ctsk Goblet_cells Ear1 Lrrc58
Interstitial macrophages 20 S100a11 Cd103+/Cd11b-_dendritic_cells Fabp1 mt-Nd4 p-value [-log10] p-value Alox5ap mt-Cytb Awat1 Eosinophil_granulocyte Ctss 15 Tyrobp Fstl1 Il18 Ctsd Cd4+_T_cells Slc25a5 S100a10 Ctsz Spp1 Atp1b1 H2-K1 Gpnmb CD8+_T_cells F7 Lgmn Ctsb S100a1 10 Marco Fxyd5 Gpx1 Igf1 Mki67+_proliferating_cells Ccl3 Natural_Killer_cells Mfge8
Interstitial_Fibroblast 5 Lipofibroblast
lymphatic_endothelial_cells 0 Type1_pneumocytes Vcam1+_endothelial_cells -0.8 -0.4 0 0.4 0.8 1 1.4 1.8 Smooth_muscle_cells Average fold change [log2; old / young] Fn1+_macrophage Plasma_cells c 0 10 100 classical_monocyte_(Ly6c2+) Differential gene expression Ccl17+/Cd103-/Cd11b-_dendritic_cells - alveolar type 2 pneumocytes % FDR CD209+/Cd11b+_dendritic_cells Megakaryocytes mt-Nd3
Neutrophil_granulocytes 60 Scd1 Gm26924
H2-K1 50 Malat1 H2-Q7
deH2−K1 Scd1 40 B2m 5 6 Sftpb Atp1b1 4 30 mt-Nd2 Ctsh Npc2
4 p-value [-log10] Sfta2 Sftpa1 H2-D1 3 Chchd10 Lyz1 Sftpd Ddx3y
20 Ppp1r14c Pigr App H2-Q6 2 Sparc Eif2s3y Gm10800 2 S100g Napsa Wfdc17 Ly6i Atrx Lrg1
UMI counts [log2] 1 Suclg1 Nrn1 Srgn UMI counts[log2] Ppbp 10 Cd59a Retnlg Il1bScd2 0 0 Inmt young old young old 0
-0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 Average fold change [log2; old / young]
fg40 Epithelium Endothelium Stromal cells old 0.1404 young 30
20 Epcam
H2-K1 [MFI] 0.0553 10 0.0261 CD31 H2-K1 0
cells l a Epithelium rom Endothelium St
Figure 4. Cell type specific differential gene expression plots depict normalized expression values of two exemplary analysis. (a) Heatmap displays fold changes derived from the genes H2‐dK1 (d) an Scd1 (e) across age groups for type‐2 cell type resolved differential expression analysis. Rows and pneumocytes. (f) The indicated cell lineages were gated by flow columns correspond to cell types and genes, respectively. cytometry as shown in the left panel in a CD31 and Epcam co‐ Negative fold change values (blue) represent higher expression staining and evaluated for H2‐K1 expression on protein level. in young compared to old. Positive fold change values are The histograms show fluorescence intensity distribution of eth colored in pink. (b, c) Volcano plots visualize the differential H2‐K1 cell surface staining for the indicated lineages and age gene expression results in Alveolar macrophages (b) and groups. (g) Boxplot shows mean fluorescence intensity for H2‐K1 alveolar type‐2 pneumocytes (c). X and Y axes show average in the indicated cell lineages across four replicates of mice. P‐ log2 fold change and ‐log10 p‐value, respectively. (d, e) Violin values are from an unpaired, two‐tailed t‐test.
The example of basement membrane collagen IV genes upregulated on the protein level serves as a reminder that were all downregulated on the mRNA level but that mRNA analysis should always be taken with a grain of
9 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
salt. Next to mass spectrometry based methods, novel age, indicating that aging‐related ECM remodeling does single cell methods combining mRNA and protein not involve large differences in covalent protein analysis, such as CITE‐seq43, will become ever more crosslinks. However, we observed a few very strong important in the near future. We show that the changes in the ECM, which were completely novel ine th combination of single cell‐resolved mRNA analysis and context of aging and are open for future investigation into bulk proteomics is highly complementary by using the their functional implications. single cell expression data to understand the most likely cellular origin of proteins that showed altered abundance In order to stabilize the alveolar structure during with age. Spatial transcriptomics methods for high breathing‐induced expansion and contraction, type‐2 throughput detection of transcripts in single cells in situ pneumocytes produce and secrete pulmonary surfactant, are currently quickly evolving44, 45. Traditional antibody which is a thin film of phospholipids and surfactant based methods for single cell protein analysis in situ are proteins41. The lipid composition of pulmonary surfactant however not well multiplexed and do not easily scale for has been shown to change with age50, and electron high throughput. Thus, to fully develop the enormous microscopy of surfactant and the lipid loaded lamellar potential of single cell multi‐omics data integration bodies in type‐2 pneumocytes revealed ultrastructural (M.Colomé‐Tatché & F.J.Theis 2018; disorganization with age51. This may be related to our https://doi.org/10.1016/j.coisb.2018.01.003), the field finding that cholesterol biosynthesis and neutral lipid depends on current and future developments in methods content is upregulated in type‐2 cells of old mice. It is for simultaneous protein and mRNA analysis on single cell currently unclear at which level the homeostasis of lipid level in situ46, 47. metabolism is altered in the aged lungs. We found strong similarity of the aged type‐2 phenotype with the We analyzed the foundations of lung tissue architecture phenotype in Insig1/2 knock out mice that accumulated by quantifying compositional and structural changes in neutral lipids, accompanied by lipotoxicity‐related lung the aged extracellular matrix using state of the art mass inflammation and tissue remodeling42. Thus, it is possible spectrometry workflows. The ECM is not only key as a that part of the chronic inflammation we observed in the scaffold for the lungs overall architecture, but also an aged lung is influenced by deregulation of lipid important instructive niche for cell fated an phenotype24, homeostasis. The inflammatory phenotype may also be 48. Recent proteomic studies have demonstrated that at related to epithelial senescence, as mice with a type‐2 least 150 different ECM proteins, glycosaminoglycans and pneumocyte specific deletion of telomerase, and thus modifying enzymes are expressed in the lung, and these premature aging with increased senescence in these cells, assemble into intricate composite biomaterials that are developed a pro‐inflammatory tissue microenvironment characterized by specific biophysical and biochemical and were less efficient in resolving acute lung injury52. properties27, 28, 49. Due to this complexity of the ECM, both in terms of composition and posttranslational In summary, we have demonstrated that the lung aging modification, and the assembly of ECM proteins into atlas presented here contains a plethora of novel supramolecular structures, it is presently unclear on information on molecular and cellular scale and serves as which level and how exactly the aging process affects the a reference for the large community of scientists studying lungs´ ECM scaffold. We used detergent solubility chronic lung diseases and the aging process. The data also profiling to screen for differences in protein crosslinking serves as analysis template efor th currently ongoing and complex formation within the ECM. Surprisingly, worldwide collection of data for a Human Lung Cell Atlas, most solubility profiles were not significantly altered with which will hopefully enable analysis of aging effects in human lungs.
10 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
abAnnotation enrichment - alveolar macrophages (FDR < 5%) Annotation enrichment - type-II pneumocytes (FDR < 5%)
antigen processing and presentation via MHC1 MHCI Peptidetransport Cholesterolbiosynthesis proton-transporting V-type ATPase Bromodomain positive regulation of T cell mediated cytotoxicity phagocytic cup Sterolbiosynthesis Glycosphingolipid biosynthesis - globo series Steroidbiosynthesis endocytic vesicle membrane Integrin brush border membrane Adaptiveimmunity Other glycan degradation Antiviraldefense MHC class I protein complex Chemotaxis protein phosphatase type 2A complex Nitration Myristate positive regulation of osteoblast differentiation Proto-oncogene regulation of leukocyte mediated cytotoxicity Immunity protein serine/threonine phosphatase complex Proteinphosphatase Graft-versus-host disease Methylation MHC protein complex Lipidbiosynthesis unsaturated fatty acid metabolic process icosanoid metabolic process SH2domain negative regulation of NF-kappaB Innateimmunity pigment granule Inflammatoryresponse positive regulation of proteasomal degradation Mitochondrion cellular response to organic cyclic compound Ribonucleoprotein ceramide metabolic process Mitochondrioninnermembrane translation Proteasome Transitpeptide cellular component biogenesis at cellular level Extracellularmatrix Propanoate metabolism Ribosomalprotein Mismatch repair Sensorytransduction Ribosomebiogenesis Collagen RNA polymerase Vision mitochondrial proton-transporting ATP synthase complex Basementmembrane DNA-directed RNA polymerase II, core complex organellar small ribosomal subunit Respiratorychain mitochondrial small ribosomal subunit Hydrogeniontransport peroxisomal matrix Keratin microbody lumen Selenium Gamma-carboxyglutamicacid small nucleolar ribonucleoprotein complex -0.8 -0.4 0 0.4 0.8 proteasome core complex, alpha-subunit complex aminoacyl-tRNA synthetase multienzyme complex Enrichment score -1 -0.5 0 0.5 1 [old / young] Enrichment score [old / young] d c Epithelium - Upstream regulator analysis Mesenchyme - Upstream regulator analysis
Type 2 pneumocytes capillary EC Ciliated cells Lipofibroblast Goblet cells Mesothelial cells Club cells Interstitial fibroblast Vcam1+ EC IL3 IL4 RB1 TNF APP CD3 IL1B AGT MYC
IFNG TP53 Vascular EC CST5 SCAP MTOR TGFB1 INSIG1 ERBB2 IL10RA PPARA PPARG NFE2L2 RICTOR SREBF2 Lymphatic EC Smooth muscle cells e FN1 TNF ERK FOS IL1B
Myeloid cells - Upstream regulator analysis IGF1 TP73 E2F1 KLF3 CST5 KRAS HRAS IKBKB MYCN TGFB1 IL10RA CEBPB RICTOR PDGF BB
Alveolar macrophages Non-classical monocytes Ccl17+ dendritic cells -4 0 4 Interstitial macrophages Classical monocytes Z score CD209+ dendritic cells [old / young] Fn1+ macrophages CD103+ dendritic cells Cg TNF APP Ccl2 Ifnar IL1B OSM TP53 IFNG NFkB CST5 PTEN STAT1 TCL1A STAT3 TGFB1 PSEN1 ERBB2 HNF4A IL10RA MYD88 ERK1/2 RICTOR
Figure 5. Cell type specific pathway and upstream regulator the observed gene expression changes for (c) epithelial, (d) analysis. (a, b) The bar graph shows the result of a gene mesenchymal, and (e) myeloid cells. Cell types and regulators annotation enrichment analysis for (a) alveolar macrophages, were grouped by unsupervised hierarchical clustering (Pearson and (b) type‐II pneumocytes, respectively. Gene categories with correlation) and the indicated transcriptional regulators and positive (upregulated in old) and negative scores cytokines, growth factors and ECM proteins are color coded (downregulated in old) are highlighted in red d an blue based on the activation score as shown. respectively. (c‐e) Upstream regulators are predicted based on
11 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
abINSIG1 upstream regulator network Type 2 pneumocyte c in type-II pneumocytes Lipofibroblast higher in old Annotation enrichment analysis (FDR<5%)
Cholesterolbiosynthesis
Sterolbiosynthesis
Myristate
Immunity
Secreted
Mitochondrion
Ribonucleoprotein
Mitochondrioninnermembrane
translation
Ribosomalprotein
higher in old Extracellularmatrix lower in old Collagen
-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 Enrichment score [old / young]
d e CD45- cells young old FSC hi FSC lo 100 100 old 80 80 young
60 60
40 40
20 20
0 0 2 3 4 5 0102 10 3 10 4 10 5 010 10 10 10 Nile Red Nile Red Nile Red
f CD45- / FSC hi g CD45- / FSC lo 3000 25000 p = 0.07 p = 0.01
2500 20000 2000
15000 1500 NileRed (MFI) NileRed (MFI)
10000 1000 young old young old
Figure 6. Increased cholesterol biosynthesis in type 2 nodes. (d) Nile red staining of neutral lipids in cells from a whole pneumocytes and lipofibroblasts in aged mice. (a) The graph suspension depleted for CD45+ leukocytes (CD45 lineage shows genes known to be negatively regulated by Insig1 that negative cells). (e) Quantification of nile red stainings using flow were found to be upregulated in type‐2 pneumopcytes of old cytometry. Histograms show flow cytometry analysis of Nile Red mice. (b) Selected gene categories found to be significantly (FDR in aged (red) and young (blue) mice, unstained control is <5%) up‐ (positive enrichment scores) or downregulated represented in gray. Cells were stratified by size in bins of large (negative enrichment scores) in the indicated cell‐types. (c) (FSC hi) and small (FSC lo) cells using the forward scatter. (f, g) Segment of the cholesterol biosynthesis pathway. Diamant Nile Red mean fluorescence intensity (MFI) quantification across shaped nodes represent enzymes that were found to be three individual mice for (f) CD45 negative and forward scatter upregulated in type‐2 pneumocytes of old mice. The (FSC) high, and (g) FSC low cells. P‐values are from an unpaired, biochemical intermediates are named in between the enzyme two ‐tailed t‐test.
12 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
Supplementary data and Methods
abyoung muc3838 old muc3839 muc3840 muc3841 muc4166 muc4167 muc4168
muc4169 tSNE 2 tSNE 2 tSNE muc4170 muc4172 muc4173 muc4174 muc4175 muc4654 muc4657
tSNE 1 tSNE 1 c T Cell_Cd8b1 high 00.020.01000.01000.150.2300.0100000000.01000.010.02000000000 Stromal cell_Inmt high 0 0 0.01 0 0.04 0 0 0 0 0 0.01 0 0.02 0.02 0.02 0.01 0.35 0.44 0 0.05 0.02 0.15 0 0 0 0 0 0 0.18 0 0.04 0.03 0.05 Stromal cell_Dcn high 0.01 0 0.01 0 0.03 0 0 0 0.01 0 0.01 0.01 0 0.02 0.02 0 0.54 0.23 0 0.03 0.01 0.17 0 0 0.02 0 0 0 0.12 0 0.03 0.03 0.06 Stromal cell_Acta2 high 0 0.01 0.02 0 0.03 0 0.01 0.01 0.02 0 0.01 0.02 0 0.01 0.02 0 0.15 0.14 0 0.05 0.03 0.1 0.01 0.01 0.02 0.01 0 0 0.34 0 0.03 0.02 0.05 Plasmacytoid dendritic cell 0.07 0.02 0.12 0.04 0 0.01 0.06 0.09 0.03 0.01 0 0.43 0 0.08 0 0 0.01 0.01 0.01 0.01 0 0.02 0 0.03 0.05 0.18 0 0 0.01 0 0 0 0.01 Nuocyte 0 0.05 0.02 0 0 0.01 0.02 0.01 0.35 0.1 0 0.02 0 0 0.01 0 0 0.01 0 0.01 0 0 0 0.08 0.02 0.02 0 0 0.01 0 0 0 0 NK Cell 0.01 0.03 0.01 0.01 0 0.01 0.01 0 0.15 0.08 0 0.01 0 0 0.01 0 0 0.01 0 0 0 0 0.01 0.49 0.01 0.02 0 0 0.01 0 0 0 0 Neutrophil granulocyte 0.02 0.01 0.02 0.29 0 0.02 0.02 0.03 0.01 0.01 0.01 0.05 0 0.04 0 0.01 0.01 0 0.01 0 0.01 0.01 0.01 0.01 0.15 0.03 0.01 0 0 0.01 0 0 0 Monocyte progenitor cell 0 0.01 0.01 0.09 0 0 0 0 0 0.01 0 0.03 0 0 0 0 0 0 0 0 0 0 0.16 0.01 0.01 0.01 0.01 0 0 0 0 0 0 Interstitial macrophage 0.03 0.03 0.45 0 0 0.05 0.05 0.12 0.01 0 0 0.05 0 0.12 0.01 0 0 0 0.01 0.01 0.01 0 0 0 0.02 0.03 0 0 0 0 0 0 0 Ig−producing B cell 0 0.04 0 0 0 0.01 0.01 0.01 0 0 0 0 0 0.01 0 0 0 0 0.06 0 0 0 0 0 0 0 0.4 0.02 0 0 0 0 0.01 Eosinophil granulocyte 0.04 0 0.03 0.03 0 0.04 0 0.01 0 0 0 0.06 0 0.06 0.01 0 0.01 0.01 0.01 0.01 0.03 0.01 0 0 0.28 0.06 0.01 0.01 0 0 0.01 0.01 0.01 Endothelial cells_Vwf high 0.01 0 0 0 0.19 0 0 0 0 0 0.01 0.01 0 0.03 0.01 0 0.03 0.01 0 0.08 0.02 0.05 0 0 0.02 0.02 0 0 0.03 0 0.03 0.31 0.44 matchSCore Endothelial cell_Tmem100 high 0 0 0 0 0.29 0 0 0 0 0 0.01 0 0 0.02 0.01 0 0.02 0.01 0 0.09 0.02 0.02 0 0 0.01 0.01 0 0 0 0 0.02 0.43 0.28 Endothelial cell_Kdr high 0 0 0.01 0 0.49 0 0 0 0 0 0 0 0 0.01 0.02 0 0.02 0.02 0 0.08 0.02 0.04 0 0 0 0.01 0 0 0.01 0 0.05 0.31 0.18 0.5 Dividing T cells 0 0 0.01 0.04 0 0.01 0.01 0.01 0.17 0.1 0 0.01 0 0.01 0.01 0 0 0 0 0.01 0 0 0.3 0.05 0 0.01 0.01 0 0.01 0 0 0 0 0.4 Dividing dendritic cells 0.02 0.02 0.06 0.04 0 0.06 0.25 0.11 0.01 0.01 0.01 0.03 0 0.02 0.01 0 0 0 0 0.01 0 0 0.22 0.02 0.01 0.02 0 0 0.01 0 0 0 0 0.3 Dividing cells 0000.03000000.010.01000000000.01000.350000000000 0.2 Dendritic cell_Naaa high 0.04 0.04 0.07 0.01 0 0.08 0.48 0.21 0.03 0.01 0 0.06 0 0.02 0 0 0 0 0 0 0.01 0.01 0 0.02 0.01 0.02 0 0 0.01 0 0 0 0.01 0.1 Conventional dendritic cell_Tubb5 high 0.01 0.03 0.06 0.03 0 0.09 0.1 0.18 0.01 0 0 0.02 0 0.02 0 0 0 0 0 0.01 0 0 0.18 0 0.01 0.01 0 0 0.01 0 0 0 0 0.0 Conventional dendritic cell_Mgl2 high 0.02 0.03 0.09 0.01 0 0.18 0.13 0.26 0.01 0 0 0.05 0 0.02 0 0 0.01 0 0.01 0 0 0 0 0 0.03 0.03 0 0 0 0 0 0 0 Conventional dendritic cell_H2−M2 high 0.01 0.02 0.03 0 0 0.4 0.06 0.05 0.03 0.03 0.01 0.02 0 0 0 0 0 0.01 0 0 0.01 0.01 0 0.03 0 0.04 0 0 0 0.01 0.02 0 0 Conventional dendritic cell_Gngt2 high 0.07 0.01 0.09 0.04 0.01 0.02 0.02 0.02 0.02 0 0 0.23 0 0.08 0 0 0 0 0 0.01 0 0.01 0 0.01 0.07 0.37 0 0 0.01 0 0 0.01 0.01
MouseCell Atlas (lung) MouseCell Atlas Clara Cell 0.01 0 0 0.01 0 0 0 0 0 0 0.03 0 0.38 0.02 0 0.33 0 0.02 0.01 0.01 0 0.01 0 0 0.01 0 0 0 0 0.01 0.01 0 0 Ciliated cell 0 0 0 0 0.01 0 0 0 0 0 0.46 0 0.03 0 0.01 0.03 0.01 0 0 0 0 0.01 0 0 0 0 0 0 0 0 0.01 0.01 0.01 Basophil 0.01 0 0.02 0.01 0 0.02 0 0.01 0.02 0.01 0 0.02 0 0.01 0.01 0 0 0 0 0 0.01 0.01 0.01 0.01 0.06 0.02 0 0 0 0 0 0 0 B Cell 00.240000.0100.010.020.01000000000.0400000.010.0100.08000000 AT 2 Cell 0 0 0 0.01 0 0.01 0 0 0 0 0.01 0 0.01 0 0 0.01 0 0 0 0 0 0 0 0 0.01 0 0 0 0.01 0.57 0.01 0 0.01 AT 1 Cell 0 0 0 0 0.07 0.01 0 0 0 0 0.03 0 0.01 0 0.02 0 0.03 0.05 0 0.03 0.01 0.07 0 0 0.01 0.01 0 0 0.02 0.03 0.49 0.04 0.03 Alveolar macrophage_Pclaf high 0.23 0 0.02 0.04 0.01 0.01 0 0 0 0 0 0.02 0 0.02 0 0 0.01 0 0.02 0 0 0 0.17 0 0.01 0 0 0 0 0 0 0.01 0 Alveolar macrophage_Ear2 high 0.47 0 0.02 0.01 0 0.01 0 0 0 0 0 0.03 0 0.05 0 0 0.01 0 0.02 0 0 0.01 0 0 0 0.01 0 0 0 0 0 0 0.01 Alveolar bipotent progenitor 00000.010.0200000.010000000000.010.010000000.010.060.0100 t t ge lls es es lls lls lls lls lls lls lls +) lls ge lls s s lls es lls lls lls te −) lls lls lls es es lls lls a ce ag yt ce ce ce ce ce ce ce c2 ce a ce bla bla ce yt ce ce ce cy c2 ce ce ce yt yt ce ce ph _ h loc l_ c_ c_ c_ _ _ _ y6 _ ph t_ ro ro l_ oc l_ g_ r_ lo y6 a_ _ e_ oc oc l_ l_ ro B op u lia iti iti iti _T _T ted (L lub ro le ib ib lia ry lia in lle nu (L m od cl m m lia lia ac cr an e d r dr dr 4+ 8+ ia _ C ac ob F of e ka he at Ki ra _ s lo us eu eu e e m a gr oth n n n d D il te m G al_ ip oth a ot er l_ _g te la _b m n n oth oth r_ l m _ d de de de C C C cy +_ titi L d eg s lif ra il cy P ed _ _p _p d d la tia p+ en −_ −_ +_ no 1 rs en M Me ro tu ph no r th 2 1 en en eo ti g y_ b b b o Fn te c_ _p a tro o oo e_ pe r_ _ lv rs /N r 11 11 11 _m In ti + N u m m yp Ty la 1+ A te p+ illa d d d al ha i67 Ne al_ S T cu m In m ap /C /C /C ic p k ic as ca Ca C 3− 3+ 9+ ss lym M ss v V 10 10 20 cla cla Cd Cd CD n− +/ no l17 Cc Cell types in this study
Supplementary Figure S1. Excellent technical reproducibility visualization of overlap between cell type markers derived from and correct cell type annotation. (a, b) tSNE visualization the Mouse Cell Atlas (lung) and this study (c) . Red colors colored by mouse sample (a) and age group (b). machSCore indicate higher matchSCore values.
13 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
a real values b 5138 proteins from whole lung tissue (old n=4, young n=4) data imputation Frem2 rep1 rep1 8
7 Pbx1 Counts Counts
6 Get4 rep2 rep2 Nnt Ecscr Spcs3 Ccdc77
5 Ehbp1 Cd55 A1bg Ankle2 Lrg1 C1qtnf5 C9 Counts Counts Upk3b Hpx Serpina1e Sept1 Pclo Col16a1 4 Serpinb9b Col14a1 Flt1 Ebna1bp2 Sacs Frem1 C8b rep3 rep3 Col6a4
P-value [-log10] C6 3 Fmod Wnt9a Mcpt8
Counts Counts Fras1 Nfia C7 2 MgpPrdm16 Cxcl12 rep4 rep4 1 Counts Counts 0
3025 35 2530 35 -8 -6 -4 -2 0 2 4 6 8 MS-intensity [old] MS-intensity [young] Student's T-test Difference [old / young; log2] c Cilp2 Lox Agrn Fmod Serpinb9b Ecm2 Sdc4 Col14a1 Frem2 Col6a4 Frem1 Ltbp2 Col6a6 Fras1 Mfap5 Col6a5 Cxcl12 Mgp C1qc Ambp Itih3 Sema7a Vtn F13b Mfge8 Col4a3 Hpx Lrg1 Col16a1 Serpina1e Wnt9a C1qtnf5 old_1 old_2 old_3 -1.5 0 1.5 old_4 young_1 Z score young_2 [old / young] young_3 young_4
32 significantly regulated Matrisome proteins (FDR <10%)
Supplementary Figure S2. Whole lung tissue proteomes of discovery rate < 10% are highlighted in red in the vulcano plot mouse lung aging. (a) The histogram shows the normal showing the indicated fold changes and p‐values from t‐test distribution of mass spectrometry (MS) intensities for the statistic. Matrisome proteins are shown as rectangles and all quantified proteins in all replicates and experimental groups. other proteins as filled circles. (c) The z‐score values of 32 For t‐test statistics the missing values (proteins not detected in significantly regulated extracellular matrix proteins were the respective sample) were replaced by a normal distribution grouped by unsupervised hierarchical clustering (Pearson of random values at the lower end of the distribution of real correlation) values (data imputation). (b) Proteins regulated with a false .
14 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
a bcP < 0.005 Relative difference in proportion (%) − − 10 10 0 5 5
Club cells 0.2 0.0 0.2 MDS 2 MDS MDS 1 MDS −
0.4 0.0 0.4 Ciliated cells − 0.4 young
− old
−0.6 0.0 0.6 0.8
− more more MDS 1 3m 24m in young in old
d young
CC10 Foxj1 old
CC10 Foxj1
efgCiliated cells Club cells p = 0.0001 p = 0.025 p = 0.001 4 80 120 100 3
60 80 2 60 40 1
# cells / 100 µm / cells # 40 # cells / 100 µm / # cells
20 20 Ratio Club/Ciliatedcells 0 young old young old young old
Supplemental Fig S3. Cell type frequency analysis reveals an ciliated cells were stained using a CC10 and Foxj1 antibody altered ratio of club to ciliated cells in airways of old mice. (a) respectively. (e, f) The box plots show the quantification of (e) The MDS plot shows the mouse‐wise euclidean distances of cell Ciliated and (f) Club cells from counting 2647 cells in 14 airways type proportions for the two age groups (b) The box plot shows of two mice of each age group. (g) Ratio of ciliated to club cells the significant difference in euclidean distance of cell type in 14 airways. P‐values are from an unpaired, two‐tailed t‐test proportions. (c) Relative difference in proportion in percent of using Welch´s correction. all cell in the dataset for the indicated cell types. (d) Club and
15 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
a Lung cell type signatures - cell type and gene specifc queries, list of most significant marker genes for 30 cell types
b Lung aging protein - cellular source, protein abundance, and protein solubility
c Lung aging mRNA - cellular source, mRNA volcano plot [old/young], and UMI counts
d Lung aging - annotation enrichments and pathways
Supplementary Figure S4. A user friendly and interactive interest. A cell type query produces a list of top marker genes webtool to navigate the Lung Aging Atlas. (a) The first tab `Lung for the cell type of interest (right panel). (b) The panel `Lung cell type signatures´ provides a cell type dotplot (left panel) and aging protein´ features a dot plot to illustrate the most likely a color coded tSNE map (middle panel) for gene specific queries cellular source of the protein of interest (left panel), a box plot and illustrates cell type specific expression of any gene of to show alterations in total lung tissue protein abundance in old
16 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
mice (middle panel), and a line plot to show protein solubilities. illustrating mRNA abundance in young and old mice. The dot Protein solubility is measured by relative quantification of and violin plot are navigated with the gene specific query, while protein abundance across four fractions. Fraction 1 (FR1) the volcano plot requires navigation via the cell type query. The contains proteins with highest and fraction 4 (ECM) with lowest volcano plot has a toggle over function that allows identification solubility. Curves that peak on the right (ECM) thus represent of genes and can thus be used to browse through differential insoluble proteins. (c) The tab `Lung aging mRNA´ again features gene expression between young and old cells of any cell type of the dotplot (left panel), a volcano plot that shows fold changes interest. (d) In the tab `Lung aging ‐ annotation enrichments´, [old/young] on the x‐axis and ‐log10 p‐values on the y‐axis the gene annotation enrichments between old and young can (middle panel), and a violin plot of the log2 UMI counts be browsed for all 30 cell types.
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
Methods purified twice by 0.6x clean‐up beads (CleanNA). Prior to tagmentation, cDNA samples were loaded on a DNA High Generation of single cell suspensions from whole mouse Sensitivity Chip on the 2100 Bioanalyzer (Agilent) to lung ensure transcript integrity, purity, and amount. For each Lung tissue was perfused with sterile saline from the right sample, 1 ng of pre‐amplified cDNA from an estimated to the left ventricle of the heart and subsequently inflated 1000 cells was tagmented by Nextera XT (Illumina) with a via a catheter in the trachea, by an enzyme mix custom P5 primer (Integrated DNA Technologies). Single containing dispase (50 caseinolytic units/ml), collagenase cell libraries were sequenced in a 100 bp paired‐end run (2 mg/ml), elastase (1 mg/ml), and DNase (30 ug/ml). on the Illumina HiSeq4000 using 0.2 nM denatured After tying off the trachea, the lung was removed and sample and 5% PhiX spike‐in. For priming of read 1, 0.5 immediately minced to small pieces (approx. 1 mm2). The µM Read1CustSeqB (primer sequence: tissue was transferred into 4 ml of enzyme mix for GCCTGTCCGCGGAAGCAGTGGTATCAACGCAGAGTAC) was enzymatic digestion for 30 min at 37°C. Enzyme activity used. was inhibited by adding 5 ml of PBS supplemented with 10% fetal calf serum. Dissociated cells in suspension were Bioinformatic processing of scRNA‐seq reads passed through a 70 μm strainer and centrifuged at 500 x The Dropseq core computational pipeline was used for g for 5 min at 4°C. Red blood cell lysis (Thermo Fischer 00‐ processing next generation sequencing reads of the 4333‐57) was done for 2 min and stopped with 10% FCS scRNA‐seq data, as previously described15. Briefly, STAR in PBS. After another centrifugation 5 min at 500 x g (4°C) (version 2.5.2a) was used for mapping54. Reads were the cells were counted using a Neubauer chamber and aligned to the mm10 genome reference (provided by critically assessed for single cell separation and viability. Drop‐seq group, GSE63269). For barcode filtering, we 250.000 cells were aliquoted in 2.5 mL of PBS excluded barcodes with less than 200 genes detected. supplemented with 0.04% of bovine serum albumin and loaded for Drop‐Seq at a final concentration of 100 Single cell data analysis cells/μL. After constructing the single cell gene expression count matrix, we used the R package Seurat55 and custom Single cell RNA sequencing scripts for analysis. Drop‐seq experiments were performed largely as described previously15, 16, 53, with few adaptations during Quality control. A high proportion of transcript counts the single cell library preparation. Briefly, using a derived from mitochondria‐encoded genes can indicate microfluidic PDMS device (Nanoshift), single cells (100/µl) low cell quality. Similarly cells with unusually high UMI from the lung cell suspension were co‐encapsulated in count levels may represent cell doublets. Therefore we droplets with barcoded beads (120/µl, purchased from removed cells with >10% mitochondria derived counts ChemGenes Corporation, Wilmington, MA) at rates of and >5000 total UMI counts from the downstream 4000 µl/hr. Droplet emulsions were collected for 15 analysis. min/each prior to droplet breakage by perfluorooctanol (Sigma‐Aldrich). After breakage, beads were harvested Unsupervised clustering and visualization. Highly variable and the hybridized mRNA transcripts reverse transcribed genes were defined within each mouse sample (young, (Maxima RT, Thermo Fisher). Unused primers were n=8; old n=7) separately following Seurat standard removed by the addition of exonuclease I (New England approach. Next, genes appearing in >4 mouse samples in Biolabs), following which beads were washed, counted, the set of highly variable genes were defined as a set of and aliquoted for pre‐amplification (2000 beads/reaction, consensus highly variable genes. To minimize the effect of equals ~100 cells/reaction) with 12 PCR cycles (primers, cell‐cycle on clustering we removed cell‐cycle genes56 chemistry, and cycle conditions identical to those from the set of consensus highly variable genes. All 14813 previously described15. PCR products were pooled and cells passing quality control were merged into one count
17 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
matrix and normalized and scaled using Seurat’s between gene signatures was evaluated using Fisher’s NormalizeData() and ScaleData() functions. The reduced Exact test. set of consensus highly variable genes was used as the feature set for independent component analysis using Mouse Cell Atlas integration. Cell type marker signatures Seurat’s RunICA() function. The first 50 independent in our data (Table S1) were compared to cell type marker components were used for tSNE visualization and Louvain signatures in the Mouse Cell Atlas (MCA)13. MatchSCore clustering using the Seurat functions RunTSNE() and (Mereu et al 2018, bioRxiv doi: FindClusters(), respectively. https://doi.org/10.1101/314831) was used to quantify overlap between cell type marker signatures derived from Cell type marker discovery. The Seurat FindAllMarkers() our study and the MCA. Marker genes with adjusted p‐ function was used to identify cluster specific marker value < 0.1 and average log fold change > 1 were genes. Based on manual annotation and with guidance of considered. the enrichment analysis (see below) the 36 clusters were assigned to 30 cell type identities. Using the annotation of Quantifying transcriptional noise. Transcriptional noise in cell type identities the FindAllMarkers() function was the gene expression profiles was quantified in a called to identify the final set of celle typ markers used multivariate fashion following previous work21. Briefly, throughout this analysis. the euclidean distance between the gene expression profile of each cell and the average expression profile of Ambient RNA identification. An important technical detail the respective cell type was calculated for each mouse. needed our attention and is briefly described here. As Median euclidean distance was used to summarize infrequently discussed in the community but not yet transcriptional noise at the mouse level. To statistically addressed, we also observed `ambient mRNA´ effects, assess the association between transcriptional noise and which we believe are the consequence of free mRNA age within each cell type Wilcoxon’s rank sum test was released from dying cells hybridizing with beads in used. To statistically assess the global association droplets during the microfluidic capture of single cells in between transcriptional noise and age, analysis of the Dropseq workflow. The ambient mRNAs are typically variance was used In particular, transcriptional noise was derived from highly abundant transcripts and this artefact modelled as the dependent variable with cell type as a is inherent to all droplet based methods (including the covariate and age as the explanatory variable. commercially available 10x platform). Here, it can be exemplified by the Scgb1a1 gene in Figure 1C that is Cell type yfrequenc analysis. Cell type frequencies were known to be highly specific for Club and Goblet cells but calculated based on the counts of cells annotated to each was nevertheless detected in almost 100% of the cells in cell type for each mouse. Counts were transformed to our data. However, the UMI count levels were much proportions using the DR_data() function of the higher in Club and Goblet cells (representing the real DirichletReg R package which causes the values to shrink source of expression) indicating that the mRNA counts away from extreme values of 0 and 1. Next, the mouse‐ observed in all other clusters were ambient mRNA wise euclidean distances were calculated based on these background. To independently confirm this we therefore proportions using the dist() R function followed by determined all genes that showed ambient mRNA multidimensional scaling using the isoMDS() R function. background by analyzing the identity of genes on beads at To statistically assess the association between age and the tailend of the total UMI count distribution (on the first coordinate derived from the multidimensional average 10 UMIs per barcode), representing empty beads scaling Wilcoxon test was applied. Relative changes in cell that were never in contact with a real cell but type frequencies were calculated by subtracting the nevertheless contain information from free floating median ecell typ proportion of the young mice from the ambient mRNA. We identified 153 genes (Table S8) with cell type proportions of the old mice. an `ambient mRNA´ effect and accounted for this effect in Cell type resolved differential expression analysis. Cell downstream analysis. type resolved differential expression analysis was performed using the Seurat differential gene expression Cell type annotation. To aid the assignment of cell type to testing framework. Within each cell type cells were clusters derived from unsupervised clustering, we grouped by age and differential testing performed using performed cell type enrichment analysis. Cell type gene the Seurat FindMarkers() function. To account for the signatures obtained from bulk level gene expression were effect of ambient mRNAs, testing of ambient mRNAs was downloaded from the ImmGen and xCell resources. Each limited to cell types in which the ambient mRNA was a gene signature obtained from our clustering was cell type marker gene. statistically evaluated for overlap with gene signatures contained in these two resources. Mouse gene symbols Cell type resolved pathway analysis. The 1D annotation were capitalized to map to human gene symbols. Overlap enrichment analysis23 was done with the freely available
18 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
software package Perseus57, as previously described28. To coupled to a Q‐Exactive Mass Spectrometer (Thermo predict the activity of upstream transcriptional regulators Scientific). Peptides were loaded with buffer A (0.1% (v/v) and growth factors based on the observed gene formic acid) and eluted with a nonlinear 240‐min gradient expression changes, we used the Ingenuity® Pathway of 5–60% buffer B (0.1% (v/v) formic acid, 80% (v/v) Analysis platform (IPA®, QIAGEN Redwood City, acetonitrile) at a flow rate of 250 nl/min. After each www.qiagen.com/ingenuity) as previously described28. gradient, the column was washed with 95% buffer B and The analysis uses a suite of algorithms and tools re‐equilibrated with buffer A. Column temperature was embedded in IPA for inferring and scoring regulator kept at 50 °C by an in‐house designed oven with a Peltier networks upstream of gene‐expression data based on a element58 and operational parameters were monitored in large‐scale causal network derived from the Ingenuity real time by the SprayQc software59. MS data were Knowledge Base. The analytics tool ‘Upstream Regulator acquired with a shotgun proteomics method, where in Analysis’22 was used to compare the known effect each cycle a full scan, providing an overview of the full (transcriptional activation or repression) of a complement of isotope patterns visible at that particular transcriptional regulator on its target genes to the time point, is follow by up‐to ten data‐dependent MS/MS observed changes to assign an activation Z‐score. Since it scans on the most abundant not yet sequenced isotopes is a priori unknown which causal edges in the master (top10 method)60. Target value for the full scan MS network are applicable to the experimental context, the spectra was 3 × 106 charges in the 300−1,650 m/z range ‘Upstream Regulator Analysis’ tool uses a statistical with a maximum injection time of 20 ms and a resolution approach to determine and score those regulators whose of 70,000 at m/z 400. Isolation of precursors was network connections to dataset genes as well as performed with the quadrupole at window of 3 Th. associated regulation directions are unlikely to occur in a Precursors were fragmented by higher‐energy collisional random model22. In particular, the tool defines an overlap dissociation (HCD) with normalized collision energy of 25 P‐value measuring enrichment of network‐regulated % (the appropriate energy is calculated using this genes in the dataset, as well as an activation Z‐score percentage, and m/z and charge state of the precursor). which can be used to find likely regulating molecules MS/MS scans were acquired at a resolution of 17,500 at based on a statistically significant pattern match of up‐ m/z 400 with an ion target value of 1 × 105, a maximum and down‐regulation, and also to predict the activation injection time of 120 ms, and fixed first mass of 100 Th. state (either activated or inhibited) of a putative Repeat sequencing of peptides was minimized by regulator. In our analysis we considered genes with an excluding the selected peptide candidates for 40 seconds. overlap P‐value of >7 (log10) that had an activation Z‐ score > 2 as activated and those with an activation Z‐ Mass spectrometry raw data processing: MS raw files score < −2 as inhibited. were analyzed by the MaxQuant61 (version 1.4.3.20) and peak lists were searched against the human Uniprot
FASTA database (version Nov 2016), and a common Proteomics and multi‐aomics dat integration contaminants database (247 entries) by the Andromeda Detergent solubility profiling. For proteome analysis search engine62 as previously described28. ~100mg of fresh frozen total tissue (wet weight) of mouse
lungs was homogenized in 500 µl PBS (with protease QDSP data analysis: The quantitative detergent solubility inhibitor cocktail) using an Ultra‐turrax homogenizer. profiling (QDSP) analysis was done as previously After centrifugation the soluble proteins were collected described28. Briefly, intensities were first normalized such and proteins were extracted from the insoluble pellet in 3 that the mean log2 intensities of the young and the old steps using buffers with increasing stringency as samples are zero, respectively. Using the normalized describede in th QDSP protocol28. Peptides from LysC and intensities, a two‐way ANOVA with the two‐factor trypsin proteolysis of the four protein fractions in treatment (old/young) and solubility fraction (FR1, FR2, guadinium hydrochloride (enzyme/protein ratio 1:50), FR3, INSOL) and the corresponding interaction term was were purified as previously described on SDB‐RPS performed using the R function aov(). Proteins significant material stage‐tips28. in the interaction term correspond to proteins for which
the solubility profile changes between young and old Mass spectrometry: Data was acquired on a mice. Therefore, the corresponding P‐value was used for Quadrupole/Orbitrap type Mass Spectrometer (Q‐ filtering the significantly changed profiles after FDR Exactive, Thermo Scientific) as previously described28. correction. Approximately 2 μg of peptides were separated in a four
hour gradient on a 50‐cm long (75‐μm inner diameter) In silico bulk differential expression analysis. In silico bulk column packed in‐house with ReproSil‐Pur C18‐AQ 1.9 μm samples were generated by summing UMI counts across resin (Dr. Maisch GmbH). Reverse‐phase chromatography all cells within one sample. Differential gene expression was performed with an EASY‐nLC 1000 ultra‐high pressure system (Thermo Fisher Scientific), which was
19 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
analysis of in silico bulk samples was performed using the Immunofluorescence and histology R package DESeq2 (v1.20.0)63. For immunofluorescence microscopy, mouse lungs were perfused with PBS, fixed in 4% paraformaldehyde (pH 2D annotation enrichment analysis. Statistical and 7.0), and embedded in paraffin for FFPE sections. The bioinformatics operations, such as normalization, pattern paraffin sections (3.5 µm) were deparaffinized and recognition, cross‐omics comparisons and multiple‐ rehydrated, and the antigen retrieval was accomplished hypothesis testing corrections, were performed with the by pressure‐cooking (30 s at 125°C and 10 s at 90°C) in Perseus software package57. The 2D annotation citrate buffer (10 mM, pH 6.0). After blocking for 1 h at enrichment test used to compare proteome and room temperature with 5% BSA, the lung sections were transcriptome, is based on a two‐dimensional incubated with the primary antibodies overnight at 4°C, generalization of the nonparametric two‐sample test. The incubated with the secondary antibodies (1:250) for 2 false discovery rate is stringently controlled by correcting h, followed by DAPI (Sigma‐Aldrich, 1:2,000) for 20 min at for multiple hypothesis testing23. room temperature. Images were acquired with an LSM 710 microscope (Zeiss). The following primary (1) and Flow cytometry secondary (2) antibodies were used: (1) CC10 rabbit Isolated total lung cell suspensions were used to detect (Santa Cruz, sc‐ 25554), Foxj1 mouse (Santa Cruz, sc‐ and quantify cell populations and activation by flow 53139), Collagen IV rabbit (Abcam, ab6586), Podoplanin cytometry. We depleted red blood cells by positive goat (R&D Systems, AF3244); (2) donkey anti‐mouse selection of Ter199 cells, followed by CD45 bead Alexa Fluor (AF) 647 (Invitrogen, A21447), donkey anti‐ separation (Miltenyi Biotec; Bergish Gladbach, Germany). rabbit AF 568 (Invitrogen, A10042), donkey anti‐goat AF Next, we analyzed cells by FACS cell suspensions before 488 (Invitrogen, A21202). and after CD45 separation and stained cells suspensions The frequency of ciliated (nuclear Foxj1+) and club cells with anti‐mouse CD31, EpCAM, Podoplanin, (Biolegend; (CC10+) were quantified by counting 2647 cells, covering San Diego, CA, USA), H2‐K1 (Thermo Fisher Scientific; a total length of 22 mm airway in 28 individual airways Waltham, Massachusetts, USA). Cells were stained in the (young, n=14; old n=14) of two mice of each age group. dark at 4°C, for 20 minutes. CD45 lineage negative cells We normalized cell numbers to the total length of their were stained with Nile Red (Santacruz Biotechnology), in respective airway using the ZEN 2.3 SP1 software for a 1:1000 dilution for 10 min at 4°C, as previously image processing. reported64. Data acquisition was performed in a BD Fortessa flow cytometer (Becton Dickinson; Heidelberg, Data availability Germany). Data were analyzed using the FlowJo software (TreeStart Inc; Ashland, OR, USA). Data were reported as Proteome raw data can be downloaded from the PRIDE absolute numbers (cells/uL), normalized by beads counts repository under the accession number XXX. scRNA‐seq (BD Truecount TM Beads tubes; BD Biosciences, raw data can be downloaded from the Gene Expression Heidelberg, Germany). For H2‐K1 and Nile Red, data were Omnibus under the accession number XXX and accessed analyzed by mean fluorescence intensity (MFI). Negative via the EBIs single cell expression atlas under the link XXX. thresholds for gating were set according to isotype‐ The whole lung aging atlas can be accessed via an labeled, and unstained controls. interactive user‐friendly webtool at: https://theislab.github.io/LungAgingAtlas
‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
20 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
Acknowledgments Author contributions
We thank Silvia Weidner and Daniela Dietel for excellent H.B.S. conceptualized and supervised the entire project technical assistance. We also thank Gabi Sowa, Igor Paron and wrote the paper. F.J.T. supervised single cell analysis and Korbinian Mayr for expert support of the proteomics and multi‐omics data integration. I.A. and M.S. performed pipeline. We thank Sandy Lösecke for technical assistance single cell transcriptomics experiments. L.M.S., I.A. and in next generation sequencing and Thomas Schwarzmayr H.B.S. analyzed single cell transcriptomics data. H.B.S. for support with High‐seq 4000 sequencing raw data. performed proteomics experiments. H.B.S. and L.M.S. L.M.S. acknowledges funding from the European Union’s analyzed the proteomics data and performed Horizon 2020 research and innovation programme under transcriptomics and proteomics data integration. I.A. and the Marie Sklodowska‐Curie grant agreement No 753039. C.H.M. performed histology and immunofluorescence This work was supported by the German Center for Lung microscopy. I.E.F. and F.R.G. performed flow cytometry Research (DZL), the Helmholtz Association, and the experiments. E.G. and T.M.S. performed next generation German Federal Ministry of Education and Research sequencing of single cell libraries. L.M.S. and G.T. set up (BMBF), project Single Cell Genomics Network Germany. the interactive webtool. O.E. assisted in data interpretation and M.M. provided important support with
mass spectrometry equipment. All authors read, edited
and approved the manuscript.
Conflict of Interest
The authors have no conflict of interest.
21 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
References
1. Hogan, B.L. et al. Repair and Regeneration of Using Nanoliter Droplets. Cell 161, 1202‐1214 the Respiratory System: Complexity, Plasticity, (2015). and Mechanisms of Lung Stem Cell Function. 16. Ziegenhain, C. et al. Comparative Analysis of Cell stem cell 15, 123‐138 (2014). Single‐Cell RNA Sequencing Methods. Mol Cell 2. Meiners, S., Eickelberg, O. & Konigshoff, M. 65, 631‐643 e634 (2017). Hallmarks of the ageing lung. Eur Respir J 45, 17. Lefrancais, E. et al. The lung is a site of platelet 807‐827 (2015). biogenesis and a reservoir for haematopoietic 3. Lopez‐ Otin, C., Blasco, M.A., Partridge, L., progenitors. Nature (2017). Serrano, M. & Kroemer, G. The hallmarks of 18. Heng, T.S., Painter, M.W. & Immunological aging. Cell 153, 1194‐1217 (2013). Genome Project, C. The Immunological 4. Lowery, E.M., Brubaker, A.L., Kuhlmann, E. & Genome Project: networks of gene expression Kovacs, E.J. The aging lung. Clinical in immune cells. Nature immunology 9, 1091‐ interventions in aging 8, 1489‐1496 (2013). 1094 (2008). 5. Bailey, K.L. et al. Aging causes a slowing in 19. Aran, D., Hu, Z. & Butte, A.J. xCell: digitally ciliary beat frequency, mediated by portraying the tissue cellular heterogeneity PKCepsilon. American journal of physiology. landscape. Genome biology 18, 220 (2017). Lung cellular and molecular physiology 306, 20. Martinez‐Jimenez, C.P. et. al Aging increases L584‐589 (2014). cell‐to‐cell transcriptional variability upon 6. Panda, A. et al. Human innate immune stimulation. Science 355, 1433‐1436 immunosenescence: causes and (2017). consequences for immunity in old age. Trends 21. Enge, M. et al. Single‐Cell Analysis of Human in immunology 30, 325‐333 (2009). Pancreas Reveals Transcriptional Signatures of 7. Jane‐Wit, D. & Chun, H.J. Mechanisms of Aging and Somatic Mutation Patterns. Cell dysfunction in senescent pulmonary 171, 321‐330 e314 (2017). endothelium. The journals of gerontology. 22. Kramer, A., Green, J., Pollard, J., Jr. & Series A, Biological sciences and medical Tugendreich, S. Causal analysis approaches in sciences 67, 236‐241 (2012). Ingenuity Pathway Analysis. Bioinformatics 30, 8. Brandenberger, C. & Muhlfeld, C. Mechanisms 523‐530 (2014). of lung aging. Cell and tissue research 367, 23. Cox, J. & Mann, M. 1D and 2D annotation 469‐480 (2017). enrichment: a statistical method integrating 9. Sicard, D. et al. Aging and anatomical quantitative proteomics with complementary variations in lung tissue stiffness. American high‐throughput data. BMC Bioinformatics 13 journal of physiology. Lung cellular and Suppl 16, S12 (2012). molecular physiology 314, L946‐L955 (2018). 24. Burgstaller, G. et al. eTh instructive 10. Franks, T.J. et al. Resident cellular extracellular matrix of the lung: basic components of the human lung: current composition and alterations in chronic lung knowledge and goals for research on cell disease. Eur Respir J 50 (2017). phenotyping and function. Proc Am Thorac 25. Budinger, G.R.S. et al. The Intersection of Soc 5, 763‐766 (2008). Aging Biology and the Pathobiology of Lung 11. Svensson, V., Vento‐Tormo, R. & Teichmann, Diseases: A Joint NHLBI/NIA Workshop. The S.A. Exponential scaling of single‐cell RNA‐seq journals of gerontology. Series A, Biological in the past decade. Nature protocols 13, 599‐ sciences and medical sciences 72, 1492‐1500 604 (2018). (2017). 12. Tanay, A. & Regev, A. Scaling single‐cell 26. Wierer, M. et al. Compartment‐resolved genomics from phenomenology to Proteomic Analysis of Mouse Aorta during mechanism. Nature 541, 331‐338 (2017). Atherosclerotic Plaque Formation Reveals 13. Han, X. et al. Mapping the Mouse Cell Atlas by Osteoclast‐specific Protein Expression. Microwell‐Seq. Cell 172, 1091‐1107 e1017 Molecular & cellular proteomics : MCP 17, (2018). 321‐334 (2018). 14. Aebersold, R. & Mann, M. Mass‐spectrometric 27. Schiller, H.B. et al. Deep Proteome Profiling exploration of proteome structure and Reveals Common Prevalence of MZB1‐Positive function. Nature 537, 347‐355 (2016). Plasma B Cells in Human Lung and Skin 15. Macosko, E.Z. et al. Highly Parallel Genome‐ Fibrosis. American journal of respiratory and wide Expression Profiling of Individual Cells critical care medicine 196, 1298‐1310 (2017).
22 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
28. Schiller, H.B. et al. Time‐ and compartment‐ and p35 mRNAs in macrophages. Journal of resolved proteome profiling of the immunology 168, 4055‐4062 (2002). extracellular niche in lung injury and repair. 39. Zhou, F. Molecular mechanisms of IFN‐gamma Mol Syst Biol 11, 819 (2015). to up‐regulate MHC class I antigen processing 29. Kiyozumi, D., Sugimoto, N. & Sekiguchi, K. and presentation. International reviews of Breakdown of the reciprocal stabilization of immunology 28, 239‐260 (2009). QBRICK/Frem1, Fras1, and Frem2 at the 40. Koeberle, A., Loser, K. & Thurmer, M. basement membrane provokes Fraser Stearoyl‐CoA desaturase‐1 and adaptive stress syndrome‐like defects. Proceedings of the signaling. Biochimica et biophysica acta 1861, National Academy of Sciences of the United 1719‐1726 (2016). States of America 103, 11981‐11986 (2006). 41. Whitsett, J.A., Wert, S.E. & Weaver, T.E. 30. Petrou, P., Pavlakis, E., Dalezios, Y., Diseases of pulmonary surfactant Galanopoulos, V.K. & Chalepakis, G. Basement homeostasis. Annual review of pathology 10, membrane distortions impair lung lobation 371‐393 (2015). and capillary organization in the mouse model 42. Plantier, L. et al. Activation of sterol‐response for fraser syndrome. The Journal of biological element‐binding proteins (SREBP) in alveolar chemistry 280, 10350‐10356 (2005). type II cells enhances lipogenesis causing 31. Ricard‐Blum, S. The collagen family. Cold pulmonary lipotoxicity. The Journal of Spring Harbor perspectives in biology 3, biological chemistry 287, 10099‐10114 (2012). a004978 (2011). 43. Stoeckius, M. et al. Simultaneous epitope and 32. Ehnis, T., Dieterich, W., Bauer, M., Kresse, H. transcriptome measurement in single cells. & Schuppan, D. Localization of a binding site Nature methods 14, 865‐868 (2017). for the proteoglycan decorin on collagen XIV 44. Lein, E., Borm, L.E. & Linnarsson, S. The (undulin). The Journal of biological chemistry promise of spatial transcriptomics for 272, 20414‐20419 (1997). neuroscience in the era of molecular cell 33. Kolb, M., Margetts, P.J., Sime, P.J. & Gauldie, J. typing. Science 358, 64‐69 (2017). Proteoglycans decorin and biglycan 45. Moor, A.E. & Itzkovitz, S. Spatial differentially modulate TGF‐beta‐mediated transcriptomics: paving the way for tissue‐ fibrotic responses in the lung. American level systems biology. Current opinion in journal of physiology. Lung cellular and biotechnology 46, 126‐133 (2017). molecular physiology 280, L1327‐1334 (2001). 46. Schulz, D. et al. Simultaneous Multiplexed 34. Yamaguchi, Y., Mann, D.M. & Ruoslahti, E. Imaging of mRNA and Proteins with Negative regulation of transforming growth Subcellular Resolution in Breast Cancer Tissue factor‐beta by the proteoglycan decorin. Samples by Mass Cytometry. Cell systems 6, Nature 346, 281‐284 (1990). 531 (2018). 35. Cormier, S.A. et al. T(H)2‐mediated pulmonary 47. Mondal, M., Liao, R., Xiao, L., Eno, T. & Guo, J. inflammation leads to the differential Highly Multiplexed Single‐Cell In Situ Protein expression of ribonuclease genes by alveolar Analysis with Cleavable Fluorescent macrophages. American journal of respiratory Antibodies. Angewandte Chemie 56, 2636‐ cell and molecular biology 27, 678‐687 (2002). 2639 (2017). 36. Palecanda, A. et al. Role of the scavenger 48. Hynes, R.O. Stretching the boundaries of receptor MARCO in alveolar macrophage extracellular matrix research. Nat Rev Mol Cell binding of unopsonized environmental Biol 15, 761‐763 (2014). particles. The Journal of experimental 49. Decaris, M.L. et al. Proteomic Analysis of medicine 189, 1497‐1506 (1999). Altered Extracellular Matrix Turnover in 37. Ruffell, D. et al. A CREB‐C/EBPbeta cascade Bleomycin‐induced Pulmonary Fibrosis. induces M2 macrophage‐specific gene Molecular & cellular proteomics : MCP 13, expression and promotes muscle injury repair. 1741‐1752 (2014). Proceedings of the National Academy of 50. Moliva, J.I. et al. Molecular composition of the Sciences of the United States of America 106, alveolar lining fluide in th aging lung. Age 36, 17475‐17480 (2009). 9633 (2014). 38. Gorgoni, B., Maritano, D., Marthyn, P., Righi, 51. Walski, M. et al. Pulmonary surfactant: M. & Poli, V. C/EBP beta gene inactivation ultrastructural features and putative causes both impaired and enhanced gene mechanisms of aging. Journal of physiology expression and inverse regulation of IL‐12 p40 and pharmacology : an official journal of the
23 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
Angelidis et al 2018, Multi‐omics systems biology of lung aging
Polish Physiological Society 60 Suppl 5, 121‐ maximize uptime using off the shelf 125 (2009). components. J Proteome Res 11, 3458‐ 3466 52. Alder, J.K. et al. Telomere dysfunction causes (2012). alveolar stem cell failure. Proceedings of the 60. Michalski, A. et al. Mass spectrometry‐based National Academy of Sciences of the United proteomics using Q Exactive, a high‐ States of America (2015). performance benchtop quadrupole Orbitrap 53. Shekhar, K. et al. Comprehensive Classification mass spectrometer. Molecular & cellular of Retinal Bipolar Neurons by Single‐Cell proteomics : MCP 10, M111 011015 (2011). Transcriptomics. 6Cell, 16 1308‐1323 e1330 61. Cox, J. & Mann, M. MaxQuant enables high (2016). peptide identification rates, individualized 54. Dobin, A. et al. STAR: ultrafast universal RNA‐ p.p.b.‐range mass accuracies and proteome‐ seq aligner. Bioinformatics 29, 15‐21 (2013). wide protein quantification. Nature 55. Satija, R., Farrell, J.A., Gennert, D., Schier, A.F. biotechnology 26, 1367‐1372 (2008). & Regev, A. Spatial reconstruction of single‐ 62. Cox, J. et al. Andromeda: a peptide search cell gene expression data. Nature engine integrated into the MaxQuant biotechnology 33, 495‐502 (2015). environment. Journal of proteome research 56. Kowalczyk, M.S. et al. Single‐cell RNA‐seq 10, 1794‐1805 (2011). reveals changes in cell cycle and 63. Love, M.I., Huber, W. & Anders, S. Moderated differentiation programs upon aging of estimation of fold change and dispersion for hematopoietic stem cells. Genome research RNA‐seq data with DESeq2. Genome biology 25, 1860‐1872 (2015). 15, 550 (2014). 57. Tyanova, S. et al. The Perseus computational 64. Perico, M.E., Crivellente, F., Faustinelli, I., platform for comprehensive analysis of Suozzi, A. & Cristofori, P. Flow cytometry, with (prote)omics data. Nature methods (2016). double staining with Nile red and anti‐CD3 58. Thakur, S.S. et al. Deep and highly sensitive antibody, to detect phospholipidosis in proteome coverage by LC‐MS/MS without peripheral blood lymphocytes of rats treated prefractionation. Molecular & cellular with amiodarone. Cell biology and toxicology proteomics : MCP 10, M110 003699 (2011). 725, 58 ‐598 (2009). 59. Scheltema, R.A. & Mann, M. SprayQc: a real‐ time LC‐MS/MS quality monitoring system to
24 bioRxiv preprint doi: https://doi.org/10.1101/351353; this version posted June 20, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.