Systems level identification of a matrisome-associated macrophage polarisation state in multi-organ fibrosis
Curation statements for this article:-
Curated by eLife
eLife assessment
This important study deepens our understanding of macrophage phenotypes in pathological contexts and identifies a new macrophage state associated with tissue fibrosis, as well as putative drivers of this cellular state. The authors provide convincing evidence and performed a well-thought-out and thoroughly described computational analysis of single-cell RNA-sequencing data. This work will be of broad interest to the fields of tissue inflammation, fibrosis, macrophage biology, and immunology.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Computational and Systems Biology (eLife)
Abstract
Tissue fibrosis affects multiple organs and involves a master-regulatory role of macrophages which respond to an initial inflammatory insult common in all forms of fibrosis. The recently unravelled multi-organ heterogeneity of macrophages in healthy and fibrotic human disease suggests that macrophages expressing osteopontin (SPP1) associate with lung and liver fibrosis. However, the conservation of this SPP1 + macrophage population across different tissues and its specificity to fibrotic diseases with different etiologies remain unclear. Integrating 15 single-cell RNA-sequencing datasets to profile 235,930 tissue macrophages from healthy and fibrotic heart, lung, liver, kidney, skin, and endometrium, we extended the association of SPP1 + macrophages with fibrosis to all these tissues. We also identified a subpopulation expressing matrisome-associated genes (e.g., matrix metalloproteinases and their tissue inhibitors), functionally enriched for ECM remodelling and cell metabolism, representative of a matrisome-associated macrophage (MAM) polarisation state within SPP1 + macrophages. Importantly, the MAM polarisation state follows a differentiation trajectory from SPP1 + macrophages and is associated with a core set of regulon activity. SPP1 + macrophages without the MAM polarisation state (SPP1 + MAM - ) show a positive association with ageing lung in mice and humans. These results suggest an advanced and conserved polarisation state of SPP1 + macrophages in fibrotic tissues resulting from prolonged inflammatory cues within each tissue microenvironment.
Article activity feed
-
-
Author Response
eLife assessment
This important study deepens our understanding of macrophage phenotypes in pathological contexts and identifies a new macrophage state associated with tissue fibrosis, as well as putative drivers of this cellular state. The authors provide convincing evidence and performed a well-thought-out and thoroughly described computational analysis of single-cell RNA-sequencing data. This work will be of broad interest to the fields of tissue inflammation, fibrosis, macrophage biology, and immunology.
We thank eLife reviewing editors as well as the two Reviewers for their supportive, constructive and insightful assessment of the manuscript. We apologize for the time that has taken us to submit the revisions. The main reason for this delay was the integration of newly published scRNA-seq datasets that were …
Author Response
eLife assessment
This important study deepens our understanding of macrophage phenotypes in pathological contexts and identifies a new macrophage state associated with tissue fibrosis, as well as putative drivers of this cellular state. The authors provide convincing evidence and performed a well-thought-out and thoroughly described computational analysis of single-cell RNA-sequencing data. This work will be of broad interest to the fields of tissue inflammation, fibrosis, macrophage biology, and immunology.
We thank eLife reviewing editors as well as the two Reviewers for their supportive, constructive and insightful assessment of the manuscript. We apologize for the time that has taken us to submit the revisions. The main reason for this delay was the integration of newly published scRNA-seq datasets that were relevant for gaining further power and reproducibility for our analyses, especially for refining the transcriptomics resolution of SPP1+MAM- and SPP1+MAM+ cells and their respective correlation with ageing. Specifically, we have added new datasets from NASH [1] and endometrium [2] patients so that each human tissue comprises scRNA-seq data derived from at least 2 independent studies (revised Table 1). Crucially, as the human lung cell atlas got published recently (after receipt of our decision letter) [3], we investigated in greater detail (increased N numbers and co-variates), the association of SPP1+ macrophages and homeostatic ones with lung ageing.
This new undertaking was not directly asked by reviewers/editors, but instead, was suggested as informal feedback received after posting our manuscript into biorxiv repository. Importantly, these revisions together with the corrections asked by the two reviewers made the conclusions of the manuscript stronger (and more robust as we increased the number of samples) by refining (i) the regulons that associate with SPP1+MAM+ differentiation and (ii) subset-specific association with human and mice lung ageing, a finding that suggests MAM polarization state is acquired when there is prominent tissue fibrosis. Lung aging is significantly associated with SPP1+MAM- state, which represents the inflammatory/secretory phenotype that yet to be polarized to the fibrotic one seen in the disease state.
Reviewer #1 (Public Review):
Huang, Kevin Y. et al. perform a meta-analysis of single-cell RNA-seq (scRNA-seq) data derived from 11 studies and across six tissues (liver, lung, heart, skin, kidney, endometrium) to address a focused hypothesis: pro-fibrotic SPP1+ macrophages that have been found in liver and lung tissue of idiopathic pulmonary fibrosis patients exist in other human tissues which can result in broader fibrotic disease states. The authors use existing, state-of-the-art single-cell analysis tools to perform the meta-analysis. They convincingly show that the SPP1+ macrophage population can be identified in lung, liver, heart, skin, uterus (endometrium), and kidney clusters derived from each tissues' scRNA-seq data. They further identify three subpopulations of the SPP1+ macrophages: a matrisome-associated macrophages (MAMs) defined as SPP1+MAM+ and two others enriched for inflammatory and ribosomal processes which they group together and define as SPP1+MAM-. Pathway analysis of genes unregulated in SPP1+MAM+ vs SPP1+MAM- cells yields significant enrichment of extracellular matrix remodeling and metabolism-related pathways and genes. This allows them to arrive at SPP1+MAM+ and SPP1+MAM- gene expression signature scores to further highlight the upregulation of these pathways in SPP1+MAM+ macrophages and their role in fibrosis. They explicitly show enrichment for SPP1+MAM+ macrophages in disease compared to healthy control subjects in a variety of tissues and their associated fibrosis-related diseases. Cell differentiation trajectory analysis identified 2 main trajectories: both starting from FCN1+ infiltrating monocytes/macrophages with one moving toward a homeostatic state and another toward SPP1+MAM+. They verified this using an alternative trajectory analysis approach. Importantly, for all tissues and fibrotic diseases, they found SPP1+MAM+ were at the end of the trajectory preceded by the SPP1+MAM- state, suggesting SPP1+MAM+ represents a common polarization state of SPP1+ macrophages. They develop a probability-based score that estimates the propensity of SPP1+MAM- macrophages to differentiate into SPP1+MAM+ and show that this was significantly higher in fibrotic disease subjects compared to healthy controls. They go on to identify the transcription factor networks (regulons) associated with SPP1+MAM+ differentiation and activation. They find a number of enriched regulons/transcription factors and through a linear-modeling trajectory analysis highlight the regulons that are associated specifically with the SPP1+MAM- to SPP1+MAM+ transition. In this way, they prioritize the NFATC1 and HIVEP3 regulations as driving the differentiation of SPP1+MAM- macrophages toward the SPP1+MAM+ polarization state. Finally, given that age is a risk factor for fibrotic disease, they assessed the association of SPP1+MAM+ and SPP1+MAM- gene signatures in healthy control old and young human subjects as well as old and young mice and found SPP1+MAM+ was either exclusively (human) or more significantly (mice) elevated in old versus young compared to SPP1+MAM-.
The strengths of this paper are the authors gathered a number of relevant single-cell RNA-seq data sets from fibrosis-focused studies to address a highly focused hypothesis (stated above). They gained the power to detect the population of SPP1+MAM+ cells by integrating these datasets. The analysis is carried out well using existing state-of-the-art tools. With whatever metric or single cell analysis-based discovery they make about the SPP1+MAM+ subpopulations (e.g., gene signatures, endpoint of trajectory analysis, associated regulons, etc), they compare the relevant scoring metrics in fibrosis and control subjects at every stage of the meta-analysis and find the SPP11+MAM+ is consistently higher across tissues and fibrosis-related diseases.
There are only minor weaknesses in this paper. One is that some of the most highly significant or simply significant results are not shown in main figures but are summarized in supplementary tables (e.g., MYC TARGETS V1 would have appeared as the most significant, highest enriched, and among the largest in terms of set size). Another is analysis criteria that may not yield the most biologically relevant or impactful conclusion (e.g., while the regulon THRA does not display a shift in slopes it shows the strongest, progressive increase going toward the SPP1+MAM+ state).
We thank the Reviewer for his very accurate summary of our findings. We agree with the Reviewer regarding all points and provide the answers to the suggested minor points as per below.
Reviewer #2 (Public Review):
In the past few years, single-cell transcriptomics analysis has uncovered cellular states associated with disease in experimental models and humans, revealing previously unrecognized disease-associated macrophage states. In particular, a macrophage state characterized by high expression of SPP1 (encoding osteopontin), and by a specific gene expression signature including the expression of TREM2, has been observed in various pathologies and given various names depending on the context e.g. TREM2hi macrophages, lipid-associated macrophages (LAM), disease-associated microglia (DAM), Scar-associated macrophages (SAM), etc... However, a focused investigation and comparison of SPP1+ macrophages across disease contexts were lacking. Here, the authors aimed to systematically analyze SPP1+ macrophages in the context of tissue fibrosis, and integrated single-cell RNA-seq data of >200,000 human macrophages in 6 organs in health and tissue fibrosis.
Beyond confirming the presence of SPP1+ macrophages with a conserved gene expression module (TREM2, CD9, GPNMB, etc...) across tissues and their association with fibrosis, the authors identified a previously unknown cell subset within SPP1+ macrophages, that was enriched for the expression of genes involved in remodeling of the extracellular matrix, which they termed SPP1+ matrisome-associated macrophages (SPP1+MAM+). The authors further used computational tools to compare these SPP1+MAM+ macrophages to previously described SPP1+ macrophage states (LAM, DAM, SAM), investigate the differentiation and activation trajectory of SPP1+MAM+ macrophages, and identify potential transcriptional regulators involved in their differentiation. Finally, the authors show that SPP1+MAM+ macrophages are associated with ageing in both humans and mice.
Overall, the conclusions of the authors are well supported by the data. The authors made excellent use of available computational tools, and the figures are clear and informative. The methods are well-described and appropriately used. In particular, the authors made a nice effort in explaining and justifying some key decisions in their scRNA-seq data analysis workflow, including a data-driven approach to decisions in the clustering analysis.
The author's findings are of broad interest to the fields of tissue inflammation, fibrosis, macrophage biology, and immunology, and their report constitutes a valuable resource, and a basis for further investigations of macrophage differentiation mechanisms in tissue fibrosis, and how macrophages could be targeted to alleviate pathological tissue fibrosis.
We thank the reviewer for finding our work valuable and for carefully assessing the manuscript. We agree with the Reviewer regarding all points.
-
eLife assessment
This important study deepens our understanding of macrophage phenotypes in pathological contexts and identifies a new macrophage state associated with tissue fibrosis, as well as putative drivers of this cellular state. The authors provide convincing evidence and performed a well-thought-out and thoroughly described computational analysis of single-cell RNA-sequencing data. This work will be of broad interest to the fields of tissue inflammation, fibrosis, macrophage biology, and immunology.
-
Reviewer #1 (Public Review):
Huang, Kevin Y. et al. perform a meta-analysis of single-cell RNA-seq (scRNA-seq) data derived from 11 studies and across six tissues (liver, lung, heart, skin, kidney, endometrium) to address a focused hypothesis: pro-fibrotic SPP1+ macrophages that have been found in liver and lung tissue of idiopathic pulmonary fibrosis patients exist in other human tissues which can result in broader fibrotic disease states. The authors use existing, state-of-the-art single-cell analysis tools to perform the meta-analysis. They convincingly show that the SPP1+ macrophage population can be identified in lung, liver, heart, skin, uterus (endometrium), and kidney clusters derived from each tissues' scRNA-seq data. They further identify three subpopulations of the SPP1+ macrophages: a matrisome-associated macrophages (MAMs) …
Reviewer #1 (Public Review):
Huang, Kevin Y. et al. perform a meta-analysis of single-cell RNA-seq (scRNA-seq) data derived from 11 studies and across six tissues (liver, lung, heart, skin, kidney, endometrium) to address a focused hypothesis: pro-fibrotic SPP1+ macrophages that have been found in liver and lung tissue of idiopathic pulmonary fibrosis patients exist in other human tissues which can result in broader fibrotic disease states. The authors use existing, state-of-the-art single-cell analysis tools to perform the meta-analysis. They convincingly show that the SPP1+ macrophage population can be identified in lung, liver, heart, skin, uterus (endometrium), and kidney clusters derived from each tissues' scRNA-seq data. They further identify three subpopulations of the SPP1+ macrophages: a matrisome-associated macrophages (MAMs) defined as SPP1+MAM+ and two others enriched for inflammatory and ribosomal processes which they group together and define as SPP1+MAM-. Pathway analysis of genes unregulated in SPP1+MAM+ vs SPP1+MAM- cells yields significant enrichment of extracellular matrix remodeling and metabolism-related pathways and genes. This allows them to arrive at SPP1+MAM+ and SPP1+MAM- gene expression signature scores to further highlight the upregulation of these pathways in SPP1+MAM+ macrophages and their role in fibrosis. They explicitly show enrichment for SPP1+MAM+ macrophages in disease compared to healthy control subjects in a variety of tissues and their associated fibrosis-related diseases. Cell differentiation trajectory analysis identified 2 main trajectories: both starting from FCN1+ infiltrating monocytes/macrophages with one moving toward a homeostatic state and another toward SPP1+MAM+. They verified this using an alternative trajectory analysis approach. Importantly, for all tissues and fibrotic diseases, they found SPP1+MAM+ were at the end of the trajectory preceded by the SPP1+MAM- state, suggesting SPP1+MAM+ represents a common polarization state of SPP1+ macrophages. They develop a probability-based score that estimates the propensity of SPP1+MAM- macrophages to differentiate into SPP1+MAM+ and show that this was significantly higher in fibrotic disease subjects compared to healthy controls. They go on to identify the transcription factor networks (regulons) associated with SPP1+MAM+ differentiation and activation. They find a number of enriched regulons/transcription factors and through a linear-modeling trajectory analysis highlight the regulons that are associated specifically with the SPP1+MAM- to SPP1+MAM+ transition. In this way, they prioritize the NFATC1 and HIVEP3 regulations as driving the differentiation of SPP1+MAM- macrophages toward the SPP1+MAM+ polarization state. Finally, given that age is a risk factor for fibrotic disease, they assessed the association of SPP1+MAM+ and SPP1+MAM- gene signatures in healthy control old and young human subjects as well as old and young mice and found SPP1+MAM+ was either exclusively (human) or more significantly (mice) elevated in old versus young compared to SPP1+MAM-.
The strengths of this paper are the authors gathered a number of relevant single-cell RNA-seq data sets from fibrosis-focused studies to address a highly focused hypothesis (stated above). They gained the power to detect the population of SPP1+MAM+ cells by integrating these datasets. The analysis is carried out well using existing state-of-the-art tools. With whatever metric or single cell analysis-based discovery they make about the SPP1+MAM+ subpopulations (e.g., gene signatures, endpoint of trajectory analysis, associated regulons, etc), they compare the relevant scoring metrics in fibrosis and control subjects at every stage of the meta-analysis and find the SPP11+MAM+ is consistently higher across tissues and fibrosis-related diseases.
There are only minor weaknesses in this paper. One is that some of the most highly significant or simply significant results are not shown in main figures but are summarized in supplementary tables (e.g., MYC TARGETS V1 would have appeared as the most significant, highest enriched, and among the largest in terms of set size). Another is analysis criteria that may not yield the most biologically relevant or impactful conclusion (e.g., while the regulon THRA does not display a shift in slopes it shows the strongest, progressive increase going toward the SPP1+MAM+ state).
-
Reviewer #2 (Public Review):
In the past few years, single-cell transcriptomics analysis has uncovered cellular states associated with disease in experimental models and humans, revealing previously unrecognized disease-associated macrophage states. In particular, a macrophage state characterized by high expression of SPP1 (encoding osteopontin), and by a specific gene expression signature including the expression of TREM2, has been observed in various pathologies and given various names depending on the context e.g. TREM2hi macrophages, lipid-associated macrophages (LAM), disease-associated microglia (DAM), Scar-associated macrophages (SAM), etc... However, a focused investigation and comparison of SPP1+ macrophages across disease contexts were lacking. Here, the authors aimed to systematically analyze SPP1+ macrophages in the context …
Reviewer #2 (Public Review):
In the past few years, single-cell transcriptomics analysis has uncovered cellular states associated with disease in experimental models and humans, revealing previously unrecognized disease-associated macrophage states. In particular, a macrophage state characterized by high expression of SPP1 (encoding osteopontin), and by a specific gene expression signature including the expression of TREM2, has been observed in various pathologies and given various names depending on the context e.g. TREM2hi macrophages, lipid-associated macrophages (LAM), disease-associated microglia (DAM), Scar-associated macrophages (SAM), etc... However, a focused investigation and comparison of SPP1+ macrophages across disease contexts were lacking. Here, the authors aimed to systematically analyze SPP1+ macrophages in the context of tissue fibrosis, and integrated single-cell RNA-seq data of >200,000 human macrophages in 6 organs in health and tissue fibrosis.
Beyond confirming the presence of SPP1+ macrophages with a conserved gene expression module (TREM2, CD9, GPNMB, etc...) across tissues and their association with fibrosis, the authors identified a previously unknown cell subset within SPP1+ macrophages, that was enriched for the expression of genes involved in remodeling of the extracellular matrix, which they termed SPP1+ matrisome-associated macrophages (SPP1+MAM+). The authors further used computational tools to compare these SPP1+MAM+ macrophages to previously described SPP1+ macrophage states (LAM, DAM, SAM), investigate the differentiation and activation trajectory of SPP1+MAM+ macrophages, and identify potential transcriptional regulators involved in their differentiation. Finally, the authors show that SPP1+MAM+ macrophages are associated with ageing in both humans and mice.
Overall, the conclusions of the authors are well supported by the data. The authors made excellent use of available computational tools, and the figures are clear and informative. The methods are well-described and appropriately used. In particular, the authors made a nice effort in explaining and justifying some key decisions in their scRNA-seq data analysis workflow, including a data-driven approach to decisions in the clustering analysis.
The author's findings are of broad interest to the fields of tissue inflammation, fibrosis, macrophage biology, and immunology, and their report constitutes a valuable resource, and a basis for further investigations of macrophage differentiation mechanisms in tissue fibrosis, and how macrophages could be targeted to alleviate pathological tissue fibrosis.
-
