Targeting immune–fibroblast cell communication in heart failure

Existing snRNA-seq studies of the human heart rely on nuclear RNA and lack cellular transcriptomic, epigenetic and proteomic profiling needed for a granular picture of immune and fibroblast populations in the failing heart.

“However, these studies rely on nuclear RNA information and lack cellular transcriptomic, epigenetic and proteomic profiling information necessary to obtain a granular picture of the precise immune and fibroblast populations that emerge in the failing human heart and collaborate to drive cardiac fibrosis.”

• cites [18] Tucker, N. R. et al. Transcriptional and cellular diversity of the human heart. Circulation 142, 466–482 (2020). sourceNot checked — no full text of the source
• cites [19] Chaffin, M. et al. Single-nucleus profiling of human dilated and hypertrophic cardiomyopathy. Nature https://doi.org/10.1038/S41586-022-04817-8 (2022). sourcePartly supported

  The Methods describe single-nucleus RNA sequencing, supporting the claim that this study relies on nuclear RNA rather than integrated cellular transcriptomic, epigenetic and proteomic profiling. But the ‘Changes in cell sub-populations’ and ‘Activated fibroblasts in DCM and HCM’ sections identify macrophage sub-populations and a disease-associated activated fibroblast population. The paper therefore does not support the implication that its approach cannot provide a granular picture of those populations, or establish that the additional profiling is necessary to do so. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [20] Koenig, A. L. et al. Single-cell transcriptomics reveals cell-type-specific diversification in human heart failure. Nat. Cardiovasc. Res. 1, 263–280 (2022). sourceDisputed

  The Results section (‘snRNA-seq and scRNA-seq reveal the cellular landscape of the human heart’; page unavailable) says this study sequenced both nuclei and whole cells, contrary to the claim that it relied only on nuclear RNA and lacked cellular transcriptomic profiling. The Discussion (page unavailable) also describes distinct fibroblast populations in healthy and failing hearts. The paper does not establish the claim about missing epigenetic or proteomic profiling. (LLM, conf=0.89)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured
• cites [21] Litviňuková, M. et al. Cells of the adult human heart. Nature 588, 466–472 (2020). sourceDisputed

  The Main section (page unavailable) says this study combines snRNA-seq with scRNA-seq to enrich immune-cell populations; the Discussion likewise describes combined single-cell and single-nucleus data. Thus, contrary to the claim, this source does not rely only on nuclear RNA or lack cellular transcriptomic profiling. It studies the healthy adult heart rather than profiling immune and fibroblast populations emerging in heart failure. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured
• cites [22] Reichart, D. et al. Pathogenic variants damage cell composition and single cell transcription in cardiomyopathies. Science 377, eabo1984 (2022). sourcePartly supported

  The Abstract and Methods describe a study based on single-nucleus RNA sequencing, and the Discussion calls for future epigenomic and proteomic work. But the Results already distinguish fibroblast states and 14 myeloid subclusters and discuss fibroblast–macrophage interactions in fibrosis. The source therefore supports the missing-modality point, not the implication that its nuclear-RNA data cannot provide a granular picture of those populations. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

No approved treatments directly target cardiac fibrosis.

“there are at present no approved treatments that directly target cardiac fibrosis3,4.”

• cites [3] Rurik, J. G. et al. CAR T cells produced in vivo to treat cardiac injury. Science 375, 91–96 (2022). sourceCould not verify

  The Abstract says trials of anti-fibrotic therapeutics have shown at most modest effects and that these interventions aim to limit progression rather than remodel established fibrosis. It does not state whether any treatment that directly targets cardiac fibrosis has been approved; searches for approved anti-fibrotic treatments found no passage resolving that point. (LLM, conf=0.72)

  Experimental — accuracy not yet measured
• cites [4] Aghajanian, H. et al. Targeting cardiac fibrosis with engineered T cells. Nature 573, 430–433 (2019). sourceSupported

Spatial transcriptomics from human MI at days 2, 11, 62, and 153 post-MI found FAP expression was most prominent acutely after MI and diminished at later time points.

“To validate our findings, we used spatial transcriptomics data from human MI25 at days 2, 11, 62 and 153 post-MI and found FAP expression was most prominent acutely after MI and diminished at later time points (Fig. 2e, bottom).”

• cites [25] Kuppe, C. et al. Spatial multi-omic map of human myocardial infarction. Nature https://doi.org/10.1038/s41586-022-05060-x (2022). sourceSupported

A large clinical trial targeting the IL-1β pathway with the neutralizing antibody canakinumab led to a reduction in cardiac events.

“In support of this concept, a large clinical trial targeting the IL-1β pathway with the neutralizing antibody canakinumab led to a reduction in cardiac events31.”

• cites [31] Ridker, P. M. et al. Antiinflammatory therapy with canakinumab for atherosclerotic disease. N. Engl. J. Med. 377, 1119–1131 (2017). sourceSupported

Canakinumab, an anti-IL-1β monoclonal antibody, has clinical efficacy.

“We chose to focus on IL-1β signalling given the availability and clinical efficacy of the anti-IL-1β monoclonal antibody (mAb) canakinumab31.”

• cites [31] Ridker, P. M. et al. Antiinflammatory therapy with canakinumab for atherosclerotic disease. N. Engl. J. Med. 377, 1119–1131 (2017). sourceSupported

Cardiac FAP-positive fibroblasts resemble POSTN-expressing matrifibrocytes.

“and resemble POSTN-expressing matrifibrocytes7.”

• cites [7] Fu, X. et al. Specialized fibroblast differentiated states underlie scar formation in the infarcted mouse heart. J. Clin. Invest. 128, 2127–2143 (2018). sourcePartly supported

  The Results section on cardiac fibroblast proliferation following MI reports Postn expression in activated cardiac fibroblasts, and the Discussion describes matrifibrocytes as a later fibroblast state in the scar (pages not provided). Our copy does not identify FAP-positive fibroblasts or compare them with matrifibrocytes, so the claimed resemblance is not established. (LLM, conf=0.79)

  Experimental — accuracy not yet measured

Cultured human primary cardiac and dermal fibroblasts and immortalized cardiac fibroblasts were relatively poor models of human cardiac fibroblast diversity under resting or stimulated conditions.

“By contrast, cultured human primary cardiac and dermal fibroblasts and immortalized cardiac fibroblasts were relatively poor models of human cardiac fibroblast diversity under resting or stimulated (TGFβ (refs. 19,39,40), IL-1β (ref. 41)) conditions.”

• cites [19] Chaffin, M. et al. Single-nucleus profiling of human dilated and hypertrophic cardiomyopathy. Nature https://doi.org/10.1038/S41586-022-04817-8 (2022). sourcePartly supported

  The Cardiac fibrosis assay and Methods sections (pages unavailable) describe stimulating cultured human cardiac fibroblasts with TGFβ1, supporting the TGFβ condition cited in the sentence. The paper does not compare cultured cardiac, dermal, and immortalized fibroblasts as models of human cardiac fibroblast diversity. Its Discussion instead says it is not yet known how well the cultured-cell transition reflects the trajectory inferred from single-nucleus data. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites [39] Khalil, H. et al. Fibroblast-specific TGF-β–Smad2/3 signaling underlies cardiac fibrosis. J. Clin. Invest. 127, 3770–3783 (2017). sourceSupported
• cites [40] Stratton, M. S. et al. Dynamic chromatin targeting of BRD4 stimulates cardiac fibroblast activation. Circ. Res. 125, 662 (2019). sourcePartly supported

  The Results section, “BRD4 inhibition blocks TGF-β-mediated cardiac fibroblast activation” (page unavailable), documents TGF-β stimulation of cultured primary adult rat ventricular fibroblasts. The cited paper therefore supports the stimulation condition, but not its application to cultured human fibroblasts. The assessment that the cultures were poor models of human cardiac fibroblast diversity is the citing authors’ own conclusion, not a finding established by this source. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites [41] Bujak, M. & Frangogiannis, N. G. The role of interleukin-1 in the pathogenesis of heart disease. Arch. Immunol. Ther. Exp. (Warsz.) 57, 165 (2009). sourceNot checked — no full text of the source

The Srivastava laboratory showed that Brd4 deletion in macrophages abrogates IL-1β signalling and suppresses fibroblast activation and fibrosis through a MEOX1-dependent regulatory mechanism in a pressure-overload model of heart failure.

“Consistent with our findings, the Srivastava laboratory has shown that Brd4 deletion in macrophages abrogates IL-1β signalling and suppresses fibroblast activation and fibrosis through a MEOX1-dependent regulatory mechanism in a pressure-overload model of HF30.”

• cites [30] Alexanian, M. et al. Chromatin remodelling drives immune–fibroblast crosstalk in heart failure pathogenesis. Nature https://doi.org/10.1038/s41586-024-08085-6 (2024). sourcePartly supported

  In “BRD4 function in Cx3cr1+ cells during heart failure,” Brd4 deletion in Cx3cr1-positive cells reduced fibrosis after pressure overload; “BRD4 in Cx3cr1+ cells activates fibroblasts” reports reduced profibrotic fibroblast programs. The paper also links IL1B to MEOX1-dependent fibroblast activation, but “BRD4 drives proinflammatory transcription” reports decreased Il1b expression after Brd4 deletion, not that IL-1β signalling was abrogated. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Concurrent work established a role for MEOX1 in the pathogenesis of cardiac fibrosis.

“These findings are underscored by concurrent work establishing a role for MEOX1 in the pathogenesis of cardiac fibrosis30.”

• cites [30] Alexanian, M. et al. Chromatin remodelling drives immune–fibroblast crosstalk in heart failure pathogenesis. Nature https://doi.org/10.1038/s41586-024-08085-6 (2024). sourceSupported

IL-1β was selectively expressed by CCR2-positive monocytes and macrophages in the human heart.

“We demonstrated that IL-1β was selectively expressed by CCR2+ monocytes and macrophages in the human heart, consistent with our earlier mouse studies42,43.”

• cites [42] Bajpai, G. et al. Tissue resident CCR2− and CCR2+ cardiac macrophages differentially orchestrate monocyte recruitment and fate specification following myocardial injury. Circ. Res. 124, 263–278 (2019). sourcePartly supported

  The Results section on monocyte-derived macrophages reports higher Il1β expression in recruited CCR2+ macrophages from injured mouse hearts. It does not establish selective expression by CCR2+ monocytes and macrophages in the human heart; the paper mentions human cardiac macrophage subsets but does not report this IL-1β finding in humans. No page numbers are available in our copy. (LLM, conf=0.84)

  Experimental — accuracy not yet measured
• cites [43] Bajpai, G. et al. The human heart contains distinct macrophage subsets with divergent origins and functions. Nat. Med. 24, 1234–1245 (2018). sourcePartly supported

  The paper reports that human cardiac CCR2+ macrophages expressed more IL-1β than CCR2− macrophages (section ‘CCR2+ macrophages represent an inflammatory population’; Fig. 5a–c). It also identifies CCR2+ monocytes in the heart, but the passages do not establish that those monocytes selectively expressed IL-1β. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

Cardiac FAP-positive fibroblasts display similarity to cancer-associated fibroblasts.

“Using cross-tissue omic integration, we show that cardiac FAP+ fibroblasts display similarity to cancer-associated fibroblasts38”

• cites [38] Xin, L. et al. Fibroblast activation protein-α as a target in the bench-to-bedside diagnosis and treatment of tumors: a narrative review. Front. Oncol. 11, 3187 (2021). sourceCould not verify

  The review’s ‘Expression of FAP in Tumors’ section describes FAP as a marker of cancer-associated fibroblasts. It does not address cardiac fibroblasts or compare them with cancer-associated fibroblasts; searches for cardiac FAP-positive fibroblasts and heart or myocardial fibroblasts found no passage establishing the claimed similarity. The citing sentence frames the cross-tissue comparison as the citing authors’ own finding. (LLM, conf=0.73)

  Experimental — accuracy not yet measured

The Ang II/PE infusion model used to study cardiac fibrosis has inherent limitations.

“We used an Ang II/PE infusion model to study cardiac fibrosis which has inherent limitations3,4,7,39”

• cites [3] Rurik, J. G. et al. CAR T cells produced in vivo to treat cardiac injury. Science 375, 91–96 (2022). sourcePartly supported

  The passage labeled Abstract describes using AngII/PE delivered by osmotic mini-pumps to induce cardiac injury in mice and studying the resulting fibrosis. It does not identify inherent limitations of that infusion model, so it supports the model’s use but not the stated qualification. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites [4] Aghajanian, H. et al. Targeting cardiac fibrosis with engineered T cells. Nature 573, 430–433 (2019). sourcePartly supported

  The Fig. 3 caption (page unavailable) confirms that the paper used continuous Ang II/PE infusion in mice to induce cardiac injury and fibrosis. Neither that passage nor the searched passages on model limitations state that the infusion model has inherent limitations, so the citation supports the model identification but not the limitation claim. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites [7] Fu, X. et al. Specialized fibroblast differentiated states underlie scar formation in the infarcted mouse heart. J. Clin. Invest. 128, 2127–2143 (2018). sourcePartly supported

  The Results section (‘Cardiac fibroblasts in the scar have a unique differentiated state’; page unavailable) describes using Ang II/PE infusion and reports that it induced proliferation in fibroblasts outside the infarct scar, but not within it. This supports identification of the model and shows a context-dependent response; the paper does not state that the infusion model has inherent limitations. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites [39] Khalil, H. et al. Fibroblast-specific TGF-β–Smad2/3 signaling underlies cardiac fibrosis. J. Clin. Invest. 127, 3770–3783 (2017). sourceCould not verify

  The Results (p. 3) describe a transverse aortic constriction model, while the Introduction (p. 1) identifies a limitation of earlier global Smad3-deletion studies: their lack of cell-type specificity. Searches for Ang II/PE, phenylephrine, infusion, and model limitations found no passage addressing inherent limitations of an Ang II/PE infusion model. (LLM, conf=0.77)

  Experimental — accuracy not yet measured

CellOracle gene regulatory network analysis associated with HF fibroblasts identified enriched connectivity around RUNX1.

“We then used CellOracle to build a gene regulatory network associated with HF fibroblasts, which identified enriched connectivity around RUNX1 (Extended Data Fig. 5c).”

KEGG pathway analysis using F9 marker genes revealed enrichment for extracellular matrix remodelling, cell-cell adhesion, protein digestion, PI3K-Akt signalling, and AGE-RAGE pathways.

“Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis using F9 marker genes revealed enrichment for extracellular matrix remodelling, cell–cell adhesion, protein digestion, PI3K-Akt signalling and AGE-RAGE pathways (Fig. 2f and Supplementary Table 10).”

AMI and chronic heart failure hearts contained the greatest proportion of F9 (FAP/POSTN) and F2 (myofibroblast) fibroblasts.

“AMI and chronic HF hearts contained the greatest proportion of F9 (FAP/POSTN) and F2 (myofibroblasts) fibroblasts (Fig. 2b and Extended Data Fig. 3h).”

Markers of F2, F4, and F9 fibroblasts increased over pseudotime in a lineage-specific manner.

“Notably, the markers of F2, F4 and F9 fibroblasts increased over pseudotime in a lineage-specific manner (Extended Data Fig. 5b, bottom).”

Genes linked to differentially accessible peaks using paired RNA-ATAC data yielded 3,167 unique linked genes.

“We then identified genes linked to these differentially accessible peaks using paired RNA–ATAC data and found 3,167 unique linked genes (Fig. 2q and Supplementary Table 14).”

scVelo velocity vector field projection suggested convergence toward F2, F4, and F9 fibroblast states.

“Projection of the velocity vector field suggested convergence toward the F2, F4 and F9 fibroblast states (Fig. 2i).”

Different fibroblast cell states were differentially enriched across myocardial regions: infarct zone (F2, F8, F9), border zone (F5, F6, F7, F10, F13), and remote zone (F11, F12).

“We found that different fibroblast cell states were differentially enriched across myocardial regions: infarct zone (F2, F8, F9), border zone (F5, F6, F7, F10, F13) and remote zone (F11, F12) (Extended Data Fig. 4f).”

Differential accessibility in fibroblasts between non-failing donors and failing hearts found 15,373 peaks increased and 12,701 peaks reduced in HF.

“We found 15,373 peaks increased and 12,701 peaks reduced in HF.”

Candidate gene predictors of chromatin composed of transcription factors were linked to dynamic regulatory elements.

“Intriguingly, a number of candidate GPCs composed of transcription factors were linked to dynamic regulatory elements, highlighting the potential for Multiome sequencing to identify cell type-specific gene regulatory networks.”

Transcription factor enrichment analysis using marker genes identified MEOX1 as the top transcription factor driving the F9 lineage.

“This analysis identified MEOX1 as the top transcription factor driving the F9 lineage (Fig. 2o and Supplementary Tables 12 and 13).”

Differential gene expression analysis identified 11 distinct cell types with canonical gene and protein signatures.

“Differential gene expression (DGE) analysis identified 11 distinct cell types (Fig. 1d and Extended Data Fig. 1b–d) with canonical gene (Fig. 1c, Extended Data Fig. 1a–c and Supplementary Table 2) and protein (Fig. 1d, Extended Data Fig. 1f and Supplementary Table 3) signatures.”

Most cardiac fibroblasts differed from fibroblasts found in other tissues and diseases.

“To assess similarity among integrated fibroblast clusters, we calculated Pearson correlation coefficients between pairwise cell states and found that most cardiac fibroblasts differed from fibroblasts found within other tissues and diseases.”

Palantir identified F2, F4, and F9 as terminal states with low entropy across simulation parameters.

“Next, we used Palantir to predict fibroblast differentiation endpoints, which also identified F2, F4 and F9 as terminal states with low entropy values across a range of simulation parameters (Fig. 2j and Extended Data Fig. 5a).”

Cell interaction analysis and spatial transcriptomics in human heart tissue identified immune-fibroblast neighbourhoods within fibrotic areas that were predicted to be driven by IL-1β signalling from macrophages to activated fibroblasts.

“Using cell interaction analysis and spatial transcriptomics in human heart tissue, we identified immune–fibroblast neighbourhoods within areas of fibrosis that were predicted to be driven by interleukin-1β (IL-1β) signalling from macrophages to activated fibroblasts.”

Donor hearts contained the greatest number of F4 (APOE/AGT) and F5 (DLK1/GPX3) fibroblasts.

“Donor hearts contained the greatest number of F4 (APOE/AGT) and F5 (DLK1/GPX3) fibroblasts.”

Merging this gene list with pseudobulk RNA differential expression comparing donor and HF fibroblasts found 234 overlapping genes, including POSTN, MEOX1, and RUNX1.

“We merged this gene list with the pseudobulk RNA differential expression data comparing fibroblasts in donor and HF samples and found 234 overlapping genes (Fig. 2q), which included POSTN, MEOX1 and RUNX1 (Supplementary Table 15).”

Pseudobulk differential gene expression at the patient level showed strong differences between healthy donor and heart failure conditions within fibroblasts, endothelial cells, smooth muscle cells, pericytes, and myeloid cells.

“Pseudobulk DGE analysis at the patient level showed strong differences between healthy donor and heart failure (HF) (AMI, ICM, NICM) conditions within fibroblasts, endothelial cells, smooth muscle cells (SMCs) and pericytes, and myeloid cells (Fig. 1f and Supplementary Table 4).”

RUNX1 was co-expressed with POSTN at single-cell and patient levels.

“Consistent with this finding, RUNX1 was co-expressed with POSTN at the single-cell and patient levels (Extended Data Figs. 4b and 5d) and we previously showed expansion of RUNX1-expressing cells in human HF29.”

JQ1 treatment, which improved left ventricle systolic function and abrogated fibrosis, was associated with loss of F9 fibroblasts.

“Intriguingly, we observed that JQ1 treatment, which resulted in improved left ventricle systolic function and abrogation of fibrosis, was associated with the loss of F9 fibroblasts.”

Mapping mouse fibroblasts onto human CITE-seq fibroblasts using label-transfer showed strong mapping scores across most cell states except F5, F6, and F7, which were poorly represented in the mouse data.

“Mapping scores were strong across most cell states except for F5, F6 and F7 fibroblasts, which were poorly represented in the mouse data (Fig. 3a and Extended Data Fig. 6a).”

Unbiased clustering of CITE-seq data identified 13 distinct fibroblast cell states marked by specific gene signatures.

“Unbiased clustering of our CITE-seq data identified 13 distinct fibroblast cell states (Fig. 2a), marked by Pl16/CCDC80/FBLN1 (F1, ground state), ACTA2/TAGLN (F2, myofibroblasts), CCL2/THBS1 (F3), APOE/AGT (F4), DLK1/GPX3 (F5), PTGDS/GPC3 (F6), APOD/IGFBP5 (F7), GDF15/ATF5 (F8), FAP/POSTN (F9), ISG15/MX1 (F10), PLA2G2A (F11), PCOLCE2/MFAP5 (F12) and PRG4/CXCL14 (F13) (Fig. 2b).”

Genes differentially expressed in more abundant cell types mapped to unique ATAC clusters.

“Genes differentially expressed in the more abundant cell types (cardiomyocytes, myeloid, fibroblasts, endothelium, pericytes, T and natural killer cells, and endocardium) mapped to unique ATAC clusters (Extended Data Fig. 2d–f).”

FAP expression progressively increases along the F9 lineage.

“FAP expression progressively increases along the F9 lineage.”

MEOX1 was co-expressed with RUNX1 and POSTN along the F9 lineage.

“Moreover, MEOX1 was co-expressed with RUNX1 and POSTN along the F9 lineage.”

Ang II/PE infusion resulted in diminished interstitial fibrosis in IL-1βflox/flox CCR2creERT2 mice compared with IL-1βflox/flox mice.

“Ang II/PE infusion resulted in diminished interstitial fibrosis in IL-1βflox/flox CCR2creERT2 mice compared with IL-1βflox/flox mice (Fig. 5e).”

Strong correlation was found between macrophage IL-1β and fibroblast POSTN expression in individual cardiac tissue specimens.

“We found strong correlation between macrophage IL-1β and fibroblast POSTN expression in individual cardiac tissue specimens and detected enriched NF-kB pathway signalling in corresponding regions (Extended Data Fig. 9h).”

SMCs, pericytes, and endothelial cells co-localized in a distinct perivascular niche.

“SMCs, pericytes and endothelial cells co-localized in a distinct perivascular niche (Fig. 4e, Extended Data Fig. 9a–g and Supplementary Tables 21 and 22).”

Flow cytometry showed an increase in the proportion of FAP+ fibroblasts in Ang II/PE-treated animals that received isotype antibodies compared with corresponding controls.

“Quantification of flow cytometry data revealed an increase in the proportion of FAP+ fibroblasts in Ang II/PE-treated animals that received isotype antibodies compared with corresponding controls.”

Ang II/PE-treated mice had greater FAPi signal compared with sham.

“Ang II/PE-treated mice had greater fibroblast activator protein inhibitor (FAPi) signal compared with sham.”

In human hearts, CCR2-positive monocytes and macrophages co-localize with FAP/POSTN fibroblasts, forming an immune-fibroblast niche.

“Using spatial transcriptomic data from human hearts, we showed that CCR2+ monocytes and macrophages co-localize with FAP/POSTN fibroblasts, forming an immune–fibroblast niche.”

Anti-IL-1β mAb treatment also reduced expression of the myofibroblast marker Acta2.

“Anti-IL-1β mAb treatment also reduced the expression of the myofibroblast marker, Acta2.”

Ang II/PE infusion resulted in increased interstitial fibrosis in control mice compared with shams.

“Ang II/PE infusion resulted in increased interstitial fibrosis in control mice compared with shams.”

There were fewer FAP+ fibroblasts in IL-1Rflox/flox Dermo1Cre mice compared with control mice after Ang II/PE.

“Importantly, there were fewer FAP+ fibroblasts in IL-1Rflox/flox Dermo1Cre mice compared with control mice (Fig. 4k and Supplementary Fig. 6).”

After deconvolution, fibroblasts most strongly co-localize with myeloid cells and T cells, forming an immune-fibroblast niche.

“After deconvolution, we calculated cell–cell correlations and found fibroblasts most strongly co-localize with myeloid cells and T cells, forming an immune–fibroblast niche.”

The anti-IL-1β mAb group had fewer Postn+ cells (Fib 1) compared with isotype-treated mice.

“Notably, the anti-IL-1β mAb group had fewer Postn+ cells (Fib 1; Extended Data Fig. 10e, bottom panel) and decreased expression of key pro-fibrotic genes such as Aebp1, Fap, Postn, Thbs1, Meox1, Runx1, Clip2, Comp, Col1a1, Thy1 and Cthrc1 compared with isotype-treated mice.”

CCR2+ monocytes/macrophages and DC2s co-localized in the infarct region using spatial transcriptomics.

“each of which co-localized in the infarct region using spatial transcriptomics data (Fig. 4c and Extended Data Fig. 8g,h).”

POSTN immunostaining and quantitative PCR of left-ventricle tissue showed decreased POSTN expression at both protein and RNA levels, corroborating the single-cell RNA sequencing results.

“POSTN immunostaining and quantitative PCR from left-ventricle tissue showed decreased POSTN expression at the protein and RNA levels, corroborating our scRNA-seq (Extended Data Fig. 10k,l).”

Mice treated with anti-IL-1β mAb demonstrated less 68Ga-FAPi-46 uptake compared with mice treated with isotype control.

“Mice treated with anti-IL-1β mAb demonstrated less 68Ga-FAPi-46 uptake compared with mice treated with isotype control (Fig. 5h).”

Increased cardiomyocyte cross-sectional area was found in Ang II/PE-treated controls compared with shams.

“We found increased cardiomyocyte cross-sectional area in Ang II/PE-treated controls compared with shams.”

Immunofluorescence for CCR2/FAP/CD68 in human left-ventricle sections showed co-localization of FAP+ fibroblasts with CCR2+ macrophages in the failing left ventricle.

“We used immunofluorescence for CCR2/FAP/CD68 in human left-ventricle sections to show co-localization of FAP+ fibroblasts with CCR2+ macrophages in the failing left ventricle (Fig. 4g).”

With Ang II/PE infusion, there were fewer FAP+ fibroblasts in IL-1βflox/flox CCR2creERT2 mice compared with IL-1βflox/flox mice at day 7.

“Notably, with Ang II/PE infusion, there were fewer FAP+ fibroblasts in IL-1βflox/flox CCR2creERT2 mice compared with IL-1βflox/flox mice at day 7 (Fig. 5b).”

Anti-IL-1β mAb treatment reduced expression of key pro-fibrotic genes including Aebp1, Fap, Postn, Thbs1, Meox1, Runx1, Clip2, Comp, Col1a1, Thy1, and Cthrc1 compared with isotype-treated mice.

“decreased expression of key pro-fibrotic genes such as Aebp1, Fap, Postn, Thbs1, Meox1, Runx1, Clip2, Comp, Col1a1, Thy1 and Cthrc1 compared with isotype-treated mice.”

Reduced trichrome staining was observed in Ang II/PE-treated IL-1Rflox/floxDermo1Cre mice compared with Ang II/PE-treated controls.

“we observed reduced trichrome staining in Ang II/PE-treated IL-1Rflox/floxDermo1Cre mice compared with Ang II/PE-treated controls (Fig. 4l).”

IL-1R was expressed within cardiac fibroblasts including F9 fibroblasts as well as the endocardium and a subset of endothelial cells.

“IL-1R was expressed within cardiac fibroblasts including F9 fibroblasts as well as the endocardium and a subset of endothelial cells (Extended Data Fig. 9i).”

Single-cell RNA sequencing of FAP-positive and FAP-negative fibroblasts from mice infused with Ang II/PE for 7 days and treated with isotype or anti-IL-1β antibody resolved ten transcriptionally distinct fibroblast cell states.

“Cells were clustered into ten transcriptionally distinct fibroblast cell states (Extended Data Fig. 10f–h and Supplementary Table 26).”

scRNA-seq on sorted fibroblasts at day 7 found seven transcriptionally distinct fibroblast cell states.

“We found seven transcriptionally distinct fibroblast cell states (Fig. 5i, Extended Data Fig. 10a–c and Supplementary Table 24) and expansion of Postn+ fibroblasts in Ang II/PE relative to sham (Extended Data Fig. 10d,e).”

PHCFs and PHDFs recovered a broader diversity of cell states compared with iHCFs.

“PHCFs and PHDFs recovered a broader diversity of cell states compared with iHCFs.”

RNAscope for POSTN and CCR2 in post-MI hearts showed more CCR2+ cells in areas with high numbers of POSTN+ cells compared with areas with fewer POSTN+ cells.

“Quantification showed more CCR2+ cells in areas with high numbers of POSTN+ cells compared with areas with fewer POSTN+ cells (Fig. 4i).”

Both mouse models displayed diminished fibrosis.

“Both mouse models displayed reduced FAP+ fibroblasts, diminished fibrosis and improved cardiac function.”

Anti-IL-1β antibody treatment produced differences in gene expression in both FAP-positive and FAP-negative fibroblasts.

“We detected differences in gene expression in both the FAP+ and FAP− fibroblasts following anti-IL-1β mAb treatment (Extended Data Fig. 10j).”

Echocardiography after 28 days of Ang II/PE showed significant decrease in E/e′ and end systolic volume with significant improvement in fractional shortening and ejection fraction in IL-1βflox/flox CCR2creERT2 mice compared with IL-1βflox/flox mice.

“We observed a significant decrease in E/e′ and end systolic volume with a significant improvement in fractional shortening and ejection fraction in IL-1βflox/flox CCR2creERT2 mice compared with IL-1βflox/flox mice (Fig. 5c,d).”

Linear regression showed a strong correlation between pseudobulk macrophage IL-1β expression and fibroblast POSTN expression.

“Linear regression analysis showed a strong correlation between pseudobulk macrophage IL-1β expression with fibroblast POSTN expression (Fig. 4h).”

Mouse fibroblasts across cardiac injury models showed substantially greater transcriptional similarity to human cardiac fibroblasts relative to human cell culture preparations.

“Collectively, mouse fibroblasts present across cardiac injury models showed substantially greater transcriptional similarity to human cardiac fibroblasts relative to human cell culture preparations (Fig. 3h and Extended Data Fig. 7d,e).”

Mouse models recapitulated the expansion of Fap- and Postn-expressing fibroblasts following injury.

“and recapitulated expansion of Fap- and Postn-expressing fibroblasts following injury.”

All cultured cell lines displayed a limited ability to recapitulate human cardiac fibroblast states.

“All cultured cell lines displayed a limited ability to recapitulate human cardiac fibroblast states.”

FAP-negative fibroblasts showed reduced expression of ApoE, Gdf15, Ptgds, Ccl2 and Prg4 following anti-IL-1β antibody treatment.

“FAP− fibroblasts showed reduced expression of ApoE, Gdf15, Ptgds, Ccl2 and Prg4 following anti-IL-1β mAb treatment (Extended Data Fig. 10j).”

Reduction in FAP+ fibroblasts was observed in Ang II/PE-treated mice that received anti-IL-1β mAb compared with Ang II/PE-treated mice that received isotype antibody.

“Importantly, we observed a reduction in FAP+ fibroblasts in Ang II/PE-treated mice that received anti-IL-1β mAb compared with Ang II/PE-treated mice that received isotype antibody (Fig. 5g).”

Both mouse models displayed improved cardiac function.

“Both mouse models displayed reduced FAP+ fibroblasts, diminished fibrosis and improved cardiac function.”

The gene set downregulated in the anti-IL-1β mAb treatment group was expressed predominately in Postn+ fibroblasts.

“This gene set was expressed predominately in Postn+ fibroblasts (Fig. 5i).”

Tagln and Sparc expression were not significantly impacted by anti-IL-1β mAb treatment.

“However, Tagln and Sparc were not significantly impacted (Fig. 5j).”

MEOX1 was the most enriched transcription factor among putative epigenetic regulators of FAP/POSTN fibroblasts.

“We further characterized putative epigenetic regulators of FAP/POSTN fibroblasts and found MEOX1 to be the most enriched transcription factor.”

Anti-IL-1β antibody treatment increased expression of myofibroblast markers including Tagln, Acta2 and Svep1 in FAP-positive fibroblasts.

“Anti-IL-1β mAb treatment increased the expression of myofibroblast markers (including Tagln, Acta2 and Svep1) in FAP+ fibroblasts (Extended Data Fig. 10j)”

Pathway analysis suggested reductions in TGFβ regulation of extracellular matrix in the anti-IL-1β mAb group relative to isotype controls.

“Pathway analysis suggested reductions in TGFβ regulation of extracellular matrix in the anti-IL-1β mAb group relative to isotype controls (Fig. 5k and Supplementary Table 25).”

Mouse cardiac injury models contained many of the fibroblast populations found in the human heart.

“We found that mouse models contained many of the fibroblast populations found in the human heart”

In vivo, IL-1β expression was conserved in mouse CCR2+ monocytes and macrophages in the Ang II/PE infusion model.

“In vivo, IL-1β expression was conserved in mouse CCR2+ monocytes and macrophages in the Ang II/PE infusion model (Fig. 4d).”

Enriched NF-kB pathway signalling was detected in regions corresponding to macrophage IL-1β and fibroblast POSTN correlation.

“and detected enriched NF-kB pathway signalling in corresponding regions (Extended Data Fig. 9h).”

Cell-cell interaction analysis identified enrichment of IL-1β and TGFβ signalling in fibroblasts in heart failure.

“Using cell–cell interaction analysis, we identified enrichment of IL-1β and TGFβ signalling in fibroblasts in HF.”

In the FAP-positive population, isotype-treated mice showed increased expression of Fap, Thbs4, Isg15, Col1a1 and Cola3.

“Within the FAP+ population, isotype-treated animals displayed increased expression of Fap, Thbs4, Isg15, Col1a1 and Cola3.”

Flow cytometry showed an increase in the proportion of FAP+ fibroblasts in Ang II/PE-treated animals compared with sham surgeries.

“Quantification of flow cytometry data revealed an increase in the proportion of FAP+ fibroblasts in Ang II/PE-treated animals compared with sham surgeries.”

Expansion of Postn+ fibroblasts occurred in Ang II/PE relative to sham.

“and expansion of Postn+ fibroblasts in Ang II/PE relative to sham (Extended Data Fig. 10d,e).”

After 28 days of Ang II/PE infusion via osmotic minipumps, anti-IL-1β antibody treatment decreased cardiac fibrosis.

“after 28 d of Ang II/PE infusion via osmotic minipumps, anti-IL-1β mAb treatment decreased cardiac fibrosis (Fig. 5l and Extended Data Fig. 10m).”

NicheNet analysis identified TGFβ and IL-1β as the strongest predicted signals received by fibroblasts from any cell type.

“We focused on signals received by fibroblasts from any cell type and identified TGFβ and IL-1β as the strongest predicted signals (Fig. 4a).”

There was a significant decrease in cardiomyocyte size in Ang II/PE-treated IL-1Rflox/floxDermo1Cre mice compared with Ang II/PE-treated controls.

“However, there was a significant decrease in cardiomyocyte size in Ang II/PE-treated IL-1Rflox/floxDermo1Cre mice compared with Ang II/PE-treated controls (Fig. 4m).”

Blocking IL-1β signalling does not fully ameliorate cardiac fibrosis.

“Blocking IL-1β signalling does not fully ameliorate cardiac fibrosis”

These findings predict FAP/POSTN (F9) fibroblasts are regulated by RUNX1 and MEOX1.

“Collectively, these findings predict that FAP/POSTN (F9) fibroblasts are regulated by RUNX1 and MEOX1.”

These data suggest FAP/POSTN (F9) fibroblasts were distinct from myofibroblasts (F2) and other predicted terminal fibroblast cell states.

“These data suggest that FAP/POSTN (F9) fibroblasts were distinct from myofibroblasts (F2) and other predicted terminal fibroblast cell states.”

Fibroblast cell state clustering was validated by gene set and surface protein signatures.

“Fibroblast cell state clustering was validated by gene set and surface protein signatures (Extended Data Fig. 3a–e and Supplementary Tables 8 and 9).”

IL-1β signalling from CCR2+ macrophages to fibroblasts is causally linked to cardiac fibrosis in vivo.

“Finally, we demonstrated that IL-1β signalling from C-C chemokine receptor type 2 (CCR2) macrophages to fibroblasts is causally linked to cardiac fibrosis in vivo.”

RNAscope and immunofluorescence for POSTN and ACTA2 in human infarct tissue validated that F2 and F9 fibroblast populations are distinct cell types.

“We then validated that F2 and F9 fibroblast populations are distinct cell types using RNAscope and immunofluorescence for POSTN and ACTA2, respectively, in the human infarct tissue (Fig. 2k).”

These findings implicate F9 fibroblasts as a potential pathogenic cell state.

“These findings implicate F9 fibroblasts as a potential pathogenic cell state.”

The FAP/POSTN fibroblast trajectory contributes to cardiac fibrosis.

“and contributes to cardiac fibrosis.”

The study did not have adequate sample size to independently assess the effects of ethnicity, sex and age in the human dataset.

“Although we chose to include patients of diverse ethnicities, sexes and ages in the generation of the human dataset, we did not have adequate an sample size available to independently assess the effects of these variables.”

IL-1β signalling plays a selective role in the FAP/POSTN fibroblast lineage.

“highlighting a selective role of IL-1β signalling in the FAP/POSTN fibroblast lineage.”

The FAP/POSTN fibroblast trajectory is governed by inflammatory cytokines derived from monocytes and macrophages.

“governed by inflammatory cytokines derived from monocytes and derived macrophages”

CCR2-positive monocytes and macrophages are key regulators of fibrosis in heart failure.

“Collectively, these findings implicate CCR2+ monocytes and macrophages as key regulators of fibrosis in HF”

Using CITE-seq, Multiome and spatial transcriptomics, the authors identified a fibroblast trajectory marked by FAP and POSTN expression that is independent of myofibroblasts.

“We used CITE-seq, Multiome and spatial transcriptomics to dissect fibroblast cell states that emerge following AMI and in chronic HF and identified a fibroblast trajectory marked by FAP and POSTN expression that is independent of myofibroblasts”

Tamoxifen dosing in CCR2creERT2 mice specifically targets blood monocytes and monocyte-derived macrophages.

“and used a tamoxifen dosing strategy to specifically target blood monocytes and monocyte-derived macrophages34,35.”

• cites [34] Kadyrov, F. F. et al. Hypoxia sensing in resident cardiac macrophages regulates monocyte fate specification following ischemic heart injury. Preprint at bioRxiv https://doi.org/10.1101/2022.08.04.502542 (2023). sourceNot checked — no full text of the source
• cites [35] Koenig, A. L. et al. Genetic mapping of monocyte fate decisions following myocardial infarction. Preprint at bioRxiv https://doi.org/10.1101/2023.12.24.573263 (2023). sourceNot checked — no full text of the source

Targeting inflammatory mediators including IL-1β has potential as a therapeutic approach.

“and highlight the potential for targeting inflammatory mediators including IL-1β.”

Quantitative PCR verified decreased IL-1R expression in fibroblasts but not macrophages from IL-1Rflox/floxDermo1Cre mice 7 days after Ang II/PE infusion.

“Quantitative PCR verified decreased IL-1R expression in fibroblasts but not macrophages (Extended Data Fig. 9j).”

These data indicate mouse cardiac injury models are a reasonable experimental model to explore mechanisms involved in FAP/POSTN (F9) fibroblast emergence and cardiac fibrosis.

“These data indicate that mouse cardiac injury models are a reasonable experimental model to explore mechanisms involved in FAP/POSTN (F9) fibroblast emergence and cardiac fibrosis.”

Immunomodulators have a promising role in targeting cardiac fibrosis.

“Collectively, our findings highlight a promising role for immunomodulators in targeting cardiac fibrosis.”

IL-1β produced by CCR2-positive monocytes and macrophages regulates FAP/POSTN fibroblast specification by directly signalling to cardiac fibroblasts.

“We demonstrated that IL-1β produced by CCR2+ monocytes and macrophages regulates FAP/POSTN fibroblast specification by directly signalling to cardiac fibroblasts.”

Factors other than IL-1β signalling may contribute to cardiac fibrosis.

“suggesting there may be more factors that contribute to cardiac fibrosis.”