CoChat

PRRX1 is a master transcription factor of stromal fibroblasts for myofibroblastic lineage progression

Open access

The DOI matches its Crossref record; 66 of the 66 references it cites were found in a registry or an index.

Abstract

Although stromal fibroblasts play a critical role in cancer progression, their identities remain unclear as they exhibit high heterogeneity and plasticity. Here, a master transcription factor (mTF) constructing core-regulatory circuitry, PRRX1, which determines the fibroblast lineage with a myofibroblastic phenotype, is identified for the fibroblast subgroup. PRRX1 orchestrates the functional drift of fibroblasts into myofibroblastic phenotype via TGF-β signaling by remodeling a super-enhancer landscape. Such reprogrammed fibroblasts have myofibroblastic functions resulting in markedly enhanced tumorigenicity and aggressiveness of cancer. PRRX1 expression in cancer-associated fibroblast (CAF) has an unfavorable prognosis in multiple cancer types. Fibroblast-specific PRRX1 depletion induces long-term and sustained complete remission of chemotherapy-resistant cancer in genetically engineered mice models. This study reveals CAF subpopulations based on super-enhancer profiles including PRRX1. Therefore, mTFs, including PRRX1, provide another opportunity for establishing a hierarchical classification system of fibroblasts and cancer treatment by targeting fibroblasts.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-022-30484-4

Cites

66 of 66 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Discovery of PRRX1 as a highly CAF-specific TF with significant correlation with poor prognosis in diverse cancer types. A Schematic diagram showing the method used to identify activated CAF-specific TFs in six cancer tissues compared with that in normal fibroblasts using large-scale single-cell RNA sequencing (scRNA-seq) datasets. PRRX1 was identified as a CAF-specific TF and was highly correlated with the CAF activation score. B Level of PRRX1 expression in normal fibroblasts and CAFs in colorectal, stomach, and lung cancers (two-sided Wilcoxon rank-sum test, ***p < 2e−16, Colorectal cancer n = 2047, Stomach cancer n = 1821, Lung cancer n = 2002). Others are shown in Supplementary Fig. 1B. C Bar plots of the proportion of CAFs with high-PRRX1 expression in the eight scRNA-seq datasets. The criteria used for categorizing high (+)- and low (−)-PRRX1 CAFs are shown in Supplementary Fig. 1C. D Analysis of eight scRNA-seq datasets revealed that PRRX1 expression is highly restricted to CAFs across 11 cell types (two-sided Wilcoxon rank-sum test, ***p < 2e−16). Others are shown in Supplementary Fig. 2E. E Immunohistochemical analyses revealed that PRRX1 protein is specifically detected in CAFs in all five cancer types. Representative Images from five FFPE tumor microarray sections (TMA). F Kaplan–Meier (KM) survival analyses using immunohistochemistry revealed that a high level of CAF-specific PRRX1 expression is significantly associated with unfavorable overall survival in the colon, stomach, and esophageal cancers. p value was calculated using log-rank test: *p < 0.01, **p < 0.001, ***p < 0.0001. G, H (i) Human stomach cancer cells (MKN28) alone, (ii) mixed with SNF32 (human stomach normal fibroblasts)-FLAG (control), or either (iii) Prrx1a- or (iv) Prrx1b-overexpressing SNF32 were co-injected subcutaneously into NSG mice. Measurement of primary tumor volume and H&E images of tumors in each group. G Data are presented as the mean ± SEM, n = 8 tumors per group, two tailed t test: *p < 0.005, **p < 0.0001. H Box plot showing number of lung metastasis(n = 6 mice/group). p value was calculated using two tailed t test: *p < 0.05. Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). I Schematic summary of the results. Source data and exact p values are provided as a Source Data file.
Fig. 1 Discovery of PRRX1 as a highly CAF-specific TF with significant correlation with poor prognosis in diverse cancer types. A Schematic diagram showing the method used to identify activated CAF-specific TFs in six cancer tissues compared with that in normal fibroblasts using large-scale single-cell RNA sequencing (scRNA-seq) datasets. PRRX1 was identified as a CAF-specific TF and was highly correlated with the CAF activation score. B Level of PRRX1 expression in normal fibroblasts and CAFs in colorectal, stomach, and lung cancers (two-sided Wilcoxon rank-sum test, ***p < 2e−16, Colorectal cancer n = 2047, Stomach cancer n = 1821, Lung cancer n = 2002). Others are shown in Supplementary Fig. 1B. C Bar plots of the proportion of CAFs with high-PRRX1 expression in the eight scRNA-seq datasets. The criteria used for categorizing high (+)- and low (−)-PRRX1 CAFs are shown in Supplementary Fig. 1C. D Analysis of eight scRNA-seq datasets revealed that PRRX1 expression is highly restricted to CAFs across 11 cell types (two-sided Wilcoxon rank-sum test, ***p < 2e−16). Others are shown in Supplementary Fig. 2E. E Immunohistochemical analyses revealed that PRRX1 protein is specifically detected in CAFs in all five cancer types. Representative Images from five FFPE tumor microarray sections (TMA). F Kaplan–Meier (KM) survival analyses using immunohistochemistry revealed that a high level of CAF-specific PRRX1 expression is significantly associated with unfavorable overall survival in the colon, stomach, and esophageal cancers. p value was calculated using log-rank test: *p < 0.01, **p < 0.001, ***p < 0.0001. G, H (i) Human stomach cancer cells (MKN28) alone, (ii) mixed with SNF32 (human stomach normal fibroblasts)-FLAG (control), or either (iii) Prrx1a- or (iv) Prrx1b-overexpressing SNF32 were co-injected subcutaneously into NSG mice. Measurement of primary tumor volume and H&E images of tumors in each group. G Data are presented as the mean ± SEM, n = 8 tumors per group, two tailed t test: *p < 0.005, **p < 0.0001. H Box plot showing number of lung metastasis(n = 6 mice/group). p value was calculated using two tailed t test: *p < 0.05. Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). I Schematic summary of the results. Source data and exact p values are provided as a Source Data file.CC BY · PubMed Central
Fig. 2 Prrx1 expression in fibroblasts is essential for in vivo tumorigenicity and metastasis in genetically engineered immunocompetent mice. A Mouse model in which fibroblast-specific Prrx1 was depleted using sgRNA against Prrx1 (sgPrrx1) before LLC1 transplantation. Bioluminescence imaging of representative mice in each group by date. sgNS was used as the control (n = 6 mice: sgNS group, n = 10 mice: sgPrrx1 group, *p < 0.005 by chi-squared test). B Transgenic mice (Col1a2;rtTATetO7-Prrx1Luc) capable of fibroblast-specific Prrx1 induction at a desired time point using the Tet-on system. Tumor growth curves; Data are presented as the mean ± SEM; n = 7 mice: Dox- group, n = 10: Dox+ group (two tailed t test: *p < 0.005). Table comparing the incidence of secondary metastases. H&E staining of representative lung and liver metastasis in Prrx1-induced mice. *p < 0.05, **p < 0.01, ***p < 0.005 by chi-squared test. C Prrx1 induction in fibroblasts after cutaneous transplantation of mouse colon cancer cells MC38-Luc-GFP (tumor formation, skin incision, and tumor removal, followed by MRI for 4 weeks). Representative MRI (yellow arrow indicates metastatic growth near the lung) and H&E stained images of metachronous metastasis from Prrx1-induced mice (n = 10 mice: Dox− group, n = 15 mice: Dox+ group). D Timeline for in vivo model of AOM/DSS-induced colorectal cancer. (n = 8 mice: AOM/DSS+ Dox- group, n = 13 mice: AOM/DSS+ Dox− group) Masson’s trichrome staining for assessing desmoplasia in mice and quantification of collagen deposition. Box plots showing quantification of collagen deposition(n = 6 measurements, *p < 0.001 by two tailed t test). Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). Immunohistochemical staining was conducted using AOM/DSS treatment in WT and fibroblast-specific Prrx1-inducible (Col1a2;rtTATetO7-Prrx1Luc) mouse tissues. Visualization using ImageJ Fiji (DAPI [hematoxylin]-blue, αSMA-green, Prrx1-red) and quantification of staining images. Scale bar, 100 μm. Bar plot showing quantification of aSMA and Prrx1 intensity. Data are presented as the mean ± SEM; n = 5 measurements (two tailed t test: *p < 0.0001). Source data and exact p values are provided as a Source Data file.
Fig. 2 Prrx1 expression in fibroblasts is essential for in vivo tumorigenicity and metastasis in genetically engineered immunocompetent mice. A Mouse model in which fibroblast-specific Prrx1 was depleted using sgRNA against Prrx1 (sgPrrx1) before LLC1 transplantation. Bioluminescence imaging of representative mice in each group by date. sgNS was used as the control (n = 6 mice: sgNS group, n = 10 mice: sgPrrx1 group, *p < 0.005 by chi-squared test). B Transgenic mice (Col1a2;rtTATetO7-Prrx1Luc) capable of fibroblast-specific Prrx1 induction at a desired time point using the Tet-on system. Tumor growth curves; Data are presented as the mean ± SEM; n = 7 mice: Dox- group, n = 10: Dox+ group (two tailed t test: *p < 0.005). Table comparing the incidence of secondary metastases. H&E staining of representative lung and liver metastasis in Prrx1-induced mice. *p < 0.05, **p < 0.01, ***p < 0.005 by chi-squared test. C Prrx1 induction in fibroblasts after cutaneous transplantation of mouse colon cancer cells MC38-Luc-GFP (tumor formation, skin incision, and tumor removal, followed by MRI for 4 weeks). Representative MRI (yellow arrow indicates metastatic growth near the lung) and H&E stained images of metachronous metastasis from Prrx1-induced mice (n = 10 mice: Dox− group, n = 15 mice: Dox+ group). D Timeline for in vivo model of AOM/DSS-induced colorectal cancer. (n = 8 mice: AOM/DSS+ Dox- group, n = 13 mice: AOM/DSS+ Dox− group) Masson’s trichrome staining for assessing desmoplasia in mice and quantification of collagen deposition. Box plots showing quantification of collagen deposition(n = 6 measurements, *p < 0.001 by two tailed t test). Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). Immunohistochemical staining was conducted using AOM/DSS treatment in WT and fibroblast-specific Prrx1-inducible (Col1a2;rtTATetO7-Prrx1Luc) mouse tissues. Visualization using ImageJ Fiji (DAPI [hematoxylin]-blue, αSMA-green, Prrx1-red) and quantification of staining images. Scale bar, 100 μm. Bar plot showing quantification of aSMA and Prrx1 intensity. Data are presented as the mean ± SEM; n = 5 measurements (two tailed t test: *p < 0.0001). Source data and exact p values are provided as a Source Data file.CC BY · PubMed Central
Fig. 3 Prrx1 orchestrates myofibroblast-like functions in CAFs across diverse cancer types, enhancing EMT and stemness of tumor cells. A In public scRNA-seq data, human CAFs were virtually sorted into high (+)- and low (−)-PRRX1 CAFs. Both high- and low-Prrx1 CAFs were compared using module score analysis. Ex vivo fibroblast cultures of either Prrx1 WT or Prrx1 KO for each type of mouse fibroblast were established. Then, the two groups were compared for each type of fibroblast using GSVA analysis. Heatmap shows the module score from scRNA-seq of human CAFs and GSVA score from bulk RNA-seq of mouse fibroblast models. B After establishing PRRX1 promoter-driven model of SCAF36, apoptosis analysis was performed at each group (PRRX1 low and PRRX1 high). Experiments were conducted in triplicate. Bar plot showing percentage of cell apoptosis(%). Data are presented as the mean ± SEM; n = 3 independent experiments (two tailed t test: *p < 0.0001). C Gel contraction assay, depending on the expression of Prrx1 in fibroblasts. Representative images are shown. Data are presented as the mean ± SEM; n = 3 independent experiments (two tailed t test: *p < 0.005, **p < 0.0001). D Establishment of hanging drop direct 3D spheroid coculture system. Breast cancer cells 168FARN-Luc-GFP were cocultured with either MMTV-CAFs sgNS or sgPrrx1, and lung cancer cells LLC1-Luc-GFP were also cocultured with either wound-healing fibroblasts sgNS or sgPrrx1, using 3D spheroid systems. E H&E and immunohistochemical staining of representative spheroids in each group (DAPI: nuclei, GFP: Luc-GFP-tagged cancer cells). Bar plot showing quantification of spheroid size. Data are presented as the mean ± SEM; n = 7 independent experiments (two tailed t test: *p < 0.001). F GFP+ cancer cells were isolated from spheroids using FACS, and RNA-seq of these cancer cells was performed. G 168FARN-Luc-GFP+ cancer cells were isolated from spheroids using FACS, and then invasion, migration, and sphere forming assays were performed on isolated cancer cells. Data are presented as the mean ± SEM; n = 4 independent experiments (two tailed t test: *p < 0.005, **p < 0.0001). Source data and exact p values are provided as a Source Data file.
Fig. 3 Prrx1 orchestrates myofibroblast-like functions in CAFs across diverse cancer types, enhancing EMT and stemness of tumor cells. A In public scRNA-seq data, human CAFs were virtually sorted into high (+)- and low (−)-PRRX1 CAFs. Both high- and low-Prrx1 CAFs were compared using module score analysis. Ex vivo fibroblast cultures of either Prrx1 WT or Prrx1 KO for each type of mouse fibroblast were established. Then, the two groups were compared for each type of fibroblast using GSVA analysis. Heatmap shows the module score from scRNA-seq of human CAFs and GSVA score from bulk RNA-seq of mouse fibroblast models. B After establishing PRRX1 promoter-driven model of SCAF36, apoptosis analysis was performed at each group (PRRX1 low and PRRX1 high). Experiments were conducted in triplicate. Bar plot showing percentage of cell apoptosis(%). Data are presented as the mean ± SEM; n = 3 independent experiments (two tailed t test: *p < 0.0001). C Gel contraction assay, depending on the expression of Prrx1 in fibroblasts. Representative images are shown. Data are presented as the mean ± SEM; n = 3 independent experiments (two tailed t test: *p < 0.005, **p < 0.0001). D Establishment of hanging drop direct 3D spheroid coculture system. Breast cancer cells 168FARN-Luc-GFP were cocultured with either MMTV-CAFs sgNS or sgPrrx1, and lung cancer cells LLC1-Luc-GFP were also cocultured with either wound-healing fibroblasts sgNS or sgPrrx1, using 3D spheroid systems. E H&E and immunohistochemical staining of representative spheroids in each group (DAPI: nuclei, GFP: Luc-GFP-tagged cancer cells). Bar plot showing quantification of spheroid size. Data are presented as the mean ± SEM; n = 7 independent experiments (two tailed t test: *p < 0.001). F GFP+ cancer cells were isolated from spheroids using FACS, and RNA-seq of these cancer cells was performed. G 168FARN-Luc-GFP+ cancer cells were isolated from spheroids using FACS, and then invasion, migration, and sphere forming assays were performed on isolated cancer cells. Data are presented as the mean ± SEM; n = 4 independent experiments (two tailed t test: *p < 0.005, **p < 0.0001). Source data and exact p values are provided as a Source Data file.CC BY · PubMed Central
Fig. 4 Prrx1 of master transcription factor in activated murine fibroblast. A Identification of candidate mTFs in three different fibroblast types. Total 18 highly expressed mTFs in MEF, 13 in wound-healing fibroblasts, and nine in MMTV-CAF were identified. B Among the mTFs, eight common mTFs in the three fibroblast types were detected. C Prrx1 was one of the eight common mTFs, highly expressed in activated fibroblasts. Gene ontology using target genes with three or more binding motifs for the eight common TFs showed enriched ECM stiffness, TGF-β pathway, angiogenesis, and chromatin remodeling pathway. D Clustering of single cells from mouse skin wound tissue (day 10 after post wound) revealed two clusters of the activated fibroblast group and inactivated fibroblast group, respectively (left). Prrx1 was the most significant highly expressed gene in the activated fibroblast group compared with the inactivated fibroblast group (right) among the candidate mTFs. The detail analysis results of these scRNA-seq were described in supplementary figure 10. E Prrx1 binds to SE with a higher rate than to TE in all three types of fibroblasts. F Super-enhancer peaks in a representative gene such as Col1a1 that are involved in ECM and contraction as well as Prrx1 in all three types of fibroblasts. There were other ChIP-seq profiles for Nfix, Klf13, Tead1, Fosl2 in Supplementary Fig. 11A–D. Source data are provided as a Source Data file.
Fig. 4 Prrx1 of master transcription factor in activated murine fibroblast. A Identification of candidate mTFs in three different fibroblast types. Total 18 highly expressed mTFs in MEF, 13 in wound-healing fibroblasts, and nine in MMTV-CAF were identified. B Among the mTFs, eight common mTFs in the three fibroblast types were detected. C Prrx1 was one of the eight common mTFs, highly expressed in activated fibroblasts. Gene ontology using target genes with three or more binding motifs for the eight common TFs showed enriched ECM stiffness, TGF-β pathway, angiogenesis, and chromatin remodeling pathway. D Clustering of single cells from mouse skin wound tissue (day 10 after post wound) revealed two clusters of the activated fibroblast group and inactivated fibroblast group, respectively (left). Prrx1 was the most significant highly expressed gene in the activated fibroblast group compared with the inactivated fibroblast group (right) among the candidate mTFs. The detail analysis results of these scRNA-seq were described in supplementary figure 10. E Prrx1 binds to SE with a higher rate than to TE in all three types of fibroblasts. F Super-enhancer peaks in a representative gene such as Col1a1 that are involved in ECM and contraction as well as Prrx1 in all three types of fibroblasts. There were other ChIP-seq profiles for Nfix, Klf13, Tead1, Fosl2 in Supplementary Fig. 11A–D. Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 5 Prrx1/Smad3 complex executes a myofibroblastic program induced by TGF-β in fibroblasts. A Total 515 common binding sites of Prrx1 in three different fibroblasts by Prrx1 Chip-seq. The genes with TSS within 1 kb from the 515 binding sites were enriched in various pathways such as ECM remodeling, TGF-β, and angiogenesis pathways, which were features of activated cancer-associated fibroblasts. B 3D matrix remodeling capacity of MEFs depending on the presence of Prrx1 was assessed by the alignment of the cytoskeleton (F-actin) and ECM deposition (fibronectin). TGF-β mediated Prrx1-induced remodeling of. No exhibitions affected Prrx1-induced ECM remodeling in TGF-β untreated groups (Red: fibronectin, Green: F-actin, Blue: DAPI). Box plot showing quantification of fibronection intensity (n = 6 independent experiments, *p < 0.0001 by two tailed t test). Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). C Predicted binding of Prrx1 and Smad3 at 75-bp interval between them using motif analysis. D Tables based on Prrx1 ChIP-seq motif analysis show that Prxx1 cobinds with Smad complexes. E Precipitation of Prrx1 after sonication of MEF genome 100 bp in length, confirmed that Smad3 was located near Prrx1. F, G Prrx1 and Smad3 were simultaneously expressed in the nucleus in the MEF WT group (Prrx1 normal expression) when TGF-β was treated with the MEFs, but not in the MEF Prrx1-mutant group (Impaired Prrx1 expression due to mutation). For (E, F): this experiment was independently repeated three times with similar results. H Schematic model of activated fibroblast induced angiogenesis, ECM stiffness, and TGF-β signaling via Prrx1-Smad3 complex. Source data and exact p values are provided as a Source Data file.
Fig. 5 Prrx1/Smad3 complex executes a myofibroblastic program induced by TGF-β in fibroblasts. A Total 515 common binding sites of Prrx1 in three different fibroblasts by Prrx1 Chip-seq. The genes with TSS within 1 kb from the 515 binding sites were enriched in various pathways such as ECM remodeling, TGF-β, and angiogenesis pathways, which were features of activated cancer-associated fibroblasts. B 3D matrix remodeling capacity of MEFs depending on the presence of Prrx1 was assessed by the alignment of the cytoskeleton (F-actin) and ECM deposition (fibronectin). TGF-β mediated Prrx1-induced remodeling of. No exhibitions affected Prrx1-induced ECM remodeling in TGF-β untreated groups (Red: fibronectin, Green: F-actin, Blue: DAPI). Box plot showing quantification of fibronection intensity (n = 6 independent experiments, *p < 0.0001 by two tailed t test). Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). C Predicted binding of Prrx1 and Smad3 at 75-bp interval between them using motif analysis. D Tables based on Prrx1 ChIP-seq motif analysis show that Prxx1 cobinds with Smad complexes. E Precipitation of Prrx1 after sonication of MEF genome 100 bp in length, confirmed that Smad3 was located near Prrx1. F, G Prrx1 and Smad3 were simultaneously expressed in the nucleus in the MEF WT group (Prrx1 normal expression) when TGF-β was treated with the MEFs, but not in the MEF Prrx1-mutant group (Impaired Prrx1 expression due to mutation). For (E, F): this experiment was independently repeated three times with similar results. H Schematic model of activated fibroblast induced angiogenesis, ECM stiffness, and TGF-β signaling via Prrx1-Smad3 complex. Source data and exact p values are provided as a Source Data file.CC BY · PubMed Central
Fig. 6 Super-enhancer mapping of human colorectal CAF identifies distinct subpopulations. A Typical enhancers (TEs) and super-enhancers (SEs) based on H3K27ac ChIP-seq were identified from 9 CAFs cultured from nine colon cancer tissues and 171 cells from ENCODE database. Many of them from the nine CAFs were overlapped with TEs and SEs from ENCODE data. B Principal component analysis (PCA) using SEs for the nine CAFs and 171 cells from ENCODE revealed two distinct subgroups of the nine CAFs into the ‘Cluster-A’ and ‘Cluster-B’. C ATAC-seq was performed for all nine CAFs and PCA using the peak intensities of ATAC-seq also revealed two subgroups, which corresponded to the two CAF clusters identified using SEs from ChIP-seq. D GO analysis of SE-associated genes (genes containing TSS within 1 kb from the SE) showed that each of the two CAF clusters is related to distinct molecular functions. For the Cluster-A group, the enriched gene ontologies are ECM remodeling, VEGF and TGF-β activation, which are representative myofibroblastic functions, whereas cellular senescence signatures were increased in Cluster-B group. E RNA-seq was performed in the nine CAFs, then RNA-seq data were analyzed by gene set enrichment analysis (GSEA) for key cellular functions. The GSEA analysis of the mRNA expression dataset of these nine CAFs confirmed the result of GO analysis of SE-associated gene shown in (D). F Gel contraction assay revealed increased contraction abilities of CAFs in Cluster-A compared with Cluster-B. Box plots showing quantification of gel contraction (n = 3 independent experiments, *p < 0.001 by two tailed t test). Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). G The invasion and migration of HCT-116 cancer cells were significantly enhanced by conditioned medium (CM) of CAFs in the Cluster-A group compared with Cluster-B (*p < 0.001 by two tailed t test). Data are presented as the mean ± SEM; n = 4 independent experiments. HCT116 derived CM without CAF was used as a control. Source data and exact p values are provided as a Source Data file. CAF, cancer-associated fibroblast; TSS, transcription start site; SE, super-enhancer.
Fig. 6 Super-enhancer mapping of human colorectal CAF identifies distinct subpopulations. A Typical enhancers (TEs) and super-enhancers (SEs) based on H3K27ac ChIP-seq were identified from 9 CAFs cultured from nine colon cancer tissues and 171 cells from ENCODE database. Many of them from the nine CAFs were overlapped with TEs and SEs from ENCODE data. B Principal component analysis (PCA) using SEs for the nine CAFs and 171 cells from ENCODE revealed two distinct subgroups of the nine CAFs into the ‘Cluster-A’ and ‘Cluster-B’. C ATAC-seq was performed for all nine CAFs and PCA using the peak intensities of ATAC-seq also revealed two subgroups, which corresponded to the two CAF clusters identified using SEs from ChIP-seq. D GO analysis of SE-associated genes (genes containing TSS within 1 kb from the SE) showed that each of the two CAF clusters is related to distinct molecular functions. For the Cluster-A group, the enriched gene ontologies are ECM remodeling, VEGF and TGF-β activation, which are representative myofibroblastic functions, whereas cellular senescence signatures were increased in Cluster-B group. E RNA-seq was performed in the nine CAFs, then RNA-seq data were analyzed by gene set enrichment analysis (GSEA) for key cellular functions. The GSEA analysis of the mRNA expression dataset of these nine CAFs confirmed the result of GO analysis of SE-associated gene shown in (D). F Gel contraction assay revealed increased contraction abilities of CAFs in Cluster-A compared with Cluster-B. Box plots showing quantification of gel contraction (n = 3 independent experiments, *p < 0.001 by two tailed t test). Data are presented as box-and-whisker plots indicating median (middle line) and minimum and maximum values (whiskers). G The invasion and migration of HCT-116 cancer cells were significantly enhanced by conditioned medium (CM) of CAFs in the Cluster-A group compared with Cluster-B (*p < 0.001 by two tailed t test). Data are presented as the mean ± SEM; n = 4 independent experiments. HCT116 derived CM without CAF was used as a control. Source data and exact p values are provided as a Source Data file. CAF, cancer-associated fibroblast; TSS, transcription start site; SE, super-enhancer.CC BY · PubMed Central
Fig. 7 Prrx1 appeared as a master TF for the Cluster-A subpopulation of human colorectal CAF. A Identification of master transcription factor (mTF) candidates constructing core-regulatory circuitry (CRC). Initially, mTF candidates were screened based on a super-enhancer profile derived from H3K27ac ChIP-seq. Then, mTF candidates with high mRNA expression were selected by RNA-seq of these nine CAFs. B Each TF in CRC between the Cluster-A group and Cluster-B group. Arrowheads indicate “expressed CRC TF” (CRC mTF with high mRNA expression and specificity for each group). Prrx1 belonged to the expressed CRC TF of the Cluster-A group. C Super-enhancer peaks in PRRX1 gene loci were shown in the CAFs of Cluster-A. Contrastly, typical enhancer peaks were shown in similar PRRX1 gene loci in the CAFs of Cluster-B. D The network similarities of each of the CRCs of the nine CAFs were calculated using a mathematical model based on graph theory. In detail, the components, the link profile, and degree distribution (the internal structure of links) of each of the CRCs were compared, and their similarities were calculated. E To identify phenotypes regulated by Cluster-A-specific CRC, target genes using motif analysis were inferred. These target genes had binding motifs of three or more TFs constituting Cluster-A-specific CRC. Then, gene ontology analysis revealed that these target genes were enriched in myofibroblastic cellular functions. F PRRX1 showed more than four times binding to SE than TE by PRRX1 ChIP-seq in CAF4 cells of Cluster-A. G Enrichment analysis performed using the genes near the peak in PRRX1 ChIP-seq revealed that PRRX1 directly bound genes were significantly enriched in myofibroblastic phenotype-associated cellular functions. H Remarkable decrease of viable cell number in all Cluster-A CAFs in response to shRNA-mediated PRRX1 knockdown, with no effect in Cluster-B CAFs. Bar plot showing percentage of cell prolifeartion. Data are presented as the mean ± SEM; n = 12 independent experiments (two tailed t test: p *<0.05, **<0.01, ***<0.001). (I) Pseudo-time trajectory of 6,083 CAFs from the scRNA-seq of colorectal cancer tissue showed two differentiated fate (two tailed t test, *p < 2e−16, n = 6083). Source data and exact p values are provided as a Source Data file.
Fig. 7 Prrx1 appeared as a master TF for the Cluster-A subpopulation of human colorectal CAF. A Identification of master transcription factor (mTF) candidates constructing core-regulatory circuitry (CRC). Initially, mTF candidates were screened based on a super-enhancer profile derived from H3K27ac ChIP-seq. Then, mTF candidates with high mRNA expression were selected by RNA-seq of these nine CAFs. B Each TF in CRC between the Cluster-A group and Cluster-B group. Arrowheads indicate “expressed CRC TF” (CRC mTF with high mRNA expression and specificity for each group). Prrx1 belonged to the expressed CRC TF of the Cluster-A group. C Super-enhancer peaks in PRRX1 gene loci were shown in the CAFs of Cluster-A. Contrastly, typical enhancer peaks were shown in similar PRRX1 gene loci in the CAFs of Cluster-B. D The network similarities of each of the CRCs of the nine CAFs were calculated using a mathematical model based on graph theory. In detail, the components, the link profile, and degree distribution (the internal structure of links) of each of the CRCs were compared, and their similarities were calculated. E To identify phenotypes regulated by Cluster-A-specific CRC, target genes using motif analysis were inferred. These target genes had binding motifs of three or more TFs constituting Cluster-A-specific CRC. Then, gene ontology analysis revealed that these target genes were enriched in myofibroblastic cellular functions. F PRRX1 showed more than four times binding to SE than TE by PRRX1 ChIP-seq in CAF4 cells of Cluster-A. G Enrichment analysis performed using the genes near the peak in PRRX1 ChIP-seq revealed that PRRX1 directly bound genes were significantly enriched in myofibroblastic phenotype-associated cellular functions. H Remarkable decrease of viable cell number in all Cluster-A CAFs in response to shRNA-mediated PRRX1 knockdown, with no effect in Cluster-B CAFs. Bar plot showing percentage of cell prolifeartion. Data are presented as the mean ± SEM; n = 12 independent experiments (two tailed t test: p *<0.05, **<0.01, ***<0.001). (I) Pseudo-time trajectory of 6,083 CAFs from the scRNA-seq of colorectal cancer tissue showed two differentiated fate (two tailed t test, *p < 2e−16, n = 6083). Source data and exact p values are provided as a Source Data file.CC BY · PubMed Central
Fig. 8 Fibroblast-specific Prrx1 expression is critical in the expansion and maintenance of myofibroblast subpopulations during wound healing. A Illustration of the mouse model in which fibroblast-specific Prrx1 depletion. B Representative images of wound area from day 0 to day 18 post-operative wound in the fibroblast-specific Prrx1-knockout (sgPrrx1, n = 6) and control (sgNS, n = 9) mice, respectively. Delayed wound healing (%) and reduced wound contraction (%) due to Prrx1 deficiency on day 10 post-wound were shown by bar plot (bottom right). Data are presented as the mean ± SEM; n = 5 measurements (two tailed t test, *p < 0.0005, **p < 0.0001). C Decreased myofibroblastic subpopulations during the wound-healing process due to a deficiency of Prrx1. H&E and immunohistochemical stainings of wound area on day 10 post-wound demonstrated the reduced epithelial and dermal layers due to Prrx1 deficiency (bottom left). Box plots showed the number of total fibroblast and fibroblasts expressing α-SMA, PDGFR α, and PCNA between WT and Prrx1 KO mice (bottom right). (n = 3, two tailed t test: *p < 0.005, **p < 0.0005, ***p < 0.0001). D, E Induction of Dox-mediated fibroblast-specific Prrx1 expression in mice 5 days after skin punching. Representative images of wound area from day 0 to day 15 after wounding in fibroblast-specific Prrx1-induced (+Dox, n = 12) and control (−Dox, n = 11) mice were shown. Bar plots indicated an increased wound-healing rate (%) and wound contraction (%) due to Prrx1 overexpression on day 10 post-wound. Data are presented as the mean ± SEM; n = 5 measurements (two tailed t test: *p < 0.0001). F H&E and immunohistochemical images of representative wound area on day 30 post-wound were visualized using ImageJ Fiji (DAPI [hematoxylin]-blue, aSMA-green, TNC-red, PDGFRa-yellow, Prrx1-magenta, PCNA-cyan) (−Dox, n = 11 mice, +Dox, n = 12 mice). The thickness of both epithelial and dermal layers was increased in fibroblast-specific Prrx1 overexpressing mice (bottom left). Box plots showed the number of total fibroblast and fibroblasts expressing α-SMA, PDGFR-α, and PCNA in the wounded tissue (bottom right) (n = 3 measurements, two tailed t test: *p < 0.005, **p < 0.0005, ***p < 0.0001). Source data and exact p values are provided as a Source Data file.
Fig. 8 Fibroblast-specific Prrx1 expression is critical in the expansion and maintenance of myofibroblast subpopulations during wound healing. A Illustration of the mouse model in which fibroblast-specific Prrx1 depletion. B Representative images of wound area from day 0 to day 18 post-operative wound in the fibroblast-specific Prrx1-knockout (sgPrrx1, n = 6) and control (sgNS, n = 9) mice, respectively. Delayed wound healing (%) and reduced wound contraction (%) due to Prrx1 deficiency on day 10 post-wound were shown by bar plot (bottom right). Data are presented as the mean ± SEM; n = 5 measurements (two tailed t test, *p < 0.0005, **p < 0.0001). C Decreased myofibroblastic subpopulations during the wound-healing process due to a deficiency of Prrx1. H&E and immunohistochemical stainings of wound area on day 10 post-wound demonstrated the reduced epithelial and dermal layers due to Prrx1 deficiency (bottom left). Box plots showed the number of total fibroblast and fibroblasts expressing α-SMA, PDGFR α, and PCNA between WT and Prrx1 KO mice (bottom right). (n = 3, two tailed t test: *p < 0.005, **p < 0.0005, ***p < 0.0001). D, E Induction of Dox-mediated fibroblast-specific Prrx1 expression in mice 5 days after skin punching. Representative images of wound area from day 0 to day 15 after wounding in fibroblast-specific Prrx1-induced (+Dox, n = 12) and control (−Dox, n = 11) mice were shown. Bar plots indicated an increased wound-healing rate (%) and wound contraction (%) due to Prrx1 overexpression on day 10 post-wound. Data are presented as the mean ± SEM; n = 5 measurements (two tailed t test: *p < 0.0001). F H&E and immunohistochemical images of representative wound area on day 30 post-wound were visualized using ImageJ Fiji (DAPI [hematoxylin]-blue, aSMA-green, TNC-red, PDGFRa-yellow, Prrx1-magenta, PCNA-cyan) (−Dox, n = 11 mice, +Dox, n = 12 mice). The thickness of both epithelial and dermal layers was increased in fibroblast-specific Prrx1 overexpressing mice (bottom left). Box plots showed the number of total fibroblast and fibroblasts expressing α-SMA, PDGFR-α, and PCNA in the wounded tissue (bottom right) (n = 3 measurements, two tailed t test: *p < 0.005, **p < 0.0005, ***p < 0.0001). Source data and exact p values are provided as a Source Data file.CC BY · PubMed Central
Fig. 9 Induced depletion of Prrx1 in fibroblasts alone induced long-term complete remission of cisplatin-resistant cancers in immunocompetent mice. A LLC1-Luc-GFP cancer cells were transplanted subcutaneously into transgenic mice (FSP1;creCAS9EGFP). Four kinds of treatments were administered (n = 10 for each of four groups; sgNS, sgPrrx1, cisplatin, and sgPrrx1+cisplatin) after tumor formation (tumor size > 30 mm3). B Comparison of tumor response in each treatment. Tumor growth curves; n = 10 mice/group. C Tumor volume was monitored using MRI for 10 weeks. Representative images of tumor growth monitored using MRI. The yellow arrows indicate tumor (green). Representative H&E images of primary tumor site in each group. D Additional MRI follow-up was performed to confirm long-term complete remission of tumors in the two groups (sgPrrx1 and sgPrrx1+cisplatin). Source data are provided as a Source Data file.
Fig. 9 Induced depletion of Prrx1 in fibroblasts alone induced long-term complete remission of cisplatin-resistant cancers in immunocompetent mice. A LLC1-Luc-GFP cancer cells were transplanted subcutaneously into transgenic mice (FSP1;creCAS9EGFP). Four kinds of treatments were administered (n = 10 for each of four groups; sgNS, sgPrrx1, cisplatin, and sgPrrx1+cisplatin) after tumor formation (tumor size > 30 mm3). B Comparison of tumor response in each treatment. Tumor growth curves; n = 10 mice/group. C Tumor volume was monitored using MRI for 10 weeks. Representative images of tumor growth monitored using MRI. The yellow arrows indicate tumor (green). Representative H&E images of primary tumor site in each group. D Additional MRI follow-up was performed to confirm long-term complete remission of tumors in the two groups (sgPrrx1 and sgPrrx1+cisplatin). Source data are provided as a Source Data file.CC BY · PubMed Central
Fig. 10 Summary of the overall findings. PRRX1 is the master transcription factor (mTF) that determines the lineage of highly pro-tumorigenic cancer-associated fibroblasts (CAFs) with a myofibroblastic phenotype through remodeling a super-enhancer landscape. Targeting the mTF, PRRX1 belonging to the CAF subgroup with a myofibroblastic phenotype may be a promising cancer treatment strategy.
Fig. 10 Summary of the overall findings. PRRX1 is the master transcription factor (mTF) that determines the lineage of highly pro-tumorigenic cancer-associated fibroblasts (CAFs) with a myofibroblastic phenotype through remodeling a super-enhancer landscape. Targeting the mTF, PRRX1 belonging to the CAF subgroup with a myofibroblastic phenotype may be a promising cancer treatment strategy.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 66 references

Cited by

See all 112

Metadata

Authors
Lee, K.-W., Yeo, S.-Y., Gong, J.-R., Koo, O.-J., Sohn, I., Lee, W. Y., Kim, H. C., Yun, S. H., Cho, Y. B., Choi, M.-A., An, S., Kim, J., Sung, C. O., Cho, K.-H., & Kim, S.-H.
Affiliations
  • Sungkyunkwan University (KR)
  • Korea Advanced Institute of Science and Technology (KR)
  • ToolGen (South Korea) (KR)
  • Samsung Medical Center (KR)
  • Asan Medical Center (KR)
Corresponding author
  • Sung, Chang Ohk
  • Cho, Kwang-Hyun
  • Kim, Seok-Hyung
Venue
Nature Communications
Published
2022-05-19
Volume
13
Issue
1
Pages
2793-2793
ISSN
2041-1723 (ISSN-L)
OpenAlex
W4280652009
Cited by
103
References
66
Citations by year
2026: 17 · 2025: 36 · 2024: 27 · 2023: 20 · 2022: 3
FWCI
7.02
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Research Foundation of Korea (2018R1C1B6006428, 2019R1I1A1A01062460, 2020R1A2B5B03094920, 2016R1A5A2945889, 2019R1A2C2003900, 2020M3E5E2040308)
  • Electronics and Telecommunications Research Institute (22ZS1100)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › TGF-β signaling in diseases
  • Health Sciences › Medicine › Oncology › Cancer Cells and Metastasis
  • Health Sciences › Medicine › Oncology › Pancreatic and Hepatic Oncology Research
Keywords
PRRX1, stromal fibroblasts, cancer-associated fibroblasts, myofibroblastic phenotype, fibroblast heterogeneity, fibroblast plasticity, TGF-beta signaling, super-enhancers, transcription factors, fibroblast lineage, tumorigenicity, chemoresistance
MeSH terms
Cancer-Associated Fibroblasts, Animals, Fibroblasts, Neoplasms, Transcription Factors, Mice, Myofibroblasts

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: PRRX1 is a master transcription factor of stromal fibroblasts for myofibroblastic lineage progression. https://app.cochat.ai/papers/10.1038/s41467-022-30484-4
The paper
Lee, K.-W., Yeo, S.-Y., Gong, J.-R., Koo, O.-J., Sohn, I., Lee, W. Y., Kim, H. C., Yun, S. H., Cho, Y. B., Choi, M.-A., An, S., Kim, J., Sung, C. O., Cho, K.-H., & Kim, S.-H. (2022). PRRX1 is a master transcription factor of stromal fibroblasts for myofibroblastic lineage progression. Nature Communications. https://doi.org/10.1038/s41467-022-30484-4

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.