Advertisement

Insights into Fibroblast Plasticity

Cellular Communication Network 2 Is Required for Activation of Cancer-Associated Fibroblasts in a Murine Model of Melanoma
Open ArchivePublished:October 11, 2019DOI:https://doi.org/10.1016/j.ajpath.2019.09.006
      Tumor stroma resembles a fibrotic microenvironment, being characterized by the presence of myofibroblast-like cancer-associated fibroblasts (CAFs). In wild-type mice injected with melanoma cells, we show that the stem cell transcription factor Sox2 is expressed by tumor cells and induced in CAFs derived from synthetic fibroblasts. These fibroblasts were labeled postnatally with green fluorescent protein using mice expressing a tamoxifen-dependent Cre recombinase under the control of a fibroblast-specific promoter/enhancer. Conversely, fibroblast activation was impaired in mice with a fibroblast-specific deletion of cellular communication network 2 (Ccn2), associated with reduced expression of α-smooth muscle actin and Sox2. Multipotent Sox2-expressing skin-derived precursor (SKP) spheroids were cultured from murine back skin. Using lineage tracing and flow cytometry, approximately 40% of SKPs were found to be derived from type I collagen-lineage cells and acquired multipotency in culture. Inhibition of mechanotransduction pathways prevented myofibroblast differentiation of SKPs and expression of Ccn2. In SKPs deleted for Ccn2, differentiation into a myofibroblast, but not an adipocyte or neuronal phenotype, was also impaired. In human melanoma, CCN2 expression was associated with a profibrotic integrin alpha (ITGA) 11–expressing subset of CAFs that negatively associated with survival. These results suggest that synthetic dermal fibroblasts are plastic, and that CCN2 is required for the differentiation of dermal progenitor cells into a myofibroblast/CAF phenotype and is, therefore, a therapeutic target in melanoma.
      Melanoma is highly metastatic and the most fatal skin cancer.
      • Uong A.
      • Zon L.I.
      Melanocytes in development and cancer.
      Although drugs, including those targeting specific mutations (eg, in BRAF) and checkpoint inhibitors, have been discovered that can retard melanoma progression in some patients, ultimately patients develop resistance to these regimens.
      • Kakadia S.
      • Yarlagadda N.
      • Awad R.
      • Kundranda M.
      • Niu J.
      • Naraev B.
      • Mina L.
      • Dragovich T.
      • Gimbel M.
      • Mahmoud F.
      Mechanisms of resistance to BRAF and MEK inhibitors and clinical update of US Food and Drug Administration-approved targeted therapy in advanced melanoma.
      ,
      • Silva I.P.
      • Long G.V.
      Systemic therapy in advanced melanoma: integrating targeted therapy and immunotherapy into clinical practice.
      Although much work is currently being expended on how tumors evade immunosurveillance, alternative strategies to block metastasis are needed, yet such novel approaches are relatively underresearched.
      Fibroblasts, the cells responsible for generating and maintaining extracellular matrix (ECM), are fundamental contributors to diverse physiological and pathologic processes, including cancer. Within the tumor stroma, fibroblasts, intimately associated with cancer cells at all stages of cancer progression, actively contribute to both tumor growth and metastasis.
      • Pickup M.W.
      • Mouw J.K.
      • Weaver V.M.
      The extracellular matrix modulates the hallmarks of cancer.
      More specifically, in melanoma, the stiffness of the surrounding ECM has been linked to tumor aggressiveness and the acquisition of drug resistance.
      • Pickup M.W.
      • Mouw J.K.
      • Weaver V.M.
      The extracellular matrix modulates the hallmarks of cancer.
      • Werb Z.
      • Lu P.
      The role of stroma in tumor development.
      • Zhou L.
      • Yang K.
      • Andl T.
      • Wickett R.R.
      • Zhang Y.
      Perspective of targeting cancer-associated fibroblasts in melanoma.
      This phenomenon can be conceptually described as follows. When located within the tumor, cancer cells attach to each other via cell/cell contacts. In this environment, mutations (eg, those activating BRAF) that activate prosurvival and proliferative signals (eg, extracellular signal-regulated kinase and Akt) can be directly targeted by inhibitors targeting these mutations (eg, BRAF inhibitors). Conversely, when cells leave this niche and migrate on a stiff ECM, cancer cells now attach to the surrounding ECM via focal adhesion kinase/integrin β 1; this engagement now results in the activation of downstream prosurvival signals (eg, extracellular signal-regulated kinase and Akt) in an adhesion/stiffness-dependent yet BRAF-independent manner, resulting in resistance to anti-BRAF inhibitor chemotherapy.
      • Taylor K.N.
      • Schlaepfer D.D.
      Adaptive resistance to chemotherapy, a multi-FAK-torial linkage.
      ,
      • Hirata E.
      • Girotti M.R.
      • Viros A.
      • Hooper S.
      • Spencer-Dene B.
      • Matsuda M.
      • Larkin J.
      • Marais R.
      • Sahai E.
      Intravital imaging reveals how BRAF inhibition generates drug-tolerant microenvironments with high integrin beta1/FAK signaling.
      Thus, a stiff ECM may be sufficient for the activation and metastasis of melanoma cells. This stiff ECM is generated by so-called cancer-associated fibroblasts (CAFs) that are similar to the myofibroblasts seen in fibrotic conditions, as these CAFs express the highly contractile protein α-smooth muscle actin (α-SMA).
      • Polanska U.M.
      • Orimo A.
      Carcinoma-associated fibroblasts: non-neoplastic tumour-promoting mesenchymal cells.
      Proteins secreted by CAFs into the tumor microenvironment are essential mediators of cellular signaling responses that contribute to cancer cell proliferation and metastasis.
      • Gandellini P.
      • Andriani F.
      • Merlino G.
      • D'Aiuto F.
      • Roz L.
      • Callari M.
      Complexity in the tumour microenvironment: cancer associated fibroblast gene expression patterns identify both common and unique features of tumour-stroma crosstalk across cancer types.
      Therefore, understanding how CAFs originate and promote melanoma progression is of major importance.
      Matricellular proteins, nonstructural ECM components that are secreted by fibroblasts into the microenvironment during conditions of pathologic tissue remodeling, are emerging therapeutic targets in fibrotic disease and cancers, including in melanoma.
      • Vincent K.M.
      • Postovit L.M.
      Matricellular proteins in cancer: a focus on secreted Frizzled-related proteins.
      ,
      • Tsang M.
      Mesenchymal cells emerge as primary contributors to fibrosis in multiple tissues.
      Specifically, the central communication network (CCN) family of matricellular proteins is of current therapeutic interest.
      • Riser B.L.
      • Barnes J.L.
      • Varani J.
      Balanced regulation of the CCN family of matricellular proteins: a novel approach to the prevention and treatment of fibrosis and cancer.
      Cellular communication network 2 (CCN2), a CCN family member formerly referred to as connective tissue growth factor,
      • Perbal B.
      • Tweedie S.
      • Bruford E.
      The official unified nomenclature adopted by the HGNC calls for the use of the acronyms, CCN1-6, and discontinuation in the use of CYR61, CTGF, NOV and WISP 1-3 respectively.
      is not normally expressed by dermal fibroblasts, but is potently induced in normal tissue repair and pathologies, including fibrosis and cancers.
      • Ramazani Y.
      • Knops N.
      • Elmonem M.A.
      • Nguyen T.Q.
      • Arcolino F.O.
      • van den Heuvel L.
      • Levtchenko E.
      • Kuypers D.
      • Goldschmeding R.
      Connective tissue growth factor (CTGF) from basics to clinics.
      In human melanoma patients, CCN2 expression does not correlate with BRAF mutational status, but instead correlates with expression of stroma and angiogenic gene subsets, and negatively correlates with survival.
      • Hutchenreuther J.
      • Vincent K.M.
      • Carter D.E.
      • Postovit L.M.
      • Leask A.
      CCN2 expression by tumor stroma is required for melanoma metastasis.
      ,
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      In mice, conditional knockout strategies, using a tamoxifen-dependent Cre recombinase expressed under the control of fibroblast-specific Col1a2 promoter/enhancer,
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      have shown that loss of CCN2 expression by fibroblasts results in resistance to melanoma metastasis and tumor-induced neovascularization, concomitant with a reduced production of the proangiogenic mediators periostin and vascular endothelial growth factor.
      • Hutchenreuther J.
      • Vincent K.M.
      • Carter D.E.
      • Postovit L.M.
      • Leask A.
      CCN2 expression by tumor stroma is required for melanoma metastasis.
      ,
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      Accordingly, CCN2 represents a novel therapeutic target for BRAF inhibitor–resistant melanoma.
      Although much work has been done to investigate the plasticity of tumor cells and their acquisition of stem-like characteristics in cancer progression, relatively little is known about the specific mechanisms through which resident fibroblasts become recruited and activated within the tumor stroma to become CAFs. In a model of bleomycin-induced skin fibrosis, it has previously been shown that dermal cells expressing the progenitor cell marker Sox2 represent an important source of α-SMA–expressing myofibroblasts and differentiate in a CCN2-dependent manner to result in pathologic skin thickening.
      • Liu S.
      • Herault Y.
      • Pavlovic G.
      • Leask A.
      Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
      Interestingly, in human melanoma, expression of Sox2 correlates with tumor thickness and is markedly induced in spindle-like cells occurring at the tumor/stroma interface (ie, in cells resembling fibroblasts),
      • Laga A.C.
      • Lai C.Y.
      • Zhan Q.
      • Huang S.J.
      • Velazquez E.F.
      • Yang Q.
      • Hsu M.Y.
      • Murphy G.F.
      Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
      suggesting the intriguing possibility that Sox2 may be expressed in CAFs and, as in during skin fibrosis, may be involved with the acquisition of a myofibroblast phenotype.
      In this report, conditional knockout and lineage tracing strategies using a tamoxifen-dependent Cre recombinase expressed under the control of a fibroblast-specific promoter/enhancer derived from the Col1a2 gene
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      were used to investigate how dermal fibroblasts contribute, in a CCN2-dependent manner, to the appearance of progenitor-like cells and CAFs in tumor stroma. Skin-derived precursor (SKP) spheroids were further used as an in vitro model of fibroblast plasticity to demonstrate the mechanistic role of CCN2 in myofibroblast differentiation of progenitor cells. These data provide new and valuable insights into the potential role of fibroblast plasticity in melanoma progression.

      Materials and Methods

      Generation of Transgenic Mice

      For lineage tracing, mice hemizygous for a tamoxifen-dependent Cre recombinase under the control of a Col1a2
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      or Sox2 (Institut Clinique de la Souris, Alsace, France) promoter [Col1a2-Cre(ER)-T; Sox2-Cre(ER)-T] were bred with mice homozygous for a double-fluorescent reporter transgene (mTmG) integrated into the Gt(ROSA) 26Sor locus (Jackson Laboratories, Bar Harbor, ME), thereby generating Col1a2-Cre(ER)T; Rosa26mTmG or Sox2-Cre(ER)T; Rosa26mTmG mice. Please note Col1a2-Cre(ER)-T mice contain a fibroblast-specific far upstream enhancer, initially identified in the laboratory of Benoit de Crombrugghe, subcloned upstream of the Col1a2 minimal promoter; previous publications have extensively shown that this construct permits transgene expression only in fibroblasts and not in other collagen-expressing cells, such as bone and cartilage.
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      ,
      • Bou-Gharios G.
      • Garrett L.A.
      • Rossert J.
      • Niederreither K.
      • Eberspaecher H.
      • Smith C.
      • Black C.
      • Crombrugghe B.
      A potent far-upstream enhancer in the mouse pro alpha 2(I) collagen gene regulates expression of reporter genes in transgenic mice.
      • Tanaka S.
      • Antoniv T.T.
      • Liu K.
      • Wang L.
      • Wells D.J.
      • Ramirez F.
      • Bou-Gharios G.
      Cooperativity between far upstream enhancer and proximal promoter elements of the human {alpha}2(I) collagen (COL1A2) gene instructs tissue specificity in transgenic mice.
      • Ponticos M.
      • Abraham D.
      • Alexakis C.
      • Lu Q.L.
      • Black C.
      • Partridge T.
      • Bou-Gharios G.
      Col1a2 enhancer regulates collagen activity during development and in adult tissue repair.
      In these mice, when Cre recombinase is induced via tamoxifen, the mTmG transgene results in permanent green fluorescent protein (GFP) expression in cells expressing the promoter construct (and all of their progeny), whereas all other cells express tdTomato. To generate Ccn2 conditional knockout mice, Col1a2-Cre(ER)T; Rosa26mTmG mice were bred for two generations with mice homozygous for a floxed Ccn2 allele
      • Hutchenreuther J.
      • Vincent K.M.
      • Carter D.E.
      • Postovit L.M.
      • Leask A.
      CCN2 expression by tumor stroma is required for melanoma metastasis.
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      • Liu S.
      • Herault Y.
      • Pavlovic G.
      • Leask A.
      Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
      to produce Col1a2-Cre(ER)T; Rosa26mTmG; CCN2fl/fl mice. In these mice, induction of Cre via tamoxifen results in a Col1a2-specific Ccn2 knockout (Ccn2−/−) in addition to activating the mTmG reporter described above. To induce Cre recombinase, a stock solution of 4-hydroxy-tamoxifen in ethanol (Sigma, St. Louis, MO; 100 mg/mL) was diluted in corn oil (Sigma) to 10 mg/mL. Three-week–old littermate mice were given 0.1 mL i.p. injections of the tamoxifen solution (for lineage tracing and Ccn2−/−) or corn oil (for Ccn2fl/fl) for 5 consecutive days. Two weeks after tamoxifen injection, loss of Ccn2 was confirmed by PCR genotyping, and mice were sacrificed for experimental use or injected with melanoma cells.

      Melanoma Mouse Model

      B16F10 murine melanoma cells (ATCC, Manassas, VA) were injected subcutaneously (330,000 cells) into the right flank of transgenic mice described above. Tumors grew for 14 days after the appearance of a palpable tumor, and skin samples containing tumor and tumor stroma were collected. Samples were put in OCT compound (Sakura Finetek, Torrance, CA), frozen at −80°C, and stained as described below.

      Cell Extraction and Culture

      Full-thickness dorsal skin was dissected from sacrificed mice, washed in phosphate-buffered saline (PBS), and then placed into high-glucose Dulbecco's modified Eagle's medium (DMEM; Invitrogen, Burlington, ON, Canada) containing 2 mg/mL collagenase type II and incubated for 3 hours at 37°C. Dermal tissue was scraped off into DMEM and mechanically dissociated by pipetting. Samples were centrifuged for 1 minute at 55 × g, and supernatant containing cells was collected and spun down again for 5 minutes at 250 × g. Cell pellets were collected and resuspended in media, filtered with a 40-μm cell strainer, and seeded in a T75 cell culture flask. Fibroblasts were cultured in high-glucose DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS; Invitrogen) and 1% antibiotic-antimycotic (Invitrogen). SKP spheroids were cultured in SKP base media: three parts DMEM glutaMAX and one part Hams F12 (Invitrogen); supplemented with 1% antibiotic-antimycotic, 2% B27 (Invitrogen; A35828-01), 1% N2 (Invitrogen; 17502048), epidermal growth factor (20 ng/mL; Invitrogen; PHG0311), and basic fibroblast growth factor (20 ng/mL; Invitrogen; PHG0026).
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      ,
      • Liu S.
      • Herault Y.
      • Pavlovic G.
      • Leask A.
      Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
      Alternatively, to generate larger quantities of SKPs, similar to a previously described report,
      • Laga A.C.
      • Lai C.Y.
      • Zhan Q.
      • Huang S.J.
      • Velazquez E.F.
      • Yang Q.
      • Hsu M.Y.
      • Murphy G.F.
      Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
      confluent 10-cm plates of fibroblasts were removed with 0.25% trypsin (Invitrogen) and resuspended in SKP base media, as described above. Cells were then seeded into 3 wells of a 6-well plate per 10-cm plate of fibroblasts. Basic fibroblast growth factor (20 ng/mL) supplement was added Monday, Wednesday, and Friday for 2 weeks. Melanoma B16(F10) (ATCC) cells were cultured in high-glucose DMEM supplemented with 10% FBS and 1% antibiotic-antimycotic.

      SKP Differentiation

      SKP differentiation was conducted exactly as previously described.
      • Fernandes K.J.
      • Miller F.D.
      Isolation, expansion, and differentiation of mouse skin-derived precursors.
      Briefly, plates were coated with 0.001% poly-l-lysine (Sigma) and laminin (2.5 mg/mL; Sigma) in PBS (Invitrogen) overnight. SKPs were collected and plated on coated plates in various media.

      SKP to Myofibroblasts

      SKPs were transferred to DMEM supplemented with 0.5% FBS for 24 hours, at which point images were taken and samples were collected for further analysis. For inhibitor experiments, inhibitors were added 30 minutes before FBS exposure: PP2 (Sigma; 10 μm), PF228 (Tocris, Oakville, ON, Canada; 10 μm), and CCG1423 (Tocris; 50 μm).

      SKP to Neurons

      SKPs were differentiated in medium containing 5% FBS for 5 days, then medium was replaced and supplemented with NGF (40 ng/mL; Invitrogen; 13257-019) for 5 additional days.

      SKP to Adipocytes

      SKPs were differentiated in medium containing 5% FBS for 24 hours, then medium was replaced with adipocyte differentiation medium (A1007001; ThermoFisher, St. Laurent, QC, Canada) every approximately 5 days and stained with oil red O after a total of 14 days (Sigma).

      Histology

      Cryosections of mouse dorsal skin were acquired by freezing tissue in OCT compound at −80°C and dividing into sections (20 μm thick) at −22°C using a Leica CM1900 UV microtome cryostat (Leica, Concord, ON, Canada). To detect endogenous GFP and tdTomato expression in mTmG reporter mice, slides were washed in PBS, mounted with mounting media containing DAPI (Vector Laboratories, Burlington, ON, Canada), and imaged using a Zeiss Imager M1 fluorescence microscope and Northern Eclipse software version 8 (Empix, Mississauga, ON, Canada).

      Immunofluorescence

      Cultured cells or skin tissue cryosections were fixed with 4% paraformaldehyde for 10 minutes, followed by permeabilization with 0.5% PBS/0.05% Tween 20 (Sigma) for 10 minutes. Samples were incubated in blocking solution (0.5% PBS/0.05% Tween 20 with 10% donkey serum; Jackson Immunoresearch, West Grove, PA) for 30 minutes, followed by incubating with primary antibody in the blocking solution overnight. The following primary antibodies were used: anti–α-SMA (1:500; Sigma; A5228), anti-CCN2 (1:100; Santa Cruz Biotechnology, Dallas, TX; sc-14939) anti–βIII-tubulin (1:1000; PhosphoSolutions, Aurora, CO), anti-GFP (1:100; Santa Cruz Biotechnology; sc-9996), antinestin (1:100; Santa Cruz Biotechnology; sc-21248), and anti-Sox2 (1:100; Santa Cruz Biotechnology; sc-17320). Appropriate IgG controls were from Jackson Immunoresearch. Appropriate fluorophore-conjugated secondary antibodies (raised in donkey; 1:5000; Jackson Immunoresearch) diluted in blocking solution were added (1 hour) and washed, and cells were mounted using DAPI-containing mounting media (Vector Laboratories). Imaging was performed using a Leica Microsystems DMI6000B inverted microscope. Images were analyzed using ImageJ software version 1.49 (NIH, Bethesda, MD; https://imagej.nih.gov/ij), and cells within the tumor stroma staining positively for GFP were manually counted. Of those cells, those that were coexpressing α-SMA, SOX2, both α-SMA and SOX2, or SOX2 without α-SMA were manually counted.

      RNA Analysis

      RNA was extracted from cell cultures using a TRIzol-Chloroform (Invitrogen) method and analyzed as described previously.
      • Bou-Gharios G.
      • Garrett L.A.
      • Rossert J.
      • Niederreither K.
      • Eberspaecher H.
      • Smith C.
      • Black C.
      • Crombrugghe B.
      A potent far-upstream enhancer in the mouse pro alpha 2(I) collagen gene regulates expression of reporter genes in transgenic mice.
      ,
      • Tanaka S.
      • Antoniv T.T.
      • Liu K.
      • Wang L.
      • Wells D.J.
      • Ramirez F.
      • Bou-Gharios G.
      Cooperativity between far upstream enhancer and proximal promoter elements of the human {alpha}2(I) collagen (COL1A2) gene instructs tissue specificity in transgenic mice.
      Microarray analysis was performed using Affymetrix GeneChip Mouse Gene 2.0 ST arrays (Affymetrix, Santa Clara, CA), as described previously
      • Ramazani Y.
      • Knops N.
      • Elmonem M.A.
      • Nguyen T.Q.
      • Arcolino F.O.
      • van den Heuvel L.
      • Levtchenko E.
      • Kuypers D.
      • Goldschmeding R.
      Connective tissue growth factor (CTGF) from basics to clinics.
      ,
      • Bou-Gharios G.
      • Garrett L.A.
      • Rossert J.
      • Niederreither K.
      • Eberspaecher H.
      • Smith C.
      • Black C.
      • Crombrugghe B.
      A potent far-upstream enhancer in the mouse pro alpha 2(I) collagen gene regulates expression of reporter genes in transgenic mice.
      ,
      • Tanaka S.
      • Antoniv T.T.
      • Liu K.
      • Wang L.
      • Wells D.J.
      • Ramirez F.
      • Bou-Gharios G.
      Cooperativity between far upstream enhancer and proximal promoter elements of the human {alpha}2(I) collagen (COL1A2) gene instructs tissue specificity in transgenic mice.
      ,
      • Shi-wen X.
      • Parapuram S.K.
      • Pala D.
      • Chen Y.
      • Carter D.E.
      • Eastwood M.
      • Denton C.P.
      • Abraham D.J.
      • Leask A.
      Requirement of transforming growth factor beta-activated kinase 1 for transforming growth factor beta-induced alpha-smooth muscle actin expression and extracellular matrix contraction in fibroblasts.
      • Guo F.
      • Carter D.E.
      • Leask A.
      Mechanical tension increases CCN2/CTGF expression and proliferation in gingival fibroblasts via a TGFbeta-dependent mechanism.
      • Chen S.
      • McLean S.
      • Carter D.E.
      • Leask A.
      The gene expression profile induced by Wnt 3a in NIH 3T3 fibroblasts.
      • Liu S.
      • Xu S.W.
      • Kennedy L.
      • Pala D.
      • Chen Y.
      • Eastwood M.
      • Carter D.E.
      • Black C.M.
      • Abraham D.J.
      • Leask A.
      FAK is required for TGFbeta-induced JNK phosphorylation in fibroblasts: implications for acquisition of a matrix-remodeling phenotype.
      • Kuk H.
      • Hutchenreuther J.
      • Murphy-Marshman H.
      • Carter D.
      • Leask A.
      5Z-7-oxozeanol inhibits the effects of TGFbeta1 on human gingival fibroblasts.
      by the London Regional Genomics Center; cluster analysis was performed using DAVID Functional Annotation Software version 6.7 (https://david.ncifcrf.gov).
      • Huang D.W.
      • Sherman B.T.
      • Lempicki R.A.
      Systematic and integrative analysis of large gene lists using DAVID Bioinformatics Resources.
      Experiments were performed twice; the complete data set after average values was calculated, is shown in Supplemental Table S1, and as such represents an initial approach to identifying genes whose expression is potentially CCN2 dependent (fold change > |1.4|; N = 2). Data were analyzed as previously described.
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      ,
      • Shi-wen X.
      • Parapuram S.K.
      • Pala D.
      • Chen Y.
      • Carter D.E.
      • Eastwood M.
      • Denton C.P.
      • Abraham D.J.
      • Leask A.
      Requirement of transforming growth factor beta-activated kinase 1 for transforming growth factor beta-induced alpha-smooth muscle actin expression and extracellular matrix contraction in fibroblasts.
      • Guo F.
      • Carter D.E.
      • Leask A.
      Mechanical tension increases CCN2/CTGF expression and proliferation in gingival fibroblasts via a TGFbeta-dependent mechanism.
      • Chen S.
      • McLean S.
      • Carter D.E.
      • Leask A.
      The gene expression profile induced by Wnt 3a in NIH 3T3 fibroblasts.
      • Liu S.
      • Xu S.W.
      • Kennedy L.
      • Pala D.
      • Chen Y.
      • Eastwood M.
      • Carter D.E.
      • Black C.M.
      • Abraham D.J.
      • Leask A.
      FAK is required for TGFbeta-induced JNK phosphorylation in fibroblasts: implications for acquisition of a matrix-remodeling phenotype.
      • Kuk H.
      • Hutchenreuther J.
      • Murphy-Marshman H.
      • Carter D.
      • Leask A.
      5Z-7-oxozeanol inhibits the effects of TGFbeta1 on human gingival fibroblasts.
      Real-time PCR using 40-ng RNA samples was run in triplicate on an ABI Prism 7900 HT PCR machine (ThermoFisher). 18S was used as internal control, and fold change was calculated using the ΔΔCt method compared with one RNA sample derived from wild-type cells treated with serum.

      Western Blot Analysis

      Protein was harvested by cell lysis by incubating cultured cells in radioimmunoprecipitation assay buffer supplemented with complete protease inhibitor cocktail (Sigma). Protein samples (50 μg) were run on 10% SDS-acrylamide gels. Protein was transferred to a nitrocellulose membrane and blocked in 5% milk in 0.01% Tris-buffered saline/Tween 20 for 1 hour. Primary antibodies were incubated in blocking milk overnight at 4°C: anti–α-SMA (1:2500), anti-CCN2 (1:100), anti–βIII-tubulin (1:10,000), or anti–β-actin (1:5000; Sigma). Horseradish peroxidase–conjugated secondary antibodies were incubated for 1 hour at room temperature after washes in 0.01% Tris-buffered saline/Tween 20, and membranes were treated with SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific, St. Laurent, QC, Canada) and visualized using X-ray film. Densitometry quantification was performed using ImageJ software.

      Flow Cytometry

      Cells were brought to single-cell suspension by incubation with 1 U/μL collagenase type IV (Invitrogen; 17104-019) for 1 hour at 37°C in Hanks’ balanced salt solution, followed by mechanical dissociation. Cells were then filtered through a 40-μm cell strainer and incubated in Zombie Aqua dye (423101; Biolegend, San Diego, CA) and PBS for 20 minutes in the dark. Cells were fixed and permeabilized using nuclear fixation kit (Biolegend), followed by staining with 647-fluorophore–conjugated Sox2 antibody (1:50) in cell staining buffer (Biolegend). Cells were washed three times in cell staining buffer, and 50,000 cells were run through the FACSCanto flow cytometer (BD Biosciences, San Jose, CA) to determine the number of tdTomato-, GFP-, and Sox2-positive cells using FACSDiva software version 8.01 (BD Biosciences). Analysis was performed on FlowJo software version 10.6.1 (Tree Star Inc., Ashland, OR).

      Data Sets and Survival Analysis

      Single-cell RNA-sequencing data, derived from 6879 melanoma tumor cells, were obtained from Gene Expression Omnibus GSE115978 in July 2019 (https://www.ncbi.nlm.nih.gov/geo; accession number GSE115978). Please note that the data set analyzed is distinct from that analyzed previously.
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      For single-cell analysis, transcript-per-million (TPM) values were divided by 10, given that the complexity of single-cell libraries is estimated to be in the order of 100,000 transcripts. Level 3 The Cancer Genome Atlas RNAseqV2 gene expression was obtained from The Cancer Genome Atlas Data Portal in January 2016. RNA-Seq by Expectation Maximization (RSEM) expression data were used in downstream analyses.
      • Li B.
      • Dewey C.N.
      RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome.
      Patient survival information was downloaded in December 2017. Z-score unsupervised hierarchical clustering was conducted using 1-c (where c is the Pearson correlation coefficient) as the distance and the Ward agglomeration method (ward.D2).
      CAF-specific CCN2 scores were calculated for tumors by summing Z-scores for genes that were i) strongly correlated with CCN2 (Pearson r > 0.5), ii) had 10-fold higher mean expression in CAFs than any other cell type, and iii) had mean expression <0.1 TPM/10 in malignant cells. This resulted in a nine-gene score composed of the following genes: ZFHX4, DCN, ITGA11, COL6A3, COL1A1, ITGBL1, COL8A1, INHBA, and, MEG3. CAF-specific CCN2 scores were dichotomized with receiver operating characteristic curves to determine the optimal cutoff for the end point of overall survival censorship. Survival curves were constructed using the Kaplan-Meier method on primary tumor samples that had survival information available from The Cancer Genome Atlas. Significance was determined by log-rank test.
      All analyses and visualizations were conducted in the RStudio programming environment version 1.1.463 (RStudio, Inc., Boston, MA). R/Bioconductor packages ggplot2, plyr, pROC, gplots, and survival were used, where appropriate.

      Statistical Analysis

      Statistical analysis using either unpaired t-test or one-way analysis of variance with Tukey multiple comparison post-hoc testing, as appropriate, was performed using GraphPad Prism version 7 (GraphPad Software, La Jolla, CA).

      Results

      In a Murine Melanoma Model, Dermal Fibroblasts Acquire Expression of Sox2 and α-SMA in a CCN2-Dependent Manner

      To address the potential role of Sox2-expressing progenitor cells in the Ccn2-mediated activation of CAFs, a syngeneic mouse model was used in which Col1a2-CreER(T); Rosa26mTmG C57/BL6 mice deleted (Ccn2−/− or knockout) or not (Ccn2fl/fl or wild type) for Ccn2 in Col1a2-expressing fibroblasts were injected with B16F10 melanoma cells 2 weeks after induction of Cre recombinase. In this model, host cells are permanently labeled with either GFP or tdTomato, whereas tumor-derived cells express neither GFP nor tdTomato. To test the hypothesis that the induction of α-SMA–expressing CAFs might be occurring via a Sox2-expressing intermediate in the host tumor stroma, it was examined whether, in principle, there existed any cells in our tumor model that labeled with GFP that were also expressing both Sox2 and α-SMA. Tumors were allowed to grow for 14 days after detection of a palpable tumor. Mice were sacrificed, and tissue sections were stained for GFP (to detect cells derived from Col1a2-expressing fibroblasts at the time of tamoxifen injection and before the s.c. injection of tumor cells), Sox2 (to detect progenitor-like cells), and α-SMA (to detect CAFs; ie, myofibroblasts) (Figure 1, A and B ).
      Figure thumbnail gr1
      Figure 1Col1a2-specific Ccn2 knockout decreases expression of Sox2 and α-smooth muscle actin (α-SMA) in Col1a2-derived melanoma tumor stroma cells. Col1a2-CreER(T); Rosa26mTmG; Ccn2fl/fl and Ccn2−/− mice were implanted with B16F10 melanoma cells. In this model, host-derived cells express green fluorescent protein (GFP) or tdTomato, whereas tumor cells express neither GFP nor tdTomato. GFP+ refers to cells expressing GFP and are, therefore, host cells that were expressing the fibroblast-specific Col1a2 promoter/enhancer at the time of tamoxifen injection, before the injection of tumor cells. A and B: Ccn2fl/fl (A) and Ccn2−/− (B) tissue sections were stained for GFP (green), α-SMA (red), and Sox2 (gray) with representative low-power (left panel) and high-power (right panel) magnification images shown. C: Col1a2-derived GFP+ cells, expressed as percentage of total tumor stroma cells in Ccn2fl/fl and Ccn2−/− sections. D and E: Percentage of GFP+ cells coexpressing either α-SMA or Sox2 in Ccn2fl/fl and Ccn2−/− sections. F and G: Percentage of GFP+ cells either coexpressing both α-SMA and Sox2 (arrow in A) or expressing Sox2 but negative for α-SMA, in Ccn2fl/fl and Ccn2−/− sections. Data are expressed as means ± SEM (CG). N = 3 mice with tumors (AG). *P < 0.05, ***P < 0.001 versus Ccn2fl/fl (t-test). Scale bars: 100 μm (A and B, left two columns); 25 μm (A and B, right two columns).
      GFP was readily detected in the tumor stroma of both wild-type and knockout mice, indicating that these cells were host derived and contained an active Col1a2 promoter at the time of tamoxifen injection (Figure 1C). In wild-type mice, most GFP+ cells also expressed either Sox2 (Figure 1D) or α-SMA (Figure 1E). Notably, a subset of GFP+ cells coexpressed both Sox2 and α-SMA (Figure 1F). In contrast, in the tumor stroma of Ccn2-deficient mice, cells coexpressing GFP and α-SMA were undetected, resulting in an absence of cells triple positive for GFP, Sox2, and α-SMA (Supplemental Figure S1). Ccn2-deficient mice also displayed a marked reduction in the overall proportion of Sox2-expressing GFP+ cells (P < 0.05). When measuring the proportion of GFP+ cells that were α-SMA and Sox2+, there was no significant difference between wild-type and knockout mice (P = 0.97) (Figure 1G). These data suggested that, in a murine model of melanoma, host stromal cells derived from Col1a2-expressing cells (ie, synthetic fibroblasts) are plastic and can express the progenitor cell marker Sox2 in vivo. Moreover, expression of Ccn2 appeared necessary for the activation and differentiation of these Sox2-expressing cells into α-SMA–expressing CAFs (myofibroblasts) in the stroma.

      A Subset of Multipotent Dermal Skin Progenitor Cells (SKPs) in Vitro Is Derived from Col1a2-Expressing Fibroblasts

      On the basis of these observations, it was tested if GFP-expressing (ie, synthetic) dermal fibroblasts could acquire progenitor cell–like features in vitro, and if their ability to differentiate into myofibroblasts was CCN2 dependent. To test this hypothesis, mouse dermal cells were cultured in stem cell media containing basic fibroblast growth factor, which results in the formation of free-floating spheroid aggregates that express Sox2 and nestin (Figure 2A). These spheroids, which can be readily isolated directly from the dermis or from established fibroblast monolayer cultures, are well described in the literature, possess multipotent differentiation potential, and are often referred to as SKPs.
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      • Liu S.
      • Herault Y.
      • Pavlovic G.
      • Leask A.
      Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
      • Laga A.C.
      • Lai C.Y.
      • Zhan Q.
      • Huang S.J.
      • Velazquez E.F.
      • Yang Q.
      • Hsu M.Y.
      • Murphy G.F.
      Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
      Figure thumbnail gr2
      Figure 2Col1a2-expressing dermal fibroblasts express Sox2 when cultured as skin-derived precursor (SKP) spheroids. A: Expression of neural crest stem cell markers Sox2 and nestin by SKP spheroids; representative images shown. Lineage tracing was performed using Col1a2-CreER(T); Rosa26mTmG and Sox2-CreER(T); Rosa26mTmG reporter mice that were injected with tamoxifen 3 weeks after birth. AG: Expression of green fluorescent protein (GFP) and tdTomato was assessed in skin histologic sections via fluorescence microscopy (A, B, and E) and in cultured SKPs via fluorescence microscopy (C and F) and flow cytometry (D and G). Representative images are shown. H: Percentage of GFP+ cells in SKPs cultured from Sox2 and Col1a2 lineage reporter mice via quantification of flow cytometry. I: Flow cytometry performed on Col1a2-GFP SKPs detecting expression of Sox2 with (left panel) representative flow cytometry quadrants (right panel) graph representing percentage of Sox2+ cells in the GFP+ and GFP populations. Data are expressed as means ± SEM (H and I). N = 4 mice (H and I). **P < 0.01 versus Col1a2 (t-test). Scale bars: 100 μm (A, B, and E); 300 μm (C and F). Q, quartile.
      To test if cells derived from synthetic fibroblasts could display progenitor cell–like features, it was tested if cells labeled postnatally with GFP using a fibroblast-specific Col1a2-derived promoter/enhancer (ie, cells derived from synthetic fibroblasts) could be incorporated into SKP spheroids and differentiate into other cell types in vitro. Alternatively, it was tested if SKPs were derived from dermal cells that had previously expressed Sox2 in vivo. Accordingly, SKP spheroids were cultured from the dermis of Col1a2-CreER(T); Rosa26mTmG or Sox2-CreER(T); Rosa26mTmG promoter mice, where Col1a2- or Sox2-expressing cells were permanently labeled with GFP 3 weeks after birth, respectively. All other cells were labeled with tdTomato. Visualization of promoter activity in skin demonstrated Col1a2 lineage labeling in most dermal fibroblasts (Figure 2B); when fibroblasts were cultured in vitro from dermal and s.c. tissue, approximately 25% of monolayer cells expressed GFP (Supplemental Figure S2). Conversely, Sox2 lineage labeling was largely seen in the hair follicle dermal papilla and dermal sheath, and in the rare cell at the dermal-epidermal junction adjacent to hair, likely representing Merkel cells (Figure 2E), consistent with the known expression pattern of Sox2 in skin.
      • Lesko M.H.
      • Driskell R.R.
      • Kretzschmar K.
      • Goldie S.J.
      • Watt F.M.
      Sox2 modulates the function of two distinct cell lineages in mouse skin.
      Thus, the Col1a2 and Sox2 promoters were faithfully expressed. Interestingly, it was found via fluorescence microscopy and quantified via flow cytometry (Figure 2, C, D, and F–H) that a large portion of SKP cells originated from Col1a2-expressing fibroblasts, whereas few were derived from cells that expressed Sox2. Indeed, flow cytometry further confirmed that Sox2 was readily induced in cells of Col1a2 origin in response to culturing in stem cell medium (Figure 2I).
      Standard phenotypic assays have been established to assess the capability of SKPs to differentiate into neural and mesodermal lineages.
      • Fernandes K.J.
      • Miller F.D.
      Isolation, expansion, and differentiation of mouse skin-derived precursors.
      ,
      • Toma J.G.
      • McKenzie I.A.
      • Bagli D.
      • Miller F.D.
      Isolation and characterization of multipotent skin-derived precursors from human skin.
      ,
      • Mirakhori F.
      • Zeynali B.
      • Rassouli H.
      • Shahbazi E.
      • Hashemizadeh S.
      • Kiani S.
      • Salekdeh G.H.
      • Baharvand H.
      Induction of neural progenitor-like cells from human fibroblasts via a genetic material-free approach.
      Neuronal and adipogenic differentiation can be detected by the appearance of βIII-tubulin protein expression and the presence of oil red O–staining lipid droplets, respectively, whereas myofibroblast differentiation is measured by the appearance of α-SMA protein.
      • Fernandes K.J.
      • Miller F.D.
      Isolation, expansion, and differentiation of mouse skin-derived precursors.
      ,
      • Toma J.G.
      • McKenzie I.A.
      • Bagli D.
      • Miller F.D.
      Isolation and characterization of multipotent skin-derived precursors from human skin.
      ,
      • Budel L.
      • Djabali K.
      Rapid isolation and expansion of skin-derived precursor cells from human primary fibroblast cultures.
      ,
      • Biernaskie J.
      • Paris M.
      • Morozova O.
      • Fagan B.M.
      • Marra M.
      • Pevny L.
      • Miller F.D.
      SKPs derive from hair follicle precursors and exhibit properties of adult dermal stem cells.
      To test whether the subset of fibroblast-derived SKPs possessed multipotent differentiation potential, SKPs cultured from Col1a2-CreER(T); Rosa26mTmG promoter mice were exposed to various differentiation media; and it was found that differentiated cells expressing myofibroblast, neuronal, and adipocyte markers could all be traced back to a Col1a2 origin via GFP expression (Figure 3). These data collectively support the notion that fibroblasts are highly plastic and can acquire aspects of a progenitor cell phenotype, and they further suggest that SKP spheroids can be used to model fibroblast plasticity similar to what is seen in melanoma.
      Figure thumbnail gr3
      Figure 3Fibroblast-derived skin-derived precursors (SKPs) possess multipotent differentiation potential. SKPs cultured from the dorsal skin of Col1a2-CreER(T); Rosa26mTmG reporter mice were grown in media containing 0.5% fetal bovine serum (FBS), 40 ng/mL NGF, or adipogenesis supplemented, allowing for differentiation into myofibroblast, neuronal, or adipocyte (Adipo)–like cells, respectively. Indirect immunofluorescence analysis was performed to assess for colocalization (arrow) of green fluorescent protein (GFP) and markers of myofibroblast [α-smooth muscle actin (α-SMA); A], neuronal (βIII-tubulin; B), and adipocyte differentiation (oil red O; C) to indicate a Col1a2 origin. Representative images are shown. Arrows denote Col1a2-lineage cells that have differentiated into myofibroblast- (A), neuronal- (B) or adipocyte-like (C) cells. N = 3 independent experiments (AC). Scale bars = 100 μm (AC).

      The Ability of SKPs to Differentiate into Myofibroblasts in Response to Serum Is Impeded by Focal Adhesion Kinase/Src or Myocardin-Related Transcription Factor/Serum Response Factor Inhibition and Is Dependent on CCN2

      Mechanotransduction, mediated through focal adhesions, has been shown, in other systems, to mediate myofibroblast differentiation and to be important for the initiation and maintenance of fibrogenic responses in vitro and in vivo as well as in the acquisition of drug resistance in melanoma.
      • Taylor K.N.
      • Schlaepfer D.D.
      Adaptive resistance to chemotherapy, a multi-FAK-torial linkage.
      ,
      • Hirata E.
      • Girotti M.R.
      • Viros A.
      • Hooper S.
      • Spencer-Dene B.
      • Matsuda M.
      • Larkin J.
      • Marais R.
      • Sahai E.
      Intravital imaging reveals how BRAF inhibition generates drug-tolerant microenvironments with high integrin beta1/FAK signaling.
      ,
      • Lagares D.
      • Busnadiego O.
      • Garcia-Fernandez R.A.
      • Kapoor M.
      • Liu S.
      • Carter D.E.
      • Abraham D.
      • Shi-Wen X.
      • Carreira P.
      • Fontaine B.A.
      • Shea B.S.
      • Tager A.M.
      • Leask A.
      • Lamas S.
      • Rodriguez-Pascual F.
      Inhibition of focal adhesion kinase prevents experimental lung fibrosis and myofibroblast formation.
      In response to serum, SKP spheroids attach to and spread on the surrounding matrix, forming a monolayer. To assess for pathways essential for SKP differentiation into myofibroblasts, SKPs were treated with FBS to induce myofibroblast differentiation in the presence or absence of the Src-family kinase inhibitor PP2, the focal adhesion kinase inhibitor PF228, or the myocardin-related transcription factor/serum response factor inhibitor CCG-1423.
      • Budel L.
      • Djabali K.
      Rapid isolation and expansion of skin-derived precursor cells from human primary fibroblast cultures.
      • Biernaskie J.
      • Paris M.
      • Morozova O.
      • Fagan B.M.
      • Marra M.
      • Pevny L.
      • Miller F.D.
      SKPs derive from hair follicle precursors and exhibit properties of adult dermal stem cells.
      • Lagares D.
      • Busnadiego O.
      • Garcia-Fernandez R.A.
      • Kapoor M.
      • Liu S.
      • Carter D.E.
      • Abraham D.
      • Shi-Wen X.
      • Carreira P.
      • Fontaine B.A.
      • Shea B.S.
      • Tager A.M.
      • Leask A.
      • Lamas S.
      • Rodriguez-Pascual F.
      Inhibition of focal adhesion kinase prevents experimental lung fibrosis and myofibroblast formation.
      These inhibitors were chosen as focal adhesion kinase/Src mediates adhesive signaling, resulting from integrin-mediated attachment to ECM; and the myocardin-related transcription factor/serum response factor pathway mediates responses to mechanotransduction.
      • Lagares D.
      • Busnadiego O.
      • Garcia-Fernandez R.A.
      • Kapoor M.
      • Liu S.
      • Carter D.E.
      • Abraham D.
      • Shi-Wen X.
      • Carreira P.
      • Fontaine B.A.
      • Shea B.S.
      • Tager A.M.
      • Leask A.
      • Lamas S.
      • Rodriguez-Pascual F.
      Inhibition of focal adhesion kinase prevents experimental lung fibrosis and myofibroblast formation.
      • Yoon H.
      • Dehart J.P.
      • Murphy J.M.
      • Lim S.T.
      Understanding the roles of FAK in cancer: inhibitors, genetic models, and new insights.
      • Johnson L.A.
      • Rodansky E.S.
      • Haak A.J.
      • Larsen S.D.
      • Neubig R.R.
      • Higgins P.D.
      Novel Rho/MRTF/SRF inhibitors block matrix-stiffness and TGF-beta-induced fibrogenesis in human colonic myofibroblasts.
      All three inhibitors blocked spreading of spheres into monolayers and prevented myofibroblast differentiation, as revealed by immunofluorescence and real-time PCR analysis examining expression of α-SMA and CCN2 (Figure 4), suggesting that cell attachment to ECM mediates progenitor cell myofibroblast differentiation.
      Figure thumbnail gr4
      Figure 4Inhibition of Src, focal adhesion kinase (FAK), and serum response factor/myocardin-related transcription factor (SRF/MRTFA) prevents myofibroblastic differentiation of skin-derived precursors (SKPs) and suppresses the activation of CCN2. SKP spheroids were pretreated with dimethyl sulfoxide (DMSO), PP2 (Src-family kinase inhibitor; 10 μmol/L), PF228 (FAK inhibitor; 10 μmol/L), or CCG-1432 (SRF/MRTF inhibitor; 50 μmol/L) for 30 minutes before 24 hours of treatment with or without (negative control) 0.5% fetal bovine serum (FBS). Representative phase contrast and immunofluorescence images for α-smooth muscle actin (α-SMA) and CCN2 are shown. RNA was harvested and subjected to real-time quantitative PCR analysis for α-SMA and CCN2 expression; fold change is shown. Data are expressed as means ± SEM. N = 3. *P < 0.05, **P < 0.01, and ***P < 0.001 (one-way analysis of variance, Tukey post-hoc test). Scale bars = 300 μm.

      CCN2 Is Required for the Ability of SKPs to Differentiate into Myofibroblasts

      The role of CCN2 in SKP differentiation was further elucidated. Although SKPs could readily be isolated from both wild-type CCN2-deficent skin, CCN2 deficiency impaired the ability of SKPs to differentiate into α-SMA–expressing cells (ie, myofibroblasts), as shown previously,
      • Liu S.
      • Herault Y.
      • Pavlovic G.
      • Leask A.
      Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
      as visualized by immunofluorescence analysis (Figure 5A). This effect was independent of cell proliferation, demonstrated by quantification of Ki-67 staining. Results were further confirmed by Western blot and real-time quantitative PCR analysis examining α-SMA and CCN2 expression, demonstrating significant decreases in both (Figure 5, B and C). SKPs exhibit baseline expression of α-SMA before differentiation in serum, as the protein is expressed by cells on the outer surface of the spheroids. The standard procedure in the literature for measuring neural differentiation and adipogenic differentiation is the appearance of βIII-tubulin protein expression and the presence of lipid droplets, respectively; and myofibroblast differentiation is measured by the appearance of α-SMA protein,
      • Li I.M.H.
      • Horwell A.L.
      • Chu G.
      • de Crombrugghe B.
      • Bou-Gharios G.
      Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
      • Liu S.
      • Herault Y.
      • Pavlovic G.
      • Leask A.
      Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
      • Laga A.C.
      • Lai C.Y.
      • Zhan Q.
      • Huang S.J.
      • Velazquez E.F.
      • Yang Q.
      • Hsu M.Y.
      • Murphy G.F.
      Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
      ,
      • Toma J.G.
      • McKenzie I.A.
      • Bagli D.
      • Miller F.D.
      Isolation and characterization of multipotent skin-derived precursors from human skin.
      as these are measures of the overall phenotype of the cells. Differentiation into neuronal or adipocyte-like cells was unaffected by loss of CCN2, as demonstrated by βIII-tubulin expression via immunofluorescence and Western blot analysis (Figure 6A) and quantification of oil red O staining of lipid droplets (Figure 6B), respectively.
      Figure thumbnail gr5
      Figure 5Col1a2-specific CCN2 knockdown impairs myofibroblast skin-derived precursor (SKP) differentiation. SKPs cultured from Col1a2-CreER(T); Rosa26mTmG; Ccn2fl/fl or Col1a2-CreER(T); Rosa26mTmG; Ccn2−/− mice were grown in the presence [+ fetal bovine serum (FBS)] or absence (− FBS) of myofibroblast differentiation media. A: Representative immunofluorescence images for α-smooth muscle actin (α-SMA), CCN2, and Ki-67 expression in response to myofibroblast differentiation media are shown. Immunostaining for Ki-67 is expressed as percentage of proliferating cells. B and C: Real-time quantitative PCR (B) and Western blot analysis (C) were performed to assess α-SMA and CCN2 expression. Data are expressed as means ± SEM (B and C). N = average of 10 fields (A); N = 3 independent experiments (B and C). *P < 0.05, **P < 0.01, and ***P < 0.001 (one-way analysis of variance, Tukey post-hoc test). Scale bars = 300 μm (A).
      Figure thumbnail gr6
      Figure 6Col1a2-specific Ccn2 knockdown does not impair neuronal or adipocytic skin-derived precursor (SKP) differentiation. SKPs cultured from Col1a2-CreER(T); Rosa26mTmG; Ccn2fl/fl or Col1a2-CreER(T); Rosa26mTmG; Ccn2−/− mice were grown in neuronal (NGF) and adipocytic (Adipo) differentiation media. A: Neuronal differentiation in the presence or absence of Ccn2 was assessed through βIII-tubulin and Ccn2 expression via immunofluorescence and Western blot analysis. B: Adipocytic differentiation was assessed via quantification of oil red O staining in green fluorescent protein–positive (GFP+; cells with activated Cre) in Ccn2fl/fl and Ccn2−/− SKPs. Representative images are shown. N = 3 independent experiments (A); N = average of 10 fields (B). *P < 0.05 versus Ccn2fl/fl + NGF. Scale bars = 300 μm (A and B).
      Microarray analysis of Ccn2−/− versus Ccn2fl/fl SKPs treated with FBS revealed 145 differentially expressed genes (Supplemental Table S1). Cluster analysis was performed using the DAVID Functional Annotation Tool, which demonstrated alterations in gene clusters related to the extracellular matrix, cytoskeleton, cell adhesion, and cell migration (Table 1). Notably, five genes encoding contractile proteins were down-regulated (α-skeletal muscle actin, α-smooth muscle actin, myosin light chain 9, myosin light chain 10, and transgelin). Real-time quantitative PCR was performed to verify the differential expression of select genes associated with each altered gene cluster (Figure 7). In Ccn2-deficient SKPs, significant decreases in expression in response to serum were confirmed for transgelin (TAGLN), integrin α-11 (ITGA11), Thy1 cell surface antigen (THY1), epidermal growth factor receptor (EGFR), and metalloproteinase inhibitor 3 (TIMP3), whereas significantly increased expression was demonstrated for matrix metallopeptidase 9 (MMP9). Briefly, TAGLN encodes an actin-binding protein that is an early marker of smooth muscle differentiation.
      • Liu R.
      • Hossain M.M.
      • Chen X.
      • Jin J.P.
      Mechanoregulation of SM22alpha/Transgelin.
      ITGA11 is a profibrotic gene involved in mediating ECM stiffness and metastasis in cancer models.
      • Navab R.
      • Strumpf D.
      • To C.
      • Pasko E.
      • Kim K.S.
      • Park C.J.
      • Hai J.
      • Liu J.
      • Jonkman J.
      • Barczyk M.
      • Bandarchi B.
      • Wang Y.H.
      • Venkat K.
      • Ibrahimov E.
      • Pham N.A.
      • Ng C.
      • Radulovich N.
      • Zhu C.Q.
      • Pintilie M.
      • Wang D.
      • Lu A.
      • Jurisica I.
      • Walker G.C.
      • Gullberg D.
      • Tsao M.S.
      Integrin alpha11beta1 regulates cancer stromal stiffness and promotes tumorigenicity and metastasis in non-small cell lung cancer.
      THY1 mediates cell adhesion and is also a mesenchymal stem cell marker whose expression in cancer is associated with disease progression.
      • Numakura S.
      • Uozaki H.
      • Kikuchi Y.
      • Watabe S.
      • Togashi A.
      • Watanabe M.
      Mesenchymal stem cell marker expression in gastric cancer stroma.
      EGFR plays a critical role in transforming growth factor-β–dependent fibroblast to myofibroblast differentiation and also drives cancer cell proliferation when overexpressed.
      • Kato T.
      • Nishio K.
      Clinical aspects of epidermal growth factor receptor inhibitors: benefit and risk.
      MMP9 and its inhibitor TIMP3 are known to play pivotal roles in regulating extracellular matrix degradation, thereby mediating physiological processes of wound healing and angiogenesis as well as driving pathologic processes, including cancer.
      • Gill S.E.
      • Parks W.C.
      Metalloproteinases and their inhibitors: regulators of wound healing.
      Table 1Microarray Was Performed on SKPs Cultured from Ccn2fl/fl and Ccn2−/− Mice before and after 24 Hours of Treatment with 0.5% FBS to Induce Myofibroblast Differentiation
      Gene nameGene symbolFold change
      Extracellular matrix
       Matrix metallopeptidase 9Mmp91.83033
       Matrix metallopeptidase 1aMmp1a1.77374
       Connective tissue growth factorCtgf−3.03684
       Tissue inhibitor of metalloproteinase 3Timp3−1.77853
       ADAMTS-like 1Adamtsl1−1.69146
       Collagen, type XII, α 1Col12a1−1.42349
       AggrecanAcan−1.41547
      Cytoskeleton
       Myosin, light chain 9Myl9−2.38173
       Actin, α 2, smooth muscleActa2−1.64337
       TransgelinTagln−1.52242
       Actin, α 1, skeletal muscleActa1−1.52047
       A kinase anchor protein 12Akap12−1.51423
       NexilinNexn−1.41814
      Cell adhesion
       Connective tissue growth factorCtgf−3.03684
       Sorbin and SH3 domain containing 1Sorbs1−1.96531
       LIM and senescent cell antigen–like domains 2Lims2−1.68726
       Multiple EGF-like domains 10Megf10−1.59757
       Thymus cell antigen 1, θThy1−1.58788
       Contactin 1Cntn1−1.50956
       DermatopontinDpt−1.45043
       Integrin α 11Itga11−1.41949
      Cell migration
       Connective tissue growth factorCtgf−3.03684
       Nerve growth factor receptorNgfr−1.50912
       Epidermal growth factor receptorEgfr−1.42589
       Integrin α 11Itga11−1.41949
      Cluster analysis using DAVID Functional Annotation Software version 6.7 showed alterations in cytoskeleton, extracellular matrix, cell adhesion, and cell migration gene clusters in Ccn2−/− versus Ccn2fl/fl SKPs in response to FBS. A negative fold change value indicates lower expression in Ccn2−/−, whereas a positive value indicates increased expression in Ccn2−/−.
      ADAMTS, a disintegrin and metalloproteinase with thrombospondin motifs; EGF, epidermal growth factor; FBS, fetal bovine serum; SKP, skin-derived precursor.
      Figure thumbnail gr7
      Figure 7Loss of Ccn2 alters cytoskeleton, extracellular matrix, cell adhesion, and cell migration gene expression to impair myofibroblastic differentiation. Microarray was performed on skin-derived precursors (SKPs) cultured from Ccn2fl/fl and Ccn2−/− mice before and after 24 hours treatment with 0.5% fetal bovine serum (FBS) to induce myofibroblast differentiation. Real-time quantitative PCR was performed to verify the differential expression of Tagln, Itga11, Thy1, Egfr, Mmp9, and Timp3. Graphs represent RNA fold change. Data were calculated relative to sample 1 of Ccn2fl/fl SKPs treated with FBS. Data are expressed as means ± SEM. N = 3 independent experiments. *P < 0.05, **P < 0.01 versus Ccn2fl/fl + FBS (t-test).
      Collectively, these results emphasize the selectivity of CCN2 action and that CCN2 may represent a specific profibrotic mediator through its requirement for progenitor cells to differentiate CAFs into activated cells through regulation of extracellular matrix, cytoskeleton, cell adhesion, and cell migration genes.

      In Melanoma Patients, CCN2 Expression in CAFs Correlates with Expression of a Fibrotic Signature

      In cancers, including in melanoma, α-SMA–expressing CAFs are a heterogeneous population, composed of subsets differing on the basis of their expression profile. To provide a clinical context for these studies, a single-cell RNA-sequencing data set was used to determine which genes correlated with CCN2 expression in CAF single cells from human melanoma tumors. CCN2 expression correlated with expression of FAP expression and several profibrotic genes, including ITGA11, COL1A1, P4HA2, and LOX (Supplemental Table S2 and Figure 8A). These genes are known to contribute to the generation of a stiff, fibrotic ECM, and, hence, to drug resistance.
      • Hirata E.
      • Girotti M.R.
      • Viros A.
      • Hooper S.
      • Spencer-Dene B.
      • Matsuda M.
      • Larkin J.
      • Marais R.
      • Sahai E.
      Intravital imaging reveals how BRAF inhibition generates drug-tolerant microenvironments with high integrin beta1/FAK signaling.
      ,
      • Lee H.O.
      • Mullins S.R.
      • Franco-Barraza J.
      • Valianou M.
      • Cukierman E.
      • Cheng J.D.
      FAP overexpressing fibroblasts produce an extracellular matrix that enhances invasive velocity and directionality of pancreatic cancer cells.
      ,
      • Schnittert J.
      • Bansal R.
      • Mardhian D.F.
      • van Baarlen J.
      • Östman A.
      • Prakash J.
      Integrin α11 in pancreatic stellate cells regulates tumor stroma interaction in pancreatic cancer.
      As expected, ACTA2 (the gene encoding α-SMA) expression correlated with a broader set of CAFs (Figure 8A).
      Figure thumbnail gr8
      Figure 8CCN2 expression in human melanoma tumor single cells. A: Heat map representation of expression of select genes in 106 human cancer-associated fibroblast (CAF) single cells from melanoma tumors depicts CCN2 expression in a particular subset of CAFs (https://www.ncbi.nlm.nih.gov/geo; accession number GSE115978). Columns represent single CAF cells, and rows represent Z-score mRNA expression (TPM/10). Horizontal bar denotes three sample clusters. B: Boxplot shows CCN2 gene expression values (TPM/10) in melanoma tumor single cells by cell type [B cells, CAFs, endothelial cells, macrophages, malignant cells, natural killer (NK) cells, CD4+ T cells, CD8+ T cells, and other T cells]. Boxes represent interquartile ranges, and points represent individual sample values (triangles depict outliers). Please note that the data set analyzed is distinct from that previously analyzed.
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      C: Kaplan-Meier curves of overall survival (OS) and disease-free survival (DFS) for patients with high CAF-specific CCN2 scores and those with low CAF-specific CCN2 scores. A nine-gene set of CAF-specific genes correlating with CCN2 (ZFHX4, DCN, ITGA11, COL6A3, COL1A1, ITGBL1, COL8A1, INHBA, and MEG3), as described in , was analyzed. n = 818 B cells (B); n = 106 CAFs (B); n = 104 endothelial cells (B); n = 420 macrophages (B); n = 2018 malignant cells (B); n = 92 NK cells (B); n = 856 CD4+ T cells (B); n = 1759 CD8+ T cells (B); n = 706 other T cells (B); n = 59 OS for patients with high CAF-specific CCN2 scores (C); n = 51 DFS for patients with high CAF-specific CCN2 scores (C); n = 21 OS for patients with low CAF-specific CCN2 scores (C); n = 16 DFS for patients with low CAF-specific CCN2 scores (C). ***P < 0.001 versus high CAF-specific CCN2 score (OS) (log-rank test); ††P < 0.01 versus high CAF-specific CCN2 score (DFS). TPM, transcript-per-million.
      When the expression of CCN2 was examined by cell type in melanoma tumors, median CCN2 expression was highest in CAFs, similar to what was found previously using a smaller data set (Figure 8B).
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      Biologically, CCN2 produced by CAFs may have a different functional impact than CCN2 expressed by other cell types in the tumor microenvironment. However, as CCN2 is expressed in additional cell types in the melanoma tumor environment, it is difficult to associate CAF-specific CCN2 expression with patient outcomes using bulk tumor RNA sequencing. Therefore, a surrogate score was generated for CAF-specific CCN2 expression. Using single-cell RNA sequencing in CAF single cells (https://www.ncbi.nlm.nih.gov/geo; accession number GSE115978),
      • Jerby-Arnon L.
      • Shah P.
      • Cuoco M.S.
      • Rodman C.
      • Su M.J.
      • Melms J.C.
      • et al.
      A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade.
      a cohort of nine genes (ZFHX4, DCN, ITGA11, COL6A3, COL1A1, ITGBL1, COL8A1, INHBA, and MEG3) was identified that strongly correlated with CCN2 expression in CAF cells, and was minimally expressed in other cell types within the tumor microenvironment. These nine genes were used as a surrogate score for CAF-specific CCN2 expression in bulk tumor RNA sequencing. CAF-specific CCN2 scores were dichotomized with receiver operating characteristic curves to ascertain the optimal cutoff for the end point of overall survival censorship. In this data set, a high CAF-specific CCN2 score was associated with poor overall and disease-free survival (P < 0.01) (Figure 8C). These data are consistent with the notions that CCN2 is expressed in a profibrotic subset of CAFs and, hence, may mediate a profibrotic and prometastatic phenotype in melanoma patients and targeting CCN2 may be a novel therapeutic strategy.

      Discussion

      Cancer biologists have proposed a cancer stem cell theory of cancer, which suggests that a subpopulation of self-renewing tumor cells is responsible for tumorigenesis
      • Carruba G.
      • Trosko J.E.
      The long evolutionary journey of cancer from ancestor to modern humans.
      In human melanoma, Sox2, an embryonic neural crest stem-cell transcription factor, is abundantly expressed,
      • Laga A.C.
      • Lai C.Y.
      • Zhan Q.
      • Huang S.J.
      • Velazquez E.F.
      • Yang Q.
      • Hsu M.Y.
      • Murphy G.F.
      Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
      with the 3-year median survival for patients with Sox2-expressing metastatic tumors being 145 days less than those with metastatic tumors that do not express Sox2.
      • Schoenhals M.
      • Kassambara A.
      • De Vos J.
      • Hose D.
      • Moreaux J.
      • Klein B.
      Embryonic stem cell markers expression in cancers.
      CAFs, one of the most prominent cell types present in the tumor stroma, and of likely heterogeneous origin, are now recognized as playing an active role in cancer progression and metastasis, and have been implicated in providing a survival niche for cancer stem cells.
      • Xiong S.
      • Wang R.
      • Chen Q.
      • Luo J.
      • Wang J.
      • Zhao Z.
      • Li Y.
      • Wang Y.
      • Wang X.
      • Cheng B.
      Cancer-associated fibroblasts promote stem cell-like properties of hepatocellular carcinoma cells through IL-6/STAT3/Notch signaling.
      • Su S.
      • Chen J.
      • Yao H.
      • Liu J.
      • Yu S.
      • Lao L.
      • Wang M.
      • Luo M.
      • Xing Y.
      • Chen F.
      • Huang D.
      • Zhao J.
      • Yang L.
      • Liao D.
      • Su F.
      • Li M.
      • Liu Q.
      • Song E.
      CD10(+)GPR77(+) cancer-associated fibroblasts promote cancer formation and chemoresistance by sustaining cancer stemness.
      • Herrera M.
      • Islam A.B.
      • Herrera A.
      • Martin P.
      • Garcia V.
      • Silva J.
      • Garcia J.M.
      • Salas C.
      • Casal I.
      • de Herreros A.G.
      • Bonilla F.
      • Pena C.
      Functional heterogeneity of cancer-associated fibroblasts from human colon tumors shows specific prognostic gene expression signature.
      • Sukowati C.H.
      • Anfuso B.
      • Croce L.S.
      • Tiribelli C.
      The role of multipotent cancer associated fibroblasts in hepatocarcinogenesis.
      CAFs have also been proposed to possess inherent stemness characteristics similar to mesenchymal stem cells, and Sox2 has been noted to be expressed in tumor stroma cells of human melanoma
      • Laga A.C.
      • Lai C.Y.
      • Zhan Q.
      • Huang S.J.
      • Velazquez E.F.
      • Yang Q.
      • Hsu M.Y.
      • Murphy G.F.
      Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
      and colorectal cancer.
      • Xiong S.
      • Wang R.
      • Chen Q.
      • Luo J.
      • Wang J.
      • Zhao Z.
      • Li Y.
      • Wang Y.
      • Wang X.
      • Cheng B.
      Cancer-associated fibroblasts promote stem cell-like properties of hepatocellular carcinoma cells through IL-6/STAT3/Notch signaling.
      • Su S.
      • Chen J.
      • Yao H.
      • Liu J.
      • Yu S.
      • Lao L.
      • Wang M.
      • Luo M.
      • Xing Y.
      • Chen F.
      • Huang D.
      • Zhao J.
      • Yang L.
      • Liao D.
      • Su F.
      • Li M.
      • Liu Q.
      • Song E.
      CD10(+)GPR77(+) cancer-associated fibroblasts promote cancer formation and chemoresistance by sustaining cancer stemness.
      • Herrera M.
      • Islam A.B.
      • Herrera A.
      • Martin P.
      • Garcia V.
      • Silva J.
      • Garcia J.M.
      • Salas C.
      • Casal I.
      • de Herreros A.G.
      • Bonilla F.
      • Pena C.
      Functional heterogeneity of cancer-associated fibroblasts from human colon tumors shows specific prognostic gene expression signature.
      • Sukowati C.H.
      • Anfuso B.
      • Croce L.S.
      • Tiribelli C.
      The role of multipotent cancer associated fibroblasts in hepatocarcinogenesis.
      In this study, using lineage tracing in a murine model of melanoma, it was shown that resident Col1a2-expressing (ie, synthetic) dermal fibroblasts recruited adjacent to tumor can become induced to express the stem cell marker Sox2 (also expressed by tumor cells) and that these cells are activated into an α-SMA–expressing CAF phenotype, all in a CCN2-dependent manner. It should be emphasized that by using GFP to specifically label host fibroblasts, this process could be distinguished from tumor cells potentially undergoing epithelial-mesenchymal transition
      • Kalluri R.
      • Weinberg R.A.
      The basics of epithelial-mesenchymal transition.
      or host endothelial cells underdoing endothelial-mesenchymal transition,
      • Zeisberg E.M.
      • Potenta S.
      • Xie L.
      • Zeisberg M.
      • Kalluri R.
      Discovery of endothelial to mesenchymal transition as a source for carcinoma-associated fibroblasts.
      which in other models have also been shown to express α-SMA and have been proposed to contribute to CAF populations.
      Col1a2-expressing dermal fibroblasts were shown to be capable of being induced to express Sox2 when cultured as multipotent SKP spheroids. Lineage tracing experiments demonstrated that these cells could differentiate into neuronal, adipocytic, and myofibroblastic phenotypes. In this model, CCN2 was required for the differentiation of progenitor cells into contractile myofibroblasts through regulation of extracellular matrix, cytoskeleton, cell adhesion, and cell migration genes. These data support the notion that host fibroblast–derived CAFs are capable of exhibiting stem cell–like characteristics and emphasize an underappreciated phenotypic plasticity of dermal fibroblasts. Specifically, when first described, SKPs were thought to originate from endogenous Sox2-expressing hair follicle dermal papilla and dermal sheath.
      • Biernaskie J.
      • Paris M.
      • Morozova O.
      • Fagan B.M.
      • Marra M.
      • Pevny L.
      • Miller F.D.
      SKPs derive from hair follicle precursors and exhibit properties of adult dermal stem cells.
      Alternatively, it has been suggested that a subpopulation of progenitor cells exists in the dermis, and this subpopulation is being selected for in stem cell media.
      • Fernandes K.J.
      • Miller F.D.
      Isolation, expansion, and differentiation of mouse skin-derived precursors.
      ,
      • Toma J.G.
      • McKenzie I.A.
      • Bagli D.
      • Miller F.D.
      Isolation and characterization of multipotent skin-derived precursors from human skin.
      ,
      • Budel L.
      • Djabali K.
      Rapid isolation and expansion of skin-derived precursor cells from human primary fibroblast cultures.
      ,
      • Takahashi K.
      • Yamanaka S.
      Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors.
      However, the data showing that, in tumor stroma, cells derived from Col1a2-expressing fibroblasts can acquire the expression of Sox2 suggest that cultured SKPs can partly originate from Col1a2-expressing fibroblasts. These data are consistent with a prior experiment suggesting that SKPs can be isolated from primary cultures of dermal fibroblasts; however, in that study, it was unclear if synthetic (ie, collagen-producing) fibroblasts could actively behave like progenitor cells or if an already existing progenitor cell population was being selected.
      • Hill R.P.
      • Gledhill K.
      • Gardner A.
      • Higgins C.A.
      • Crawford H.
      • Lawrence C.
      • Hutchison C.J.
      • Owens W.A.
      • Kara B.
      • James S.E.
      • Jahoda C.A.
      Generation and characterization of multipotent stem cells from established dermal cultures.
      ,
      • Krause M.P.
      • Dworski S.
      • Feinberg K.
      • Jones K.
      • Johnston A.P.
      • Paul S.
      • Paris M.
      • Peles E.
      • Bagli D.
      • Forrest C.R.
      • Kaplan D.R.
      • Miller F.D.
      Direct genesis of functional rodent and human schwann cells from skin mesenchymal precursors.
      In this study, the ability of fibroblasts to behave like progenitor cells in vivo was triggered by the presence of the tumor microenvironment. It remains unclear if this property is unique to dermal fibroblasts or if it might be also observed in other tissues, although the latter seems likely.
      CCN2 is under clinical development as a therapeutic target in idiopathic pulmonary fibrosis and pancreatic cancer.
      • Riser B.L.
      • Barnes J.L.
      • Varani J.
      Balanced regulation of the CCN family of matricellular proteins: a novel approach to the prevention and treatment of fibrosis and cancer.
      • Perbal B.
      • Tweedie S.
      • Bruford E.
      The official unified nomenclature adopted by the HGNC calls for the use of the acronyms, CCN1-6, and discontinuation in the use of CYR61, CTGF, NOV and WISP 1-3 respectively.
      • Ramazani Y.
      • Knops N.
      • Elmonem M.A.
      • Nguyen T.Q.
      • Arcolino F.O.
      • van den Heuvel L.
      • Levtchenko E.
      • Kuypers D.
      • Goldschmeding R.
      Connective tissue growth factor (CTGF) from basics to clinics.
      ,
      • Raghu G.
      • Scholand M.B.
      • de Andrade J.
      • Lancaster L.
      • Mageto Y.
      • Goldin J.
      • Brown K.K.
      • Flaherty K.R.
      • Wencel M.
      • Wanger J.
      • Neff T.
      • Valone F.
      • Stauffer J.
      • Porter S.
      FG-3019 anti-connective tissue growth factor monoclonal antibody: results of an open-label clinical trial in idiopathic pulmonary fibrosis.
      Recent evidence has emerged that targeting CCN2 may also be of value in melanoma.
      • Hutchenreuther J.
      • Vincent K.M.
      • Carter D.E.
      • Postovit L.M.
      • Leask A.
      CCN2 expression by tumor stroma is required for melanoma metastasis.
      ,
      • Hutchenreuther J.
      • Vincent K.
      • Norley C.
      • Racanelli M.
      • Gruber S.B.
      • Johnson T.M.
      • Fullen D.R.
      • Raskin L.
      • Perbal B.
      • Holdsworth D.W.
      • Postovit L.M.
      • Leask A.
      Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
      ,
      • Hutchenreuther J.
      • Leask A.
      Why target the tumor stroma in melanoma?.
      As a member of the CCN family of matricellular proteins, CCN2 has limited in vitro activity as a proadhesive molecule; however, because of its ability to interact with a wide variety of proteins and receptors, CCN2 has profound effects in development and pathologic processes in vivo.
      • Perbal B.
      The concept of the CCN protein family revisited: a centralized coordination network.
      Uncovering these in vivo roles has involved the use of genetic (whole-body and conditional knockout) approaches.
      • Falke L.L.
      • Leeuwis J.W.
      • Lyons K.M.
      • Mummery C.L.
      • Nguyen T.Q.
      • Goldschmeding R.
      CCN2 reduction mediates protective effects of BMP7 treatment in obstructive nephropathy.
      ,
      • Petrosino J.M.
      • Leask A.
      • Accornero F.
      Genetic manipulation of CCN2/CTGF unveils cell-specific ECM-remodeling effects in injured skeletal muscle.
      Previously, in a mouse model of bleomycin-induced skin fibrosis,
      • Liu S.
      • Herault Y.
      • Pavlovic G.
      • Leask A.
      Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
      and now in a model of cutaneous melanoma, we have shown that CCN2 is likely necessary for the appearance of myofibroblasts involving progenitor cell–like intermediates. Conversely, this mechanism may not play an appreciable role in cutaneous wound healing,
      • Liu S.
      • Thompson K.
      • Leask A.
      CCN2 expression by fibroblasts is not required for cutaneous tissue repair.
      • Liu S.
      • Taghavi R.
      • Leask A.
      Connective tissue growth factor is induced in bleomycin-induced skin scleroderma.
      • Tsang M.
      • Leask A.
      CCN2 is required for recruitment of Sox2-expressing cells during cutaneous tissue repair.
      suggesting that CCN2 represents a specific target for pathologic fibrosis through differentiation of various types of precursor cells into myofibroblasts. Consistent with this notion, in CAFs of melanoma patients, CCN2 expression correlated with a subset of CAFs, expressing specific fibrogenic genes, including ITGA11, which negatively correlates with survival. Conversely, the related protein CCN1 appears to mediate collagen fiber alignment and stability in fibrotic conditions.
      • Quensel K.
      • Shiwen X.
      • Hutchenreuther J.
      • Xiao Y.
      • Liu S.
      • Peidl A.
      • Naskar D.
      • Siqueira W.L.
      • O'Gorman D.B.
      • Hinz B.
      • Stratton R.J.
      • Leask A.
      CCN1 expression by fibroblasts is required for bleomycin-induced skin fibrosis.
      It is interesting to note that in a murine model of melanoma, there were fewer fibroblasts expressing Sox2 in CCN2-deficient stroma. Conversely, fibroblasts ex vivo were able to form multipotent SKP spheroids in the absence of CCN2. These results suggest that the ability of fibroblasts to express Sox2 does not absolutely require CCN2; rather, CCN2 is required for the differentiation of fibroblasts to α-SMA–expressing CAFs via a Sox2 (progenitor cell–like) intermediate. Alternatively, as there were fewer Sox2-expressing cells in the absence to CCN2, these results also suggest that the mechanism of how fibroblasts acquire a Sox2+ phenotype may differ in vitro versus in vivo. Several different methods have been developed that show dermal fibroblasts transition into progenitor cells/neuropheres.
      • Laga A.C.
      • Lai C.Y.
      • Zhan Q.
      • Huang S.J.
      • Velazquez E.F.
      • Yang Q.
      • Hsu M.Y.
      • Murphy G.F.
      Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
      ,
      • Sukowati C.H.
      • Anfuso B.
      • Croce L.S.
      • Tiribelli C.
      The role of multipotent cancer associated fibroblasts in hepatocarcinogenesis.
      ,
      • Kalluri R.
      • Weinberg R.A.
      The basics of epithelial-mesenchymal transition.
      In vivo, it is likely that activation of fibroblasts to a Sox2-expressing phenotype involves cross talk from tumor (melanoma) cells. CCN2 might represent a competency factor, allowing fibroblasts to maximally respond to these yet unidentified signals.
      Overall, the current data show that cells derived from Col1a2-expressing cells are major contributors to SKP spheres and progenitor-like tumor stromal cells that differentiate into CAFs. Collectively, these data point to the fact that at least a subset of fibroblasts, notably those derived from type I collagen-expressing cells, are plastic, can readily acquire a progenitor cell phenotype, and may in the future represent a valuable source of progenitor cells for therapeutic applications. Moreover, the ability of collagen-expressing fibroblasts in vivo to acquire a progenitor cell–like phenotype (eg, the acquisition of Sox2 expression) may indicate a pathologic phenotype and an intermediate in the differentiation of fibroblasts to myofibroblasts. That Sox2-expressing cells represent a transient intermediate cell type in CAF differentiation is consistent with prior studies performed in CAFs derived from human colorectal tumors, which have demonstrated expression of Sox2 and other stem cell markers, in association with promigratory effects,
      • Herrera M.
      • Islam A.B.
      • Herrera A.
      • Martin P.
      • Garcia V.
      • Silva J.
      • Garcia J.M.
      • Salas C.
      • Casal I.
      • de Herreros A.G.
      • Bonilla F.
      • Pena C.
      Functional heterogeneity of cancer-associated fibroblasts from human colon tumors shows specific prognostic gene expression signature.
      high proliferation potential, and poor prognosis.
      • Su S.
      • Chen J.
      • Yao H.
      • Liu J.
      • Yu S.
      • Lao L.
      • Wang M.
      • Luo M.
      • Xing Y.
      • Chen F.
      • Huang D.
      • Zhao J.
      • Yang L.
      • Liao D.
      • Su F.
      • Li M.
      • Liu Q.
      • Song E.
      CD10(+)GPR77(+) cancer-associated fibroblasts promote cancer formation and chemoresistance by sustaining cancer stemness.
      Furthermore, Sox2 is a recognized marker of both cancer and mesenchymal stem cells; thus, these data are consistent with the notion that mesenchymal stem cells or mesenchymal stem cell–like cells play roles in fibroproliferative responses, including in cancers.
      • Katoh M.
      • Nakagama H.
      FGF receptors: cancer biology and therapeutics.
      • Mamun M.A.
      • Mannoor K.
      • Cao J.
      • Qadri F.
      • Song X.
      SOX2 in cancer stemness: tumor malignancy and therapeutic potentials.
      • Pelizzo G.
      • Veschi V.
      • Mantelli M.
      • Croce S.
      • Di Benedetto V.
      • D'Angelo P.
      • Maltese A.
      • Catenacci L.
      • Apuzzo T.
      • Scavo E.
      • Moretta A.
      • Todaro M.
      • Stassi G.
      • Avanzini M.A.
      • Calcaterra V.
      Microenvironment in neuroblastoma: isolation and characterization of tumor-derived mesenchymal stromal cells.
      • Pitrone M.
      • Pizzolanti G.
      • Tomasello L.
      • Coppola A.
      • Morini L.
      • Pantuso G.
      • Ficarella R.
      • Guarnotta V.
      • Perrini S.
      • Giorgino F.
      • Giordano C.
      NANOG plays a hierarchical role in the transcription network regulating the pluripotency and plasticity of adipose tissue-derived stem cells.
      • Riekstina U.
      • Cakstina I.
      • Parfejevs V.
      • Hoogduijn M.
      • Jankovskis G.
      • Muiznieks I.
      • Muceniece R.
      • Ancans J.
      Embryonic stem cell marker expression pattern in human mesenchymal stem cells derived from bone marrow, adipose tissue, heart and dermis.
      Finally, these data suggest that cultured SKPs have a predominantly fibroblast origin and that the use of SKPs appears to be a useful model system for studying fibroblast plasticity in melanoma. Intriguingly, these data extend recent lineage tracing analysis showing mesenchymal precursors (Twist2-expressing cells) contribute to SKPs and can generate a peripheral neural cell type believed to derive directly only from the neural crest.
      • Krause M.P.
      • Dworski S.
      • Feinberg K.
      • Jones K.
      • Johnston A.P.
      • Paul S.
      • Paris M.
      • Peles E.
      • Bagli D.
      • Forrest C.R.
      • Kaplan D.R.
      • Miller F.D.
      Direct genesis of functional rodent and human schwann cells from skin mesenchymal precursors.
      Finally, these data suggest the intriguing possibility that at least a subset of so-called fibroadipogenic precursors, which are proposed to play a key role in fibroproliferative disorders,
      • Takahashi K.
      • Yamanaka S.
      Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors.
      may be synthetic fibroblasts.
      In summary, these data collectively suggest that fibroblasts are plastic in vitro and in vivo and can contribute to the appearance of myofibroblasts, possibly through a progenitor cell intermediate and CCN2. These observations may be of particular relevance in pathologic fibroproliferative conditions involving myofibroblast differentiation, and are consistent with the notion that targeting CCN2 represents a novel therapeutic target.

      Suppemental Data

      Figure thumbnail figs1
      Supplemental Figure S1Col1a2-specific Ccn2 knockout decreases expression of Sox2 and α-smooth muscle actin (α-SMA) in melanoma tumor stroma cells. Col1a2-CreER(T); Rosa26mTmG; Ccn2fl/fl and Ccn2−/− mice were implanted with B16F10 melanoma cells. Sections were stained for green fluorescent protein (GFP; green), α-SMA (blue), and Sox2 (magenta). High-magnification images of stromal cells in Ccn2fl/fl (A) and Ccn2−/− (B) mice are shown to reveal nuclear staining of Sox2. No triple-positive cells were detected in any Ccn2−/− sections. Representative images are shown. N = 3 mice with tumors (A and B).
      • Supplemental Figure S2

        Approximately 25% of dermal cells cultured in monolayer are derived from Col1a2-expressing fibroblasts. Lineage tracing was performed using Col1a2-CreER(T); Rosa26mTmG reporter mice that were injected with tamoxifen 3 weeks after birth. Dermal fibroblasts from these mice were cultured in monolayer, as described in Materials and Methods, and subjected to fluorescence microscopy (A) and flow cytometry with quantification (B). Representative image is shown. Graphs represent percentage of monolayer cells expressing green fluorescent protein (GFP; Col1a2 derived) and tdTomato (all other cells). Data are expressed as means ± SEM. N = 7 independent experiments. Q, quartile.

      References

        • Uong A.
        • Zon L.I.
        Melanocytes in development and cancer.
        J Cell Physiol. 2010; 222: 38-41
        • Kakadia S.
        • Yarlagadda N.
        • Awad R.
        • Kundranda M.
        • Niu J.
        • Naraev B.
        • Mina L.
        • Dragovich T.
        • Gimbel M.
        • Mahmoud F.
        Mechanisms of resistance to BRAF and MEK inhibitors and clinical update of US Food and Drug Administration-approved targeted therapy in advanced melanoma.
        Onco Targets Ther. 2018; 11: 7095-7107
        • Silva I.P.
        • Long G.V.
        Systemic therapy in advanced melanoma: integrating targeted therapy and immunotherapy into clinical practice.
        Curr Opin Oncol. 2017; 29: 484-492
        • Pickup M.W.
        • Mouw J.K.
        • Weaver V.M.
        The extracellular matrix modulates the hallmarks of cancer.
        EMBO Rep. 2014; 15: 1243-1253
        • Werb Z.
        • Lu P.
        The role of stroma in tumor development.
        Cancer J. 2015; 21: 250-253
        • Zhou L.
        • Yang K.
        • Andl T.
        • Wickett R.R.
        • Zhang Y.
        Perspective of targeting cancer-associated fibroblasts in melanoma.
        J Cancer. 2015; 6: 717-726
        • Taylor K.N.
        • Schlaepfer D.D.
        Adaptive resistance to chemotherapy, a multi-FAK-torial linkage.
        Mol Cancer Ther. 2018; 17: 719-723
        • Hirata E.
        • Girotti M.R.
        • Viros A.
        • Hooper S.
        • Spencer-Dene B.
        • Matsuda M.
        • Larkin J.
        • Marais R.
        • Sahai E.
        Intravital imaging reveals how BRAF inhibition generates drug-tolerant microenvironments with high integrin beta1/FAK signaling.
        Cancer Cell. 2015; 27: 574-588
        • Polanska U.M.
        • Orimo A.
        Carcinoma-associated fibroblasts: non-neoplastic tumour-promoting mesenchymal cells.
        J Cell Physiol. 2013; 228: 1651-1657
        • Gandellini P.
        • Andriani F.
        • Merlino G.
        • D'Aiuto F.
        • Roz L.
        • Callari M.
        Complexity in the tumour microenvironment: cancer associated fibroblast gene expression patterns identify both common and unique features of tumour-stroma crosstalk across cancer types.
        Semin Cancer Biol. 2015; 35: 96-106
        • Vincent K.M.
        • Postovit L.M.
        Matricellular proteins in cancer: a focus on secreted Frizzled-related proteins.
        J Cell Commun Signal. 2018; 12: 103-112
        • Tsang M.
        Mesenchymal cells emerge as primary contributors to fibrosis in multiple tissues.
        J Cell Commun Signal. 2014; 8: 3-4
        • Riser B.L.
        • Barnes J.L.
        • Varani J.
        Balanced regulation of the CCN family of matricellular proteins: a novel approach to the prevention and treatment of fibrosis and cancer.
        J Cell Commun Signal. 2015; 9: 327-339
        • Perbal B.
        • Tweedie S.
        • Bruford E.
        The official unified nomenclature adopted by the HGNC calls for the use of the acronyms, CCN1-6, and discontinuation in the use of CYR61, CTGF, NOV and WISP 1-3 respectively.
        J Cell Commun Signal. 2018; 12: 625-629
        • Ramazani Y.
        • Knops N.
        • Elmonem M.A.
        • Nguyen T.Q.
        • Arcolino F.O.
        • van den Heuvel L.
        • Levtchenko E.
        • Kuypers D.
        • Goldschmeding R.
        Connective tissue growth factor (CTGF) from basics to clinics.
        Matrix Biol. 2018; 68-69: 44-66
        • Hutchenreuther J.
        • Vincent K.M.
        • Carter D.E.
        • Postovit L.M.
        • Leask A.
        CCN2 expression by tumor stroma is required for melanoma metastasis.
        J Invest Dermatol. 2015; 135: 2805-2813
        • Hutchenreuther J.
        • Vincent K.
        • Norley C.
        • Racanelli M.
        • Gruber S.B.
        • Johnson T.M.
        • Fullen D.R.
        • Raskin L.
        • Perbal B.
        • Holdsworth D.W.
        • Postovit L.M.
        • Leask A.
        Activation of cancer-associated fibroblasts is required for tumor neovascularization in a murine model of melanoma.
        Matrix Biol. 2018; 74: 52-61
        • Li I.M.H.
        • Horwell A.L.
        • Chu G.
        • de Crombrugghe B.
        • Bou-Gharios G.
        Characterization of mesenchymal-fibroblast cells using the Col1a2 promoter/enhancer.
        Methods Mol Biol. 2017; 1627: 139-161
        • Liu S.
        • Herault Y.
        • Pavlovic G.
        • Leask A.
        Skin progenitor cells contribute to bleomycin-induced skin fibrosis.
        Arthritis Rheum. 2014; 66: 707-713
        • Laga A.C.
        • Lai C.Y.
        • Zhan Q.
        • Huang S.J.
        • Velazquez E.F.
        • Yang Q.
        • Hsu M.Y.
        • Murphy G.F.
        Expression of the embryonic stem cell transcription factor SOX2 in human skin: relevance to melanocyte and merkel cell biology.
        Am J Pathol. 2010; 176: 903-913
        • Bou-Gharios G.
        • Garrett L.A.
        • Rossert J.
        • Niederreither K.
        • Eberspaecher H.
        • Smith C.
        • Black C.
        • Crombrugghe B.
        A potent far-upstream enhancer in the mouse pro alpha 2(I) collagen gene regulates expression of reporter genes in transgenic mice.
        J Cell Biol. 1996; 134: 1333-1344
        • Tanaka S.
        • Antoniv T.T.
        • Liu K.
        • Wang L.
        • Wells D.J.
        • Ramirez F.
        • Bou-Gharios G.
        Cooperativity between far upstream enhancer and proximal promoter elements of the human {alpha}2(I) collagen (COL1A2) gene instructs tissue specificity in transgenic mice.
        J Biol Chem. 2004; 279: 56024-56031
        • Ponticos M.
        • Abraham D.
        • Alexakis C.
        • Lu Q.L.
        • Black C.
        • Partridge T.
        • Bou-Gharios G.
        Col1a2 enhancer regulates collagen activity during development and in adult tissue repair.
        Matrix Biol. 2004; 22: 619-628
        • Fernandes K.J.
        • Miller F.D.
        Isolation, expansion, and differentiation of mouse skin-derived precursors.
        Methods Mol Biol. 2009; 482: 159-170
        • Shi-wen X.
        • Parapuram S.K.
        • Pala D.
        • Chen Y.
        • Carter D.E.
        • Eastwood M.
        • Denton C.P.
        • Abraham D.J.
        • Leask A.
        Requirement of transforming growth factor beta-activated kinase 1 for transforming growth factor beta-induced alpha-smooth muscle actin expression and extracellular matrix contraction in fibroblasts.
        Arthritis Rheum. 2009; 60: 234-241
        • Guo F.
        • Carter D.E.
        • Leask A.
        Mechanical tension increases CCN2/CTGF expression and proliferation in gingival fibroblasts via a TGFbeta-dependent mechanism.
        PLoS One. 2011; 6: e19756
        • Chen S.
        • McLean S.
        • Carter D.E.
        • Leask A.
        The gene expression profile induced by Wnt 3a in NIH 3T3 fibroblasts.
        J Cell Commun Signal. 2007; 1: 175-183
        • Liu S.
        • Xu S.W.
        • Kennedy L.
        • Pala D.
        • Chen Y.
        • Eastwood M.
        • Carter D.E.
        • Black C.M.
        • Abraham D.J.
        • Leask A.
        FAK is required for TGFbeta-induced JNK phosphorylation in fibroblasts: implications for acquisition of a matrix-remodeling phenotype.
        Mol Biol Cell. 2007; 18: 2169-2178
        • Kuk H.
        • Hutchenreuther J.
        • Murphy-Marshman H.
        • Carter D.
        • Leask A.
        5Z-7-oxozeanol inhibits the effects of TGFbeta1 on human gingival fibroblasts.
        PLoS One. 2015; 10: e0123689
        • Huang D.W.
        • Sherman B.T.
        • Lempicki R.A.
        Systematic and integrative analysis of large gene lists using DAVID Bioinformatics Resources.
        Nature Protoc. 2009; 4: 44-57
        • Li B.
        • Dewey C.N.
        RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome.
        BMC Bioinformatics. 2011; 12: 323
        • Lesko M.H.
        • Driskell R.R.
        • Kretzschmar K.
        • Goldie S.J.
        • Watt F.M.
        Sox2 modulates the function of two distinct cell lineages in mouse skin.
        Dev Biol. 2013; 382: 15-26
        • Toma J.G.
        • McKenzie I.A.
        • Bagli D.
        • Miller F.D.
        Isolation and characterization of multipotent skin-derived precursors from human skin.
        Stem Cells. 2005; 23: 727-737
        • Mirakhori F.
        • Zeynali B.
        • Rassouli H.
        • Shahbazi E.
        • Hashemizadeh S.
        • Kiani S.
        • Salekdeh G.H.
        • Baharvand H.
        Induction of neural progenitor-like cells from human fibroblasts via a genetic material-free approach.
        PLoS One. 2015; 10: e0135479
        • Budel L.
        • Djabali K.
        Rapid isolation and expansion of skin-derived precursor cells from human primary fibroblast cultures.
        Biol Open. 2017; 6: 1745-1755
        • Biernaskie J.
        • Paris M.
        • Morozova O.
        • Fagan B.M.
        • Marra M.
        • Pevny L.
        • Miller F.D.
        SKPs derive from hair follicle precursors and exhibit properties of adult dermal stem cells.
        Cell Stem Cell. 2009; 5: 610-623
        • Lagares D.
        • Busnadiego O.
        • Garcia-Fernandez R.A.
        • Kapoor M.
        • Liu S.
        • Carter D.E.
        • Abraham D.
        • Shi-Wen X.
        • Carreira P.
        • Fontaine B.A.
        • Shea B.S.
        • Tager A.M.
        • Leask A.
        • Lamas S.
        • Rodriguez-Pascual F.
        Inhibition of focal adhesion kinase prevents experimental lung fibrosis and myofibroblast formation.
        Arthritis Rheum. 2012; 64: 1653-1664
        • Yoon H.
        • Dehart J.P.
        • Murphy J.M.
        • Lim S.T.
        Understanding the roles of FAK in cancer: inhibitors, genetic models, and new insights.
        J Histochem Cytochem. 2015; 63: 114-128
        • Johnson L.A.
        • Rodansky E.S.
        • Haak A.J.
        • Larsen S.D.
        • Neubig R.R.
        • Higgins P.D.
        Novel Rho/MRTF/SRF inhibitors block matrix-stiffness and TGF-beta-induced fibrogenesis in human colonic myofibroblasts.
        Inflamm Bowel Dis. 2014; 20: 154-165
        • Liu R.
        • Hossain M.M.
        • Chen X.
        • Jin J.P.
        Mechanoregulation of SM22alpha/Transgelin.
        Biochemistry. 2017; 56: 5526-5538
        • Navab R.
        • Strumpf D.
        • To C.
        • Pasko E.
        • Kim K.S.
        • Park C.J.
        • Hai J.
        • Liu J.
        • Jonkman J.
        • Barczyk M.
        • Bandarchi B.
        • Wang Y.H.
        • Venkat K.
        • Ibrahimov E.
        • Pham N.A.
        • Ng C.
        • Radulovich N.
        • Zhu C.Q.
        • Pintilie M.
        • Wang D.
        • Lu A.
        • Jurisica I.
        • Walker G.C.
        • Gullberg D.
        • Tsao M.S.
        Integrin alpha11beta1 regulates cancer stromal stiffness and promotes tumorigenicity and metastasis in non-small cell lung cancer.
        Oncogene. 2016; 35: 1899-1908
        • Numakura S.
        • Uozaki H.
        • Kikuchi Y.
        • Watabe S.
        • Togashi A.
        • Watanabe M.
        Mesenchymal stem cell marker expression in gastric cancer stroma.
        Anticancer Res. 2019; 39: 387-393
        • Kato T.
        • Nishio K.
        Clinical aspects of epidermal growth factor receptor inhibitors: benefit and risk.
        Respirology. 2006; 11: 693-698
        • Gill S.E.
        • Parks W.C.
        Metalloproteinases and their inhibitors: regulators of wound healing.
        Int J Biochem Cell Biol. 2008; 40: 1334-1347
        • Lee H.O.
        • Mullins S.R.
        • Franco-Barraza J.
        • Valianou M.
        • Cukierman E.
        • Cheng J.D.
        FAP overexpressing fibroblasts produce an extracellular matrix that enhances invasive velocity and directionality of pancreatic cancer cells.
        BMC Cancer. 2011; 11: 245
        • Schnittert J.
        • Bansal R.
        • Mardhian D.F.
        • van Baarlen J.
        • Östman A.
        • Prakash J.
        Integrin α11 in pancreatic stellate cells regulates tumor stroma interaction in pancreatic cancer.
        FASEB J. 2019; 33: 6609-6621
        • Jerby-Arnon L.
        • Shah P.
        • Cuoco M.S.
        • Rodman C.
        • Su M.J.
        • Melms J.C.
        • et al.
        A cancer cell program promotes T cell exclusion and resistance to checkpoint blockade.
        Cell. 2018; 175: 984-997
        • Carruba G.
        • Trosko J.E.
        The long evolutionary journey of cancer from ancestor to modern humans.
        Crit Rev Oncog. 2017; 22: 323-352
        • Schoenhals M.
        • Kassambara A.
        • De Vos J.
        • Hose D.
        • Moreaux J.
        • Klein B.
        Embryonic stem cell markers expression in cancers.
        Biochem Biophys Res Commun. 2009; 383: 157-162
        • Xiong S.
        • Wang R.
        • Chen Q.
        • Luo J.
        • Wang J.
        • Zhao Z.
        • Li Y.
        • Wang Y.
        • Wang X.
        • Cheng B.
        Cancer-associated fibroblasts promote stem cell-like properties of hepatocellular carcinoma cells through IL-6/STAT3/Notch signaling.
        Am J Cancer Res. 2018; 8: 302-316
        • Su S.
        • Chen J.
        • Yao H.
        • Liu J.
        • Yu S.
        • Lao L.
        • Wang M.
        • Luo M.
        • Xing Y.
        • Chen F.
        • Huang D.
        • Zhao J.
        • Yang L.
        • Liao D.
        • Su F.
        • Li M.
        • Liu Q.
        • Song E.
        CD10(+)GPR77(+) cancer-associated fibroblasts promote cancer formation and chemoresistance by sustaining cancer stemness.
        Cell. 2018; 172: 841-856.e16
        • Herrera M.
        • Islam A.B.
        • Herrera A.
        • Martin P.
        • Garcia V.
        • Silva J.
        • Garcia J.M.
        • Salas C.
        • Casal I.
        • de Herreros A.G.
        • Bonilla F.
        • Pena C.
        Functional heterogeneity of cancer-associated fibroblasts from human colon tumors shows specific prognostic gene expression signature.
        Clin Cancer Res. 2013; 19: 5914-5926
        • Sukowati C.H.
        • Anfuso B.
        • Croce L.S.
        • Tiribelli C.
        The role of multipotent cancer associated fibroblasts in hepatocarcinogenesis.
        BMC Cancer. 2015; 15: 188
        • Kalluri R.
        • Weinberg R.A.
        The basics of epithelial-mesenchymal transition.
        J Clin Invest. 2009; 119: 1420-1428
        • Zeisberg E.M.
        • Potenta S.
        • Xie L.
        • Zeisberg M.
        • Kalluri R.
        Discovery of endothelial to mesenchymal transition as a source for carcinoma-associated fibroblasts.
        Cancer Res. 2007; 67: 10123-10128
        • Takahashi K.
        • Yamanaka S.
        Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors.
        Cell. 2006; 126: 663-676
        • Hill R.P.
        • Gledhill K.
        • Gardner A.
        • Higgins C.A.
        • Crawford H.
        • Lawrence C.
        • Hutchison C.J.
        • Owens W.A.
        • Kara B.
        • James S.E.
        • Jahoda C.A.
        Generation and characterization of multipotent stem cells from established dermal cultures.
        PLoS One. 2012; 7: e50742
        • Krause M.P.
        • Dworski S.
        • Feinberg K.
        • Jones K.
        • Johnston A.P.
        • Paul S.
        • Paris M.
        • Peles E.
        • Bagli D.
        • Forrest C.R.
        • Kaplan D.R.
        • Miller F.D.
        Direct genesis of functional rodent and human schwann cells from skin mesenchymal precursors.
        Stem Cell Reports. 2014; 3: 85-100
        • Raghu G.
        • Scholand M.B.
        • de Andrade J.
        • Lancaster L.
        • Mageto Y.
        • Goldin J.
        • Brown K.K.
        • Flaherty K.R.
        • Wencel M.
        • Wanger J.
        • Neff T.
        • Valone F.
        • Stauffer J.
        • Porter S.
        FG-3019 anti-connective tissue growth factor monoclonal antibody: results of an open-label clinical trial in idiopathic pulmonary fibrosis.
        Eur Respir J. 2016; 47: 1481-1491
        • Hutchenreuther J.
        • Leask A.
        Why target the tumor stroma in melanoma?.
        J Cell Commun Signal. 2018; 12: 113-118
        • Perbal B.
        The concept of the CCN protein family revisited: a centralized coordination network.
        J Cell Commun Signal. 2018; 12: 3-12
        • Falke L.L.
        • Leeuwis J.W.
        • Lyons K.M.
        • Mummery C.L.
        • Nguyen T.Q.
        • Goldschmeding R.
        CCN2 reduction mediates protective effects of BMP7 treatment in obstructive nephropathy.
        J Cell Commun Signal. 2017; 11: 39-48
        • Petrosino J.M.
        • Leask A.
        • Accornero F.
        Genetic manipulation of CCN2/CTGF unveils cell-specific ECM-remodeling effects in injured skeletal muscle.
        FASEB J. 2019; 33: 2047-2057
        • Liu S.
        • Thompson K.
        • Leask A.
        CCN2 expression by fibroblasts is not required for cutaneous tissue repair.
        Wound Repair Regen. 2014; 22: 119-124
        • Liu S.
        • Taghavi R.
        • Leask A.
        Connective tissue growth factor is induced in bleomycin-induced skin scleroderma.
        J Cell Commun Signal. 2010; 4: 25-30
        • Tsang M.
        • Leask A.
        CCN2 is required for recruitment of Sox2-expressing cells during cutaneous tissue repair.
        J Cell Commun Signal. 2015; 9: 341-346
        • Quensel K.
        • Shiwen X.
        • Hutchenreuther J.
        • Xiao Y.
        • Liu S.
        • Peidl A.
        • Naskar D.
        • Siqueira W.L.
        • O'Gorman D.B.
        • Hinz B.
        • Stratton R.J.
        • Leask A.
        CCN1 expression by fibroblasts is required for bleomycin-induced skin fibrosis.
        Matrix Biol Plus. 2019; 3: 100009
        • Katoh M.
        • Nakagama H.
        FGF receptors: cancer biology and therapeutics.
        Med Res Rev. 2014; 34: 280-300
        • Mamun M.A.
        • Mannoor K.
        • Cao J.
        • Qadri F.
        • Song X.
        SOX2 in cancer stemness: tumor malignancy and therapeutic potentials.
        J Mol Cell Biol. 2018; ([Epub ahead of print] doi:10.1093/jmcb/mjy080)
        • Pelizzo G.
        • Veschi V.
        • Mantelli M.
        • Croce S.
        • Di Benedetto V.
        • D'Angelo P.
        • Maltese A.
        • Catenacci L.
        • Apuzzo T.
        • Scavo E.
        • Moretta A.
        • Todaro M.
        • Stassi G.
        • Avanzini M.A.
        • Calcaterra V.
        Microenvironment in neuroblastoma: isolation and characterization of tumor-derived mesenchymal stromal cells.
        BMC Cancer. 2018; 18: 1176
        • Pitrone M.
        • Pizzolanti G.
        • Tomasello L.
        • Coppola A.
        • Morini L.
        • Pantuso G.
        • Ficarella R.
        • Guarnotta V.
        • Perrini S.
        • Giorgino F.
        • Giordano C.
        NANOG plays a hierarchical role in the transcription network regulating the pluripotency and plasticity of adipose tissue-derived stem cells.
        Int J Mol Sci. 2017; 18: E1107
        • Riekstina U.
        • Cakstina I.
        • Parfejevs V.
        • Hoogduijn M.
        • Jankovskis G.
        • Muiznieks I.
        • Muceniece R.
        • Ancans J.
        Embryonic stem cell marker expression pattern in human mesenchymal stem cells derived from bone marrow, adipose tissue, heart and dermis.
        Stem Cell Rev Rep. 2009; 5: 378-386

      Linked Article

      • This Month in AJP
        The American Journal of PathologyVol. 190Issue 1
        • Preview
          The following highlights summarize research articles that are published in the current issue of The American Journal of Pathology.
        • Full-Text
        • PDF
        Open Archive