Enter the Matrix: Fibroblast-immune cell interactions shape extracellular matrix deposition in health and disease.

ECM-based regulation of fibroblast-macrophage interactions

Fibroblasts produce cytokines and chemokines which bind the ECM and are critical for immune cell recruitment and function in steady-state and disease^2,16 ( Figure 1). For example, multiple in vitro and tissue-restricted in vivo studies have demonstrated that a relationship exists between fibroblasts and macrophages, an innate immune cell which orchestrates long-lived adaptive immune responses through phagocytosis, antigen presentation, and immunological mediator secretion.¹⁶ Fibroblasts and macrophages are found in almost all tissues and multiple studies indicate that fibroblast-derived CSF1, which can decorate the ECM, is a critical factor for macrophage survival. Using in vitro studies, Zhou et al. demonstrated that macrophages and fibroblasts form stable cell circuits that are resistant to perturbations and that cell-to-cell contact increased cell circuit survival because of local exchange of growth factors, including CSF1.¹⁷ In addition, in vivo studies have demonstrated that fibroblastic reticular cells (FRC) are required to establish the lymph node (LN) macrophage niche. D’Rozario et al. demonstrated that genetic ablation of FRC using a genetic tool in which Chemokine (C-C motif ) ligand 19 (Ccl19)-expressing cells could be depleted using diphtheria toxin resulted in rapid loss of monocytes and macrophages from LN in two separate in vivo models.¹⁸ Moreover, single-cell RNA sequencing (scRNA-seq) of murine brachial lymph nodes revealed that FRC subsets broadly expressed master macrophage regulator CSF1 and functional assays containing purified FRC and monocytes showed that CSF1R signaling was sufficient for macrophage survival.¹⁸ Comparative analysis demonstrated these effects were conserved between mice and humans.¹⁸

Figure 1. Fibroblast influence macrophage accumulation by altering the composition of the ECM.

Fibroblasts use several mechanisms to recruit macrophages to the ECM. Fibroblast-derived cytokines and chemokines bind the ECM and enhance accumulation of ECM-associated macrophages. Fibroblast contraction remodels collagen fibers to enhance macrophage recruitment. Metabolic factors alter macrophage population dynamics and function. Fibroblasts secrete MMPs to cleave ECM- and ECM bound proteins to enhance macrophage accumulation in the ECM.

Monocytes, which can give rise to macrophages, have also been shown to require stromal cell derived CSF1. Emoto et al. demonstrated that CSF1 produced by sinusoidal endothelial cells was required for the survival of Ly6C- monocytes. Conversely, CSF1 produced by both endothelial and Lepr+ perivascular stromal cells were required for the survival of Ly6C^hi monocytes.¹⁹ Taken together, these studies demonstrated that macrophage survival in specific tissues is regulated by stromal-derived CSF1. While all fibroblasts can generate CSF1, tissue-specific fibroblast subsets do exhibit different levels of CSF1 mRNA,²⁰ so additional studies are required to delineate which subsets are critical for macrophage homeostasis.

Further investigation is required to determine the functional role of macrophages sustained by fibroblast-derived CSF1 in steady-state and disease. Previous work by Zhu et al. has demonstrated that a pharmacologically induced CSF1 blockade in mouse pancreatic tumour models results in increased antigen presentation and productive antitumour T cell responses.²¹ Despite this, the precise contribution of fibroblast-derived CSF1 in antitumour immunity is not well understood. Furthermore, the functional contributions of these macrophages in chronic inflammation and cancer also require subsequent study.

In addition to cytokines, activated fibroblasts, including myofibroblasts, produce ECM-binding macrophage chemoattractants indicating that fibroblasts are capable of recruiting monocytes and macrophages during injury, infection, and inflammation.^22,23 For example, the bacterial ligand lipopolysaccharide (LPS), which activates toll-like receptor (TLR)4, has been shown to enhance expression of macrophage chemoattractants Ccl2 and Ccl4 in myofibroblast progenitor cells in a mouse model of liver fibrosis.²² As LPS-mediated TLR4 activation occurs widely in anti-bacterial immunity, this suggests that fibroblasts can respond to pathogen-associated molecular patterns (PAMP) to recruit immune cell subsets, including macrophages, in a variety of anti-infective contexts. However, the role of fibroblast-mediated macrophage recruitment in additional anti-infective immune responses, as well as chronic inflammation, requires future investigation.

Fibroblasts influence macrophage migration through mechanical cues. Previous studies have also demonstrated that fibroblasts can influence macrophage dynamics by regulating the physical mechanics of the ECM ( Figure 1). Hinz and colleagues showed that contractile myofibroblasts drive mechanical cues through local remodeling of collagen fibers resulting in increased macrophage migration directly towards myofibroblasts using 3D collagen gels in vitro.²⁴ Migration occurred independently of chemotaxis and required macrophage-mediated attachment to collagen via the α2β1 integrin and stretch-sensitive ion channels.²⁴ Similar observations have been described in several fibroblast-like cells,^25–27 indicating the importance of these biomechanical mechanisms for cell communication and movement across tissues.

Fibroblast-derived matrix metalloproteinases degrade ECM proteins to enhance macrophage recruitment. Matrix metalloproteinases (MMPs) are a family of enzymes which degrade ECM proteins. These enzymes are either secreted or attached to cell surfaces, confining their activity to membranes, secreted proteins, or proteins within the extracellular space.²⁸ MMPs have a complex role in regulating inflammation by acting on a multitude of immunological mediators, including antimicrobial host defense peptides, cytokines, chemokines, and ECM proteins as reviewed in detail here.²⁸ In addition, matrix metalloproteinases are enhanced in stromal cells, including fibroblasts, in response to pro-inflammatory cytokines.^29–31 As a result, modulation of the ECM during inflammation alters immune cell migration and ECM resident cell types, including macrophages ( Figure 1). For example, Shubayev et al. demonstrated that TNF-α-mediated MMP9 (also known as gelatinase B) production enhances macrophage recruitment in a model of peripheral nerve injury.³² Gong et al. demonstrated that mice deficient for the ECM-bound protease plasminogen (Plg) had decreased trans-ECM macrophage migration and decreased MMP9 activation.³³ In addition, the authors demonstrated that MMP9 administration to Plg deficient mice resulted in increased macrophage accumulation in the ECM,³³ suggesting that Plg activates MMP9 to increase ECM resident macrophages. Moreover, Tan et al. demonstrated that MMP9 can also cleave the ECM protein osteopontin (OPN) to enhance macrophage recruitment in a mouse model of renal fibrosis.³⁴ Despite this, additional studies are required to understand how MMP-mediated degradation of other ECM components alters immune cell recruitment and fibrotic disease development.

In addition to cytokines, mechanical cues, and enzymatic means, fibroblast can regulate immune cell regulation through metabolic cues. Previous studies have demonstrated that metabolic factors are required for continuing collagen deposition during fibrotic disease. Schworer et al. demonstrated that the metabolic enzyme pyruvate carboxylase (PC) is required anaplerosis in tumour-associated fibroblasts for continuing collagen deposition in the glutamine and glucose scare tumour microenvironment (TME).³⁵ Kay et al. also demonstrated that collagen deposition requires de novo synthesis of the amino acid proline, a major component of multiple collagens, from glutamine, by the enzyme pyrroline-5-carboxylate reductase 1 (PYCR1).³⁶ Despite these findings, the impact that this had on immune cell, and specifically macrophage, infiltration into the TME was not investigated. To this end, some studies have investigated the interplay between metabolic factors and macrophage accumulation in various tissues. Buechler et al. demonstrated that Wt1+ stromal cells produce retinoic acid (RA) to maintain a population of GATA6+ Large Cavity Macrophages (LCM) in the peritoneal, pleural, and pericardial spaces.³⁷ Similarly, Pucino et al. demonstrated that physiological concentrations of lactate in RA joints induces the IL-6 production in fibroblasts,³⁸ which may enhance the differentiation of monocytes into macrophages.³⁹ Taken together, these studies provide the rationale for future investigations on how metabolic factors alter the relationship between fibroblasts, ECM deposition, and macrophage accumulation.

Fibroblast-derived CTHRC1 as a regulator of ECM deposition

Collagen triple helix repeat containing 1 (CTHRC1), a secreted ECM protein, has recently been identified as a critical modulator of ECM protein modulation and wound healing.^54,55 Human CTHRC1 contains a N-terminal 30 amino acid hydrophobic signal peptide secretory domain, a short collagen triple helix repeat (CTHR) domain consisting of 12 repeats of the Gly-X-Y motif, and a highly conserved C-terminal domain with similar structure of the globular C1q domain of Collagen VIII.^56,57 CTHRC1 is expressed in macrophages, myofibroblasts, endothelial cells as well as mesenchymal-derived cells⁵⁸ and as a result, CTHRC1 expression occurs in multiple tissues, including the bone,⁵⁹ the lung,⁶⁰ and tumours.⁶¹

Previous studies suggest that CTHRC1 modulates TGF-β signaling ( Figure 3A). For example, Cthrc1 modulates TGF-β signaling by promoting the degradation of canonical signaling intermediates, including Smad2/3, and has been suggested to influence Wnt and b-integrin signaling to affect cancer cell proliferation and metastasis.⁵⁸ Guo et al. showed that Cthrc1 modulated β-integrin signaling by enhancing phosphorylation of Focal adhesion kinase (FAK) to promote metastasis of epithelial ovarian cancer (EOC) cells.⁶² In fibroblast and smooth muscle cells, CTHRC1 levels are associated with increased cell migration in vitro.⁶³ Other reports suggest that CTHRC1 may negatively regulate TGF-β-mediated signaling pathways. For example, Leclair et al. demonstrated that overexpression of CTHRC1 in smooth muscle cells reduced levels of phosphorylated Smad2/3, which is required for TGF-β-mediated gene expression and collagen production.⁶⁴ As Cthrc1 expression is enhanced in response to TGF-β in vitro,⁶³ mechanistic evidence suggests that TGF-β and CTHRC1 may act in a negative-feedback loop to limit TGF-β-induced ECM deposition. Due to conflicting reports on the role of CTHRC1 in a disease context, additional studies are required. Future queries delineating the precise relationship between CTHRC1 in TGF-β-mediated signaling and the role of CTHRC1 in a disease context must focus on cell- and tissue-specific mechanisms.

Figure 3. Proposed mechanisms of CTHRC1- and BIGH3-mediated ECM modulation.

(Left) TGF-β signaling enhances production of extracellular matrix proteins, as well as CTRHC1, resulting in negative feedback loop limiting its own production. Secreted CTHRC1 can bind to collagens and integrins and has been implicated in cancer cell proliferation and migration by influencing Wnt and integrin β signaling pathways. (Right) TGF-β signaling enhances BIGH3 production which binds integrin receptors. Proposed pathways of BIGH3-mediated collagen production include activation of Snail and/or PI3K, transcription factors which influences collagen production as well as suppression of MMP14, which promotes collagen turnover. Legend: dashed lines represent prospective/predicted interactions.

CTHRC1 has been shown to mark lung myofibroblasts in the context of models of lung fibrosis in murine systems and idiopathic pulmonary fibrosis (IPF) in humans.^{20,60,65–67} Cthrc1 expression may also distinguish two disease-associated myofibroblast clusters across tissues.²⁰ One disease-associated myofibroblast cluster is marked by the simultaneous expression of Lrrc15 (leucine-rich repeat 15) and Cthrc1 expression, whereas the other is only marked by Cthrc1 and not Lrrc15 expression.²⁰ scRNA-seq-based characterization of collagen-producing cells in fibrotic mouse lungs demonstrated that the Cthrc1+ fibroblast cluster expresses the highest levels of ECM proteins, including Col1a1 and Col3a1.⁶⁰ Tsukui et al. used transgenic mice to track and ablate Cthrc1-expressing fibroblasts in lung fibrosis. Lineage tracing using Ctrhrc1-CreER mice crossed to Rosa26-tdTomato mice demonstrated that Cthrc1-expressing lung fibroblasts expand in a mouse model of bleomycin-induced pulmonary fibrosis. Cthrc1-Cre-ER-labeled cells showed elevated expression of ECM genes, including Col1a1 and Postn (periostin). Further, depletion of Cthrc1-expressing lung fibroblasts using Cthrc1-CreER mice crossed to Rosa26-tdTomato and Diphtheria toxin subunit A expressing mice attenuated hydroxyproline, a marker of collagen deposition following administration of bleomycin.⁶⁶ Taken together, these results suggest that Cthrc1 expressing cells drive ECM deposition and may promote fibrosis.

Conversely, different studies have demonstrated that CTHRC1 may limit ECM deposition and fibrosis. Global Cthrc1 knock-out (KO) resulted in increased hydroxyproline levels and tissue remodeling in the lungs of mice following bleomycin administration.⁶⁵ In a mouse model of vascular fibrosis, Cthrc1 was transiently expressed by adventitial fibroblasts following injury and was associated with decreased collagen deposition in cross-sections of carotid arteries.⁶³ The discrepancies between these studies may highlight the cell-, tissue-, and disease-specific role of Cthrc1. As such, further studies are needed to determine the precise role of CTHRC1 in diverse fibrotic diseases. In addition, specific genetic tools, including fibroblast-specific Cthrc1 KO mice or fluorescently-labelled Cthrc1-expressing cells are required to properly assess the pathogenicity of Cthrc1+ myofibroblasts in different tissues and disease models.

BIGH3 is associated with increased collagen deposition

Transforming growth factor beta inducible protein (TGFBI), is also known as keratoepithelian,⁶⁸ and is referred to here as beta-inducible growth hormone 3 (BIGH3) is produced by fibroblasts and macrophages.⁶⁹ BIGH3 was discovered in 1992 as a gene highly upregulated by A594 lung adenocarcinoma cells following TGF-β stimulation.⁷⁰ The structure of BIGH3 has not been solved; however, early characterization of this protein using cDNA sequence analysis suggested that it contains a secretory signal peptide, four fasciclin 1 (FAS1) domains that mediate binding to extracellular matrix proteins, such as collagens, and a RGD motif that enables the binding of this protein to integrin receptors.^70–72 Due to structural similarities, BIGH3 has been hypothesized to be a paralog of Periostin (Postn),⁷³ which drives collagen production in the presence of TGF-β.⁷⁴ BigH3 has been shown to be expressed by fibroblasts in response to stimulation with TGF-β⁷⁵ and by macrophages in ‘M2’ polarization conditions in vitro^60,61,76 or following ingestion of apoptotic cells.⁶⁹ Upon secretion by fibroblasts and macrophages, BIGH3 binds ECM components, such as collagens,⁷⁷ and subsequently binds the integrin a_vb₅⁷⁸ and promotes adhesion and migration of mesenchymal and ectodermal-lineage cells, including fibroblasts, chrondrocytes, osteoblasts, keratinocytes, and endothelial cells.^79–81

Studies investigating the role of BIGH3 in disease indicated tissue-specific functions. In the central nervous system, Peng et al. demonstrated that BigH3 is elevated in the serum and cerebrospinal fluid (CSF) of glioblastoma patients and that TGFBI-expressing macrophages promoted the survival of glioma stem cells by avB5-Src-Stat3 signaling.⁷⁸ In the heart, Schwanekamp et al. demonstrated that BigH3 was enhanced in response to myocardial infarction. Despite this, cardiac-specific deletion of BigH3 did not alter cardiac disease or fibrosis post-myocardial infarction.⁸² In the colon, TGF-β signalling plays a major role in promoting fibrotic lesions in inflammatory bowel diseases (IBD).⁵³ Aligned with this BigH3, as well as ECM genes COL1A1, COL1A2, and COL3A1 are upregulated in patients with fibrotic IBD, suggesting that combined BIGH3 and TGF-β signalling may accelerate colon fibrosis.^53,83 In the lungs, studies by Yang et al. and Xu et al. have independently demonstrated that BIGH3 is elevated in patients with interstitial lung disease.^84,85 Ahlfeld et al. demonstrated that BigH3 is elevated in response to lung injury and BigH3 KO mice at early ages had documented lung developmental abnormalities, lack of elastin-positive tips, reduced proliferation, and abnormally persistent aSMA myofibroblasts, which resolve by adulthood.⁸⁶ The authors also demonstrated that lungs in BigH3 deficient mice had reduced elastic recoil and gas exchange efficiency.⁸⁶ It remains to be seen if adult animals exhibit these defects or compensatory effects mask phenotypes seen in younger animals.

In addition to impacting lung development, in vitro studies have demonstrated that BIGH3 can promote collagen deposition in lung fibroblasts.⁸⁷ Merl-Pham et al. demonstrated that TGF-β enhances the production of BigH3 in primary human lung fibroblasts.⁸⁸ Yang et al. found that siRNA-mediated knockdown (KD) of BigH3 in human lung fibroblast cultures resulted in decreased TGF-β-mediated collagen 1 (Col 1) and aSMA production, demonstrating that BIGH3 is necessary to produce these ECM proteins,⁸⁷ potentially through autocrine stimulation. Despite these studies, the molecular mechanism by which BIGH3 mediates Col1a1 deposition downstream of TGF-β is unclear. It has been posited based upon in vitro studies using lung fibroblast cell lines that BIGH3 mediates its effects via a multi-part mechanism. Briefly, TGF-β first elicits BIGH3 expression, then BIGH3 binds a multitude of integrin receptors, including a₁b₁,⁸⁹ a₃b₁,^90,91 a_vb₃⁷⁹ or a_vb₅⁷⁹ integrin. BigH3 then drives activation of PI3K-signaling⁹¹ and increased Snail expression, encoded by Snail1, via downregulation of a Snail negative regulator G-protein signaling modulator 2 (GPSM2).^79,87 Taken together, these studies revealed that BIGH3 is both necessary and sufficient to induce collagen production in lung fibroblasts. Therefore, mechanistic studies indicate that BIGH3 modulates TGF-β-mediated signal transduction pathways ( Figure 3B).

Alternative mechanisms for BIGH3-mediated regulation of collagen deposition in the lung have also been proposed. Zhang et al. demonstrated that BigH3 deficient mice had stunted growth in vivo, as well as enhanced proliferation and cyclin D1 expression ex vivo, suggesting that BIGH3 may limit cellular proliferation.⁹² Nacu et al. demonstrated that primary lung fibroblasts stimulated with BIGH3 had increased Col 1 protein abundance in the absence of transcription, and subsequently demonstrated that BigH3 suppresses the transcription of MMP14, which degrades collagen.⁶⁹ These results suggest that BIGH3 reduces collagen turnover by decreasing ECM degradation. As a result of these conflicting reports, further study is required to characterize the full scope of ECM deposition changes in response to BIGH3. In addition to studies defining the tissue-specific nature of BIGH3 and the mechanisms of its effect on ECM deposition, further investigation is required into additional roles of BIGH3 in the context of fibroblasts and macrophage spatiality, including mechanical sensing, and/or monocyte/macrophage chemotaxis. As such, cell- and tissue- restricted BIGH3 KO models paired with high-resolution techniques, such as scRNA-seq are required to examine the role that BIGH3 plays in modulating ECM deposition, development, and disease.