TGF-β is believed to play a crucial role in promoting fibrotic responses. Three isoforms of TGF-β exist, and in the context of fibrosis most of our knowledge is centered on TGF-β1 [61, 62]. TGF-β affects the behavior of a number of cellular types either of hematopoietic, mesenchymal, or epithelial origin. TGF-β is a powerful immunosuppressant, regulates inflammation, variably promotes or suppresses tumor growth, and is essential in wound healing [63]. The biological activities of TGF-β are mostly regulated by modifications of the released protein rather than by modifications of its synthesis. Indeed, TGF-β exists in a latent complex bound to the ECM. Following tissue injury, TGF-β is released from ECM and influences the cells in the immediate vicinity. Within the ECM, mature TGF-β exists as a dimer, forming a small latent complex with the latency-associated peptide (LAP), itself bound to a larger complex directly bound to ECM proteins. To become biologically active, the latent TGF-β complex needs to be activated by proteolytic mechanisms or by tractional forces mediated by integrins – most notably, the αVβ6, αVβ5, and αVβ3 integrins [64]. Following receptor binding, TGF-β implicates canonical and noncanonical signaling pathways. The SMAD pathway implicates the phosphorylation of SMAD2 and SMAD3 proteins and their association with the co-SMAD4 for nuclear translocation. This signaling cascade is negatively regulated by SMAD7 [65]. In fibrosis development, the noncanonical TGF-β signaling pathways may play a major role and involve MAPK, PI3K, and NFkB pathways as well as NADPH oxidase 4 (NOX4)/reactive oxygen species (ROS) pathways [66]. Key to avoid excessive ECM deposition, TGF-β signaling needs to be terminated once tissue is repaired and the synthesis of extracellular matrix should return to normal levels. Finely tuned homeostatic mechanisms are thought to be at play, but the molecular mechanisms that underlie the failure to limit TGF-β activity are poorly understood. However, recent work has highlighted the role of the orphan nuclear receptor NR4A1 in downregulating TGF-β signaling and participating to a negative feedback loop that limits excessive TGF signaling and could be exploited to attenuate fibrotic disorders [67].

PDGF signals through two different PDGF receptors, alpha and beta, and comprises a family of homo- or heterodimeric growth factors including PDGF-AA, PDGF-AB, PDGF-BB, PDGFCC, and PDGF-DD [68]. After wounding, the expression of PDGF is increased; this growth factor participates in the recruitment of neutrophils, macrophages, fibroblasts, and smooth muscle cells into the wound [69]. PDGF also stimulates the formation of granulation tissue with direct effects on fibroblasts, resulting in collagen matrix contraction and fibroblast differentiation into myofibroblasts in vitro [70]. Tyrosine kinase inhibitors such as imatinib mesylate, dasatinib, and nilotinib exert potent anti-fibrotic effects in experimental models of fibrosis, most likely by inhibiting signaling induced by PDGF-receptor engagement [71].

CTGF is a member of the CCN proteins that are key signaling and regulatory molecules involved in cell proliferation, angiogenesis, tumorigenesis, wound healing, and fibrosis [72]. CTGF is a modular matricellular component produced by fibroblasts (among other cells) and induced in response to various stimuli including TGF-β, endothelin-1, and angiotensin-II and is thought to act mainly by favoring an environment which promotes fibrogenesis [73]. It acts mostly in an autocrine manner interacting with other matrix components and adhesion molecules [74]. It has been implicated in several fibrotic conditions including scleroderma, chronic heart failure, and proliferative diabetic retinopathy.

PAI-1 is a member of the serine protease inhibitor (serpin) gene family and the major physiologic inhibitor of the serine proteases urokinase plasminogen activator (uPA) and tissue plasminogen activator (tPA). Inhibition of uPA/tPA results in the inhibition of plasminogen-to-plasmin conversion as well as plasmin-dependent MMP activation. Thus, PAI-1 protects ECM proteins from proteolytic degradation and helps accelerate wound healing. However, sustained PAI-1 activation may contribute to excessive collagen accumulation [75]. Fibroblasts obtained from fibrotic tissues spontaneously produce higher amounts of PAI-1 compared to control fibroblasts, and inhibition of PAI-1 protects against fibrosis in the lung [76], heart, and kidney but not the skin [77] in animal models of fibrosis. Factors that may induce high production of PAI-1 are VEGF and TGF-β1. Thus, PAI-1 participation to fibrosis development exemplifies the role of an excessive brake directed against events aimed at ECM reabsorption.

Classical Th2 cell products, such as IL-4 and particularly IL-13, have been shown to have direct and indirect roles in fibrosis development. IL-4 and IL-13 enhance type I collagen production in dermal fibroblasts [78, 79]. IL-13 in vivo acts by inducing TGF-β production by macrophages and directly stimulating myofibroblast and fibroblast synthetic activities. IL-13 function is regulated by the relative expression of the two IL-13 receptor subunits, IL-13Rα1 (signaling component) and IL-13Rα2 (decoy component). When IL-13Rα2 levels are low, fibrotic responses are enhanced [80]. Consistent with the importance of Th2-like response in fibrotic pathologies, the report shows increased eotaxin/CXCL13 serum levels in idiopathic retroperitoneal fibrosis [81].

IL-6 is rather a newcomer in the fibrosis field, whose importance has been highlighted by the availability of targeted biological therapies. IL-6 is a pleiotropic cytokine with direct pro-angiogenic and pro-fibrotic activities [82]. Its levels are increased in the serum and skin of scleroderma patients [83], and its inhibition attenuates bleomycin-induced experimental fibrosis in the murine lung and skin [84, 85]. Of importance, IL-6 blockade has shown beneficial effects in humans affected by idiopathic retroperitoneal fibrosis [86].

Historically, TNF has been associated with the development of fibrosis in experimental models dominated by inflammation. In a pioneering work, Piguet and coworkers demonstrated that TNF blockade strongly attenuated the fibrotic response in the mouse with bleomycin-induced lung fibrosis [87]. However, when TNF inhibitors have become available in the clinics, their use in humans was not always associated with fibrosis halting; conversely, in some cases worsening of the fibrotic process was observed, particularly in individuals suffering from systemic sclerosis [88]. On the one hand, TNF is particularly important among inflammatory cytokines because it enhances the recruitment of inflammatory cells which mandatorily precede fibrosis; on the other hand, TNF inhibits the transcription of type I and type III procollagen mRNA [89, 90] and is a strong inducer, in fibroblasts and macrophages, of MMP production, in particular of the interstitial collagenase MMP1, which participates in ECM degradation [91] [92]. Thus, the net effect on fibrosis development of TNF may very well depend on the context and timing in which it operates, variably resulting in enhanced or decreased ECM deposition.