The majority of the fibrosis-related extracellular factors are receptor-binding ligands, such as growth factors and cytokines. These factors target adjacent and distant cells in autocrine, paracrine, or endocrine signaling pathways. Then they bind to specific receptors on cell membrane and trigger the intracellular signaling, leading to pro-fibrotic cellular responses. Other extracellular factors, mainly enzymes such as matrix metalloproteinases (MMPs) could degrade ECM to prevent its excessive accumulation.

Growth factors contain a huge family of proteins that stimulate cell growth and proliferation. They are secreted by fibroblasts, immune cells, and epithelial/endothelial cells, and are able to orchestrate cellular responses. While epithelium/endothelium are damaged, cells in these tissues massively upregulate the production of growth factors to promote the proliferation of immune cells and fibroblasts. Among growth factors, TGF-β is the “master” modulator in fibrogensis (Meng et al., 2016), as it could provoke fibrosis through SMAD-dependent pathway (Lan, 2011) and SMAD-independent pathway related to a number of other pro-fibrotic reactions (Zhang, 2009). TGF-β signaling cascade results in differentiation of effector cells via inducing the expression of myofibroblasts hallmark α-hallmar muscle actin(α-SMA) (Sebe et al., 2008). In addition, TGF-β signaling leads to the transcription of collagen I and III genes (Fine and Goldstein, 1987; Chen et al., 1999) contributing to ECM accumulation. Interacting with TGF-β, many other growth factors have distinct roles in pro-fibrotic process. For example, PDGF induces HSC proliferation and type I collagen expression via downstream focal adhesion kinase/phosphoinositide 3-kinase/protein kinase B signaling (Reif et al., 2003). On the contrary, some growth factors have anti-fibrotic property, such as hepatocyte growth factor (HGF), the overexpression of which alleviates fibrosis in cardiomyopathic hamster through activation of MMP-1 and urokinase-type plasminogen activator (Taniyama et al., 2002). Excessive cytokines are usually secreted by immune cells, such as macrophages, neutrophils, and T cells in inflammation phase of wound healing. Compared with growth factors, cytokines tend to act as cell signaling transmitters to augment immunological responses and then lead to inflammation. One major type of cytokines is chemokines, which guide the recruitment of immune cells and fibroblasts to injury sites. Different immune cells are recruited by different chemokines. Neutrophils could be recruited by chemokine (C-X-C motif) ligand 1 and chemokine (C-X-C motif) ligand 8 by binding glycosaminoglycans with receptors in slightly different ways (Sawant et al., 2016). Chemokine (C-C motif) ligand 5 has been a major factor to induce the migration of HSC in liver fibrosis (Seki et al., 2009). Another type of cytokines is T cell cytokines, which are mainly secreted by activated T lymphocytes. They include interleukins, interferons and tumor necrosis factors (TNF), mediating adaptive immune responses and inflammation that might promote fibrosis. For example, IL-6 has been reported to shift the tissue repair to a chronic inflammatory state by signal transducer and activator of transcription 3(STAT3) signaling pathway in peritoneal fibrosis (Fielding et al., 2014). In contrast, some cytokines have anti-fibrotic effects. It has been reported that interferon-γ down-regulated the Adenosine A2A receptor signaling to prevent the production of type I collagen in HSC (Block and Cronstein, 2010). Moreover, some T cell cytokines, such as TNF-α, exhibit two-sided effects, pro-fibrotic or anti-fibrotic effect depending on the alternative status of macrophages and micro-environment (Redente et al., 2014).

MMPs are the extracellular endopeptidases degrading ECM including collagens, proteoglycans, laminins, and fibronectin. Tissue inhibitor of metalloproteinases (TIMPs) work as the inhibitors of MMPs. The balance of MMPs and TIMPs modulates the process throughout fibrosis development, including the formation of multiple cell injuries, the activation of latent cytokines and myofibroblasts and mainly, the maintenance of the homeostasis of ECM (Giannandrea and Parks, 2014). Some MMPs have pro-fibrotic functions whereas some have anti-fibrotic according to cell types and phases. The dysregulation of MMP-19 has been proved to cause the degradation of normal liver ECM and initiate liver injury (Jirouskova et al., 2012). Conversely, MMP-2 has been reported to cleave type I collagen and attenuate collagen deposition by HSC, inhibiting liver fibrosis (Radbill et al., 2011). As to TIMPs, they could inhibit or activate fibrosis via MMPs. For example, TIMP-3 inhibits MMPs to induce inflammation (Gill et al., 2010) in lung injury, and TIMP-1 has been found to play a dual role in liver fibrosis (Wang H. et al., 2011).

Other extracellular factors include a wide range of proteins and peptides. They mainly guide the differentiation of myofibroblasts, and are closely related to each other and growth factors. For example, the hedgehog (Hh) signaling pathway mediates EMT during the fetal development, and responds to injury through the repression of epithelial marker epithelial-cadherin by Snail and Twist. Overactivation of Hh signaling pathway contributes to biliary fibrosis and related liver fibrosis (Omenetti et al., 2008). In fibrotic kidney, the upregulated Wnt signaling has been reported to result in abundant of β-catenin. The signaling regulates genes such as Twist, LEF1 to induce EMT, thus to aggravate disease (He et al., 2009).

Intracellular factors, mainly multiple kinases, propagate the signaling received by cells through phosphorylation and other pathways. A common consequence of the signaling is that, transcription activators or inhibitors translocate into nucleus to regulate fibrosis-related gene expression and cell responses. In certain inflammatory pathways, intracellular factors also modulate the expression of extracellular factors, such as growth factors and cytokines, and secrete them out of cells to amplify inflammatory responses. Besides, epigenetic factors are emerging as a new way to affect fibrosis-related gene expression.

Intracellular factors include a number of kinases. In inflammation phase, many upstream factors, such as TGF-β, TNF-α, and epidermal growth factors initiate mitogen-activated protein kinase (MAPK) pathway. An element of MAPK pathway, mitogen-activated protein kinase-activated protein kinase-2 mediates myofibroblasts differentiation and regulates the gene expression of several matrix proteins such as col1a2, col3a1, and lox (Vittal et al., 2013). Another important intracellular signaling, mechanistic target of rapamycin (mTOR) pathway is activated by Wnt and TNF-α. Then, mTOR pathway activates ribosomal protein S6 kinase β-1 and modulates protein p21 or p27, which regulates the cell cycle of many cells including fibroblasts. In addition, the inhibition of mTOR pathway reduces collagen deposition and cardiac fibrosis (Chen et al., 2012). Besides serving as amplifiers of signal transduction, some intracellular enzymes such as cathepsin K, which belongs to lysosomal cysteine proteases, have been proved to be able to degrade ECM in lysosome after phagocytosis (Fukumori et al., 2003; Buhling et al., 2004).

Nuclear receptors are receptors located in cytoplasm and nucleus that could receive signals from intracellular ligands and bind to DNA to regulate gene expression. For example, peroxisome proliferator activated receptor γ(PPAR-γ), may directly regulate type I collagen gene (Yang et al., 2006) and block TGF-β signaling (Ghosh et al., 2009). Another nuclear receptor, farnesoid-X receptor (FXR), exhibits anti-fibrotic effect via the reduction of proliferating cholangiocytes and subsequent reduction of TGF-β (Liu et al., 2003). The activation of FXR also decreases a series of pro-fibrotic factors including TIMP-1, collagens, α-SMA, and MMP-2 (Zhang et al., 2009).

Nowadays, epigenetics including microRNAs, DNA methylation and lncRNAs, are found involved in machinery of pro-fibrotic process mainly through regulating fibrosis-related gene expression. Some microRNAs are found to negatively regulate translation of ECM components. Among them, miR-21 induces extracellular-signal regulated kinase/MAPK activity via the inhibition of Spry1 to protect cardiac fibroblasts survival (Thum et al., 2008). It has been reported that the levels of DNA methylation at specific CpG sites of pro-fibrotic genes (PPARα, PPARδ, TGFβ1, Collagen1A1, and PDGFα) differ among different fibrosis stages in NAFLD (Zeybel et al., 2015).

There are factors that affect fibrosis process both extra- and intracellularly, including reactive oxygen species (ROS), galectin-3 and lysyl oxidase homolog 2 (LOXL2). ROS can be generated through tissue injuries, cell damages and NADPH oxidase activities. Extracellular ROS targets latency-associated peptides and then activates TGF-β signaling while intracellular oxidative stress induces p53-dependent apoptosis in lung fibrosis via the caspases-9/3 activation in mitochondria (Cheresh et al., 2013). As to galectin-3, extracellular galectin-3 induces T cell apoptosis and plays a dual function inside and outside cells (Li et al., 2014). Another factor LOXL2 is generally considered as extracellular enzyme that promotes collagen production and crosslink with collagen fibers in response to mechanical stress (Yang et al., 2016). While on the other hand, intracellular LOXL2 has been reported to induce EMT in carcinoma progression (Peinado et al., 2005).

Many studies proved that different fibroproliferative diseases share common underlying mechanisms (Wenzke et al., 2012). The existence of common mechanisms facilitates the complete interpretation of fibrosis pathogenesis and enhances our understanding of fibrosis-related diseases. On this basis, it is necessary to reconsider targets involved in these mechanisms and evaluate their potential roles in fibrosis treatment across tissues and organs.

Nowadays, most approved and investigational drugs are single-component drugs, which only contain one organic component and have distinct targets. As we described previously, fibrosis progression results from a combination of the abnormalities of extracellular and intracellular factors. Drugs targeting extracellular factors are prevalent, about 60% of known targets are receptors located on cell membrane mainly because the extracellular targets are accessible and serve as upstream signals (Overington et al., 2006), and so are the targets of fibrosis-related drugs. The binding of receptors and ligands triggers the downstream signaling, thus the blockade of receptors or ligands is considered to be an effective choice to alleviate fibrosis (Table 1).

The majority of anti-fibrosis drugs targeting extracellular factors are inhibitors of ligands such as growth factors, cytokines and MMPs. Most inhibitors could directly bind to the active sites of targets. The majority of approved inhibitor drugs in Table 1 target TNF, which are widely used in fibrosis-related diseases and could suppress the action of TNF through multiple mechanisms. These drugs include small molecules Thalidomide and Pomalidomide (Weingartner et al., 2012), recombinant protein Etanercept and monoclonal antibodies Belimumab. Thalidomide and Etanercept have completed the phase II trial in IPF (Raghu et al., 2008; Horton et al., 2012) while Belimumab has completed the phase II trial in SSc. Besides, some inhibitor drugs targeting growth factors are still under investigation. For example, Disitertide, a synthetic peptide derived from TGF-β type III receptor, inhibits the binding of TGF-β and its receptor and exhibit anti-fibrotic function (Ezquerro et al., 2003). Drugs that inhibit interleukins are always monoclonal antibodies. Tralokinumab, a human IgG4 monoclonal antibody, shows pro-apoptotic effects via IL-induced apoptotic factors in IPF (Murray et al., 2014). Many natural products act as inhibitor drugs like Dioscin, which is a monomer extracted from Dioscoreae Rhizoma and could ameliorate liver fibrosis (Liu et al., 2015; Zhang et al., 2015a,b; Gu et al., 2016; Xu et al., 2017; Yin et al., 2017). In addition, some preclinical drugs inhibit MMPs, like Marimastat, which simultaneously down-regulates MMPs gene expression and MMPs activities. However, Marimastat reduces inflammation and liver injuries while increases fibrosis in mice model. This may result from the indiscriminative inhibition of MMPs, some of which function to degrade ECM (de Meijer et al., 2010).

Compared with inhibitor drugs, antagonists achieve the same inhibition effect by targeting cell membrane receptors to dampen downstream signaling. These small molecules bind to receptors without effectively activating them. The receptors of PDGF, vascular endothelial growth factor, endothelin (ET), and angiotensin all have approved antagonist drugs. An antagonist for tyrosine kinase receptors of PDGF, Imatinib, has showed protective effect by reducing differentiation of resting fibroblasts in SSc mice model (Akhmetshina et al., 2009). However, it did not show efficacy in phase II clinical trial in IPF (Daniels et al., 2010) and the high-dose of Imatinib may lead to severe adverse events (Khanna et al., 2011). Macitentan, a dual antagonist of ET_A and ET_B receptor, is beneficial for lung fibrosis. Similarly, another drug Losartan targeting angiotensin II receptor has been evaluated in IPF patients (Couluris et al., 2012). Some preclinical antagonist drugs including Maraviroc (Gonzalez et al., 2014), Atrasentan (Ritter et al., 2014), and PAR1 antagonists (Fiorucci et al., 2004) were under investigation for fibrosis treatment.

On the contrary, many drugs exert their therapeutic effects by activating their targets. Many anti-fibrotic receptors can be targets of these exogenous agonists that augment the downstream biological responses to suppress fibrosis. An approved agonist drug in this category is Iloprost, which can reverse right ventricle fibrosis by re-establishing collagen balance (Gomez-Arroyo et al., 2015). Another agonist of vasoactive intestinal peptide, Treprostinil, reduces inflammation and collagen deposition (Manitsopoulos et al., 2015). Other anti-fibrotic agonists for cell membrane receptors include Aviptadil, INT-767 (Baghdasaryan et al., 2011) and Beraprost (Kaneshige et al., 2007).

Moreover, a few drugs are synthetic proteins that bind receptors to serve as stimulants and perform the same functions as native proteins. An approved drug, synthetic interferon-γ, Actimmune, has completed phase II or phase III study in multiple fibrosis including IPF (Skaria et al., 2015), liver fibrosis (Muir et al., 2006), and cystic fibrosis (Moss et al., 2005). Another stimulant Refanalin, a HGF mimetic, is a potential drug for liver fibrosis (Fallowfield, 2011; Pellicoro et al., 2014).

Compared with extracellular factors, intracellular targets are less popular owing to their inaccessibility. Drugs targeting intracellular factors are less varied because most of them are small molecules. Small molecules could readily translocate into cytoplasm while large molecules such as monoclonal antibodies face more challenges to cross the plasma membrane (Imai and Takaoka, 2006). Nevertheless, more and more studies concerned intracellular factors as targets in recent years, and numerous candidate targets are identified in cytoplasm, nucleus, and mitochondrion. Many approaches, including increasing membrane permeation, combination with supercharged proteins and activating transport through receptors, were implemented to deliver drugs across cell membrane (Mitragotri et al., 2014).

Fibrosis drugs targeting intracellular factors are summarized and classified into four categories: enzymes, nuclear receptors, other proteins, and epigenetics (Table 2). Many drugs acting through intracellular factors are also inhibitors. These drugs inhibit a wide range of kinases located in cytoplasm, and consequently suppress the translocation of transcription factors that drive the expression of pro-fibrotic genes. Rapamycin and Sirolimus are approved drugs that inhibit mTOR. Rapamycin prevents the activation of macrophages and myofibroblasts and the subsequent release of TGF-β in chronic kidney disease (CKD) (Chen et al., 2012). Sirolimus shows anti-inflammatory and anti-fibrotic effects in IPF (Tulek et al., 2011). Besides kinases, there are many other intracellular proteins that serve as potential targets for fibrosis management. Pirfenidone, one of the IPF drugs, has completed phase III trial in IPF patients with alleviated disease progression and acceptable side effects (King et al., 2014). The potential mechanism of Pirfenidone is inhibiting the nuclear accumulation of intracellular proteins SMAD2/3 to regulate TGF-β signaling (Choi et al., 2012). Other approved inhibitor drugs include Ruxolitinib for bone marrow fibrosis (Wilkins et al., 2013), Paquinimod for SSc (Stenstrom et al., 2016), and Pentoxifylline (Okunieff et al., 2004) combined with vitamin E (Jacobson et al., 2013).

In addition, there are also some nuclear receptors located in cytoplasm and nucleus, which could be activated by small molecule agonists, such as PPAR. Rosiglitazone, a PPAR-γ agonist, has anti-fibrotic effect as a consequence of activating MMP-1 and elevating HGF expression in patients with systemic sclerosis-related interstitial lung disease (Bogatkevich et al., 2012). Other approved PPAR targeting drugs like Elafibranor and Pioglitazone, have entered phase II and phase I studies, respectively. Obeticholic acid is an approved drug entering phase III study that decreases inflammation and fibrosis in NASH patients via activating FXR signaling (Verbeke et al., 2016).

Epigenetics are very different types of fibrosis therapies. The most studied epigenetics-based therapy for fibrosis is microRNA. MicroRNA could be neutralized by specific anti-miRNA oligonucleotides delivered into cells. Among them, anti-miR-21 has been reported to inhibit miR-21 activity and ameliorate fibrosis progression through PPAR signaling in CKD (Chau et al., 2012). Another strategy is the intervention of DNA methylation for proteins such as BMP endothelial cell precursor-derived regulator (BMPER), which acts as the regulator of fibroblasts activation. Altering methylation on BMPER gene has been reported to decrease BMPER level and thus to inhibit fibroblasts activity (Huan et al., 2015). Besides, some drugs targeting both extracellular and intracellular factors are also incorporated (Table 3). A majority of them are antioxidants, including an approved drug N-acetylcysteine (Zhang et al., 2014).

Differed from single-component drugs that target a single protein or other simple targets, multi-component drugs contain more than one active ingredient. Traditional Chinese medicines (TCM) therapies, usually appeared as herbal formula, have been studied for thousands of years as multi-component drugs (Wang et al., 2012). Nowadays, single compound acting on multiple targets and multiple compounds acting on multiple targets are popular strategies in drug development (Hopkins, 2008). Fibrosis includes numerous complicated pathological pathways. Multi-component drugs, aiming at different targets, have the advantage in modulating these pathways simultaneously and producing synergistic effects. Moreover, multi-component drugs are expected to provide great resources for discovering new effective drug molecules. Many studies have revealed the pharmacology of multi-component drugs in the fibrosis treatment (Feng et al., 2009; Yang et al., 2009; Li and Kan, 2017) (Table 4). For example, Fuzhenghuayu capsule (FZHY), a well-known multi-component drug for treating liver fibrosis, inhibits liver fibrosis and improves liver function in patients via inhibition of nuclear factor kappa-B kinase subunit β/nuclear factor κF and TGF-β signaling (Liu et al., 2005). Another emerging multi-component drug, Qishenyiqi (QSYQ), is under phase II clinical trial for ischemic heart failure. QSYQ attenuates cardiac fibrosis via IL-6/STAT3 and TNF-α/nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB) signalings and anti-apoptosis activities (Wang et al., 2017). Multi-component drugs act on different physiological reactions associated with fibrosis, such as inflammation and angiogenesis, leading to a systematic improvement of disease. Qushi Huayu Decoction (QHD) is a multitargeting drug that alleviates fibrosis by reducing ROS via the induction of glutathione and modulating lipid metabolism and gut barrier function (Feng et al., 2017).

With multiple targets being discovered, it becomes apparent that more common targets will be shared across many fibrotic diseases with common mechanisms, though some proteins will not express in special conditions. Meanwhile, there will be more chance for drug repositioning, which indicate common drugs will be shared across different targets andfibroproliferative diseases.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer XT and handling Editor declared their shared affiliation.