Pathogenesis of hepatic fibrosis

INTRODUCTION — Fibrosis is a wound healing response in which damaged regions are encapsulated by an extracellular matrix or scar. It develops in almost all patients with chronic liver injury at variable rates depending in part upon the cause of liver disease and host factors [1-4]. In contrast, for unclear reasons, patients with self-limited injury (such as fulminant hepatitis) do not develop scarring despite an abundance of fibrogenic stimuli, unless they go on to develop chronic injury.

The composition of the hepatic scar is similar regardless of the cause of injury. Furthermore, hepatic fibrosis represents a paradigm for wound healing in other tissues, including skin, lung, and kidney, since it involves many of the same cell types and mediators.

Fibrosis occurs earliest in regions where injury is most severe, particularly in chronic inflammatory liver disease due to alcohol or viral infection. As an example, pericentral injury is a hallmark of alcoholic hepatitis; the development of pericentral fibrosis (also known as sclerosing hyaline necrosis or perivenular fibrosis) is an early marker of likely progression to panlobular cirrhosis [5].

The development of fibrosis usually requires several months to years of ongoing injury. Three exceptions in adults are veno-occlusive disease, mechanical biliary obstruction, and fibrosing cholestatic hepatitis associated with viral hepatitis in which (for unclear reasons) fibrosis can progress more rapidly.

While fibrosis is reversible in its initial stages, progressive fibrosis can lead to cirrhosis. The exact point when fibrosis becomes irreversible is incompletely understood. However, increasing evidence suggests that even early stages of cirrhosis may be reversible [6-11]. Furthermore, an understanding of the molecular mechanisms involved in fibrogenesis has a number of clinical implications, including the development of interventions designed to impede or reverse hepatic fibrosis, some of which are already available [12,13]. Despite significant advances in understanding hepatic fibrosis and defining targets of therapy, there are no antifibrotic drugs yet approved for clinical use in patients with advanced liver disease.

This topic review will discuss the mechanisms underlying hepatic fibrosis.

EXTRACELLULAR MATRIX COMPOSITION OF THE NORMAL AND FIBROTIC LIVER — Extracellular matrix (ECM) refers to a group of macromolecules that comprise the scaffolding of normal and fibrotic liver. These include collagens, noncollagen glycoproteins, matrix-bound growth factors, glycosaminoglycans, proteoglycans, and matricellular proteins. A great deal of progress has been made in identifying new members of these families and in understanding how these molecules interact [14-16]. In addition to providing the scaffolding of the liver, matrix molecules are now recognized to have a variety of other functions. As an example, some serve as transmembrane transducers of extracellular signals.

Marked heterogeneity exists within different tissue regions in matrix composition with respect to the variety of isoforms within each class of molecules, their stoichiometry, and their intermolecular interactions. Furthermore, hybrid molecules have been identified that contain both collagenous and proteoglycan domains.

●In the normal liver, collagen types I, III, V, and XI (sometimes referred to as "fibril-forming" collagens) are principally found in the capsule, around large vessels, and in the portal triad, while only scattered fibrils containing types I and III collagen can be found in the subendothelial space. Smaller amounts of other collagens including types IV, VI, XIV (previously called "undulin"), and XVIII can also be found.

●Also present are glycoproteins and matricellular proteins including subendothelial deposits of fibronectin, laminin, tenascin, SPARC, and von Willebrand factor. The proteoglycans consist primarily of heparan sulfate proteoglycans such as perlecan, as well as small amounts of decorin, biglycan, fibromodulin, aggrecan, glypican, syndecan, and lumican [14].

●The appearance of elastin may mark the conversion of reversible to irreversible fibrosis, and its degradation by macrophages may be an important determinant of elastic accumulation [17-21].

As the liver becomes fibrotic, significant changes occur in the ECM quantitatively and qualitatively. The total collagen content increases 3- to 10-fold [22]. There is also a marked increase in the ECM to levels and composition typically seen in wound healing. These include an increase in fibril-forming collagens (ie, types I, III, and IV), some non-fibril forming collagens (types IV and VI), a number of glycoproteins (cellular fibronectin, laminin, SPARC, osteonectin, tenascin, and von Willebrand factor), proteoglycans, and glycosaminoglycans (perlecan, decorin, aggrecan, lumican, and fibromodulin) [23]. Particularly notable is a shift from heparan sulfate-containing proteoglycans to those containing chondroitin and dermatan sulfates. These processes represent a change in the type of ECM in subendothelial space from the normal low-density basement membrane-like matrix to the interstitial type [23].

The replacement of the low-density matrix with the interstitial type has consequences on the function of hepatocytes, hepatic stellate cells, and endothelial cells, in part explaining the synthetic and metabolic dysfunction observed in patients with advanced fibrosis. The high-density matrix also activates hepatic stellate cells leading to the loss of hepatocyte microvilli and disappearance of endothelial fenestrations, which impairs transport of solutes from the sinusoid to the hepatocytes, thereby further contributing to the hepatocyte dysfunction [24]. At the same time, altered function of liver sinusoidal endothelial cells can promote injury and impede regeneration [25]

The liver responds to injury with angiogenic stimulation, with evidence of new blood vessel formation, sinusoidal remodeling, and pericyte (ie, stellate cell) amplification [26,27]. Thus, angiogenic mediators are involved, including platelet-derived growth factor, vascular endothelial growth factor (VEGF) and their cognate receptors, as well as vasoactive mediators that include nitric oxide and carbon monoxide. Increased VEGF concentrations may be particularly important in progression of fibrosis in smokers who have hepatitis C [28].

Progressive accumulation of ECM composition provokes positive feedback pathways that further amplify fibrosis.

●First, changes in membrane receptors, in particular integrins, sense altered matrix signals that provoke stellate cell activation and migration through focal adhesion disassembly [29-31]. Matrix-provoked signals also engage membrane-bound GTP-binding proteins, in particular Rho [32] and Rac [33].

●Second, activation of cellular matrix metalloproteases leads to release of fibrogenic and proliferative growth factors from matrix-bound reservoirs in the extracellular space [14,34,35].

●Third, the enhanced density of ECM leads to increasing matrix stiffness, which may stimulate stellate cell activation through integrin signaling [36]. These experimental findings explain the increasing use of FibroScan [37] and magnetic resonance elastography [38], two clinical techniques which noninvasively assess hepatic stiffness as a reflection of ECM content. (See "Noninvasive assessment of hepatic fibrosis: Overview of serologic tests and imaging examinations".)

BIOLOGIC ACTIVITY OF EXTRACELLULAR MATRIX — Matrix alterations observed during fibrogenesis alter cellular behavior by processes involving cell membrane receptors. One of the best characterized are integrins, which are a large family of homologous membrane linker proteins that control several cellular functions including gene expression, growth, and differentiation. A variety of other cytokines and adhesion proteins involved in hepatic fibrogenesis have also been described.

Integrins — Integrins are composed of alpha and beta subunits whose ligands are matrix molecules rather than cytokines [39]. In particular, integrin ligands contain an Arg-Gly-Asp (RGD) tripeptide sequence. Several integrins and their downstream effectors have been identified in stellate cells, including alpha-1-beta-1, alpha-2-beta-1, alpha-5-beta-1, and alpha-6-beta-4. The common presence of RGD with many integrin ligands has raised the possibility of using competitive RGD antagonists to block integrin-mediated pathways in fibrogenesis.

Additional studies have identified the integrin subunits alpha-1, alpha-V, beta-1 and beta-6 as potentially important therapeutic targets, not only in liver diseases [40] but also in fibrosis of other organs [41-43].

Integrin signaling across the plasma membrane permits communication between the extracellular matrix (ECM) and cytoskeleton. Signaling occurs in conjunction with the phosphorylation of several of intracellular substrates, which is typical of most transmembrane receptors. However, in addition to the "outside-in" signaling pathway, integrins can also signal in the opposite direction (ie, from the inside to the outside of the cell), thereby mediating cytoskeletal changes that can lead to altered conformation of ECM molecules such as fibronectin.

Several integrin and non-integrin receptors have been described in situ on hepatocytes and non-parenchymal cells [29,30,44-47]. Upregulation of alpha-6-beta-1 and alpha-2-beta-1 receptors [5], both of which bind laminin, has been reported in experimental fibrosis. Studies have also defined the integrin phenotypes of isolated cell types from liver. In particular, stellate cells express integrin receptors for collagen and laminin [29,44,45,48], which may contribute to their activation and proliferation in response to deposition of these matrix components during injury.

Integrins are an attractive target for antifibrotic therapies because several integrin heterodimers can activate the fibrogenic cytokine TGF-beta at the cell surface, and therefore, integrin antagonists may interrupt this effect [41] (see 'Soluble growth factors' below). Clinical trials exploring the role of integrins are underway in patients with fibrosis of the liver and other tissues [49].

Other adhesion proteins and cell matrix receptors — A growing number of adhesion proteins and cell matrix receptors other than integrins have been characterized, including cadherins and selectins, which mediate interactions between inflammatory cells and the endothelial wall [50-52]. As an example, upregulation of a tyrosine kinase receptor, discoidin domain receptor 2 (DDR2), has been observed during stellate cell activation leading to enhanced matrix metalloproteinase expression and cell growth [53]. DDR2 is the only tyrosine kinase receptor whose ligand is an ECM molecule rather than a peptide ligand. Its upregulation may be a critical step toward perpetuating liver fibrosis. Cell culture studies suggest this might be a potential therapeutic target [54].

Soluble growth factors — The ECM can also affect cell function indirectly by the release of soluble growth factors (cytokines). The soluble growth factors are controlled by local metalloproteinases (a family of zinc-dependent enzymes) [55]. The cytokines include platelet-derived growth factor, HGF, connective tissue growth factor, tumor necrosis factor alpha, basic fibroblast growth factor, and vascular endothelial cell growth factor [14]. Controlled release of these cytokines from the ECM is a key mechanism for regulating cytokine activity since it provides a local, accessible source of cytokines that can be regulated tightly through the actions of proteases and their inhibitors. In addition, ECM can regulate the activity of proteases through specific binding to collagens or fibronectins [14].

Transforming growth factor (TGF)-beta-1, derived from both paracrine and autocrine sources, remains the classic fibrogenic cytokine [56,57]. Signals downstream of TGF-beta converge upon Smad proteins, which fine-tune and enhance the effects of TGF-beta during stellate cell activation; Smads 2 and 3 are stimulatory, whereas Smad7 is inhibitory [56,58,59] and is antagonized by Id1 [60]. TGF-beta-1 also stimulates collagen transcription in stellate cells through a hydrogen peroxide- and C/EBP-beta-dependent mechanism [61]. The response of Smads in stellate cells evolves as injury becomes chronic, further enhancing fibrogenesis [58,62]. Studies have identified an upstream signaling intermediate, IQGAP1, that attenuates TGF-beta signaling [63].

CELLULAR SOURCES OF ECM IN NORMAL AND FIBROTIC LIVER — A major advance in the understanding of hepatic fibrosis has been the identification of the cellular sources of extracellular matrix (ECM). The hepatic stellate cell (previously called the lipocyte, Ito, fat-storing, or perisinusoidal cell) is the primary source of ECM in normal and fibrotic liver. In addition, related mesenchymal cell types from a variety of sources may have measurable contributions to total matrix accumulation, including classical portal fibroblasts [64-66] (especially in biliary fibrosis), bone marrow-derived cells [67], as well as fibroblasts derived from epithelial-mesenchymal transition (EMT) [68]. EMT is a well-characterized response of the kidney [68] to injury, but its role in liver injury has been less convincing.

Hepatic stellate cells, located in subendothelial space of Disse between hepatocytes and sinusoidal endothelial cells, represent one-third of the non-parenchymal population or approximately 15 percent of the total number of resident cells in normal liver (figure 1) [69]. In normal liver, they are the principal storage site for retinoids (vitamin A metabolites), which accounts for 40 to 70 percent of retinoids in the body. Most of the retinoids are in the form of retinyl esters and are confined to cytoplasmic droplets. Stellate cells actually comprise a somewhat heterogeneous group of cells that are functionally and anatomically similar but differ in their expression of cytoskeletal filaments, their retinoid content, and in their potential for activation [70-72]. A series of studies have used single-cell sequencing to fully characterize the heterogeneity of hepatic stellate cells in mouse models and human liver [73-78]. These studies underscore the heterogeneity of stellate cells, but also reinforce the conclusion that activated stellate cells/myofibroblasts are the primary fibrogenic cell in injured liver. Stellate cells with fibrogenic potential have been identified in locations other than the liver, such as the pancreas [79]. Furthermore, remarkable plasticity of stellate cell phenotype has been documented in vivo and in culture, precluding a strict definition based only on cytoskeletal phenotype [47,80]. Indeed, single cell analyses indicate there are several subtypes, whose distinct functional roles are not yet fully clarified. Stellate cells with fibrogenic potential are not confined to the liver and have been identified in other organs such as the pancreas, where they contribute to primary or metastatic cancers as cancer-associated fibroblasts [81-83] or to desmoplasia in chronic pancreatitis [84].

Animal and human studies have defined a process of changes within stellate cells that collectively are termed "activation." During activation, stellate cells undergo a transition from a quiescent vitamin A-rich cell into proliferative, fibrogenic, and contractile myofibroblasts (figure 2). This change is characterized morphologically by enlargement of rough endoplasmic reticulum, diminution of vitamin A droplets, a ruffled nuclear membrane, appearance of contractile filaments, and proliferation. Cells with features of both quiescent and activated cells are sometimes referred to as "transitional" cells. Proliferation of stellate cells occurs predominantly in regions of greatest injury. Their function has been characterized in a variety of human diseases including alcohol-associated liver disease, viral hepatitis, hepatocellular carcinoma, vascular diseases, hematologic malignancy, biliary disease, mucopolysaccharidosis, acetaminophen overdose, leishmaniasis, allograft rejection, and in substance use disorder [85-89].

Sinusoidal endothelial cells are also an important contributor to early fibrosis. Similar to stellate cells, there is considerable heterogeneity of this cell type in normal and fibrotic liver [72]. Endothelial cells from normal liver produce types III and IV collagen, laminin, syndecan, and fibronectin [90-93]. Increased expression of cellular isoforms of fibronectin by these cells is a key early event following acute liver injury because their appearance creates a microenvironment that activates stellate cells.

Other cell types may also contribute to fibrosis. CD8+ lymphocytes are potentially profibrogenic cells based upon their ability to induce fibrogenesis after adoptive transfer from animals with liver injury [94]. However, in hepatitis C virus infection, direct acting antiviral treatment may not resolve the increase in this cell type [95]. Cross-talk between sinusoidal endothelial cells and stellate cells or hepatocytes regulate both fibrosis and regeneration [96,97]. Cross talk between macrophages and stellate cells is a key driver of fibrogenesis as well [73].

Remarkably, very few studies have defined the cellular or matrix composition of congenital hepatic fibrosis, an entity whose pathogenesis is unclear. Current theories suggest that as in adults, congenital fibrosis represents a final common pathway of fetal hepatic injury, whether from biliary malformations, viral infections (especially cytomegalovirus) or other insult, with the stellate cell playing a significant role [98].

Fibrogenic cells in the liver derive not only from resident stellate cells, but also from portal fibroblasts [64,99,100], and potentially from circulating fibrocytes [101], bone marrow [67], and epithelial-mesenchymal cell transition [102,103]. Portal fibroblasts may be particularly important in cholestatic liver diseases and ischemia [100], in which paracrine interactions between cholangiocytes and fibroblasts involve both chemokines [104] and extracellular nucleotides [105]. Although controversial, progressive recruitment of bone marrow derived cells may occur over time, but do not represent a major fraction of the total fibrogenic population in chronic injury. Studies using genetic tracing and single-cell sequencing techniques emphasize that, while other sources including bone marrow are possible, the vast bulk of fibrotic gene expression occurs in myofibroblasts derived from activated stellate cells [73-76,106].

The development of technologies to sequence the full transcriptome of individual cells has transformed our understanding of hepatic fibrogenesis by defining an unprecedented level of heterogeneity among stellate cells and other fibrogenic cells in human and rodent livers [74,75,107,108]. These studies may allow us to target specific subsets of cells that are especially fibrogenic and pro-inflammatory. For example, one study used chimeric-antigen receptor (CAR) T cells engineered to deplete only senescent stellate cells in a fibrosis model in mice, which markedly reduced fibrosis and improved liver function [109].

DEGRADATION OF EXTRACELLULAR MATRIX — Fibrosis reflects a balance between matrix production and degradation. The degradation of extracellular matrix is a key event in hepatic fibrosis. Early disruption of the normal hepatic matrix by matrix proteases hastens its replacement by scar matrix, which has deleterious effects on cell function. As a result, such degradation has been referred to as being "pathologic." On the other hand, resorption of excess matrix in patients with chronic liver disease provides the opportunity to reverse hepatic dysfunction and portal hypertension [110].

An understanding of mechanisms involved in matrix remodeling has evolved significantly in the past several years. A critical element in matrix remodeling is a family of matrix metalloproteinases (MMPs; also known as matrixins). These are calcium-dependent enzymes that specifically degrade collagens and noncollagenous substrates [111,112]. As a general rule, the MMPs fall into five categories based upon their substrate specificity:

●Interstitial collagenases (MMP-1, -8, -13), which degrade interstitial collagens

●Gelatinases (MMP-2, -9 and fibroblast activation protein), which generally degrade basement membrane collagens and denatured interstitial collagen

●Stromelysins (MMP-3, -7, -10, 11), which have a broad substrate range

●Membrane type (MMP-14, 15, -16, -17, -24, -25), which are primarily interstitial collagenases

●Metalloelastase (MMP-12), which degrades elastin [113]

Metalloproteinases are regulated at many levels, permitting their activity to be restricted to discrete regions within the pericellular environment. Inactive metalloproteinases can be activated through proteolytic cleavage by either membrane-type MMP-1 (MT1-MMP) or plasmin, and inhibited by binding to specific inhibitors known as tissue inhibitors of metalloproteinases (TIMPs). As an example, MT1-MMP and TIMP-2 form a ternary complex with MMP-2, possibly including avb3 integrin, which is essential for optimal MMP-2 activity [111]. Plasmin activity is controlled by its activating enzyme uroplasminogen activator (uPA) and a specific inhibitor, plasminogen activator inhibitor 1 (PAI-1), and can be stimulated by active transforming growth factor beta-1 (TGF-beta-1).

Thus, collagenase activity reflects the relative amounts of activated metalloproteinases and their inhibitors, especially TIMPs. Other protease inhibitors (such as a2-macroglobulin) may also affect net degradative activity.

Fibrosis regression is increasingly observed as treatments for chronic liver disease have improved, especially for hepatitis B virus and hepatitis C virus infection. Thus, interest is focused on mediators of fibrosis regression. In animal models, macrophage subsets have emerged as an important determinant, especially those that have low-cell surface expression of Ly6C [114], but their corresponding human counterparts are not yet clarified. In mouse studies, triggering receptors expressed on myeloid cells 2 (Trem2) is a cell surface molecule that marks macrophages that are more pro-inflammatory and pro-fibrogenic [73].

Pathologic matrix degradation — "Pathologic" matrix degradation refers to the early disruption of the matrix of the normal space between hepatocytes and endothelial cells. Degradation occurs through the actions of at least four enzymes:

●Matrix metalloproteinase 2 (MMP-2) (also called "gelatinase A" or "72 kDA type IV collagenase") and MMP-9 ("gelatinase B" or "92 kDa type IV collagenase"), which degrade type IV collagen

●Membrane-type metalloproteinase-1 or -2, which activate latent MMP-2

●Stromelysin-1, which degrades proteoglycans and glycoproteins and also activates latent collagenases

Stellate cells are the principal source of MMP-2 [115,116] and stromelysin [117]. Activation of latent MMP-2 may require interaction with hepatocytes [118,119]. Markedly increased expression of MMP-2 is characteristic of cirrhosis [120]. MMP-9 is secreted locally by Kupffer cells and macrophages [111]. Disruption of the normal liver matrix is also a requirement for tumor invasion and desmoplasia [121].

As noted above, a paradigm is emerging in which inflammatory cell subsets are the primary source of degradative enzymes that can degrade scar. In particular, macrophage subsets, including one called "Ly6C-lo," are increased when liver fibrosis regression is maximal [122]. Additionally, dendritic cells in liver also harbor proteolytic activity [123]. An increasing focus on inflammatory and immunoregulatory cells in the liver is likely to yield important advances in understanding matrix degradation during fibrosis regression, with clear therapeutic implications [124-126].

Failure to degrade the increased interstitial or scar matrix is a major determinant of progressive fibrosis. MMP-1 is the main protease that can degrade type I collagen, the principal collagen in fibrotic liver. However, sources of this enzyme are not as clearly established as for the type IV collagenases. Stellate cells express MMP-1 mRNA, but little enzyme can be detected [127]. More importantly, progressive fibrosis is associated with marked increases in TIMP-1 [128,129] and TIMP-2 [130], leading to a net decrease in protease activity and therefore more unopposed matrix accumulation. Stellate cells are the major source of these inhibitors [131]. Sustained TIMP-1 expression is emerging as a key reason for progressive fibrosis, and its diminution is an important prerequisite to allow for reversal of fibrosis (see below). Unique mechanisms of TIMP-1 regulation in stellate cells [132] offer the potential for selective inhibition of TIMP-1 expression in order to accelerate resorption of scar matrix in patients with liver disease.

The cross-linking of collagen by lysyl oxidase and tissue transglutaminase, and the "maturation" of hepatic scar through the action of ADAMTS2 (A Disintegrin and Metalloproteinase with ThromboSpondin type repeats metalloproteinase with thrombospondin type I motif) may regulate hepatic fibrosis reversibility. In animal models, even advanced fibrosis is reversible, limited by the extent of collagen cross-linking due to tissue transglutaminase [133] and lysyl oxidase-2 [134]. As advanced fibrosis resolves, the micronodules typical of active cirrhosis dissolve, coalescing into macronodules [133]. This finding correlates with clinical data demonstrating that increased septal thickness and smaller nodule size are significant predictors of poorer clinical outcomes [135]. Clinical trials targeting inhibition of lysyl oxidase 2 with a monoclonal antibody have been negative [136], but the molecule still remains a potential target for antifibrotic therapy using small molecule inhibitors that may more readily access the target molecule in fibrotic liver.

STELLATE CELL ACTIVATION, THE CENTRAL EVENT IN HEPATIC FIBROSIS — Hepatic stellate cell activation is the common pathway leading to hepatic fibrosis. Once they are activated, they release chemokines and other leukocyte chemoattractants while upregulating the expression of important inflammatory receptors such as ICAM-1, chemokine receptors, and mediators of lipopolysaccharide signaling [137].

Activation consists of two major phases:

●Initiation (also called a "preinflammatory stage"), which refers to early changes in gene expression and phenotype that render the cells responsive to other cytokines and stimuli [70]. Initiation results mostly from paracrine stimulation.

●Perpetuation results from the effects of these stimuli on maintaining the activated phenotype and generating fibrosis. Perpetuation involves autocrine as well as paracrine loops.

Initiation — The earliest changes observed during stellate activation result from paracrine stimulation by all neighboring cell types, including sinusoidal endothelium, Kupffer cells and infiltrating inflammatory monocytes, hepatocytes, and platelets. As noted above, early injury to endothelial cells stimulates production of cellular fibronectin, which has an activating effect on stellate cells [93]. Endothelial cells are also likely to participate in conversion of transforming growth factor (TGF)-beta from the latent to active, profibrogenic form. Platelets are another important source of paracrine stimuli, including platelet-derived growth factor (PDGF), TGF-beta-1, and epidermal growth factor (EGF) [138]. TGF-beta-1, derived from both paracrine and autocrine sources, is the best characterized and most potent fibrogenic cytokine.

Monocyte infiltration and activation also contribute to stellate cell activation. Kupffer cells stimulate matrix synthesis, cell proliferation, and release of retinoids by stellate cells through the actions of cytokines (especially TGF-beta-1) and reactive oxygen intermediates/lipid peroxides [139]. On the other hand, activated macrophages can also lead to stellate cell apoptosis by different mechanisms [140]. Growing appreciation for macrophage heterogeneity indicates that specific subsets serve important and sometimes divergent roles in driving or regressing fibrosis [114,141-144].

Hepatocytes are also a potent source of fibrogenic lipid peroxides. Hepatocyte apoptosis following injury also promotes stellate cell initiation through a process mediated by Fas (a protein involved in causing apoptosis) [145,146]. Apoptosis of parenchymal cells is also an important inflammatory stimulus [147]. The response of stellate cells to apoptotic hepatocytes in part reflects the interaction of hepatocyte DNA with Toll-like receptor 9 (TLR9) expressed on stellate cells [148]. A profibrogenic response can also be elicited by hepatocyte apoptosis following disruption of the antiapoptotic mediator Bcl-xL and by Fas [145,149].

Thus, efforts to block hepatocyte apoptosis therapeutically are being developed as a potential antifibrotic strategy [150]. In contrast, selective stimulation of apoptosis in stellate cells by either TRAIL [151], gliotoxin [152], or proteasome inhibitors [153] is antifibrotic. While hepatocyte necrosis associated with lipid peroxidation is considered a classical inflammatory and fibrogenic stimulus, apoptosis, or programmed cell death, has also been implicated in the fibrogenic response. Apoptotic fragments released from hepatocytes are fibrogenic towards cultured stellate cells [154], and activate Kupffer cells [155] Also, Fas-mediated hepatocyte apoptosis in vivo in experimental animals is fibrogenic [145].

The cytochrome CYP2E1 may have an important role in the generation of reactive oxygen species that stimulates hepatic stellate cells [156]. Cultured hepatic stellate cells grown in the presence of a cell that expresses CYP2E1 (E47 cells) increase the production of collagen, an effect prevented in the presence of antioxidants or a CYP2E1 inhibitor [156]. These data suggest that the CYP2E1-derived reactive oxygen species are responsible for the increased collagen production. In similar experiments using co-cultured hepatic stellate and E47 cells, the addition of arachidonic acid plus ferric nitrilotriacetate (agents that potentiate oxidative stress) further induced collagen protein synthesis [157]. These findings may help to explain the pathogenesis of liver injury in alcoholic liver disease since CYP2E1 is alcohol-inducible.

As noted above, reactive oxygen species (ROS) are generated through lipid peroxidation from hepatocytes, macrophages, stellate cells, and other inflammatory cells [158,159]. In alcohol-associated or nonalcohol-associated steatohepatitis, ROS generation in hepatocytes results from induction of cytochrome P450 2E1 [160,161], leading to pericentral (zone 3) injury. Also, NADPH oxidase (NOX) mediates fibrogenic activation in hepatic stellate cells, as well as in Kupffer cells or resident liver macrophages through generation of oxidant stress [162]. Increasing knowledge about NOX isoforms and their cell-specific activities is leading to their emergence as a therapeutic target [163]. Hepatitis C virus (HCV) as well as HCV/HIV coinfection also produces oxidant stress molecules that amplify inflammation and fibrosis [164,165].

Nitrosative stress from hepatocyte mitochondrial injury and induction of nitric oxide synthase 2 (NOS2) may also be an overlooked pathway of liver injury [166,167]. Hypoxia is typically also a component of the injury milieu and elicits fibrogenic and angiogenic signaling [27].

The identification of toll-like receptor 4 (TLR4), the receptor for bacterial lipopolysaccharide, on Kupffer cells and stellate cells introduces a role for the innate immune system in hepatic fibrosis [168]. Furthermore, signaling by stellate cells in response to LPS and possibly endogenous ligands of TLR4 (eg, high-mobility group box 1 [169,170], biglycan and heparan sulfate) may be more important than in Kupffer cells in eliciting a fibrogenic response by downregulating BAMBI, a transmembrane suppressor of TGF-beta-1, which is the major fibrogenic cytokine in the liver [171]. This finding correlates with evidence that specific single nucleotide polymorphisms of TLR4 contribute to the rate of fibrosis progression in HCV infection [172], thereby linking a genetic risk marker to disease pathogenesis. TLR2 has also been identified as a regulator of inflammation and fibrosis in fatty liver diseases [173,174].

Direct effects of hepatotrophic viruses may also contribute to stellate cell activation. In particular, HCV proteins can activate stellate cells through several mechanisms [175-177], whereas accelerated fibrosis in chronic hepatitis B virus may engage specific immune cell types, especially natural killer cells [178]. Some of the signals driving fibrosis may derive from the gut microbiome.

Gene regulation during stellate cell activation — Attention has focused on regulatory pathways that can respond quickly to injurious stimuli, either by activating or repressing gene transcription, by epigenetic regulation, or by posttranscriptional control. A number of potential antifibrotic targets are present both on the cell surface and within hepatic stellate cells (figure 3) [137]. As fibrosis advances, the collection of potential therapeutic targets evolves from those engaged in paracrine stimulation towards hepatic stellate cells, to largely autocrine interactions by this cell type [179]. These findings suggest that drugs that may be effective in earlier fibrosis may be less useful as the disease advances because the pathways and signals driving disease evolve.

Among the many target genes of transcription factors described in stellate cells, those most comprehensively characterized include type I collagen (alpha 1 and alpha 2 chains), alpha-SMA, TGF-beta-1 and TGF-beta receptors, matrix metalloproteinase (MMP)-2, and tissue inhibitors of metalloproteinases 1 and 2 [180-183]. Nuclear hormone receptors, in particular the vitamin D receptor, peroxisome proliferator-activated receptor gamma, and LXR, have been implicated in fibrogenic gene regulation in hepatic stellate cells [184-186]. Similarly, autophagy, an intracellular pathway that provides metabolic energy, is required for stellate cells to activate and may explain the hydrolysis of retinyl esters that occurs during stellate cell activation [187-189]. Its effects may be mediated in part through enhanced endoplasmic reticulum stress involving the inositol-requiring enzyme 1 (IRE1) branch of the unfolded protein response [189].

Epigenetic regulation is a tightly controlled pathway that modulates stellate cell activation in part through induction of the molecules CBF1 and MeCP2 [181,190-192].

mRNA stabilization also contributes to increased gene expression during stellate cell activation. Specifically, there is a 16-fold increase in collagen alpha 1(I) mRNA stabilization during stellate cell activation due to interaction of a specific protein, alpha CP, to a specific sequence in the 3' untranslated region of the mRNA [193], and also involving the interaction of a 120 kDa protein with the 5' stem-loop structure [194].

Perpetuation — Perpetuation of stellate cell activation involves at least seven discrete changes in cell behavior: proliferation, chemotaxis, fibrogenesis, contractility, matrix degradation, retinoid loss, and white blood cell chemoattractant and cytokine release. The net effect of these changes is to increase accumulation of extracellular matrix. As an example, proliferation and chemotaxis lead to increased numbers of collagen-producing cells as well as more matrix production per cell. Cytokine release by stellate cells can amplify the inflammatory and fibrogenic tissue responses, and matrix proteases may hasten the replacement of normal matrix with one typical of the wound "scar."

Proliferation — Platelet-derived growth factor (PDGF) is the most potent stellate cell mitogen identified [195]. Induction of PDGF receptors early in stellate cell activation increases responsiveness to this potent mitogen [196]. A co-receptor, neuropilin-1, may enhance PDGF signaling at the cell membrane of stellate cells [197]. Downstream pathways of PDGF signaling have been carefully characterized in stellate cells [198]. In addition to proliferation, PDGF stimulates Na+/H+ exchange, providing a potential site for therapeutic intervention by blocking ion transport [199]. Other compounds with mitogenic activity in stellate cells and a potential role in fibrogenesis include vascular endothelial cell growth factor [200], thrombin and its receptor [201,202], EGF, TGF-alpha, keratinocyte growth factor [203] and basic fibroblast growth factor [204]. Signaling pathways for these and other mitogens have been greatly clarified in stellate cells, offering many potential sites for therapeutic intervention [198,205]. In addition, cell cycle engagement has been characterized during stellate cell activation in response to PDGF and other mitogens [206].

PDGF-C and -D isoforms have also been described [207]. PDGF-D may be the most potent and physiologically relevant PDGF subunit in stellate cell activation [208]. Furthermore, whereas both mice with transgenic expression of either PDGF-B [207] or PDGF-C have hepatic fibrosis [200], the PDGF-C transgenic animals also develop hepatocellular carcinoma (HCC) [200], mimicking the progression from fibrosis to cancer that occurs in humans.

Chemotaxis — Stellate cells can migrate towards cytokine chemoattractants [209,210], explaining in part why stellate cells align within inflammatory septae in vivo. These chemoattractants include PDGF [211,212], MCP-1 [213], and CXCR3 [214].

In contrast to PDGF, adenosine blunts chemotaxis, thereby providing a counter-regulatory pathway that fixes cells at sites of injury [215]. Paradoxically, enhanced adenosine signaling may also contribute to alcoholic fibrosis by stimulating stellate cell fibrogenesis [216], which not only represents a potential fibrogenic mechanism, but also may explain the protective effect of caffeine (which inhibits adenosine generation) reported in epidemiologic studies [217].

Fibrogenesis — The most direct way that stellate cells influence fibrosis is by increasing matrix production and scar formation. The best-studied component of hepatic scar is collagen type I, the expression of which is regulated posttranscriptionally in hepatic stellate cells. The most potent stimulus for collagen I production is TGF-beta, which is derived from both paracrine and autocrine sources; TGF-beta also stimulates the production of other matrix components including cellular fibronectin and proteoglycans [23,218]. Other factors that stimulate collagen I by activated stellate cells in culture include retinoids, angiotensin II [219], interleukin-1 beta, tumor necrosis factor, and acetaldehyde. However, none of these is as potent as TGF-beta-1. The expression of type I collagen can be inhibited when protein factors responsible for increasing expression are blocked from binding to a conserved stem-loop in the 5' untranslated region of type I collagen mRNA, suggesting that this region is important in the regulation of collagen synthesis by stellate cells [220].

TGF-beta-1 stimulates collagen in stellate cells through a hydrogen peroxide and C/EBPb-dependent mechanism [61], which also involves Rho kinase [221]. Signals downstream of TGF-beta include a family of bifunctional molecules known as SMADs, upon which many extracellular and intracellular signals converge to fine-tune and enhance the effects of TGF-beta during fibrogenesis [198]. The response of SMADs in stellate cells differs between acute and chronic injury to further favor matrix production [222].

As mentioned above, lipid peroxidation products are emerging as important stimuli to extracellular matrix production [223] (see 'Cellular sources of ECM in normal and fibrotic liver' above). Their effects may be amplified by loss of antioxidant capacity of stellate cells as they activate [224]. These important insights have provided the rationale for the study of antioxidants in the treatment of a variety of liver diseases.

Hepatic iron concentration may also influence fibrogenesis, at least in patients with hepatitis C. Hepatic iron stores in such patients correlated with histologic disease severity and with stellate cell numbers [225].

Connective tissue growth factor (CTGF/CCN2) is also a potent fibrogenic signal towards stellate cells [226-229] that is upregulated by hyperglycemia and hyperinsulinemia [230]. Interestingly, TGF-beta stimulates CTGF primarily in hepatocytes, not stellate cells [231,232], a notable exception to the general rule that cytokine signaling in stellate cell activation is typically autocrine.

Neurohumoral signaling contributes to stellate cell responses [233]. Specifically cannabinoids are potent mediators of hepatic steatosis, stellate cell activation and fibrosis (reviewed in [234]), as well as provoking the hemodynamic alterations associated with advanced liver disease [235]. Two receptors, CB1 and CB2, exert opposing effects, with CB1 a fibrogenic pathway and CB2 antifibrotic. Thus, antagonism of CB1 signaling in stellate cells represents a promising antifibrotic strategy [236]. In addition, neurotrophin receptor signaling has emerged as a potential target based on its contributions to stellate cell activation and its regulation by thyroid hormone [237-239].

In contrast, agonism of CB2 receptors, which are also expressed by stellate cells, reverses fibrosis in experimental animals [240]. The fundamental challenge of developing cannabinoid therapeutics for liver disease is to minimize central nervous system effects, since CB1 and CB2 receptors are abundantly expressed in brain. Similarly, opioids signal in stellate cells and promote fibrogenesis [241,242], which is antagonized by naltrexone. Finally, sympathetic neurotransmitters also contribute to activation pathways [243].

Leptin, a ubiquitous adipokine, is profibrogenic, and its levels are elevated in patients who are obese [244-246]. In contrast, adiponectin is a natural counter-regulator to leptin whose activity is diminished in advanced liver diseases and obesity, raising the prospect that it may have therapeutic value as an antifibrotic [246-248]. In addition, ghrelin, an orexigenic hormone, may attenuate hepatic fibrosis based on animal studies [249].

Components of Hedgehog signaling, a potent developmental pathway, have been identified in hepatic stellate cells and contribute to their activation and fibrogenesis [250-252].

Contractility — Contractility of stellate cells may be a major determinant of early and late increases in portal resistance during liver fibrosis. The collagenous bands typical of end-stage cirrhosis contain large numbers of activated stellate cells [253]. These impede portal blood flow by constricting individual sinusoids and by contracting the cirrhotic liver. The acquisition of a contractile phenotype during stellate cell activation has been documented in culture and in vivo, and is mediated in part by receptors which interact with the extracellular matrix [198].

In the process of becoming contractile, stellate cells develop increased expression of the cytoskeletal protein alpha smooth muscle actin. If smooth muscle actin is required for contraction, then inactivating it could represent a therapeutic target for treating portal hypertension.

The major stimulus for stellate cell contraction is endothelin-1, whose receptors are expressed on quiescent and activated stellate cells [254]. Endothelin activity is antagonized by the nuclear receptor FXR [255]. With activation, receptor expression does not increase (unlike PDGF receptors), but there is a shift in the predominant type of endothelin receptor and increased sensitivity to autocrine endothelin-1 [253,256]. Locally produced vasodilator substances (particularly nitric oxide) may counteract the constrictive effects of endothelin-1 [253]. In vivo studies suggest that carbon monoxide also mediates sinusoidal relaxation through its effects on stellate cells [257].

Matrix degradation — As discussed above, quantitative and qualitative changes in matrix protease activity have an important role in extracellular matrix remodeling accompanying fibrosing liver injury. Because stellate cells express virtually all the components required for pathologic matrix degradation, they have a key role not only in matrix production, but also matrix degradation.

Retinoid loss — Activation of stellate cells is accompanied by the loss of the characteristic perinuclear retinoid (vitamin A) droplets. In culture, retinoid is stored as retinyl esters, whereas the form of retinoid released outside the cell during activation is retinol, suggesting that there is intracellular hydrolysis of esters prior to export [85]. Whether retinoid loss is required for stellate cells to activate, and which retinoids might accelerate or prevent activation, is incompletely understood. The enzyme PNPLA3 has been proposed as a "gatekeeper" of stellate cell retinoid content, and interestingly, polymorphisms in this gene have also been linked to risk of nonalcoholic fatty liver disease [258]. Free cholesterol may also have an activating role towards stellate cells [259].

Several nuclear retinoid receptors have been identified in stellate cells [260]. These molecules bind intracellular retinoid ligands and regulate gene expression, but it is uncertain whether they have a regulatory role in fibrogenesis. The question has important clinical implications since efforts are being made to use retinoids therapeutically.

Peroxisome proliferator-activated receptors (PPAR), in particular PPARg, have been identified in stellate cells [261,262]. Their expression decreases with activation [261,262]. Ligands for this newly identified nuclear receptor family downregulate stellate cell activation [262].

The intracellular lipid storage protein, adipose differentiation-related protein (ADRP), has been uncovered in stellate cells; its expression is reduced during cellular activation, and it is induced by retinoid exposure, suggesting that ADRP may have a regulatory role linking lipid content to cellular activation through an unknown mechanism [263].

Elegant mechanisms of gene regulation mediated by the vitamin D receptor point towards a potential therapeutic target since vitamin D appears to be antifibrotic [184,264].

Inflammatory signaling and WBC chemoattraction — Stellate cells are assuming an increasingly central role in our understanding of hepatic inflammation. They can amplify the inflammatory response by inducing infiltration of mono- and polymorphonuclear leukocytes. Activated stellate cells produce chemokines that include MCP-1 [265], CCL21 [266], RANTES, and CCR5 [267], among several others [268-271]. These molecules are particularly attractive therapeutic targets, in part because chemokine receptors are G-coupled protein receptors, which are especially targetable by antagonists [272]. Stellate cells also express toll-like receptors (see above) [168], indicating a capacity to interact with bacterial lipopolysaccharide, which in turn stimulates stellate cells [273,274]. Stellate cells can also function as antigen-presenting cells [275] that can stimulate lymphocyte proliferation or apoptosis [276]. Stellate cells produce neutrophil chemoattractants, which could contribute to the neutrophil accumulation characteristic of alcoholic liver disease.

In addition to regulating leukocyte behavior, stellate cells may in turn be affected by specific lymphocyte populations [277]. As an example, CD8 cells harbor more fibrogenic activity towards stellate cells than CD4 cells [94], which may explain in part the increased hepatic fibrosis seen in patients with HCV/HIV coinfection, where CD4/CD8 ratios are reduced, than in patients mono-infected with HCV alone.

Links between stellate and progenitor cells, fibrosis, and cancer — The observation that stellate cells express the stem cell marker CD133 raises the possibility that they are true progenitor cells [278-280]. Importantly, the requirement for fibrosis to occur before HCC develops in patients with chronic HCV remains unexplained. Potential explanations have included the presence of secreted survival factors that prevent apoptosis of DNA-damaged hepatocytes and activated stellate cells (eg, Gas6 [281]), reduced tumor surveillance due to decreasing NK cell number and function, and/or the accelerated shortening of telomeres that accompanies progressive fibrosis. How fibrosis promotes HCC is an important, unanswered question. An intimate, yet unexplained, physical link between activated stellate cells, the ductal reaction (clusters of bile ductular expansion), and fibrosis raises important questions about the interactions of these cell types and their roles in fibrosis and regeneration [282].

Reversion of stellate cell activation and the role of senescence — The fate of activated stellate cells as liver injury resolves has been an ongoing focus of investigation. Earlier studies implicated apoptosis, or programmed cell death, as an important pathway by which the liver clears activated stellate cells [129]. An additional pathway has emerged in which activated cells "revert" back to a quiescent phenotype [283-285]. Interestingly, reverted cells harbor the capacity to reactivate more quickly and extensively than cells that have never been activated. Thus, efforts to "turn off" stellate cell activation represent a therapeutic focus in efforts to regress fibrosis. A growing list of "de-activating" factors and epigenetic modifications underlying de-activation/reversion offers clues to potential therapeutic targets, including GATA4, Lhx2, Tcf21, NF1 and others [179,286].

Senescence of hepatic stellate cells has also emerged as an important property of this cell type. The senescence-associated secretory phenotype is a distinct phenotype of activated stellate cells that can both limit the extent of fibrosis and alter the microenvironment, thereby regulating the polarity of macrophages and the ability to resist the development of HCC [287-289]. As noted above, selective clearance of senescent stellate cells is antifibrotic in mouse models of fibrosis and may be a potential strategy for human disease [109]. Interleukin-22, BMP4, and Gremlin 1 have been identified as potential regulators of stellate cell senescence [290,291].

SUMMARY

●Background – Emerging insights into the pathobiology of hepatic fibrosis point to new targets for antifibrotic therapy and heightened prospects for success. (See 'Introduction' above.)

Extracellular matrix (ECM) refers to a group of macromolecules that comprise the scaffolding of normal and fibrotic liver. These include collagens, noncollagen glycoproteins, matrix-bound growth factors, glycosaminoglycans, proteoglycans, and matricellular proteins. A great deal of progress has been made in identifying new members of these families and in understanding how these molecules interact. (See 'Extracellular matrix composition of the normal and fibrotic liver' above.)

Matrix alterations observed during fibrogenesis alter cellular behavior by processes involving cell membrane receptors. (See 'Biologic activity of extracellular matrix' above.)

●Cellular sources of extracellular matrix – A major advance in the understanding of hepatic fibrosis has been the identification of the cellular sources of ECM. The hepatic stellate cell (previously called the lipocyte, Ito, fat-storing, or perisinusoidal cell) is the primary source of ECM in normal and fibrotic liver. (See 'Cellular sources of ECM in normal and fibrotic liver' above.)

●Degradation of extracellular matrix – Fibrosis reflects a balance between matrix production and degradation. The degradation of extracellular matrix is a key event in hepatic fibrosis. Early disruption of the normal hepatic matrix by matrix proteases hastens its replacement by scar matrix, which has deleterious effects on cell function. Macrophages are emerging as important determinants of fibrosis regression. (See 'Degradation of extracellular matrix' above.)

●Hepatic stellate cell activation – Hepatic stellate cell activation is the common pathway leading to hepatic fibrosis. Once activated, hepatic stellate cells release chemokines and other leukocyte chemoattractants, while upregulating the expression of important inflammatory receptors such as ICAM-1, chemokine receptors, and mediators of lipopolysaccharide signaling. (See 'Stellate cell activation, the central event in hepatic fibrosis' above.)