During the fibrotic process, ECM components present an obvious change as the advanced fibrotic hepatic tissue contains approximately 6 times more ECM than the normal, including collagen type I, III, and IV, fibronectin, undulin, elastin, laminin, hyaluronan, and proteoglycans (Bataller and Brenner, 2005). Collagens are mainly produced by hepatic stellate cells and portal fibroblasts (Karsdal et al., 2020). In a carbon tetrachloride (CCl4) induced rat liver fibrosis model, the accumulation of total collagen increased as fibrosis progressed, which is 4.94% of the liver sections at 4 weeks, 8.25% at 6 weeks, and 9.11% at 8 weeks (Vassiliadis et al., 2011). Further measurements revealed that the liver matrix was significantly stiffer in regions approaching fibrillar collagen deposition and returned to near normal rigidity in areas remote from it within the same lobule (Desai et al., 2016). Collagen type I is the most common protein in the body and associated with stiffness across tissues generally (Swift et al., 2013). It’s the most excessive ECM components in fibrotic hepatic tissues (Kisseleva and Brenner, 2013), above, the massive deposition of collagen especially collagen type I is an essential contributing factor in the local stiffening of the ECM in progressive liver fibrosis. In addition to collagen accumulation, the stiffness of the ECM may also be influenced by the orientation of collagen. As Desai et al. (2016) found that the collagen bridges have more dense collagen but also more aligned collagen which can influence stiffness.

Elastic fiber is a compound consisted of amorphous elastin and fibrous fibrillin, which both are mainly made of smaller amino acids including glycine, valine, alanine, and proline (Kielty et al., 2002). It is actively synthesized by hepatic stellate cells and portal fibroblasts in diseased liver (Kanta 2016). In the CCl4-treated liver damage mice model, the increased stiffness of damaged liver ECM is accompanied with decreased elastic fibers. While in the diethoxycarbonyl dihydrocollidine (DDC)-treated livers, the expression of elastic fibers showed no alteration (Klaas et al., 2016). From the study conducted on patients with hepatitis C, the elastin increased from stage F2-F3 (Yasui et al., 2019). These studies indicate that elastic fibers may play a complex role in the stiffness of liver fibrosis, which may be associated with the causes of disease.

Fibronectin (Fn), mainly secreted and synthesized by endothelial cells, Kupffer cells, and hepatocytes, is a large dimeric glycoprotein existing in its insoluble form as a part of the ECM(Younesi and Parsian, 2019). The livers of Fn-null mice present extensive fibrosis, which is accompanied by elevated hepatic matrix rigidity and impaired liver functions. Mechanistically, mutant livers show increased TGF-β bioactivity and TGF-β-regulated lysyl oxidase (LOX) expression, which in turn promote ECM remodeling and stiffening (Iwasaki et al., 2016). These studies indicate a potential association between Fn-mediated control of TGF-β bioavailability and collagen fibril stiffness regulated by LOX.

Proteoglycans (PGs) are formed by glycosaminoglycans (GAGs) bound to a core protein, which can be a part of the ECM (Amendum et al., 2021). PGs have been demonstrated enhancement in remodeling liver tissues (Kimura et al., 2008). The amount of PGs increased significantly in stiffened liver tumors, especially perlecan and decorin (Kovalszky et al., 1993). The knockout of perlecan and decorin results in decreased matrix stiffness in murine model (Costell et al., 1999; Robinson et al., 2017).

MMPs and their inhibitors metalloproteinase tissue inhibitors (TIMPs) are normally in a dynamic balance to keep the expression and deposition levels of the ECM in the liver. While fibrosis occurs, their balance will be interrupted and the profibrogenic effects of TIMP-1 are thought to be mediated via preventing collagen degradation through inhibition of MMPs (Knittel et al., 2000), which changes mechanical properties of the liver matrix. Studies confirmed that increasing expression of TIPMs were detected in patients with liver injury and cirrhosis (Dechêne et al., 2010; Latronico et al., 2016). The imbalance secretion of MMP-9 and TIMPs mediated by rigid matrix perpetuate liver fibrosis (Lachowski et al., 2019). These finding shows that change of the deposition of ECM components in liver matrix, eventually altering the matrix stiffness of liver and affecting the process of hepatic fibrosis.

In addition to the promotion of matrix stiffness by the change of ECM components, the degree of cross-linking of the ECM affects the mechanical characteristics of the matrix.

Enzymatic cross-linking is major mediated by LOX, which increases the cross-linking of collagens and subsequently elevate matrix stiffness (Barker et al., 2012). The LOX family, comprising LOX and four lysyl oxidase-like proteins (LOXL1-4), are secreted copper-dependent amine oxidases, which catalyze the oxidative deamination of primary amine groups into reactive aldehyde condense with other aldehydes or ε-amino groups of selected lysine and hydroxylysine residues to form covalent intra- and inter-molecular crosslinks (Shah et al., 1993; Eekhoff et al., 2018). The stellate cells and portal fibroblasts were the major cellular sources of LOXs in CCl4-induced model of liver fibrosis, and LOX-mediated deoxypyridinoline and pyridinoline cross-links increased correspondingly with liver stiffness (Perepelyuk et al., 2013). Previous studies found that expression of LOX, LOXL1, and LOXL2 are increased in patients with liver fibrosis (Vadasz et al., 2005; Chen et al., 2020). The inhibition of LOX suppressed the cross-linking of collagen and attenuated CCl4-induced advanced hepatic fibrosis (Liu et al., 2016). Besides the expression, the increased activity of LOX family members are also observed in sera of patients with liver fibrosis cirrhosis (Mesarwi et al., 2015). Studies have shown that higher expression of LOX is a predictor of progression and poor prognosis in patients with liver cancers (Lin et al., 2020; Sun et al., 2022). In addition to collagen, LOX1 could also increase the cross-linking of elastin and promote the liver cirrhosis (Zhao et al., 2018).

Transglutaminase (TGs) are calcium-dependent enzymes that catalyze covalent crosslinking between the γ-carboxy-amine group of glutamine residues and the ε-amino group of lysine residues, generating ε-(γ-glutamyl)lysine isopeptide crosslinks (Griffin et al., 2002). The expression of TG is upregulated in experimental liver fibrogenesis (Mirza et al., 1997; Chen et al., 2008). Further study indicated that TG-mediated cross-linking occurs in the liver extracellular matrix during the early, inflammatory, stage of liver fibrosis, whereas cross-linking by pyridinoline occurs mostly later in the fibrotic process⁸. The inhibition of TG has been demonstrated to ameliorate liver fibrogenesis in animal models (Qiu et al., 2007; Tang et al., 2014).

Advanced glycation end products (AGEs) are non-enzymatic modifications of protein crosslinking. It differs from enzymatic process as AGEs can form crosslinks throughout the collagen molecule not only at the N- and C-terminal ends of the molecule (Brownlee et al., 1988). Accumulation of AGEs induces abnormal matrix crosslinking that ultimately leads to a progressive increase in tissue stiffness. The upregulation of AGEs accelerates liver injury and fibrosis (Goodwin et al., 2013). In a rat liver fibrosis model, the contents of AGEs in their blood sera and liver homogenates were increased in the late phase, and it could reflect the fibrosis degree of the rat livers (Gao et al., 2006). In patients with hepatitis C-related cirrhosis, AGEs are powerful parameters associated with the HCC occurrence (Abdel-Razik et al., 2021). These data suggest that the AGEs accumulation may take a function in the late stage of liver fibrosis and the inhibition of AGEs may provide a novel target for late patients with cirrhosis.

The cells’ physical properties and interaction with microenvironment could alter the stiffness of the ECM. In the in vitro model for liver cirrhosis, the stiffness of human hepatoma-derived HepG2 cells was significant increased after treatment with three known induction factors (collagen substrates, alcohol, and CCl4) (Sun et al., 2014). The lung with idiopathic pulmonary fibrosis remains only 44% of the original lung stiffness after decellularized treatment (Booth et al., 2012). These evidence suggest cell stiffness modulates the mechanical property of ECM. Active microrheology (AMR) analysis showed that cells grown in 3D type 1 collagen gels modulate the mechanical microenvironment in a cellular contraction-dependent manner. Inhibition of cellular contractility instantaneously softens the pericellular space and reduces stiffness heterogeneity (Keating et al., 2017). Rocky et al. further demonstrated that promotion of contraction of hepatic stellate cells contribute to the cirrhotic liver (Rockey and Weisiger, 1996). Furthermore, proteoglycans in the ECM can influence the cell contraction based alignment (Chen et al. Sci Reports. 2020). Together, these data suggest a central role for cellular physical property and contractility modifying the ECM stiffness.

TGF-β1, a 25-kDa homodimeric protein, has been involved in hepatic fibrogenesis, regulation of liver cell growth, tumor development, and induction of hepatocellular apoptosis (Gressner et al., 1997; Flisiak et al., 2000). The TGF-β1-promoted HSCs activation is a dominant signal in liver fibrosis (Guvendiren et al., 2014). Increased TGF-β1 expression has been found in patients with hepatic fibrosis (Li et al., 2015; Kotsiri et al., 2016). TGF-β1 was secreted as a large latent complex (LLC), consisting of TGF-β1, latency-associated pro-peptide (LAP), and the latent TGF-β1 binding protein (LTBP-1). LTBP-1 is an ECM protein binding to several other ECM components, including fibrillin-1, Fn, and vitronectin, thereby depositing latent TGF-β1 within the ECM(Annes et al., 2003; Robertson et al., 2015). Cells contraction activates latent TGF-β1 from a preformed ECM, and this TGF-β1 activation increases with augmented matrix stiffness (Wipff et al., 2007; Giacomini et al., 2012). Together, the efficiency of latent TGF-β1 activation depends on the mechanical properties of the ECM.

The small Rho GTPase RhoA/Rho kinase (ROCK)-mediated mechanotransduction pathway is activated by matrix stiffness (Ohashi et al., 2017). The substrate stiffness could directly activate FAK, and induce the downstream cascades of intracellular signals of the RhoA/ROCK pathway (Peng et al., 2019). Specifically, there are two ROCK isoforms (ROCK1 and ROCK2), and the matrix stiffness-activated ROCK isoforms differentially regulated the pathways of RhoA/ROCK1/p-MLC and RhoA/ROCK2/p-cofilin in a coordinate fashion to modulate cell motility (Peng et al., 2019). Recently, Dou et al. demonstrated that matrix stiffening can activate AKT through inducing phosphorylation and accumulation of p300 to the nucleus of HSCs and promote cancer metastasis by elevating the expression of paracrine factors such as CXCL12. In this process, the RhoA pathway of human HSCs was activated by the matrix stiffness from 0.4 kPa to 25.6 kPa and accelerated the downstream nuclear translocation of p300 leading to HSCs activation, which suggests that RhoA is required of substrate stiffness-promoted phosphorylation of p300 (Dou et al., 2018). Inhibition of the RhoA/ROCK pathway has been demonstrated effective in attenuating fibrogenesis in several animal hepatic fibrosis models (Xie et al., 2018; Lai et al., 2019).

Yes-associated protein (YAP) and its close paralog transcriptional coactivator with PDZ-binding motif (TAZ) are activated in HSCs in response to matrix stiffening (Caliari et al., 2016). Immunohistochemical examination of fibrotic liver tissues from patients with hepatitis C found that the myofibroblasts in the fibrotic region with strong nuclear staining of YAP and TAZ (Mannaerts et al., 2015). The stiff substrates promote YAP/TAZ nuclear translocation from the cytoplasm (Dupont et al., 2011), and induce the expression of profibrotic genes and increase the expression of α-SMA and excessive matrix accumulation (Panciera et al., 2017). The inhibition of YAP by verteporfin reduced HSC activation and fibrogenesis (Zhang et al., 2016). RhoA/ROCK signaling is crucial for YAP/TAZ activation, which has been experimentally demonstrated in various systems, either genetically or by using inhibitors (Panciera et al., 2017). Though evidence showed that YAP/TAZ activity in other liver cells including hepatocytes, Kupffer cells and liver sinusoidal endothelial cells are increased and promotes inflammation and fibrosis in multiple models of non-alcoholic steatohepatitis and chemical-induced liver fibrosis, the relationship between matrix stiffness and its influence on these cells needs further elucidation (Zhang et al., 2016). Mechanistic studies suggested that CXCR4 is a critical intracellular signal transducer that relays matrix stiffness signals to inhibit ubiquitin domain-containing protein 1 (UBTD1), which then activates YAP signaling pathway (Yang et al., 2020). And the TAZ could further target gene AREG and CTGF to promote liver fibrogenesis (Yang et al., 2012).

The pathological stage of hepatic fibrosis is generally divided into 5 grades: F0 is non-hepatic fibrosis, F1 is mild hepatic fibrosis, F2 is moderate hepatic fibrosis, F3 is severe hepatic fibrosis, and F4 is cirrhosis. Golo Petzold examined patients with healthy liver or cirrhosis with 2-dimensional shear wave elastography and found that the normal liver stiffness to be 4.93 ± 0.83 kPa and F4 cirrhotic liver to be 13.29 ± 3.27 kPa (Petzold et al., 2019). The higher liver stiffness indicates a higher possibility of clinical complications and a worse prognosis in patients with advanced chronic liver disease (Vergniol et al., 2011; Singh et al., 2013). Even a study found that the increased stiffness precedes fibrosis and potentially promotes fibrotic pathogenesis in an animal liver injury model. This early change appears to result from matrix cross-linking mediated by an imbalanced LOX family (Georges et al., 2007). Cell types including hepatocytes, hepatic stellate cells, and liver sinusoidal endothelial cells, isolated from rats with cirrhosis and cultured on a rigid matrix, presented an altered nuclear morphology compared with those on soft matrix, which suggests mechanotransduction could potentially influence the phenotype of various liver cells and accelerate cirrhosis process (Guixé-Muntet et al., 2020). The expression of mechanotransducer including TAZ and serum osteopontin (OPN) is associated with increased liver stiffness in patients with liver cirrhosis, especially in autoimmune- and alcohol-related cirrhosis (Khajehahmadi et al., 2020). The nuclear translocation of TAZ could target genes AREG and CTGF, and the overexpression of the two genes promotes the liver cirrhosis process (Mitani et al., 2009; Yang et al., 2012; Mohagheghi et al., 2019). In liver cirrhosis, the mechanical transduction pathway RhoA/ROCK was activated and led to increased vascular contractility and portal pressure (Hennenberg et al., 2006; Trebicka et al., 2007), moreover, the matrix stiffness-induced RhoA overexpression also interacts with the cytosolic tyrosine kinase c-SRC and decreases the c-SRC activity to activate HSC to mediate liver cirrhosis (Görtzen et al., 2015). The expression of TIMP-1 is increased in hepatic cirrhosis patients, which is partly caused by matrix stiffness-induced HSC’s elevated exocytosis and secretion of TIMP-1 in a caveolin-1-and dynamin-2-dependent manner (Lachowski et al., 2022).

Portal hypertension is a complication for patients with compensated advanced chronic liver disease or decompensated cirrhosis. Liver stiffness measurements (LSM) offer valuable alternatives to detect and monitor portal hypertension. The thresholds of liver stiffness>15 kPa is highly suggestive of advanced chronic liver disease, and>20–25 kPa indicates a high likelihood of clinically significant portal hypertension (Reiberger 2022). Moreover, LSM has become an important instrument to assess the clinical course of portal hypertension, as patients with baseline LSM >14 kPa have a worse prognosis regarding both development of complications or survival (Stefanescu and Procopet, 2014).

Increased matrix stiffness is a mechanical feature of solid tumors, including liver cancer. HCC cells cultured on stiff matrix showed a significantly higher proliferative index than that on soft matrix (Schrader et al., 2011). A rigid matrix promotes proliferation of HCC cells by stimulating expression of integrin β1, activating the PI3K/Akt pathway, and upregulating VEGF expression (Dong et al., 2014). ECM-derived mechanical signals promote HCC cell invasion and metastasis, which is also mediated by integrin β1 (Zhao et al., 2010). Moreover, higher stiffness-stimulated HCC cells exhibited chemotherapeutic resistance to cisplatin (Schrader et al., 2011), paclitaxel, 5-FU (Liu et al., 2015), weakened oxaliplatin-induced apoptosis (You et al., 2016), and attenuated metformin-inhibited invasion and metastasis (Gao et al., 2020). Studies above indicate that matrix stiffness accelerated the pathogenesis of liver cancer through mediating HCC cells phenotype. Epithelial-mesenchymal transition (EMT) is a phenomenon that HCC cells presents fibroblast-like morphology, cytoskeleton remodeling, protrusive and invasive pseudopodial structure, similar to morphology and great migration capability, which promotes the development of liver cancer (Giannelli et al., 2016). High stiffness could induce EMT in HCC cell and is related to overexpression of Snail, which is converged by three signaling pathways including S100A11 membrane translocation, eIF4E phosphorylation, and TGF-β1 autocrine (Dong et al., 2019). Besides that, the activated NEAT1-WNT/β-catenin pathway is confirmed in substrate stiffness-regulated EMT in HCC cells (Dong et al., 2019). Liver cancer cells with stemness properties (LCSCs) comprise a small portion of HCC cells. This small proportion of HCC cells has great self-renewal and metastatic potential. Research has shown that the stiffnesses of the liver tumor invasive front and core are different, as the core is softer than the invasive front (Sun et al., 2021). Boren Tian et al. demonstrated that a soft matrix increases the stemness of HCC cells, which may be beneficial to maintain dormancy and adaptation to the surrounding environment and then initiate tumorigenesis when the microenvironment is suitable (Tian et al., 2019). Furthermore, the soft spot-enhanced CSCs stemness is related to drug resistance and HCC metastasis (Ng et al., 2021).

Activated HSCs in the liver microenvironment regulate HCC growth by paracrine mechanisms, including secreting growth factors, ECM, and cytokines (Zhang and Friedman, 2012). The mechanical stimulation of ECM can induce HSCs to differentiate into cancer-associated fibroblasts (CAFs) (Yeung et al., 2005). Changwei Dou et al. demonstrated that substrate stiffness-activated HSCs through a RhoA-AKT-p300 signaling pathway (Dou et al., 2018). The other study indicated that stiffness-activated CD36-AKT-E2F3 signaling also takes a role in the HSCs transdifferentiation to CAFs by FGFR1-mediated PI3K/AKT and MEK/ERK signaling (Liu et al., 2021). All above suggest matrix stiffness-promoted HSCs activation is enrolled in HCC growth and metastasis.

Ultrasound elastography (USE) is an imaging technology sensitive to organ mechanical property that was first introduced in the 1990s (Gennisson et al., 2013). Shear wave imaging (SWI) employs a dynamic stress to generate shear waves in the parallel or perpendicular dimensions. There are currently three technical approaches for SWI: 1) point shear wave elastography (pSWE), 2) 1 dimensional transient elastography (1D-TE), and 3) 2 dimensional shear wave elastography (2D-SWE) (Barr et al., 2015). pSWE can be performed on conventional US machine but it could be only stressed in a single focal location. 1D-TE can estimate the stiffness along ultrasonic A-line, but it is not user adjustable and no image-guided. 2D-SWE is the newest SWI method which can test stiffness in a rapid speed in multiple focal zones (Ferraioli et al., 2015). However, there are some common technical limitations for SWI, such as shadowing, reverberation, clutter artifacts, decreased ultrasound signal for deep tissues, or the operator-dependent nature of free-hand ultrasound system (Palmeri and Nightingale, 2011).

Magnetic resonance elastography (MRE) uses a modified phase-contrast imaging sequence to detect propagating shear waves within the liver. It provides parameters with high sensitivity to elasticity, viscosity, and poroelastic properties for the evaluation of structural variations in livers at multiple scales (Singh et al., 2016). MRE has been demonstrated a highly accurate non-invasive diagnostic method to detect and monitor various liver pathophysiologic states (Li et al., 2020). MRE is limited by long operating times, high costs, difficulty in distinguishing tissue stiffening between fibrosis and inflammation or altered perfusion, such as portal venous hypertension in cirrhosis (Salameh et al., 2009).

One of the major drawbacks of common liver stiffness measurements including USE and MRE is that they do not provide cellular-level resolution of stiffness heterogeneity in the liver (Ojha et al., 2022). During the liver disease development, areas containing rich collagens show higher stiffness compared to the surrounding parenchyma (Calò et al., 2020). Thus, changes in local rigidity need to be characterized on a microscopic level to better understand the progression of fibrosis. AFM could detect the liver mechanical properties with high resolution and high force sensitivity. AFM indents the tissue surface with the tip of a cantilever, and causes deformation at the microscopic or nanoscopic level according to the geometric shape and size of the tip. The force response of the tested sample to the applied strain is then calculated as the deflection in the cantilever (Binnig et al., 1986). The force displacement curve is collected from the approach and retraction of the cantilever, which can be fitted with an appropriate contact mechanical model to evaluate the local stiffness of the sample. AFM can also provide topographic information about specific features in the sample, such as the structure of collagen (Amabili et al., 2021). Moreover, AFM has also been used in vitro to determine the stiffness of cells and extracellular protein scaffolds (Norman et al., 2021).