MSP-RON signaling in liver pathobiology and as an emerging therapeutic target: a review of the current evidence

Abstract

The liver is a crucial organ in the human body and is responsible for various functions, including digestion, detoxification, metabolism, and immune response. Proper hepatic function is vital for maintaining systemic homeostasis, and dysregulation of liver signaling pathways contributes to various diseases. Recepteur d’Origine Nantais (RON) is a transmembrane receptor tyrosine kinase that is activated by macrophage-stimulating protein (MSP) and coordinates cell fate decisions through the activation of downstream signaling cascades. As the predominant source of MSP in humans, the liver establishes a liver-specific MSP‒RON autocrine‒paracrine signaling axis that contributes to hepatic regeneration, metabolism, and immune functions. Extensive research has demonstrated that MSP-RON signaling is involved in steatotic liver diseases, hepatitis, cirrhosis, cholestatic liver disease, and liver cancer, highlighting the importance of RON in the development of liver diseases. This review demonstrates the role of the MSP-RON pathway both in maintaining liver homeostasis and in driving disease onset and progression while exploring its signaling mechanisms and therapeutic potential for liver disorders.

Background

Macrophage-stimulating protein (MSP) was first reported in 1976—E. J. Leonard and A. Skeel discovered a protein in human and fetal bovine serum that promotes the chemotaxis, movement, and spread of mouse peritoneal macrophages [1]. Because of its structural similarities to hepatocyte growth factor (HGF), it was designated hepatocyte growth factor-like protein (HGFL) [2, 3]. Compared with mRNAs from other organs, MSP cDNA was noted to hybridize strongly with mRNAs from the liver, indicating that hepatocytes are the primary source [4]. As the exclusive receptor for MSP, Recepteur d’Origine Nantais (RON) is a receptor tyrosine kinase that regulates downstream signaling to control cell fate decisions, metabolism, immune responses, and cancer progression [5, 6]. Research has shown that RON is expressed in both hepatocytes and Kupffer cells of the liver, thereby establishing a specific autocrine‒paracrine loop in the liver [7, 8]. Loss of RON receptor signaling exacerbates high-fat, high-cholesterol diet-induced nonalcoholic steatohepatitis in apolipoprotein E-deficient mice [9]. Transfection of the murine MSP gene into human SCLC cell lines significantly increased liver metastasis, whereas the bone and lung metastatic burdens were not significantly different [10]. Recent evidence has also demonstrated that MSP-RON signaling modulates hepatic fibrogenesis by activating transforming growth factor-beta (TGF-β)-dependent epithelial‒mesenchymal transition (EMT) mechanisms [11, 12]. Collectively, these findings establish the MSP-RON pathway as an important regulator of hepatic homeostasis and liver disease. This review focuses on the roles of MSP-RON signaling in various liver disorders and the pivotal role of this pathway in maintaining liver homeostasis. Given the established evidence that RON is a promising therapeutic target in oncology [13–15], we discuss its potential for pharmacological intervention in hepatic disorders.

Overview of the MSP-RON signaling pathway

Components of the MSP-RON signaling pathway

The MSP-RON signaling pathway is primarily composed of the ligand MSP, the receptor RON, and downstream effectors triggered by receptor activation. The genes encoding human MSP and RON are both located in the chromosome 3p21 region [16]. MSP expression in the liver is primarily controlled by transcription factor-coactivator complexes. Hepatocyte nuclear factor 4 (HNF4) is essential for the liver-specific expression of MSP, while nuclear transcription factor Y (NF-Y) cooperates with HNF4 to activate its transcription [5, 17]. It is initially synthesized by the liver as an inactive precursor (pro-MSP) and is secreted into systemic circulation [2, 4]. Multiple members of the trypsin family of serine proteases capable of promoting MSP maturation have been identified across diverse tissues, such as hepatocyte growth factor activator (HGFA) [18, 19], membrane-type serine protease 1 (MT-SP1/matriptase) [20], and hepsin (Fig. 1a) [21]. Proteolytic cleavage generates mature MSP as a disulfide-bonded α/β heterodimer, a process fundamentally required for its biological functionality [22, 23].

Fig. 1.

Schematic of MSP, RON, and RON variants. a MSP and its maturation process. b Mature RON. c All 20 exons encoding RON and their corresponding structure. d RON and its major variants. Exon 11 deletion prevents RON from maturing from RONΔ165 (E11) and RONΔ155 (E5/6/11) into the α/β two-chain form, leading to protein accumulation in the cytoplasm. HL hairpin loop, K1-4 kringle domains 1–4, HGFA hepatocyte growth factor activator, MT-SP1 membrane-type serine protease 1, SEMA semaphoring, PSI plexin-semaphorin-integrin, IPT immunoglobulin-like plexin and transcription, TM transmembrane, TK tyrosine kinase

RON is widely expressed on various epithelial cells [24], tissue macrophages [7, 8, 25, 26], and multiple types of cancer cells [5, 27–30]. Human RON contains 20 exons that encode pro-RON, an approximately 180-kDa glycosylated single-chain protein consisting of 1,400 amino acids (Fig. 1b) [31]. Similar to MET, pro-RON needs to undergo proteolytic cleavage to become a biologically active mature RON [31]. The structure of it closely resembles that of MET, sharing key domains such as the SEMA domain; however, their roles in cellular signaling differ significantly [32]. Mature RON consists of a 35-kDa α-chain and 145-kDa β-chain; furthermore, the α-chain of RON includes the N-terminus of the SEMA domain, whereas its β-chain contains the C-terminus of the SEMA domain, a PSI motif, four IPT domains, a transmembrane region, an intracellular kinase domain, and a carboxy-terminal tail (Fig. 1c) [31]. MSP and RON are highly conserved across species [4, 33–35]. In mice, a primary model for investigating MSP-RON signaling biology, both genes are located on chromosome 9qF1 [16]. Transgenic mice overexpressing human wild-type RON developed multiple lung adenomas. This suggests functional similarity between human and rodent RON, given its role in human cancers including lung cancer [36]. However, human and murine RON proteins exhibit 75.42% sequence similarity [37], suggesting potential structural variations between these orthologs. Thus, certain therapeutic antibodies targeting human RON show limited binding to murine RON [38], which impedes the development of RON-directed antibody therapeutics.

Canonical MSP-RON signaling pathway

The canonical MSP-RON signaling cascade is initiated through three sequential steps: ligand-induced receptor dimerization upon MSP binding to the RON extracellular domain, trans-autophosphorylation of specific tyrosine residues within the intracellular kinase domain, and recruitment of adaptor proteins to activate downstream effectors [5]. MSP contains two structurally distinct receptor-binding domains. The β-chain of MSP contains a high-affinity binding site that specifically interacts with the RON SEMA domains. Figure 2 illustrates the structure of the MSP β-chain (Fig. 2a), RON SEMA domains (Fig. 2b), and partial structures of the MSP-RON complex, with a focused annotation of critical residues involved in the binding interface (Fig. 2c) [31, 39]. The low-affinity site in the α-chain of MSP subsequently binds to RON [23, 31]. This sequential engagement triggers conformational rearrangements in the RON β-chain, leading to the autophosphorylation of Tyr1238 and Tyr1239 within the intracellular kinase domain [40] and then promoting the phosphorylation of Tyr1353 and Tyr1360 located on the C-terminal docking site on the intracellular tail [41]. The docking site recruits downstream signaling effectors, thereby activating diverse intracellular cascades, including the Src [42, 43], Ras/Erk [44, 45], phosphoinositide 3-kinase (PI3K)/AKT [46, 47], β-catenin [48, 49], NF-κB [7, 50, 51], JAK/STAT [52, 53], and transforming growth factor (TGF) β/Smad [5, 54] pathways, which coordinate critical cellular responses (Fig. 3).

Fig. 2.

Crystal structure of the MSP–RON complex. The image includes color-coded domains for the MSP β-chain (forest green) and RON domains, including SEMA (sky blue), PSI (deep purple), and IPT₁ (pink). a Crystal structure of the MSP β-chain (PBD ID: 2ASU). Residues involved in the MSP–RON interaction are marked in red. b Crystal structure of RON-SEMA-PSI (PBD ID: 4FWW). Residues involved in the MSP–RON interaction are marked in red. c Crystal structure of the MSP-RON complex (PBD ID: 4QT8). The zoomed-in view shows the spatial arrangement of residues at the interaction interface

Fig. 3.

Overview of MSP-RON signaling in epithelial and cancer cells. RON exhibits various forms of activation, including MSP-induced RON dimerization, spontaneous dimerization due to RON overexpression, spontaneous activation caused by RON variants and point mutations, and crosstalk with other RTKs. RON activation induces the activation of downstream signaling cascades and various biological processes. GRB2 growth factor receptor-bound protein 2, SOS son of sevenless, STAT signal transducer and activator of transcription, GS3Kβ glycogen synthase kinase 3β, mTOR mammalian target of rapamycin, p70S6K p70S6 kinase, VHL von Hippel–Lindau, HIF1 hypoxia-inducible factor 1, TFs transcription factors, RSK2 p90 ribosomal S6 kinase 2, LKB1 liver kinase B1, AMPK AMP-activated protein kinase, USF1 upstream transcription factor 1, SHP small heterodimer partner, HNF4-α hepatocyte nuclear factor 4α, IKK inhibitor of κB kinase, EMT epithelial–mesenchymal transition. Figure created with BioRender.com

Noncanonical MSP-RON signaling pathway

In some instances, RON can be aberrantly activated in a ligand-independent manner, typically manifested as its overexpression [55, 56], the production of variants [57, 58], and crosstalk with other RTKs. These forms of the noncanonical MSP-RON signaling pathway are key in tumorigenesis, malignant progression, angiogenesis, and chemoresistance in many cancers [5, 27–30].

RON overexpression on certain cancer cell surfaces results in spontaneous dimerization, leading to continuous phosphorylation [5]. Notably, increased RON expression is associated with the production of its variants [59]. The three main mechanisms underlying the production of RON variants are alternative mRNA splicing, protein truncation, and alternative transcription [5, 27, 60]. The identified variants have been named on the basis of the predicted molecular weight of the protein and the exon regions affected by alternative mRNA splicing. Thus far, the following major variants have been characterized (Table 1): RONΔ170(E19) [61, 62], RONΔ165(E11) [57, 63–66], RONΔ165(E2) [67], RONΔ160(E5/6) [30, 49, 57, 62, 64, 65, 68], RONΔ160(E2/3) [69], RONΔ155(E5/6/11) [57, 64, 65], RONΔ110 [27, 70], RONΔ90 [71], RONΔ85 [72], and RONΔ55/SF-RON [58, 73–78].

Table 1.

Summary of structural bases and functional characteristics of human RON variants

RON Variant	Exons Involved	Structural Characteristics	Key Functional Characteristics	References
RON170	Deletion in exon 19	Frameshift mutation generating a premature termination codon, resulting in deletion of the kinase domain and multifunctional docking site.	Lacks kinase activity.	[61, 62]
RON165E11	Deletion in exon 11	Deletion of β-chain IPT domain 4 impairs proteolytic processing, causing intracellular accumulation of single-chain RON165.	Exhibits constitutive phosphorylation.	[57, 63–66, 217]
RON165E2	Deletion in exon 2	Partial deletion of the SEMA domain.	Lacks tyrosine kinase activity but constitutively activates PI3K/AKT via PTEN hyperphosphorylation.	[67]
RON160E5E6	Deletion in exon 5,6	Deletion of β-chain IPT domain 1.	Exhibits constitutive phosphorylation.	[30, 49, 57, 62, 64, 65, 68, 218]
RON160EP5P6	Partial splicing of exons 5 and 6	Partial deletion of β-chain IPT domain 1.	Exhibits constitutive phosphorylation.	[73, 219]
RON160E2E3	Deletion of exon 2, 3	Partial deletion of the SEMA domain.	Lacks tyrosine kinase activity but constitutively activates PI3K/AKT.	[69]
RON155	Deletion in exon 5,6,11	Deletion of β-chain IPT domain 1 and 4, causing intracellular accumulation of single-chain RON155.	Exhibits constitutive phosphorylation.	[57, 64, 65]
RON110	Insertion between exons 5 and 6	Trypsin-like protease cleavage at residues Arg631–Lys632 in the extracellular domain of the β-chain.	Exhibits constitutive phosphorylation.	[27, 70]
RON90	Deletion in exon6	Frameshift and premature termination codon in exon 7 generate a truncated soluble protein with an intact ligand-binding domain.	Competitively inhibiting MSP-mediated RON activation.	[71]
RON85	Insertion between exons 5 and 6	Frameshift and premature termination codon generate a truncated soluble protein with an intact ligand-binding domain.	Competitively inhibiting MSP-mediated RON activation and blocking dimerization/signaling of full-length RON and variants.	[72]
RON55/sf-RON	Alternative transcription in exon 11	Lacking most extracellular domains.	Exhibits constitutive phosphorylation.	[74, 83, 220]

Most RON variants are generated via alternative mRNA splicing and have constitutive kinase activity (Fig. 1d). For example, the deletion of exon 11 results in the absence of 49 amino acids located in the fourth IPT domain in the extracellular sequence of the its β-chain, allowing RONΔ165(E11) to undergo constitutive phosphorylation and maintain continuous activation of downstream signaling cascades [57, 63]. However, the variant RONΔ170(E19), with deletion of exon 19, which encodes 46 amino acids leading to elimination of a multifunctional docking site and truncation of the C-terminus, lacks kinase activity [62]. Although the mechanism underlying alternative splicing of RON mRNA remains unclear, the heterogeneous nuclear ribonucleoprotein (hnRNP) family may be widely involved in this process [79–82]. The splicing of exon 11 is controlled by silencers and enhancers located in exons 11 and 12, and hnRNP H promotes the skipping of exon 11 by binding to the exon splicing silencer (ESS) within exon 11 [81]. Moreover, SRSF1 (SF2/ASF) in exon 12 facilitates exon 11 skipping by binding to the adjacent exon splicing enhancer (ESE) [79]. In contrast, hnRNP A1 binds to ESS, inhibiting the interaction between SRSF1 and ESE and promoting exon 11 inclusion [66]. In addition, hnRNP A2/B1 can increase exon 11 skipping; however, the underlying mechanisms must be characterized [80]. A recent study revealed that hnRNP A1 is also critical for promoting exon 5/6 inclusion, reducing RONΔ160(E5/6) production [82].

RONΔ110 is a typical truncated RON variant produced by digestion with the RONE5/6in protease [70]. RONE5/6in is derived from a transcript featuring a 20-amino acid insertion between exons 5 and 6. This modification renders the protein highly susceptible to protease digestion, resulting in the formation of the truncated variant RONΔ110. Notably, two other soluble truncated RON variants exist [71, 72]. The insertion and deletion at exon 6 lead to a frameshift mutation, followed by the generation of a premature termination codon, resulting in RONΔ85 and RONΔ90, respectively [71, 72]. The absence of part of the extracellular domains and the entire transmembrane and intracellular domains renders these proteins unable to anchor to the surface of the cell membrane and then become soluble splice variants of RON. However, the retention of the SEMA and PSI domains and part of the first IPT domain of the RON extracellular domain allows them to bind competitively to MSP, inhibiting MSP-RON signaling [71, 72].

SF-RON/RONΔ55, one of the most widely studied RON variants, is produced through alternative transcription, which initiates at an intragenic promoter located in exon 10, resulting in a truncated RON lacking most extracellular amino acids [74, 83]. The generation of two distinct transcripts may be associated with the methylation status of two CpG islands in the proximal classic RON promoter [83]. Hypomethylation of CpG island 1 and methylation of CpG island 2 are associated with the promotion of FL-RON transcription. In contrast, hypermethylation of CpG island 1 and strong methylation of CpG island 2 correlate with the absence of FL-RON and the presence of SF-RON [83]. Moreover, SF-RON has constitutive kinase activity mediated by spontaneous dimerization [83, 84]. The expression distribution, formation mechanism, and biological characteristics of RON variants warrant further research.

RON can interact with other RTKs in many tumor cells, including MET [85–87], epidermal growth factor receptor (EGFR) [88], platelet-derived growth factor receptor (PDGFR) [89], and insulin-like growth factor (IGF) 1 receptor (IGF1R) [90, 91]. For example, RON can form heterodimers with MET on the cell surface, producing synergistic effects through mutual phosphorylation [85]. Therefore, HGF (the ligand of MET) can induce the phosphorylation of RON and vice versa [85]. In addition, RON variants can crosstalk with full-length RON [62, 83], other variants [62, 72], and even other tyrosine kinase receptors [86, 92]. This mechanism expands the cellular functions and regulatory manners of RON signaling (Fig. 3). Notably, MET [85, 87], EGFR [93], PDGFR [94], and IGF1R [95] also play crucial roles in various physiological functions and diseases of the liver. Thus, the crosstalk between these receptors may be a potential mechanism underlying RON signaling–associated liver homeostasis regulation. However, research on RON crosstalk with other RTKs has thus far been conducted primarily in oncology [5, 96]. The interactions among these receptors in the liver warrant further investigation. Understanding the expression and biological properties of RON and its variants in different cells, as well as RON’s interactions with other receptors, is essential for the development of rational therapeutic strategies [5].

MSP-RON signaling in liver repair and regeneration

MSP-RON signaling is strongly involved in tissue repair. RON is expressed in wound healing response components, including keratinocytes, macrophages, and capillaries [3, 97]. In wound exudates, more pro-MSP is cleaved into mature MSP, promoting RON activation [97]. This activation enhances fibroblast migration and significantly upregulates the mRNA expression of type I and III collagen, as well as matrix metalloproteinase 1, all of which are major markers of wound healing [3, 98]. In addition, MSP can also regulate α6β4 integrin-mediated keratinocyte adhesion and migration during wound healing by activating the PI3K-AKT pathway [99].

The liver has a strong regenerative capacity and can restore its function by repairing and regenerating damaged tissues after injury [100, 101]. The MSP-RON pathway is also involved in liver damage repair. Enzo et al. reported that RON activation can promote the division, migration, invasion, scattering growth, and polarization of hepatic progenitor cells [102]. RON activation can also induce complex morphological changes in hepatic progenitor cells, including the formation of cord-like branching structures similar to ducts [102]. Some studies have demonstrated the dynamic modulation of MSP expression levels during liver injury. Compared with those in sham-operated controls, MSP mRNA levels in rats after 70% partial hepatectomy were transiently increased 93% within 1 h and returned to baseline within 2 h [103]. However, the early upregulation of mRNA expression was not directly associated with DNA synthesis, suggesting that MSP is involved in the initial stage of liver regeneration and promotes the entry of liver cells into the cell cycle. In addition, in fulminant liver failure, massive liver cell necrosis leads to a decrease in MSP synthesis, which may be detrimental to early liver repair by impairing the phagocytosis of macrophages [104]. This may be because the liver cells remaining after liver failure exhibit extensive necrosis and functional impairment, in contrast to the liver cells remaining after partial hepatectomy, which retain normal function [103, 104]. An increase in MSP expression is a compensatory mechanism that may be impeded in severely damaged livers. In general, the activation of the MSP-RON signaling pathway contributes to the early stage of liver regeneration, and early administration of MSP may be a feasible strategy for alleviating liver injury. Although MSP-RON signaling engages in extensive crosstalk with key hepatic developmental regulators like the HGF-MET signaling [85, 105], direct functional evidence for its role in liver development and growth remains lacking, warranting further investigation.

MSP-RON signaling in modulating liver immune responses

Modulation of immune cell function by MSP-RON signaling

MSP-RON signaling is a basic component of liver innate immunity (Fig. 4a) [106]. MSP was originally described as a serum protein that stimulates the movement, chemotaxis, and spread of peritoneal resident macrophages. It enhances the ability of these macrophages to bind via αMβ2 integrin to intercellular adhesion molecule 1 (ICAM1) and to phagocytose C3bi-coated red blood cells [107]. Notably, Atsushi et al. analyzed the phenotype of exudative macrophages on consecutive days after a stimulus was elicited and reported a continuous increase in RON expression. Thus, RON may be a marker for the terminal differentiation of macrophages [108]. In addition to being expressed in peritoneal resident macrophages, RON is expressed in certain tissue macrophages, such as Kupffer cells [7, 8], osteoclasts [25], and microglia [26]. Kupffer cells play a critical role in liver tissue homeostasis and disease [109]. RON signaling activation can result in macrophage polarization from the M1 to the M2 phenotype [2, 110]. On the basis of their activation state and function, macrophages can be divided into M1 (classically activated) and M2 (alternatively activated) types [111]. M1 macrophages, which are characterized by high inducible nitric oxide synthase (iNOS) expression, enhance the inflammatory response and pathogen-killing ability. In contrast, M2 macrophages, characterized by high ARG1 expression, promote inflammation resolution, tissue repair, and tumor growth [112]. In macrophages, RON inhibits iNOS expression through the PI3K-AKT pathway, thereby reducing nitric oxide synthesis [113]. Moreover, MSP activates the MAPK pathway, upregulates FOS expression, and enhances the binding of FOS to the AP-1 site in the ARG1 promoter, ultimately leading to increased ARG1 expression [110]. MSP-RON signaling can also inhibit type I interferon (IFN) expression through the regulation of macrophage polarization and the toll-like receptor (TLR) 4 signaling pathway, affecting innate immunity in humans (Fig. 4b) [114, 115].

Fig. 4.

MSP-RON signaling in the modulation of liver immunity. a Overall landscape of MSP-RON signaling in regulating interactions among liver immune cells. Most MSP, produced by the liver, circulates throughout the body via the bloodstream to regulate systemic RON signaling. MSP secreted by hepatocytes can also act on RON receptors expressed by hepatocytes themselves, forming an autocrine loop. Moreover, because tissue-resident macrophages express FL-RON, MSP can modulate Kupffer cell function, thereby influencing the activity of other immune cells and cytokine release. In T cells, RON is expressed as SF-RON, which is critical for T-cell function regulation. b MSP-RON signaling in Kupffer cells. MSP-mediated activation of full-length RON is the main form of RON signaling in macrophages. The MSP-RON pathway plays crucial regulatory roles in macrophage functions, including chemotaxis, phagocytosis, polarization, and cytokine release. PDL1 programmed death-ligand 1, MCP1 monocyte chemoattractant protein-1, MIP2 macrophage inflammatory protein-2, iC3b inactivated C3b, ICAM1 intercellular adhesion molecule 1, CR3 complement receptor 3, PKCζ protein kinase c zeta, iNOS inducible nitric oxide synthase, ARG1 arginase 1, CIITA class II major histocompatibility complex transactivator, MHCII major histocompatibility complex class II, LPS lipopolysaccharide, TLR4 toll-like receptor 4, MYD88 myeloid differentiation primary response 88, IRAK1 interleukin-1 receptor-associated kinase 1, TRAF6 tumor necrosis factor receptor associated factor 6. Figure created with BioRender.com

MSP-RON signaling can influence the function of immune cells (other than macrophages) through its effects on cytokine and chemokine secretion in macrophages [116]. For example, RON may facilitate neutrophil mobilization and activation. Targeted deletion of the RON tyrosine kinase domain (RON TK^−/−) in murine sepsis models significantly reduces the secretion of mediators critical for neutrophil recruitment, including monocyte chemoattractant protein 1 (MCP1/CCL2), macrophage inflammatory protein 2 (MIP2/CCL3), and interleukin-6 (IL-6) [117]. Furthermore, neutrophils from TK^−/− mice demonstrated a significant decrease in spontaneous oxidative burst activity in vitro [117]. Similarly, reduced neutrophil migration to and translocation within the liver are accompanied by considerable increases in liver tissue injury and systemic organ colonization [117]. In the context of endotoxin-induced innate immune responses, RON activation reduces the production of IL-12 by macrophages, which is crucial for IFN-γ production in natural killer cells [52]. Atakan Ekiz et al. reported that after 7 h of MSP stimulation, the mRNA levels of programmed death ligand 1 (PDL1), PDL2, B7H3, and CD80 increased significantly in macrophages [118]. Furthermore, MSP reduces IFN-γ-induced STAT1 phosphorylation and CIITA expression, reducing the levels of MHC class II molecules on the cell surface and potentially diminishing the T-cell activation capacity [52].

RON expression in the immune system was initially believed to be restricted to certain types of tissue macrophages. However, recent studies have shown that SF-RON mRNA is also expressed in CD4⁺ and CD8⁺ T cells [119]. Using embryonic stem cell-based homologous recombination gene targeting, researchers generated SF-RON isoform-deficient mice [119, 120]. These animals presented a marked increase in CD4⁺ and CD8⁺ T cells accompanied by a considerable decrease in regulatory T cells (Tregs) in the spleen [119]. SF-RON, a truncated variant of RON, has strong intrinsic tyrosine kinase activity [74] with a protective role in T-cell-mediated acute liver injury [120]. T-cell activation with either concanavalin A (an APC-dependent activator) or an anti-CD3/T-cell receptor (TCR) antibody (an APC-independent activator) leads to SF-RON-dependent inhibition of IFN-γ production, suggesting that SF-RON might suppress T-cell effector function directly or indirectly [120]. Furthermore, the inhibitory effects of SF-RON on IFN-γ production indicate RON’s potential to regulate the balance between type I and II immune responses [120, 121].

MSP-RON signaling in liver inflammation

MSP-RON signaling activation negatively regulates the production of inflammatory factors and plays a major role in anti-inflammatory processes [24, 52, 122, 123]. RON signaling forms the molecular basis of MSP-mediated anti-inflammatory effects, primarily by blocking the transcription of inflammatory factor genes [124] and inhibiting the function of inflammatory signaling molecules, such as nuclear factor κB (NF-κB) [125, 126]. MSP-RON signaling activates the LKB1-AMP-activated protein kinase (AMPK) signaling pathway, further inducing the transcription of orphan nuclear receptor small heterodimer partner (SHP). By blocking the entry of p65 into the nucleus and preventing the ubiquitination and degradation of TRAF6, SHP effectively inhibits NF-κB signaling and reduces the TLR-mediated inflammatory response [127].

Given the anti-inflammatory role of RON, TK^−/− mice typically display exaggerated inflammatory responses across various inflammation models [24, 52, 122, 123]. Paradoxically, the inhibition of RON signaling may contribute to liver protection in endotoxin-induced acute liver failure. Leonis et al. reported that the loss of RON kinase activity led to a protective effect in a lipopolysaccharide (LPS)/GalN-induced acute liver failure model [128]. Compared with control TK^+/+ mice, TK^−/− mice presented lower serum levels of aminotransferase (a liver injury marker) after LPS/GalN injection. Histological analysis further revealed reduced hepatocyte apoptosis in these mice [128]. In contrast, the levels of tumor necrosis factor-α (TNF-α), the key inflammatory mediator of this model [129, 130], were significantly elevated in TK^−/− mice after LPS/GalN injection [128]. This result is supported by the observation that, compared with control media, conditioned media from LPS-treated TK^−/− Kupffer cells exhibited greater toxicity to hepatocytes [8]. Notably, TK^−/− hepatocytes demonstrated increased resistance to death compared with TK^+/+ hepatocytes, indicating that RON plays a role in both epithelial and inflammatory cell compartments in acute liver injury regulation [8]. Further study revealed that the lack of RON relieves the inhibition of certain survival signals, including NF-κB, B-cell lymphoma 6, and the suppressor of cytokine signaling, reducing hepatocyte sensitivity to Kupffer cell products and increasing hepatocyte resistance to death [8, 53, 128]. Researchers verified this by using the Alb-Cre and Lys-Cre systems to construct hepatocyte-specific and myeloid cell-specific RON knockout mouse models and conducted LPS stimulation experiments [8]. The results demonstrated that the loss of RON signaling in hepatocytes alleviated liver injury and significantly prolonged survival in mice, whereas the absence of it in myeloid cells led to the opposite effects [8].

MSP-RON signaling and infections

Considering its complex regulatory roles in the immune system and inflammation, the MSP-RON signaling pathway may be involved in the infection processes of different pathogens, such as HIV-1 [131, 132] and Listeria monocytogenes [133]. In murine models of polymicrobial peritonitis induced by cecal ligation and puncture (CLP) and systemic Listeria monocytogenes infection, compared with TK^−/− mice, TK^+/+ mice presented a significantly reduced hepatic bacterial burden and increased survival rates [117, 133].

Viral hepatitis, which leads to liver inflammation, is a major issue in health care worldwide. While the interplay between hepatitis viruses and RON signaling remains unexplored, some evidence suggests that RON may modulate virus‒host immune interactions and contribute to chronic viral hepatitis progression through its immunoregulatory functions, including but not necessarily limited to regulating cytokine balance and macrophage polarization, as previously described [24, 52, 122]. Given the prevalence of hepatitis B virus (HBV)-related chronic hepatitis, we exemplify this mechanism through HBV pathogenesis. In HBV infection, effective antiviral defense by the immune system depends on the coordination of antiviral mediators produced by hepatocytes and Kupffer cells, including IFN-α/β, IFN-γ, IL-18, and TNF-α [134–139]. Through complex interactions with the host, HBV disrupts the liver’s natural immune response by interfering with antiviral mediator production [134–139], leading to persistent viral infection. Several studies have indicated that HBV promotes M2 macrophage activation [140, 141]. Moreover, increased M2 macrophage infiltration has been associated with the progression of HBV-associated liver diseases, such as liver fibrosis and hepatocellular carcinoma (HCC) [142, 143]. As previously discussed, MSP-RON signaling critically modulates hepatic cytokine profiles and macrophage polarization. Notably, RON’s involvement in oncogenic pathways of other viruses also suggests potential mechanisms in viral hepatitis-associated carcinogenesis [78, 144]. In conclusion, elucidating the role of RON signaling in viral hepatitis pathogenesis and progressive liver disease is particularly interesting and promising.

MSP-RON signaling in liver fibrosis

RON mediates liver fibrosis progression by promoting epithelial‒mesenchymal transition (EMT) [145], the process through which polarized epithelial cells with intercellular connections transform into motile, invasive mesenchymal cells. EMT plays a crucial role in the fibrosis of various organs [146–148], contributes to liver fibrosis and serves as an important source of liver myofibroblasts [149, 150].

In a previous study, the human renal proximal tubular epithelial cell line HK-2 and the interstitial fibroblast line NRK49F transiently transfected with RON demonstrated significant EMT, as evidenced by the upregulation of the fibrosis marker proteins N-cadherin, vimentin, TGF-β, α smooth muscle actin (SMA), and fibronectin [43]. Similar results were obtained by Weng et al. [11] In the normal human liver cell line, RON, ERK1/2, and Smad2/3 phosphorylation increased (in the indicated order of magnitude) within 1 h of stimulation with exogenous MSP. After 48 h, N-cadherin, vimentin, α-SMA, and type I collagen levels increased, whereas E-cadherin levels decreased. Notably, galunisertib, a TGF-β pathway inhibitor, inhibited this effect [11]. In summary, MSP-RON signaling can promote the transformation of hepatocytes into mesenchymal cells through the TGF-β-related EMT pathway, promoting liver fibrosis.

Weng et al. reported that cirrhosis patients with higher MSP expression had poorer clinical outcomes [11]. These patients presented decreased platelet counts and estimated glomerular filtration rates, increased blood creatinine, urea nitrogen, and uric acid levels, increased international normalized ratios, and prolonged prothrombin times. Notably, serum MSP levels demonstrated comparable accuracy [area under the curve (AUC) = 0.769] to Model for End-Stage Liver Disease (AUC = 0.825) and Child‒Pugh classification (AUC = 0.799) scores in predicting poor prognosis in patients with cirrhosis [11]. A murine model of CCl₄-induced liver fibrosis demonstrated elevated liver RON expression. Compared with control mice, mice treated with antibodies against MSP and small-molecule inhibitors of RON exhibited significant reductions in liver fibrosis [11]. Thus, MSP is a potential diagnostic and prognostic marker for cirrhosis. Liver fibrosis is a critical stage in the progression from chronic liver disease to cirrhosis, making it a key determinant of disease prognosis. Targeting the MSP-RON signaling pathway is a promising strategy for the reversal of liver fibrosis and cirrhosis.

Metabolic functions of MSP-RON signaling in the liver

MSP-RON signaling-mediated regulation of glucose and lipid metabolism

The AMPK pathway is not only a key metabolism regulator but also an important participant in anti-inflammatory processes [151]. It also plays a significant role in steatotic liver diseases pathogenesis [152]. One study evaluated the potential role of MSP in liver gluconeogenesis in primary human hepatocytes (PHHs) exposed to cAMP/Dex treatment [153]. The results showed that MSP negatively regulates the expression levels of two key gluconeogenesis enzymes, namely, PEPCK and Glc-6-Pase, and significantly inhibits gluconeogenesis in PHHs. This mechanism depends on the AMPK-SHP pathway [153]. MSP treatment significantly increased SHP mRNA expression within 3 h by increasing gene promoter activity in a time- and dose-dependent manner [153]. MSP increased the phosphorylation of AMPK and acetyl-CoA carboxylase (a downstream AMPK target) within 15 min of treatment. However, pretreatment with compound C (an AMPK inhibitor) or Ad-siSHP reversed MSP-mediated reductions in cAMP/Dex-mediated glucose production [153]. These results reconfirm that the MSP-RON-AMPK-SHP pathway is crucial for the regulation of liver gluconeogenesis in hepatocytes [153]. Liver gluconeogenesis and hepatic glucose production are closely related regulatory processes. Glucagon and glucocorticoids exert their effects through the cAMP-dependent pathway, whereas insulin exerts its effects through the PI3K pathway. Thus, the crosstalk between RON and IGFR and the crucial effects of classical MSP-RON signaling on PI3K may represent potential forms of RON signaling involved in liver glucose metabolism [90].

In 1998, Jorge et al. developed MSP^−/− mice to assess the role of it in growth [154]. Compared with wild-type mice, MSP^−/− mice demonstrated lipid accumulation in the cytoplasmic vacuoles of hepatocytes, even when both groups of mice were fed a normal diet [154]. Thus, MSP may play a significant role in liver lipid metabolism. Because palmitic acid (PA) is the major fatty acid in circulation, researchers have exposed HepG2 (a human liver cancer cell line) to PA with or without MSP to assess the effects of it on PA-induced lipid accumulation [155]. MSP stimulation was noted to significantly reduce lipid accumulation; this effect was potentially dependent on the AMPK pathway [155]. Treatment with MSP or AICAR (a well-established AMPK activator) resulted in considerable reductions in the expression of key lipogenic proteins, sterol regulatory element-binding protein 1, fatty acid synthase, and peroxisome proliferator-activated receptor α [155]. However, the inhibition of AMPK by Comp. C significantly attenuated this antilipogenic effect of MSP (or AICAR). In contrast, MSP or AICAR treatment considerably reversed PA- and LPS-induced inhibition of key fatty acid oxidation genes, such as aconitase, carnitine palmitoyl transferase 1, and peroxisome proliferator-activated receptor γ coactivator 1α [155]. Therefore, MSP can improve liver lipid metabolism by activating the AMPK pathway, enhancing fatty acid oxidation, and inhibiting adipogenesis and the inflammatory response [155].

In summary, MSP-RON signaling positively regulates liver glucose and lipid metabolism (Fig. 5). Furthermore, in breast cancer, RON promotes MAPK-dependent MYC expression and β-catenin-dependent SREBP2 expression, both of which are critical for tumor metastasis and recurrence [156]. SREBP2 is a key cholesterol biosynthesis regulator [157]. MYC can also regulate liver metabolic homeostasis through mechanisms such as promoting glycolysis and inducing endoplasmic reticulum (ER) stress [158, 159]. Therefore, SREBP2 and MYC may represent potential underlying mechanisms in RON-mediated hepatic metabolic regulation.

Fig. 5.

MSP-RON signaling in liver metabolism regulation. RON primarily modulates metabolism by regulating the transcription and activity of key enzymes involved in carbohydrate and lipid metabolism, predominantly facilitating glycemic and lipid reduction. PFKFB3 phosphofructokinase-2/fructose-2,6-bisphosphatase 3, GYS glycogen synthase, Fas fatty acid synthase, SREBF1 sterol regulatory element-binding protein 1, ACO aconitase, CPT1 carnitine palmitoyltransferase 1, PGC-1α peroxisome proliferator-activated receptor γ coactivator 1α, PECK phosphoenolpyruvate carboxykinase, Glc-6-pase glucose-6-phosphatase. Figure created with BioRender.com

MSP-RON signaling in steatotic liver diseases

In response to an updated understanding of the disease and the need to reduce stigma, international liver societies have recently revised the definition of NAFLD and renamed it metabolic dysfunction-associated steatotic liver disease (MASLD) [160]. Although this change is controversial, the research outcomes related to NAFLD obtained before the aforementioned revision remain valid [161]. NAFLD is a complex disease influenced by various mechanisms involving metabolic, genetic, environmental, and gut microbial factors [162]. The disease can be divided into two histological categories according to the presence of features of hepatocellular injury: nonalcoholic fatty liver (NAFL) and nonalcoholic steatohepatitis (NASH) [163, 164]. Further progression of these histological categories leads to the development of fatty liver fibrosis, cirrhosis, and even liver cancer [165]. Notably, RON is involved in almost all steps of this process.

NASH, which is generally considered the liver manifestation of metabolic syndrome, is characterized by lipid accumulation within hepatocytes, which ultimately induces lipotoxicity and inflammatory injury [166, 167]. However, research on the involvement of the MSP-RON signaling pathway in NASH has yielded conflicting results. Apolipoprotein E (APOE)-knockout (KO) mice are frequently used as metabolic syndrome models in NASH research [168]. Some researchers have employed double KO (DKO) mice lacking both APOE and RON (in which the ligand binding domain of RON is deleted) to investigate the role of RON in NASH. These results demonstrated that MSP-RON signaling plays a protective role in NASH [9, 169, 170]. High-fat diet (HFD) consumption has been noted to lead to paler-appearing livers, along with greater lipid deposition and liver damage (as indicated by hematoxylin‒eosin and Oil Red O staining), in DKO mice than in APOE-KO mice [169, 170]. Moreover, a significant increase in the expression of proinflammatory cytokines and ARG1 in DKO mice livers indicates that the absence of RON causes Kupffer cell transformation from the M2 phenotype to the M1 phenotype, exacerbating liver inflammation [169, 170].

In contrast to prior findings, a recent investigation utilizing RON TK^−/− mouse models demonstrated diametrically opposed phenotypic outcomes [171]. A HFD was noted to lead to 50% less weight gain in TK^−/− mice than in controls; moreover, TK^−/− mice’s livers were considerably less steatotic and weighed significantly less than the livers of TK^+/+ mice. Compared with control mice, TK^−/− mice also presented significantly higher serum levels of proinflammatory cytokines, such as IL-6 and IFN-γ [171]. These conflicting results may be attributed to the differences between experimental models. The deletion of Ron tyrosine kinase signaling domain and the ligand binding domain of Ron will lead to different effects on the MSP-RON signaling pathway [5]. In RON TK^−/− mouse models, RON may retain intact transmembrane and extracellular domains despite the deletion of the tyrosine kinase domain. In mice with targeted deletion of the RON ligand-binding domain, RON may retain partial extracellular segments, an intact transmembrane domain, and a functional intracellular tyrosine kinase domain, even including the variants that naturally lack this domain [27]. Notably, distinct RON-deficient models may implicate differential outcomes of targeting specific components within the MSP-RON axis. The mechanisms underlying these discrepant outcomes and their implications require further investigation.

To elucidate the effects of MSP in the early stages of NASH, a study fed mice lacking low-density lipoprotein receptor expression a high-fat, high-cholesterol diet for 7 days and administered MSP via minipumps in the final 4 days [172]. Compared with saline-treated control mice, MSP-treated mice presented increased gene expression of proinflammatory and proapoptotic mediators in the liver; however, the underlying mechanisms remain unclear. Some scholars believe that the aforementioned studies did not report the dose of MSP incorporated into the liver, which may have affected the experimental results [6]. Future studies should explore the long-term effects of MSP on the development of NASH, as well as the effects of MSP in an advanced-stage NASH model.

The presence of liver fibrosis is the most important determinant of NAFLD prognosis [173]. Some studies in which RON’s ligand-binding domain was deleted in mice demonstrated that RON signaling has an antifibrotic effect on a NASH model [169, 170]. This finding contrasts with the aforementioned finding that RON activation promotes liver fibrosis development [11]. This might be due to the complex and diverse biological effects of RON activation in NAFLD. The protective effect of RON on liver fibrosis may have originated from the positive effects of RON on liver lipid metabolism and inflammation inhibition [169, 170].

Taken together, these findings highlight the importance of considering the stages and systemic effects of NAFLD when designing treatment strategies [165]. Future studies investigating the potential for regulating MSP-RON signaling at different stages of NAFLD development are warranted.

MSP-RON signaling in cholestatic liver disease

Cholestatic liver disease, caused by bile stasis, results in liver damage and fibrosis [174]. Currently, the evidence regarding MSP-RON in cholestatic liver disease primarily focuses on primary sclerosing cholangitis (PSC). PSC is characterized by idiopathic, progressive inflammation and fibrosis of the intrahepatic and extrahepatic bile ducts, and occurs in association with inflammatory bowel disease (IBD) in 70–80% of patients [175]. In 2011, Espen Melum et al. identified SNP rs3197999 in MST1 as a susceptibility locus for PSC through genome-wide association analysis (P = 1.1 × 10^− 16). The A→G nucleotide substitution at this locus results in a p.Arg689Cys missense variant, replacing arginine with cysteine at position 689 of the MSP protein [176]. This missense variant is situated within a critical region of the high-affinity receptor-binding interface, suggesting its potential to functionally impair protein activity [39, 177]。In 2012, Häuser et al. demonstrated that the p.Arg689Cys variant substantially enhances MSP’s stimulatory effects on macrophage chemotaxis and proliferation [177]. Notably, this non-synonymous coding variant has been robustly replicated as a risk variant for IBD in multiple independent studies [177–181]. This association underscores the pathogenic significance of the rs3197999 variant in PSC patients with IBD, although the precise mechanistic role of the MSP-RON signaling axis in PSC pathogenesis remains incompletely defined. Future research should prioritize elucidating disease-relevant biological processes mediated by dysregulated MSP signaling and developing targeted therapeutic strategies to modulate this pathway.

MSP-RON signaling in liver cancer

MSP-RON signaling in HCC progression

Aberrant RON activation is key in the tumorigenesis, malignant progression, angiogenesis, and chemoresistance of many cancers [5, 27–29]. RON overexpression is observed in only a minority of HCC cases (5.7%), in stark contrast to its higher prevalence in breast cancer (55.6%) and colorectal cancer (51.1%) cohorts [55, 182]. However, RON plays a critical role in HCC. A study employing siRNA to investigate its function in HCC cell lines revealed that RON is associated with the invasive and carcinogenic phenotypes of cancer cells, including migration, invasion, antiapoptotic behavior, and cell cycle arrest, through the regulation of AKT, RAF, and ERK signaling molecules [183]. To date, few studies have systematically evaluated both the expression profiles of RON variants in HCC and the fundamental differences in RON signaling activation between malignant and normal hepatocytes. These knowledge gaps underscore the need for comprehensive investigations into the molecular heterogeneity of RON during hepatic carcinogenesis.

MSP expression is markedly elevated in HCC. Zhu et al. demonstrated that MSP mRNA expression is higher in clinical samples of HCC than in normal liver tissue cells [184, 185]. Research using the antibody microarray technique indicates that MSPα is a serum biomarker for predicting HBV-related HCC [186]. Thus, the MSP-RON signaling pathway may play a role in HBV-related HCC progression. Recently, Xi et al. reported that MSP activates γδT cells within the peripheral blood mononuclear cells (PBMCs) of patients with HCC [187]. These activated γδT cells exhibit cytotoxic activity against HepG2 cells [187]. Further characterization of these γδT cells revealed that T cells primarily recognize MSP through their TCRs [185]. Compared with those from healthy controls, γδT cells from the PBMCs of patients with HCC significantly increased the expression of antigen-presenting molecules, including MHC-I, MHC-II, CD86, and CD11a, after MSP stimulation [185]. Although the underlying mechanisms remain unclear, MSP has potential as a therapeutic target for HCC and in γδT-cell–based treatment for HCC [188–190].

Emerging evidence implicates RON signaling in the pathogenesis of cholangiocarcinoma in addition to HCC. Preclinical studies have demonstrated that the small-molecule inhibitor BMS-777607 suppresses cholangiocarcinoma cell proliferation [191]. However, as BMS-777607 exhibits multikinase inhibitory activity targeting both RON and MET receptors, the specific mechanistic contribution of RON requires further validation.

RON and MET in the prognosis of patients with hepatobiliary carcinoma

RON and MET expression are strongly correlated with the prognosis of patients with hepatobiliary cancer. Compared with MET-negative HCC patients, MET-positive HCC patients presented higher 5-year overall recurrence rates (65.3% vs. 54.3%, P = 0.041) [192]. In another study, advanced perihilar cholangiocarcinoma (CCA) patients with both MET and RON positivity presented worse 5-year survival rates than those with other expression patterns (16.7% vs. 39.8%, P = 0.021) [193]. Compared with the MET/RON intermediate group (48.5%), the MET/RON negativity and MET/RON positivity groups of patients with extrahepatic CCA had significantly lower overall 5-year survival rates (28.3% and 32.4%, respectively; P = 0.01) [194]. Notably, MET or RON negativity was associated with nodal metastasis in extrahepatic CCA [193]. Krawczyk et al. conducted genotyping on the rs3197999 variant of MSP in 223 patients with CCA and discovered a tendency for overexpression of the MSP variant in these patients [195]. Thus, MSP may play a role in modulating the genetic risk of CCA [195].

Taken together, these results reveal that RON and MET may be prognostic biomarkers for clinical hepatobiliary cancer. The examination of RON and MET expression may provide a novel method for patient classification than traditional pathological classification systems.

RON as a potential target in liver metastasis therapy

Accumulating evidence highlights the crucial role of MSP-RON signaling in promoting cancer metastasis [5, 96]. RON activation can promote the invasive growth of cancer cells through various signaling pathways [48, 58, 145]. MSP-RON signaling also promotes cancer metastasis by modulating the tumor microenvironment (TME) [5]. RON is expressed in various cellular components of the TME and regulates their functions, including tumor-associated macrophages [110], myeloid-derived suppressor cells [116], and cancer-associated fibroblasts [196]. Current evidence indicates that MSP-RON activation generates a complex immunoregulatory gene signature, ultimately suppressing CD8 + T-cell activity to promote tumor progression and metastasis [114, 116, 118, 197]. Notably, RON signaling in T cells is mediated by SF-RON. Lai et al. reported that the absence of SF-RON is sufficient to completely abolish breast cancer metastases in mice. In mice with SF-RON deletion, spleen cells show a decrease in Treg and CD11b + Ly6G + myeloid cells, along with an increase in stem-like CD4 + T cells and tumor-specific CD8 + T cells within the TME. RON activation also promotes angiogenesis in prostate cancer through the upregulation of VEGFR, CXCL1, CXCL2, CXCL5, and CXCL8 expression [50, 51].

In tumor cells, MSP alone is sufficient to induce liver metastasis of lung cancer. Sato et al. transfected SBC-5 cells with MSP to explore its effects on tumor metastasis behavior and the tumor immune microenvironment [10]. SBC-5 is a small-cell lung cancer cell line that does not express RON; therefore, MSP does not affect these cells directly [10]. The authors reported that SBC-5 cells expressing MSP predominantly metastasized to the liver rather than to bones or the lungs [10]. Immunohistochemical staining indicated that macrophage infiltration in liver metastases of SBC-5 expressing MSP was significantly greater than that in controls, with a greater density of tumor-associated microvessels—which may facilitate liver metastasis [10]. However, the mechanisms underlying MSP-dependent hepatic tropism in tumor metastasis remain unclear. One possible explanation is that MSP, RON, and HGFA expression in the liver is widely believed to be higher than that in other organs [10, 198]. HGFA is a trypsin-like serine protease that can activate the single-chain precursor pro-MSP into mature form [199]. MSP and HGFA are synthesized mainly by the liver, and the constitutive expression of RON in hepatocytes and Kupffer cells generates the most advanced environment for receiving MSP signals [10, 18]. Certain cancer cells with high RON expression have been observed to escape paracrine regulation by establishing an autocrine circuit through autonomous MSP expression, thereby enabling self-sustaining oncogenic signaling [196, 200, 201]. In addition, analysis of The Cancer Genome Atlas PanCancer data revealed that MSP expression is significantly and positively correlated with RON expression in almost all available cancer datasets [201]. Therefore, the liver may be susceptible to cancer cells with high RON and MSP expression. In other words, cancer cells with strong RON and MSP expression may have an increased risk of liver metastasis. Overall, targeting MSP-RON signaling may represent a potential strategy to simultaneously suppress primary tumors and their liver metastases.

Therapeutic strategies targeting MSP-RON signaling

To date, therapeutic strategies targeting RON in oncology have focused primarily on three modalities: small-molecule inhibitors, monoclonal antibodies, and antibody‒drug conjugates (ADCs). Table 2 summarizes the current therapeutic strategies targeting the MSP-RON signaling pathway by the end of 2024.

Table 2.

Summary of drugs targeting the RON receptor tyrosine kinase

Agents	Type	Mechanism and Target Receptor	Phase	Characteristics	References
Crizotinib (PF-2341066)	Small Molecule Inhibitor	ATP-competitive inhibitor of RON, Met, ALK and ROS	FDA-approved (NSCLC)	Inhibits ATP binding, effectively suppressing cancer cell growth and metastasis.	[221, 222]
Foretinib (GSK1363089)	Small Molecule Inhibitor	Multi-target tyrosine kinase inhibitor of RON, Met, VEGFRs, AXL, ROS, and TIE-2	Phase I/II Clinical Trial	Shows antitumor activity in various cancers, including HCC.	[223, 224]
BMS-777607	Small Molecule Inhibitor	ATP-competitive inhibitor of RON, Met, and other RTKs	Phase I/II Clinical Trial	Demonstrates potential in inhibiting tumor growth and metastasis across different cancers.	[203, 225]
Golvatinib (E7050)	Small Molecule Inhibitor	Multi-target tyrosine kinase inhibitor of RON, Met, and c-Kit	Phase I/II Clinical Trial	Inhibits tumor growth and metastasis in various cancer models; effectively blocks HAV infection.	[226–228]
Glesatinib (MGCD265)	Small Molecule Inhibitor	Multi-target tyrosine kinase inhibitor including MET, RON, VEGFR1, VEGFR2, VEGFR3, and TIE2	Phase I/II Clinical Trial	Showing antitumor activity in various cancers.	[229, 230]
Compound I	Small Molecule Inhibitor	Specific inhibitor of MET and RON	Preclinical Studies	Effective in xenograft models, showing potential in various cancer cell lines.	[231]
MK-2461	Small Molecule Inhibitor	ATP-competitive inhibitor of RON and Met	Phase I/II Clinical Trial	Effective in xenograft models, showing potential in various cancer cell lines.	[232]
MK-8033	Small Molecule Inhibitor	ATP-competitive inhibitor of RON and Met	Phase I Clinical Trial	Effective in xenograft models, showing potential in various cancer cell lines.	[233, 234]
Lcrf-0004	Small Molecule Inhibitor	Selective inhibitor of RON	Preclinical Studies	Showing potential in various cancer cell lines.	[235]
WM-S1-030	Small Molecule Inhibitor	Selective inhibitor of RON and RON variants	Phase I Clinical Trial	More potent against the RON variants, including RON, RON Δ155, Δ160, and Δ165.	[205]
IMC-41A10	Monoclonal antibody	Binds to the β-chain extracellular domain of RON, prevents MSP binding, and blocks MSP-RON signaling	Preclinical Studies	Effective in xenograft models, showing potential in various cancer cell lines.	[236]
Narnatumab (IMC-RON8)	Monoclonal Antibody	Binds to RON, prevents MSP binding, and blocks MSP-RON signaling	Phase I Clinical Trial	Well-tolerated with limited antitumor activity.	[208]
29B06/07F01 (humanized)	Monoclonal Antibody	Humanized monoclonal antibodies targeting RON	Preclinical Studies	Effective in xenograft models, potential in various cancer cell lines.	EORTC- NCI-AACR International Symposium‡
Ig4/Ig7/Ig10 (human)	Monoclonal Antibody	Binds to the SEMA domain of RON, and blocks MSP-RON signaling	Preclinical Studies	Effective in various cancer cell lines, but not in xenograft models.	[237]
Zt/f2 (mouse)	Monoclonal Antibody	Binds to the 49 amino acid sequence coded by exon 11 in the RON β-chain extracellular sequences. Does not compete with MSP for binding to RON. Induces RON internalization, reducing RON expression and impairing downstream signaling.	Preclinical Studies	Effective as a single agent or in combination with 5-FU in xenograft tumors, showing potential in various cancer cell lines.	[206, 238]
Zt/g4 (mouse)	Monoclonal Antibody	Binds to the SEMA domain of RON. Induces RON internalization, reducing RON expression and impairing downstream signaling.	Preclinical Studies	Effectively induces RON internalization, diminishes RON expression, and impairs downstream signaling.	[62, 206, 207]
Zt/g4-MMAE	Antibody-Drug Conjugate	Zt/g4 antibody conjugated with MMAE, targeting RON receptor	Preclinical Studies	Shows antitumor activity in PDAC, and TNBC models.	[38, 209, 210]
Zt/g4-DM1	Antibody-Drug Conjugate	Zt/g4 antibody conjugated with DM1, targeting RON receptor	Preclinical Studies	Shows antitumor activity in PDAC, CRC, and NSCLC models.	[14, 239, 240]
H5B14-MMAE	Antibody-Drug Conjugate	H5B14 antibody conjugated with MMAE, targeting RON PSI domain	Preclinical Studies	Effective against various cancer cell lines and xenograft models.	[241]
H5B14-DCM	Antibody-Drug Conjugate	H5B14 antibody conjugated with MMAE, targeting RON PSI domain	Preclinical Studies	Effective against various cancer cell lines and xenograft models. Stronger tumor elimination ability than H5B14-MMAE.	[241]

NSCLC non-small cell lung cancer, HCC hepatocellular carcinoma, HAV hepatitis A virus, PDAC pancreatic ductal adenocarcinoma, TNBC triple-negative breast cancer, CRC colorectal cancer

‡Data presented at the EORTC-NCI-AACR International Symposium, 16–19 Nov 2010, Berlin, Germany

Owing to the resemblance of ATP pocket structures across various kinases, many inhibitors that target the ATP binding sites of tyrosine kinases, such as the RON inhibitor BMS-777607, impact multiple tyrosine kinases and exhibit off-target effects [202, 203]. Moreover, variants constitute a major oncogenicity component of RON signaling [96]. Given the structural heterogeneity of RON variants, which exhibit varying divergence from FL-RON, constitutes a major obstacle for targeted therapies. Many small-molecule inhibitors frequently demonstrate selective efficacy against FL-RON while showing minimal activity against variants [73, 204]. Comfortingly, the recent discovery of WM-S1-030 has afforded a paradigm for the development of small-molecule inhibitors that target RON variants [205]. By potently suppressing phosphorylation in RONΔ155, Δ160, and Δ165 variants, this compound exhibits enhanced antitumor efficacy over conventional RON inhibitor BMS-777607 [205].

RON monoclonal antibodies specifically bind to the extracellular domain, inhibiting MSP-RON signaling through antibody-mediated receptor internalization and lysosomal degradation [62, 206, 207]. Despite demonstrating favorable tolerability in clinical trials, the therapeutic efficacy of narnatinib appears to be limited [208]. A potential explanation is that monoclonal antibodies fail to effectively target RON variants. The structural alterations between FL-RON and its variants result in a lack of shared antibody recognition epitopes, which is the primary reason for the failure of monoclonal antibodies to target RON during cancer therapy [5]. Recent advances in RON-targeted antibody delivery platforms have demonstrated significant therapeutic potential. A prime example is the RON-directed antibody‒drug conjugate Ztg4-MMAE, which achieves complete tumor eradication in RON-overexpressing xenograft models, thereby expanding the therapeutic potential of RON-targeted strategies through precision payload delivery [209, 210]. Notably, ADCs directly eliminate RON-overexpressing cells, largely overcoming the limitation of conventional targeted therapies in suppressing oncogenic signaling from RON variants. In summary, how to effectively target variants remains one of the greatest difficulties faced by RON-targeted therapies. In the future, screening and characterizing more small-molecule drugs targeting pathogenic RON variants and developing more antibody-based delivery therapies might be the key to overcoming this difficulty.

Furthermore, emerging strategies targeting the MSP-RON signaling axis warrant attention. For example, identifying natural inhibitors from traditional Chinese medicine may be a promising research strategy [211–213]. Targeting enzymes such as MT-SP1/matriptase and HGFA, which are involved in the MSP maturation process, could be novel strategies for modulating the MSP-RON signaling pathway [18–21, 198, 214]. Notably, given the crosstalk between RON and multiple RTKs, the exclusive use of RON inhibitors may lead to compensatory activation of other signaling pathways. The combination of other kinase inhibitors may lead to improved therapeutic effects for cancer [92]. RON is also likely involved in resistance to treatments targeting receptors engaging in crosstalk with it [92]. Table 3 summarizes liver treatment strategies targeting these receptors.

Table 3.

Summary of mainly small molecule inhibitors and monoclonal antibodies of key receptors involved in RON crosstalk in the field of liver diseases

Agents	Type	Target Receptor	Phase	Characteristics	References
MET
Cabozantinib (INC280)	Small Molecule Inhibitor	MET, VEGFR1/2/3, RET, KIT, FLT3, AXL, NTRK, ROS1	FDA-approved (NSCLC)	Effective in HCC. Demonstrates preventive effects on NASH. Inhibits HBV RNA transcription.	[242–246]
Tepotinib (Tepmetko)	Small Molecule Inhibitor	MET	FDA-approved (NSCLC)	Effective in HCC and ICC with high-level MET amplification.	[246, 247]
Tivantinib (ARQ 197)	Small Molecule Inhibitor	MET	Phase III clinical trial	Almost no effect on RON. Effective in cirrhotic patients with HCC.	[248]
Capmatinib	Small Molecule Inhibitor	MET	FDA-approved (NSCLC)	Effective in HCC with high-level MET amplification.	[249, 250]
Savolitinib (Volitinib)	Small Molecule Inhibitor	MET	Phase III clinical trial	Effective in HCC with high-level MET amplification.	[251, 252]
EGFR
Erlotinib (Tarceva)	Small Molecule Inhibitor	EGFR	FDA-approved (NSCLC)	Well tolerated with modest disease-control benefits in HCC. Demonstrated a comprehensive anti-HBV potential. Reduction in fibrogenesis.	[253–258]
Gefitinib (Iressa)	Small Molecule Inhibitor	EGFR	FDA-approved (NSCLC)	Induces growth inhibition, apoptosis, and cell cycle arrest in human HCC cells, attributed to blunted HSC activation.	[259–261]
Osimertinib (AZD9291)	Small Molecule Inhibitor	EGFR	FDA-approved (NSCLC)	Significantly induces apoptosis in a panel of HCC cell lines, irrespective of EGFR expression levels. Targets both tumor cells and angiogenesis.	[262, 263]
Afatinib (Gilotrif)	Small Molecule Inhibitor	EGFR, Her2	FDA-approved (NSCLC)	Inhibits the viability, migration, and invasion of HCC cells.	[264–266]
Pelitinib (EKB-569)	Small Molecule Inhibitor	EGFR, Her2	Phase II clinical trial	Inhibition of HCC cell migration and invasion occurs through suppression of EMT. Combination with sorafenib may overcome HCC resistance to EGFR-TK inhibitors.	[267, 268]
Cetuximab (Erbitux)	Monoclonal Antibody	EGFR	FDA-approved (mCRC)	Inhibits the growth of HCC cell lines and has the potential for combination therapy with other drugs for HCC.	[269–275]
Panitumumab	Monoclonal Antibody	EGFR	FDA-approved (mCRC)	Induces significant pre-G1 and G2/M cell cycle arrest in HepG2 cells when combined with sorafenib.	[276]
Nimotuzumab	Monoclonal Antibody	EGFR	Approved (III/IV NPC)	Effective in certain HCC cell lines. Showing potential in HCC.	[277]
PDGFR
Nintedanib	Small Molecule Inhibitor	PDGFRα, PDGFRβ, VEGFR1, VEGFR2, VEGFR3	FDA-approved (IPF)	Reduces steatohepatitis and hepatic fibrosis pathology. Inhibits ICC aggressiveness by suppressing cytokines derived from activated cancer-associated fibroblasts.	[278–280]
Regorafenib (Stivarga)	Small Molecule Inhibitor	PDGFRβ, VEGFR1/2/3, Kit, RET, Raf-1, BRAF, TIE2	FDA-approved (HCC)	Prevented the formation of liver metastases in colon cancer models. Demonstrated acceptable tolerability and evidence of antitumor activity in patients with HCC.	[281–286]
Sorafenib (Nexavar)	Small Molecule Inhibitor	PDGFRβ, VEGFR2, VEGFR3, FLT3, c-Kit	FDA-approved (HCC)	Demonstrates antitumor activity in HCC and potential for combination with other anticancer agents. Efficiently blocks HCV replication in vitro	[287–290]
Sunitinib (Sutent)	Small Molecule Inhibitor	PDGFRβ, VEGFR2, FLT3	FDA-approved (RCC, GIST, PNET)	Exhibits modest antitumor activity in advanced HCC with manageable adverse effects. Reduces inflammatory infiltrate, fibrosis, and portal pressure in cirrhotic rats. Weakly reactivates HBV.	[291–296]
Pazopanib (Votrient)	Small Molecule Inhibitor	PDGFRβ, VEGFR1, VEGFR2, VEGFR3, c-Kit, FGFR1, c-Fms	FDA-approved (RCC)	Demonstrates a manageable safety profile. Antitumor activity in patients with advanced HCC is attributed to antiangiogenic effects. In vitro, the proliferation of various HCC cell lines was not inhibited.	[297, 298]
Lenvatinib	Small Molecule Inhibitor	PDGFRα/β, VEGFR1/2/3, FGFR1	FDA-approved (HCC)	Demonstrates an antitumor activity in advanced HCC with manageable adverse effects. Exhibits preventive effects on experimental liver fibrosis and sinusoidal capillarization, as well as on the in vitro phenotypes of HSC.	[299–302]
Olaratumab	Monoclonal Antibody	PDGFRα	FDA-approved (STS)	Contributes to human HSC proliferation and migration. Showing potential in HEHE.	[303, 304]
IGFR
Linsitinib (OSI-906)	Small Molecule Inhibitor	IGF-1R, IR, INSR	Phase III clinical trial	Effective against certain HCC cell lines. Strongly enhances the inhibitory effect of Sorafenib and/or vitamin K1 on HCC cell migration.	[305–307]
BMS-754,807	Small Molecule Inhibitor	IGF-1R, INSR, Met, Aurora A/B, TrkA/B, Ron	Phase II clinical trial	Effective in certain HCC cell lines.	[308, 309]
NVP-AEW541 (AEW541)	Small Molecule Inhibitor	IGF-1R, INSR	Preclinical Studies	Induces growth inhibition, apoptosis, and cell cycle arrest in human HCC cell lines without accompanying cytotoxicity.	[310, 311]

NSCLC non-small cell lung cancer, HCC hepatocellular carcinoma, NASH non-alcoholic steatohepatitis, HBV hepatitis B virus, HSC hepatic stellate cell, mCRC metastatic colorectal cancer, IPF idiopathic pulmonary fibrosis, ICC intrahepatic cholangiocarcinoma, HCV hepatitis C virus, RCC renal cell carcinoma, GIST gastrointestinal stromal tumor, PNET pancreatic neuroendocrine tumor, STS soft tissue sarcoma, HEHE hepatic epithelioid hemangioendothelioma

Conclusions

In conclusion, this comprehensive review highlights the multiple involvement of MSP-RON signaling in liver pathobiology and its therapeutic potential in liver disease. In bacterial peritonitis, MSP-RON signaling activation significantly reduces bacterial loads in the liver [117]. It may also be associated with HBV infection [110, 140, 141], highlighting the need for further relevant research. By regulating lipid metabolism and suppressing inflammation, the MSP-RON signaling pathway might provide potential benefits to patients with early NAFLD [114, 115, 153]. Moreover, RON inhibitors can reverse liver fibrosis progression, and serum MSP levels may be effective prognostic markers in patients with liver cirrhosis [11]. Given the strong association of the MSP rs3197999 variant with PSC and CCA, further research is warranted to define its pathogenic mechanisms. However, given the intricate complexity of hepatic MSP-RON signaling, therapeutic strategies must be contextually tailored to specific disease etiologies and their associated molecular landscapes. Future research must elucidate its cell type-selective roles and spatiotemporal dynamics, but developing cell type-selective targeting approaches remains a critical challenge.

Targeting MSP-RON signaling may be a novel, promising strategy for the treatment of primary tumors and their liver metastases [5, 10]. However, from the therapeutic perspective of targeting primary cancers with RON overexpression, RON-targeted therapies face persistent challenges, including off-target effects and limited efficacy against RON variants [215, 216]. Recent advancements in the structural biology of the MSP-RON signaling pathway have provided a theoretical foundation, enhancing the current understanding of its complex signaling mechanisms. Moreover, AlphaFold and RosettaFold are anticipated to further accelerate the development of drugs that target MSP-RON signaling and targeted degradation technologies, such as molecular glues and PROTACs. These advancements are likely to accelerate the translation of basic research into clinical application.

Abbreviations

ACO

Aconitase

ADCs

Antibody‒drug conjugates

AMPK

AMP-activated protein kinase

APOE

Apolipoprotein E

ARG1

Arginase 1

AUC

Area under the curve

CCA

Cholangiocarcinoma

CIITA

Class II major histocompatibility complex transactivator

CLP

Ligation and puncture

CPT1

Carnitine palmitoyltransferase 1

CR3

Complement receptor 3

DKO

Double knockout

EGFR

Epidermal growth factor receptor

EMT

Epithelial–mesenchymal transition

ER

Endoplasmic reticulum

ESE

Exon splicing enhancer

ESS

Exon splicing silencer

Fas

Fatty acid synthase

Glc-6-pase

Glucose-6-phosphatase

GRB2

Growth factor receptor-bound protein 2

GS3Kβ

Glycogen synthase kinase 3β

GYS

Glycogen synthase

HBV

Hepatitis B virus

HCC

Hepatocellular carcinoma

HFD

High-fat diet

HGF

Hepatocyte growth factor

HGFA

Hepatocyte growth factor activator

HGFL

Hepatocyte growth factor-like protein

HIF1

Hypoxia-inducible factor 1

HL

Hairpin loop

HNF4

Hepatocyte nuclear factor 4

HNF4-α

Hepatocyte nuclear factor 4α

iC3b

Inactivated C3b

ICAM1

Intercellular adhesion molecule 1

IFN

Interferon

IGF1R

Insulin-like growth factor 1 receptor

IKK

Inhibitor of κB kinase

IL-6

Interleukin-6

iNOS

Inducible nitric oxide synthase

IPT

Immunoglobulin-like plexin and transcription

IRAK1

Interleukin-1 receptor-associated kinase 1

K1-4

Kringle domains 1–4

KO

Knockout

LKB1

Liver kinase B1

LPS

Lipopolysaccharide

MASLD

Metabolic dysfunction associated steatotic liver disease

MCP

Monocyte chemoattractant protein

MHCII

Major histocompatibility complex class II

MIP2

Macrophage inflammatory protein 2

MSP

Macrophage-stimulating protein

mTOR

Mammalian target of rapamycin

MT-SP1/matriptase

Membrane-type serine protease 1

MYD88

Myeloid differentiation primary response 88

NAFL

Nonalcoholic fatty liver

NASH

Nonalcoholic steatohepatitis

NF

Nuclear factor

NF-Y

Nuclear factor Y

p70S6K

p70S6 kinase

PBMCs

Peripheral blood mononuclear cells

PDGFR

Platelet-derived growth factor receptor

PDL1

Programmed death ligand 1

PECK

Phosphoenolpyruvate carboxykinase

PFKFB3

Phosphofructokinase-2/fructose-2,6-bisphosphatase isoform 3

PGC-1α

Peroxisome proliferator-activated receptor γ coactivator 1α

PHHs

Primary human hepatocytes

PI3K

Phosphoinositide 3-kinase

PKCζ

Protein kinase C zeta

PSI

Plexin-semaphorin-integrin

RON

Recepteur d’Origine Nantais

RSK2

p90 ribosomal S6 kinase 2

SEMA

Semaphorin

SF2/ASF

also known as serine/arginine-rich splicing factor 1 (SRSF1)

SHP

Small heterodimer partner

SMA

Smooth muscle actin

SOS

Son of sevenless

SREBF1

Sterol regulatory element transcription factor 1

STAT

Signal transducer and activator of transcription

TCR

T-cell receptor

TFs

Transcription factors

TGF-β

Transforming growth factor-beta

TK

Tyrosine kinase

TLR

Toll-like receptor

TLR4

Toll-like receptor 4

TM

Transmembrane

TME

Tumor microenvironment

TNF

Tumor necrosis factor

TRAF6

Tumor necrosis factor receptor-associated factor 6

Tregs

Regulatory T cells

USF1

Upstream transcription factor 1

VHL

Von Hippel–Lindau

Author contributions

HPY and MHW conceived the central concept of this review and delineated its overall structure. KW and HPY wrote the original version of the manuscript. JJ and KW created the figures. JYP, MJZ, JLZ, TS, DL, and MDW participated in the reference collection and table organizing. HPY and MHW provided guidance and supervision during the writing process and critically reviewed the manuscript. All the authors have read and approved the article.

Funding

This work was supported by grants from the National Natural Sciences Foundation of China [grant number 82473937], the Zhejiang Plan for the Special Support for Top-notch Talents in China [grant number 2022R52029], and Fundamental Research Funds for the Central Universities [grant number 2022ZFJH003].

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.