Type VI Collagen Receptor

Type VI collagen is a minor but important matrix protein in the liver. In a normal liver, type VI collagen is mainly distributed in the portal areas and at the membranes of endothelial cells, hepatocytes, and HSCs. The expression of type VI collagen is markedly elevated during liver fibrogenesis (Schuppan et al., 1985; Loreal et al., 1992; Takahara et al., 1995; Beljaars et al., 2000). Type VI collagen is one of the most important matrix proteins involved in cell adhesion to the surrounding matrix. It also modulates the homeostasis of matrix by maintaining the interaction between matrix molecules and the cells (Li et al., 2015). Type VI collagen receptor contains three α chains and is mainly expressed on activated HSCs but not in normal liver cells (Popov and Schuppan, 2009). Type VI collage receptor has thus been exploited for targeted delivery of antifibrotic agents to HSCs (Beljaars et al., 2000; Du et al., 2007).

A cyclic arginylglycylaspartic acid (RGD) peptide C*GRGDSPC* (* denotes the cyclizing cysteine residue) was discovered as a ligand for the type VI collagen receptor (Marcelino and McDevitt, 1995). The cyclic peptide was conjugated to human serum albumin (HSA) in a 10:1 molar ratio (Fig. 2). In vitro studies showed that the cyclic peptide-modified HSA would specifically bind to activated rat HSCs and enter the cells via internalization. A biodistribution study in rats with induced liver fibrosis demonstrated high accumulation of the peptide-modified HSA in activated HSCs (Beljaars et al., 2000). The cyclic RGD peptide was also attached to a biodegradable polymersome encapsulating the antifibrotic agent oxymatrine. The polymersome significantly inhibited the proliferation of activated HSCs and reduced the expressions of α smooth muscle actin and collagen in the cells. The peptide-modified polymersome exhibited higher antifibrotic activity in bile duct-ligated (BDL) rats than did the unmodified polymersome (Yang et al., 2014).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Cyclic RGD peptide C*GRGDSPC* conjugated human serum albumin for type VI collagen receptor.

Another cyclic RGD peptide, C*GRGDSPK*, was also exploited as a targeting ligand for the type VI collagen receptor. The peptide preferentially binds to activated HSCs rather than hepatocytes. An interferon-α1b loaded liposome was coupled with the RGD peptide to achieve targeted delivery to fibrotic liver via recognition of type VI collagen receptors on HSCs. In rats with BDL-induced liver fibrosis, the RGD peptide-coupled liposome exhibited 10-fold higher accumulation in HSCs than the unmodified liposome. Moreover, the RGD peptide-modified liposome demonstrated significant higher antifibrotic activity in the rats with liver fibrosis (Du et al., 2007).

Retinol-Binding Protein (RBP) Receptor

A key role of quiescent HSCs in the liver is to store up to 80% of vitamin A as retinyl palmitate in lipid droplets. Vitamin A binds to RBP in the systematic circulation and is taken up by HSCs via RBP receptors (Higashi et al., 2005). In addition, vitamin A is taken up more efficiently by activated HSCs rather than quiescent HSCs (Senoo et al., 2007; Sato et al., 2008). As a result, vitamin A and RBP have been used as ligands for HSC-specific delivery.

In one study, the domain III of an albumin was fused to RBP for HSC-specific delivery. The fusion protein can be efficiently taken up by HSCs in vitro and in vivo. More importantly, the fusion protein ameliorated liver fibrosis in rats induced by carbon tetrachloride (CCl₄) or BLD (Park et al., 2012; Lee et al., 2015). Using a similar concept, Zhang et al. (2015) developed a retinol-modified nanoparticle encapsulating an antifibrotic antisense oligonucleotide. The nanoparticle binds to RBP in the serum and is selectively taken up by HSCs via the RBP receptor. The nanoparticles successfully suppressed the expression of type I collagen in the liver and reversed BDL- and CCl₄-induced liver fibrosis in mice (Zhang et al., 2015).

In a pioneering in vivo study using siRNA for liver fibrosis therapy, a gp46 siRNA was encapsulated in a liposome coupled with vitamin A. The functionalized liposome efficiently silenced gp46, inhibited collagen secretion in HSCs, and reversed CCl₄- and BDL-induced liver fibrosis in rats. This was one of the first animal studies demonstrating the therapeutic potential of siRNAs in an animal model (Sato et al., 2008).

Similar strategies have been adopted by scientists to deliver various antifibrotic agents to HSCs (Sauvant et al., 2011; Duong et al., 2015; Lee et al., 2015; Pan et al., 2016). A phase 1b/2 clinical trial was recently conducted to evaluate the safety and tolerability of a vitamin A–coupled lipid nanoparticle containing siRNA targeting heat shocking protein 47 (HSP47) in patients with moderate to extensive liver fibrosis (ClinicalTrials.gov Identifier: NCT02227459).

Platelet-Derived Growth Factor Receptor-β (PDGFR-β)

Platelet-derived growth factor (PDGF) plays a critical role in the initiation and progression of liver fibrogenesis (Beljaars et al., 2003). It binds to PDGFR and regulates the migration, proliferation, and survival of HSCs. PDGFR is not only up-regulated in activated HSCs in animals but also is up-regulated in human fibrotic livers (Campbell et al., 2005; Borkham-Kamphorst and Weiskirchen, 2016). Particularly, PDGFR-β is dramatically overexpressed on activated HSCs, and its expression is much higher than that on other PDGFR-β-positive cells (Friedman, 2003). Therefore, several PDGFR-β-targeted delivery systems have been investigated for antifibrotic agents (Bansal et al., 2011a,b; Li et al., 2012).

The cyclic peptide C*SRNLIDC* (* denotes the cyclizing cysteine residue) was designed based on the receptor-binding residues of the PDGF B-chain. The cyclic peptide-conjugated HSA highly accumulated in the liver of the rats with BDL-induced liver fibrosis. Moreover, an intrahepatic distribution study showed that the majority of the peptide-conjugate HSA is located in HSCs, indicating selective binding of the cyclic peptide to PDGFR-β on activated HSCs (Beljaars et al., 2003).

In another study, the cyclic peptide C*SRNLIDC* was fused to a single-chain antibody fragment targeting the knob of a recombinant adenovirus. After binding to the adenovirus, the fusion protein retargeted the adenovirus to activated HSCs and abolished the virus’s natural tropism for hepatocytes and Kupffer cells (Fig. 3). This assembly provides a platform to selectively deliver antifibrotic genes to activated HSCs without nonspecific uptake by neighboring cells, such as Kupffer cells and hepatocytes (Schoemaker et al., 2008).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

A fusion protein to change the tropism of adenovirus to HSCs. The PDGFRβ-specific cyclic peptide C*SRNLIDC* was fused to a single-chain antibody fragment targeting the knob of a recombinant adenovirus. After binding to the adenovirus, the fusion protein retargets the adenovirus to activated HSCs and abolishes its natural tropism for hepatocytes and Kupffer cells.

The cyclic peptide C*SRNLIDC* was also conjugated to interferon-γ (IFNγ) using polyethylene glycol (PEG) as a linker. The peptide-conjugated IFNγ showed specific binding to PDGFR-β on cultured HSCs and inhibited their activation. It also significantly inhibited fibrogenesis in acute liver injury and in CCl₄-induced liver fibrosis animal models without inducing IFNγ-related side effects. By contrast, unmodified IFNγ failed to inhibit fibrogenesis in both animal models (Bansal et al., 2011b).

Instead of direct conjugation to IFNγ, the cyclic peptide C*SRNLIDC* was also coupled to a liposome encapsulating IFNγ, leading to higher uptake in activated HSCs. Compared with free IFNγ and unmodified liposome encapsulating IFNγ, the cyclic peptide-coupled liposome demonstrated higher inhibitory effect on the proliferation of HSCs. In vivo studies in rats with thioacetamide-induced liver fibrosis showed a high accumulation of the peptide-coupled liposome in activated HSCs. Moreover, an enhanced antifibrotic effect of the peptide-coupled liposome was observed (Li et al., 2012).

Synaptophysin

Synaptophysin (SYN) is a membrane glycoprotein present in the presynaptic vesicles of neurons, adrenal medulla, and other neuroendocrine epithelial cells (Wiedenmann et al., 1986). Its main function involves exocytosis of neurotransmitter. HSCs also express several neuroectodermal differentiation markers such as glial fibrillary acidic protein (GFAP) and nestin (Cassiman et al., 1999). SYN was found present in HSCs in both human and rat liver biopsies. After only 36 hours of galactosamine intoxication, rat liver biopsies revealed significant SYN expression in comparison with normal rat liver biopsies. This result suggests that SYN is a potential marker for HSCs (Cassiman et al., 1999).

Researchers have used SYN as a marker to differentiate portal fibroblasts from HSCs of fibrotic livers of humans, rats, and mice (Iwaisako et al., 2014). However, studies have found that canine HSCs (quiescent and activated) are negative for SYN.

Ijzer et al. (2006) investigated the morphologic characteristics of canine HSCs and myofibroblasts. They reported that muscle-specific actin clone HHF35 (HHF35) and α smooth muscle actin were potent markers for canine HSCs and myofibroblasts, but liver sections were consistently negative for SYN and GFAP. SYN has been used as a diagnostic marker because of its variable expressions with respect to different liver stresses and injuries. In a study, rats with nonalcoholic fatty liver disease (NAFLD) showed high hepatic inflammation, necrosis and fatty infiltration. Apart from Sirius Red staining, which is a primary staining method for collagen fibers, SYN staining was used to evaluate HSC activation and the fibrogenesis. Xiao et al. (2014) reported distinct increase in SYN positive areas in a rat model of NAFLD and subsequent decline in SYN expression after therapeutic intervention with Epigallocatechin gallate.

Despite its unique expression patterns during liver fibrosis and NAFLD, SYN has not been thoroughly investigated as a ligand for targeted drug delivery. However, SYN is advantageous over other markers for targeted delivery of antifibrotic agents because it forms part of endocytosing vesicles, leading to increased chance of endocytosis of its ligand and associated cargos (Douglass et al., 2008a). A group of researchers discovered a single-chain antibody (scAb), C1-3, that binds to SYN on activated HSCs in fibrotic mouse liver. C1-3 scAb was generated by phage display biopanning against the peptide sequence that is part of the extracellular domain of SYN (Elrick et al., 2005). Studies have shown that C1-3 scAb binds specifically to activated HSCs but not hepatocytes. The C1-3 scAb-conjugated tributyl tin was delivered into HSCs and exert its activity (Douglass et al., 2008a). This demonstrated the feasibility of using C1-3 scAb as a targeting ligand for HSC-specific delivery. One potential limitation of this strategy is the nonspecific delivery to neuroendocrine and neural cells. However, there are no evidence showing that scAb can cross the blood–brain barrier, which addresses the nonspecific issue to some degree (Douglass et al., 2008b).

Insulin-Like Growth Factor II Receptor (IGFIIR)

Also known as the cation-independent mannose-6-phosphate receptor (M6PR), the 300 kDa IGFIIR is a member of the IGF signaling family (Morgan et al., 1987). The major functions of IGFIIR include regulating the activity of insulin like growth factor II (IGFII); transportation of M6P-tagged lysosomal enzymes from Golgi network and outside of cells to lysosomes; and activation of TGF-β. M6P-tagged proteins are transported to lysosomes for degradation by lysosomal acid hydrolases (Puxbaum et al., 2012). It is noteworthy that IGFIIR is unrelated to tyrosine kinase activity and proliferation (Sedlaczek et al., 2003). As a result, IGFIIR can be exploited as a surface maker for targeted drug delivery without inducing the proliferation of HSCs.

While IGFIIR is expressed in quiescent HSCs, its expression is strongly upregulated (∼20 folds) in activated HSCs relative to quiescent HSCs (Beljaars et al., 1999; Mousavi et al., 2013). The expression of IGFIIR is often positively correlated with chronic liver injury (Hellemans et al., 2004). Moreover, IGFIIR-mediated ligand internalization is 3 times faster than cellular endocytosis in activated HSCs (York et al., 1999). Therefore, IGFIIR is considered as a promising receptor for HSC-targeted drug delivery (Ye et al., 2005; Gary-Bobo et al., 2007; Mousavi et al., 2013). Additionally, upregulation of IGFIIR is also a signature feature of nephroblastoma, hepatocellular carcinoma and rhabdomyosarcoma (Sedlaczek et al., 2003).

Mannose 6 phosphate (M6P) is the natural ligand for IGFIIR. IGFIIR consists of two M6P binding extracellular domains. Generally, the affinity of M6P toward IGFIIR is low, but it can be enhanced by oligomerization of the receptor (Byrd and MacDonald, 2000). Dimerization of a ligand has also proven to increase its binding affinity toward IGFIIR (Chen et al., 2015). Researchers have conjugated multiple M6Ps to HSA to deliver triplex-forming oligonucleotide (TFO) to activated HSCs in rats. This delivery system demonstrated high specificity toward HSCs and efficiently delivered the therapeutic load to target cells. Moreover, in vivo studies also revealed that TFO was selectively distributed in HSCs of fibrotic rats (Ye et al., 2006).

Similarly, the Rho-kinase inhibitor (+)-(R)-trans-4-(1-aminoethyl)-N-(4-pyridyl) cyclohexanecarboxamide dihydrochloride (Y27632) was conjugated with M6P/HSA, and its in vivo biodistribution was studied (van Beuge et al., 2011a,b). In comparison with free Y27632, the Y27632-M6P/HSA conjugate demonstrated exclusive distribution in HSCs. The conjugated Y27632 significantly reduced the collagen deposition in the extracellular matrix in contrast to the free Y27632. M6P-GlcNAc is another such conjugate, which has shown high binding affinity toward IGFIIR (Liu et al., 2013). Despite the high binding affinity of the M6P conjugate toward IGFIIR, the greatest limitations remain the difficulty in conjugation and relatively low binding affinity (approximately K_D 23 µM) of M6P (Jeanjean et al., 2006).

Using a novel combinatorial biopanning strategy, Chen et al. (2015) discovered an IGFIIR-specific peptide, peptide-431. Peptide-431 demonstrated high specificity toward IGFIIR expressed on human and rat HSCs. It has been used as a promising ligand for antifibrotic siRNA nanocomplexes (Fig. 4).(Zhao et al., 2018) This novel peptide is a promising ligand for the delivery of a therapeutic payload to activated HSCs.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Peptide-431, an IGF2R-specific peptide, enhances the accumulation of a siRNA nanocomplex in fibrotic liver. (A) Structure of the peptide-431 dimer. (B) Biodistribution of the siRNA nanocomplexes modified with different ligands including cholesterol, vitamin A, and peptide-431. Adapted from Zhao et al. (2018).

Chen et al. (2017) also discovered IGFIIR-specific aptamer using systematic evolution of ligands exponential enrichment (SELEX). The aptamer was annealed to the 3′ end of an siRNA and effectively delivered the siRNA to rat and human HSCs. The aptamer-siRNA chimera also exhibited high accumulation in the liver of the rats with CCl₄-induced liver fibrosis.

Low-Density Lipoprotein Receptor (LDLR)

The liver is the major organ for cholesterol homeostasis by controlling lipid synthesis and uptake in the body (van de Sluis et al., 2017). The expressions of LDLR and high-density lipoprotein receptor-scavenger receptor class B type1 (SR-B1) are high in the liver (Gao and Brigstock, 2003; Wolfrum et al., 2007). Clearance of plasma lipids in the liver is mediated by LDLR and LDLR-related protein 1 (van de Sluis et al., 2017).

LDLR plays a critical role in cholesterol metabolism via recognizing the apolipoprotein B100 of cholesterol-rich LDL and mediating a rapid endocytosis recycling process (Rudenko et al., 2002). LDLR binds to cholesterol-rich LDL at neutral pH through the “ligand-binding domain” with the help of Ca²⁺ but rapidly dissociates from these ligands at acidic pH in the endosome (Rudenko et al., 2002; Rudenko and Deisenhofer, 2003). LDLR also induces receptor-mediated endocytosis for very-low-density lipoprotein and chylomicron remnants through binding with apolipoprotein E (Rudenko et al., 2002).

Cholesterol was found accumulated in HSCs of the mice with CCl₄- and BDL-induced liver fibrosis. Accumulation of cholesterol in HSCs leads to sensitization of the cells to TGF-β and subsequently aggravates fibrogenesis (Teratani et al., 2012). Furthermore, the exacerbation of liver fibrosis and activation of HSCs also enhance the accumulation of cholesterol in HSCs, which is partially correlated with the increased expression of LDLR (Tomita et al., 2014).

A triplex-forming oligonucleotide targeting the genomic DNA of Type I collagen was conjugated with cholesterol and showed higher uptake in the liver of the rats with dimethylnirosamine-induced liver fibrosis (Cheng et al., 2006). In another study, cholesterol was used as a targeting ligand and was coupled to a streptavidin-based siRNA nanocomplex. The cholesterol-modified nanocomplex showed a higher uptake in HSC-T6 (LDLR+) cells rather than in Caco-2 (LDLR-) cells. Meanwhile, incubation of HSC-T6 cells with puromycin, an LDLR inhibitor, suppressed the cellular uptake of the cholesterol-modified nanocomplex (Shukla et al., 2013).

Cluster of Differentiation 44 (CD44)

CD44 is the major receptor for hyaluronic acid (HA), which is an important component of the extracellular matrix. CD44 was initially found overexpressed in liver biopsies from patients with alcoholic liver disease (Urashima et al., 2000). Similar results were observed in rats with dimethylnitrosamine-induced liver fibrosis. CD44 was highly expressed in fibrotic liver but only weakly expressed in normal liver. Although HA was localized in HSCs and endothelial cells, CD44 was mainly detected in infiltrating lymphocytes, Kupffer cells, and some endothelial cells (Satoh et al., 2000). In another study, researchers specifically investigated the expression of CD44 in HSCs from normal and fibrotic rats. The CD44 splice variant, CD44v6, is highly expressed in activated HSCs and facilitates the migration of the HSCs (Kikuchi et al., 2005).

HA is a well-established ligand for CD44 and has been used for HSC-targeted delivery of antifibrotic agents. In one study, curcumin was encapsulated in HA-modified polylactide nanoparticles. The nanoparticle induced cell death of activated HSCs without affecting quiescent HSCs and hepatocytes (Chen et al., 2016). In another study, HA was conjugated to 5β-cholanic acid to form micelles and encapsulate losartan. The HA micelles showed specific accumulation in fibrotic liver and ameliorated liver fibrosis in mice (Thomas et al., 2015).

HA was also used for targeted delivery to cirrhotic livers in mice. Quantum dots (QDots) were conjugated to HA and demonstrated higher uptake by HSCs than by normal hepatocytes. Biodistribution study showed specific delivery of the QDots to HSCs in mice with liver cirrhosis (Kim et al., 2010). The same group then developed HA-coupled polyethyleneimine to deliver a TGF-β siRNA for the treatment of liver cirrhosis. The siRNA/HA-polyethyleneimine nanocomplex entered HSCs via CD44-mediated endocytosis and significantly attenuated liver cirrhosis (Park et al., 2011).

Tumor Necrosis Factor–Related Apoptosis-Inducing Ligand Receptor (TRAILR)

The tumor necrosis factor receptor superfamily (TNFRSF) are cytokine receptors, including TRAILR1 (death receptor 4), TRAILR2 (death receptor 5), TRAILR3 (decoy receptor 1), and TRAILR4 (decoy receptor 1) (Locksley et al., 2001). TNF-related apoptosis-inducing ligand (TRAIL) is a known anticancer agent that induces apoptosis of cancer cells by specifically targeting TRAILRs (Singh et al., 2017). It was recently found that some immune cells, such as natural killer cells, play important roles in preventing liver fibrosis by killing activated HSCs in a TRAIL-dependent manner. Once activated, HSCs are more susceptible to apoptotic factors, such as TRAIL.

TRAILR1/2 contain a death domain, which plays an essential role in the apoptosis of HSCs (Ashkenazi and Dixit, 1998; Singh et al., 2017). Moreover, TRAILR1/2 are highly expressed in activated HSCs (Taimr et al., 2003; Yang et al., 2015). For example, during the activation of human HSCs (LX-2), the mRNA expression of TRAILR1 and TRAILR2 was increased by 18-fold and 17.6 fold, respectively. Similar results were observed in murine HSCs (Taimr et al., 2003).

In addition to binding to TRAILR1/2 to induce apoptosis of HSCs, TRAIL also binds to TRAILR3/4, which are also expressed on human HSCs (Singh et al., 2017). Because they lack a death domain, TRAILR3/4 act as a decoy receptors to limit the engagement of TRAIL with TRAILR1/2 (Falschlehner et al., 2007). A recent study has shown that silencing the expression of TRAILR3/4 in human HSCs sensitizes the cells to apoptosis (Singh et al., 2017).

Despite its therapeutic potential, TRAIL’s clinical applications are limited because of its very short half-life in the body. HA was conjugated to the N-terminal amino group of TRAIL to achieve targeted delivery to HSCs because CD44 is overexpressed in fibrotic livers. The TRAIL-HA conjugate showed similar biologic activity as native TRAIL but a higher accumulation in the liver. Moreover, the conjugate exhibited long circulation times in the blood for more than 4 days and reversed liver fibrosis in a rat model (Yang et al., 2015). In another study, PEG was conjugated to TRAIL to prolong its half-life. The PEG-conjugated TRAIL showed an extended half-life not only in rodents but also in nonhuman primates. PEG-TRAIL also attenuated CCl₄-induced liver fibrosis in rats (Oh et al., 2016).

Although TRAILR has yet to be exploited as a marker for HSC-specific drug delivery, it is possible to discover artificial ligands of TRAILR using in vitro selection procedures, such as phage display biopanning or SELEX. The artificial ligands can then be conjugated to antifibrotic agents to increase their uptake by HSCs.