Simulation

Docking

To investigate potential binding interactions between key overexpressed membrane receptors in hepatocellular carcinoma (HCC) cells and all-trans retinoic acid (ATRA) and its corresponding ligands, we performed computational docking simulations using AutoDock.

For the receptors ASGPR (asialoglycoprotein receptor), CD44, GPC3 (glycosaminoglycan protein-3), and EGFR (epidermal growth factor receptor), we customised dedicated PDB files using structures from the Protein Data Bank (PDB) or, where necessary, AlphaFold prediction technology.

We first performed molecular docking simulations between ASGPR, CD44, GPC3, and EGFR with ATRA.

For the ligands, galactose (Gal) was used for ASGPR, and a tetrasaccharide fragment of hyaluronic acid (HA) for CD44.

Docking models were evaluated based on binding affinity, intermolecular contacts, and key residue interactions to predict stable complexes suitable for targeted drug delivery or immunotherapy in hepatocellular carcinoma.

Liver cancer cell membrane receptor-ATRA Docking

GPC3-ATRA Docking

To explore potential therapeutic interactions, we performed docking simulations between all-trans retinoic acid (ATRA) and GPC3. GPC3 is a proteoglycan overexpressed in hepatocellular carcinoma that regulates the Wnt signaling pathway and serves as a biomarker. ATRA, a vitamin A derivative with anticancer properties, was docked against the extracellular domain of GPC3 using AutoDock. The optimal model exhibited a binding energy of -6.06 kcal/mol, indicating moderate affinity. Although ATRA primarily targets nuclear retinoic acid receptors, emerging studies suggest its potential application in HCC by regulating cellular differentiation. This docking complex provides a basis for exploring potential off-target binding mechanisms and may be utilized to develop GPC3-targeted ATRA delivery systems, thereby enhancing the specificity of liver cancer therapy.

Model	Binding Energy (kcal/mol)
1	-6.06
2	-5.85
3	-5.27
4	-5.16
5	-5.13
6	-4.75
7	-4.55
8	-4.19

CD44-ATRA Docking

CD44 participates in hepatocellular carcinoma metastasis by binding hyaluronan. Its docking with ATRA was evaluated to assess potential inhibitory or regulatory effects. Using AutoDock, ATRA was positioned within the hyaluronan-binding pocket of the extracellular domain of CD44 (PDB: 1UUH). The optimal conformation achieved a binding energy of -7.75 kcal/mol. Existing literature indicates ATRA influences extracellular matrix interactions; this model proposes ATRA as a competitive inhibitor of CD44-hyaluronan binding, potentially reducing tumor invasiveness in HCC. This finding may guide ATRA-conjugated therapies targeting CD44-overexpressing cells.

Model	Binding Energy (kcal/mol)
1	-7.75
2	-7.24
3	-7.18
4	-7.10
5	-6.60
6	-6.53
7	-6.49
8	-6.17

EGFR-ATRA Docking

As a key driver of HCC proliferation, the extracellular domain of EGFR (PDB: 1IVO) underwent ATRA binding analysis via AutoDock. In the optimal model, ATRA docked near the EGF binding site with a binding energy of -7.32 kcal/mol. Although ATRA primarily acts intracellularly, computational studies suggest it may regulate membrane receptors. These docking results indicate ATRA could influence EGFR signaling via allosteric effects, opening new avenues for combination therapy in EGFR-amplified hepatocellular carcinoma.

Model	Binding Energy (kcal/mol)
1	-7.32
2	-7.10
3	-6.98
4	-6.34
5	-6.01
6	-5.56
7	-5.45
8	-5.11

ASGPR-ATRA Docking

ASGPR (responsible for glycoprotein clearance in HCC) was docked with ATRA via AutoDock using its CRD (PDB: 1DV8). The dominant model yielded a binding energy of -6.19 kcal/mol. The glycosylation simulation potential of ATRA promotes HCC cell uptake mediated by ASGPR. These simulation results support designing ATRA derivatives for liver-specific delivery, enhancing therapeutic efficacy by leveraging ASGPR's endocytic function.

Model	Binding Energy (kcal/mol)
1	-6.19
2	-6.08
3	-5.96
4	-5.90
5	-5.60
6	-5.09
7	-5.02
8	-4.99

Liver cancer cell membrane receptor-corresponding ligand Docking

ASGPR-Galactose Docking

ASGPR-galactose docking simulations focused on the carbohydrate recognition domain (CRD) of ASGPR, which specifically binds galactose residues at the terminal end of desialylated glycoproteins, thereby promoting receptor-mediated endocytosis in hepatocytes and hepatocellular carcinoma cells. AutoDock calculations revealed a highest-scoring model with a binding energy of -3.69 kcal/mol, indicating strong binding affinity. Key interactions include hydrogen bonds formed between the galactose hydroxyl groups (O3, O4, O6) and ASGPR residues Arg16, Asn127, and Gln128, along with hydrophobic contacts involving Trp39 and Phe75. These interactions align with crystallographic data (PDB: 9G76), confirming the role of this binding site in targeting HCC with glycosylated therapeutics.

Literature confirms that HCC-overexpressed ASGPR enables selective ligand uptake. The docking conformation suggests an efficient internalization process, providing structural rationale for applying galactose-conjugated nanoparticles in HCC therapy.

Model	Binding Energy (kcal/mol)
1	-3.69
2	-3.54
3	-3.50
4	-3.44
5	-3.42
6	-2.60
7	-2.40
8	-2.23

CD44-Hyaluronic Acid Docking

CD44, a hyaluronan-binding protein overexpressed in hepatocellular carcinoma, promotes tumour invasion and metastasis by binding to hyaluronan (HA) via its extracellular hyaluronan-binding domain.

We employed AutoDock to dock a tetrasaccharide hyaluronan fragment (GlcNAc-GlcUA-GlcNAc-GlcUA) into the hyaluronan-binding groove of CD44. The optimal conformation exhibited a binding energy of -5.48 kcal/mol, forming an extensive hydrogen-bond network: the glucuronic acid carboxyl group of HA and the acetamid group of N-acetylglucosamine interacted with CD44 residues Arg78, Tyr79, Arg120, and Tyr183, respectively. Furthermore, van der Waals interactions with Leu111 and Val136 further stabilise the complex.

This model validates molecular dynamics findings: HA binding induces a conformational shift in CD44, thereby enhancing cancer cell signalling. Its high affinity provides theoretical support for HA-based HCC targeting strategies, such as drug-loaded HA nanoparticles.

Model	Binding Energy (kcal/mol)
1	-5.48
2	-5.40
3	-4.85
4	-4.72
5	-4.57
6	-4.39
7	-4.39
8	-4.11