Exploration in the mechanism of fucosterol for the treatment of non-small cell lung cancer based on network pharmacology and molecular docking

Fucosterol, a sterol isolated from brown algae, has been demonstrated to have anti-cancer properties. However, the effects and underlying molecular mechanism of fucosterol on non-small cell lung cancer remain to be elucidated. In this study, the corresponding targets of fucosterol were obtained from PharmMapper, and NSCLC related targets were gathered from the GeneCards database, and the candidate targets of fucosterol-treated NSCLC were predicted. The mechanism of fucosterol against NSCLC was identified in DAVID6.8 by enrichment analysis of GO and KEGG, and protein–protein interaction data were collected from STRING database. The hub gene GRB2 was further screened out and verified by molecular docking. Moreover, the relationship of GRB2 expression and immune infiltrates were analyzed by the TIMER database. The results of network pharmacology suggest that fucosterol acts against candidate targets, such as MAPK1, EGFR, GRB2, IGF2, MAPK8, and SRC, which regulate biological processes including negative regulation of the apoptotic process, peptidyl-tyrosine phosphorylation, positive regulation of cell proliferation. The Raf/MEK/ERK signaling pathway initiated by GRB2 showed to be significant in treating NSCLC. In conclusion, our study indicates that fucosterol may suppress NSCLC progression by targeting GRB2 activated the Raf/MEK/ERK signaling pathway, which laying a theoretical foundation for further research and providing scientific support for the development of new drugs.

. A stepwise workflow showed the mechanism of fucosterol against NSCLC through network pharmacology.
Prediction of binding between fucosterol and GRB2. In this work, the initial structure of the ligand-free structure of the GRB2 (PDB code: 6SDF), was obtained from X-ray diffraction in Protein Data Bank, and the 3D structure of fucosterol (compound CID: 5281328) and Lymecycline (compound CID: 54707177) were obtained from Pubchem. Lymecycline was used as a positive control. Then, we used AutoDock 4 which based on the standard docking procedure for a rigid protein and a flexible ligand. At first, we used pymol2.4 to remove the water molecules and the original ligands from the GRB2, and then used AutoDock4 to add hydrogen and used the vina to dock the GRB2 with fucosterol and Lymecycline, and then selected the compound with the best binding effect. Finally, we used pymol2.4 and ligplot2.3 to observe and analyze the interaction and binding mode between ligand and receptor.
Molecular dynamics simulations. Lymecycline and fucosterol with GRB2 were rationally docked in our previous work. This complex was analyzed by all atom molecular dynamics (MD) simulations for 50 ns. All the simulations were carried out with Gromacs 2020.2 software package and gromos54a7_atb.ff force field. The force field parameters of fucosterol and Lymecycline was generated by ATB (https ://atb.uq.edu.au/regis ter.py). To ensure the total charge neutrality of the simulated system, the corresponding amount of sodium ions were added to the system to replace water molecules to produce solvent boxes of the appropriate size. Initial the energy of 50,000 steps of the whole system was minimized by (EM) at 300 K. Subsequently, the systems were equilibrated by position restraint simulations of NVT and NPT ensembles. Equilibrated systems were used to simulate a 50 ns no restraint production run. In the end MD analyses were performed, this includes root mean square deviation (RMSD), root mean square fluctuations (RMSF) and Hydrogen bond analysis. The data of RMSD, RMSF and Hydrogen bond analysis is generated by Origin2019b (https ://www.origi nlab.com/). GRB2 expression and immune infiltrates. In order to explore the correlation between GRB2 expression and immune infiltrates in LUAD, the TIMER2.0 database (http://timer .cistr ome.org), a friendly comprehensive tool Scientific Reports | (2021) 11:4901 | https://doi.org/10.1038/s41598-021-84380-w www.nature.com/scientificreports/ set for systematical analysis of immune infiltrates across diverse cancer types [36][37][38] , was carried out. There were many immune cells in the TIMER database. We chose CD4 + T cells, CD8 + T cells, B cells, neutrophils, macrophages, myeloid dendritic cells to conduct a new analysis. Besides, we performed the association between gene GRB2 and EGFR after adjusting tumor purity in the correction section in TIMER2.0 database.
Prognostic values of GRB2. The correlation between GRB2 expression and survival in lung adenocarcinoma was analyzed by Kaplan-Meier plotter (http://kmplo t.com/analy sis/), a tool for assessing the functions of 54,675 genes and 10,188 tumor tissue samples, including breast cancer, ovarian cancer, lung cancer, and gastric cancer.

Results
Predicted target screening of fucosterol. A total 210 predictive targets were identified from the PharmMapper database, and ultimately, 135 official symbols of fucosterol related targets for limiting to "Homo sapiens" were obtained in the UniProtKB database after filtering by z-score > 0 (Table 1).
Cluster analysis for NSCLC-related targets. Although we selected high-relevance score of NSCLC targets from GeneCards, the NSCLC bio-network was still huge. To further analyze biological processes of the function module in NSCLC network, we obtained protein-protein interaction network (PPI) data for NSCLC, and clustered to find out the topology module of NSCLC's PPI network. The cluster 1 was retained since it is the most significant for PPI network of NSCLC (Fig. 2), and the genes of cluster 1 was shown in Table1. Functional enrichment analyses of GO biological process on the genes of cluster 1 were performed by DAVID6.8, showing cell proliferation, apoptosis, cell cycle, angiogenesis, NSCLC gene expression, invasion and migration, signal transduction, and NSCLC related signaling pathways (Fig. 3).

Analysis of fucosterol-NSCLC PPI network.
In order to construct the interaction network between proteins and dig out the core regulatory genes, the fucosterol-NSCLC PPI network of the candidate targets was performed (Fig. 4). Fucosterol shared 37 targets with NSCLC, and the network consisted of 36 nodes (one of the target proteins does not interact with others) and 177 edges. The color of each node is related to its degree; the darker nodes have the larger value of Degree. The size of the node is linked to its Edge; the bigger nodes have the larger value of Edge Betweenness. Based on the network topology analysis, the betweenness centrality is 0.0261, the average node degree is 9.83, the average closeness centrality is 0.535, which suggests the presence of a central hub between candidate targets. Six hub genes were extracted according to betweenness centrality, node degree and closeness centrality. The hub genes were speculated to play a significant role in fucosterol treated NSCLC, including EGFR, MAPK8, MAPK1, GRB2, SRC, IGF1 (Table 2).
Gene ontology enrichment analysis for candidate targets. GO enrichment yields a deeper understanding on the gene function and biological significance of the candidate targets of fucosterol treated NSCLC on a systematic level. To obtain the biological processes, molecular functions, and cellular components of the candidate targets, we performed a GO enrichment analysis, and displayed the top 20 significantly terms (p-value ≤ 0.05) of each module in Fig. 5. It is suggested that the candidate targets could act through protein tyrosine kinase activity, protein phosphatase binding, negative regulation of apoptotic process, peptidyl-tyrosine phosphorylation, positive regulation of cell proliferation in the nucleus, cytosol, extracellular space, nucleoplasm, extracellular region. Among them, 14 vital biological processes directly affect cluster 1 of NSCLC disease module were presented independently by connecting with the candidate targets ( Fig. 6). This network diagram revealed candidate targets were mainly involved in cell proliferation and apoptosis, angiogenesis, cell migration and signal transduction, and illustrated that hub genes were strongly associated with various biological processes.
Evaluation of target-pathway network. To explain better the mechanisms of fucosterol treatment of NSCLC at the pathway level, 73 pathways were obtained by mapping 37 candidate targets (Table 3). After getting rid of generalized and other disease terms, a total of 45 pathways and 37 candidate targets constituted the target-pathway interaction network (Fig. 7). Some targets are mapped to multiple pathways and multiple targets also regulated various pathways suggesting that candidate targets may mediate interaction and crosstalk of different pathways. The pathways may be the major factor for fucosterol's resistance to NSCLC, that is, PI3K-Akt signaling pathway, VEGF signaling pathway, ErbB signaling pathway. The PI3K-Akt signaling pathway is widely recognized as a prominent cancer signaling pathway, and is closely related to affect the proliferation, survival and apoptosis of NSCLC cells [39][40][41] . Also, VEGF signaling pathway involves in tumor cell-dependent continuous vascular supply and thus has a profound effect on tumor cell growth and metastasis 42 . In addition, owing to interaction of ErbB receptors with many signal transduction molecules, which can activate multiple intracellular pathways, the ErbB signaling pathway plays a significant role in the development of cancer 43 . As mentioned, the three pathways are closely relevant to NSCLC treatment. Besides, it has been proved that fucosterol exerts antiproliferative effects to achieve the therapeutic purpose of lung cancer through targeting Raf/MEK/ERK signaling pathway 44    www.nature.com/scientificreports/ eration and migration, and play a crucial role in the development and metastasis of tumors [45][46][47] . Meanwhile, the PI3K-Akt signaling pathway participates in the regulation of cell cycle progression [48][49][50] . According to this map, the three signaling pathways can jointly activate the Raf/MEK/ERK signaling pathway reported to play an antiproliferation role in curing cancer 44,51,52 . More importantly, from the observation in this map, GRB2 can activate the downstream pathways-Raf/MEK/ERK signaling pathway through ErbB signaling pathway and PI3K-Akt signaling pathway, which may make GRB2 one of the potential therapeutic targets for fucosterol in the treatment of NSCLC. Moreover, GRB2 was presumed to be one of the six hub genes of fucosterol therapy for NSCLC base on fucosterol-NSCLC PPI network topology analysis. Collectively, GRB2 is speculated to play an important role in fucosterol treated NSCLC. We will further discuss GRB2 targeted by fucosterol for treating NSCLC.
Molecular docking validation of GRB2. Molecular docking simulation was utilized to verify the binding ability of the fucosterol to the hub genes GRB2, and Lymecycline was selected as the positive control. The docking mode and hydrogen bonding residues of fucosterol and Lymecycline with GRB2 after docking are shown in Fig. 9. Then, the binding sites of ligands, GRB2 and the surrounding residues by software pymol2.4 and ligplot2.3 are shown in Fig. 9A. In the docking model of fucosterol and GRB2, the action site are ASP-16, LYS-11, LYS-21, Ser-33, ASN-30, which forms hydrogen bonds with Ser-33, while the interaction between ligands and surrounding residues is analyzed in Fig. 9B. (binding energy is − 9.9 kcal/mol) It was indicated that the effector-binding energy of fucosterol is mainly provided by hydrophobic interactions rather than hydrogen interactions and it also showed that GRB2 was able to provide a decent hydrophobic environment for the substrate binding. And for positive control Lymecycline, it forms hydrogen bonds with THR-13, LYS-21, ASP-24. The others are hydrophobic interaction (Fig. 9C,D) and the binding energy is − 9.7 kcal/mol. Therefore, the docking effect of fucosterol and GRB2 is not much different from that of Lymecycline, and the binding energy of fucosterol is lower than that of Lymecycline. Figure 10 shows the molecular dynamics of Lymecycline and fucosterol with GRB2 protein ligand complex and GRB2 protein (RMSD, RMSF and hydrogen bond analysis). For RMSD (Fig. 10A), in the 100 ns simulation, all the three systems reached the equilibrium state quickly. In the end, the RMSD value of the complex formed by GRB2 and fucosterol was the lowest (0.25 nm), while the RMSD value of the positive control compound Lymecycline was 0.325 nm, and the RMSD value of the separate GRB2 protein system was 0.4 nm. This suggests that in terms of stability, the fucosterol compound formed a more stable combination in the active pocket of GRB2. For RMSF (Fig. 10B), the GRB2 and fucosterol system showed lowest RMSF, in the range between 0.5 and 1.25 nm, while the other two system between 0.75 and 2.0 nm, which in the docking results fucosterol formed hydrogen bond with SER-33. In RMSF, GRB2 and fucosterol system showed the fluctuations of SER-33 decreased significantly, which means that the hydrogen bonding interaction reduces the flexibility of the system, and Lymecycline formed hydrogen bond with THR-13, LYS-21, ASP-24, the hydrogen bond also reduced the flexibility of this system, but it was not obvious to GRB2 and fucosterol. In terms of hydrogen bond analysis (Fig. 10C,D), the number of hydrogen bonds between fucosterol and GRB2 system was 0 between 45 and 63 ns, while hydrogen bonds were absent in the positive control Lymecycline for a long time. In comparison of the number and existence of hydrogen bonds, the number and duration of hydrogen bonds in Lymecycline and GRB2 system is obviously better than that in fucosterol and GRB2 system. But not only by the hydrogen bonding interaction to determine the combination of ligand and receptor, also other interactions are needed to take into consideration.

Expression level and immune infiltrates analyses of GRB2 in NSCLC.
To further explore the direct relationship between GRB2 and NSCLC, we conducted the analysis of the relationship between GRB2 expression and immune infiltrates on TIMER(2.0) database. As shown in Fig. 11A Fig. 11D that the lower expression level of GRB2 had a better overall survival rate and gene GRB2 was an independent prognosis indictor for OS of patients with LUAD. Hence, we speculated that GRB2 exerted a more significant effect on the prognosis of LUAD, for it was highly associated with various immune cells in LUAD.

Discussion
NSCLC is characterized by high malignancy, low 5-year survival rate and poor prognosis. However, for patients with advanced lung cancer, having specific predictive biomarkers and receiving targeted therapy or immunotherapy significantly improve quality of life and progression-free survival (PFS) compared to chemotherapy [53][54][55][56][57][58] . Hence, finding effective targeted drugs to anti-lung cancer has become an urgent problem to be solved. Network pharmacology can analyze and explain the complexity between biological systems, diseases, and drugs from a network perspective, becoming a frontier method of drug discovery 59,60 . Molecular docking is a structurebased drug design method, which can predict the affinity and binding pattern through the interaction between ligands and receptors, accelerate the design and screening of drugs, and provide a basis for future experimental detection 61,62 . Molecular dynamics simulate ligand target complexes in a given system to explore their stability and flexibility 63,64 . Therefore, based on network pharmacology and molecular docking simulation ligand target binding, the mechanism research of fucosterol therapy for NSCLC may serve as the foundation for the development of NSCLC targeted drugs in the future. Based on the reverse of molecular docking technology for predicting fucosterol targets, and compare with NSCLC related targets, we obtained 37 candidate targets. Fucosterol-NSCLC PPI network analysis uncovered that fucosterol probably exerted pharmacological effects on NSCLC via 37 candidate targets, including 6 hub genes: GRB2, EGFR, MAPK1, SRC, IGF2, MAPK8. According to GO biological process analysis, 37 candidate targets were found to mainly responsible for cell proliferation, apoptosis, migration, signal transduction and angiogenesis. In addition, KEGG analysis of candidate targets disclosed that fucosterol probably has therapeutic effect on NSCLC through multiple pathways like PI3K-Akt signaling pathway, VEGF signaling pathway, ErbB signaling pathway. Furthermore, it is reported that fucosterol has antiproliferative effects on human lung cancer cells by inducing apoptosis, cell cycle arrest and targeting of Raf/MEK/ERK signaling pathway, such as A549 and SK-LU-1 cancer cells, in addition, fucosterol could also inhibit the growth of xenografted tumours in mice 44 . Through the NSCLC-related pathway map, we discover that GRB2 can be acted as the trigger or initiation signal of the Raf/MEK/ERK pathway by PI3K-Akt signaling pathway and ErbB signaling pathway, which may be crucial for fucosterol in the treatment of NSCLC.
Based on molecular docking, the combination of fucosterol and GRB2 can be discussed. At the same time, Lymecycline was selected as the positive control to provide the basis for future experiments. We chose to complete many key structures with higher resolution crystal structures, and then the docking site was selected in the SH3 region of GRB2, so as to influence the cell signal transduction events and achieve the therapeutic effect of cancer. RMSD, RMSF and hydrogen bond analysis were calculated using molecular dynamics simulations to infer the basic properties of ligand-target complexes-stability and flexibility. The stability of RMSD and RMSF was essential to infer good binding affinity, while hydrogen bond analysis was to compare the binding of phytosterol and Lymecycline to GRB2. The results show that the combination of GRB2 and fucosterol is stable.
Growth factor receptor-bound protein 2 (Grb2), a universally expressed adaptor protein, which plays a pivotal downstream mediator role in a variety of oncogenes signaling pathways and has a significant effect on signal transduction in normal and cancer cells 65,66 . Grb2 is a downstream protein of epidermal growth factor receptor (EGFR) that is known to be closely related to NSCLC, and may serve as an adaptor protein to bind to phosphorylated tyrosine in the EGFR, thereby linking receptor activation to intracellular signaling cascade [67][68][69] .
Simultaneously, recent research has demonstrated that the balance of Grb2 monomer-dimer is a determinant of its normal and carcinogenic functions 70 . Grb2 regulating angiogenesis and cell movement is highly overexpressed Figure 6. The network made in Cytoscape3.7.2 (https ://cytos cape.org/) depicted the relationship of candidate targets and important biological processes of fucosterol treated NSCLC. (Node in red is the hub gene, node in blue is the candidate targets, and yellow is the specific biological processes.)

Conclusion
In summary, our study indicated the molecular and pharmacological mechanism of fucosterol against NSCLC from a systematic perspective. We unveil that GRB2 can serve as an anticancer target in fucosterol to initiate the Raf/MEK/ERK pathway for treating NSCLC. This strategy provides a new idea of anti-NSCLC and lays a foundation for the development of new medicines. Nonetheless, network pharmacology has certain limitations, and more experiments are needed to verify the validity of our findings. Moreover, we hope that our study will be useful for fostering innovative research of marine drugs against cancers.

Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request. License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.