Anti-breast cancer drugs targeting cell-surface glucose-regulated protein 78: a drug repositioning in silico study

Abstract Breast cancer (BC) is prevalent worldwide and is a leading cause of death among women. However, cell-surface glucose-regulated protein 78 (cs-GRP78) is overexpressed in several types of cancer and during pathogen infections. This study examines two well-known BC drugs approved by the FDA as BC treatments to GRP78. The first type consists of inhibitors of cyclin-based kinases 4/6, including abemaciclib, palbociclib, ribociclib, and dinaciclib. In addition, tunicamycin, and doxorubicin, which are among the most effective anticancer drugs for early and late-stage BC, are tested against GRP78. As (−)-epiGallocatechin gallate inhibits GRP78, it is also being evaluated (used as positive control). Thus, using molecular dynamics simulation approaches, this study aims to examine the advantages of targeting GRP78, which represents a promising cancer therapy regime. In light of recent advances in computational drug response prediction models, this study aimed to examine the benefits of GRP78 targeting, which represents a promising cancer therapy regime, by utilizing combined molecular docking and molecular dynamics simulation approaches. The simulated protein (50 ns) was docked with the drugs, then a second round of dynamics simulation was performed for 100 ns. After that, the binding free energies were calculated from 30 to 100 ns for each complex during the simulation period. These findings demonstrate the efficacy of abemaciclib, ribociclib, and tunicamycin in binding to the nucleotide-binding domain of the GRP78, paving the way for elucidating the mode of interactions between these drugs and cancer (and other stressed) cells that overexpress GRP78. Communicated by Ramaswamy H. Sarma


Introduction
Breast cancer (BC) is the most prevalent malignancy in women and is a heterogeneous disease on the molecular level . Most cases are at an advanced stage with dire consequences (Abdelaziz et al., 2020). BC is considered the most prevalent cancer worldwide, with over 2 million new cases expected to be diagnosed in females by 2020. Due to a shift in risk factor profiles, cancer registration, and cancer detection, its incidence and mortality rates have increased over the past three decades. Approximately 80% of patients with BC today are over the age of 50. The organism's survival depends on its developmental stage and molecular subtypes. Invasive BCs consist of a broad spectrum of tumors that differ in clinical indication, behavior, and morphology. Based on mRNA gene expression levels, BC can be split into molecular subtypes (Luminal A,Luminal B,. The molecular subtypes provide insights into new treatment strategies and patient stratifications that influence the management of BC patients (Łukasiewicz et al., 2021).
The use of computer simulations to detect drug response prediction has resulted in highly ambitious and all-encompassing goals to increase the likelihood of effective patient recovery. Due to recent rapid improvements and advancements in deep learning, a new trend has emerged in the search for computational drug response prediction models, resulting in more accurate medication response mechanisms (Adam et al., 2020). Exploration of diverse forms of genomewide molecular data and computational prediction of drug responses are essential for delivering the promise of precise functional therapy in oncology. This will benefit cancer patients by matching their tumor characteristics with the optimal treatment (Azuaje, 2017).
Drug repositioning is the process of identifying new therapeutic applications for existing medications and developing treatments for diseases that are currently untreated. Consequently, compared to standard de novo drug discovery methods, drug repositioning plays a crucial role in optimizing the pre-clinical process of developing innovative medications by saving both time and money (Abosheasha et al., 2022;Ashburn & Thor, 2004). Drug repositioning is based on publicly available data for existing medications and disorders, large-scale biological, biomedical, and electronic healthrelated databases, and high-performance computing, which have accelerated the development of computational drug repositioning techniques (Ashburn & Thor, 2004;Jarada et al., 2020). Molecular docking and dynamics simulation are extensively used to study many pathogens (viral, bacterial, and fungal) and cancer Kumar et al., 2022;Singh et al., 2022). The development of software and hardware in the last years enhanced the application of computer-aided drug design. Many medications are available based on these techniques against emerging pathogens such as SARS-CoV-2 (Adem et al., 2022;Elfiky, 2022;Elgohary et al., 2022).
The glucose-regulated protein of 78 kD (GRP78) or BiP is an endoplasmic reticulum-localized chaperone protein that is a member of the heat shock protein 70 family (member HSPA5). This protein is normally present in small amounts in adult cells, but it is activated by endoplasmic reticulum (ER) stress, such as glucose deprivation, hypoxia, and viral infection. It is overexpressed in tumor cells, with a proportion of the protein present on the cell surface (Delie et al., 2012;Elfiky et al., 2021). GRP78 comprises two functional domains: a 44-kDa N-terminal ATPase and a 30-kDa polypeptide-binding domain at the C-terminus (Delie et al., 2012). GRP78 is an indispensable component in protein synthesis. It facilitates the folding and assembly of newly synthesized proteins and prevents intramolecular and intermolecular aggregation under stress (Hendershot, 2004;Ni & Lee, 2007). Environmental and physiological stressors influence GRP78 composition, resulting in a loss of essential ER activities and homeostasis required to protect organs and tissues against apoptosis (Lee, 2001). According to developmental stages and tissue type, its expression varies. In the majority of adult tissues, the basal level is low, but it is expressed extensively in cancer (Dong et al., 2004;Li & Lee, 2006). Hypoxia and nutritional deprivation activate GRP78, which explains why it is so abundant in tumor cells (Lee, 2007). Additionally, GRP78 is reported to be overexpressed in the case of viral and fungal infections, and it has a role in the process of internalization of SARS-CoV-2 and Mucormycosis, causing fungus Khater & Nassar, 2022;Sabirli et al., 2021). SARS-CoV-2 spike can recognize cell-surface GRP78 as an auxiliary internalization route (Carlos et al., 2021). This recognition could be inhibited by using anti-GRP78 compounds (Elshemey et al., 2022;Palmeira et al., 2020).
Numerous studies reported various cellular and microenvironmental disturbances and numerous pharmacological interventions that could lead to increased GRP78 expression and exacerbated ER stress (Lee, 2001;Luo & Lee, 2013). Furthermore, GRP78 becomes the master initiator of early ER stress/unfolded protein response (UPR) signaling as a result (Serrano-Negr� on et al., 2018). Therefore, GRP78 is a promising target for a variety of antibodies or peptides that could be used for targeted cancer therapy or imaging (Farshbaf et al., 2020).
It has been shown that cell-surface GRP78 (cs-GRP78) makes cells more susceptible to viral and fungal pathogens (Elfiky, 2020a(Elfiky, , 2020b(Elfiky, , 2020cElfiky & Ibrahim, 2021a, 2021bGebremariam et al., 2014;Ibrahim et al., 2021). As a result, anti-infection drugs that target the substrate-binding domain (SBD) or nucleotide-binding domain (NBD) of GRP78 have been proposed as promising anti-infection therapies. Furthermore, combining drugs that inhibit GRP78 may be a novel strategy for enhancing the efficacy of microtubule-targeted chemotherapy (Wang et al., 2009). In the current study, we will investigate the efficacy of anticancer drugs against GRP78 in an effort to comprehend the mechanism by which these compounds eradicate cancer and to repurpose these drugs against pathogenic infections that use cs-GRP78 as an internalization route. The selection of the drugs is based on the cyclin-based kinases (CDK) 4/6 anticancer drugs that have recently yielded promising results or been approved by the FDA against many cancer types. This would enable the repurposing of anticancer drugs to treat a variety of fungal (Mucormycosis) and viral diseases (human coronaviruses, human papillomavirus, Zika virus, hepatitis B and C viruses) (Elfiky, 2020a(Elfiky, , 2020b(Elfiky, , 2020cElfiky & Ibrahim, 2021b;Elgohary et al., 2021;Gebremariam et al., 2014;Suwanmanee et al., 2021;Waris et al., 2002).

Structural retrieval
The structures of the positive control (À )-epigallocatechin gallate (EGCG) and the drugs Abemaciclib, Tunicamycin, Doxorubicin, Palbociclib, Ribociclib, and Dinaciclib were extracted from the PubChem database as 3D structure data file format (Kim et al., 2016). Drugs were prepared for the docking by adding Gasteiger charges and saved as PDBQT files utilizing AutoDock Tools software (Morris et al., 2009). GRP78 structure has a movable arm (SBDa), so it is important to explore the available states of this arm and equilibrate the system before docking. Therefore, GRP78 different conformations are used, which were generated in a previous study after molecular dynamics simulation (MDS) run for 50 ns (PDB ID: 5E84, Chain A, (see the supplementary file)), then the trajectories were clustered utilizing the pairwise best-fit RMSD of Chimera 1.14 software (Pettersen et al., 2004;Yang et al., 2015). The simulation was performed using TIP3P water model with 0.154 M NaCl as a buffer at physiological temperature (37 0 C) and 1 atm pressure (NVT ensemple). CHARMM27 force field was the choice in the nanoscale molecular dynamics (NAMD) 2.13 software to equibrate the GRP78 system for 50 ns (Elfiky, 2021;Huang & MacKerell, 2013;Phillips et al., 2005). In the present study, four conformations representing the most populated clusters of GRP78 trajectories were utilized. In order to prepare the GRP78 conformations for docking, any missing polar Hydrogen atoms, Kollman, and Gasteiger charges were added, while water molecules were eliminated. Both GRP78 and ligand files were saved in the PDBQT file format for docking calculations.

Molecular dynamics simulation of GRP78 NBD-drug complexes
Ligands-GRP78 NBD complexes were subjected to a molecular dynamics simulation run for a period of 100 ns at the Normal Volume and Temperature ensemble. The production run was preceded by a 1 ns equilibration run at the same ensemble. Next, using the TIP3P water model, the protein complexes were solvated and ionized with 154 mM NaCl at 310 K (Mark & Nilsson, 2001). The MDS was run using the Nanoscale Molecular Dynamics (NAMD) 2.13 software with the CHARMM 36 force field, while the Visualizing Molecular Dynamics 1.9.3 (VMD) software was used to prepare the system and analyze the data (Huang & MacKerell, 2013;Humphrey et al., 1996;Phillips et al., 2005). The CHARMM-GUI web server (Jo et al., 2008;Lee et al., 2016) was used to prepare the MDS input files. The binding free energy for complex formation is then calculated using Amber tools 20 and Molecular Mechanics Generalized Born Surface Area (MM-GBSA) method (Genheden & Ryde, 2015). In order to ensure the accuracy of the output values, the MM-GBSA calculations were performed using the frames corresponding to the equilibrated trajectories (after removal of the first 30 ns of the simulation, see Figure 4(A)). In addition, free energy decomposition was performed to determine each complex's residual contribution to the binding. The graphical abstract summarizes the steps taken in this study to test the anticancer drugs Doxorubicin, Tunicamycin, Abemaciclib, Palbociclib, Ribociclib, and Dinaciclib against GRP78.

Results and discussion
GRP78 was reported to be of paramount importance in cancer (Ibrahim et al., 2019;Lee, 2007). A greater understanding of GRP78 could provide additional strategies to enhance the effectiveness of GPR78 inhibitors and provide a greater understanding of novel resistance mechanisms to help target cancer cells with drugs after modification through repositioning to reduce cancer cell reproduction and drug resistance. In vitro studies suggested that GRP78 promotes chemoresistance to topoisomerase inhibitors like doxorubicin . A cohort study reported an association between GRP78 overexpression in BC patients and a shorter time to recurrence .
Many estrogen receptor-positive cancers lose medication sensitivity, making endocrine resistance a serious clinical problem. According to research into the molecular underpinnings of endocrine resistance, the UPR is a causative factor in antiestrogen resistance. Notably, GRP78, the master regulator of the UPR, was unexpectedly found to be elevated in endocrine-resistant cancer and directly influences antiestrogen treatment response. GRP78 has been found to influence multiple cellular processes, impacting cancer cells' survival. Furthermore, numerous substances that inhibit GRP78 activity have been identified, suggesting that combining these medications with antiestrogens could help overcome endocrine therapy drug resistance (Cook et al., 2013). GRP78 is a possible master regulator of endoplasmic reticulum (ER) functions and acts as a chaperone protein that contributes to chemoresistance. It is frequently overexpressed at the cancer cell surface. It is a well-studied ER stress signal, but it is also an unexplored drug development target. A variety of drugs that activate or inhibit GRP78, as well as compounds that bind to its NBD, are described. This domain is important for the protein function as it binds ADP or ATP and confers structural changes in the SBD and hence binds to or releases the substrate protein, respectively. There are numerous alternatives to positively or negatively influence chaperone expression or interfere with cellular trafficking (Bailly & Waring, 2019). Due to a lack of vascularization and the resulting hypoxia and glucose deprivation, tumor cells are susceptible to ER stress and UPR. GRP78 is a protein that prevents ER-stress-induced apoptosis in cancer cells (Dudek et al., 2009). CDK4/6 inhibitors, including abemaciclib, palbociclib, and ribociclib, interfere with molecular cycle progression, induce molecular senescence, and may promote molecular cancer disruption via a cytotoxic T cells-mediated effect. In phases I, II, and III clinical trials, the role of CDK4/6 inhibitors in monotherapy for numerous stable tumors is currently being evaluated. Abemaciclib is currently the most effective of the three inhibitors approved as a single-agent treatment for pretreated hormone receptor-positive (HRþ) HER2-negative (HER2À ) metastatic BC (Schettini et al., 2018). Dinaciclib (MK-7965, formerly SCH727965) is an extremely novel multi-CDK inhibitor (1/2/5/9) with promising pre-clinical results and an acceptable safety profile in Phase I clinical trials. Currently, it is undergoing clinical trials to treat hematological and solid cancers, including BC (Criscitiello et al., 2014;Saqub et al., 2020). In contrast, tunicamycin inhibited tumor cell growth in BC ER stress and regulated the expression of the GRP78 protein and mRNA in vitro (Serrano-Negr� on et al., 2018).
HER2-positive and -negative MBC are currently treated with CDK4/6 inhibitors and endocrine therapy (Shikanai et al., 2022). Neutropenia and gastrointestinal side effects are among the most common side effects of combination therapy based on CDK4/6 inhibitors. Abemaciclib is specifically associated with gastrointestinal toxicity. As it pertains to chemotherapy, dose reductions and dose adjustments can aid in reducing toxicity. The key to successfully treating patients, reducing side effects and treatment interruptions, and preventing a lack of confidence in this novel treatment (Thill & Schmidt, 2018) is, therefore, early and adequate monitoring with regular clinical assessments and management of side effects. Doxorubicin has numerous side effects, including exhaustion, baldness, nausea and vomiting, mouth sores, suppression of bone marrow, and an increased risk of subsequent cancers (Johnson-Arbor and Dubey). Furthermore, Tunicamycin disrupts hepatic energy homeostasis by causing triglyceride buildup and glycogen deficiency. This is due to abnormalities in hepatic energy metabolism, which are regulated by ER stress and lead to metabolic illnesses such as hepatic steatosis and hypoglycemia (Feng et al., 2017). Intravenous extravasation of doxorubicin can result in progressive tissue necrosis and severe ulceration. Doxorubicin is also associated with significant heart damage, which restricts the drug's long-term application. The mechanism by which doxorubicin induces cardiac toxicity is distinct from its anticancer mechanism. It increased oxidative stress, downregulated cardiac-specific genes, and induced cardiac myocyte apoptosis due to doxorubicin (Johnson-Arbor and Dubey).
Due to its lack of tumor specificity and narrow safety margins, chemotherapy faces the challenge of achieving higher efficacy and toxicity levels. A number of anticancer studies have demonstrated that phytochemicals or bioactive compounds, such as polyphenol and carotenoids, have promising or adjuvant efficacy for gynecological cancer with rare or mild side effects. Due to the side effects of established cytotoxic agents, numerous researchers have investigated the anticancer and synergistic effects of EGCG (Huang et al., 2020). EGCG could elevate the GRP78 levels in the ER while lowering its expression at the cell surface in a mesothelioma cell line (Martinotti et al., 2018). Thus, the use of EGCG as a drug delivery target is capable of lessening the chemoresistance to traditional anticancer drugs. A modeling study showed that EGCG attaches directly to GRP78, interacting with amino acid residues in the ATPase domain of the protein, stabilizing and folding the protein steadily (Gurusinghe et al., 2018). The binding affinity and selectivity of EGCG toward GRP78 were higher than that toward another HSP70 protein. Interestingly, inhibition of GRP78 activity by EGCG also enriches the chemo-sensitivity of both glioma cells and human brain endothelial cells (Virrey et al., 2008).
Palbociclib, ribociclib, and abemaciclib (see Figure 1(A)) are third-generation CDK4/6 inhibitors that the FDA has approved in the last decade for BC associated with endocrine therapy (Braal et al., 2021;Eggersmann et al., 2019;Petrelli et al., 2019). These compounds compete with ATP and have the pyridineamine-pyrimidine scaffold that represents the activation of the estrogen receptor pathway in the cell. Serine/threonine kinases CDK4 and CDK6 are activated by their association with cyclin D and inhibited by INK4 family CDK inhibitors (Braal et al., 2021;Du et al., 2020). Alternatively, tunicamycin (Figure 1(A)) induces ER stress and inhibits the proliferation of ERÀ /PRÀ / HER2À and ER þ human BC cells (Serrano-Negr� on et al., 2018). Additionally, dinaciclib (SCH727965) is a potent and selective small-molecule inhibitor of CDKs. At nanomolar concentrations, this compound inhibits CDK1, CDK2, CDK5, and CDK9 (Saqub et al., 2020). It is active against various human cancer cell lines and inhibits tumor growth in animal models, as it has an excellent safety profile in mice (Saqub et al., 2020).

Molecular docking
Prior to performing the docking calculations, a redocking study was conducted in which ATP was docked into the NBD of GRP78. Then, using PyMOL, we superimpose the docked complex onto the solved structure (PDB ID: 5E84). The complex has a root-mean-square displacement value of zero, and the total number of comparison atoms is 4741. Figure  1(B) depicts the superposition of the docked complex and the solved structure, with the enlarged panel illustrating how the positions of the ligand atoms in the two complexes are distinct.
Prior to this, we simulated the dynamics of the GRP78 structure (PDB ID: 5E84) for 50 ns, clustering the trajectories into four groups (Elfiky, 2021). In the present study, one representative conformation is chosen to test the binding affinities of the investigated compounds (at 17.8, 26.2, 31.8, and 37.8 ns of the simulation). Figure 2 displays the average binding affinities (in kcal/mol) of EGCG (positive control) (red column) and the drugs abemaciclib, tunicamycin, doxorubicin, palbociclib, ribociclib, and dinaciclib (blue columns) for GRP78 NBD (A) and SBD (B). Error bars represent the standard deviations. Abemaciclib and tunicamycin have a lower (better) average binding affinity to GRP78 NBD than EGCG, as shown in Figure 2A (À 10.45 and À 10.43 kcal/mol, respectively). Additionally, the average binding affinity of doxorubicin to GRP78 NBD is comparable to that of EGCG (À 9.90 kcal/mol). Alternatively, palbociclib, ribociclib, and dinaciclib have marginally higher average binding affinities for GRP78 NBD than EGCG (À 8.90, À 8.75, and À 8.60 kcal/ mol, respectively).
The majority of drugs have lower binding affinities for the GRP78 NBD than for the SBD domain. Furthermore, the positive control compound (EGCG) binds with 14% lower affinity to the NBD domain than to the SBD domain. The average affinity of abemaciclib, tunicamycin, doxorubicin, palbociclib, ribociclib, and dinaciclib for GRP78 NBD is lower by 20.4%, 22.7%, 21.8%, 4.1%, À 0.3%, and 7.5%, respectively, than that for GRP78 SBD. Additionally, we used the DoGSiteScorer tool of Proteins Plus-Structure Base Modelling Support Server (https://proteins.plus/) to evaluate the drugability of EGCG in a pocket within GRP78 structure (PDB ID: 5E84, Chain A). Based on DoGSiteScorer, the NBD has a volume of 2328.56 Å 3 , a surface area of 2324.81 Å 2 , a drug score of 0.81, and a simple score of 0.6. On the other hand, the SBD has a volume of 131.6 Å 3 , a surface area of 285.96 Å 2 , a drug score of 0.25, and a zero simple score. We can conclude that GRP78 NBD is more enticing to the studied drugs and may serve as the docking platform for these drugs in stressed cells (expressing more GRP78). The results are comparable with the previous results of docking the EGCG to NBD of GRP78 (Gurusinghe et al., 2018). In that study, EGCG has a binding affinity value of À 9.9 kcal/mol, while in the current study, it has an average value of À 9.875 ± 0.676 kcal/mol (Figure 2(A)). Table 1 lists the established interactions formed upon docking the drugs against both NBD and SBD of GRP78. The main types of interactions that are established are the formation of H-bonds, hydrophobic interactions, and a few salt bridges. These bold residues are the active site residues that formed direct contacts between the two domains (D34, T37, T38, Y39, K96, K296, R297, S300, G364, and R367 for GRP78 NBD and I426, T428, V429, V432, T434, F451, S452, V457, and I459 for GRP78 SBD). All drugs except dinaciclib were able to bind to the active site (bold residues) of the GRP78 NBD with excellent average binding affinities (À 9.69 ± 0.73 kcal/ mol). The most recorded residues from GRP78 NBD that interact with the drugs were R297 (11), R367 (7), T38 (6), Y39 (5), D391 (5), D259 (4), E293 (4), D34 (3), and K296 (3), with the values between the brackets representing the number of occurrences according to Table 1. However, only palbociclib could bind to the GRP78 SBD active site (bold residues) with an average binding affinity of À 8.775 ± 0.29 kcal/mol. Figure 3 shows established interactions in selected ligands-GRP78 NBD complexes depicted by PyMOL software. Blue lines represent H-bonds, while dashed gray lines represent hydrophobic interactions. Yellow dashed lines represent salt bridges, whereas yellow spheres represent atoms that interact. As shown in Table 1 and Figure 3, the GRP78 NBD forms at least two H-bonds and one hydrophobic contact with the ligands.

Molecular dynamics simulation of the complexes
To further investigate the binding behavior of the ligands, we decided to perform 100 ns of molecular dynamics Red and blue residues represent the residues that interact with p-stacking and p-cation interactions, respectively. Bold residues are the active residues according do the literature.  the previous study by Gurusinghe et al., the RMSD was between 6 and 11 Å. In addition, the RoG and SASA values for the complexes demonstrate that the systems were stable throughout the duration of the MDS, with average RoG and SASA values of 30 Ð and 32,500 Ð 2 , respectively. The simulation reveals that the systems are stable, as the total number of H-bonds averages approximately 980. Figure 4(E) also illustrates the per-residue root-mean-square fluctuation (RMSF) in Ð for the seven ligand-GRP78 NBD complexes. The NBD structure of abemaciclib-GRP78 is depicted in color in the illustration at the top of the figure.
As reflected by the enlarged RMSF curves in Figure 4(G), the tunicamycin-GRP78 complex (gray curve) has a lower RMSF value (2.9 Ð) than other complexes for the D131-K138 (marked by an asterisk in Figure 4(E)), Y270-D281 region, and I309-E320 region. For the D131-K138 region, the dinaciclib-GRP78 (blue curve) and abemaciclib-GRP78 (orange curve) complexes demonstrate greater fluctuations (RMSF of 5.2 Ð) than the positive control EGCG-GRP78 (black curve) complex (RMSF of 4.2 Ð). In the D408-G412 (yellow) region, abemaciclib-GRP78 has higher RMSF values than other complexes. The same trend of higher fluctuation is observed for dinaciclib-GRP78 (blue curve) in comparison to other complexes in the T428-M433 and L468-D471 regions (Figure 4(F), arrow) (RMSF of 5.4 and 6.6 Ð versus 4.25 and 3.67 Ð, respectively). This indicates the instability of dinaciclib-GRP78, as the active site region displays significant fluctuations. In contrast, the tunicamycin-GRP78 complex continues to exhibit lower fluctuations in all regions, as indicated by the enlarged RMSF panels, when compared to other complexes. The dynamics of the various drug-GRP78 complexes showed that tunicamycin-GRP78 was the most stable complex, followed by doxorubicin-GRP78 and ribociclib-GRP78 complexes.
Again, these values are close to the previously reported values for GRP78, GRP78-ATP, and GRP78-EGCG systems simulated for 50 ns by Gurusinghe et al. (2018). The SBDa has the highest fluctuations (RMSF <7.5 Å) for all the systems except tunicamycin-GRP78 and doxorubicin-GRP78 complexes, which have lower RMSF values (5 Å).

MM-GBSA of the best complexes
Using Amber tools 20, we performed MM-GBSA calculations to further analyze the complexes (Genheden & Ryde, 2015). The MM-GBSA has performed on the 30 À 100 ns trajectories because the first 30 ns of the simulation were required to equilibrate the systems based on the RMSD curves. Table 2 displays the residual contribution to the binding of EGCG, abemaciclib, ribociclib, and tunicamycin to GRP78 NBD (upper portion of Table 2) and the binding energy (kcal/mol) decomposition into its components: Van der Waal's term (DEVDW), Electrostatic term (DEELE), generalized Born (DGGB), and surface area (DGSA) (lower part of the table). These four drugs listed in Table 2 are the most stable GRP78-binding complexes based on the VMD software-examined binding trajectories. In contrast, the doxorubicin-GRP78, dinaciclib-GRP78, and palbociclib-GRP78 complexes are not stable throughout the duration of the simulation, as the drugs leave the protein after 95, 85, and 55 ns, respectively (see also Figure 4(E)).
In Table 2, amino acids are listed and ranked (in ascending order) according to their binding affinity values (in kcal/ mol). In addition, red-colored residues contributed negatively (positive binding energy) to the binding, whereas residues in bold are the active site residues. Among these active site residues, Y39 is the most significant contributor to binding in all complexes, contributing 9.4%, 13.1%, 16.2%, and 12.6% of the total binding energy in EGCG-GRP78, abemaciclib-GRP78, ribociclib-GRP78, and tunicamycin-GRP78, respectively. These percentages were calculated by dividing the Y39 binding contribution by the total binding energy of the complex. Additionally, other residues contribute enormously to the binding in the four complexes. For example, R297 has a high contribution in EGCG (À 2.08 kcal/mol) and tunicamycin (À 4.61 kcal/mol) binding to GRP78, but a negative contribution in the binding in the case of ribociclib (þ0.40 kcal/mol). On the other hand, T37 has a moderate positive contribution in binding abemaciclib (À 0.77 kcal/mol), ribociclib (À 0.95 kcal/mol), and tunicamycin (À 1.20 kcal/mol) to GRP78. At the same time, it does not contribute at all in the case of EGCG. Additionally, T38 has a moderate positive contribution in binding EGCG (À 0.44 kcal/mol), ribociclib (À 0.80 kcal/mol), and tunicamycin (À 0.72 kcal/mol) to GRP78, while it does not contribute at all in the case of the abemaciclib-GRP78 complex. G364 has a moderate positive binding contribution in EGCG (À 1.30 kcal/mol) and ribociclib (À 1.28 kcal/mol) to GRP78. In comparison, it has a weak positive contribution in binding abemaciclib (À 0.24 kcal/mol) and tunicamycin (À 0.42 kcal/mol) to GRP78. The UNK is the drug molecule that has a paramount effect in binding in the case of the positive control (À 3.08 kcal/mol) and the two drugs ribociclib, and tunicamycin (À 9.67 and À 5.17 kcal/mol, respectively), while in abemaciclib, it has a powerfully negative effect on the binding (þ2.91 kcal/mol), but balanced by other binding contributors.
Conclusively, we can suggest the three drugs, abemaciclib, ribociclib, and tunicamycin, can tightly bind to GRP78 NBD, and the binding complexes are stable during the 100 ns simulation. Furthermore, the decomposition of the binding energy indicates that the residues Y39, R297, T37, T38, and G364 are the primary driving force of the binding among the drug molecule itself (in the case of ribociclib and tunicamycin). So the latter two drugs may successfully target GRP78 either in cancer or pathogenic infections that use GRP78 as an internalization route . However, these results need to be verified experimentally by in vitro and in vivo assays before moving to clinical trials. Also, ribociclib and tunicamycin can be used as a seed for effective new compounds blocking GRP78 in stressed cells.

Conclusion
GRP78 is a crucial component of host cells in cancer and pathogen infection (fungal and viral). Based on our modeling study, various FDA-approved anticancer agents are potential candidates against GRP78. Abemaciclib, ribociclib, and tunicamycin are potent anticancer agents that are able to target GRP78 NBD and therefore can oppose GRP78 function in stressed cells. This inhibition of the GRP78 function suppresses cancer resistance and prevents cs-GRP78 from recognizing pathogens. In contrast, since cs-GRP78 is one of COVID-19's confirmed receptors (recognizing its spike protein), these anticancer agents may be repurposed as possible therapeutic agents against the current COVID-19 pandemic. Moreover, it may be optimal to use these candidates in the case of the COVID-19-cancer correlation.