Proximicin A-C as Prospective HER2-Positive and Negative Breast Cancer Drugs: Molecular Docking and In silico ADME Modeling

24 History Breast cancer is the second leading cause of cancer-related death among women worldwide. Proximicin A-C are bioactive chemicals produced by the marine Verrucosispora strain, which have been shown to have a potent cytostatic effect against human breast cancer [MCF 7]. HER2 (Human epidermal growth factor receptor 2) is a gene that has been linked to breast cancer development. The study’s goals are to (1) forecast the intensity of binding affinity and interactions between HER2 and proximicin A-C, and (2) analyze ADME characteristics of proximicin A-C (Absorption, Distribution, Metabolism, and Excretion). The CB-Dock web service was used to dock proximicin A-C and commonly used breast cancer medicines Neratinib (Nerlynx) and Talazoparib against HER2, and protein-ligand interaction findings were collected via the protein- ligand interaction profiler (PLP) web server. The SwissADME web server was used to investigate ADME properties of the substances. In terms of docking, proximicin A has a vina score of -8.6, proximicin B and C has a score of -10, Talazoparib has a vina score of -8.5, and Neratinib (Nerlynx) has a vina score of -10.2 on CB-Dock. This means that proximicin B and C bind to HER2 more strongly than proximicin A and Talazoparib. Furthermore, their high binding affinity is nearly equivalent to Neratinibs (Nerlynx). Talazoparib has a lower binding affinity for HER2 than proximicin A. With HER2, all three chemicals have a strong hydrogen bond and hydrophobic contact. SwissADME estimated that all three substances follow the Lipinski rule (RO5) and have a bioavailability score of 0.55. They don’t have any structural issue in medicinal chemistry (no alerts in PAINS and Brenk forecasts), and their synthetic accessibility scales range from 3 to 3.5. Only proximicin A, on the other hand, has the leadlikeness feature. All three drugs failed to cross the blood-brain barrier (BBB) in terms of pharmacokinetics. Proximicin A has a high absorption rate in the GI tract, whereas proximicin B-C has a low absorption rate in the GI tract (GI). Similarly, proximicin A is neither a P-gp substrate nor a CYP1A2, CYP2C19, CYP2C9, CYP2D6, or CYP3A4 inhibitor. Proximicin B-C, on the other hand, are P-gp substrates, and proximicin C is an inhibitor of all provided CYP enzymes, whilst proximicin B inhibits only three. Overall, proximicin A-C could be used as a possible breast cancer therapeutic candidate. Proximicin B-C will outperform proximicin A in terms of therapeutic efficacy. Proximicin A, on the other hand, will have better ADME qualities than Proximicin B-C. This study will provide the lead information for developing a new breast cancer medication with a good pharmacological


Introduction
The uncontrolled development and duplication of tissue cells in the breast causes breast cancer. Breast cancer is still the main cause of cancer death in women around the world today. One in every eight women will be diagnosed with breast cancer at some point in their lives, and one in every 38 will succumb to the disease. [1] Breast cancer is classified into two categories based on the type of proteins found in breast cells that cause it: (1) Human epidermal growth factor receptor 2 (HER2)-positive and HER2-negative breast cancers, as well as hormone-receptive breast cancers (either estrogen-receptive or progesterone receptive cancer). [2] HER2-positive is a breast cancer that tests positive for the protein human epidermal growth factor receptor known as human epidermal growth factor receptor 2 breast cancer (HER2). This protein encourages cancer cell proliferation. Extra copies of the gene that produces the HER2 protein are found in around one out of every five breast tumors. Breast cancers that are HER2-positive are more aggressive than other kinds of breast cancer. [3] HER2-negative breast cancer suggests that the malignant cells do not have This work is published open access under the Creative Commons Attribution License 4.0, which permits free reuse, remix, redistribution and transformation provided due credit is given. high amounts of the protein HER2. There are a variety of therapy options for this form of breast cancer, however the prognosis varies. [4] Over the last two decades, medical advancements have resulted in the development of new, effective treatments. Trastuzumab (Herceptin) was the first FDA-approved targeted therapy for HER2-positive breast cancer. Trastuzumab is administered intravenously (via an IV) once a week or once every three weeks. [5] A marine Verrucosispora strain produces a family of three novel aminofuran antibiotics known as proximicin. Proximicins B, which has a molecular mass of about 413 kDa, can inhibit the growth of gram-positive bacteria, whereas proximicins C, which has a molecular mass of about 436 kDa, can only inhibit the growth of Brevibaccillus brevis. Surprisingly, gram-negative bacteria like E. coli K12, Pseudomonas fluorescens, and Proteus mirabilis, as well as yeasts like Saccharomyces cerevisiae, are all resistant to proximicins. Proximicins have a molecular structure that includes 4-amino-furan-2-carboxylic acid, a previously discovered -amino acid. Although they are mild bactericidal peptides, they have a strong cytostatic effect on human breast cancer (MCF 7). Proximicin C can produce cell cycle arrest at the G0/G1 phase after 24 hours and enhance the number of apoptotic cells after 40 hours, according to an in vivo study on gastric adenocarcinoma (AGS) cells. In AGS cells, it can also increase intracellular levels of p53 and the cyclin kinase inhibitor p21. [6] The molecular docking analysis of proximicin A-C against human epidermal growth factor receptor 2 (HER2) and the computational assessment of pharmacokinetic or ADME features for proximicin A-C are both part of this research project. The most promising route for drug design and discovery is through computational molecular docking and scoring. As a result, this research could lead to the development of a novel breast cancer treatment.

Materials and Methods Selection of ligands and protein
Three bioactive compounds, proximicin A-C from the Verrucosispora strain were retrieved in SDF format from pubChem (http://pubchem.ncbi.nlm.nib.gov/) based on public databases and published research articles. In addition, two standard medicines, Neratinib (Nerlynx) and Talazoparib, were downloaded in 3D SDF format for comparison with the phytochemicals, having Pubmed IDCID 135565082 and CID 9915743, respectively. PDB ID 3PP0 was used to get the 3D structure of the human endothelium receptor HER-2 from the RCSB protein data bank.

Molecular Docking study
The CB-Dock web server (http://cao.labshare.cn/cb-dock/) was used to analyze the docking of the compounds. The CB-Dock online server simply requires a protein file in PDB format and a ligand file in MOL2, MOL, or SDF format as input. Following submission, the CB-Dock examines the input files and uses OpenBabel and MG Tools to convert them to pdbqt formatted files. CB-Dock then predicts the protein's cavities and determines the centres and sizes of the top N (by default, n=5) cavities. AutoDock Vina is used to dock the centres and sizes, as well as the pdbqt files. After computing, the final findings are displayed. In the table, users can look up binding scores, cavity size, and docking parameters for projected binding modes. Users can also examine the 3D structures of any binding modes on the web page by clicking on the structures in the associated table. [7] Only the best pose of binding modes for each molecule was evaluated in this investigation, as shown in the Table  1.

Protein-ligand interaction
For protein-ligand interaction, the best poses of protein-ligand complex pdb file obtained from CB-Dock web server were submitted to protein-ligand interaction profiler (PLP) web server (http:plip-tool.biotec.tu-dresden.de) for each molecule. Protein-ligand interaction profiler (PLP) is a novel web service for fully automated detection and visualization of relevant noncovalent protein-ligand contacts in 3D structures, freely available at projects.biotec.tu-dresden.de/plip-web. It returns a list of detected interactions on single atom level, covering seven interaction types (hydrogen bonds, hydrophobic contacts, pi-stacking, pi-cation interactions, salt bridges, water bridges and halogen bonds. [8] In silico ADME Study SwissADME (http:// www. swiss adme. ch/ index. php) was used to investigate pharmacokinetics parameters in the ligands, such as absorption, distribution, metabolism, and excretion. Lipinski's rule of five (RO5), Ghose's, Egan's, Veber's, and Muegee's criteria were used to assess the drug-likeness of compounds. The PAINS and Brenk filters were used to determine whether chemicals were promiscuous. The leadlikeness approach was designed because it is critical to determine whether a given molecule is suitable for the start of lead optimization. The SwissADME online programme also generated in silico data for the key human cytochrome P450 (CYP) isoforms involved in drug metabolism, such as CYP2C9, CYP2D6, and CYP3A4. SwissADME can additionally provide the BOILED-Egg model, a synthetic accessibility score, and computational filters like Ghose, Egan, Veber, and Muegee PAINS, as well as Brenk. The SwissADME online tool was also used to determine molecular parameters like MW (molecular weight), HBD (hydrogen bond donor), HBA (hydrogen bond acceptor), log P (lipophilicity log), log S (aqueous solubility), TPSA (topological polar surface area), MW, nRot (number of rotatable bonds), and MR (molar refractivity). [9][10] Results and Discussion Molecular docking study Molecular docking is a method for predicting a ligand's optimal orientation, affinity, and interaction in a protein's binding site. Using scoring functions, the preferred orientation can be utilized to predict the strength of binding affinity between the therapeutic target and the ligand molecule. Because of its medical uses, the protein-ligand interaction is the most intriguing instance. [11] The three bioactive molecules, proximicin A-C, were tested against the human epidermal growth factor receptor 2 in molecular docking studies (HER2). HER2 proteins aid in the rapid growth of breast cancer cells. Neratinib (Nerlynx) and Talazoparib, two commonly used breast cancer medicines, were also docked against the target receptor to compare their binding affinity to that of proximicin A-C.
Neratinib (Nerlynx) is a kinase inhibitor that prevents cancer cells from growing by blocking the action of HER2 and other kinase proteins. After a year of trastuzumab, it's utilised to treat HER2 positive early breast cancer. It is a tablet that must be taken on a daily basis. Talazoparib is a drug that is used to treat advanced or metastatic HER2-negative breast cancer in women who have had chemotherapy and have a BRCA gene mutation. [5] The CB-Dock web server calculates vina ratings for binding affinity and binding modes for each chemical ligand. Only the best position of each molecule was taken into account, as shown in the table. Vina score is a quantitative criterion for determining the affinity, or efficiency, of proteinligand interactions. [12] The vina scores of proximicin B and C are the same [-10], and they are much lower than those of proximicin A [-8.6] ( Table 1). Proximicin A-C has a lower score than Talazoparib [-8.5], and Neratinib (Nerlynx) has a poorer score than all three substances when compared to the conventional medications. A strong binding affinity between protein and ligand is indicated by a low vina score. [13] As a result of their vina score, it may be deduced that proximicin A-C have a high binding affinity for HER2. Furthermore, when compared to proximicin A and Talazoparib, Proximicin B and C show a greater binding affinity for HER 2. The strength of the connection between the drug and its receptor is defined by affinity. The stronger the connections, the more the ligand will alter the physiological function of the target proteins; thus, drug candidates are chosen from ligands that bind strongly to the target protein. Because the anticipated binding affinity of ligands in a library can be used for virtual screening or lead optimization, precise binding affinity prediction can lower the cost of a de novo drug design. A low dosage required is usually accompanied with a high affinity (compared with low affinity for the same receptor). [14]  If ligands can make hydrophobic interactions with the binding site's hydrophobic amino acids, the binding affinity will be higher. This is why, in comparison to proximicin B and C, proximicin A has a lesser affinity. [15] Proximicin B and C will have a higher therapeutic efficacy than proximicin A, according to the findings of this study. However, because their binding affinity is comparable to that of typical medications, proximicin A-C have the potential to act as breast cancer treatments.

In silico ADME Study
The flow of a medicine through the body's biological systems is referred to as pharmacokinetics. Absorption, distribution, metabolism, and excretion (ADME) are the four processes of the pharmacokinetic phase [1]. The ADME study has recently been added to incorporate the toxicological evaluation of new medication candidates. The distribution and fate of pharmacological chemicals within an organism, particularly in the human body, is described by ADME. The biggest cause of failure during the drug development phase is the candidate compound's poor pharmacokinetics (PK) and toxicity in compared to its efficacy. [16] Many bioactive chemicals, for example, have demonstrated substantial anticancer activity, but their use is restricted because to their harmful and lifethreatening side effects and toxic consequences. [17] Both high-throughput experimental and computational (in silico) directions are now available. Drugtransporter interactions involving P-glycoprotein transporters in the intestines frequently result in poor absorption and low oral bioavailability because the drug is easily effluxed backed into methods have become well-known for obtaining ADME or PK properties of a large number of compounds during the early stages of drug discovery. [18] In the gastrointestinal system, proximicin A was found to have a high absorption rate. Given the advantages of the oral mode of administration, this means that Proximicin A has a highly favourable feature of a medication candidate. The microvascular endothelial cell layer of the brain that separates the brain from the blood is known as the blood brain barrier (BBB). Only chemicals that target the central nervous system are required to penetrate the BBB (CNS). [19] The target chemicals do not appear to be capable of bridging the BBB based on the computational findings. Because none of the chemicals showed the ability to cross the BBB, this may be a benefit because they are less likely to cause deleterious effects in the CNS (Table 5).
P-glycoprotein (P-gp) is a type of membrane transporter that transports chemicals in the intracellular or extracellular lumen of the intestine and excretes them. Furthermore, P-glycoprotein inhibits the uptake of a wide range of structurally and functionally varied drugs, including the majority of cancer treatments, resulting in multidrug resistance. P-glycoprotein is also overexpressed in cancer cells, posing a substantial treatment hurdle by causing drug efflux and rendering chemotherapy ineffective. [20] Compound 1 was determined to be P-gp non-substrates. This means that the compounds would be unaffected by P-efflux gp's action, which causes compounds to be eliminated from cells, resulting in therapeutic failure due to lower concentrations than expected. Proximicin B and C are thought to be P-gp substrates. This means that they may have poor absorption and oral bioavailability, resulting in multidrug resistance as a result of their potency. Because it is not a P-gp substrate, proximicin A is the best candidate for cancer treatment against multidrug-resistant cancer cells ( Table 5).
The basic enzymes for drug biotransformation are the cytochrome P450 (CYP) enzymes. CYP1A2 inhibitor, CYP2C19 inhibitor, CYP2C9 inhibitor, CYP2D6 inhibitor, and CYP3A4 inhibitor are the most important inhibitors in biotransformation. CYP isoenzymes are involved in xenobiotic detoxification, cellular metabolism, and homeostasis. As a result, drug metabolism via CYP isoenzymes is a key determinant of drug interactions, which can result in drug toxicity and a reduction in pharmacological activity [21][22] . CYP3A4 is not only the most common CYP enzyme in the liver, but it is also utilised by more than half of all drugs on the market for metabolism and elimination from the body. Both proximicin B and C are predicted to inhibit CYP3A4, which could be a disadvantage because this CYP isoform is involved in the metabolism and elimination of the majority of clinically used drugs, including calcium channel blockers, some stains, immunosuppressors, macrolides, and atypical antipsychotics, among others. [23] Proximicin A would have no effect on CYP1A2, CYP2C19, CYP2C9, CYP2D6, or CYP3A4, however proximicin C would inhibit all of the CYP enzymes. CYP1A2 and CYP2C19 were not inhibited by proximicin B, although it did inhibit CYP2C9, CYP2D6, and CYP3A4. Because proximicin A has no inhibitory effect on these enzymes, the chemicals have a high chance of being converted and hence accessible after oral treatment ( Table 5).
Inhibition of the CYP isomers by proximicin B and C, on the other hand, can result in poor bioavailability due to failure to be metabolised and hazardous side effects due to their buildup. Proximicin B, on the other hand, may be processed by more than one enzyme, lowering the probability of a drug-drug interaction. The skin acts as a selective barrier, allowing various chemicals to enter at different rates based on their physicochemical qualities. As a result, skin permeability (LogKp) is an important characteristic to consider when evaluating medicines that may require transdermal delivery. [23] Table 5 shows the LogKp of the various compounds. Because they had negative values between -8.7 and -6.2 cm/s, all of the chemicals should be accessible. This means that none of the three chemicals could be delivered efficiently through the skin.
The term "drug-likeness" refers to how likely a molecule is to become an oral drug in terms of bioavailability. Structure or physicochemical inspections of research compounds progressed enough to be deemed oral drug candidates were used to determine drug-likeness. [22] As a result, swissADME was used to assess the physiochemical characteristics and lipophilicity of compounds in Tables 7 and 8. The Lipinski (Pfizer) filter was the first of its kind in terms of predicting oral medication candidates. Many extensions have been provided to improve the predictions of drug candidates for oral administrations, such as The Ghose (Amgen), Veber (GSK), Egan (Pharmacia), and Muegge (Bayer) guidelines. There are five different types of rule-based filters, as follows: 1. Molecular weight 500, MLOGP (lipophilicity) 4.15, hydrogen bond acceptors 10, and hydrogen bond donors 5 are all included in Lipinski's filter. [24] 2. Ghose's filter has a molecular weight of 480, a lipophilicity of 0.4 WLOGP (5.6), a molar refractivity of 40, and a number of atoms of 20. [25] 3. The number of rotatable bonds is 10 and the total polar surface area is 140 in Veber's filter. [26] 4. WLOGP (Lipophilicity) 5.88 and total polar surface area 131.6 are included in Egan's filter. [27] 5. Muegge's filter includes the following parameters: 200 molecular weight 600, 2 XLOGP3 (lipophilicity) 5, total polar surface area 150, number of rings 7, number of carbon > 4, number of heteroatoms > 1, number of rotatable bonds 15, hydrogen bond acceptors 10, and hydrogen bond donors 5. [28] The Lipinski's rule, Ghose's rule, and Muegge's rule are all in agreement with proximicin A-C. Proximicin A met Veber's qualifying conditions, however proximicin B broke two regulations (Rotors>10 and TPSA>140) and proximicin C broke one rule (Rotors>10), according to Veber's rule. All of the compounds in the Egan's rule cases had one violation (TPSA>131.6) ( Table  6). If a substance violates more than one of Ro5, poor oral absorption of medicinal molecules is found. A compound is predicted to be a non-orally accessible medication if two or more of these conditions are violated. [23] All of the compounds were orally bioavailable, meeting the requirements of the Lipinski's, Ghose's, Egan's, and Muegge's filters. According to Veber's filter, proximicin B is expected to have some oral bioavailability issues, whereas proximicin A and C are expected to have acceptable oral bioavailability. Proximicin A-C, on the other hand, has a bioavailability of 0.55. This indicates that the chemicals have a 55% chance of being bioavailable. The bioavailability score is an indicator of drug material oral absorption. Any medication molecule with a BA score of 0.55 that meets the rule of five is regarded suitably absorbable via oral route. [23] Proximicin A-C had a score of 0.55, indicating that it has good oral bioavailability (Table 6).
There are some basic guidelines for determining the medicinal chemistry of produced substances. The physicochemical filters PAINS and Brenk are used to anticipate the compounds that have poor pharmacokinetic properties. PAINS is a filter that helps to identify a molecule by determining if it is a reaction to biological assays or not, and Brenk is a filter that helps to identify compounds with the acceptable hazardous level, chemical reactivity, and metabolic instability. [29][30] Proximicin A-C are structurally sound and unlikely to be reactive or hazardous.
The synthetic accessibility value is a number based on fragmental structural research; the more molecular fragments there are, the easier the molecule is to produce. Descriptors for molecule size and complexity correct this fragmental contribution approach. The score ranges from 1 (easy synthesis) to 10 (very difficult synthesis). [22][23][24][25][26][27][28][29][30][31] Smaller synthetic accessibility numbers are often easier to synthesise than bigger values between 1 and 10. Proximicin A-C may be produced with synthetic accessibility scales of 3.01, 3.31, and 3.42, according to the findings ( Table 6).
The concept of leadlikeness is comparable to that of druglikeness. Leads are subjected to chemical changes in order to improve their size and lipophilicity. As a result, leads must be less hydrophobic and smaller than drug-like compounds. The following are the qualifying rules for the lead likeness property: 250 = molecular weight=350, XLOGP=3.5, and number of rotatable bonds=7. [22] Only Proximicin A has no violations in the lead likeness property, indicating that it meets the standards for a lead compound, but Proximicin B-C have two violations in the lead likeness parameter (MW>350 and Rotors>7).

Conclusion
In this study, when compared to standard medications (Neratinib (Nerlynx) and Talazoparib), docking studies demonstrated that proximicin A-C have a decent ability to operate as possible breast cancer drugs since they all have a comparable high binding affinity with human epidermal growth factor receptor 2 (HER2). Proximicin B and C, on the other hand, will have a stronger therapeutic efficacy than proximicin A since they have a higher binding affinity for HER 2. All three compounds, proximicin A-C, had good oral bioavailability and they all met Lipinski's rule of five for oral drugability (RO5). However, none of the compounds were able to cross the blood-brain barrier (BBI). Because proximicin A is the only one of these three compounds that isn't a P-gp substrate and doesn't inhibit any cytochrome P450 (CYP) enzymes, it will be the most suitable oral medication choice. It also has a high gastrointestinal absorption rate and follows the druglikeness and leadlikeness principles. This research can be used to develop a new breast cancer medication with a favourable pharmacological profile.