Abstract
Background/Objectives: Pomegranate (Punica granatum) peel, often
discarded as waste, contains abundant bioactive compounds such as
polyphenols, vitamins, flavonoids, tannins, anthocyanins, and many
more. This contributes to remarkable bioactivities, including
anticancer, anti-inflammatory, antioxidant, antibacterial, and
antifungal properties. Pancreatic cancer is a deadly cancer with a 9%
survival rate. Its aggressiveness, invasiveness, quick metastasis, and
poor prognosis significantly decrease the survival rate. Thus, we aim
to explore pomegranate peel as a possible alternative medication for
treating pancreatic cancer through virtual methods. Methods: Firstly,
bioactive compounds were collected from multiple databases and screened
for oral bioavailability (OB) ≥ 0.3 and drug likeness (DL) ≥ 0.18
scores. Simultaneously, network pharmacology was employed to extract
the most probable targets for pancreatic cancer. Further computational
analyses were performed, including molecular docking, molecular
dynamics simulation, and in silico pharmacokinetics evaluation.
Results: Consequently, the top 10 key targets from network analysis
were AKT1, IL6, TNF, SRC, STAT3, EGFR, BCL2, HSP90AA1, HIF1A, and
PTGS2. However, only AKT1, EGFR, BCL2, HSP90AA1, and PTGS2 exhibited
strong binding affinities with pomegranate compounds, which are
significantly declared in affected cells to enhance cancer progression.
Outcomes from molecular dynamics simulations, particularly RMSD, RMSF,
hydrogen bonding, and radius of gyration (Rg), confirmed stable
interactions between 1-O-Galloyl-beta-D-glucose, epicatechin,
phloridzin, and epicatechin gallate with respective target proteins.
Conclusions: This suggests that pomegranate peels hold anticancer
bioactive compounds for treating pancreatic cancer. Surprisingly, most
compounds adhere to Lipinski’s and Pfizer’s rules and display no
toxicity. However, as this study relies entirely on computational
methods, experimental validation is necessary to confirm these findings
and assess real-world efficacy and potential side effects.
Keywords: pomegranate peel, pancreatic cancer, network pharmacology,
molecular docking, molecular dynamics simulation
1. Introduction
Pancreatic cancer (PC) is a fatal malignancy where an uncontrolled mass
of cells is found in pancreatic tissue. Globally, PC is regarded as
seventh most prevalent type of cancer, with a larger percentage of
cases occurring in developed countries [[30]1]. Several factors
contribute to PC in humans, including abnormal microbiological
metabolism, blood glucose, aging, alcohol consumption, smoking,
obesity, diabetes, hereditary factors, polluted air and water, and poor
diet [[31]2,[32]3]. Due to the absence of specific symptoms, most
patients are found to be diagnosed in the late stages of the disease,
which is commonly known as late-stage PC. The survival rate of PC
patients is less than 10 % compared to other types of cancers [[33]4].
Certain treatment methods, such as surgery and chemotherapy, have shown
limited success in improving prognosis. As per GLOBOCAN 2020,
approximately 495,773 new pancreatic cancer cases were recorded in
2020, which is 2.6 percent of all new cancer cases in 2020, as reported
by the Agency for Research on Cancer. In total, 466,003 deaths were
reported, representing 4.7% of the total cancer-related mortality rate
[[34]5].
Natural plant-based compounds have become a viable substitute due to
their distinct biological and pharmacological properties [[35]6,[36]7].
In recent years, people have been widely utilizing prior knowledge
about bioactive products because of their anti-inflammatory,
antibacterial, anticancer, radioprotective, anti-tumor properties and
lesser toxicity [[37]8]. It has also been found that natural products
can cause apoptosis in PC cells as they regulate the expression of
protein kinases, mitochondrial pathway, or genes that are involved in
apoptosis. They can also inhibit the angiogenic process, limiting the
growth and spread of cancer cells by obstructing the blood supply to
the tumor cells [[38]7].
Pomegranate (Punica granatum) is widely recognized for its health
benefits, primarily attributed to higher concentrations of
phytochemicals [[39]9,[40]10,[41]11]. However, this research focused on
pomegranate peels (PPs), which are considered to be non-edible parts of
pomegranates (PG) and treated as waste but have potential sources of
bioactive ingredients such as polyphenols, flavonoids, anthocyanins,
organic acids, alkaloids, and hydrolyzable tannins [[42]12,[43]13].
Previous research has demonstrated that the peels of pomegranates
possess anti-inflammatory, antimicrobial, anticancer, antioxidant, and
cardioprotective properties [[44]14,[45]15,[46]16]. Moreover, it has
nutraceutical properties contained in minerals, fats, carbohydrates,
proteins, and fibers [[47]8]. There is evidence that pomegranate peel
polyphenols, flavonoids, and anthocyanin impact several diseases,
including diabetes and metabolic syndrome, cardiovascular diseases,
male infertility, Alzheimer’s disease, gastrointestinal diseases, and
cancer inflammation [[48]17]. Hence, this study involves network
pharmacology to find the most relevant pancreatic cancer target protein
for pomegranate-derived compounds and to investigate drug interactions
at different target sites and related diseases by constructing
“drug-gene target disease” and its associated pathways. Additionally,
pharmacokinetic properties and toxicity assessments of all compounds
are carried out. Molecular docking is employed to predict the binding
pose between two molecules, a ligand and receptor. Further, we
determine how well molecules fit into the receptor pocket by
calculating the binding affinity, which is expressed as docking score
along with key interacting amino acids of proteins with compounds.
Simultaneously, molecular dynamic simulation is used to verify the
stability of protein and compound binding affinity.
2. Results
2.1. Finding Active Compounds and Common Target Proteins
We collected 143 compounds from pomegranates through various databases
and literature reviews. The final compounds obtained were 75 after
screening from oral bioavailability (OB) and drug likeness (DL),
presented in [49]Table S1. This is crucial for early drug design
because it provides information about the effectiveness of the drug in
various body systems and physicochemical properties, whether it reaches
the target site or not [[50]18]. Furthermore, toxicity was analyzed,
and the final compounds were 15, as shown in [51]Table 1. Swiss target
prediction and STICH databases resulted in 698 target genes of
compounds. Similarly, from the OMIM and gene cards databases, we gained
233 and 13,780 target genes related to pancreatic cancer, respectively.
Then, duplicate genes in OMIM, gene cards, and compound targets were
removed. Venn analysis of compound targets and PC targets was then
carried out.
Table 1.
List of top 15 compounds and their relative information.
Compounds CID No. Molecular Formula Molecular Weight (gm/mol) Drug
Likeness (DL) ≥ 0.18 Oral Bioavailabilty (OB) ≥ 0.3
Catechin 9064 C[15]H[14]O[6] 290.27 0.64 0.55
Epicatechin 72276 C[15]H[14]O[6] 290.27 0.64 0.55
Naringenin 439246 C[15]H[12]O[5] 272.25 0.82 0.55
Phloridzin 6072 C[21]H[24]O[10] 436.41 0.66 0.55
Genistein 5280961 C[15]H[10]O[5] 270.24 0.44 0.55
Gamma-Tocopherol 92729 C[28]H[48]O[2] 416.68 0.48 0.55
Daidzein 5281708 C[15]H[10]O[4] 254.24 0.29 0.55
Quinic acid 6508 C[7]H[12]O[6] 192.17 0.19 0.56
1-O-Galloyl-beta-D-glucose 124021 C[13]H[16]O[10] 332.26 0.81 0.55
Palmitelaidic acid 5282745 C[16]H[30]O[2] 254.41 0.87 0.56
Epicatechin gallate 107905 C[22]H[18]O[10] 442.37 0.93 0.55
Alpha-Zearalanol 2999413 C[18]H[26]O[5] 322.4 0.5 0.55
Beta-Zearalanol 65434 C[18]H[26]O[5] 322.4 0.5 0.55
Astragalin 5282102 C[21]H[22]O[11] 448.4 0.67 14.03
Kaempferol 5280863 C[15]H[10]O[6] 286.24 0.5 0.55
[52]Open in a new tab
As a result, [53]Figure 1a illustrates the interactions of predicted
protein targets (blue nodes) and the PG compounds identified as shades
of red to orange, indicating that each edge represents an interaction,
which suggests that the compound may have multiple biological targets
that it may act upon. Also, the different degrees of connectivity
between the central hubs are differentiated (yellow to dark red):
darker nodes represents greater interaction with target proteins.
Therefore, PG compounds exhibit multi-target properties, which are
important characteristics of complex disease treatment approaches the
top ten compound targets outlined in [54]Figure 1a. Similarly, Venn
analysis was developed to compare the overlap between disease-related
genes and PG target genes. Out of the 13,519 disease-associated genes,
297 were found to have cross-relationships with the 332 PG target
genes, which indicates a significant link between PG compounds and
disease-relevant genes. Due to their role in disease processes, the
intersecting 297 genes share a great deal of therapeutic potential.
These overlapping genes could be used to identify potentially drug
candidate and disease-associated targets, which could guide future
studies for the validation as depicted in [55]Figure 1b.
Figure 1.
[56]Figure 1
[57]Open in a new tab
Pomegranate–Pancreatic Cancer. (a) Top 10 compound targets highlighted
in yellow, red, orange and target proteins are highlighted in blue
colors; (b) Intersection of target genes between disease and
pomegranate.
2.2. Investigation of Protein–Protein Network Interactions
The 297 common genes from the Venn analysis were imported into STRING
web server. These common genes were from pomegranate compound targets
and pancreatic cancer targets. Our data demonstrated 297 nodes and 4415
interaction edges as shown in [58]Figure 2a. The network was linked
with Cytoscape software 3.10.2 for topological analysis and to obtain
the top key genes responsible for pancreatic cancer treatment. The
significance of each node in protein–protein interaction was
characterized using parameters of degree, closeness centrality (CC),
and betweenness centrality (BC) and degree of connection presented in
[59]Table 2. The top 30 and core target proteins shorted from 30
targets are demonstrated in [60]Figure 2b and [61]Figure 2c,
respectively.
Figure 2.
[62]Figure 2
[63]Open in a new tab
Protein–protein interaction network. (a) Interaction nodes and edges;
(b) Top 30 gene interactions obtained from (a); (c) Top 10 genes
responsible for PC management.
Table 2.
Top 10 target proteins by gene degree.
Target Gene UniProt ID Protein Name Degree BC CC
AKT1 [64]P31749 RAC-alpha serine/threonine-protein kinase 158 0.067
0.678
IL6 [65]P05231 Interleukin-6 157 0.066 0.678
TNF [66]P01375 Tumor Necrosis Factor 152 0.058 0.669
SRC [67]P12931 Proto-oncogene tyrosine-protein kinase Src 134 0.081
0.633
STAT3 [68]P40763 Signal transducer and activator of transcription 3 130
0.032 0.633
EGFR [69]P00533 Epidermal growth factor receptor 129 0.037 0.630
BCL2 [70]P10415 Apoptosis Regulator Bcl2 121 0.020 0.617
HSP90AA1 [71]P07900 Heat shock protein HSP90-alpha 119 0.030 0.612
HIF1A [72]Q16665 Hypoxia inducible factor 1-alpha 117 0.026 0.608
PTGS2 [73]P35354 Prostaglandin G/H synthase 2 114 0.040 0.612
[74]Open in a new tab
BC = betweenness centrality, CC = closeness centrality.
2.3. Go and KEGG Terms Analysis
Go and KEGG analyses aim to collect biological information and
important signaling pathways that control biological processes. Go
enrichment analysis is illustrated in [75]Figure 3a. In biological
process (BP), most genes are involved in decreasing of apoptotic
process and signal transduction. Assembled genes were present in
cytosol, plasma membrane, cytoplasm, and nucleus. [76]Figure 3b
illustrates the higher number of genes participating in metabolic
pathways and pathways in cancer, which accounted for 40 and 28 gene
numbers, respectively. Lastly, the genes involved in molecular function
were protein binding, ATP binding, identical protein binding, protein
kinase activity, and zinc ion binding. Furthermore, the top 30
signaling pathways shown in [77]Figure 3c participate in pancreatic
cancer and the P13/Akt signaling pathaway. The relations diagram
between the top 20 pathways and the number of significant gene counts
are mostly related to metabolic pathways, pathway in cancer, and
P13/Akt pathways, as shown in [78]Figure 3d. The key pathways and genes
involved in pancreatic cancer are depicted in [79]Figure 4, which
offers information about crucial target proteins for treating
pancreatic cancer.
Figure 3.
[80]Figure 3
[81]Open in a new tab
GO and KEGG analysis. (a) Top 10 Go enrichment terms; (b) The histogram
diagram of genes in each pathway; (c) Top 30 KEGG pathways involved in
PC; (d) Sanky and Bubble diagram of KEGG pathways for hub targets.
Figure 4.
[82]Figure 4
[83]Open in a new tab
Pancreatic cancer pathways and key proteins highlighted in red color.
2.4. Docking Score of Core Targets with Compounds
Molecular docking of selected pomegranate compounds was performed with
the help of Schrodinger platform. It provides information about binding
affinity and specific interactions between proteins and compounds.
Among 15 compounds of pomegranate peels, only 11 compounds showed
docking score of less than (<−6 kcal/mol). We took a score of less than
or equal to −6 kcal/mol as initial filtering criteria to prioritize a
manageable number of ligands for further analysis [[84]19]. Moreover, a
docking score of less than −8 kcal/mol was obtained in catechin,
genistein, epicatechin gallate, and astragalin complex with HSP90AA1,
EGFR, and protein PTGS2. However, other compounds did not show any good
binding affinity scores. The docking scores of all compounds are shown
in [85]Table 3. However, proteins such as IL6, TNF, SRC, STAT3, and
HIF1A could not bind with any selected compounds of pomegranate peels,
and compounds that showed promising interactions with proteins are
mentioned in [86]Figure 5. In [87]Figure 5a, the ligand forms multiple
H-bonds with residues TRP387, TYR385, and PHE210, while the surrounding
residues provide hydrophobic support and a polar environment that may
enhance specificity. These types of interactions are a hint of deep and
specific binding within the binding pocket which is likely an active
region of the receptor. Also, as depicted in [88]Figure 5b, the ligand
forms four key hydrogen bonds with negatively charged residues (ASP93,
ASP102), polar residues ASN51, and water-bridge with residues GLY108,
reinforcing strong and stable binding to the ligand. The surrounding
hydrophobic bond highlighted in green and polar residues (light blue)
help in maintaining an optimal binding environment, potentially
contributing to selectivity and binding stability. The ligand is
connected to polar and charged residues like ASN842, ARG841, and LYS745
by multiple hydrogen bonds, as shown in [89]Figure 5c. Hydroxyl groups
are primarily responsible for these interactions. The ligand aliphatic
side chains are surrounded by a potent hydrophobic environment, which
improves nonpolar stabilization. The diverse set of interactions
suggests a highly complementary binding pocket, enhancing both polar
and nonpolar contacts for high binding affinity. Moreover, in
[90]Figure 5d, the ligand is docked with several hydrogen bonds,
especially with charged and polar residues (ASP800, ASN842, LYS745, and
SER720). Hydrophobic residues provide additional stabilization via
nonpolar interactions. The ligand is well positioned within a pocket
formed by a mix of polar and hydrophobic environments, resulting in
enhanced binding affinity and specificity. Finally, [91]Figure 5e shows
that the ligand binds to residues ALA230, GLU228, ASN279, GLU234, and
ASP292 to form five strong hydrogen bonds. The ligand is stabilized in
the binding pocket by the hydrophobic residues TYR229 and MET227. The
binding pocket offers a well-balanced mix of polar and nonpolar
environments which enhances the binding affinity, specificity, and also
overall stability of protein–ligand complexes which is crucial for
molecular identification and therapeutic targeting. Among
aforementioned interactions, the AKT1-1-O-galloyl-beta-D-glucose
complex displays extensive interaction compared to other interactions.
Table 3.
Docking score (kcal/mol) of all compounds.
Compounds AKT1 EGFR BCL2 HSP90AA1 PTGS2
Catechin −6.66 −6.45 −5.61 −8.18 −7.28
Epicatechin −6.53 −6.63 −5.82 −7.93 −7.06
Naringenin −6.33 −6.91 −6.00 −7.85 −7.44
Phloridzin −6.92 −6.44 −6.02 −7.08 −6.73
Genistein −6.25 −8.10 −5.40 −5.61 −7.34
Daidzein −5.73 −7.56 −5.11 −5.84 −7.34
Quinic acid −4.92 −6.54 −5.04 −5.77 −6.47
1-O-Galloyl-beta-D-glucose −6.37 −7.11 −4.92 −7.65 −6.35
Epicatechin gallate −5.63 −6.39 −6.32 −7.92 −7.92
Astragalin −4.42 −5.81 −6.09 −7.92 −9.00
Kaempferol −6.25 −6.99 −6.67 −7.11 −6.96
[92]Open in a new tab
Figure 5.
[93]Figure 5
[94]Open in a new tab
Molecular interactions between protein and ligands. (a)
1-O-Galloyl-beta-D-glucose compound interaction with PTGS2 protein (PDB
ID: 5F19); (b) Epicatechin compound interaction with HSP90AA1 protein
(PDB ID: 6TN5); (c) Phloridzin compounds interaction with EGFR protein
(PDB ID: 8A27); (d) Epicatechin gallate interaction with EGFR protein
(PDB ID:8A27); (e) 1-O-Galloyl-beta-D-glucose compound interaction with
AKT1 protein (PDB ID:4GV1).
2.5. Molecular Dynamic Simulation
Molecular dynamics simulation confirms the stability between compounds
and proteins by evaluating RMSD, protein–ligand contact, hydrogen
bonding, RMSF, and radius of gyration. We performed molecular dynamics
simulations of all top-ranked docking scores less than or equal to
−6.00 kcal/mol as a threshold; they are given in [95]Table 3. The lower
the RMSD value, the higher the stability between complexes. Among many,
some protein–ligand complexes reveal promising outcomes, which is
explained in a section below. Firstly, in [96]Figure 6a, the
PTGS2-1-O-Galloyl-beta-D-glucose complex exhibited initial fluctuations
between 1.5–2.3 Å over 15ns. Afterwards, protein RMSD graph stabilized
gradually and levitated between 2.0 and 2.7 Å until the end of the
simulation. So, an RMSD of 2.7 Å means the protein, especially the
backbone of the protein, has shifted to 2.7 Å from its initial position
throughout the time of simulation, which means the binding ligand did
not undergo extreme conformation changes [[97]20,[98]21]. The value
obtained from RMSD tells us about the nature and stability of the
protein–ligand complex. In the initial time, when the ligand came in
contact with the protein, it underwent some adjustment and finally
bound to the pocket of protein either strongly or weakly. Therefore, it
suggests that there were minor fluctuations but an overall stable trend
after 30 ns. However, the ligand RMSD started with high fluctuations,
dropped as low to 0.6 Å, and spiked over 5 Å before 40 ns. After that,
the ligand stabilized within the 4.2–4.8 Å range, illustrating moderate
mobility with in the pocket with some conformational shifts. Secondly,
the complex in [99]Figure 6b was initially low around 1.6 Å, and it
rose and stabilized near 2 Å at around 20 ns. There was minor
fluctuation observed but the graph maintained equilibrium after 30 ns
suggesting the stability of protein throughout the simulation. However,
looking at the ligand graph, there was a sharp rise from 0.6 Å to 3.6 Å
in the first 20 ns. Afterwards, the graph maintained stability around
3.6–4.2 Å. Similarly, the complex of EGFR- phloridziz fluctuated
between 1.2 and 2.4 Å. The protein was observed to be fluctuated
throughout time. However, the fluctuation was minor with stable RMSD
average range from 1.6 to 2.2 Å. The ligand maintained a stable RMSD
graph, within 2.0–3.6 Å most of the time, as shown in [100]Figure 6c.
Moreover, in [101]Figure 6d, the RMSD graph reached its highest
fluctuation up to 2.4 Å. However, the graph initially elevated from
approximately 1.2 Å to 2.2 Å. Afterwards, the protein reached its
stability around 2.0–2.3 Å after 25 ns with minor fluctuations.
Therefore, the protein structure reached equilibrium and remained
steady thereafter. Looking at the ligand graph, from 40 ns, the graph
settled in the 2.0–3.0 Å range. The interaction fraction exceeded 1.6.
Lastly, in the complex in [102]Figure 6e, the RMSD graph displays a
steady increase from 1.8 Å to 2.6 Å by 40 ns. After 40 ns, the graph
stabilized between 2.4 and 2.8 Å with minor fluctuation which is an
acceptable range. Moreover, the ligand RMSD graph remained stable
throughout between 1.0 and 1.4 Å, suggesting strong binding affinity.
These indicate that pomegranate compounds form strong, stable
interactions with target pancreatic cancer proteins. The results from
post-molecular-dynamics simulations showed that reorientation of the
ligand occurred during the simulations, forming new interactions like
water-mediated hydrogen bonds and additional residual contacts compared
to original docking poses. This might be due to the movement and
adjustments of the ligand in protein residues in physiological
simulation conditions ([103]Figure S7).
Figure 6.
[104]Figure 6
[105]Open in a new tab
RMSD of all compounds. (a) 1-O-Galloyl-beta-D-glucose with PTGS2
protein; (b) Epicatechin with HSP90AA1 protein; (c) Phloridzin with
EGFR protein; (d) Epicatechin gallate with EGFR protein; (e)
1-O-Galloyl-beta-D-glucose with AKT1 protein.
The second criterion for stability observation is an RMSF graph, which
offers information of individual amino acid flexibility throughout
simulation. Our data revealed that both stable and flexible segments in
are present in the structure. The majority of fluctuations occurred in
the terminal regions, where the RMSF values were frequently in the
range of 3.5 to 4.8, indicating extremely high mobility in these
regions. On the other hand, the structural parts of the proteins,
indicated by the green bars, displayed lower RMSF values (around 0.5 to
1.5 Å), reflecting their structural stability. As seen in [106]Figure
7e, the core was more stable, especially between residues 50 and 200,
whereas the values in [107]Figure 7a,d are more flexible. On the other
hand, in [108]Figure 7c, a change in motion appears around residue 150
as opposed to [109]Figure 7b, where the data are relatively rigid
except at the ends. Therefore, the overall structure of the protein
remains intact, with the majority of flexibility concentrated in the
terminal and loop regions of the protein. [110]Figure 8 illustrates the
different bonding types between ligand and receptor along with the
frequency of interactions during simulation. Initially, the binding in
[111]Figure 8a was facilitated by THR 206, Thr 210, and ASN 382 that
formed hydrogen bond frequently and maintained interaction fractions
above 1.0, indicated in green and blue colors, respectively. Further,
GLN 203, ALA 199, PHE 213 presented mixed bonding. In addition, in
[112]Figure 8b, several residues (LYS58, ASP93, and ASP102) exceed the
1.0 interaction fraction and have multiple interactions; ASP 93 mostly
made a hydrogen bond and some hydrophobic angle. Other significant
contributers, such as ASN 51, LYS58, GLY 97, and ASPO 102, have an
interaction fraction around 0.6. In addition, residue ASP855, which has
an interaction fraction over 2.0, was highly responsible for stability
maintenance between EGFR and phloridzin as shown in [113]Figure 8c.
Other residues involved in binding include MET793, ARG 841, and ASN
842. Residues like MET 793 and ASP 100 reach interaction fraction above
1 in [114]Figure 8d. They are the most prominanat binders, forming
hydrogen along with hydrobhobic contacts. Meanwhile, diverse
interaction types including ionic and hydrophobic with multiple
residues and frequency of interaction above 1.5 are illustrated in
[115]Figure 8e. GLU 278,GLU 234, and ASP 292 have higher interaction
fractions and involve ionic bridges. Further, THR 291 retained
stability by hydrogen bond generation. So, all these key residues play
a crucial role in binding 1-O-Galloyl-beta-D-glucose with AKT1.
Figure 7.
[116]Figure 7
[117]Open in a new tab
RMSF (a–e) illustrates the protein–ligand interaction, highlighting
(green lines) the key residues that are involved in the stabilizing
ligand within the binding pocket of the protein. (a)
1-O-Galloyl-beta-D-glucose with PTGS2 protein; (b) Epicatechin with
HSP90AA1 protein; (c) Phloridzin with EGFR protein; (d) Epicatechin
gallate with EGFR protein; (e) 1-O-Galloyl-beta-D-glucose with AKT1
protein.
Figure 8.
[118]Figure 8
[119]Open in a new tab
Protein–ligand contact: Green (H-bonds); Grey (Hydrophobic); Pink
(Ionic); Blue (Water bridge); (a) 1-O-Galloyl-beta-D-glucose with PTGS2
protein; (b) Epicatechin with HSP90AA1 protein; (c) Phloridzin with
EGFR protein; (d) Epicatechin gallate with EGFR protein; (e)
1-O-Galloyl-beta-D-glucose with AKT1 protein.
Additionally, protein structure analysis data reflect the stability and
behavior of protein during simulation. Notably, graphs (b–d) of
[120]Figure S1 have higher proportion of secondary structure elements
(SSE) around 47–50%, signifying the maintenance of the native secondary
structure, whereas graphs (a,d), [121]Figure S1, have SSE significantly
below 41%, suggesting flexibility and partial unfolding. Moreover, the
distributions of helix and strand are uniform and stable in the
simulation. In a similar way, the radius of gyration (Rg) illustrates a
minor orientation shift but no major folding or collapse and structural
stability; see [122]Figure S1, right side, green-color graph.
2.6. Ligand-Free Protein RMSF
From root square fluctuation (RMSF) of apo (unbound) protein data, the
apo protein was more flexible than the ligand-bound protein especially
in loop regions and the N terminal. For instance, ligands like
1-O-Galloyl-beta-D-glucose and epicatechin reduced the variation in
binding residues, meaning it improved the structural stability of
binding proteins. Due to ligand interactions, some regions showed
increased flexibility, while others displayed more rigidity, as
represented in [123]Figure S6. This suggests that interacting ligands
cause dual effects in proteins. These results demonstrate that ligand
binding not only stabilizes the specific binding sites but also induces
long-range dynamic changes across the protein structure, highlighting
the allosteric nature of interactions.
2.7. In Silico ADME, Toxicity, and Lipinski and Pfizer’s Rule
ADME analysis offers valuable information on compounds in the early
stage of drug development. [124]Table 4 demonstrates the
pharmacokinetic characteristics of compounds. Our result illustrates
that all compounds are water soluble, ranging from 0.045 to −7.641.
Additionally, the volume distribution of drugs was found to be between
0.04 and 20 L/kg, indicating that our compounds have greater
distribution in tissue rather than plasma. Interestingly, toxicity
results show that all compounds are free from hepatotoxicity,
immunotoxicity, mutagenicity, and cytotoxicity. Lastly, we observed
Lipinski’s and Pfizer’s rules to ensure whether our molecules hold
drug-like properties concerning physicochemical properties. [125]Table
5 shows Lipinski and Pfizer’s rules and the toxicity of the selected
compounds.
Table 4.
In silico ADME analysis of all compounds.
Compounds Absorption Distribution Metabolism (CYP Inhibitor) Excretion
Log S
CaCO[2] P-gp HIA BBB VD[ss] 1A2 2C19 2C9 2D6 3A4 CL T1/2
Catechin −6.04 – — — 1.15 — — — — — 14.9 2.14 −2.28
Epicatechin −6.04 – — — 1.15 — — — — — 14.9 2.14 −2.28
Phloridzin −6.20 - — — 0.68 — — — — ++ 3.76 2.26 −2.53
Daidzein −4.69 - — — 0.62 +++ +++ +++ +++ ++ 7.85 1.17 −3.79
Quinic acid −6.31 — + — 0.37 — — — — — 1.57 3.35 −0.10
1-O-Galloyl-beta-D-glucose −6.39 – + — 0.37 — — — — — 3.68 2.37 −1.29
Epicatechin gallate −6.51 — — — 0.44 — — ++ — +++ 9.65 2.08 −3.70
Kaempferol −5.97 – — — 0.15 +++ – ++ — +++ 5.69 1.33 −3.65
[126]Open in a new tab
0–0.1 (—), 0.1–0.3 (–), 0.3–0.5 (-), 0.5–0.7 (+), 0.7–0 (++), 0.9–1.0
(+++). HIA-Human Intestinal Absorption, BBB-Blood Brain Barrier,
VD[ss]-Volume distribution, CL-Clearance, Log S- Water solubility.
Table 5.
Toxicity and Lipinski’s rule analysis of pomegranate compounds.
Compounds Carcinogenicity Immunotoxicity Mutagenicity Cytotoxicity
Lipinski Pfizer PAINS
Catechin Inactive Inactive Inactive Inactive Yes Yes 1
Epicatechin Inactive Inactive Inactive Inactive Yes Yes 1
Phloridzin Inactive Inactive Inactive Inactive Yes Yes 0
Daidzein Inactive Inactive Inactive Inactive Yes Yes 0
Quinic acid Inactive Inactive Inactive Inactive Yes Yes 0
1-O-Galloyl-beta-D-glucose Inactive Inactive Inactive Inactive Yes Yes
1
Epicatechin gallate Inactive Inactive Inactive Inactive Yes Yes 1
Kaempferol Inactive Inactive Inactive Inactive Yes Yes 0
[127]Open in a new tab
3. Discussion
Pomegranate is recognized as a fruit with excellent health benefits due
to its bioactive properties. There is no doubt that pomegranate fruit
has many advantages, but the peel also provides a wide variety of
benefits, just like the fruit [[128]22]. Thus, this fascinating
attribute could contribute to the treatment of certain diseases like
pancreatic cancer (PC), a highly aggressive malignancy with limited
therapeutic options [[129]13,[130]23]. The treatment of PC remains
challenging due to extreme aggressiveness, poor prognosis, and lack of
early-stage symptoms. A network-based approach was performed to
identify bioactive compounds from pomegranate peels that could serve as
potential therapeutic agents for PC. Before choosing the final PG
compounds, we screened the compounds DL and BO. This is an important
step in determining the properties of our compounds of interest with
positive and negative sets of drugs and whether our products will reach
the site of drug action [[131]24]. Our analysis identified 297 common
targets shared between pomegranate-derived compounds and PC-related
proteins ([132]Figure 1b). This common target is useful for finding the
best target proteins for further analysis. Protein–protein interaction
data revealed the top 10 key target proteins ([133]Table 2), indicating
these targets might be responsible for pancreatic cancer. However, the
docking reports demonstrated that five proteins (AKT1, EGFR, BCL2,
HSP90AA1, PTGS2) exhibited strong binding with pomegranate compounds
with good scores. Previous studies have explored the effects of
catechin and its derivatives on disease-relevant targets including BCL2
and HIF1A as a key target. This research mentioned molecular docking of
catechins with receptors similar to those we collected [[134]25].
Similarly, Rehan et al. [[135]26] investigated cancer therapeutic
properties of catechin and epicatechin with epidermaal growth factor
receptor (EGFR) involved in cancer progression. This study sketched
that pomegranate peel compounds like 1-O-galloyl-beta-D-glucose,
Phloridzin, epicatechin gallate, and epicatechin show promising
interaction with AKT1, EGFR, HSP90AA2, and PTGS2, as confirmed by RMSD,
RMSF, and interaction fractions ([136]Figure 6, [137]Figure 7 and
[138]Figure 8). For other compounds like quinic acid and kaempferol,
interactions are given in [139]Figure S2. Most compounds possess
binding energy ≤−6.0 kcal/mol [[140]27] while the reference drugs
tanespimycin, erlotinib, afuresertib, and celecoxib docking score
ranges from −5.48 to 8.07 kcal/mol with proteins HSP90AA1, EGFR, AKT1,
and PTGS2, respectively, as shown in [141]Table S2. The RMSD graph of
the selected complex indicates less fluctuation or movements of
proteins and ligands during contact, and deviation is in acceptable
range from 2.25 to 3.2 Å compared to the reference drug; [142]Figure
S3. When the protein structure is dramatically changed, the protein
loses its native structure (secondary and tertiary structure),
indicating possible denaturation and incomplete equilibration. Loss of
stability can also be confirmed by higher fluctuation and flexibility
in protein structure, losing stable hydrogen bond, which is not good
for complex stability maintenance. So, our protein conformational
changes are not drastic, and stability is maintained. Comparing the
RMSD value with that of reference drugs, the range is between 2.0 to
3.5 Å, occasionally reaching 4 Å, which reflects noticeable
confrontational shifts. In contrast, our results displayed lower and
more consistent values, signifying enhance structural stability.
Further, the RMSF result of reference drugs indicated that some binding
residues have higher fluctuation compared to test PG compounds, and
protein–ligand contact revealed that while both the control and test
compounds displayed various interactions, the pomegranate ligands
demonstrated higher interaction fraction, often reaching a value of
one. These data suggest more stable anchoring in binding residues. A
significant number of hydrogen bonds, hydrophobic bonds, and water
bridges can be observed in the interaction diagram ([143]Figure 5 and
[144]Figure 8). Interestingly, epicatechin, 1-O-galloyl-beta-D-glucose,
epicatechin gallate, and pholoridzin can bind with this carcinogenic
protein, which might play a pivotal role in PC suppression ([145]Figure
6 and [146]Figure S1). The EGFR-epicatechin gallate and
1-O-galloyl-beta-D-glucose-AKT1 complex have widespread interaction
with cancer-associated targets, while others have fewer interactions.
This bonding is crucial for effective stability and good affinity in
drug discovery.
Additionally, our GO and KEGG analyses highlight several pathways and
the involvement of genes in various functions. The high frequency of
genes in the biological process mainly participate in apoptotic
regulation process and cellular signaling along with significant
molecular functions like binding and enzymatic activity. The major
process primarily takes place in the cytosol. The highest number of
genes were associated with metabolic and cancer-related pathways like
P13/Akt, which is known for cell growth regulation, apoptosis,
survival, metabolism, and proliferation [[147]28,[148]29,[149]30]. The
top enriched pathways include AGE-RAGE, epidermal growth factor (EGFR),
and VEGF, which are highly expressed in pancreatic cancer. The outcomes
from our analysis indicate their importance in tumor growth and drug
resistance because these pathways are linked to cancer progression and
angiogenesis ([150]Figure 3). So, targeting these pathways ultimately
initiates apoptosis and reduces proliferation, and this can be achieved
using pomegranate compounds, which could be a good therapeutic
candidate in cancer treatment. Oncogenic genes like EGFR and AKT1 are
responsible for metastasis, cell proliferation, and blood vessel
formation and are highly expressed in pancreatic cells, leading to poor
prognosis [[151]31]. The triggering of EGFR causes the activation of
myofibroblasts, leads to the secretion of various factors, and causes
the progression of cancer cells [[152]32]. Similarly, HSP90AA1 is known
to enhance tumor aggressiveness, while PTGS2 (encoding COX-2) promotes
apoptosis resistance, proliferation, inflammation, and metastasis of
cancer cells [[153]33]. Additionally, some investigations into the role
of flavonoids in modulating cancer signaling cascade also exemplifies
that this natural compounds plays a significant role in disrupting the
cancer signaling cascades, causes inflammation, and is involved in the
apoptosis process to destroy the cancer cells [[154]34,[155]35]. This
can be achieved by suppressing vital pathways like P13/Ak and EFGR,
whose main role is enhancement of cancer development. Further, research
findings suggest that pomegranate compounds can inhibit or suppress
this carcinogenic protein and halt cancer cell progression.
Finally, to assess the drug-like character of compounds, ADMET
(Adsorption, Distribution, Metabolism, Excretion, and Toxicity)
analysis was conducted. Remarkably, most of the compounds of
pomegranate are free from toxicity. This property make compounds more
valuable for the drug development process. Besides this, evaluation of
compounds via Lipinski’s and Pfizer’s rules was achieved; 15 compounds
met these criteria [156]Table 5. However, epicatechin gallate and
kaempferol displayed moderate solubility but high tissue distribution
(−4–0.5 log mol/L) [[157]36]. The rest of the compounds had good water
solubility. For optimal absorption, the drug should be dissolved
sufficiently in the blood circulation system; otherwise, it affects the
drug discovery process [[158]37]. All compounds exhibited no BBB
permeability and higher human intestinal absorption ≥ 30% (HIA),
indicating higher bioavailability [[159]38]. Certain selected compounds
have a high clearance rate, which means they can be eliminated from the
body quickly. Conversely, 1-o-galloyl-beta showed limited clearance,
indicating prolonged retention within the body. As a result, a lower
dosage is required to achieve therapeutic efficacy. Most compounds
displayed poor intestinal permeability (low CaCO[2]) value.
Additionally, our promising compounds, epicatechin, phlorizin, and
1-O-galloyl-beta-D-glucose, acted as anti-inhibitors of multiple CYP
enzymes, except epicatechin gallate and kaempferol. All these
satisfactory data from our research indicate that each compound of
pomegranate peels carries highly valuable drug-like properties,
signifying strength in the drug development process as well as for
treatments of various diseases, including highly aggressive pancreatic
cancer, by inducing apoptosis or autophagy and participating in
cancer-related pathways. Additionally, natural-based drugs mitigate the
health side effects compared to synthetic drugs that are available on
the market. Moreover, people treat pomegranate peels as waste;
consequently, this research minimizes the negative impact on the
environment and concurrently adds the valorization of pomegranate
peels.
Despite the promising computational outcomes, it is critical to
consider the limitation of computational methodologies. Furthermore,
the verification of computational outcomes can be achieved through in
vivo and in vitro studies. Thus, based on our data, pomegranate
compounds may interact with several pathways to play a significant role
in PC inhibition. Because of the promising benefits of PG in various
cancer inhibitions, the above compounds may be suitable for PC.
4. Materials and Methods
4.1. Screening of Bioactive Compounds
Bioactive compounds from pomegranate were extracted from different
databases like Traditional Chinese Medicine Systems Pharmacology
(TCMSP) [160]https://www.tcmsp-e.com/tcmsp.php (accessed on 12
September 2024), Indian Medicinal Plants, Phytochemistry, Therapeutics
[161]https://cb.imsc.res.in/imppat/ (accessed on 12 September 2024),
and literature review. The data obtained were filtered using oral
bioavailability score (OB) ≥ 0.3 and drug likeness (DL) ≥ 0.18
thresholds [[162]18]. Furthermore, using PubChem-ID, we recorded each
compound’s detailed information.
4.2. Identification of Pancreatic Cancer and the Compound’s Target Genes
The compound target protein was collected by inserting canonical SMILE
into Swiss Target Prediction [163]http://www.swissadme.ch (accessed on
19 September 2024) and STITCH data [164]http://stitch.embl.de/
(accessed on 25 September 2024). These two databases present each
compound’s target proteins. Furthermore, the compound target
probability was filtered with a value of 0.1. Similarly, disease target
proteins were gathered from two online databases, Online Mendelian
Inheritance in Man (OMIM) and Gene Cards. These databases have a
collection of all genes related to disesaes. They were used to identify
genes associated with pancreatic cancer. All the related target excel
files were downloaded for further analysis.
4.3. Exploration of Overlapping Genes
Firstly, data retrieved from OMIM, gene cards, and pomegranate targets
were combined and inserted into the Venn analysis website,
[165]https://bioinformatics.psb.ugent.be/webtools/Venn/ (accessed on 2
October 2024). The website provides information on common genes between
OMIM, gene cards, and compound target genes.
4.4. Hub Gene Analysis Through Protein–Protein Network
The overlapping genes identified from the Venn analysis were analyzed
using STRING 12.0 to see every interaction of each gene with high
confidence scores. STRING is a biological database that is used to
image protein–protein interactions and to obtain key proteins with
biological pathways. To achieve this, selection parameters were set as
multiple proteins and organism homo sapiens, and interactions were
visualized through Cytoscape 3.10.2 which delivers the topological
properties including nodes and relationship within the networks such as
degree and betweenness closeness. The plugin Cytohubba identified the
top 10 genes, and Centiscape calculated their network parameters.
4.5. Bioinformatics Technique for GO and KEGG Analysis
Observation of biological phenomenon on molecular-level function can be
accomplished using Gene Ontology (GO) that delivers all the related
reports on the role of protein. This was achieved by applying the DAVID
database [166]https://david.ncifcrf.gov/tools.jsp (accessed on 10
October 2024), which is the bioinformatics tool that interprets
biological and functional annotation of the protein list. In the same
way, important pathways involved during pancreatic cancer treatments
were also found through DAVID, and results were looked at in SRplot.
Screening condition p < 0.05 was used to obtain the main pathways and
roles of the collected genes.
4.6. Molecular Docking of Hub Genes
Molecular docking helps to anticipate the most probable binding pose of
a ligand with in the associated target receptor, and this can be
achieved using Schrodinger (Schrödinger, 2024). The 3D structures of
each compound and related proteins were downloaded from PubChem and
RSCB-PDB databases, respectively. Regarding protein selection, we
gathered high-resolution crystallographic structure without mutation.
The higher resolution provided the co-ordinates of each atom in detail
and aided in generating trustworthy binding affinity. The PDB IDs used
were 4GV1 (AKT1), 1ALU (IL6), 2AZ5 (TNF), 1043 (SRC), 6NJS (STAT3),
8A27 (EGFR), 4MAN (BCL2), 6TN5 (HSP9OOA1), 7LVS (HIF1A), and 5F19
(PTGS2). Firstly, in PyMOL and PUResNet, a ligand binding pocket was
chosen. The binding residues were selected by creating a 4 Å distance
from a co-crystallized ligand in PyMol [[167]39].
4.6.1. Ligand Preparation
Using the LigPrep module, the imported structures were refined
geometrically into Maestro, creating a single low energy. Consequently,
the primary states of ionization and chiralities were determined at
this stage. The maximum atom size and possible ionization state were
set to 500 and Epik at pH 7.0 ± 2, respectively. Tautomer and
steroisomer generation was allowed to generate different forms of
ligand by position and different 3D arrangement of atoms. The
confirmation of ligand structure was minimized using the OPLS4 force
field. This provides improved parameterization for both proteins and
small molecules, enhancing protein–ligand interactions and
conformational flexibility for charged and polar groups. It also
alleviates steric hindrance and achieves an ideal geometry by adjusting
bond lengths, angles, and torsions, thereby indicating a more accurate
receptor conformation. Also, this helps in energy minimization before
the commencement of docking simulations [[168]40,[169]41].
4.6.2. Protein Preparation and Docking Setup
At last, the imported protein was prepared in Maestro using the default
option provided by the protein preparation wizard in Schrodinger, which
includes the addition of missing hydrogen atoms, missing chain filling,
assignment of bond order, protonation state adjustment using PROPKA at
pH7.4, and Epik to predict protonation states and tautamers of
molecules. The protein energy was suppressed with an OLSP4 force field.
Furthermore, the grid dimension was generated at the centroid of
selected residues. And inner and an outer box were set to 10 Å and 20 Å
from the ligand center. This allowed the entire ligand to fit inside
the grid [[170]42]. This was achieved using the receptor grid
generation module. Following grid formation, the standard precision
mode (SP) docking was used with van der Waals scaling factor 0.80 and
partial charge cutoff 0.15 for non-polar atoms. Moreover, nitrogen
inversion, ring flexibility, torsion bias for functional groups, and
the Epik tool were applied during docking. Numerous binding poses for
individual ligands were maintained by setting a 30 kcal/mol window
energy. Lastly, 10 poses per ligand were created to collect top glide
score.
4.7. Validation of Molecular Docking
Using the Pymol visualization tool, the co-crystallized ligand attached
in the active sites of proteins was separated and redocked into the
same active sites of that protein. This was performed to validate the
docking score and for reliability of the results.
4.8. Assessment of Protein–Ligand Complex Stability
Molecular dynamics (MD) simulation was performed using Desmond through
the Schrodinger Maestro interface. Simulation offers information on how
much the ligands and proteins are bound with each other and their
stability, along with various inter-molecular interactions. Firstly,
the top ranked docking score was taken for MD simulation, and during
this, solvent MD package, a fixed OPLS4 force field, and simulation
period at 100 ns were set. The protein–ligand complex was maintained
with an orthorhombic periodic box. The buffer distance was set to 10 ×
10 × 10 Å, and water was added to molecules using the TIP3P model. The
sufficient charges like 0.15 M NaCl, Na^+, and Cl^− were added to
balance the complex systems. A constant number of particles as well as
the pressure and temperature (NPT) ensemble method was used at 300.0 K
temperature and 1.013 bar pressure, together with the relax model
system before simulation [[171]43,[172]44]. The relax model was
incorporated with a thermostat (Nose-hoover Chain with relaxation time
1.0 ps) and a barostat method (Martyna–Tobias Klein with relaxation
time 2.0 ps). In addition, we applied an RESPA multiple time step
integrator with a 2 fs time for bonded, 2 fs for near, and 6 fs for far
steps, and a 9.0 Å cutoff was used as a short-range integrator. At
last, a trajectory interval of 4.8 ns, 10 ps of energy, and a solvent
box of a 10 Å size were set up [[173]45]. Afterward, a simulation was
run usingthe Desmond module of Schrodinger, and the stability of
complexes was assessed using graphical parameters obtained from
simulation results.
4.9. In Silico Pharmacokinetics (ADME) and Toxicity Assessment
ADMET is a crucial property required before developing therapeutic
drugs. Drugs must have a good pharmacokinetic profile to be effective
for diseases. For pharmacokinetic analysis, we employed ADMET Lab 3.0.
[[174]38,[175]46]. It screened the compound properties like absorption,
distribution, metabolism, excretion, and toxicity in the body after
consumption and also offered ideas for the early drug discovery
process. The ProTox 3.0 server screened information about the toxicity
of selected pomegranate compounds. Furthermore, we checked whether
compounds followed Lipinski’s and Pfizer rule for drug-like molecular
properties.
4.10. Data Analysis
Standard bio-informatics tools and statistical parameters were
implemented to guarantee the accuracy of results. All compounds were
filtered through bio-availability (OB ≥ 0.3) and drug likeness (DL ≥
0.18) thresholds. Target genes were screened for a probability of 0.1,
and KEGG pathway enrichment analysis was performed with significance
thresholds of p < 0.05. Binding affinities were recorded from the
Schrodinger Glide docking score; stability and flexibility of the
protein–ligand complex were analyzed through RMSD and interaction
fraction, which was run for a 100 ns simulation time. Potential drug
candidate toxicity and pharmacokinetic properties were evaluated using
ADMET Lab 3.0 and ProTox 3.0. Also, drug-like compounds were compared
with non-drug-like ones based on Lipinski’s and Pfizer’s rules.
5. Conclusions
Our study demonstrates that the bioactive compounds in pomegranate
peels can be one of the potential therapeutic agents for pancreatic
cancer. With the in silico approach, certain key compounds identified
were 1-o-galloyl-beta-d-glucose, Phloridzin, epicatechin, and
epicatechin gallate, which exhibit optimistic binding affinity with
proteins such as AKT1, EGFR, HSP9OAA1, and PTGS2. These interactions
thus indicate that bioactive compounds in pomegranate peels attack and
inhibit the P13/AKT and EGFR signaling pathway, suppressing the
formation of cancer cells. The significance of this study lies in the
dual benefits of using pomegranate peels for the treatment of cancer
with a lower risk of side effects, as well as promoting the sustainable
utilization of biological waste. This study also contributes to the
growing field of natural compound-based drug discovery and opens the
door for new treatment strategies for pancreatic cancer, a disease with
extremely poor prognosis and limited treatment options, as well as new
avenues for alternative treatment strategies. Moreover, these natural
compounds also demonstrated positive pharmacokinetic profiles, which
complied with Lipinski criteria for drug likeness, as well as
non-toxicity, for which they were evaluated as viable candidates for
future drug development pipelines. By highlighting the therapeutic
relevance of a widely available and underutilized natural resource,
this study contributes meaningfully to the ongoing search for safer,
more effective, and more accessible treatments for pancreatic cancer.
However, to confirm the anticancer effects and safety profile of this
compound, future research should focus on experimental validation of
this compound, including studies performed both in vitro and in vivo.
Further, exploring the possibility of developing these compounds into
effective delivery systems and evaluating their synergistic potential
with existing chemotherapies could help to further advance their
development as potential candidates for clinical trials.
Abbreviations
The following abbreviations are used in this manuscript:
PC Pancreatic cancer
PG Pomegranate
PP Pomegranate peel
GO Gene ontology
KEGG Kyoto Encyclopedia of genes and genomes
DL Drug likeness
BO Bio-availability
PPI Protein–protein interaction
STRING Search tool for the retrieval of interacting genes/proteins
DC Degree of connectivity
CC Closeness centrality
BP Biological processes
MF Molecular function
DAVID Database for annotation, visualization, and integrated discovery
RCSB Research collaboratory for structural bioinformatics
PDB Protein data bank
RMSD Root mean square deviation
RMSF Root mean square fluctuation
AKT1 RAC-alpha serine/threonine-protein kinase
IL6 Interleukin-6
TNF Tumor necrosis factor
STAT3 Signal transducer and activator of transcription 3
EGFR Epidermal growth factor receptor
BCL2 Apoptosis regulator Bcl2
HSP90AA1 Heat shock protein HSP90-alpha
HIF1A Hypoxia-inducible factor 1-alpha
PTGS2 Prostaglandin G/H synthase 2
[176]Open in a new tab
Supplementary Materials
The following supporting information can be downloaded at:
[177]https://www.mdpi.com/article/10.3390/ph18060896/s1, Table S1. List
of 75 compounds. Table S2. Docking score of reference drugs. Figure S1.
Protein secondary structure (right);
[MATH: α :MATH]
-helices (red);
[MATH: β :MATH]
-strands (blue) and rGyr (left) of five different complexes: (a)
1-O-Galloyl-beta-D-glucose with PTGS2 protein; (b) Epicatechin with
HSP90AA1 protein; (c) Phloridzin with EGFR protein; (d) Epicatechin
gallate with EGFR protein; (e) of 1-O-Galloyl-beta-D-glucose with AKT1
protein. Figure S2. MD simulation RMSD and Interaction fraction diagram
of pomegranate peel compounds. (a–b) Quinic acid with EGFR; (c–d)
Kaempferol with EGFR. Figure S3. Reference drug RMSD: (a) PTGS2 protein
with Celecoxib drug; (b) HSP90AA1 protein with Tanespimycin drug; (c)
EGFR protein with Erlotinib drug; (d) AKT1 protein with Afuresertib
drug. Figure S4. Reference drug RMSF: (a) PTGS2 protein with Celecoxib
drug; (b) HSP90AA1 protein with Tanespimycin drug; (c) EGFR protein
with Erlotinib drug; (d) AKT1 protein with Afuresertib drug. Figure S5.
Reference drug hydrogen bonding: (a) PTGS2 protein with Celecoxib drug;
(b) HSP90AA1 protein with Tanespimycin drug; (c) EGFR protein with
Erlotinib drug; (d) AKT1 protein with Afuresertib drug. Figure S6. Root
Mean Square Fluctuation (RMSF) of protein alone (a) PTGS2 protein; (b)
HSP90AA1 protein; (c) EGFR protein; (d) AKT1 protein. Figure S7.
Molecular interaction of ligand-proteins after molecular dynamics
simulation (MDS) for 5 ligands. (a) 1-O-Galloyl-beta-D-glucose with
PTGS2 protein; (b) Epicatechin with HSP90AA1 protein; (c) Phloridzin
with EGFR protein; (d) Epicatechin gallate with EGFR protein; (e) of
1-O-Galloyl-beta-D-glucose with AKT1 protein.
[178]pharmaceuticals-18-00896-s001.zip^ (1.7MB, zip)
Author Contributions
Conceptualization, R.M. and S.K.; methodology, R.M. and S.K.;
writing—original draft, review and editing, R.M. and S.K.;
conceptualization, supervision, reviewing and editing, project
administration, H.T., Supervision and Fund acquisition, K.T.C. and H.T.
All authors have read and agreed to the published version of the
manuscript.
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
Data are contained within the article and [179]Supplementary Materials.
Conflicts of Interest
The authors declare no conflicts of interest.
Funding Statement
This work was supported in the part by the National Research Foundation
of Korea (NRF) grant funded by the Korean government (MSIT) (No.
2020R1A2C2005612) AND (No. 2022R1G1A1004613) and in part by the Korean
Big Data Station (K-BDS) with computing resources including technical
support.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data
contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI
and/or the editor(s) disclaim responsibility for any injury to people
or property resulting from any ideas, methods, instructions or products
referred to in the content.
References