New Anticancer Theobromine Derivative Targeting EGFRWT and EGFRT790M: Design, Semi-Synthesis, In Silico, and In Vitro Anticancer Studies

Based on the pharmacophoric features of EGFR inhibitors, a new semisynthetic theobromine-derived compound was designed to interact with the catalytic pocket of EGFR. Molecular docking against wild (EGFRWT; PDB: 4HJO) and mutant (EGFRT790M; PDB: 3W2O) types of EGFR-TK indicated that the designed theobromine derivative had the potential to bind to that pocket as an antiangiogenic inhibitor. The MD and MM-GBSA experiments identified the exact binding with optimum energy and dynamics. Additionally, the DFT calculations studied electrostatic potential, stability, and total electron density of the designed theobromine derivative. Both in silico ADMET and toxicity analyses demonstrated its general likeness and safety. We synthesized the designed theobromine derivative (compound XI) which showed an IC50 value of 17.23 nM for EGFR inhibition besides IC50 values of 21.99 and 22.02 µM for its cytotoxicity against A549 and HCT-116 cell lines, respectively. Interestingly, compound XI expressed a weak cytotoxic potential against the healthy W138 cell line (IC50 = 49.44 µM, 1.6 times safer than erlotinib), exhibiting the high selectivity index of 2.2. Compound XI arrested the growth of A549 at the G2/M stage and increased the incidence of apoptosis.


Introduction
World Health Organization, WHO, statistics indicate that cancer will become the major cause of death within a few years [1]. In response, developing safe and effective cancer therapies that bind with a specific molecular target and destroy cancer cells represents a challenging problem for scientists [2]. Angiogenesis contributes to tumor growth and reproduction. Hence, stopping angiogenesis is considered one of the possible strategies to treat cancer [3]. Angiogenesis and cancer cell growth depend greatly on the epidermal growth factor receptor (EGFR) [4]. Overexpression of EGFR promotes cellular proliferation, differentiation, and survival through downstream signaling pathways. The EGFR gene promotes the growth of various solid tumors and is elevated in multiple types of cancer [5].

Rationale
The ATP-binding active pocket of the EGFR comprises five essential regions including the adenine and ribose binding pockets, in addition to the phosphate binding region,

Rationale
The ATP-binding active pocket of the EGFR comprises five essential regions including the adenine and ribose binding pockets, in addition to the phosphate binding region, hydrophobic regions I and II. The ribose binding pocket and phosphate binding region are not essential for EGFR-TKIs intrinsic activity [47][48][49][50].
In continuation of our efforts in the discovery of EGFR-2 inhibitors [46,[55][56][57], we designed and synthesized a new theobromine derivative as a potential compound that may exert a marked EGFR inhibitory activity and consequently inhibit the growth of tumor cells. The designed compound is a modified analog of 1H-pyrazolo [3,4-d]pyrimidine derivatives.
From a drug design point of view, the theobromine moiety was utilized in the current work to engage the adenine binding pocket. This moiety has six hydrogen acceptor atoms that can bind efficiently with the essential amino acid at the adenine binding pocket. In addition, the acetamide moiety was used as a linker. Furthermore, the hydrophobic head is to be inserted in the hydrophobic region I. Lastly, one of the two methyl groups at the 3-and 7-positions of theobromine moiety was used proposed to occupy the hydrophobic region II (Figure 1). The binding mode of the synthesized compound confirmed the design as each feature occupied its target pocket in the ATP binding site.
In continuation of our efforts in the discovery of EGFR-2 inhibitors [46,[55][56][57], we designed and synthesized a new theobromine derivative as a potential compound that may exert a marked EGFR inhibitory activity and consequently inhibit the growth of tumor cells. The designed compound is a modified analog of 1H-pyrazolo [3,4-d]pyrimidine derivatives.
From a drug design point of view, the theobromine moiety was utilized in the current work to engage the adenine binding pocket. This moiety has six hydrogen acceptor atoms that can bind efficiently with the essential amino acid at the adenine binding pocket. In addition, the acetamide moiety was used as a linker. Furthermore, the hydrophobic head is to be inserted in the hydrophobic region I. Lastly, one of the two methyl groups at the 3-and 7-positions of theobromine moiety was used proposed to occupy the hydrophobic region II (Figure 1). The binding mode of the synthesized compound confirmed the design as each feature occupied its target pocket in the ATP binding site.
The re-docking validation step of the co-crystallized ligands (erlotinib and TAK-285) showed acceptable root-mean-square deviation, RMSD, values of (1.40 and 1.10 Å, respectively), as presented in Figures 2 and 3.   Erlotinib's binding interactions with EGFR WT revealed that it occupied the major pockets (affinity value of −20.35 kcal/mol). A key hydrogen bond (H-B) with Met769, besides four hydrophobic interactions (H-I) with Leu694, Ala719, and Leu820, was observed in the adenine pocket. This was achieved via the quinazoline moiety. Moreover, the hydrophobic pocket I was occupied by the ethynylphenyl moiety through a network of H-Is with Ala719 and Val702, and Lys721 ( Figure 4). Erlotinib's binding interactions with EGFR WT revealed that it occupied the major pockets (affinity value of −20.35 kcal/mol). A key hydrogen bond (H-B) with Met769, besides four hydrophobic interactions (H-I) with Leu694, Ala719, and Leu820, was observed in the adenine pocket. This was achieved via the quinazoline moiety. Moreover, the hydrophobic pocket I was occupied by the ethynylphenyl moiety through a network of H-Is with Ala719 and Val702, and Lys721 ( Figure 4).  Erlotinib's binding interactions with EGFR WT revealed that it occupied the major pockets (affinity value of −20.35 kcal/mol). A key hydrogen bond (H-B) with Met769, besides four hydrophobic interactions (H-I) with Leu694, Ala719, and Leu820, was observed in the adenine pocket. This was achieved via the quinazoline moiety. Moreover, the hydrophobic pocket I was occupied by the ethynylphenyl moiety through a network of H-Is with Ala719 and Val702, and Lys721 ( Figure 4).  The designed theobromine derivative gave a comparable affinity value to erlotinib (−20.11 kcal/mol). In addition, it took the same orientation and interacted with EGFR WT active site similar to erlotinib. The 7-methyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purine of the designed compound was oriented into the adenine pocket of the EGFR WT with a H-B against the essential amino acid, Met769, and seven H-Is with Cys733, Leu820, Leu694, Leu786, Ala719, and Val 702. The NH group of the acetamide moiety was incorporated in an electrostatic attraction with Thr830. On the other side, the benzyl moiety occupied the hydrophobic pocket I via two H-Is with Leu764 and Lys721. The methyl group at the 4position of xanthine moiety was oriented into the hydrophobic pocket II with two H-Is against Leu694 and Val 702 ( Figure 5). The designed theobromine derivative gave a comparable affinity value to erlotinib (−20.11 kcal/mol). In addition, it took the same orientation and interacted with EGFR WT active site similar to erlotinib. The 7-methyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purine of the designed compound was oriented into the adenine pocket of the EGFR WT with a H-B against the essential amino acid, Met769, and seven H-Is with Cys733, Leu820, Leu694, Leu786, Ala719, and Val 702. The NH group of the acetamide moiety was incorporated in an electrostatic attraction with Thr830. On the other side, the benzyl moiety occupied the hydrophobic pocket I via two H-Is with Leu764 and Lys721. The methyl group at the 4-position of xanthine moiety was oriented into the hydrophobic pocket II with two H-Is against Leu694 and Val 702 ( Figure 5).  The designed theobromine derivative gave a comparable affinity value to erlotinib (−20.11 kcal/mol). In addition, it took the same orientation and interacted with EGFR WT active site similar to erlotinib. The 7-methyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purine of the designed compound was oriented into the adenine pocket of the EGFR WT with a H-B against the essential amino acid, Met769, and seven H-Is with Cys733, Leu820, Leu694, Leu786, Ala719, and Val 702. The NH group of the acetamide moiety was incorporated in an electrostatic attraction with Thr830. On the other side, the benzyl moiety occupied the hydrophobic pocket I via two H-Is with Leu764 and Lys721. The methyl group at the 4position of xanthine moiety was oriented into the hydrophobic pocket II with two H-Is against Leu694 and Val 702 ( Figure 5).  The docking outputs of the mutant EGFR (EGFR T790M ) were investigated to support the docking results of the wild type (EGFR WT ). Figure 6 explains the binding of TAK-285, the co-crystallized ligand, (TAK-285) to the EGFR T790M active site. The obtained findings and the reported data were identical [55]. The docking outputs of the mutant EGFR (EGFR T790M ) were investigated to support the docking results of the wild type (EGFR WT ). Figure 6 explains the binding of TAK-285, the co-crystallized ligand, (TAK-285) to the EGFR T790M active site. The obtained findings and the reported data were identical [55]. As displayed in Figure 7, the designed theobromine derivative was bound to the catalytic site of the EGFR T790M in similar to TAK-285. The 7-methyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purine moiety was successfully buried in the adenine pocket to form one H-B with Met793 and four p-p bonds with Leu844, Leu718, Ala743, and Met793. Moreover, the terminal benzyl moiety interacted with the hydrophobic pocket I, forming an electrostatic interaction with Lys745. The methyl group at the 4-position of xanthine moiety was oriented into the hydrophobic pocket II, forming three H-Is with Leu844, Ala743, and Met793. As displayed in Figure 7, the designed theobromine derivative was bound to the catalytic site of the EGFR T790M in similar to TAK-285. The 7-methyl-2,6-dioxo-2,3,6,7tetrahydro-1H-purine moiety was successfully buried in the adenine pocket to form one H-B with Met793 and four p-p bonds with Leu844, Leu718, Ala743, and Met793. Moreover, the terminal benzyl moiety interacted with the hydrophobic pocket I, forming an electrostatic interaction with Lys745. The methyl group at the 4-position of xanthine moiety was oriented into the hydrophobic pocket II, forming three H-Is with Leu844, Ala743, and Met793.

MD Simulations
The MD analyses carried out on the production run showed that, overall, the designed theobromine derivative-EGFR system was stable. The RMSD plot ( Figure 8A) showed a stable average after the first 8 ns at 1.93 Å for the EGFR, blue curve, and the obtained complex, green curve. Furthermore, the RMSD of the designed theobromine derivative, red curve, showed a variation in values ranging from 0.5 Å to, approximately, 3.0 Å. The RMSD of the designed theobromine derivative can be divided into three parts. During the first 20 ns, it shows a large variation (from 0.

MD Simulations
The MD analyses carried out on the production run showed that, overall, the designed theobromine derivative-EGFR system was stable. The RMSD plot ( Figure 8A) showed a stable average after the first 8 ns at 1.93 Å for the EGFR, blue curve, and the obtained complex, green curve. Furthermore, the RMSD of the designed theobromine derivative, red curve, showed a variation in values ranging from 0.5 Å to, approximately, 3.0 Å. The RMSD of the designed theobromine derivative can be divided into three parts. During the first 20 ns, it shows a large variation (from 0.     686  693  700  707  718  725  732  739  746  753  760  767  774  781  788  795  802  809  816  823  830  837  844  851  858  865  872  879  886  893  900  907  914  921  928  935  942  949

MM-GBSA
The binding exact free energy investigation through molecular mechanics with generalized born and surface area, MM-GBSA, (Figure 9) indicates the different energy components that contribute to the process of binding. The designed theobromine derivative showed a total binding energy with the average value of −30.72 kcal/mol. The most favorable calculated contribution in the energy was for the Van Der Waals interaction (average value = −44.45 kcal/mol). Coming next, the electrostatic interactions (average of −25.26 kcal/mol). Furthermore, a decomposition analysis ( Figure 10) was performed to disclose which amino acid residues within 1 nm of the designed theobromine derivative have more contribution to the binding. Leu694 (−1.1), Val702 (1.51), Ala719 (−1.02), Leu753 (−1.0), Thr766 (−1.13), Leu820 (−1.45), and Thr830 (−1.6) are the key amino acid residues that have a significant contribution of a value that is less than −1 kcal/mol.

MM-GBSA
The binding exact free energy investigation through molecular mechanics with generalized born and surface area, MM-GBSA, (Figure 9) indicates the different energy components that contribute to the process of binding. The designed theobromine derivative showed a total binding energy with the average value of −30.72 kcal/mol. The most favorable calculated contribution in the energy was for the Van Der Waals interaction (average value = −44.45 kcal/mol). Coming next, the electrostatic interactions (average of −25.26 kcal/mol). Furthermore, a decomposition analysis ( Figure 10) was performed to disclose which amino acid residues within 1 nm of the designed theobromine derivative have more contribution to the binding. Leu694 (−1.1), Val702 (1.51), Ala719 (−1.02), Leu753 (−1.0), Thr766 (−1.13), Leu820 (−1.45), and Thr830 (−1.6) are the key amino acid residues that have a significant contribution of a value that is less than −1 kcal/mol.

Protein-Ligand Interaction Profiler (PLIP) Studies
Next, the trajectories of the computed MD were clustered, providing several representative frames that represent every obtained cluster.. For every cluster, the PLIP webserver was employed to identify the number as well as the types of the interactions that occur between the designed theobromine derivative and EGFR. Table 1 denotes both number as well as types of interactions that produced from the PLIP webserver. In the first cluster representative, the predominated interaction is a H-B with three bonds. On the other hand, the two remaining cluster representatives had H-Is greater than the H-Bs.
Leu837 Free binding energy (kcal/mol) Amino acids names MMGBSA free energy decomposition of residues within 10 Å of compound XI in EGFR_compound XI complex

Protein-Ligand Interaction Profiler (PLIP) Studies
Next, the trajectories of the computed MD were clustered, providing several representative frames that represent every obtained cluster.. For every cluster, the PLIP webserver was employed to identify the number as well as the types of the interactions that occur between the designed theobromine derivative and EGFR. Table 1 denotes both number as well as types of interactions that produced from the PLIP webserver. In the first cluster representative, the predominated interaction is a H-B with three bonds. On the other hand, the two remaining cluster representatives had H-Is greater than the H-Bs.
Leu837 Free binding energy (kcal/mol) Amino acids names MMGBSA free energy decomposition of residues within 10 Å of compound XI in EGFR_compound XI complex Figure 10.
Binding-free energy decomposition of the EGFR, the designed theobromine derivative complex.

Protein-Ligand Interaction Profiler (PLIP) Studies
Next, the trajectories of the computed MD were clustered, providing several representative frames that represent every obtained cluster. For every cluster, the PLIP webserver was employed to identify the number as well as the types of the interactions that occur between the designed theobromine derivative and EGFR. Table 1 denotes both number as well as types of interactions that produced from the PLIP webserver. In the first cluster representative, the predominated interaction is a H-B with three bonds. On the other hand, the two remaining cluster representatives had H-Is greater than the H-Bs. Besides obtaining types and numbers of interactions, the PLIP also generated a.pse file to see the 3D conformation of the designed theobromine derivative and its interaction with EGFR ( Figure 11). Table 1. Interactions (number and types) detected from the PLIP webserver. Bold amino acids are common in all clusters.

Cluster Number H-Is Amino Acids in EGFR H-Bs Amino Acids in EGFR
Molecules 2022, 27, x FOR PEER REVIEW 14 of 28 Besides obtaining types and numbers of interactions, the PLIP also generated a.pse file to see the 3D conformation of the designed theobromine derivative and its interaction with EGFR ( Figure 11).    Figure 12. The polarity of any system is determined by the dipole moment (µ), which measures the interaction inside the molecule. The µ value of the title compound is 5.8158 Debye (D), and the calculated ground total energy (TE) is −30469.8 eV (Table 2).

DFT Geometry Optimization
All calculations were performed at the B3LYP/6-311G + +(d,p) level of theory. The full optimization of the compound was run without any constraints and presented in Figure 12. The bonding between the intermediate and the reactant through the C14-N2 bond is highlighted with the two angles on either side of the bond as shown in Figure 12. The polarity of any system is determined by the dipole moment (µ), which measures the interaction inside the molecule. The µ value of the title compound is 5.8158 Debye (D), and the calculated ground total energy (TE) is −30469.8 eV (Table 2).    From the FMO analysis, the calculated energy gap (E gap ) between the HOMO (highest occupied molecular orbital) and LUMO (lowest unoccupied molecular orbital) for the ligand is −5.095 eV and the schematic diagram is presented in Figure 13. The E gap value is relatively small and reflects a narrow frontier orbital HOMO-LUMO gap. Based on the Dm and E gap values, the molecule is highly polarizable and chemically reactive [63]. The calculated energy of HOMO (E HOMO ), LUMO (E LUMO ), and E gap was listed in Table 2.  From the FMO analysis, the calculated energy gap (Egap) between the HOMO (highest occupied molecular orbital) and LUMO (lowest unoccupied molecular orbital) for the ligand is −5.095 eV and the schematic diagram is presented in Figure 13. The Egap value is relatively small and reflects a narrow frontier orbital HOMO-LUMO gap. Based on the Dm and Egap values, the molecule is highly polarizable and chemically reactive [63]. The calculated energy of HOMO (EHOMO), LUMO (ELUMO), and Egap was listed in Table 2. The quantum calculations can be used effectively to obtain the global reactivity indices, which provides details about the chemical reactivity or inhibition ability of a molecule. Global reactive indices such as global chemical hardness (η) and softness (σ) describe the inhibitory ability of the molecule to be stable or reactive. A molecule is thought to be a good inhibitor when its hardness value (η) is small and softness value (σ) is high. All electronic parameters and reactivity indices of the molecule, such as ionization potential (IP), electrophilicity (ω), maximal charge acceptance (∆Nmax), chemical potential (µ), energy change (∆E), global chemical softness (σ), global chemical hardness (η), electronegativity (χ), and electron affinity (EA), were estimated according to Koopmans' theory (more information in the supporting data) Based on the calculated values of global reactive indices (Table 2), the molecule under investigation is chemically reactive and can be a good inhibitor against the EGFR protein [64].
At the border area, LUMO and HOMO may not provide a meaningful description of FMO due to the potential of quasi-degenerate levels. At equilibrium state, the product of The quantum calculations can be used effectively to obtain the global reactivity indices, which provides details about the chemical reactivity or inhibition ability of a molecule. Global reactive indices such as global chemical hardness (η) and softness (σ) describe the inhibitory ability of the molecule to be stable or reactive. A molecule is thought to be a good inhibitor when its hardness value (η) is small and softness value (σ) is high. All electronic parameters and reactivity indices of the molecule, such as ionization potential (IP), electrophilicity (ω), maximal charge acceptance (∆N max ), chemical potential (µ), energy change (∆E), global chemical softness (σ), global chemical hardness (η), electronegativity (χ), and electron affinity (EA), were estimated according to Koopmans' theory (more information in the supporting data).
Based on the calculated values of global reactive indices (Table 2), the molecule under investigation is chemically reactive and can be a good inhibitor against the EGFR protein [64].
At the border area, LUMO and HOMO may not provide a meaningful description of FMO due to the potential of quasi-degenerate levels. At equilibrium state, the product of the density of states of a chemical system and its probability distribution function gives the number of occupied states forunit volume. Such a number is utilized to study the physical properties of chemical systems. In this study, the TDOS analysis has been performed and the relative results confirmed that the compound under study had a significant small energy gap, and the highest intensity was reported for orbitals over LUMO orbital ( Figure 14). Such results emphasized the promising inhibitor's efficiency.
the density of states of a chemical system and its probability distribution function gives the number of occupied states forunit volume. Such a number is utilized to study the physical properties of chemical systems. In this study, the TDOS analysis has been performed and the relative results confirmed that the compound under study had a significant small energy gap, and the highest intensity was reported for orbitals over LUMO orbital ( Figure 14). Such results emphasized the promising inhibitor's efficiency.

Surface Potential Mapping
The potential (electron + nuclei) mapping displays the distribution, molecular shape, size, and dipole moments of the electrostatic potential of the molecule [65]. Graphically, molecular electrostatic potential is shown in Figure 15, where the color-coded red, yellow, blue, and green regions denote the electron-rich, slightly electron-rich, electron-deficient, and neutral zones, respectively. The areas around the carbonyl groups represent the most negative potential zones and the slightly electron-rich regions with yellow colors are localized over terminal phenyl and N9 of purine moiety. The electron-deficient areas of blue color are localized on the purine substituted methyl (C13) and C8H25 of purine.

Surface Potential Mapping
The potential (electron + nuclei) mapping displays the distribution, molecular shape, size, and dipole moments of the electrostatic potential of the molecule [65]. Graphically, molecular electrostatic potential is shown in Figure 15, where the color-coded red, yellow, blue, and green regions denote the electron-rich, slightly electron-rich, electron-deficient, and neutral zones, respectively. The areas around the carbonyl groups represent the most negative potential zones and the slightly electron-rich regions with yellow colors are localized over terminal phenyl and N9 of purine moiety. The electron-deficient areas of blue color are localized on the purine substituted methyl (C13) and C8H25 of purine. the density of states of a chemical system and its probability distribution function gives the number of occupied states forunit volume. Such a number is utilized to study the physical properties of chemical systems. In this study, the TDOS analysis has been performed and the relative results confirmed that the compound under study had a significant small energy gap, and the highest intensity was reported for orbitals over LUMO orbital ( Figure 14). Such results emphasized the promising inhibitor's efficiency.

Surface Potential Mapping
The potential (electron + nuclei) mapping displays the distribution, molecular shape, size, and dipole moments of the electrostatic potential of the molecule [65]. Graphically, molecular electrostatic potential is shown in Figure 15, where the color-coded red, yellow, blue, and green regions denote the electron-rich, slightly electron-rich, electron-deficient, and neutral zones, respectively. The areas around the carbonyl groups represent the most negative potential zones and the slightly electron-rich regions with yellow colors are localized over terminal phenyl and N9 of purine moiety. The electron-deficient areas of blue color are localized on the purine substituted methyl (C13) and C8H25 of purine.

Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) Profiling Study
Compounds are approved for use as drugs based on their pharmacokinetic evaluation along with their biological activity, so the pharmacokinetic evaluation of any new compound should be carried out at an early stage of its creation to help prevent the withdrawal of a drug after its approval. ADMET identifies absorption, distribution, metabolism, excretion, and toxicity, but despite the fact that in vitro studies can illustrate these properties, in silico analyses are more advantageous because of the limitations of cost, time, and effort, as well as the strict regulations regarding animal lives [66]. Computed ADMET parameters for the designed theobromine derivative against erlotinib as a reference molecule were determined using Discovery studio 4.0. According to ADMET results, the designed theobromine derivative and erlotinib exhibited similar good levels of solubility and intestinal absorption. Additionally, affinity to pass the blood-brain barrier, BBB, and to inhibit the cytochrome P450 (CYP2D6) was predicted as low and non-inhibitors, respectively. The two compounds (the designed theobromine derivative and erlotinib) showed a difference in the ability to bind with plasma protein as it was predicted to be less and more than 90%, respectively ( Figure 16). Compounds are approved for use as drugs based on their pharmacokinetic evaluation along with their biological activity, so the pharmacokinetic evaluation of any new compound should be carried out at an early stage of its creation to help prevent the withdrawal of a drug after its approval. ADMET identifies absorption, distribution, metabolism, excretion, and toxicity, but despite the fact that in vitro studies can illustrate these properties, in silico analyses are more advantageous because of the limitations of cost, time, and effort, as well as the strict regulations regarding animal lives [66]. Computed ADMET parameters for the designed theobromine derivative against erlotinib as a reference molecule were determined using Discovery studio 4.0. According to ADMET results, the designed theobromine derivative and erlotinib exhibited similar good levels of solubility and intestinal absorption. Additionally, affinity to pass the blood-brain barrier, BBB, and to inhibit the cytochrome P450 (CYP2D6) was predicted as low and non-inhibitors, respectively. The two compounds (the designed theobromine derivative and erlotinib) showed a difference in the ability to bind with plasma protein as it was predicted to be less and more than 90%, respectively ( Figure 16).

In Silico Toxicity Studies
In silico approaches to toxicity prediction have played an indispensable role in drug development, as it avoids ethical regulation, resource availability, and time-wasting in traditional in vitro and in vivo studies [67]. This approach has played an indispensable role in the discovery of new drugs and treatments [68]. In silico toxicity prediction uses structure-activity relationships (SARs) to predict toxicities, and in detail, the computer compares the chemical structural properties of the molecules to thousands of compounds that have been reported to be safe or toxic [69]. A toxicity model built in Discovery studio software was used to estimate eight toxicity parameters. Table 3 shows that the designed theobromine derivative demonstrated high levels of safety in the computed models.

In Silico Toxicity Studies
In silico approaches to toxicity prediction have played an indispensable role in drug development, as it avoids ethical regulation, resource availability, and time-wasting in traditional in vitro and in vivo studies [67]. This approach has played an indispensable role in the discovery of new drugs and treatments [68]. In silico toxicity prediction uses structure-activity relationships (SARs) to predict toxicities, and in detail, the computer compares the chemical structural properties of the molecules to thousands of compounds that have been reported to be safe or toxic [69]. A toxicity model built in Discovery studio software was used to estimate eight toxicity parameters. Table 3 shows that the designed theobromine derivative demonstrated high levels of safety in the computed models.

Chemistry
Because the designed theobromine derivative demonstrated high affinity to bind with EGFR enzyme through different computational studies and it showed a good range of druglikeness, it was semi-synthesized by the modification of theobromine. In the present work, potassium salt formation was achieved on the imidic nitrogen of theobromine VII [70] by treating it with alcoholic KOH while continuously stirring. The formed salt VIII was then reacted with N-benzyl-2-chloroacetamide X, which was formed via the reaction of phenylmethanamine IX with chloroacetylchloride in DMF, to produce the corresponding target product XI (Scheme 1).

Chemistry
Because the designed theobromine derivative demonstrated high affinity to bind with EGFR enzyme through different computational studies and it showed a good range of drug-likeness, it was semi-synthesized by the modification of theobromine. In the present work, potassium salt formation was achieved on the imidic nitrogen of theobromine VII [70] by treating it with alcoholic KOH while continuously stirring. The formed salt VIII was then reacted with N-benzyl-2-chloroacetamide X, which was formed via the reaction of phenylmethanamine IX with chloroacetylchloride in DMF, to produce the corresponding target product XI (Scheme 1).
The IR spectrum of compound XI was characterized by the appearance of carbonyl absorption bands at 1711 and 1656 cm −1 . Moreover, the 1 H NMR revealed the presence of amidic proton at 8.61 ppm. Moreover, a characteristic singlet signal appeared at 4.52 ppm corresponding to the CH2 group. The 13 C NMR spectrum, which displayed distinct peaks at approximately 43.44, 42.52, 33.67, and 29.91 ppm corresponding to the two CH3 and two CH2 groups, respectively, further confirmed the validity of the proposed structure. The IR spectrum of compound XI was characterized by the appearance of carbonyl absorption bands at 1711 and 1656 cm −1 . Moreover, the 1 H NMR revealed the presence of amidic proton at 8.61 ppm. Moreover, a characteristic singlet signal appeared at 4.52 ppm corresponding to the CH 2 group. The 13 C NMR spectrum, which displayed distinct peaks at approximately 43.44, 42.52, 33.67, and 29.91 ppm corresponding to the two CH 3 and two CH 2 groups, respectively, further confirmed the validity of the proposed structure.
The chemical structure of compound XI (Figure 17) was confirmed through spectroscopic analysis. In detail, the NH proton resonated clearly at δ 8.61 as a triplet signal (J = 5.9 Hz) with the integration of one due to the presence of neighboring CH 2 . Two CH 2 groups were detected at δ 4.30 (d, J = 5.9 Hz, 2H), which is adjacent to the NH, and at 4.52 (s, 2H) which is adjacent to the carbonyl. The proton of C-8, the only proton of the theobromine rings, resonated at δ 8.07 (s, 1H). The two methyls of the theobromine moiety (3, 7) resonated as two singlet peaks at δ 3.44, and 3.90, respectively. Furthermore, a pattern of monosubstituted benzene ring was discovered at δ 7.37-7.31 (2H) and 7.26 (3H). Similarly, the 13 C NMR showed three carbonyls at δ 167.50, for the acetamide moiety, in addition to the two up-field carbonyls of the theobromine moiety because of the high anisotropic effects at δ 154. 73  The chemical structure of compound XI (Figure 17) was confirmed through spectroscopic analysis. In detail, the NH proton resonated clearly at δ 8.61 as a triplet signal (J = 5.9 Hz) with the integration of one due to the presence of neighboring CH2. Two CH2 groups were detected at δ 4.30 (d, J = 5.9 Hz, 2H), which is adjacent to the NH, and at 4.52 (s, 2H) which is adjacent to the carbonyl. The proton of C-8, the only proton of the theobromine rings, resonated at δ 8.07 (s, 1H). The two methyls of the theobromine moiety (3, 7) resonated as two singlet peaks at δ 3.44, and 3.90, respectively. Furthermore, a pattern of monosubstituted benzene ring was discovered at δ 7.37-7.31 (2H) and 7.26 (3H). Similarly, the 13

EGFR Inhibition
To examine the design and the computational results that indicated the strong binding affinity of compound XI to the EGFR enzyme, compound XI was tested in vitro against the EGFR enzyme in comparison with erlotinib. As shown in Figure 18, compound XI strongly inhibited the EGFR enzyme with 17.2 nM that was bordering erlotinib's value. The obtained results were consistent with the acquired in silico results and confirmed the strong suppressing potential of compound XI.
(A) Figure 17. Chemical structure of compound XI.

EGFR Inhibition
To examine the design and the computational results that indicated the strong binding affinity of compound XI to the EGFR enzyme, compound XI was tested in vitro against the EGFR enzyme in comparison with erlotinib. As shown in Figure 18, compound XI strongly inhibited the EGFR enzyme with 17.2 nM that was bordering erlotinib's value. The obtained results were consistent with the acquired in silico results and confirmed the strong suppressing potential of compound XI. The chemical structure of compound XI (Figure 17) was confirmed through spectroscopic analysis. In detail, the NH proton resonated clearly at δ 8.61 as a triplet signal (J = 5.9 Hz) with the integration of one due to the presence of neighboring CH2. Two CH2 groups were detected at δ 4.30 (d, J = 5.9 Hz, 2H), which is adjacent to the NH, and at 4.52 (s, 2H) which is adjacent to the carbonyl. The proton of C-8, the only proton of the theobromine rings, resonated at δ 8.07 (s, 1H). The two methyls of the theobromine moiety (3, 7) resonated as two singlet peaks at δ 3.44, and 3.90, respectively. Furthermore, a pattern of monosubstituted benzene ring was discovered at δ 7.37-7.31 (2H) and 7.26 (3H). Similarly, the 13

EGFR Inhibition
To examine the design and the computational results that indicated the strong binding affinity of compound XI to the EGFR enzyme, compound XI was tested in vitro against the EGFR enzyme in comparison with erlotinib. As shown in Figure 18, compound XI strongly inhibited the EGFR enzyme with 17.2 nM that was bordering erlotinib's value. The obtained results were consistent with the acquired in silico results and confirmed the strong suppressing potential of compound XI. (A)

Cytotoxicity and Safety
An in vitro cytotoxicity assay to evaluate compound XI's EGFR inhibition against cancer was conducted using the A549 and HCT-116 malignant cell lines against erlotinib as a reference. Compound XI demonstrated IC50 values of 21.99 and 22.02 µM, respectively ( Table 4). The anticancer potentials of compound XI were very close to those of erlotinib (6.73 and 16.35 µM, respectively).
The cytotoxic potential of compound XI against the W138 normal human cell line was evaluated to confirm its in silico safety results and determine its selectivity against cancer cell lines. Compound XI displayed excellent safety results, expressing a high IC50 value of 49.44 µM (safer than erlotinib) and very high selectivity indexes (SI) against both examined cell lines of 2.2 ( Figure 19).

Cytotoxicity and Safety
An in vitro cytotoxicity assay to evaluate compound XI's EGFR inhibition against cancer was conducted using the A549 and HCT-116 malignant cell lines against erlotinib as a reference. Compound XI demonstrated IC 50 values of 21.99 and 22.02 µM, respectively ( Table 4). The anticancer potentials of compound XI were very close to those of erlotinib (6.73 and 16.35 µM, respectively). The cytotoxic potential of compound XI against the W138 normal human cell line was evaluated to confirm its in silico safety results and determine its selectivity against cancer cell lines. Compound XI displayed excellent safety results, expressing a high IC 50 value of 49.44 µM (safer than erlotinib) and very high selectivity indexes (SI) against both examined cell lines of 2.2 ( Figure 19).

Figure 19.
In vitro anti-proliferative activities of the compound XI and erlotinib against EGFR, cancer and normal cell lines.

Cell Cycle Analysis and Apoptosis Assay
Initially, flow cytometric analysis of cell cycle phases was carried out [71,72]. Compound XI was treated with A549 cells for 72 h at a concentration of 21.99 µM. Then, the different stages of the cell cycle were examined. Compound XI decreased the growth of cells in the Sub-G1 and G1 phases from 0.87% to 0.80% and from 43.47% to 16%, respectively. Conversely, the f A549 population percentage was significantly increased from 12.10 for control cells to 19.84 for XI-treated cells in the G2/M phase (Table 5 and Figure  20). To confirm the apoptotic effects of XI, A549 cells were stained with Annexin V and PI double stains after treatment of with 21.99 µM of XI for 72 h [73,74]. Comparing the control, compound XI induced a higher number of apoptotic cells. Compound XI caused a significant increase in the apoptotic cells percentage in both early apoptosis (from 0.05% to 24.02%) and late apoptosis (from 0.49% to 41.70%). A noticeable change in the total apoptosis percentage by 65.72 was observed, compared to 0.54% in the control cells (Figure 20 and Table 6). In conclusion, compound XI primarily arrested the cancer cell cycle at the G2/M stage caused cytotoxic activities that may be due to programmed apoptosis. (See Supplementary Materials)

Cell Cycle Analysis and Apoptosis Assay
Initially, flow cytometric analysis of cell cycle phases was carried out [71,72]. Compound XI was treated with A549 cells for 72 h at a concentration of 21.99 µM. Then, the different stages of the cell cycle were examined. Compound XI decreased the growth of cells in the Sub-G1 and G1 phases from 0.87% to 0.80% and from 43.47% to 16%, respectively. Conversely, the f A549 population percentage was significantly increased from 12.10 for control cells to 19.84 for XI-treated cells in the G2/M phase (Table 5 and Figure 20). To confirm the apoptotic effects of XI, A549 cells were stained with Annexin V and PI double stains after treatment of with 21.99 µM of XI for 72 h [73,74]. Comparing the control, compound XI induced a higher number of apoptotic cells. Compound XI caused a significant increase in the apoptotic cells percentage in both early apoptosis (from 0.05% to 24.02%) and late apoptosis (from 0.49% to 41.70%). A noticeable change in the total apoptosis percentage by 65.72 was observed, compared to 0.54% in the control cells ( Figure 20 and Table 6). In conclusion, compound XI primarily arrested the cancer cell cycle at the G2/M stage caused cytotoxic activities that may be due to programmed apoptosis. (See Supplementary Materials).

Docking Studies
The molecular docking was carried out by MOE2014 software [75]. Supplementary data give a detailed explanation.

M D Simulations
CHARMM-GUI web server and GROMACS 2021 were utilized as an MD engine [76,77]. Supplementary data give a detailed explanation.

MM-GBSA
The Gmx_MMPBSA package was utilized [78]. Supplementary data give a detailed explanation.  The molecular docking was carried out by MOE2014 software [75]. Supplementary data give a detailed explanation.

MD Simulations
CHARMM-GUI web server and GROMACS 2021 were utilized as an MD engine [76,77]. Supplementary data give a detailed explanation.

MM-GBSA
The Gmx_MMPBSA package was utilized [78]. Supplementary data give a detailed explanation.

ADMET Studies
ADMET profile was carried out by Discovery Studio 4.0 [81]. Supplementary data give a detailed explanation.

Toxicity Studies
The toxicity profile was carried out by Discovery Studio 4.0. Supplementary data give a detailed explanation.

Synthesis of Compound XI
To a solution of the potassium salt of theobromine VIII (0.001 mol, 0.25 g) in dry DMF (10 mL), N-benzyl-2-chloroacetamide X (0.001 mol, 0.21 g) was added and the mixture was heated for 5 h. The mixture was then cooled and the produced precipitate was filtered, washed with water, and crystallized from ethanol to attain the final target compound XI (Scheme 2). Gaussian 09 and GaussSum3.0 programs [79] were utilized [80]. Supplementary data give a detailed explanation.

ADMET Studies
ADMET profile was carried out by Discovery Studio 4.0 [81]. Supplementary data give a detailed explanation. 3.1.6. Toxicity Studies The toxicity profile was carried out by Discovery Studio 4.0. Supplementary data give a detailed explanation.

Synthesis of Compound XI
To a solution of the potassium salt of theobromine VIII (0.001 mol, 0.25 g) in dry DMF (10 mL), N-benzyl-2-chloroacetamide X (0.001 mol, 0.21 g) was added and the mixture was heated for 5 h. The mixture was then cooled and the produced precipitate was filtered, washed with water, and crystallized from ethanol to attain the final target compound XI (Scheme 2).

In Vitro Egfr Inhibition
This was performed using the Human EGFR ELISA kit. The supplementary data provide a thorough explanation.

In Vitro Antiproliferative Activity
MTT procedure was utilized. The supplementary data provide a detailed explanation.

Safety Assay
The normal cell lines, W138, were utilized. The supplementary data provide a detailed explanation.

In Vitro Egfr Inhibition
This was performed using the Human EGFR ELISA kit. The supplementary data provide a thorough explanation.

In Vitro Antiproliferative Activity
MTT procedure was utilized. The supplementary data provide a detailed explanation.

Safety Assay
The normal cell lines, W138, were utilized. The supplementary data provide a detailed explanation.

Cell Cycle Analysis and Apoptosis
The effect of compound XI on cell cycle distribution and apoptosis was performed using flowcytometry analysis technique.
The supplementary data provide a detailed explanation.

Conclusions
Based on the essential structural properties of EGFR inhibitors, a new theobromine derivative was designed as an inhibitor. The potentiality of the designed theobromine derivative against EGFR was demonstrated by molecular docking against both wild and mutant types. The binding was confirmed by six MD (over 100 ns), two MM-GBSA, PLP, and four DFT experiments. In addition, an ADMET analysis confirmed the general likeness and safety. In vitro results were consistent with the in silico results, displaying EGFR inhibition with an IC 50  The results show that compound XI (the theobromine derivative) is a lead compound that could be employed for further modifications or in vivo and preclinical studies.