Fluorinated derivatives of tetrahydroaltersolanol molecule on COVID-19, HIV, and HTLV protease by DFT and molecular docking approaches.

Structural, optoelectronic, and biological properties of tetrahydroaltersolanol (C16H20O7) and fluorinated derivatives are calculated using density functional theory (DFT) and molecular docking approaches. It is shown that the pure C16H20O7 molecule has a direct HOMO-LUMO energy gap about 3.1 eV. The substitution of F atom at A category decreases the electronic energy gap, while it is constant at B category. In A category, the behavior of the pure molecule changed from insulator to semiconductor with various substitution of F atom. The electronic properties were depended on the F sites in the pure molecule. The molecular electrical transport properties and charge-transfer possibilities increase with decreasing energy gap. The pure C16H20O7 molecule with high energy gap has low chemical reactivity and substitution of F atom at considered molecule increases chemical reactivity. Obtained results show that F-O bonds in trifurcation bonds of C16H19O7(F14), C16H19O7(F16), and C16H19O7(F17) molecules play a key role in confronting with COVID-19, HIV, and HTLV proteases, respectively. Optical spectra, such as the dielectric functions, electron energy-loss spectroscopy, refractive index, extinction coefficient, and reflection spectra show that fluorinated derivatives of C16H20O7 at B category can be used in the new drugs.

Introduction

Fungi are found in all environments with a universal spread and grow in different habitats [1-3]. Hydroanthraquinones, alkaloids, polyketides, and peptides are bioactive materials which produce by species of fungi and some of them are toxic to plants, insects and animals. Fungal metabolites have various applications in medicine as antiviral, antibacterial, and anticancer compounds [4-6]. The reduced species of anthraquinones that are formed in fungi such as Alternaria solani [7, 8], A. porri [9], Dactylaria lutea [10], Phomopsis juniperivora [11], and Stemphylium botryosum [12] are called altersolanol, which contains many different types and exhibits interesting properties. For instance, altersolanols exhibited significant antimicrobial, antiprotozoal, and cytotoxic activities [13]. Marine fungi have recently emerged as a newfound source of fungal natural products and maybe have a potentially crucial point in drug discovery.

Genetic material (DNA and RNA) of the virus encodes viral protease as essential enzymes to complete the viral infections cycle. These enzymes cleave selectively the scissile peptide bonds with different catalytic mechanisms in viral or cellular protein. Serine, cysteine, or aspartic acid residues in the active site of viral proteases recognize the specific peptide bond that is often located within conserved sequence motifs. The 2019-nCOV as novel coronavirus has main cysteine protease named (M, 3C) [14], HIV and HTLV protease [15]. The aspartic acid protease family have potential drug targets because of their essential activities. One of the main methods commonly used to study the physical and chemical properties of materials is density functional theory (DFT) approach [16] as implemented in the Wien2k code [17]. Fluorine with high electronegativity in carbon–fluorine bond plays a key role to create a strongly polarized bond with the highest energies. A hydrogen atom or a hydroxyl group can displace with fluorine [18, 19]. Fluorine is rarely found in natural compounds, and monofluorinated compounds are produced by some plants and bacteria [20, 21]. They often play a key role in defense strategy due to their toxic properties. Nowadays, fluorinated compounds have various applications in agriculture and medicine [22]. On the other hand, hydroanthraquinone derivatives as bioactive metabolites were obtained from a soft coral–derived Alternaria sp. fungus [23]. One of them is a tetrahydroaltersolanol with C16H20O7 formula and SMILES (O = C1c2c(O)cc(cc2[C@@H](O)[C@@H]2[C@H] (O)[C@@] (O)([C@@H] (O)C[C@@H]12) C) OC) exhibited antiviral activity against porcine reproductive and respiratory syndrome virus (PRRSV) with IC50 values 65 ℳM [23].

In this study, we have investigated the structural, optoelectronic, and biological properties of tetrahydroaltersolanol (C16H20O7) [23] and fluorinated derivatives by using density functional theory (DFT) and molecular docking approaches. The 19 fluorinated derivatives of tetrahydroaltersolanol are constructed via hydrogen substitution by fluorine. The optoelectronic and ADME properties were predicted by the Wien2k code and Molsoft web server, respectively. Finally, binding mode of tetrahydroaltersolanol and designed compounds to HIV, HTLV protease, and COVID-19 protease were obtained by molecular docking and compared with the obtained DFT results.

Computational details

A. The FP-LAPW method as implemented in WIEN2k

In the first principal calculations, the full-potential linearized augmented plane wave method (FP- LAPW) is used for simulation of the structural and optoelectronic properties of pure C16H20O7 and C16H19O7(F1) to C16H19O7(F19) compounds by Wien2k code [16, 17, 24]. The exchange–correlation potentials are calculated using the generalized gradient approximation (GGA). The tetrahydroaltersolanol (C16H20O7) compound is crystalized at orthorhombic structure with the “P 21 21 21” space group, as shown in Fig. 1. The lattice constants are a = 5.6702 Å, b = 10.8579 Å, and c = 23.9550 Å. The number of k-points is selected about 1000 for self-consistent calculations. In the FP- LAPW calculations, the cut-off energy, which defines the separation of core and valence states is also selected − 6 Ry with the convergence parameter of 0.0001 Ry.

Fig. 1

The unit cells of a C16H20O7 and b C16H19O7(F1)

B. Molecular docking and molecular dynamic

Molecular docking is a computation process by which ligands are moved in a 3D space so that, finally, the best space structures of the target and ligand positions are found by getting the highest scores in the functions. Before docking process, to predict drug-likeness of compounds, ADME properties of the compounds were predicted using Molsoft web services (https://molsoft.com/mprop/) which estimates the absorption, distribution, metabolism, and excretion of candidate compounds. For docking process, proteins and ligands optimized. At the first step, primary ligand optimization was performed by Hyperchem software [25], a salvation box was created for them, and molecular dynamic simulation was done by amber force field [25] under the temperature of 300 K in 10 picoseconds in order to achieve the most sustainable state and the lowest energy of the structure of inhibitors for being used in the future studies. Then, these structures were saved as PDB files to be used in the docking process [26]. At the second step, the crystallography structure of HIV, HTLV, and COVID-19 protease enzymes were selected from the protein data bank with the respective PDB codes of 1YT9, 3LIN, and 6M03. Then, all the non-amino acid molecules (water, ion, ligand, inhibitor, etc.) were removed. Then, the structure of the HTLV and HIV protease’s optimization was done by Gromacs 5.1.2 software [27] in a quasi-natural condition in terms of temperature, atmosphere pressure, pH, and the existence of water as a solvent to get their balanced structure. First, the topology of protein was built by using the GROMOS96 53a5 force field. Then, a cubic box was considered between the protein and the edge of the box with the distance of 1.0 nm, and the water molecules were added to the box as solvent by using simple point charge (SPC) model. Then, four chlorine ions were added for neutralizing the system load. In the following, energy minimization was done by the steepest descent algorithm. Then, NVT equilibration and NPT equilibration were done each for 100 picoseconds, and finally, molecular dynamic simulation was done for 10,000 picoseconds (10 ns). The docking process was done by Auto Dock 4.2 software. In the docking process, LGA (Lamarckian genetic algorithm) was used and all the non-rotatable amide bonds were selected. The cubic box with the size of 126 × 126 × 126 Å was set to cover all the protein–ligand complexes. The number of GA runs for each 250 docks was selected by 25,000,000 energy measurement [28]. The content of hydrogen bonds and hydrophobic interactions between ligand and protein in the complex was analyzed by PLIP server. Data were drawn with PyMol software [29].

Results and discussion

Today, in medicine, plants and fungi are used as rich sources of various metabolites to treat various diseases. The resulting metabolites can have inhibitory effects on the cancer cells or viral proteins. On the other hand, rare natural fluorinated compounds are produced by plants and bacteria that play a major role in defense mechanisms. These compounds have attracted special attention as effective drug compounds. Fluorinated derivatives of chemical compounds are usually toxic; still it should also be borne in mind that the enzymes that break down fluorinated compounds are found only in small numbers in some organisms. Therefore, the necessary precautions should be taken in their therapeutic uses. One of the hydroanthraquinone derivatives extracted from marine fungi with antiviral activity against porcine reproductive and respiratory syndrome virus (PRRSV) with IC50 values 65 micromolar (μM) [23] was selected to design its fluorinated derivatives. The hydrogen atoms in hydroanthraquinones (Cif code: 1,503,469) were substituted with fluorine (Fig. 2). As the first step to design new compounds in lab, they should be energetically stable. To estimate stability of designed fluorinated substance, electronic and optical properties of them was calculated. After that, drug likeness and ADME properties of fluorinated substances was predicted. Finally, binding mode of compounds to HTLV, HIV, and COVID-19 protease was obtained.

Fig. 2

Fluorinated derivatives of hydroanthraquinones. The figures were generated by PyMol [31]

A. Optoelectronic results

The physical and chemical properties of materials are studied by investigating the behavior of electrons near the fermi level. The electrons in fermi level play a key role in the reactive properties of drugs. In this part, an attempt has been made to obtained correlation between DFT calculations and pharmaceutical properties. It is shown that several computational parameters such as highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) band gap and absorption spectra were predicted in order to describe the pharmaceutical properties of tetrahydroaltersolanol (C16H20O7) and fluorinated derivatives. The calculated electronic band structures and density of states of the C16H20O7 change by the replacement of a hydrogen atom by a fluorine in C16H19O7(F1 to F19) compounds, as shown in Figs. 3, 4, 5, and 6. The calculated bandgap values of all compounds are given in Table 1. For better presentation of the obtained results, all compounds are classified into one of two groups: category A or category B. Category A involves the fluorine atoms that joined to the oxygen atom, while in category B, the fluorine atom is attached to carbon that are located outside benzene rings with three hydrogen bonds or on benzene rings. The results show that the C16H20O7 compound has a direct bandgap about 3.1 eV (Fig. 3). The C16H20O7 bandgap is reduced upon the replacement of a hydrogen atom by a fluorine in the C16H19O7 (F1 to F5) compounds. On the other hand, obtained bandgaps for C16H19O7(F6 to F19) compounds, except for C16H19O7(F13) compound, are direct and near to the bandgap value of pure one. By decreasing the electronic band gap, the behavior of the C16H20O7 compound is altered from semiconductor to metal.

Fig. 3

The band structures of a C16H20O7, b C16H19O7(F4) from category A, c C16H19O7(F6), d C16H19O7(F13), e C16H19O7(F16), and f C16H19O7(F18) from category B. The black dash line shows Fermi energy level

Fig. 4

The calculated total density of states for a C16H20O7 and b C16H19O7(F1), c C16H19O7(F2), d C16H19O7(F3), e C16H19O7(F4), and f C16H19O7(F5) at A category of compounds. In these compounds, the fluorine atom is bonded to oxygen atom. The gray dot line shows Fermi energy level

Fig. 5

The calculated total density of states for a C16H20O7(F6), b C16H19O7(F7), c C16H19O7(F8), d C16H19O7(F9), e C16H19O7(F10), f C16H19O7(F11), g C16H19O7(F12), and h C16H19O7(F13) at B category of compounds. In these compounds, the fluorine atom is bonded to carbon inside benzene ring. The gray dot line shows Fermi energy level

Fig. 6

The calculated total density of states for a C16H20O7(F14), b C16H19O7(F15), c C16H19O7(F16), d C16H19O7(F17), e C16H19O7(F18), and f C16H19O7(F19) at B category of compounds. In these compounds, the fluorine atom is bonded to carbon outside benzene ring with three hydrogen bonds. The gray dot line shows Fermi energy level

Table 1

Obtained electrical and optical data for C16H20O7 and C16H19O7(F1) to C16H19O7(F19)

Compound	Eg	ε0x	ε0y	ε0z	n0x	n0y	n0z
C16H20O7	3.108 Γ-Γ	2.55426	3.12551	2.76888	1.59821	1.76792	1.664
C16H19O7(F1)	0.32 Γ-Γ	2.93709	5.56767	3.71813	1.7138	2.3599	1.92831
C16H19O7(F2)	0.1 Γ-Γ	3.66598	4.81036	4.25495	1.91499	2.19384	2.06315
C16H19O7(F3)	0.08 Γ-Γ	3.78383	4.31200	3.58904	1.94575	2.07696	1.89469
C16H19O7(F4)	0.2 Γ-X	6.53135	4.09997	4.71215	2.55698	2.02496	2.17119
C16H19O7(F5)	1.367 X-X	2.82586	3.63965	2.97702	1.70836	1.94341	1.75802
C16H19O7(F6)	3.037 Γ-Γ	2.57459	3.03608	2.79362	1.60455	1.74244	1.67142
C16H19O7(F7)	2.886 Γ-Γ	2.56741	3.08806	2.88624	1.60232	1.75729	1.6989
C16H19O7(F8)	3.102 Γ-Γ	2.57263	3.08293	2.86346	1.60395	1.75583	1.69218
C16H19O7(F9)	3.075 Γ-Γ	2.57579	3.00751	2.82737	1.60493	1.73422	1.68148
C16H19O7(F10)	3.168 Γ-Γ	2.56482	3.04579	2.84275	1.60152	1.74525	1.68607
C16H19O7(F11)	3.144 Γ-X	2.53929	3.08661	2.79723	1.59352	1.75688	1.6725
C16H19O7(F12)	2.883 Γ-Γ	2.65766	3.18483	2.96684	1.63024	1.78462	1.72246
C16H19O7(F13)	2.431 Γ-Γ	2.60264	3.27018	2.87743	1.61327	1.80837	1.6963
C16H19O7(F14)	2.991 Γ-Γ	2.57207	3.03811	2.80066	1.60377	1.74302	1.67352
C16H19O7(F15)	3.141 Γ-Γ	2.58228	3.03303	2.86712	1.60695	1.74156	1.69326
C16H19O7(F16)	3.016 Γ-Γ	2.56321	3.06124	2.83321	1.60101	1.74964	1.68322
C16H19O7(F17)	3.124 Γ-Γ	2.59432	3.06465	2.84178	1.61069	1.75062	1.68576
C16H19O7(F18)	3.190 Γ-Γ	2.55828	3.04666	2.83598	1.59947	1.74547	1.68404
C16H19O7(F19)	3.200 Γ-Γ	2.57751	3.00706	2.82623	1.60546	1.73409	1.68114

The HOMO–LUMO band gap describes the charge transfer interaction within the molecule. The pure tetrahydroaltersolanol (C16H20O7) and the B category of fluorinated derivatives with a high HOMO–LUMO band gap values have low chemical reactivity and high kinetic stability. The B category of fluorinated derivatives with the high HOMO–LUMO band gap are stable and hence are chemically harder than the A category of compounds. The B category compounds with the wide bandgap have semiconducting properties.

Calculated partial density of states of the pure and fluorinated-C16H20O7 are shown in Fig. 4. It is seen that the contribution of s-states is small and p-states have a remarkable contribution. The s-states of hydrogen atom and p-states of carbon atom have a small contribution in the total density of states. In addition, it is shown for pure C16H20O7, the p-states of the oxygen atom have a considerable contribution in HOMO and LUMO. In A category compounds by substitution of fluorine atom, the p-states of the fluorine atom have a remarkable contribution at the Fermi level and LUMO. On the other hand, in B category compounds such as C16H19O7(F6 to F13), the p-states of the oxygen atom play a key role at LUMO (Fig. 5).

Figure 6 shows that the p-states of the fluorine atom have the high contribution near the Fermi level at HOMO and the contribution of the p-states of the oxygen atom is small. Nevertheless, the Tables 2 and 3 indicate that the p-states of the fluorine and oxygen atoms have the main role in the reactivity of C16H19O7(F14) and C16H19O7(F16) compounds with COVID-19 and HIV proteases, respectively.

Table 2

Calculated biological properties of compounds by Molsoft tool for the Lipinski’s rule

	Molecular formula	Molecular weight	Number of hydrogen bond acceptor	Number of hydrogen bond donor	MolLogP	MolLogS	MolPSA(A2)	MolVol(A3)	pKa	BBB Score	Number of stereo centers	Drug- likeness model score
	C16 H20 O7	324.12	7	5	− 4.34	− 0.71	99.17	314.30	< 0. /8.25	2.40	6	0.45
1	C16 H19 O7(F1)	344.13	7	5	− 0.19	− 0.59	93.32	313.01	< 0. /10.03	2.40	6	0.04
2	C16 H19 O7(F2)	344.13	7	5	− 3.97	− 0.51	93.55	313.68	< 0. /7.95	2.46	7	0.14
3	C16 H19 O7(F3)	344.13	7	5	− 3.95	− 0.58	93.60	314.88	< 0. /7.95	2.46	7	0.25
4	C16 H19 O7(F4)	344.13	7	5	− 3.93	− 0.58	93.60	313.67	< 0. /7.95	2.46	7	0.24
5	C16 H19 O7(F5)	344.13	7	5	− 0.62	− 0.60	93.25	321.15	< 0. /13.35	2.46	7	− 0.10
6	C16 H19 O7(F6)	344.13	7	5	− 4.10	− 0.57	101.15	325.29	< 0.7.95	2.35	7	0.29
7	C16 H19O7(F7)	344.13	7	5	− 4.13	− 0.51	93.32	313.01	< 0. /7.95	2.46	7	0.04
8	C16 H19 O7(F8)	344.13	7	6	− 4.66	− 0.47	101.43	311.75	< 0. /7.95	2.35	8	− 0.24
9	C16 H19 O7(F9)	344.13	7	6	− 4.66	− 0.47	101.43	311.75	< 0. /7.95	2.35	8	− 0.24
10	C16H19O7(F10)	344.13	7	6	− 4.41	− 0.59	101.15	325.91	< 0. /7.95	2.35	7	− 0.12
11	C16H19O7(F11)	344.13	7	6	− 4.41	− 0.53	101.15	331.45	< 0. /7.82	2.35	7	0.32
12	C16H19O7(F12)	344.13	7	6	− 4.50	− 0.53	101.53	329.00	< 0. /7.95	2.35	7	− 0.31
13	C16H19O7(F13)	344.13	7	6	− 3.88	− 0.58	100.55	318.86	< 0. /6.75	2.36	7	0.24
14	C16H19O7(F14)	344.13	7	6	− 4.58	− 0.50	101.68	317.31	< 0. /7.95	2.34	7	0.49
15	C16H19O7(F15)	344.13	7	6	− 4.58	− 0.50	101.68	317.31	< 0. /7.95	2.34	7	0.49
16	C16H19O7(F16)	344.13	7	6	− 4.58	− 0.50	101.68	317.31	< 0. /7.95	2.34	7	0.49
17	C16H19O7(F17)	344.13	7	6	− 4.42	− 0.50	100.09	318.56	< 0. /7.95	2.37	7	0.35
18	C16H19O7(F18)	344.13	7	5	− 4.42	− 0.50	100.09	318.56	< 0. /7.95	2.37	7	0.35
19	C16H19O7(F19)	344.13	7	5	− 4.42	− 0.50	100.09	318.56	< 0. /7.95	2.37	7	0.35

Table 3

Chemical name, formula, and inhibition constant (Ki), binding energy, the number of hydrogen bond between ligand and HIV protease (HB), the names of residue that participated in hydrogen bonds

HIV protease
Name	Formula	Ki (ℳM)	Binding energy (kcal/mol)	HB	Amino acids	Hydrophobic interactions	Amino Acids
1	C16H20O7	27.32	− 6.23	10	GLN2B, THR4B (3), TRP6B, LEU10B (2), THR12B, GLU21B (2)	4	PRO1B, ILE3B, LEU10B, THR12B
4	C16H19O7(F4)	26.9	− 6.23	10	GLN7B (2), THR4B (3), TRP6B, LEU10B (2), THR12B, GLU21B	4	PRO1B, ILE3B, LEU10B, THR12B
6	C16H19O7(F6)	54.53	− 5.82	6	THR4A (2), GLN7A (2), ARG8A (2)	2	TRP6A, LEU10A
7	C16H19O7(F7)	15	− 6.58	9	GLN2B (2), THR4B (2), TRP6B, LEU10B (2), THR12B, GLU21B	4	PRO1B, ILE3B, LEU10B, THR12B
8	C16H19O7(F8)	51.88	− 5.85	10	TRP6B, GLN2B, THR4B (3), LEU10B (2), THR12B, GLU21B (2)	4	PRO1B, ILE3B, LEU10B, THR12B
9	C16H19O7(F9)	112.95	− 5.38	10	THR12B, GLN2B, THR4B (3), TRP6B, LEU10B (2), GLU21B (2)	4	PRO1B, ILEU3B, LEU10B, THR12B
10	C16H19O7(F10)	36.96	− 6.05	7	THR4A, TRP6A, GLN7A, ARG8A (3), LEU10A	4	TRP6A, GLN7A, ARG8A, LEU10A
11	C16H19O7(F11)	37.32	− 6.04	11	GLN2B (2), THR4B (3), TRP6B, LEU10B (2), THR12B, GLU21B (2)	4	PRO1B, ILE3B, LEU10B, THR12B
12	C16H19O7(F12)	31.72	− 6.14	10	GLN2B (2), THR4B (3), TRP6B, LEU10B (2), THR12B, GLU21B	4	PRO1B, ILE3B, LEU10B, THR12B
13	C16H19O7(F13)	33.76	− 6.1	11	GLN2B (2), THR4B (3), TRP6B, LEU10B (2), THR12B, GLU21B (2)	3	ILE3B, LEU10B, THR12B
14	C16H19O7 (F14)	122.34	− 5.34	9	GLN2B, THR4B (2), TRP6B, LEU10B, THR12B, GLU21B (3)	3	PRO1B, ILE3B, THR12B
15	C16H19O7(F15)	83.39	− 5.56	11	GLN2B (2), THR4B (3), TRP6B, LEU10B (2), THR12B, GLU21B (2)	3	PRO1B, ILE3B, THR12B
16	C16H19O7(F16)	684.75	− 4.32	8	ARG8A (4), GLN7A (3), LEU10A	1	LEU10A
17	C16H19O7(F17)	63.5	− 5.73	10	GLN2B (2), THR4B (2), TRP6B, LEU10B (2), THR12B, GLU21B (2)	4	PRO1B, ILE3B, LEU10B, THR12B
18	C16H19O7(F18)	33.94	− 6.1	11	GLN2B (2), THR4B (3), TRP6B, LEU10B (2), THR12B, GLU21B (2)	4	PRO1B, ILE3B, LEU10B, THR12B
19	C16H19O7 (F19)	30.93	− 6.15	2	LEU10, THR12

For the other C16H19O7(F17), C16H19O7(F18), and C16H19O7(F19) compounds, the p-states of the oxygen atom originates near the Fermi level and the contribution of the p-states of the fluorine is ignorable and it is observed in the semi-core region. Hence, in C16H19O7(F17) compound, it is clear that the p-states of the oxygen are affected in reactivity with HTLV protease. In other words, in some of the B category compounds C16H19O7(F14 to F19), the p-states of the oxygen and fluorine atoms play a key role in the pharmaceutical properties of compounds.

To analyze the optical properties of a material, we must first obtain its dielectric function, which consists of both real and imaginary parts (ε1(ω), ε2(ω)) [30]. It is the linear response of a system to electromagnetic waves. This function is calculated by considering the transition between the occupied and unoccupied states in the energy band or the transition between HOMO and LUMO states. The imaginary and real parts of the dielectric function are defined as [31]: where P represents the Cauchy integral part. and are the electronic states at the conduction band and the valence band, respectively. In Fig. 7a and b, the real and imaginary parts of the dielectric function for some selected compounds such as C16H20O7, C16H19O7(F4), C16H19O7(F6), C16H19O7(F13), C16H19O7(F16), and C16H19O7(F18) are plotted at x, y, and z directions. It is also seen from the ε1(ω) spectra that obtained values are positive for all photon energy ranges (Fig. 7a). Hence, the dielectric nature of these compounds is identified. As seen, calculated ε1x(ω) of all compounds have small variations between 2 and 4, except C16H19O7(F4), while ε1y(ω) and ε1z(ω) vary between 0 and 6. So, there is an anisotropy between three directions of the ε1(ω) spectrum corresponding to different polarizations. This polarization is due to the response of material to the electrical field of electromagnetic wave. The maximum value of polarized ε1x(ω), ε1y(ω), ε1z(ω) for the C16H20O7 and selected B category of compounds occurred at 3.1 eV. The first peaks are related to maximum values of variations ε2(ω) spectrum (Fig. 7b). For C16H20O7 and selected B category of compounds, the maximum value of ε2(ω) is related to the transitions between the O-p state on HOMO and LUMO.

Fig. 7

a The real part of dielectric spectrum in terms of energy at three directions: x, y, z. b The imaginary parts of dielectric spectrum at x, y, and z directions

For C16H19O7(F4) in A category of compounds, a maximum polarization of ε1(ω) spectra is observed at the x-direction. The first peak of ε2x(ω), ε2y(ω), and ε2z(ω) spectra is occurred around 1 eV due to the electronic transitions between the F-p state in HOMO and the O-p state in LUMO. For C16H19O7(F13) compound, the first peaks of ε1y(ω) and the ε2y(ω) spectra at energy rang of 0–3 eV appears due to the small bandgap of compound and electronic transition from valence band to the unoccupied states in the conduction band. The calculated static dielectric constants at zero energy ε1(0) are presented in Table 1.

For B category of compounds, the maximum static dielectric constant ε1(0) is observed at the y-direction, while for C16H19O7(F4) compound, the maximum value of the ε1(0) is obtained at the x-direction and it is about 6.35. It is found that the fluorination at B category of compounds does not effect on ε1(0) of compounds. As a result, a more detailed analysis showed that these compounds exhibit the same mechanism of reaction between the drug molecules and the diseases proteases.

Electron energy-loss spectroscopy (EELS) is an analytical technique that evaluates the losing energy of an electronic beam during interaction with the substance. EELS corresponded to the HOMO and LUMO transitions of electrons and their elastic scattering [30]. It is an optical spectrum [32] used to get Plasmon energy. EELS is applied to identify [33] the different (organic) compounds used in biological systems or pharmaceuticals and to detect any impurities present [34] even during the synthesis or formulation process [35]. Also, EELS is used as a technique for quantification of penetration to cell walls [36]. EELS defined as:

Calculated EELS spectra for C16H20O7, C16H19O7(F4), C16H19O7(F6), C16H19O7(F13), C16H19O7(F16), and C16H19O7(F18) compounds are shown in Fig. 8 at the x-, y-, and the z-directions. According to this figure, the prominent peaks in ELOSS spectra are located at the energy of 24 eV for pure and fluorinated compounds of B category. The maximum values of peaks are about 2.3, 2.4, and 2.2 at the x-, y-, and the z-direction, respectively. The ELOSS spectrum for B category of compounds is the same as pure (Fig. 8b). The value of main peak of C16H19O7(F4) is 1.9 at energy of 26 eV.

Fig. 8

The variation of energy electron loss spectrum L (ω) at three directions: x, y and z. a and b The comparison between C16H20O7 and the other compounds with the small bandgap

Absorption spectroscopy [37] is defined as the measurement of the interaction between electromagnetic radiation, and a substance is another important analytical chemistry technique to ensure the safety, quality, and efficacy of compounds used in the pharmaceutical industry. Especially, ultraviolet/visible (UV/Vis) light spectroscopy can be used in the manufacturing process to identify contaminants within a substance or measure the kinetics of a reaction. In this study, we have investigated the absorbance spectra of pure and fluorinated C16H20O7 compounds, as in the following:

Figure 9 displays the absorbance spectra of the pure and selected fluorinated compounds at the x-, y-, and the z-directions. The maximum absorption is in the wavelength range of 50 to 125 nm that is at the extreme ultraviolet radiation (EUV) region. As shown, all compounds have the considerably same behavior in these regions. In ultraviolet (UV) regions in the wavelength range of 200 nm to 400 nm, the absorption coefficients are less than the EUV region (inset panels of Fig. 4). Also, there are some slight differences among the absorbance spectra of compounds. High similarity in absorbance spectrum of fluorinated compounds with pure is crucial from the viewpoints of pharmaceutical usages.

Fig. 9

Calculated absorption spectra at x, y, z directions. The inset plots are the absorption spectra in terms of wavelength at ultraviolet regions. The absorption spectra are in the unit of 104 cm.−1

One of the most known physicochemical properties of substances is the refractive index [38]. The refractive index, that means the ratio of the speed of light passing in the vacuum to its passing through a substance, is complex quantity where n(ω) is the refractive index for the real part and k(ω) is the extinction coefficient for the imaginary part. We have calculated as follows:

Refractive index is mainly applied to identify a particular substance, confirm its purity, or measure its concentration. Generally, it is used to measure the concentration of a solute in an aqueous solution. So, it has the most applications to the determination of drug concentration in the pharmaceutical industry.

We have presented the refractive index and extinction coefficient spectra in terms of photon energy for C16H20O7 compound, as shown in Fig. 10a and b. The calculated spectra exhibit some anisotropic behavior at the energy range 0 to 7.5 eV. For high-energy photons (higher than 16 eV), is lesser than unity which means that the nature of the compound changes from linear to non-linear, and the superluminal effect occurs. The inset graphs are related to comparing of refraction index and extinction coefficient spectra among selected compounds at energy range 0 to 7 eV. Generally, it is shown that the refractive index and extinction coefficient spectra have the same behaviors for the pure and B category of compounds, except C16H19O7(F13) compound at the y-direction. The calculated static refractive index at the zero energy is given in Table 1. According to Table 1, for the pure C16H20O7 and B category of compounds, the maximum value of the static refractive index is about 1.7 at the y-direction. The highest value of static refractive index is related to the C16H19O7(F4) compound at the x-direction.

Fig. 10

a and b The calculated refraction index and extinction coefficient spectra for pure C16H20O7 at x, y, z directions. The inset plots are related to the comparison of refraction index and extinction coefficient spectra among selected six compounds at the energy range 0 to 7 Ev

Finally, we have calculated the reflectivity spectra R(ω) as follows:where n and k are refractive index and extinction coefficient, respectively. As seen in Fig. 11, the main peak about 0.33 at 0.5 eV, is related to the C16H19O7(F4) compound from category A of compounds at the x-direction. Other compounds have not considerable reflectivity at the x-direction. For the other compounds, the maximum value of reflectivity of 0.3 at 3.25 eV is observed at the z-direction. For pure C16H20O7 compound, large optical reflectivity anisotropy is observed at 6.5 eV in the x-, y-, z-directions. On the other hand, the behaviors of reflectivity spectra for the pure and the other B category of compounds, except C16H19O7(F13), are quite similar at the y-direction.

Fig. 11

The calculated reflectivity spectra at x, y, and z directions

In summary, in comparison with the experimental results, there is not the available experimental evidence regarding the obtained results. However, there are fewer studies for absorption spectra by Hassan like Altersolanol A and Altersolanol J compounds [39] providing stronger evidence that it makes these materials suitable for pharmaceutical applications. In Fig. 12, we have compared obtained absorption spectrum for Altersolanol L with the experimental results for the same family of compounds for Altersolanol A and Altersolanol J compounds [39]. There is a good agreement between obtained absorption spectra at 224 and 275 nm with the other experimental results.

Fig. 12

Comparison of calculated absorption spectrum Altersolanol L at the x-direction (in unit of 10–4/Cm) with the experimental absorption spectra of Altersolanol A and Altersolanol J (without unit) [39]

B. Prediction of drug-likeness

To predict drug-likeness of designed compounds, Molsoft software was used to estimate the pharmacological similarity of fluorinated derivatives with the five famous Lipinski’s rules [40]. Lipinski’s rule of five: As per Lipinski’s rule of five [40], an orally active drug has (i) not more than five hydrogen bond donors (OH and NH groups), (ii) not more than ten hydrogen bond acceptors, (iii) a molecular weight under 500, and (iv) a partition coefficient log P under 5.

The capacity of the hydrogen bond is an essential factor in the permeability of the drug to the body [41]. Drug permeation or absorption in the body decreases with increasing the number of hydrogen bond donors to more than five and the number of hydrogen bond acceptors to more than 10. The designed compounds have 7 H-bond donors and 5 to 7 H-bond acceptors (Table-2). Lipophilicity and partition coefficient (log P) values are an indicator of the tendency of the compound to dissolve in the membrane. The values describe the uptake, distribution, biotransformation, and excretion of organic chemicals in biological systems [42]. High log P values represent high solubility and good penetration of lipid membranes but low solubility in water and cytoplasm. Log P values for C16H20O7 and its fluorinated derivatives are approximately − 4 (Table 2) confirming higher affinity to the aqueous phase.

Blood–brain barrier (BBB) is a physicochemical barrier that maintains homeostasis of the central nervous system (CNS) to regulate the content of the interstitial fluid in the brain [42]. The blood–brain barrier prediction score is greater than four indicates the chemical pass the BBB. This score was calculated in Molsoft server by Gupta et al. [43]. In this study, the BBB scores for C16H20O7 and its fluorinated derivatives are 2.3 to 2.4 (Table 2) confirming them as unpenetrated chemicals to the brain.

The aqueous solubility of compounds is significant to oral absorption and distribution characteristics. All designed compounds have aqueous solubilities of MolLogS from − 0.7 to − 0.5. Based on ADME prediction, the C16H19 O7(F14), C16H19 O7(F15), and C16H19O7(F16) have the best favorite drug-likeness rather than C16H20O7. In them, fluorine is substituted with hydrogen in the methyl group (C16). To get more optimal and reliable results from the docking process, first, the 3D structure of HTLV-1 protease (PDB codes: 3LIN) and HIV protease (PDB code: 1YT9) were optimized by Gromacs in 10 ns. The RMSD results (root mean square deviation) and total energy showed that the protein structure during running time is stable (Fig. 13). HTLV-1 and HIV have a homodimer activity structurally, and each monomer of these enzymes has 125 and 99 amino acids, respectively. The active site of these proteases is located in a valley between two monomers ASP25, and ASP32 in HIV and HTLV-1 protease, respectively, in each monomer act as the main catalytic amino acids [44]. The sequence alignment evidence shows 28% identity between the primary sequence of HTLV-1 protease and HIV protease enzyme, while they are more conserved in the active site (45%) [45].

Fig. 13

RMSD and total energy for HTLV protease during 10 ns molecular dynamics simulations. a Root mean square deviations (RMSDs) of HTLV-1 protease (purple) and HIV protease (light purple). b Total energy of HTLV-1 protease (light purple) and HIV protease (purple). Both parameters confirm structural stability of protease during simulation

Ruker et al. [46] investigated similarities and differences in HTLV and HIV protease-ligand interactions on the molecular level. They showed the main protease-ligand interactions happen in similar regions of proteases: the N-terminal (HIV-PR R8; HTLV-PR R10), the active site (HIV-PR L23-D30; HTLV-PR L30-D36), the flap region (HIV-PR M46-I50; HTLV-PR S55-A59), and the lateral loops of the ligand-binding site (HIV-PR P81-I84; HTLV-PR N97-I100). Also, they showed, based on interaction energy profile, however, some well-conserved residues have strong interactions with the substrate in two proteases (R8/R10, A28/A35, D29/D36, G49, G58, and I84/I100), there are many not-conserved residues have strong interactions in the similar ligand binding pockets in HIV and HTLV protease (D30/M37, I47/V56, G48/L57, I50/A59, and V82/W98).

The C16H20O7 and fluorinated compounds interacted with residues in the N-terminal of HIV and HTLV protease that are listed in Tables 3 and 4 in detail. In designed fluorinated compounds, C16H19O7(F14), C16H19O7(F15), and C16H19O7(F16) have the best drug-likeness other than else. The C16H19O7(F16) interacted with ARG8A (4), GLN7A (3), and LEU10A via hydrogen bonds in complex with HIV protease and C16H19O7 (F14) connected to ARG10B, GLU28B (2), and ALA99B in HTLV protease. Nowadays, the world has been confronted the novel coronavirus (2019-nCOV)–dependent acute respiratory illness. It seems that imbalanced immune cells’ activity can lead to the self-destruction of the host, resulting in high severity of COVID-19. Lim et al. [47] showed that a load of the virus decreases significantly in Korean patients diagnosed with COVID-19 when treated with lopinavir and ritonavir as two potent inhibitors of HIV protease. Dipti Mothay et al. [48] analyzed some potent inhibitors of COVID-19 protease that they are able to treat the HIV patients. Based on in silico study, Arg141of protease model of COVID-19 had molecular interactions with lopinavir, ketoamide, and ritonavir; Gln175 showed molecular interactions with nelfinavir and Thr75 and His176 from COVID-19 protease bind to remdesivir via noncovalent interactions. Günther et al. [49] introduced an active site and two allosteric binding sites for potent inhibitors of COVID-19 protease: The hydrophobic pocket is created by Ile213, Leu253, Gln256, Val297, and Cys300 within the C-terminal dimerization domain; the deep groove is located between the catalytic domains (Cys145 and His 141) and the dimerization domain (Arg4, Ser10, Gly11, Glu14, Asn28, Ser139, Phe140, Ser147, Glu290, Arg298) [50]. Günther, et al. showed that AT7519 is the only compound in their screen that bound to loop 107–110, Val202, and Thr292, and the chlorinated benzene ring in AT7519known as a potent compound for human cancer treatment [49] is engaged in various van der Waals interactions to loop 107–110, Val202, and Thr292 [50, 51]. Muramatsu et al. introduced substrate-binding residues His41, Met49, Gly143, Ser144, His163, His164, Met165, Glu166, Leu167, Asp187, Arg188, Gln189, Thr190, Ala191, and Gln192 for COVID-19 protease [52]. Based on our results, C16H19O7 (F14) has a similar binding mode to COVID-19 protease with the AT7519 compound mentioned above. Two compounds are halogenated, which may effect on binding affinity to the special site of the COVID-19 protease. This site may be a target to inhibitor binding for COVID-19 protease. The C16H19O6 (F14) has hydrogen bonds and hydrophobic interactions with Arg298 in dimerization domain and loop 107–110. It may introduce as potent inhibitor to COVID-19 protease (Table 5).

Table 4

Chemical name, formula, and inhibition constant (Ki), binding energy, the number of hydrogen bond between ligand and HTLV- (HB), the names of residue that participated in hydrogen bonds

HTLV protease
Name	Formula	Ki (ℳM)	Binding energy (kcal/mol)	HB	Amino acids	Hydrophobic interactions	Amino Acids	Π-Stacking	Halogenbond
1	C16H20O7	144.77	− 5.24	6	ARG10A, VAL2A, PRO4A (2), VAL12A (2)	4	ILE3A (2), PRO4A, ASP6A
4	C16H19O7(F4)	32.53	− 6.12	4	PRO4A (2), VAL12A (2)	3	ILE3A, PRO4A, ASP6A
6	C16H19O7(F6)	77.1	− 5.61	4	ARG10B, GLU28B, ASN97B, ALA99B	3	LEU30B, ALA43B, ALA99B	TRP98B
7	C16H19O7(F7)	68.2	− 5.68	4	ARG10B, GLU28B (2), ALA99B	2	GLU28B, ALA99B
8	C16H19O7(F8)	134.12	− 5.28
9	C16H19O7(F9)	99.6	− 5.46	6	ARG10B, GLU28B (3), ASN97B, ALA99B	2	LEU30B, TRP98B
10	C16H19O7(F10)	154.6	− 5.2	8	ARG10B, GLU28B (3), LEU30B, ASN97B, ALA99B(2)	3	ILE27B, LEU30B, TRP98B
11	C16H19O7(F11)	122.06	− 5.34	8	ARG10B, GLU28B (3), LEU30B, ASN97B, ALA99B	1	LEU30B
12	C16H19O7(F12)	84.55	− 5.56	6	ARG10A, LEU57B, GLY58B, GLY60A, GLY60B, HIS66A	1	PHE67A	TRP98A	LEU57B
13	C16H19O7(F13)	83.43	− 5.56	6	GLU28B (3), ASN97B (2), ALA99B	4	27B (2), TRP98B, ALA99B
14	C16H19O7 (F14)	278.27	− 4.85	4	ARG10B, GLU28B (2), ALA99B	2	GLU28B, ALA99B
15	C16H19O7(F15)	242.2	− 4.93	3	ILE3A (2), PRO4A	6	VAL2A, PRO4A (2), ARG10A, VAL12A (2)
16	C16H19O7(F16)	79.08	− 5.6	3	ILE3A (2), PRO4A	6	VAL2A, PRO4A (2), ARG10A, VAL12A (2)
117	C16H19O7(F17)	392.96	− 4.65	5	ARG10A VAL2A, PRO4A (2), ARG10A, VAL12A	2	ILE3A, ASP6A
118	C16H19O7(F18)	306.29	− 4.79	6	VAL2A, PRO4A (2), ARG10A, VAL12A (2)	2	ILE3A, ASP6A
119	C16H19O7(F19)	102.08	− 5.44	5	PRO4A (2), ARG10A, VAL12A (2)	2	ILE3A, ASP6A

Table 5

Chemical name, formula, and inhibition constant (Ki), binding energy, the number of hydrogen bond between ligand and COVID-19 (HB), the names of residue that participated in hydrogen bonds

COVID-19 protease
Name	Formula	Ki (ℳM)	Binding Energy (kcal/ol)	HB	Amino Acids	Hydrophobic interactions	Amino acids	Π-Stacking	Alogen bod
1	C16H20O7	144.77	− 5.24	9	ARG4A (2), VAL2A, LYS5A(3), TRP207, LEU282, GLU288 (2)	3	TRP207, LEU282, GLU288
4	C16H19O6 (F4)	32.53	− 6.12	5	GLN110, THR111, ILE152, ASP153, ASP295	3	ASN151, VAL303ASP153
6	C16H19O6 (F6)	77.1	− 5.61	6	PHE66(2), VAL68, ARG76, VAL77(2)	2	HIS64, ARG76
7	C16H19O6 (F7)	68.2	− 5.68	6	PHE66, VAL68, LEU75, ARG76, VAL77(2)	2	HIS64, GLN74
8	C16H19O6(F8)	134.12	− 5.28	9	TYR37(2), GLN83, LYS88, TYR101(3), PHE103(2)	4	TYR101, LYS102, PHE103(2)
9	C16H19O6 (F9)	99.6	− 5.46	6	PHE66(2), VAL68, ARG76, VAL77(2)	2	HIS64, GLN74
10	C16H19O6 (F10)	154.6	− 5.2	5	AASN63, PHE66(2), ARG76, VAL77	3	IASN63, HIS64, GLN74
11	C16H19O6 (F11)	122.06	− 5.34	6	PHE66(3), VAL68, ARG76, VAL77	2	HIS64, LEU67
12	C16H19O6 (F12)	84.55	− 5.56	5	VAL68, LEU75, ARG76, VAL77(2)	3	HIS64, LEU67, ARG76
13	C16H19O6 (F13)	83.43	− 5.56	7	PHE66(2), VAL68, ARG76, VAL77(3)	1	HIS64		VAL68
14	C16H19O6 (F14)	278.27	− 4.85	7	GLN110, THR111, ASN151, ASP153(2), THR292, ARG298	2	PHE294(2)
15	C16H19O6 (F15)	242.2	− 4.93	5	VAL68, LEU75(2), ARG76, VAL77	2	GLN74, ARG76
16	C16H19O6 (F16)	79.08	− 5.6	5	VAL68, LEU75, ARG76, VAL77 3A (2), PRO4A	2	GLN74, ARG76
17	C16H19O6 (F17)	392.96	− 4.65	4	PHE305(3), ARG298	4	PRO9, ARG298, THR304, PHE305
18	C16H19O6 (F18)	306.29	− 4.79	6	PHE66(2), VAL68, ARG76, VAL77(2)	2	HIS64, LEU67
19	C16H19O6 (F19)	102.08	− 5.44	6	PHE66(2), VAL68, ARG76, VAL77(2)	2	HIS64, LEU67

Conclusions

First principle calculations based on DFT and molecular docking approaches are used for investigation of optoelectronic and biological properties of pure C16H20O7 and substituted by fluorine. All compounds are classified into one of two groups: category A or category B. The electronic results show that the pure C16H20O7 compound has a direct bandgap about 3.1 eV. The substitution of F atom at A category decreases the electronic bandgap, while the bandgap value of B category compounds is very similar to the pure one except for C16H19O7(F13) compound. The wide bandgap nature of pure C16H20O7 and fluorinated derivatives show that these materials can be used in biosensors applications. F-O bonds in the B category of compounds such as C16H19O7(F14), C16H19O7(F16), and C16H19O7(F17), have a considerable contribution on their stability and chemical reactivity. It is shown that fluorination at B category of compounds does not effect on intrinsic static dielectric constant of compounds. So, these compounds have the same nature in the reaction between the drug molecules and the diseases proteases. The anisotropic optical properties are observed in three directions. Because of the obtained optical spectra, the B category of compounds are predicted to be useful for different pharmaceutical usages in Covid-19, HIV, and HTLV proteases. Based on ADME prediction, the C16H19 O7(F14), C16H19 O7(F15), and C16H19O7(F16) have the best favorite drug-likeness rather than pure C16H20O7. The C16H19O6 (F14) has hydrogen bonds and hydrophobic interactions with Arg298 in dimerization domain and loop 107–110. It may introduce as potent inhibitor to COVID-19 protease.