Introduction

Cycloaddition reactions represent a valuable synthetic tool for the formation of carbo- and heterocyclic structures, often with high regio- and stereo-control [1]. Cycloaddition reactions, such as [4+2] Diels-Alder cycloaddition [2,3,4,5], [2+1] cycloaddition [6,7,8,9], and [3+2] cycloaddition [10,11,12,13,14,15,16], considered powerful tools for the construction of six-, three-, and five-membered rings, respectively, are heavily studied, and the literature contains numerous theoretical and experimental studies on these types of reactions [17,18,19,20,21,22,23,24,25]. On the contrary, there is comparatively much less work on the formation of seven-membered rings. Seven-membered rings are structural units widely present in a wide range of natural biologically active compounds. These chemicals can be found, for example, in rubriflordilactone A [26], phomarol [27], and cortistatin A [28]. Given the increase in natural compounds containing seven-membered rings which have been isolated with biological activity, it has become necessary to find and develop synthesis of these molecules utilizing successful synthetic methods, among these strategies are metathesis, cyclizations, ring isomerization, and cycloadditions [29], but synthesis of these rings is difficult and challenging, due to unfavorable entropy in cyclization processes [30]. Cycloaddition reactions represent an efficient and direct protocol for accessing seven-membered rings, primarily the [5+2] cycloaddition (52CA) reaction [31, 32] and the [4+3] cycloaddition reaction [33, 34]. However, due to the variety of forms, the final scenario is preferred to the cycloaddition [4+3] for the creation of rings with seven chains 52CA. There are several types of 52CAs reactions based on the five-membered component involved in the reaction [35]. The 52CA reactions between an oxidopyrylium and alkenes allow for the simultaneous generation of bridged bicyclic ethers and the seven-membered ring compounds, which are a structural basis present in biologically active natural products [36]. Most commonly, the oxidopyrylium cycloaddition reaction begins with the formation of an oxidopyrylium zwitterionic intermediate through the removal of the acetoxy group from the acetoxypyrone precursor to the oxidopyrylium in the presence of alkenes [37] (Scheme 1). The formation of oxidopyrylium could also be obtained from acetoxypyrone using a base [38].

Scheme 1
scheme 1

The oxidopyrylium cycloaddition reaction

Full size image

There has not been much theoretical study of the [5+2] cycloaddition reaction of oxidopyrylium with alkenes [39,40,41], compared to the [4+2] and [3+2] cycloaddition reactions, which are widely studied by various theoretical models of chemical reactivity [42,43,44,45].

One of the most widely used and robust reactivity models in recent years is Domingo’s so-called MEDT theory. In 2016, Domingo introduced the molecular electron density theory (MEDT), a novel theoretical framework for describing reactivity in organic chemistry [46], regarding organic molecules’ reactivity is related to changes in electron density, in complete disagreement with all models based on molecular orbital analysis, such as boundary molecular orbital theory [47, 48]. In addition to investigating and characterizing the energetically relevant reaction pathways associated with the process under study, the reactivity indices produced from the CDFT, topological analysis methods ELF (electron localization function) [49], QTAIM [50], and NCI [51], are employed in the current study to examine the reactivity inside the MEDT, we are particularly interested in the [5+2] cycloaddition of oxidopyrylium with an alkene (Scheme 2 and Fig. 1).

Scheme 2
scheme 2

[5+2] cycloaddition of oxidopyrylium with an alkene

Full size image
Fig. 1
figure 1

[5+2] cycloaddition of oxidopyrylium with an alkene

Full size image

Computation methods

The optimization of the geometry of reactants, products, and transition states was performed by DFT computations applying the ωB97XD functional; it is an extension of the widely used B97-D functional and includes empirical dispersion corrections, which are designed to better account for van der Waals interactions in molecular systems, coincidentally with 6-311G(d,p) basis [52]. Frequency calculations were used to characterize optimized stationary positions to make sure that none of the reactants or products had any imaginary frequencies, and that the transition states have just one imaginary frequency. Intrinsic reaction coordinate (IRC) [53] paths were plotted in both directions, with the aim of verifying the energy profiles linking each transition stage to its two corresponding minimum. By employing the Tomasi group polarizable continuum model (PCM) to re-optimize the stationary points obtained in the gas phase, the solvent impact of dichloromethane was implicitly taken into account [54].

The global electrophilicity index was calculated using the equation shown below [55]:

$$\omega =\frac{{\mu }^{2}}{2\eta }$$

wherein the electronic chemical potential and chemical hardness are represented by μ and η, respectively, the two quantities µ and η according to the aforementioned formulas, they were calculated utilizing the HOMO and LUMO values [55,56,57,58]:

$$\mu =\frac{{E}_{HOMO}+{E}_{LUMO}}{2}$$

and

$$\eta ={E}_{LUMO}-{E}_{HOMO}$$

The overall nucleophilicity index N was calculated according to the following formula:

$$N={E}_{HO}\left(Nu\right)-{E}_{HO}(TCE)$$

with EHO(Nu) corresponds to the HOMO energy of the reagent and tetracyanoethylene’s HOMO energy is known as EHO(TCE) [58].

All reactivity indices were evaluated utilizing B3LYP/6-31G(d) level in accordance with Domingo’s suggestions [59, 60].

The total of the naturally found atomic charges (q), as determined by a natural population analysis (NPA), was used to calculate the overall GEDT [61] electron density transfer of the atoms belonging to each reactant (f) at the transition state \(GEDT=\sum {q}_{f}\). The Topmod program has been employed to undertake a topological inquiry of the ELF electronic localization function [62]. The GaussView software was utilized for visualizing the locations of the ELF pool attractors [63], while the VMD program was applied to illustrate the ELF pool isosurfaces [64] with an isovalue of 0.82.

Result and even discussion

Comparison of the 3-OXP 2 and EthE ground-state electronic structures’ global and local reactivity indices obtained via CFDT.

Table 1 lists the computed global reactivity indices for 3-OXP and EthE, including electronic chemical potential (μ), chemical hardness (η), electrophilicity (ω), and nucleophilicity (N). The geometry of 3-OXP and EthE has been entirely optimized at this level since these indices have been scaled to the computational level of B3LYP/6-31G(d), with the goal of having complete consistency with the scale suggested by Domingo [59, 60].

Table 1 Derived for 3-OXP and EthE are the global electronic qualities (global electrophilicity (ω), global nucleophilicity (N), global electrochemical potential (μ), and chemical hardness (η)) calculated by B3LYP/6–31(d)
Full size table

According to an analysis of the data in Table 1, EthE (− 2.39 eV) has a lower electronic chemical potential (ECP) than 3-OXP (− 3.75 eV), which indicates that during a polar reaction, the electron density will transfer from EthE, which acts as a nucleophile, to 3-OXP, which acts as an electrophile. Domingo divided nucleophiles into three categories: strong, moderate, and marginal, with respective nucleophilicity indices of 3.00, 2.00–3.00, and less than 2.00 eV. Similar to this, electrophiles are likewise divided into three groups: strong, moderate, and marginal, depending on their electrophilicity indices, which range from 1.50 to 0.80 eV, 1.50 to 0.80 eV, and 0.80 eV and lower, respectively. This scale classifies 3-OXP as both a strong electrophile with a value of 2.09 eV and a strong nucleophile with a value of 4.09 eV. EthE is categorized as both a very poor electrophile with a value of 0.41 eV and a strong nucleophile with a value of 3.64 eV.

Regarding local reactivity, Domingo et al. suggested two new functions called electrophilic and nucleophilic Parr’s functions, to pinpoint among the most nucleophilic and electrophilic centers in the species. According to the examination of the nucleophilic Parr functions of EthE, the carbon C1, \({P}_{C1}^{-}=0.58\), is the most nucleophilic center, whereas the 3-OXP’s electrophilic Parr functions reveal that the C4 carbon is this molecule’s highest electrophilic center, with a \({P}_{C4}^{+}=0.52\) value, and is three times larger than the C6 carbon, with a \({P}_{C6}^{+}=0.17\) value (Fig. 2).

Fig. 2
figure 2

The electrophilic Parr function \({P}_{k}^{+}\) of 3-OXP and the nucleophilic Parr function \({P}_{k}^{-}\) of EthE

Full size image

The analysis of Parr functions leads to the conclusion that the C1 carbon of EthE and the C4 carbon of 3-OXP will have the best electrophile–nucleophile interaction along the nucleophilic attack of EthE on 3-OXP, suggesting that this reaction will primarily proceed via the C1-C4 regioisomeric pathway in clear agreement with experimental observations.

ELF enquiry of reagents 3-OXP and EthE

Positions of the valence attractors with their populations, localization domain, and suggested Lewis structures for the two reactants 3-OXP and EthE are shown in Fig. 3.

Fig. 3
figure 3

Positions of the 3-OXP and EthE reactants’ valence attractors, as well as their respective populations, localization domains, and Lewis structures

Full size image

Figure 3 shows that the oxygen atom of 3-OXP has two monosynaptic pools, V(O1) and V′(O1), with a combined population of 5.59 electrons. Additionally, the oxygen atom of O5 has a monosynaptic pool, V(O5), with a population of 3.67 electrons, which is indicative of two free electron doublets carried by the oxygen O5.

The presence of two disynaptic pools with electron populations of 1.88e and 1.78e, V(O5,C4) and V(O5,C6), respectively, values close to 2e, signifying an O5-C4, O5-C6 single bond, as well as a disynaptic pool V(O,C3) with a population of 2.04e, indicates an O-C3 single bond. In addition, the presence of four disynaptic pools, with electron populations of 2.54, 2.96, 2.96, and 3.27e values greater than 2e and less than 4e, suggests a single bond with a double bond character; on the other hand, the ELF analysis of EthE reveals the presence of two disynaptic pools, V(C1,C2) and V′(C1,C2), with a combined population of 3.71e, suggesting a remarkable character of a double bond. Additionally, we observe that the oxygen atom of EthE has two monosynaptic pools, V(O) and V′(O), with a combined population of 4.73e, which may be bound to two pairs of this atom’s non-bonding electrons.

The disynaptic pool V(C2,O), which has a population of 1.46e but is unexpected compared to the expected value of 2e, is present in the C2-O binding region. This depopulation of the C2-O binding region in the direction of the oxygen atom can also be attributed to the disynaptic pool, which has a population of 1.28e.

Energetics study of the 52CA reaction between 3-OXP and EthE

Due to the non-symmetry of the two reactants, there are four possible isomeric paths for the 52CA reaction of 3-OXP with EthE: two regioisomeric pathways, each of which produces two stereoisomers.

The C1-C4 single bond creation pathway and the C1-C6 single bond initiation pathway are a pair of regioisomeric pathways, whereas stereoisomeric pathways named exo and endo depending on the position between the methyl group affixed to the C4 carbon of 3-OXP and the ethoxy group of EthE, such that along the exo pathway, the two ethoxy and methyl groups are on opposite sides. This 52CA reaction exhibits a one-step mechanism, i.e., on the reaction’s surface potential energy, a single TS, TS-1x, TS-1n, TS-2x, and TS-2n, as well as their associated cycloadducts, P-1x, P-1n, P-2x, and P-2n, have been identified and characterized (Scheme 3).

Scheme 3
scheme 3

Theoretical potential regio- and stereoisomeric reaction paths for the 52CA reaction involving 3-OXP and EthE

Full size image

Table 2 lists the relative electron energies for the species involved in the reaction of 52CA between 3-OXP and EthE, both in the gas phase plus when dichloromethane is present, while the detailed calculation in gas and DCM are presented in Tables S1 and S2.

Table 2 ωB97XD/6-311G(d,p) relative electronic energies (in kcal·mol−1), in the gas phase and in the presence of dichloromethane, for the species involved in the 52CA reaction between 3-OXP and EthE
Full size table

The four competitive routes have gas phase activation energies of 2.1 (TS-1x), 5.0 (TS-1n), 3.9 (TS-2x), and 6.6 (TS-2n) kcal/mol. The results presented here indicate that the 52CA reaction is highly stereoselective and slightly regioselective because TS-1x has a lower energy than TS-2x of 1.8 kcal/mol and TS-1x has a lower energy than TS-1n of 2.9 kcal/mol. This reaction can also be regarded as irreversible because of the strong exothermic between 47.6 and 48.5 kcal/mol.

Counting dichloromethane as a solvent, the activation energies increase to 3.3 (TS-1x), 6.7 (TS-1n), 6.8 (TS-2x), and 9.0 (TS-2n) kcal/mol, due to a strong solvation of reactants with respect to transition states. The gas phase’s regioselectivity does not alter, but has improved, since the most favorable TS-1x has an energy 3.5 kcal/mol inferior than TS-2x, as well as the stereoselectivity increases, since TS-1x is more stable 3.4 kcal/mol than TS-1n.

Figure 4 illustrates the geometries of the transition states involved in the 52CA reaction between 3-OXP and EthE.

Fig. 4
figure 4

The optimized transition state geometries connected to the 52CA reaction between 3-OXP and EthE. The units of measurement are angstroms. In brackets, the dichloromethane values are listed

Full size image

The four transition states correspond to asynchronous single bond formation processes, according to an examination of the geometries of the transition states involved in the reaction of 52CA between 3-OXP and EthE. More progress has been made in the formation of the single bond involving the C1 carbon of ETHE than the C2 carbon. Compared to C2-C6, the formation of the C1-C4 bond is further along. Dichloromethane’s two formed bonds’ different lengths provide as evidence that these transition stages have grown increasingly asynchronous. The most asynchronous transition stage, TS-1x, is also the most advantageous.

By computing the GEDT values of the four transition states, it was then possible to determine whether the 52CA reaction under investigation was polar or nonpolar. Nonpolar processes are reactions with GEDT values less than 0.0e, while polar processes are reactions with GEDT values higher than 0.2e. The values of GEDT from EthE to 3-OXP are 0.18e at TS-1x, 0.12e at TS-1n, 0.08 at TS-2n, and 0.10 at TS-2x, demonstrating the weak polarity of this 52CA reaction.

ELF analysis to ascertain the 52CA reaction’s chemical mechanism

Through performing an ELF topological analysis on a few chosen spots of the IRC of the most advantageous transition state, TS-1n, it was possible to obtain insight into the chemical mechanism of the 52CA reaction involving 3-OXP and EthE.

This ELF analysis will allow us to detect successive changes in the selected points’ electron densities. The IRC profile of TS-1n consisted of a total of 250 points in the forward and reverse directions. The positions of ELF valence attractors along with their corresponding populations of relevant structures are shown in Fig. 5.

Fig. 5
figure 5

Positions of ELF valence attractors with their populations of relevant structures

Full size image

Despite the modest variations in their electron populations, the ELF valence patterns of the interacting fragments at the initial point P1—where they are significantly separated from one another—resemble those of the two separate reactants 3-OXP and EthE. The depopulation of the disynaptic basin V′(C1,C2) begins at position P2. A new monosynaptic basin V(C4) with an initial electron population of 0.31e has developed at point P3. Point P4 sees the formation of a novel monosynaptic pool, V(C1), with an initial population of 0.24e. It is confirmed that the C1-C4 bond forms before the C2-C6 bond at point P5 due to the disappearance of the two monosynaptic basins V(C1) and V(C4) and the emergence of a new disynaptic basin V(C1,C4) with an initial population of 1.23e. A new monosynaptic pool V(C6) on carbon C6 with an initial electronic population of 0.50e is formed at point P6, while a monosynaptic pool V(C2) on carbon C2 with an initial electronic population of 0.11e is formed at point P7. Due to the depopulation of V(O5) at point P8, a new monosynaptic pool V′(O5) with an initial population of 1.87e is formed on the oxygen atom O5. A second disynaptic pool, V′(C7,C8), formed at point P9 with an initial electronic population of 1.59e, demonstrating the production of the C7 = C8 double bond. The two monosynaptic pools V(C2) and V(C6) fuse at point P10, resulting in the formation of the C2-C6 single bond and a disynaptic pool V(C2,C6) with an initial population of 1.63e. According to these findings, the 52CA reaction between 3-OXP and EthE proceeds via a one-step, two-phase mechanism, in which the two new single bonds are formed between the two reactants in an uncoordinated environment and the formation of the C1-C4 bond is advanced in comparison to the formation of the C2-C6 bond.

Molecular docking

As a powerful tool for assisting the development of potential medications for a wide range of ailments, computer-aided drug design (CADD) based on ligand and structure-based processes such as 3D-QSAR, pharmacophore, molecular docking, and ADMET has evolved [65]. The retrovirus family, which includes the human immunodeficiency virus (HIV), which causes acquired immunodeficiency syndrome (AIDS), assaults immune system cells by destroying or weakening their functioning [66]. It is still crucial to find new medications with antiviral properties, as HIV mutates rapidly, leading to treatment ineffectiveness and the development of drug-resistant strains. Protein Data Bank (http://www.rcsb.org) was used to retrieve the target HIV-1 protease protein’s three-dimensional (3D) crystal structure prior to docking. To examine substituted (1–20) oxabicyclo[3.2.1]octene derivatives’ interactions with the HIV-1 protease (PDB ID: 1HSG) protein, a molecular docking simulation was conducted using the AutoDock 4.2.6 program and the graphical interface AutoDockTools (ADT) version 1.5.6 [67]. Prior to the docking process, the 1HSG protein’s water molecules and polar hydrogen atoms were removed. Next, Kollman and Gasteiger atom charges were added using ADT. The central grid box is approximately (13.073, 22.467, and 5.557) based on the ligand position in the protein, and the grid maps were constructed to 60 in the X, Y, and Z dimensions. Software programs called PyMOL and Discovery Studio Visualizer were employed to demonstrate the potential binding processes of the oxabicyclo[3.2.1]octene derivatives with their target proteins and visually check the docked molecule and its H-bond interactions in an effort to explain their anti-HIV-1 effects, respectively [68, 69].

Figures S1 and S2 depict the results obtained on docking all the (1–20) oxabicyclo[3.2.1]octene derivatives with target 1HSG protein based on docking studies, while Fig. 6 depicts docking result of compound 5. All the ligands interacted with key active site residues of the HIV-1 protease such as ASP29, ASP30, ILE50, GLY48, and ARG8. The docked molecules contain conventional hydrogen bonds that are mainly formed with oxygen atoms of oxabicyclo[3.2.1]octene and N atoms of C≡N group. The docking variables, which include intermolecular energy, inhibition constant, and binding energy of the (1–20) oxabicyclo[3.2.1]octene derivatives with the 1HSG targeted protein, were depicted in Table S3. The binding energies for all the substituted (1–20) oxabicyclo[3.2.1]octene derivatives with the HIV-1 protease (PDB ID: 1HSG) protein exhibited dock values between − 5.01 and − 8.73 kcal/mol. The order of binding energies in relation to the drug activeness against 1HSG is 15 > 2 > 11 > 17 > 14 > 19 > 12 > 7 > 20 > 13 > 8 > 16 > 6 = 18 > 10 > 5 > 9 > 3 > 1 > 4. Compound 15 formed a strong conventional hydrogen bond with ARG8 (2.1 Å) and one van der Waals with ASP29 (4.1 Å) of 1HSG protein. The obtained results indicate that the (1–20) oxabicyclo[3.2.1]octene derivatives have high potential to act as anti-HIV agents. The docking findings could influence the design and development of novel AIDS medication candidates.

Fig. 6
figure 6

2D and 3D diagrams of the (15) oxabicyclo[3.2.1]octene compounds docked in the anti-HIV-1 protein’s binding site

Full size image

Computational pharmacokinetic analysis (physicochemical and ADME properties) of the (1–20) oxabicyclo[3.2.1]octene derivatives

The two main reasons why many drug candidates still fail to become drugs are the lack of efficacy and safety, which means that at every step of drug discovery and development, the properties of chemicals that affect absorption, distribution, metabolism, excretion, and toxicity (ADMET) are essential. The ADME analysis measures absorption, distribution, metabolism, and excretion; an online software tool was employed to examine the physicochemical, lipophilicity, water solubility, and pharmacokinetics, drug-likeness plus medicinal chemistry properties of the (1–20) oxabicyclo[3.2.1]octene derivatives. Using an online program called SwissADME, which is accessible at http://www.swissadme.ch, the ADME forecasts were made [70].

The “Rule of Five (Ro5)” or Lipinski’s rule of five (5), developed by Christopher A. Lipinski in 1997 [71], is the first and best-known rule-based methodology to use when identifying chemicals as prospective therapeutic candidates. According to Lipinski’s rule, active molecules should have a molecular weight (MW) of less than 500 g/mol, an octanol/water partition coefficient (iLOGP) of less than 5, a number of hydrogen bond acceptors (NHBA) of more than 10, a number of hydrogen bond donors (NHBD) of less than 5, and a topological polar surface area (TPSA) of more than 130 g/mol.

According to the rule of five, a compound cannot be orally active when two or more of Lipinski’s rules of five are violated. All the (1–20) oxabicyclo[3.2.1]octene derivatives have no transgressions of the bioavailability score and Lipinski’s rule of all the ligands showed a score of 55% (Table S4).

The bioavailability radar is employed to provide a preliminary assessment of a molecule’s drug-likeness. The pink area must completely enclose the red line of the tested substance, indicating their drug-likelihood with a better bioavailability profile. The (1–20) oxabicyclo[3.2.1]octene derivatives are predicted orally bioavailable in Fig. S3. Six physicochemical indices, including lipophilicity (XLOGP3), size, polarity, solubility, flexibility, and saturation, were shown on the axis by the bioavailability radars. The two-dimensional bioavailability radar graphs of 15 compound show that the disadvantageous physicochemical property the deviation outside the pink area can be used to determine factors like saturation.

Solubility is an important property in the search for oral administration medications. The amount of water that may be dissolved is indicated in log (mol/l) (insoluble = 10, poorly soluble = 6, moderately soluble = 4, very soluble = 2, highly soluble = 0). The outcomes of the logS values of all oxabicyclo[3.2.1]octene derivatives in Table S2 show that they are very soluble and soluble in water except 15 compound. Thus, all the oxabicyclo[3.2.1]octene derivatives have good solubility in water and could facilitate well oral adsorption.

All the (1–20) oxabicyclo[3.2.1]octene derivatives have high gastrointestinal (GI) absorption except 3 compound and non-substrate to P-glycoprotein. Figure 4 illustrates the relationship between TPSA and LogP used in the “Boiled-Egg” model to estimate brain penetration and gastrointestinal absorption of the chosen compounds. The graphic shows that the chemicals 2, 5, 7, 8, 10, 13, 14, 16, and 18 are expected to pass through the blood–brain barrier (BBB) in the yellow zone (yolk), while the other oxabicyclo[3.2.1]octene derivatives show a negative response for BBB (in the white).

A molecule will hardly pass through the skin if skin permeability (Log Kp), an important component to take into account for enhancing medicinal efficacy, is greater than 2.5 cm/h. According to Table S2, the Log Kp for oxabicyclo[3.2.1]octene derivatives ranges from 4.55 to 8.04 cm/h (2.5). Therefore, it is reasonable to anticipate that all octene derivatives will effectively permeate skin. Additionally, for all derivatives of oxabicyclo[3.2.1]octenes, synthetic accessibility (SA) values less than 10 indicate that the compounds similar to drugs are relatively simple to synthesize.

Conclusion

The [5+2] cycloaddition reaction between oxidopyrylium 3-OXP and ethoxyethylene EthE yields as products the bridged cyclic ethers and associated seven-membered ring. Theoretical investigation into P-1x, P-1n, P-2x, and P-2n has been conducted within the MEDT framework at the computational level DFT/ωB97XD/6-311G(d,p). This 52CA reaction can proceed along 4 isomeric reaction paths and exhibits good regio- and stereoselectivity which improved greatly with the addition of the solvent. An examination of the transition state geometries for the exo and endo stereoisomers indicates that the process of formation of single bonds C1–C4 and C2–C6 and asynchronous, also the asynchronicity increases when the solvent effect of dichloromethane was taken into account. Analysis of the GEDT at the transition points to demonstrate this reaction’s weak polarity. According to an ELF topological study of specific locations on the IRC profile of the most advantageous transition state TS-1x’s electron density distribution, the reaction will proceed in a non-concerted, two-phase, one-step manner. The coupling of the pseudoradical centers C1 with C4 and C2 with C6 resulted in the development of the novel carbon–carbon, C1-C4 and C2-C6 single bonds.