Introduction

Hydrogen bonding is thought to play a critical role in stabilizing the structures of biological molecules such as proteins or nucleic acid [1, 2]. The participation of the amino acid in hydrogen bonding interaction is a well-recognized factor in many biologically relevant processes such as protein-nucleic acid interaction, protein-protein interaction, α-helices, and β-sheets formation. There are a number of amino acid residues that can form hydrogen bonds (HBs) via their side chains. Perhaps most notable of this category are side chains that contain hydroxyl (serine and threonine), acetic acid (aspartic and glutamic acids), amino (lysine and arginine), and thiol (cysteine) that are known to participate in HBs.

Among the molecules of biological interest, N-heterocyclic analogs of Indene (Indole, Indazole, Azaindole, Azaindazole) are structural constituents of many bioactive natural products and have gained much research pursuit in the field of medicinal chemistry. Such analogs of Indene are bicyclic heterocyclic compounds containing five-membered ring fused with six-membered ring. The hydrogen bonding interactions of N-heterocyclic analogs of Indene with amino acid side-chain mimics provide great opportunity for screening and discovery of diverse biologically active substances [3,4,5,6,7,8,9,10,11,12,13,14]. The choice of indole in the present study has been motivated as it is built in to proteins in the form of amino acid tryptophan, and its scaffold represents one of the most important structural subunits for the discovery of new drug candidates [15, 16]. Azaindole is another moiety of interest as it resembles molecules of the DNA base pair, and thus, is an important model system for the study of hydrogen bonding in DNA base pair. Indazole and Azaindazole moieties are important scaffolds which have attracted a considerable attention of medicinal chemists due to their pharmacological properties. The derivatives of Indazole and Azaindazole possess a variety of biological properties including anticancer, antibacterial, antiprotozoal, and anti-inflammatory activities [17,18,19].

The present study reports the hydrogen-bonded complexes in which the amino acid side-chain mimics are chosen as one partner and the N-heterocyclic analogs of Indene as another partner. The primary goal of the present study is to measure the relative strength of hydrogen bonding interactions such as N-H···O=C, N-H···O, N-H···S, N-H···N, O-H···N, S-H···N, and C-H···N in these complexes. Studying such interactions in these complexes is of great interest as their findings delineate the role of these HB interactions in the stability of protein-protein interfaces, protein-ligand complexes, protein-nucleic acid interactions, and DNA-drug complexes. It is worth noticing that the hydrogen bonding interaction N-H···S is ubiquitous in proteins [20,21,22,23] and its presence is manifested from the crystal structure data for β-lactam antibiotics such as penicillins, the active site of proteins like cytochrome-P450, and the iron-sulfur proteins [24, 25]. The study of structural, energetic, and topological parameters of this interaction will provide better insight into the understanding of the protein folding and biochemical reactions involving the formation and rupture of N-H···S HB.

Gas-phase quantum studies reveal that amino acid side-chain mimics coordinate to the N-heterocyclic analogs of Indene through N-H···O=C, N-H···O, N-H···S, and N-H···N hydrogen bonds which account for the significant changes in the structural, electronic, and energetic parameters of the Indene analogs. The results provided in this manuscript are helpful for experimental chemists working on protein structure or protein-nucleic acid-base interactions.

Results and discussions

Structural and energetic analysis

Hydrogen bonding interactions of N-heterocyclic analogs of Indene ̶ Indole (IN), Indazole (IND), Azaindole (AIN), and Azaindazole (AIND) (shown in Fig. 1) across the four amino acid side-chain mimics (Z), MeOH (mimic for serine and threonine), MeSH (mimic for cysteine), MeNH2 (mimic for lysine), and AcOH (mimic for aspartic and glutamic acid), have been studied. A total of thirty-two complexes of the selected heterocyclic molecules with amino acid side-chain mimics have been optimized and their geometries are shown in Fig. 2. Indene comprises of six-membered and five-membered nitrogen-containing rings fused to one another. Indene IN has one potential site, N-H, incorporated in the five-membered ring that has a tendency to act either as a proton donor or simultaneously as a proton donor and acceptor, whereas in IND, there are two sites, N-H and N embedded in the five-membered ring. Relative to IN and IND, the corresponding aza compounds, AIN and AIND, have one additional proton acceptor site N incorporated in the six-membered ring. Both AIN and AIND exist in two positional isomers, 4AIN/AIND and 7AIN/7AIND, according to the position of nitrogen atom in the six-membered ring with respect to N-H of the five-membered ring. In our manuscript, we have labeled the pyridinic nitrogen of the five-membered ring in the N-heterocyclic analogs of Indene as “N5” and pyridinic nitrogen of the six-membered ring as “N6” for the sake of convenience. The potential HB donor/acceptor site associated with AcOH as well as MeOH is O-H while it is N-H for MeNH2 and S-H for MeSH. AcOH also contains carbonyl oxygen as an additional HB acceptor site. Additionally, these amino acid side-chain mimics contain C-H that can also act as HB donor site towards the selected heterocyclics, forming C-H···N type interaction. The latter interaction serves as secondary interaction, offering additional stabilization to stronger HBs, and can perhaps further refinement to protein structure.

Fig. 1
figure 1

Optimized geometries of N-heterocyclic analogs of Indene at MP2/aug-cc-pVDZ level

Full size image
Fig. 2
figure 2figure 2

Optimized geometries of 1:1 hydrogen-bonded complexes of N-heterocyclic analogs of Indene with amino acid side-chain mimics at MP2/aug-cc-pVDZ level

Full size image

Hydrogen bonding interactions between N-heterocyclic analogs of Indene and amino acid side-chain mimics

(A) Indene-MeOH complexes

The hydrogen-bonded complexes formed between Indene analogs and MeOH have been divided in to three categories, I, II, and III, on the basis of different HB donor and acceptor sites of both the units involved in hydrogen bonding interaction with each other (shown in Fig. 2). The complexes in which the N-H of Indene analogs acts simultaneously as the proton donor and acceptor towards the OH and methyl groups of MeOH respectively lead to the formation of the cyclic five-membered hydrogen-bonded ring labeled as “I.” The second set of complexes designated as “II” exhibits cyclic six-membered hydrogen-bonded ring in which Indene analogs act as proton acceptor via its N5 site and proton donor via its N-H site to the methyl group and OH site of MeOH. Another set of complexes is marked as “III” in which the OH site of MeOH acts as proton acceptor as well as donor towards N-H and N5/N6 sites of Indene analogs respectively, resulting the formation of five-membered ring in IND-MeOH III and 4AIND-MeOH III and six-membered ring in 7AIN-MeOH III and 7AIND-MeOH III complexes. The complexes 7AIN-MeOH III and 7AIND-MeOH III use N-H of five-membered ring and N6 site of fused six-membered ring in hydrogen bonding with amino acid side-chain mimics whereas IND-MeOH III and 4AIND-MeOH III complexes utilize the N-H and N5 site of five-membered ring of Indene analogs. The former complexes are more stable in comparison with the latter complexes as there is strong hydrogen bond network in the former complexes. This supports the fact that the N6 site of the fused six-membered ring acts as a better HB acceptor in comparison with the N5 site of the five-membered ring.

Out of these three sets, the third set is more stable in comparison with the first and second set of complexes as indicated by their stabilization energies. The most interesting aspect which has been observed is that the Indene-MeOH complexes have relatively higher stabilization energies in comparison with their corresponding Indene-H2O complexes (shown in Table 1 and Table S1) as the additional methyl group induced the hydrogen bond switching, which can be ascribed to the electron donating property of the methyl group. This supports the assumption that additional carbon atom in the side-chain leads to significant changes in the drug-protein interactions and nucleobase-amino acid interactions.

Table 1 Hydrogen bond (HB) distances r (in Å) and angles θ (in °) of complexes of N-heterocyclic analogs of Indene with amino acid side-chain mimics at the MP2/aug-cc-pVDZ (L2) level; BSSE corrected stabilization energies (ΔEBSSE in kcal/mol) at the wB97XD/aug-cc-pVDZ (L1) and MP2/aug-cc-pVDZ (L2) levels; change in bond length ∆d (in Ǻ); and shifts of stretching frequencies Δυ (in cm−1) for the HB donor group (D-H) at the wB97XD/aug-cc-pVDZ (L1) level
Full size table

(B) Indene-MeSH complexes

MeSH binds to Indene analogs in three possible ways leading to the formation of three sets of complexes IV, V, and VI. The complexes in which Indene analogs bind to MeSH through single N-H···S HB are labeled as “IV.” For “V” and “VI” set of complexes, it has been observed that MeSH exhibits exactly similar orientation to that of MeOH in II and III set of complexes, respectively. It is worth noticing that the replacement of oxygen with sulfur leads to longer HBs with subsequent weakening of N-H···S HBs, resulting in 1.12–3.88 kcal/mol lower stabilization energies of Indene-MeSH complexes relative to their corresponding Indene-MeOH complexes.

(C) Indene-MeNH2 complexes

MeNH2 coordinates to Indene analogs in three ways turning out the formation of three sets of complexes labeled as “VII,” “VIII,” and “IX.” The “VII” set of complexes involves monodentate HB formation with N-H···N interaction. In these complexes, N-H of Indene analogs hydrogen bond to “N” of MeNH2 leading to the formation of the most linear HBs as depicted from HB angle values for IN-MeNH2 VII (176.25°) and 4AIN-MeNH2 VII (175.76°). The “VIII” set of complexes displays six-membered hydrogen-bonded ring structure, in which N-H and N5 of Indene analogs act as proton donor and acceptor respectively towards nitrogen and methyl group of MeNH2, leading to the formation of two HBs IndeneN-H···N and C-H···NIndene. In these complexes, departure of N-H···N HB angle from the linearity is seen to be almost 20°. The “IX” set of complexes involves cyclic five-membered hydrogen-bonded ring structure in IND-MeNH2 IX and 4AIND-MeNH2 IX and six-membered hydrogen-bonded ring structure in 7AIN-MeNH2 IX and 7AIND-MeNH2 IX complexes. In these complexes, the amino group solely acts as the proton donor as well as proton acceptor to the N5/N6 and N-H sites of the Indene analogs, respectively, resulting in the formation of HBs IndeneN-H···N and N-H···NIndene. Furthermore, the cyclic arrangement in these complexes results in more bent HBs; the deformation from the linearity lies in the range of 30° to 35° for IndeneN-H…NMeNH2 HB interaction, while for the MeNH2N-H···NIndene, it is 55° to 60°. The HBs of type N-H···N are expected to make a substantial contribution to protein structure and its stability and they are known to exist in nucleic acids. The stabilization energies of complexes of Indene analogs with MeNH2 is 0.39–2.37 kcal/mol higher than that of their corresponding complexes with MeOH which reflect that MeNH2 hydrogen bonds to Indene analogs more strongly than MeOH and this can be rationalized in terms of the fact that MeNH2 is more basic than MeOH.

(D) Indene-AcOH complexes

The hydroxyl group in AcOH can adopt a cis or trans orientation with respect to the carbonyl group. The more stable orientation having cis arrangement of C=O and O-H is selected for the present study and it leads to two distinct types of complexes X and XI. In the “X” set of complexes, AcOH hydrogen bonds to Indene analogs through different proton donor (OH) and acceptor (C=O) sites. In these complexes, C=O and O-H bonds of AcOH are hydrogen bonded to N-H and N5/N6 sites of Indene analogs respectively (AcOHC=O···H-NIndene and IndeneN···O-HAcOH) leading to the formation of seven-membered cyclic ring structure in IND-AcOH X and 4AIND-AcOH X and eight-membered ring structure in 7AIN-AcOH X and 7AIND-AcOH X complexes. The complexes in which AcOH coordinates to Indene analogs solely through the hydroxyl group are marked as “XI.” Indeed, in this configuration, AcOH binds to Indene analogs through N-H···O and N···H-O HBs. Likewise, similar HBs’ interactions are also present in Indene-MeOH III complexes. On comparing Indene-AcOH XI and Indene-MeOH III complexes, it is interesting to note that N···H-O HB interaction is comparatively stronger whereas N-H···O HB interaction is weaker in former complexes with respect to latter complexes which can be justified in terms of the fact that the neighboring carbonyl group in the AcOH enhances the acidity of the hydroxyl group and reduces the basicity of oxygen of hydroxyl group. The “X” set of complexes has the higher stabilization energies, almost double in comparison with the “XI” set of complexes. The reason behind is that there is probability of resonance assisted hydrogen bonding (RAHB) mechanism in the “X” set of complexes which enhances the HB strength, thus supporting the relatively higher stability of these complexes. This is also supported by NBO analysis. Moreover, in the “X” set of complexes, the two HBs closely approach linearity in the seven/eight-membered ring resulting in strong HBs whereas the five-membered cyclic arrangement in case of “XI” set of complexes causes HBs being bent more, resulting in relatively weaker HBs.

Vibrational properties of hydrogen bond donor group

Table 1 reflects the variation in bond distances (∆d in Å) and shifts of stretching vibrational frequencies (Δʋ in cm−1) of the HB donor group upon complex formation relative to monomers. In all the complexes under study, complex formation leads to red shift of conventional HB donors (N-H, O-H) and blue shift of unconventional HB donor (C-H). From the results of the vibrational analysis, it is observed that the lengthening of the HB donor is found in all red-shifting cases and the shrinking of HB donor is always associated with blue-shifted complexes. As anticipated for conventional HB donors, the O-H/N-H/S-H bonds of amino acid side-chain mimics and N-H bond of Indene analogs undergo red shifts when engaged in hydrogen bonding. The Δʋ values of O-H/N-H/S-H HB donors of amino acid side-chain mimics are found to follow the order AcOHO-H (− 310.29 to − 874.98 cm−1) > MeOHO-H (− 145.28 to − 305.89 cm−1) > MeSHS-H (− 17.08 to − 40.60 cm−1) > MeNH2N-H (− 12.31 to − 33.71 cm−1). Despite of having a similar type of hydrogen bonding interactions (O-H···NIndene and O···H-NIndene) in Indene-AcOH XI and Indene-MeOH III complexes, the red shift in the O-H stretching frequency of former complexes is found to be nearly two times higher than that of the latter complexes. The red shift stretching frequency of the N-H of Indene which occurred upon the complexation with amino acid side-chain mimics lay down in the order N-H···NMeNH2 (− 339.78 to − 424.42 cm−-1) > N-H···O=CAcOH (− 218.95 to − 276.72 cm−1) > N-H···OMeOH (− 103.81 to − 194.51 cm−1) > N-H···SMeSH (− 110.25 to − 189.23 cm−1). This is also supported from the proton affinity of acceptor atoms (O, N, S) of amino acid side-chain mimics that exactly lie in the same order MeNH2 (214.7 kcal/mol) > AcOH (187.3 kcal/mol) > MeOH (180.2 kcal/mol) > MeSH (162.9 kcal/mol). This reflects that the abovementioned HB interactions obey acid-base formalism. There is good linear correlation observed in the red shifts in the vibrational stretching frequency of the N-H bond of indene analogs with the elongation of the N-H bond upon complex formation (shown in Fig. 3). It is noteworthy that the larger red shift of N-H which has been observed in the Indene-MeOH complexes in comparison with the Indene-H2O as replacement of one of the hydrogens in the H2O by methyl group increases the electron density on the oxygen due to the positive inductive effect of methyl group, which induces a stronger interaction between the oxygen of MeOH and acidic hydrogen of N-H of Indene analogs. Red shifts in the hydrogen-bonded N-H stretching frequencies in case of the Indene-MeSH complexes are found to be comparable with those of the Indene-MeOH complexes. This infers that N-H···S HBs in proteins and biomolecules are equally strong HBs as their oxygen counterpart. Hence, their significance cannot be neglected or overlooked in comparison with the N-H···O and N-H···O=C HBs in biomolecules.

Fig. 3
figure 3

Plot of shift in vibrational stretching frequency of N-H of Indene against elongation of N-H bond. The straight lines are least squares fit to the data points of N-H···O (diamond), N-H···S (squares), N-H···N (triangle), and N-H···O=C (cross) with R2 values of 0.905, 0.932, 0.961, and 0.993 respectively

Full size image

Natural bond orbital (NBO) analysis

Second-order delocalization energy [E(2)]

NBO analysis has been carried out to investigate the donor-acceptor charge transfer delocalization interactions. These interactions can be derived from second-order perturbation energies E(2) enlisted in Table S4. It is observed that the E(2) values for nN(MeNH2) → σ*N-H(Indene) orbital interactions in Indene-MeNH2 complexes are significantly higher in comparison with nO(MeOH) → σ*N-H(Indene) in Indene-MeOH complexes which can be rationalized on the basis of the fact that HB acceptor ability of MeNH2 is better than that of MeOH. The E(2) values for nN(MeNH2) → σ*N-H(Indene) orbital interactions are remarkably higher than that of the nN(Indene) → σ*N-H(MeNH2) in Indene-MeNH2 IX complexes which can be justified in terms of the fact that the nitrogen of MeNH2 (sp3 hybridized) is a better HB acceptor in comparison with the pyridinic nitrogen (sp2 hybridized) atom of Indene analogs as increasing the s-character of nitrogen lone pair decreases the HB acceptor tendency (sp3 > sp2). The double hydrogen-bonded closed ring structured Indene-AcOH X complexes exhibit a remarkable resonance-assisted hydrogen bonding (RAHB) mechanism as a result of which there is a significant increase in E(2) values of nO(AcOH) → σ*N-H(Indene) and nN(Indene) → σ*O-H(AcOH) orbital interactions. It is interesting to note that E(2) values for nN(Indene) → σ*O-H(AcOH) orbital interactions are nearly double than that of nO(AcOH) → σ*N-H(Indene) in Indene-AcOH X complexes as σ*(O-H) orbital of AcOH has better electron accepting tendency than σ*(N-H) of indene analogs. In accord with this notion, there is a greater role of σ*(O-H) hyperconjugative interaction in O-H bond lengthening of the AcOH in comparison with the σ*(N-H) orbital of Indene analogs in the N-H bond. It is observed that there is enhancement in E(2) values of the nO/S/N → σ*N-H(Indene) orbital interactions in IND/AIND-Z complexes relative to their corresponding IN/AIN-Z complexes which reflect that the presence of an additional nitrogen atom in the five-membered ring increases the electron accepting tendency of σ* antibonding N-H orbitals that urge to the weakening and elongation of the N-H bond and is also harmonious with the concomitant increase in red shift stretching vibration of the bond. It is also noticed that the E(2) values for the nO/N/S → σ*N-H(Indene) orbital interactions in the IN/IND-Z complexes are lower in comparison with their aza analogs AIN/AIND-Z which highlight the fact that the fusion of pyridine ring to single five-membered ring makes the pyrrolic nitrogen atom function as a better HB donor in comparison with the fusion of benzene ring. The discussion of charge transfer analysis is provided in Section S1.

Topological parameters

Atoms in molecules (AIM) theory has been performed on 1:1 Indene-Z complexes to investigate the topology of electron density and intermolecular hydrogen bond properties. The electron density ρ and its Laplacian ∇2ρ at the bond critical points (BCPs) computed at the MP2/aug-cc-pVDZ level of theory for all the complexes are listed in Table S5; see Section S2 for detailed analysis. The ρ and ∇2ρ values at the BCPs are found to lie in the range of 0.002–0.051 au and 0.021–0.141 au, respectively. These values of the ρ and ∇2ρ values are well within the range specified for the existence of the HB. Moreover, ρ and ∇2ρ values at the BCPs for all the complexes are in line with the computed stabilization energies (shown in Fig. S2 and S3). Comparison of relative values of ρ and its ∇2ρ parameters for N-H···O versus N-H···S bound complexes suggests that hydrogen bonding interaction involving the sulfur center are weaker than those involving the oxygen center. These parameters also suggest that the electrostatic component of the hydrogen bonding interaction must be greater for the Indene-MeOH complexes relative to that of Indene-MeSH complexes. Larger values of ρ and ∇2ρ have been observed for IndeneN-H···NMeNH2 HBs in comparison with MeNH2N-H···NIndene in Indene-MeNH2 IX complexes, which corroborates that HB acceptor tendency strongly depends on hybridization of nitrogen atom [MeNH2 (sp3) and Indene (sp2)] with the order sp3 > sp2. Increment in ρ and ∇2ρ values for the N-H···Y (Y = O, S, N) HB interaction has been observed for IND/AIND-Z relative to that of IN/AIN-Z complexes which are consistent with the fact that incorporation of nitrogen atom in the five-membered ring boosts the HB donor tendency of ring nitrogen atom.

Summary and concluding remarks

The present work analyzed thirty-two hydrogen-bonded complexes comprised of N-heterocyclic analogs of Indene-Indole (IN), Indazole (IND), Azaindole (AIN), and Azaindazole (AIND) as one partner and amino acid side-chain mimics (Z), MeOH, MeSH, MeNH2, and AcOH, as the other partner that have been optimized at the wB97XD/aug-cc-pVDZ and MP2/aug-cc-pVDZ levels. By analyzing the geometries, vibrational frequencies, natural bond orbitals, and stabilization energies, the following inferences can be made:

  • The stabilization energies of hydrogen-bonded complexes of N-heterocyclic analogs of indene with amino acid side-chain mimics follow the order Indene-AcOH > Indene-MeNH2 > Indene-MeOH > Indene-MeSH, which can be attributed to the fact that HB interaction ability of amino acid side-chain mimics follows the similar sequence AcOH > MeNH2 > MeOH > MeSH.

  • The HB donor ability of N-H bond of N-heterocyclic analogs of Indene towards amino acid side-chain mimics follow the sequence AIND > IND > AIN > IN. This sequence reflects that the introduction of nitrogen atom in the five-membered ring of indene analogs expedites the tendency of pyrrolic nitrogen atom to act as a better HB donor. This sequence also reflects that the fusion of the pyridine ring with the five-membered ring provides higher stabilization energies in comparison with the fusion of benzene ring as it is reflected from the higher stabilization energies of the AIN/AIND-Z complexes relative to that of the IN/IND-Z.

  • Geometrical parameters indicate that the strength sequence of N-H···Y (Y = O, N, S) HBs from strongest to the weakest is N-H···O=CAcOH > N-H···NMeNH2 > N-H···OMeOH > N-H···SMeSH. From this trend, it can be inferred that carbonyl oxygen of AcOH acts as better HB acceptor in comparison with hydroxyl oxygen of MeOH. On the other way, strengths of N···H-Y (Y = O, N, S) HBs lay down in the order IndeneN···H-OAcOH > IndeneN···H-OMeOH > IndeneN···H-SMeSH > IndeneN···H-NMeNH2. This order reflects that the hydroxyl group of AcOH acts as a better HB donor in comparison with MeOH.

  • NBO analysis illustrates the effect of hybridization on HB acceptor strength of nitrogen by comparing the E(2) values of nN(MeNH2) → σ*N-H(Indene) and nN(Indene) → σ*N-H(MeNH2) orbital interactions in Indene-MeNH2 IX complexes. The comparison clearly shows that E(2) values for nN(MeNH2) → σ*N-H(Indene) orbital interaction is significantly higher in comparison with nN(Indene) → σ*N-H(MeNH2) which is consistent with our general prediction that nitrogen of MeNH2 (sp3 hybridized) is a better HB acceptor in comparison with the pyridinic nitrogen atom of indene (sp2 hybridized), and increased p-character in the nitrogen lone pair increases the HB acceptor strength (sp3 > sp2).

Computational details and choice of method

The geometry optimization of N-heterocyclic analogs of Indene and their parallel hydrogen-bonded complexes with amino acid side-chain mimics has been carried out at MP2/aug-cc-pVDZ and wB97XD/aug-cc-pVDZ levels. All the calculations of hydrogen-bonded complexes of N-heterocyclic analogs of Indene with amino acid side-chain mimics have been done by employing Gaussian 09 W package [26]. The stabilization energies of the complexes have been estimated at the abovementioned levels and are further corrected for basis set superposition error by employing counterpoise procedure. Frequency calculations have been performed at the wB97XD/aug-cc-pVDZ level to ensure that the obtained structures correspond to true energy minima [27]. The natural bond orbital analysis has been determined to obtain second-order delocalization energies, charge transfer, and atomic charges at the MP2/aug-cc-pVDZ level via the method incorporated in the Gaussian 09 W package [28]. The AIMALL program has been utilized to calculate the topological properties at bond critical points and ring critical points of the studied complexes at the MP2/aug-cc-pVDZ level [29].