3.2.1. Complexes Stoichiometry and Association Constant
The complex formation between DAL and MCD was investigated and compared to that of DAL with MCD conjugated to the quaternary ammonium chitosan (N
+-rCh-MCD) or in the presence of plain N
+-rCh (N
+-rCh / MCD). In all cases, the χ value was 0.5, corresponding to the 1:1 complex stoichiometry of DAL/MCD (
Table S8, Supplementary material). The results evidenced that the quaternary ammonium polymer was not interfering with the formation of the complex neither when dissolved together with MCD nor when covalently linked to MCD. Additionally, χ values were not affected by the incubation time, indicating a prompt complexation between MCD and DAL (
Table S8, supplementary material).
The determination of the association constant (K
a) for complexes of DAL/MCD and of DAL/N
+-rCh-MCD was performed both by UV and fluorescence spectrometry. Both techniques exploited the gradual variation of DAL absorbance or fluorescence in the presence of increasing amounts of host agent. These variations were better monitored by using DAL fluorescence, whereas UV spectrometry was poorly applicable to DAL/MCD complexes, which would have required an excessive amount of the binding partners to reveal the absorbance variation. Differently, this latter technique was successfully used only for the DAL/N
+-rCh-MCD complex. The calculated K
a values are reported in
Table 1.
The good linearity of data regression confirmed the formation of a 1:1 complex in both cases, DAL/MCD and DAL/N+-rCh-MCD. More importantly, the Ka evaluated for the peptide complexed to the N+-rCh-MCD polymer was one order of magnitude higher than that calculated for MCD alone. Despite the two different spectroscopic techniques, the values are consistent with each other.
It is worth mentioning that the spectrometric titration performed for the evaluation of K
a evidenced a clear variation of DAL absorbance peak at 227 nm in terms of intensity and wavelength, as clearly displayed by the overlay of the II derivative absorbance spectra of DAL/N
+-rCh-MCD complexes (
Figure 3A). Similarly, the greatest fluorescence variation of complexed DAL was revealed at the 304 nm emission band (
Figure 3B). Both phenomena can be related to a main variation of the chemical environment surrounding Tyr residue, which appeared more pronounced for the DAL/N
+-rCh-MCD complex. As a matter of fact, the Tyr ABS peak at 227 nm is typically exploited for protein unfolding studies [
37], and Tyr fluorescence is predominant with λ
ex 275 and λ
em peaks at 304 nm [
38].
3.2.2. NMR Interaction Studies and Complex Stereochemistry
The interaction between a small molecule (G) and a macromolecule (H) can be described by the simple equilibrium reported in Equation (2). In fast exchange conditions, the observable NMR parameters (P
obs) represent the weighted average between the bound (P
b) and free (P
f) states (Equation (3)).
where χ
f and χ
b are the molar fractions of the ligand in the free and bound states, respectively. By virtue of the marked difference between the molecular weights of H and G, it is necessary to operate with a large excess of G with respect to H in order to obtain an observable NMR signal. In these conditions, the molar fraction of the small molecule in its complexed form (χ
b) is negligible. Therefore, the choice of sensitive parameters is required to detect significant changes in the bound state with respect to the free state. Longitudinal relaxation rates represent an effective tool in the analysis of the interaction between small molecules and macromolecules. In particular, proton mono-selective relaxation rates (R
ims = 1/T
ims), measured by following the recovery of the magnetization of the selectively inverted spin
i, are remarkably more sensitive than non-selective ones [
39]. Considerable increases of R
ms are expected when the small molecule undergoes a change in molecular dynamic from fast (
ω2τc2 ≪ 0.6, where ω is the Larmor frequency and
τc is the rotational correlation time) to slow motion regimes (
ω2τc2 ≫ 0.6), which is typical when bounded to a macromolecule. The cross-relaxation term (
σij), referring to the spin
i and
j at
rij distance, is another susceptible parameter to complexation phenomena.
σij can be experimentally obtained from the difference between the bi-selective relaxation rate (detected by following the recovery of the magnetization of the nucleus
i in conditions of simultaneous inversion of the spin
j, R
ijbs = 1/ T
ijbs) and mono-selective relaxation rate, as shown in Equation (4).
In the two limit cases of fast and slow -motion regimes, the expression of
σij can be calculated by Equations (5) and (6), respectively.
Therefore, a negative σ indicates a slowdown of molecular motion, which is generated by the interaction of the active ingredient with the macromolecule.
Due to the extensive superimposition between DAL and the polymers’ resonances, only the selective relaxation rates of phenylalanine and tyrosine aromatic protons and methyl moieties of leucine of DAL (0.5 mg/mL) were measured in solutions containing pure DAL or its mixtures with polymers/cyclodextrin.
The mono-selective and bi-selective relaxation of DAL protons is collected in
Table 2, along with selected cross-relaxation terms. NMR measurements were performed both immediately after preparation of the solutions and after one week, with negligible variations in the relaxation parameters recorded.
The presence of an equimolar amount of MCD did not produce any significant variations in the relaxation parameters of DAL (
Table 2), indicating minor drug/macrocycle interactions, according to the very low amount of bound molar fraction (7%) at the concentration of 0.68 mM. The same trend was found for the binary mixture DAL/N
+-rCh, with negligible variations with respect to free DAL. Importantly, the minimum impact of the presence of N
+-rCh also rules out possible interfering effects on relaxation parameters attributable to viscosity changes caused by the presence of the polymer, rather than to chemical interactions with the macromolecule. On the contrary, outstanding increments in relaxation rates were recorded when dalargin is mixed with ammonium–chitosan covalently conjugated with MCD (DAL/N+-rCh-MCD), especially for aromatic moieties. As an example, R
1ms of H
3Phe increases from 0.62 s
−1 to a value of 4.54 s
−1. Cross-relaxation rates respectively decrease to the values of -0.88 s
−1 and -0.62 s
−1 for H
3Phe/H
4Phe and H
1Tyr/H
2Tyr proton pairs, due to the relevant slowing down of the molecular motion of DAL consequent to its interaction with N+-rCh-MCD. Interestingly, the simple physical mixture of ammonium–chitosan and cyclodextrin (DAL/N
+-rCh/MCD) produced negligible variations of the relaxation parameters compared to pure DAL, confirming the key role of the covalent grafting of cyclodextrin to the polymer (
Table 1).
To obtain a parameter that reflects the different degree of involvement of DAL fragments in the interaction with polymers, normalized relaxation rates were calculated [|R| = (R
1mix − R
1free)/R
1free], showing a major involvement of tyrosine and phenylalanine group (
Table 3). The particularly high value for the
ortho-protons to hydroxyl group of tyrosine is due to the role of the polar group in the interaction with the polysaccharide material.
The mixture DAL/MCD was employed as a model for investigating the nature of the interaction of DAL with MCD. The study was carried out at 9.8 mM concentration, which guaranteed a significant bound molar fraction of DAL (about 40%) on the basis of the value of the association constant DAL/CD (120 M
−1). Such a kind of investigation was not possible for the mixture DAL/N
+-rCh-MCD due to the extensive overlapping of polymer and cyclodextrin signals. An extended conformation of DAL was found on the basis of the analysis of 1D ROESY spectra, where any significant dipolar interactions were not detected between protons bearing different amino acidic residues. As an example, the selective perturbation of Tyr protons produces negligible ROE effects on the adjacent Ala protons (
Figure 4).
In order to point out the nature of the DAL groups that were more involved in the interaction to MCD, complexation shifts were calculated (|Δδ| = |δ
DAL/MCD – δ
MCD|, Hz). For this type of analysis, TOCSY traces were accurately analyzed: starting from the anomeric protons H
1 and H
1′ of MCD, resonances of the corresponding glucosidic rings protons were individuated. Complexation shifts were remarkably higher for H
3/H
3′ and H
5/H
5, which were located inside the hydrophobic cavity of MCD with a deep inclusion from the edge of larger diameter and supported by considerable values of Δδ for H
5/H
5′ (
Table 4).
Detailed information on the stereochemistry of the inclusion were obtained from the ROESY map, (
Figure 5 and
Figure S6, supplementary material), showing ROE effects mainly between the aromatic protons of DAL and the H
3/H
3′ and H
5/H
5′ internal protons of the cyclodextrin.
In particular, H1Tyr produces much more intense ROE effects on the H5/H5′ (narrower rim) compared with the H3/H3′ (wider rim), while H2Tyr showed an opposite trend, indicating that the phenolic group is embedded in the cavity of the cyclodextrin deeply and confirming what was observed with UV and fluorescence spectrometry.
Phenylalanine is also widely involved in the interaction with cyclodextrin, but the effects produced at H3/H3′ and H5/H5′ frequencies have comparable intensities, which is probably because of aromatic moieties mobility within the cavity of the cyclodextrin.
It is noteworthy that both methyl protons of leucine and alanine originated dipolar interactions with the internal H3/H3′ protons and external methoxy protons of the cyclodextrin, thus revealing superficial interactions at the wide rim of the cyclodextrin.
Therefore, taking into account that 1 to 1 complexation stoichiometry has been demonstrated, it can be concluded that dalargin may originate multiple interactions with the cyclodextrin, involving the deep inclusion of the aromatic groups of tyrosine or phenylalanine, as well as superficial interactions of the alkyl groups of leucine or alanine at the wide rim of the cyclodextrin.