Interaction of quinolines and artemisinin based antimalarials drugs with ferriprotoporphyrin IX

précédent sommaire suivant

Chapter 4 Results and discussion

4. 1 Choice of the medium

One of the reasons contributed to the conflicting results reported in the literature about interaction hemin-antimalarial drug interactions is the inappropriate choice of the working medium. It has been well shown that the study of Fe(III)PPPPIX in aqueous solution is problematic because of its tendency to aggregate or dimerize. As can be seen from Figure 7, the spectra (b) of hemin in alkaline aqueous solution showed a large band from 350 to 400 nm which is attributed to an oxodimere represented as (H2O)Fe-O-Fe(H2O), whereas hemin was monomeric and exhibited a sharp Soret peak with a maximum at 396-398 nm (50 % PREG or 50 % EG), at 400 nm (25 % DMSO, 80 % ethanol), at 402 nm (40 % DMSO), at 404 nm (DMSO). The slight shift observed towards longer wavelength is due to the change of medium.

Practically, propylene glycol mixture presents the same thermodynamic advantage as ethylene glycol mixture and is much less toxic than the latter. It was used to study both interaction pairs of hemin-quinoline and hemin-artesunate. Since Artemisinin and dihydroartemisinin are insoluble in both propylene glycol and 25 % DMSO, the bonding of hemin with artemisinin compounds was investigated in 40 % DMSO aqueous solutions. Because of their density, polarity, wide temperature range of the liquid state and ability to have bonding hydrogen with water molecules, DMSO and PREG mix easily with water. Particularly, DMSO is an extraordinarily efficient solvent for many kinds of substances including both organic and inorganic compounds. The heat of mixing of DMSO and water indicates there are stronger interactions between DMSO and water than between DMSO molecules (Yu and Quinn, 1994). At high DMSO concentrations, water-structure is disrupted due to the formation of the DMSO-water complexes.

Figure 4-1 Spectra of hemin solutions in different mediums at 25^oC.

The spectrum range from 300 to 500 nm and 250 to 650 nm were selected to study the interactions of hemin-quinoline and hemin-artemisinin, respectively. This is because the induced spectral modifications in the presence of the drugs are more significant in this range than in the remainder of the UV-visible region.

4. 2 Choice of buffers

Tris-HCl buffer was preferred to phosphate buffer because the latter showed some incompatibilty in terms of solubility (formation of precipitate) in 40 % DMSO and was not suitable when pH>8, although it was well used both in 25 % DMSO and water-propylene glycol mixture at pH 7.4.

4.3 Binding reaction of hemin with chloroquine, quinine and quinidine in water-propylene glycol mixture.

More specifically, the wavelength of 396 nm was selected to determine the constants of complexation because of the greatest variation of the optical density observed in the presence of the antimalarial drugs.

Titration of hemin by increasing amount of drugs in mixed water-propylene glycol solutions gives typical spectral changes as exemplified in Figure 4-2. They are similar to those observed on deuterohemin-quinine, hemin-chloroquine and hemin-quinine interactions in other mediums (Constantinidis and Satterlee, 1988; Gushimana et al., 1993, 1996).

The absorption band centered around 332 nm is from the quinoline derivative and that centered at 396 nm is from hemin. As can be seen from Figure 4-2, addition of chloroquine drug modifies markedly the hemin spectrum, but the peak maximums are still at about 396 nm. This indicates that the complexation does not involve significant modifications on the structure of the porphyrin ring of the ferriprotoporphyrin IX.

Another feature that can be seen for all the three drugs is the appearance of an isosbetic point located at around 350 nm on the titration curves. The experimental data were fitted into a 1:1 complex model as described mathematically in Eq. (3-9). What are shown in Figure are selected such results with the total drug concentration as the only changing parameter. It can be seen that the extinction of the hemin solution decreases with increasing total drug concentration.

This trend is consistent with previous results and can be attributed to complex formation between the drug and hemin (Constantinidis and Satterlee, 1988; Gushimana et al., 1993, 1996).

The solid curves in the figure are fitted data with the experimental results according to Eq. (3-9). Correlation coefficients of the nonlinear fittings are better than 0.9, which implies that the titration curves can be well described by the 1:1 complexation scheme. Similar variation in absorbance of hemin at 396 nm as function of total drug concentration has been obtained at other values of pH and the results are also consistent with the formation of 1:1 complex.

Values of binding constants at various pH obtained from these titration curves are summarized in Table 4-1. As highlighted by values of binding constants in Table 4-1, K values are in the same order of magnitude as those obtained in water-ethylene glycol mixture (Gushimana et al., 1993, 1996).

Table 4-1 Binding constant of hemin-drug complexes at various pH.

In fact, the complexation of ferriprotoporphyrin IX with the drug is believed to play the role to bring back the hemin into solution in order to prevent it from polymerization. The ability of quinoline drug to complex with hemin will inhibit the formation of hemozoin (-hematin) in vivo. The drug that has a greater affinity with hemin should maintain more hemin in solution and is thus more effective. This means that quinidine should have the highest efficiency, then comes chloroquine, and finally quinine, based on the data in Table 4-1. But in practical applications, an opposite trend is observed, probably due to the emergence of new resistant strains of malaria parasites against the existing and commonly used antimalarial drugs.

As a matter of fact, in some areas (the case in D.R.Congo, for example) quinine appears more effective than chloroquine. This proves that the strength of haematin-quinoline interactions does not directly correlate with antiplasmodial activity. This indicates that haematin binding is a necessary, but not sufficient requirement for antiplasmodial activity (Egan et al., 1994).

Scheme 4-1 Structures of three quinoline-based drugs, quinine (1), quinidine (2) and chloroquine (3).

In regard to the molecular basis of the hemin-drug interactions, rather less is known about the structures of these complexes. In fact, the complexes between 4-aminoquinolines and hemin are almost certainly p-p complexes (Egan et al., 1994).

This means that there is an interaction between the aromatic ring of the quinoline and the porphyrin structure. In addition, hydrophobic interaction, electronic and steric factors also play important roles in influencing the structures of such complexes. Results from the present study show that chloroquine interacts more strongly with ferriprotoporphyrin IX than quinine does, indicating some additional interaction of the side chain of the quinoline with Fe(III)PPIX. This finding rejoins the result of Egan and co-workers which reported the association constants of chloroquine (log K= 5.52) and quinine (log K= 4.10) in 40% aqueous DMSO at pH 7.5 (Egan et al., 2000). It is suggested that the flexible side aliphatic chain of the chloroquine structure, which is less crowded than that of the stiff quiniclidine group of the quinine structure, stabilizes hemin-chloroquine interaction. It is also supposed also that a hydrogen-bonding interaction between the side-chain amine group of chloroquine and the heme propionate group may play a role in the hemin-chloroquine complex stability. More likely, there may be some direct Van der waals interaction between the side chain of quinoline and the porphyrin ring. In addition, the stability of these complexes is supported by computational results. A molecular mechanics study of the interaction between chloroquine and an iron-porphyrin model for N-acetylmicroperoxidase-8 revealed a minimum energy arrangement with coplanar interaction of the quinoline and iron-porphyrin ring, but could not define a preferred conformation for the complex (Marques, 1996).

It is interesting to note that the conformation of drugs affects their affinity with hemin. As can be seen from Scheme 1, quinine differs from quinidine only at positions C-8 and C-9, the former has 8S9R structure and the latter has 8R9S structure (Ribeiro, 1997). The data showed significantly different affinity to hemin of the two chiral isomers.

Further more, it can also be seen that K values are pH-dependent. That dependence is probably due to acido-basic equilibrium influence on electrostatic interactions between hemin and the drugs. Due to their different pKa values, reacting partners have different electric charge at different pH values (Constantinidis and Satterlee, 1988; Gushimana et al., 1993; Kuhn, 1995).