Abstract
From the first month of the COVID-19 pandemic, the potential antiviral properties of hydroxychloroquine (HCQ) and chloroquine (CQ) against SARS-CoV-2 suggested that these drugs could be the appropriate therapeutic candidates. However, their side effects directed clinical tests towards optimizing safe utilization strategies. The noble metal nanoparticles (NP) are promising materials with antiviral and antibacterial properties that can deliver the drug to the target agent, thereby reducing the side effects. In this work, we applied both the quantum mechan
ical and classical atomistic molecular dynamics approaches to demonstrate the adsorption properties of HCQ/CQ on Ag, Au, AgAu, and Pt nanoparticles. We found the adsorption energies of HCQ/CQ towards nanoparticles have the following trend: PtNP > AuNP > AuAgNP > AgNP. This shows that PtNP has the highest affinity in comparison to the other types of nanoparticles. The (non)perturbative effects of this drug on the plasmonic absorption spectra of AgNP and AuNP with the time-dependent density functional theory. The effect of size and composition of NPs on the coating with HCQ and CQ were obtained to propose the appropriate candidate for drug delivery. This kind of modeling could help experimental groups to find efficient and safe therapies.
Introduction
Given that the process of developing new drugs to become appropriate clinical candidates is extensive, one of the most rapid and reliable treatments is drug repurposing—the examination of existing FDA approved drugs for new therapeutic purposes1. Chloroquine (CQ) and hydroxychloroquine (HCQ) have been used for many years as pharmacotherapies for malaria and were recently proposed as a potential therapeutic option against COVID-192. The pre-clinical studies have shown the prophylactic and antiviral effects of CQ and HCQ against SARS-CoV-2 (or COVID-19) The clinical safety profile for HCQ is better than that of CQ, thus allowing for long-term usage and higher daily dosage7. Some reports have mentioned that large scale (and prolonged) usage is potentially harmful and increases the risk of drug-induced torsades de pointes and may lead to cardiac death8,9,10. Therefore, different treatment regimens try to focus on efficient strategies for in vivo usage of these drugs11,12, especially the balance between the concentration of the drug in the blood and its severe potential toxicity, to ensure the safety of these therapeutic strategies7,13. Despite conflicting evidence on the efficiency of HCQ for the treatment of COVID-19, the recent clinical studies have reported no potent evidence to support the benefit of HCQ as a treatment of COVID-1914. Nanoparticles encapsulating drugs or attaching to therapeutics can be utilized as drug delivery systems to change drug biodistribution, decrease toxicity, modify drug release rate, and target affected tissues or cells15,16,17,18. However, most nanoparticles are still in the clinical trial stage, with a few having been accepted for clinical use16. In this regard, noble metal nanoparticles are well-known as promising materials that can transport drugs to specific targets in the body and be engineered to develop new delivery systems19. Notably, silver, gold, and platinum nanoparticles reveal stability in the biological environment and survive in an intracellular environment20,21,22. The stable nanoparticles with small size possess the advantage of easily interacting with biomolecules both on the surface and inside cells, thereby playing a significant role in biomedical applications such as drug vehicles in diagnosing and treating diseases.
Computational details
For HCQ and CQ molecules, the geometry optimization and frequency calculation were performed with PBE generalized gradient (GGA) exchange–correlation (xc-) density functional39 with the inclusion of the Grimme dispersion correction scheme (D3)40,41,42 applying Becke-Johnson damping and a triple-ζζ polarized (TZP) Slater type basis set (PBE-D3/TZP). The Conductor like Screening Model (COSMO)43 was considered to model the effect of water solvent. For the optimized structure, the Hirshfeld point charges44 and electrostatic potential map were obtained both with and without the solvent (in Fig. 1 and Supplementary Fig. S1). The experimental interatomic metal–metal distance was employed to create starting structures for further optimization with the LDA (local density approximation) xc-functional45 and the scalar relativistic ZORA formalism46,47. The interactions of HCQ with icosahedral silver and gold clusters with 147 atoms were investigated at the PBE-D3/TZP level of the theory under the influence of the relativistic effect (ZORA). To determine the effect of HCQ drug on the electronic structures and the plasmonic absorption spectra of noble metal particles, the recently developed time-dependent density functional approach, TD-DFT + TB method48,49 which combines a full DFT ground state with tight-binding approximations, was applied. The excited states calculations were performed at optimized geometries using the asymptotically corrected LB94 xc-functional50, and the absorption spectra were obtained in the range of 0.0–6.0 eV. All these calculations were performed with the Amsterdam Density Functional (ADF2019.1) program51.
Results and discussion
Interaction of HCQ with AgNP and AuNP and its influence on absorption spectra The charge distribution of HCQ and CQ molecules and their electrostatic potential map in Fig. 1 displays the active sites of these molecules for interaction with noble metal NPs. The initial structure of complexes was generated by placing the small silver cluster near the electron-rich sites (e.g., N-, O- and Cl- groups). These sites can donate the electron density via their lone pairs to 4d and 5s orbitals of the silver atom64,65. The nitrogen of the pyridine ring in CQ and HCQ and the oxygen of the hydroxyl group in HCQ have the highest affinity for interaction with noble metal clusters. Moreover, the optimized structure of HCQ/CQ on Ag(111), Au(111), and Pt(111) layers exhibited the highest affinity of drug molecules toward the platinum surface, and their charge density difference confirmed the transfer of charge and accumulation on the metal surface (see Fig. 2, Supplementary Fig. S2, and Supplementary Table S8).
In addition, Fig. 1 shows the stable geometry of icosahedral Ag147 and Au147 nanoparticles which are complexed with HCQ molecules (at PBE-D3/TZP level of theory). Here, the non-covalent charge-transfer interactions with partially negative charge groups of the molecule play an essential role in determining the ability of nanoparticles to bond with HCQ or CQ. The binding energy of HCQ with AgNP (at PBE-D3/TZP level of theory) is about ∆Eb = − 21.06 kcal mol−1 (per HCQ molecule), while the interaction energy with AuNP is about ∆Eb = − 29.39 kcal mol−1 more favorable than silver. The higher electron affinity of gold (EAAu = 2.31 eV) compared to silver (EAAg = 1.30 eV)66 gives rise to increasing the interaction energy of gold atoms towards the lone-pair of HCQ. This is also confirmed by the density difference map and accumulation of negative charge on the Au(111) surface. In this regard, the adsorption energy of HCQ towards Pt(111) is about 60% more than the Au(111) surface (see Fig. 2, Supplementary Fig. S2, and Supplementary Table S8).
n this part, the effect of changing the type of nanoparticles and increasing the number of HCQ molecules on the coating properties of nanoparticles are discussed based on molecular dynamics calculations and the trend of the radial distribution function (RDF). RDF depicts how the density of one molecule changes as a function of the distance from another reference molecule. Besides, RDF can be used to represent distance-dependent relative probability for observing a given site (or atom) relative to some central site (or atom). This analysis provides the microstructure information about the arrangement of HCQ/CQ molecules and their affinity for interactions with nanoparticles6
Furthermore, the O-atom of the hydroxyl group is another active site of HCQ for interaction. However, for gold, silver, and alloy nanoparticles, the g(r) values for O-group is lower than N-group, and for PtNP, it is slightly more than N-group. For PtNP–HCQ in Fig. 4d, the approximately similar RDF peaks around 3.0 Å for both O-group and N-group can be related to the high affinity of the PtNP to interact with both sides of HCQ. For gold and silver noble metals, the higher attraction of N-group (verse O-group) was established by Antusek et al. based on ab-initio calculation70.
Conclusion
In summary, the adsorption and the coating properties of noble metal nanoparticles with HCQ/CQ molecules have been studied as a potentially efficient strategy for in vivo usage of these drugs. The weak charge-transfer interaction with partially negative charge groups of drugs was investigated, and it was established by changing the type of nanoparticle elements that the affinity towards N and O groups increases as follows: AgNP < AuAgNP < AuNP < PtNP. Following the investigation of the effect of size on the coating properties, it was found that the overall affinity decreases by increasing the size from 1.6 nm (for Ag147) to 2.6 nm (for Ag561). For Ag561, Ag1415, and Ag2869 with diameters 2.6 to 4.6 nm, a nearly similar decrease in affinity was obtained. Finally, based on the quantum mechanics and molecular dynamics simulation, we can suggest these noble nanoparticles (with low toxicity and antiviral activity) as appropriate vehicles for efficient HCQ/CQ usage with decreased side effects of the drugs.