Raman spectra simulation of antiplatelet drug-platelet interaction using DFT

The paper reflects the results of molecular docking and mathematical DFT simulation for antiplatelet drugs and the target platelet receptor/ferment interaction in the limited area. The results of Raman spectra simulation are implemented and obtained from the interaction of the clopidogrel metabolite of the P2Y12 receptor. The interaction of aspirin with the COX-1 enzyme was also investigated. As a result, theoretical Raman spectra of the drug-receptor area were obtained. The theoretical data were compared with the experimental SERS results. The characteristic bands corresponding to metabolite/ferment and antiplatelet drug vibrations were clarified. The prospects of obtaining results for pathologies based on platelet conformations during cardiovascular diseases have been demonstrated.


Platelets preparation
Twenty-three healthy volunteers were involved in the study.Written informed consent had been obtained from all healthy volunteers before any study procedures.All study documents, including informed consent and protocol were approved by Immanuel Kant Baltic Federal University Independence Local Ethic Committee (Protocol No 8, 16.05.2019).The methods used were carried out in accordance with the local ethics committee of the independence of the Immanuel Kant Baltic Federal University and in accordance with the Declaration of Helsinki.Healthy volunteers aged 18-45 years without acute and chronic diseases were included in the study.All healthy volunteers were divided into 3 groups: 11 subjects without antiplatelet therapy, 8 subjects after taking 100 mg aspirin, and 4 subjects after taking 300 mg clopidogrel.The preparation protocol was based on 11 .Briefly, fresh venous blood samples were taken from healthy volunteers in vacuum tubes containing EDTA (BD Vacutainer® spray-coated K2 EDTA Tubes).It was centrifuged at 60 g for 15 min to consequentially separate platelet-rich plasma (PRP) from red blood cells (RBC) and leukocytes.After that, the PRP was collected and placed in the new tube.Finally, platelets were collected by further centrifugation of the supernatant at 1500 g for 15 min.All the centrifugations were carried out at 4 °C using Eppendorf 5702R centrifuge.After platelet preparation, samples were immediately taken for examination by Surface-enhanced Raman spectroscopy (SERS).

SERS substrates fabrication
SERS substrate fabrication was performed in three steps: Firstly, in order to create roughness, anodization of 0.1 mm thick titanium films was carried out on the laboratory hand-made equipment with a current source and a galvanic bath, in which titanium electrodes were immersed.An aqueous solution of KOH (5%) was used as the electrolyte.Anodizing was carried out at a current density of j = 30 mA cm −2 for 5 min.After anodization, the titanium surface acquired a dark blue color.In the second stage, the gold nanoparticles were prepared by the femtosecond laser ablation method.In the third step, gold nanoparticles were deposited on these surfaces.The gold plate of 99.9% purity was placed in distilled water and AVESTA femtosecond laser unit with a TETA compressor (TETA Yb amplifier system) was used.The energy and pulse duration of the laser beam at λ = 1032 nm were 15 µJ and 280 fs, respectively.The repetition frequency and the number of pulses were set by an external generator.The solution volume in the cuvette was 1.2 ml, ablation time was 5 min; the thickness of the distilled water layer over the surface of the metal plate was 2 mm.Each package of laser pulses was focused to a new location of the plate.After ablation, the solution assumed a slightly red color.The hydrodynamic radius of the obtained nanoparticles was measured by dynamic scattering light method with PhotoCorr-Complex unit (LTD "PhotoCorr", Russia) and was found to be in the range of 20-80 nm.As a last stage, deposition of ablative gold nanoparticles on titanium rough surfaces was carried out as follows: the titanium substrate was immersed in a gold nanoparticles solution, then the nanoparticles were deposited on the surface by evaporation of an aqueous colloidal gold solution at a temperature of 60 °C for 40 min.The detailed methodology of SERS substrates preparation was described in 12 .

SERS experiment
SERS spectra were obtained by Centaur U (LTD "NanoScanTechnology", Russia) Raman spectrometer, using the λ = 532 nm DPSS Cobolt Samba excitation laser with 45 mW power on sample.The optical scheme included Olympus BX 41 microscope with 100X (NA 0.9) objective.Spectrometer Shamrock 750 (Andor, UK) had a focal length of 800 mm and was equipped with 300 gr mm −1 diffraction grating with 500 nm blaze.IDus 401-DV CCD camera (Andor, UK) with 1024 × 256 pixels sensor was used for the experiments.The spectrometer had a spectral resolution of 1.5 cm −1 .The laser spot of 1 × 25 μm size was positioned at the platelets.Rayleigh scattering was eliminated by the notch filters.A 5 μl droplet of platelet-rich plasma was put on the substrate, dried for 5 min at room temperature, and then placed in the microscope holder.Three times averaged spectra from ten different places of the droplet have been collected for each sample.The signal acquisition time was 70 s.Each time before experiment, spectrometer was calibrated with silicon at a static spectrum centered at 520.1 cm −1 for 1 s.After registration, spectra were saved in .txtand specific format (.ngs) on the PC, connected to the Raman unit.

Molecular docking
Molecular docking was performed using the Molecular Operating Environment (MOE) 2014 software platform.The binding site of the ligand and the receptor for further modeling were determined.The Gaussian software package was used for further calculations.Docking was performed on platelet P2Y12 and COX-1 receptors.Its structures were obtained from the Protein Data Bank (PDB).Based on a review of the literature (it was determined that the drugs that act on these receptors are clopidogrel (for P2Y12) and aspirin (for COX-1)).Literature sources indicate that P2Y12 is affected not by clopidogrel itself but by its active metabolite 13 , so it was used in molecular docking procedures.The structures of the ligands were obtained from the online libraries DrugBank and PubChem.Before implementing molecular docking procedures, platelet receptors were prepared.The amino acid sequence of the human platelet receptor P2Y12 (PDB: 4NTJ) was used as a query sequence to search for homologue models with known structures from the Protein Data Bank (PDB), and protein similarity was assessed using NCBI-BLAST 14 .For homologous modeling, a similar human P2Y12 protein (PDB: 4PZX) was chosen to complete the artifacts and gaps in the protein crystal structure.A three-dimensional model of the human receptor was created using the MODELER program (version 10.2).For COX-1 (PDB: 6Y3C) no homologous modeling has been performed.Next, the proteins were minimized first in the gas phase and then with a solvent, using the MMFF94x force field, to obtain the most stable conformation.After that, molecular docking of the protein and drug was performed.

Simulation of Raman Spectra
The Gaussian 16 15 software package (license number: G64284555249899W-6922N) was used to calculate the theoretical Raman spectra.The Raman spectra were obtained by DFT using the selected functional (B3LYP) 16 based on optimized molecular structures.The 6-31G(d) split valence bivalent base set was chosen as the basis set.Before theoretical calculations, the analyzed structures were energetically minimized in the MOE software package.Based on the results of molecular docking, the vibrational spectra of the interaction regions of platelet receptors P2Y12 and cyclooxygenase-1 (zwitterionic forms of amino acids) and metabolites of drugs inhibiting the receptors (thiol metabolite of clopidogrel H4 and aspirin, respectively) were mathematically calculated.
In order to obtain more accurate theoretical results, linear scaling of the wave numbers was carried out.As a criterion for assessing the quality of calculation of oscillation frequencies, linear scaling of wave numbers 17 was carried out according to the formula (1): where v ex , v th -are the experimental and calculated frequency (cm −1 ), a and b are coefficients to be determined.In the framework of this work, for the 6-31G(d) basis we used, the coefficients a = − 0.0000083526 and b = 0.98134 were chosen.
The Raman activities (Ai) calculated with the Gaussian 16 program were converted to relative Raman intensities (Ii) using the formula (2): where v 0 is the exciting frequency (cm −1 ), v i is the vibrational frequency (cm −1 ) of the ith normal mode, h , c , and k are fundamental constants, and f is a suitably chosen common normalization factor for all peak intensities.The calculated spectra were reported by assigning to each normal mode a Lorentzian shape with a 5 cm −1 full width at half-maximum.For these calculations, a program was used for modeling and visualizing mixed IR and Raman spectra based on quantum mechanical calculation data 18 .

Comparison of theoretical and experimental spectral Raman data
Experimental data were divided into samples: healthy patients, healthy patients on aspirin therapy and healthy patients on clopidogrel therapy.Using the laboratory build Parcer analyzer, statistical processing was carried out.The data recorded from the device were converted into .txtand .csvformats.Further, the data were entered into tables of certain groups and a fixed frequency grid was created with a step of 5 cm −1 .The Parcer analyzer decomposes the spectrum into grid cells in the range from 400 to 1800 cm −1 in 5 cm −1 steps.Thus, all spectral fluctuations were correlated with the designated grid.Vibrational bands were used to correlate the obtained experimental data with the calculated theoretical ones.In the Origin 2021 program, experimental platelet spectra of healthy patients and patients on therapy normalized to 100 were built, after which vibrational modes were added.

Results
Experimental SERS spectra were obtained from the platelets of healthy volunteers, a healthy volunteer during clopidogrel therapy, and healthy volunteers during aspirin therapy.For a correct comparison of the results obtained, an analysis of three groups of platelets from one patient was carried out.Vibrational bands were used to correlate the obtained experimental data with the calculated theoretical ones.Figure 1 shows the Raman spectra for three groups of healthy patients: on aspirin therapy, on clopidogrel therapy, and without therapy.
Docking calculation of the P2Y12 receptor and the clopidogrel metabolite H4 was performed as the first step.As a result, diagrams of the ligand-protein interaction were obtained.The software performed 1000 iterations of ligand-to-protein attachment and sorted them by interaction enthalpy.However, variants of the docking of the metabolite and the receptor with a minimum energy were further considered (Fig. 2), since this interaction variant is the most probable.
Analyzing the obtained data, the surface contact interactions of the protein and the ligand were determined, however, these types of interactions did not form bonds and therefore were not used in further analysis.Also, the data obtained indicate that the binding of the metabolite occurs with Arginine at position 137 due to the donor-acceptor interaction, where the donor is the oxygen of the metabolite, the acceptor is the amino acid (Fig. 2).The cyclooxygenase-1 receptor and aspirin system were used for molecular docking implementation as a next step.In the course of molecular docking, diagrams of the ligand-protein interaction were obtained.The software performed 1000 iterations of ligand-to-protein attachment and sorted them by interaction enthalpy.However, variants of the docking of the metabolite and the receptor with a minimum energy were further considered (Fig. 3), since this variant of interaction is the most probable.
It was shown that aspirin binds to arginine at position 376 through a donor-acceptor bond, where the metabolite is the donor and the amino acid is the acceptor (Fig. 3).
In the third step, Raman spectra for the selected chemical compounds in the range of 0-4000 cm −1 were calculated using the Gaussian program.After that, linear scaling of the wave numbers was carried out in the area of the fingerprint 400-1800 cm −1 .
The results of the calculated Raman spectra are presented in Fig. 4 (the active metabolite of clopidogrel and arginine) and Fig. 5 (aspirin and arginine).
The comparison of the experimental and theoretical spectra was performed as a last step.In the case of considering the theoretically obtained spectra of the binding site of the P2Y12 receptor and the clopidogrel thiol metabolite H4, a comparison was made between healthy patients without therapy and with clopidogrel therapy (Fig. 6).
Using the Gaussian 16 software package, the vibrational modes of the theoretical spectrum of the metabolite of clopidogrel and arginine were deciphered.Since the obtained calculated data correlates with the experimental data, the area of the fingerprint was considered.The results were entered in Table 1. (1) kT By superimposing theoretically calculated vibrational modes on the experimentally obtained spectra of platelets from a person who did not undergo therapy, and on the spectra of a person who is under clopidogrel therapy, it can be determined that the changes occurring in the P1Y12 receptor can be explained by the binding of the clopidogrel metabolite to arginine at position 137 in the receptor.It can be seen from the graphs that the change in the intensities and spectral shifts of the spectra during therapy correlate with the theoretically calculated vibrational modes.
In the case of considering the theoretically obtained spectra of the binding site of the COX-1 receptor and aspirin, a comparison was made of healthy patients without therapy and with aspirin therapy (Fig. 7).
Using the Gaussian software package, the vibrational modes of the theoretical spectrum of aspirin and arginine were deciphered.Since the obtained calculated data correlates with the experimental data, the area of the fingerprint was considered.The results are reflected in Table 2.
By superimposing theoretically calculated vibrational modes on the experimentally obtained spectra of platelets from a person who did not undergo therapy, and on the spectra of a person who underwent aspirin therapy, it can be determined that the changes occurring in the COX-1 receptor can be explained by the binding of aspirin to arginine at the position 376 in the receptor.It can be seen from the graphs that the change in the intensities and spectral shifts of the spectra during therapy correlates with the vibrational modes calculated theoretically.

Discussions
The obtained spectra were analyzed based on the analysis of scientific literature.Most of the mentioned frequencies are experimentally obtained.From the obtained Figs. 6 and 7.The vibrational modes corresponding to CH 2 vibrations, located at 1450 cm −1 , 1451 cm −1 , 1474 cm −1 , 1475 cm −1 were revealed and reflected interaction of clopidogrel and arginine for H-O-C and CH 2 bands.Vibration at 997 cm −1 correlates with stretching of 1-azacyclohexnene-3 and scissoring H-S-C (metabolite) and wagging NH (arginine) and can also can correlates with 1001 cm −1 characteristic of the aromatic ring experimental band of phenylalanine 12,19,20 .The spectral shift by 4 cm -1 in the case of clopidogrel therapy was revealed (the mode was detected at 997 cm −1 ).When considering the region 700-950 cm −1 , several spectral differences of the groups under consideration were identified.Maximums 767, 905 describing vibrations of aromatic groups characteristic of the amino acids tyrosine, tryptophan and phenylalanine 12,[19][20][21] , when exposed to different antiplatelet drugs, change in different ways.Spectral maximums at 1017 cm −1 , 1019 cm −1 shows no correlation with the literature.Maximums 1450-1475 cm −1 and 1731-1764 cm −1 spectral regions correlate with vibrations in lipids 21 .
By superimposing theoretically calculated vibrational modes on the experimentally obtained spectra of platelets from a person who did not undergo therapy, and on the spectra of a person who underwent aspirin therapy, it can be determined that the changes occurring in the COX-1 receptor can be explained by the binding of aspirin to arginine at the position 376 in the receptor.It can be seen from the graphs that the change in intensities and spectral shifts of the spectra during therapy correlate with vibrational modes calculated theoretically.www.nature.com/scientificreports/shows the selected fluctuations for the binding sites of clopidogrel/aspirin to the platelet site, which can be potential biomarkers of the interaction of these compounds in the spectrum.During the experimental part, the spectra of human platelets were obtained without therapy, on aspirin therapy and on clopidogrel therapy.Platelet receptors and drugs that inhibit them have been investigated.
Using molecular docking, it was determined that the clopidogrel metabolite binds to arginine at position 137 in the P2Y12 receptor, and aspirin binds to arginine at position 376 in the COX-1 receptor.The correlation of theoretical and experimental data showed that the change in the SERS spectra of human platelets during therapy may be associated with a change in the conformation of amino acids (phenylalanine, tyrosine, and tryptophan).It was also determined that the changes that occur in the receptor during therapy can be explained by the binding of the drug to the receptors through arginine at different positions (137 for P2Y12 and 376 for COX-1).From the graphs obtained, it can be seen that the change in intensities and spectral shifts of the spectra during therapy correlate with vibrational modes calculated theoretically.Spectral shift at 1339 cm −1 , 814 cm −1 , 608 cm −1 , 680 cm −1 correlates with aromatics was revealed.For 1339 cm −1 and 814 cm −1 the correlation with platelets aromatics in Trp were revealed 12 .1708 cm −1 band lays near Amide I experimental band.Analyzing low-frequency 450-531 cm −1 : 465 cm −1 , 492 cm −1 , 510 cm −1 , 516 cm −1 , 541 cm −1 , 544 cm −1 bands corresponding for aspirin-arginine has been revealed and shows no direct correlation with experimental data, nevertheless 510 cm −1 and 516 cm −1 spectral bands correlates with -S-S, -C-S and -C-C vibrations 24 .Spectral bands at 544 cm −1 , 797 cm −1 , 1060 cm −1 , 1087 cm −1 , 1579 cm −1 .1621 cm −1 characterize hydrogen bonding in the complex and also can be potential biomarkers of therapy response for aspirin.Aspirin can form the assigned H-bonding interaction with different groups, including COX-1 receptor 22,23 .This fact, proved theoretically and experimentally in our paper, confirms the correctness of receptor-drug interaction DFT simulation (Supplementary Information).

Conclusions
The authors perform molecular docking and mathematical DFT simulation for antiplatelet drug and the target platelet receptor/ferment interaction in the limited area.The Raman bands, corresponding to the drug, aspirin/ clopidogrel and its interaction in the binding site have been revealed.Vibrational modes shown in Table 3 can be

Figure 1 .
Figure 1.Comparison of SERS spectra of platelets from a healthy person without therapy, on aspirin therapy and on clopidogrel therapy.Red markers indicate main experimental characteristic bands.

Figure 2 .
Figure 2. Interaction diagram of the P2Y12 receptor and clopidogrel metabolite.

Figure 3 .
Figure 3. Interaction diagram of the COX-1 receptor and aspirin.

Figure 4 .
Figure 4. Theoretical Raman spectrum of the active metabolite of clopidogrel and arginine after frequency adjustment.

Figure 5 .
Figure 5. Theoretical Raman spectrum of aspirin and arginine after frequency adjustment.Results of correlating theoretical data with experimental data.

Figure 6 .
Figure 6.Superimposed modes of the metabolite of clopidogrel and arginine on the experimentally obtained spectra of platelets from different samples.The left scale shows the intensity of vibrational modes of the experimental data, the scale on the right shows the theoretically calculated vibrational modes.The vibrational modes presented in Table3are highlighted in blue.

Figure 7 .
Figure 7. Superimposed modes of aspirin and arginine on the experimentally obtained spectra of platelets from different samples.The scale on the left shows the intensity of vibrational modes of the experimental data, the scale on the right shows the theoretically calculated vibrational modes.The vibrational modes presented in Table3are highlighted in red.

Table 3
are highlighted in blue.

(metabolite) and stretching C6=O1, scissoring C9-O3-H (arginine)Table 1 .
Calculated wavenumbers and vibrational mode assignments for the interaction site of clopidogrel thiol metabolite H4 and arginine at the P2Y12 receptor.Significant values are in[bold].Calculated wavenumbers are presented with integer precision.The atomic number is indicated after the name of the element.

Table 3
are highlighted in red.

Table 2 .
Calculated wave numbers and assignment of vibrational modes for the site of interaction between aspirin and arginine in the COX-1 receptor.Significant values are in[bold].Calculated wavenumbers are presented with integer precision.The atomic number is indicated after the name of the element.