Structural dynamics of chlorpromazine (CPZ) drug with dipalmitoylphosphatidylcholine (DPPC) lipid: a potential drug for SARS-CoV-2

Abstract There is an urgent requirement for drug discovery and more importantly drug repositioning due to infectious new Severe Acute Respiratory Syndrome coronavirus 2. As per the recent report published in the journal L'Encéphale in May 2020, there is a planned ReCoVery Study examining the repurposing the chlorpromazine for the treatment of COVID-19. Here, we apply a combined Raman microspectroscopy and DFT-MD approach to investigate the structural dynamics of the Chlorpromazine (CPZ) drug with dipalmitoylphosphatidylcholine (DPPC) lipid bilayer, identifying the specific position of the drug in the DPPC lipid bilayer. The intensity ratios of the Raman peaks I 2935/I 2880, I 1097/I 1064 and I 1097/I 1129 are representative of the interaction of drugs with lipid alkyl chains and furnish conformation of lipid alkyl chains. Raman imaging microscopy for the study of the distribution of CPZ inside the lipid vesicles is reported. We also investigated the influence of order and disorder ratio in the CPZ on the DPPC liposomes prepared on phase transition temperature. HIGHLIGHTS Drug–membrane interactions using micromolar concentrations of both lipid and drugs. Neuroleptic drug and DPPC vesicles composed of DPPC/drug mixtures reveal qualitative differences between the Raman spectra The temperature-controlled Raman microspectroscopic study has demonstrated that below phase-transition temperature, the fatty acid chains of the phospholipids are stiff and packed in a highly ordered array. DFT and MD simulations to understand molecular interactions, structural dynamics, and Raman spectra. Above phase-transition temperature, the chains are disordered and possess more motional freedom. Communicated by Ramaswamy H. Sarma


Introduction
Newly discovered coronavirus has resulted in deadly and highly infectious Coronavirus disease . The COVID-19 cases have passed 600 million and are expected to rise worldwide including India which is expected to surpass the USA with over 100 million cases. Because of this global epidemic of COVID-19 and an enormous number of increasing deaths in many countries, it is severely urgent to discover drugs which have the potential to reduce or stop the infection. Drug repositioning is an alternative to new drug discovery and this strategy is shown to be very promising, which identifies new therapeutic opportunities for already known drugs eliminating many stages of development. This technique enables to deploy of a therapy whose possible side effects are already known to human beings and can be efficiently tackled.
In vitro antiviral activity is already shown by CPZ for various viruses viz. influenza virus, HIV, Japanese Encephalitis (JE), and Chikungunya alphavirus (Nawa et al., 2003). Initially, CPZ was found to block antiviral activity in coronaviruses. Moreover, CPZ is also known to block MERS and SARS CoV replication in human cells (Thomas, 2020). CPZ was found to inhibit MERS-CoV-induced pronounced cytopathic effect (CPE) (EC 50 , 4.9 lM; SI, 4.3) with a 12 lM dose achieving complete inhibition (de Wilde et al., 2014). Very recently, M. Plaze et al. repurposed CPZ due to its anti-SARS-CoV-2 activity that can act as an alternative and quick strategy to alleviate infection severity (Plaze et al., 2020). Udrea et al. suggested phenothiazines photoproducts possess significant biological activity on SARS-CoV-2 Mpro and can be employed to treat COVID-19 disease (Udrea et al., 2020). CPZ can block the clathrin-dependent endocytosis by interacting with cell membrane protein which is necessary for coronavirus entry into the host cell. Recently, Sathyamoorthy et al. reported the plausible role of chlorpromazine hydrochloride against COVID-19 (Sathyamoorthy et al., 2020). Raman microspectroscopy is being used extensively in investigating a broad range of medical and biological problems (Kalasinsky, 1996;Carmona et al., 1997;Mantsch & Jackson, 1998;Shaw & Mantsch, 1999;Schrader et al., 1999;Mahadevan-Jansen & Puppels, 2000;Carden & Morris, 2000;Salzer et al., 2000;Gremlich & Yan, 2001;Kalasinsky & Kalasinsky, 2005). Highresolution vibrational spectroscopy not only can reveal the characteristics of a specific functional group but also identify and differentiate several functional groups altogether. Raman microspectroscopy has an immediate appeal in the study of lipid-drug interactions because it is a non-invasive and nondestructive technique and is used extensively for various diseases specifically in the detection of a variety of cancers and even pathogens present in the environment (James et al., 2006;Nerdal et al., 2000;Underhaug Gjerde et al., 2004;Mishra & Mishra, 2021a). Using Raman spectroscopic imaging techniques one can observe the spatial distribution of various molecular constituents in a sample at a microscopic scale. CPZ is the best known typical anti psych and it is widely used as a neuroleptic drug.
In the present work, we report a systematic study of the lipid drug interaction based on Raman spectroscopic, density functional theory (DFT) and molecular dynamics (MD) studies. Here, we aim to investigate CPZ drug and dipalmitoylphosphatidylcholine (DPPC) lipid interaction at the atomic (or microscopic) level. The phase-transition temperature (Tm) has an important effect on the structure of the biomembrane and individual phospholipids are known to have closely packed forming a rigidly ordered gel phase below their specific Tm (Kinnunen, 1991;Noothalapati et al., 2017). The phase-transition temperature studies have been employed to observe chemical reactions on phosphatidylcholine vesicles (Fox et al., 2007;Fox et al., 2006). The vibrational properties of CPZ, as with lipids in analysis, are the result of careful clinical observations (Jamieson & Byrne, 2017;Balan et al., 2019;Mishra & Mishra 2021b) and we present a detailed discussion on the vibrational properties of DPPC and the effect of CPZ interaction with DPPC through Raman and DFT-based calculations.

Experimental details
DPPC and Chlorpromazine hydrochloride were purchased from Sigma-Aldrich. A schematic representation of DPPC and drug is shown in Figure S1. A solution of chloroform (1.0 mL) containing DPPC and drug was dried under a stream of nitrogen. This was then placed in a vacuum desiccator vacuum for 12 h at ambient conditions to remove residual traces of the solvent. It was observed that there was no effect on whether the drug was added in chloroform before resuspension or dissolved in the resuspension medium. DPPC bilayers exist in the gel phase at temperatures below 35 � C, whereas above 42 � C they are present in the liquid crystalline phase. The nanoplasmonic sensing (NPS) methodology is a simple and valuable tool for the determination of the phase transition temperature (Tm) of phospholipids (Chen et al., 2018).
To obtain proper hydration, the lipid dispersion was repeatedly warmed to about 10 � C above the melting transition temperature, and allowed to cool below the Tm before starting the measurements. Dry drug/DPPC films were hydrated through the addition of 1 mL of water at a temperature 10 � C above the gel-liquid-crystalline phase-transition temperature of DPPC liposomes. The phase-transition temperature of DPPC liposomes is 41 � C. The very last concentration of DPPC turned into 10 mg/mL. We made DPPC solutions for different concentrations of CPZ drug. The dry DPPC/CPZ films were formed by spreading 10 ll drops of these solutions onto the hemispheric bore of an aluminum sample holder.
The Raman spectra were acquired in 200-3200 cm À 1 region with a Renishaw spectrometer equipped with a 350 MW, 785 nm diode laser, and 1200 lines per mm grating. The samples were measured in the hemispheric bore of an aluminum sample holder. Laser power of 28 mW was used to record Raman spectra. Renishaw, WiRE Version 3.2 was employed to map experiments.
The hot stage microscope system used for this study was a THMS600 stage system manufactured by Linkam Corp, UK. To have perfect heat transfer and particularly sensitive temperature measurement, samples were placed on a polished pure silver heating element loaded onto an aluminum crucible.

Raman microspectroscopy studies of drug-doped DPPC liposomes
We investigated in detail the Raman spectra changes in lipid vesicles in different regions for different DPPC/CPZ molar ratios (90/10, 80/20, 70/30, 60/40, 50/50 mol%) (Figure 1). We focus our discussion on two main spectral regions, namely region A (2800-3000 cm À 1 ) and region B (1000-1200 cm À 1 ). Region A mainly consists of methyl and methylene stretching modes, while region B includes CC stretching vibrational modes. The first important spectral region (Region A) has bands at 2935 and 2880 cm À 1 corresponding to CH 2 and CH 3 stretching vibrational bands. These CH 2 /CH 3 stretching modes are known to provide insights into the phase transition of hydrocarbons in phospholipid membranes (Bulkin & Krishnan, 1971). The important parameter here is the I 2935 / I 2880 intensity ratio values, which provide insights regarding I disorder /I order (Figure 2). The polymethylene chain state is known to govern by the peak heights I 2935 /I 2880 ratio in liquid crystals (Bulkin & Krishnamachari, 1972), and biomembranes (Snyder et al., 1978;Brown et al., 1973). The conformational disorder and insights about the lateral packing are also provided by this peak heights ratio I 2935 /I 2880 (Larsson & Rand, 1973). Adding CPZ to DPPC at room temperature results in a huge increase of the I 2935 /I 2880 ratio value as shown in Figure 2. We note that this becomes constant after 70/30 mol% (DPPC/CPZ). The peak intensity ratios of the 2935 cm À 1 and the 2880 cm À 1 features, I 2935 /I 2880 , displayed in Figure 2 reflect the measure of primarily the interchain disorder with some contribution from intrachain trans/ gauche conformer changes. Here, after adding CPZ to about 80/20 mol% (DPPC/CPZ) (Figure 2), we see a significant increase in the I 2935 /I 2880 ratio indicating the increase of interchain disorder of DPPC lipid and increase in cellular content of DPPC. This increase in intensity ratio can be seen as an increase in cellular content, where CPZ drug interaction results in an increased unsaturated/saturated lipid ratio. This is also representative of adaptive responses (Stuhne-Sekalec et al., 1987) as no further increase in I 2935 /I 2880 ratio value is observed with a further increase in CPZ after 70/30 mol%.
The second interesting region (Region B) is formed by C-C bonds stretching of phospholipids alkyl chains. The trans and gauche conformations of the alkyl chains CC stretching vibrations are reflected as peaks at 1129 and 1097 cm À 1 , respectively. The presence of C-C-C ring deformations in the CPZ molecule results in intensity change in this spectral region. This can be observed as C-C bond stretch motions related to the gauche conformations of the alkyl chains. We observe changes in the intensities of the peaks due to the presence of ring deformational modes of the drug molecule. The ratio of peak heights I 1097 /I 1064 and I 1097 /I 1129 are shown in Figure S2. As a membrane becomes more disordered, the gauche-to-trance peak intensity ratio increases and it is clear from the results in Figure S2. With the increase in the drug amount the intensity of the 1097 cm À 1 feature increases, while the intensities of the 1129 and 1064 cm À 1 transitions decrease.
Phospholipids C ¼ O stretching region in the Raman spectra is of immense importance to investigate the interfacial region molecular interactions (Brown & Brown, 1980). The peak at 1728 cm À 1 is attributed to the stretching vibration of the C ¼ O bond. The change of mC ¼ O group vibration is due to the hydrogen bond between chlorpromazine and ester groups of the DPPC and not from the changes in hydration of the polar head groups of the lipid at different concentrations of the drug.

Raman imaging experiments
To study the distribution of drugs in membranes, one of the most powerful and widely used techniques is Raman imaging. We used Raman imaging to study the drug distribution in DPPC vesicles. Figure 3 shows the results of a Raman imaging experiment on a DPPC/CPZ spectrum (70/30 mol%). Figure 3(A) presents the white light image of dried DPPC/ CPZ vesicles and Figure 3(B) shows the Raman image of ring CC stretching vibrations in CPZ drug at 1566 cm À 1 . In the Raman imaging figure red color corresponds to drug distribution inside the lipid vesicles. As a large step size was used, we couldn't get a good resolution however one can see the different distributions of the chemicals within the vesicles as we would anticipate. The vesicle consists mainly of DPPC and the distribution of CPZ within the boundary of the lipid vesicles can be observed.

Temperature-dependent Raman microspectroscopy
We collected Raman spectra at different temperatures (DPPC/CPZ (70/30 mol%)) to detect lipid phase transitions of a vesicle as shown in Figure 4. We observed considerable changes in the peak intensity ratios in the CH stretching region close to phase transition as the addition of the drug to the DPPC liposomes induced a second transition signal at a much lower temperature, indicating the presence of DPPC domains containing the drug. The cationic amphiphilic molecules intercalate because the area of the polar head groups offers a favorable atmosphere for them resulting in influencing phospholipid's bilayers phase transition temperature (Hanpft & Mohr, 1985). In the gel phase, these phospholipid molecules are packed into a highly ordered matrix with the hydrocarbon chains of fatty acids in an all-trans conformation with a limited degree of freedom below the phase transition temperature. When there is a transition into the liquid crystalline state, each hydrocarbon chain needs additional lateral space, and hydrocarbon chains gain more freedom of movement, some C-C bonds adopt gauche conformations, and the thickness of the bilayer decreases, and the bilayer expands. The phospholipid membrane's phase transition temperature is not only dependent on the properties of the fatty acid chains but also on the nature of the polar head groups (Kursch et al., 1983). The degree of hydration of the head groups (Chapman et al., 1967) or interaction with the divalent cation (Chapman et al., 1974) has been found to influence the phase-transition temperature in the case of  DPPC lipids. Recent studies have engaged in this technique to observe chemical reactions in phosphatidylcholine vesicles (Fox et al., 2007;Fox et al., 2006). A substantial residuum seemed after the increase in temperature ( Figure S3) and the existence of this residuum indicates that Raman spectra of the reported mixtures differ from the possible spectral characters predicted from the spectra of DPPC/CPZ at 0 � C.

DFT and MD simulations
As stated in the computational methodology section, we performed DFT and MD simulations to investigate the interaction between CPZ and DPPC, where we tried different configurations to identify the best interaction possible between the drug and the lipid. The most stable DFT computed structure is shown in Figure 5. MD simulations helped us to understand the dynamics and confirmed the most stable configuration possessing lowest energy ( Figure 6).
We observe an interaction between the positively charged nitrogen atom of CPZ and the negatively charged phosphate moiety of DPPC, where methylene and methyl groups connected with the N atom in CPZ form bonds with the phosphate group of DPPC (O-H bonds with a bond length of 2.28 and 2.52 � Å).We further calculated the Raman spectra of this complex and a comparison with the experimental spectrum is shown in Figure 7. We found reasonable agreement between DFT computed and experimental Raman bands, where bands in the higher region (�3000 cm À 1 ) are found to be associated with C-H stretching vibrations, band at 1729 cm À 1 in the experimental spectra matches with the calculated band at 1722 cm À 1 and is due to C ¼ O stretching vibrations. Bands in the frequency range around 1000-1100 cm À 1 are found to be associated with vibrations related to C-C-C skeleton motions. This is in accordance with our discussion of bands in Section 4.1.

Conclusions
Drug repositioning is always beneficial reducing many steps in the new drug discovery process. Understanding the molecular mechanism of melting of lipid is a challenging question. Here, the Raman image results show that CPZ distribution is within the border of the lipid vesicles. Through temperature-controlled Raman microspectroscopic studies, we found that the fatty acid chains of the phospholipids are stiff and packed in a highly ordered array below phase-transition temperature, while the chains are disordered with motional freedom above the phase transition temperature. Thus, we conclude that Raman scattering has given a detailed fingerprint spectrum of the DPPC chemical composition and insights into its structural parameters. It provides the most definitive means of identifying membrane structure and composition, bilayer assemblies, and membrane behaviours. Our combined DFT-MD simulations and experimental physical methods employed in the present work have provided valuable information regarding the interactions between the CPZ and DPPC and the drug's exact location in the lipid bilayers. CPZ drug has already been used in human patients with a safety track record and shown to be effective against various ailments, our study further can be used to develop and test in human patients suffering from the novel coronavirus disease.

Acknowledgments
Soni Mishra thanks Graphic Era Hill University for SEED financial support and Abhishek Kumar Mishra acknowledges the SEED grant (Year 2021) from UPES, Dehradun for computational resources.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The author(s) reported there is no funding associated with the work featured in this article.