Monoterpene-Containing Substituted Coumarins as Inhibitors of Respiratory Syncytial Virus (RSV) Replication

Respiratory syncytial virus (RSV) is a critical cause of infant mortality. However, there are no vaccines and adequate drugs for its treatment. We showed, for the first time, that O-linked coumarin–monoterpene conjugates are effective RSV inhibitors. The most potent compounds are active against both RSV serotypes, A and B. According to the results of the time-of-addition experiment, the conjugates act at the early stages of virus cycle. Based on molecular modelling data, RSV F protein may be considered as a possible target.


Introduction
Respiratory syncytial virus (RSV), which belongs to the Pneumoviridae family, is an enveloped negative-sense RNA virus with two major serotypes, A and B [1]. RSV infects the respiratory tract, causing annual epidemics during the cold season. Despite a relatively low variability of this virus, the immunity to it is unstable, which may cause repeated infections of the same individual throughout life. It usually causes a cold. Patients under the age of 2 years, especially those born prematurely or with a heart condition, as well as the elderly, develop different symptoms of respiratory syncytial infection [2][3][4]. These age groups present with involvement of not only the upper respiratory tract, but the lower respiratory tract as well, and develop severe bronchiolitis and pneumonia, which can lead to death. RSV is the most common cause of bronchiolitis and pneumonia in children younger than one year of age [5]. According to a meta-analysis carried out in 2010, the number of children who may die from this disease is estimated to be about 199,000 annually [6]. RSV-associated childhood respiratory illness has become a challenge since the summer of 2021, when the number of cases increased sharply, which may have been due to the relaxation of COVID-19 quarantine measures [7]. There is no vaccine for RSV. Therapy for respiratory syncytial infection is usually symptomatic. The only treatment option is a non-specific and poorly effective antiviral agent: ribavirin [8,9]. In addition, the monoclonal antibody Palivizumab is approved for prophylaxis, but it is expensive and only moderately effective at reducing hospitalization rates.
In recent decades, there has been significant progress in the identification of potential targets for RSV therapy and the search for low-molecular-weight inhibitors of RSV replication; the results of these studies were summarized in a review published Many coumarin derivatives display a variety of biological activities [16][17][18][19][20], in particular antiviral activity [21][22][23][24][25][26][27]. Tetrahydroisoquinoline 4 is the only coumarin derivatives which exhibit anti-RSV activity [27]. Attachment of monoterpenoid fragments to parent molecules is known to significantly enhance their antiviral activity [28,29]. Recently, we have found that monoterpene-containing substituted 7-hydroxycoumarins effectively inhibit H1N1 influenza virus replication, with compound 5b being most active [30]. Conjugate 5a was found to exhibit the highest activity when added to infected cell culture at early stages of viral reproduction, probably due to the interaction with viral hemagglutinin. Replacement of the monoterpenoid moiety with a benzyl substituent led to a complete loss of antiviral activity. However, there are no data on RSV inhibitory activity of coumarin derivatives comprising a terpene moiety. In this study, we revealed the ability of monoterpene-coumarin conjugates to inhibit RSV replication and suggested a possible mechanism of their action.

Chemistry
Our earlier study on the activity of monoterpene-coumarin conjugates against the influenza virus revealed that the structure and absolute configuration of a monoterpene moiety and the size of an annulated aliphatic ring had a significant effect on the antiviral activity. In addition, as the length of an aliphatic bridge between monoterpene and coumarin moieties increases from one to two CH 2 -groups, the antiviral activity enhances [30]. Given these data, in this study we synthesized a library of compounds that included both previously obtained coumarin-monoterpenoid hybrids [30,31] and new compounds with coumarin and monoterpene bicyclic fragments separated from each other by three or four CH 2 -groups, as well as a number of nitrogen-containing coumarin derivatives with an NH 2 -group instead of an OH-group. 7-Hydroxycoumarin derivatives were prepared from commercially available umbelliferone 6 and 4-methyl-7-hydroxycoumarin 7, as well as coumarins 9 and 10 (Scheme 1) synthesized from resorcinol 8 according to [31]. Bromides 11a-c were synthesized in the reaction between appropriate alcohols and PBr 3 in accordance with [31], using geraniol, (-)-myrtenol, and (+)-myrtenol produced from (+)-α-pinene as starting monoterpenoids (Scheme 2). Nopol bromide 11d was previously obtained by the method presented in [31] in a low yield; therefore, the NBS/PPh 3 system [32] was used to synthesize compound 11d. Scheme 2. Synthesis of monoterpenoid bromides 11a-d.
Derivatives of (-)-α-pinene and its homologue (-)-nopol exhibited the highest activity against the influenza virus; therefore, we synthesized compounds containing an extended aliphatic chain and an (-)-α-pinene moiety. A homolog of nopol bromide 11d, which contained an additional CH 2 -group, was synthesized according to the following scheme: (+)-Trans-pinocarveol, which was prepared according to [33], reacted with triethyl orthoacetate in the presence of hexanoic acid to form ester 12, which was then reduced by LiAlH 4 to alcohol 13 [34]. Bromide 11e was prepared by treating alcohol 13 with NBS/PPh 3 .

Scheme 3. Synthesis of bromides 11e-h.
Aurapten 16a and other coumarin derivatives 16-19 were prepared in the reaction of 7-hydroxycoumarins 6-10 with appropriate bromides 11 (Scheme 4), as described elsewhere [30]. The products were purified by recrystallization or column chromatography (yields 24-80%). The reaction of nopyl bromide 11d with methylcoumarin 7 failed due to the formation of a complex reaction mixture with a high resinification level. 7-Aminocoumarins 23 and 24 were synthesized according to the method presented in [36], starting from 3-aminophenol 20, via intermediate compound 21, and its interaction with the appropriate ketoesters (Scheme 5).

Cytotoxicity Test
The cytotoxicity was tested using a standard MTT-test (detailed below in Section 3.2.1). A series of 2-fold dilutions of compounds was made; then, each dilution was added to 24 h old cell monolayer, after which, cells were incubated for 24 h. Cell viability was assessed by adding the MTT solution. Based on the data obtained, the CC 50 was calculated.

Antiviral Activity
The antiviral activity against the respiratory syncytial virus was assessed by adding a series of 3-fold dilutions of test compounds, followed by addition of the virus in a series of 10-fold dilutions. Cells were incubated for 1 h; then, the virus was washed out and compounds were added again. Cell were incubated for 6 days, after which the viral presence was investigated, using the ELISA method.
The virus titer was calculated using the Reed and Muench method.
The obtained results are shown in Table 1. Compounds were considered promising with a selectivity index of 10 or higher. At the first stage, all the prepared compounds were tested for their ability to inhibit the replication of RSV A. Aurapten 16a, containing an acyclic monoterpene substituent, exhibited high activity against RSV A in the lower micromolar range, but its selectivity index was less than 8 due to high cytotoxicity. Among compounds 16b,c containing an α-pinene fragment, only compound 16c, derived from (+)-α-pinene, exhibited noticeable activity, with the selectivity index exceeding 11. Elongation of the hydrocarbon chain from 16b to compounds 16d-f led to a nonlinear change in the biological properties: high antiviral activity was exhibited by compounds with one (16d) and three (16f), but not two (16e) additional CH 2 -groups. Due to its lower cytotoxicity, compound 16f displayed the best selectivity index, about 65, among monoterpene-coumarin conjugates of this type. Coumarin derivative 16g, containing an aromatic substituent, showed a high selectivity index (30), which was related to its low cytotoxicity, rather than its high activity.
Among 4-methylcoumarin derivatives 17, compound 17c, containing an (+)-α-pinene moiety, exhibited high activity in the submicromolar range, whereas its (-)-isomer was 20-fold less active. Compound 17c had a high selectivity index of 90. Elongation of the hydrocarbon chain (a compound with one additional CH 2 -group was not prepared due to the formation of a complex reaction mixture) did not affect the activity much; it was almost the same, similarly to toxicity. The less cytotoxic compound 17c had a selectivity index of 62. Compound 17g, containing an aromatic substituent, did not display significant antiviral activity.
Investigation of the biological properties of compound 18, comprising a cyclopentane ring annulated with a coumarin moiety, showed that (+)-α-pinene derivative 18c with a selectivity index of 100 exhibited the highest activity and moderate toxicity. The other compounds were either toxic (18a,b,e-g) or inactive (18d); only compound 18e had a noticeable selectivity index of 18.
Among coumarin 19, containing an annulated cyclohexane ring, the highest activity was exhibited by compounds 19a and 19d, which were derived from geraniol and nopol, respectively. High selectivity indices were also displayed by compounds 19c and 19f, which were both less active and less toxic. Coumarin derivative 19g, containing a methoxybenzyl substituent, had a selectivity index of more than 10, which was due to its low cytotoxicity. Removal of the methoxy group from the aromatic ring in compound 19h led to a sharp increase in its toxicity.
Amine 25 showed significant activity and moderate toxicity, which resulted in a selectivity index of more than 10. Transition to acetamide 26 led to a loss of the antiviral effect. Compound 27 was almost one order of magnitude more active and significantly more toxic than amine 25, with the selectivity index being about 15. Compound 27 was much more active than its oxygen-containing analog 19b, which substantiates further research regarding the synthesis of N-linked monoterpene-coumarin conjugates and the investigation of their anti-RSV activity.
A significant portion of the prepared coumarin derivatives were tested for their ability to inhibit RSV replication. Among the compounds tested, 16c,g, 17c,f, 18c, 19a,c,d, and 19f had a selectivity index of more than 10, with the last two being most active. In general, activity of O-linked coumarin-monoterpene conjugates against RSV type B was similar to that against RSV type A, which indicates that the antiviral activity of these compounds lacks type specificity. At the same time, 7-aminocoumrine derivatives demonstrated some activity against RSV type A, but not type B. Among compounds with similar SI against both RSV types, 19c was chosen to investigate a possible action mechanism.
To elucidate action mechanisms of the most active compounds, we performed an experiment using the time-of-addition method. Compound 19c was added to the cells with replicating virus at different time points corresponding to different stages of viral life cycle. The results are shown in Figure 2. According to the data presented in Figure 2, compound 19c reduces the virus titer at all time points, except for prophylactic use 2 h before virus entry into cells and late introduction after 24 h. This means that action of the compound occurred within 0-6 h after infection. A slight decrease in the virus titer was also noted at time points 2 and 4. Our data suggest that the target of 19c may be the surface F and/or G proteins, viroporin SH, and, probably, L protein (its fragment responsible for transcription).

Molecular Modeling Study
According to the results of the time-of-addition experiment, the surface F protein may be considered a potential biological target. In addition, the pharmacophore features of the inhibitor of F-protein sisunatovir and compounds 19c, 19f, and 19h are similar. In all cases, there are hydrophobic parts capable of hydrophobic intermolecular contacts, and donors and acceptors of hydrogen bonds (more details are presented in the Supplementary Materials).
We suggest that compounds 19c, 19f, and 19h can bind to a fusion peptide region [38,39] ( Figure 3A), similarly to sisunatovir [15] ( Figure 3B). Several functional amino acids are located in the binding site of potential entry inhibitors ( Figure 3A). The red segment in Figure 3 is a part of the fusion peptide (137-140 amino acids), and the yellow segment corresponds to amino acids of the membrane anchor. Entry inhibitors can interact with these amino acids and inhibit the conformational transition from a pre-to post-fusion conformation. The coumarin moiety of compounds 19c, 19f, and 19h occurs in a cavity surrounded by phenylalanines and forms π-π stacking interactions with them ( Figure 3C,D). The terpene fragment of 19c and 19f occurs in a hydrophobic cavity composed of amino acids of the fusion peptide region ( Figure 3C,E). The aromatic part of 19h cannot be placed into the cavity due to a short linker between it and the coumarin fragments ( Figure 3D). Compound 19h forms intramolecular interactions with amino acids of the membrane anchor. Docking scores shown in Figure 3 may be considered as a parameter that characterizes the affinity of compounds to the binding site. The best positions are shown in Figure 3. The best ligand position was chosen based on the clustering energy and formation energy of the ligand-protein complex. For more energy parameters, see the Supplementary Materials. Affinity values of lead compounds have a slightly higher value than those for sisunatovir. Compound 19h had the lowest affinity. These results correlate with experimental data (Table S1). Thus, RSV F protein may be considered as a possible target.

General Chemical Methods
Reagents and solvents were purchased from commercial suppliers (Sigma-Aldrich, Acros, Japan) and used as received. GC-MS: Agilent 7890A gas chromatograph equipped with an Agilent 5975C quadrupole mass spectrometer as a detector; quartz column HP-5MS (copolymer 5%-diphenyl-95%-dimethylsiloxane) of length 30 m, internal diameter 0.25 mm and stationary phase film thickness 0.25 µm. Optical rotation: polAAr 3005 spectrometer. 1  Spectral and analytical investigations were carried out at the Collective Chemical Service Center of the Siberian Branch of the Russian Academy of Sciences. All product yields are given for pure compounds purified by recrystallization from ethanol or isolated by column chromatography (SiO 2 ; 60-200 µ; Macherey-Nagel). The purity of the target compounds was determined by GC-MS methods. All of the target compounds reported in this paper had a purity of no less than 95%.
(+)-Myrtenol was synthesized from the corresponding aldehyde via reduction to alcohols with NaBH 4 , as described above. NaBH 4 (10.3 mmol) was added to a cooled (0-5 • C) solution of 10.3 mmol of the appropriate aldehyde in methanol (20 mL), and the reaction mixture was stirred for 3 h at room temperature. Then, 5% aqueous HCl was added to obtain a pH of 4-5. The solvent was distilled, and the product was extracted using ether and dried with Na 2 SO 4 . The solvent was evaporated; the resulting alcohol (54% yield) was used in the synthesis without purification.
Bromide 11a was synthesized from geraniol via the reaction with PBr 3 [30]. PBr 3 (8.9 mmol) was added to cooled (0-5 • C) solution of geraniol (26.7 mmol) in dry ether (30 mL), and the reaction mixture was stirred for 2 h at r.t. Saturated aqueous NaHCO 3 was added, and the product was extracted with ether. The extracts were washed with brine, dried with Na 2 SO 4 , and evaporated.
Additional bromides 11b,c,g were synthesized as described above. Compounds 11a-c,g (with yields of 91%, 55%, 60% and 65%, respectively) were sufficiently pure and used for the next step without purification.
Triphenylphosphine (2.0 equiv., 6.1 g, 23 mmol) was dissolved in dry DCM (23 mL) under argon. N-bromosuccinimide (NBS) (2.0 equiv., 4.2 g, 23 mmol) was added to this solution in small portions over 5 min in an ice-water bath. Subsequently, the resulting deep red mixture was stirred at room temperature for 30 min. Then, pyridine (1 mL) was added, and the color turned reddish-brown. (-)-Nopol (1.0 equiv., 2.0 mL, 12 mmol) was added to the mixture dropwise over 10 min. The reaction mixture was stirred overnight at room temperature. Later, the mixture was diluted with hexane (40 mL) and filtered through a silica gel plug. Then, the reaction flask was stirred with EtOAc-hexane (1:1, 40 mL) and filtered through the silica gel plug. Later, it was concentrated in vacuo and the crude residue was purified by chromatography on SiO 2 (hexane) to obtain bromide 11d (2.3 g, 70% yield).
3.1.7. Synthesis of Compounds 16a-g, 17a-c,e-g, 18a-g and 19a-g Compounds 16a-d,g, were synthesized from coumarin 6 and the corresponding bromides 11a-d,g with the use of DBU and DMF, in accordance with [30].
DBU (1.0 mmol) and the corresponding bromides 11a-d,g (0.75 mmol) were added to compound 1 (0.5 mmol) in dry DMF (5 mL) at r.t. under stirring. The reaction mixture was stirred at r.t. for 15 min, and then heated at 60 • C for 5 h. H 2 O (15 mL) was added, and the product was extracted with ethyl acetate. The extracts were washed with brine, dried with Na 2 SO 4 , and evaporated.
Methoxycarbonyl chloride (3.6 mL 47 mmol) was added dropwise to a cooled (5-10 • C) suspension of m-aminophenol 20 (4.4 g, 40 mmol) and K 2 CO 3 (3.5 g) in 35 mL of ethyl acetate and 3 mL of water, with vigorous stirring. The mixture was stirred for 1 h; then, 10 mL of water was added, and the mixture was stirred for another 3 h. The product was extracted with ethyl acetate. The extracts were washed with water, 1 M H 2 SO 4 , and brine, dried with Na 2 SO 4 , and evaporated. The resulting solid was crystallized from benzene to give 5.7 g of 21 (77%).
A mixture of compound 21 (4.6 g, 28 mmol) and 5.5 mL acetoacetic ester was added dropwise to 12 mL H 2 SO 4 with vigorous stirring. The mixture was stirred for 2 h and diluted with 50 mL of ice-water. The precipitate was removed by filtration, washed with water, MeOH, and ether, and dried to give 4,0 g of 22 (61%).
A suspension of 2.8 g (12 mmol) of compound 22 in 6 mL of 45% KOH solution was stirred at 90 • C for 0.5 h until the solution formed. The mixture was cooled and diluted with water and acidified with concentrated HCl to pH 5-6. A solution of alkali was added to the suspension, to obtain pH 8. The mixture was stirred until crystallization ceased. The precipitate was removed by filtration, washed with water, MeOH, ether and dried to give 1.53 g of 23 (73%).
Similarly, compound 24 was synthesized from compound 21 with a yield of 64%.
Compound 25 (0.097 g,1.0 mmol), (−)-nopinal (0.112 g 1.2 mmol) (synthesized by the oxidation of (−)-nopol with IBX according to the procedure [43]) and acetic acid (100 µL) were dissolved in methanol (5 mL) and stirred at room temperature for 2.5 h. NaBH 3 CN (0.110 g, 2.0 mmol) was added, and the reaction mixture was stirred at room temperature for 1.5 h. Methanol was evaporated and the reaction mixture was extracted with CH 2 Cl 2 . The organic layer was washed with brine, dried over anhydrous Na 2 SO 4 , filtered, and evaporated. The residue was crystallized from ethanol to give 0.115 g of 25 (64%).

Cytotoxicity Test
The compounds were weighed in amounts of 2 mg and dissolved in 100 µL of DMSO. Then, the resulting solution was adjusted with the medium to a concentration of 1000 µg/mL, and a series of twofold dilutions was prepared from it. One-day culture of HEp2 cells, grown in 96-well plates, cell concentration 3 × 10 5 /well of the plate, was checked visually in an inverted microscope for the integrity of the monolayer. Plates were selected for work where the cell closure was 60-80%.
Dilutions of the compounds at the appropriate concentration were added to the plate in a volume of 100 µL in each well in two replicates for each tested concentration. The plates were incubated for 24 h at 37 • C in the presence of 5% CO 2 . Cell viability was assessed using the MTT assay.
The MTT solution was prepared on a maintenance medium at a concentration of 0.5 mg/mL. Then, 0.1 mL of MTT solution was added to each well. After 1.5 h of MTT contact at 37 • C at a CO 2 concentration of 5%, MTT was discarded with the cells of the well and 0.1 mL of ethyl alcohol 96% was poured, after which the optical density in the wells was measured at a wavelength of 535 nm. Based on the data obtained, the CC 50 was calculated.

Antiviral Activity
The antiviral activity against the respiratory syncytial virus (RSV A-strain A2, RSV Bstrain 9320) was assessed in a series of threefold dilutions of test compounds, starting from 1 2 CC 50 , which were added to HEp-2 cell culture at a double concentration, at 100 µL per well, followed by the addition of 100 µL of the virus in a series of 10-fold dilutions. Cells were incubated at 37 • C and 5% CO 2 for 1 h. Then, the virus was washed out, and the compounds were again added at a single concentration and incubated at 37 • C and 5% CO 2 for 6 days. For the enzyme-linked immunosorbent assay (ELISA), cell culture was fixed with cold 80% acetone at -20 • C for 15 min and then washed with phosphatebuffered saline containing 0.05% Tween 20. Next, a solution of primary mouse anti-RSV F protein antibodies was added to the culture and incubated at room temperature under continuous stirring for 2 h. Then, cells were again washed with buffer, secondary antimouse antibodies were added, and the cells were incubated under continuous stirring for 2 h. Then, the antibodies were washed off, and a substrate-chromogenic mixture with tetramethylbenzidine was added. After 5 min, the reaction was stopped with 0.1 M sulfuric acid, and the optical density of the solution was measured at a wavelength of 450 nm. Wells with absorbance values twofold or greater than the cell control were considered contaminated. The virus titer was calculated using the Reed and Muench method. All experiments were made in triplicate.

Time-of-Addition Assay
Compound 19c was added at different time points before, after or simultaneously with the introduction of the virus. The time of addition of the compound was counted from point 0-the time of entry of the virus into the cell. During the period (−1)-0, the cells together with the virus were incubated at 40 • C. All other experiments were carried out at 37 • C. RSV virus A 2 mL was added to the cells at a time that was conventionally designated as point −1, after which the cells were kept for an hour at a temperature of 40 • C. Then, at point 0, the virus was unbound. The cells were transferred to a thermostat at 37 • C, where they were incubated for 25 h. After this period, the medium was taken from each well and a series of ten-fold dilutions were made on a fresh cell culture and incubated for 6 days. For each compound, 2 repetitions were made by different operators. The virus titer was estimated by ELISA. The compound was added at the following times relative to the addition of the virus: point−2-the compound was introduced one hour before cell infection (prophylactic regimen); point 0-at the moment of temperature change; points 1, 2, 4, 6, 24-after 1, 2, 4, 6 and 24 h after the temperature change, respectively. In the wells marked (−2) − (25), the compound was kept throughout the experiment, starting from point −2 and until the end of the experiment −25 h. No compound was added to the control wells; instead, a similar volume of medium was added.

Receptor and Ligand Preparation
Crystallographic structures of the RSV F protein (PDB codes 7LVW [39] and 7KQD [15]) were downloaded from the Protein Data Bank database [40]. The code of the full-length protein trimer was 7LVW; 7KQD is the code of a ligand-protein complex, but only one protein from three trimeric forms. Model protein structures were prepared using the Schrodinger Protein Preparation Wizard tool (Schrodinger Suite Software): hydrogen atoms were added and minimized; missing amino acid side chains were added; bond multiplicities were restored; solvent molecules were removed; and the entire structure was optimized in the OPLS3e force field [44] at a physiological pH value. For a correct calculation procedure, we used binding site alignment procedures 7LVW and 7KQD. As results, full-size proteins and ligands (sisunatovir) in complex were obtained.
The geometric parameters of ligands (coumarin derivatives) and sisunatovir (RSV F inhibitor) were also optimized, taking into account all permissible conformations.

Molecular Docking
Molecular docking was performed using Schrodinger Suite (Release 2020-4) software. Coumarin derivatives were docked using the forced ligand positioning protocol (IFD) under the following conditions: flexible protein and ligand; grid matrix size of 15 Å; and amino acids (within a radius of 5 Å from the ligand) restrained and optimized, taking into account the influence of a ligand. Docking solutions were ranked by evaluating the following calculation parameters: docking score (based on GlideScore minus penalties); ligand efficiency (LE, which takes into account an atomic distribution of the scoring function); a model energy value (Emodel), including a GlideScore value, energy-unrelated interactions, and parameters of the energy spent in positioning of the ligand in the binding site.
Supplementary Materials: NMR 1 H and 13 C spectra of compounds 16-19, 25, and 27; Energy parameters of the docking study procedure; Figure S1-best position of Sisunatovir; Figure S2-best position of 19c; Figure S3-best position of 19f; Figure S4-best position of 19h; Figure S5-the pharmacophore features of known F-protein inhibitor of sisunatovir and compounds 19c, 19f, and 19h; 1. Figure S6-Dose-response curve and half maximal inhibitory concentration (IC 50 ) values of active compounds in Hep-2 cells against RSV A and B. Table S1-Energy parameters of docking study procedure.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.