Synthesis, characterization and evaluation of antidengue activity of enantiomeric Schiff bases derived from S-substituted dithiocarbazate

A series of Schiff bases have been successfully synthesized through the acid-catalyzed condensation of S-substituted dithiocarbazates and three enantiomerically pure monoterpenes, (1 R )-(+)-camphor, (1 S )-(-)-camphor, (1 R )-(-)-camphorquinone, (1 S )-(+)-camphorquinone, ( R )-(-)-carvone and ( S )-(+)-carvone. Spectroscopic results revealed that the Schiff bases containing camphor or carvone likely adopted an E -configuration along the characteristic imine bond while those containing camphorquinone assumed a Z -configuration. The antidengue potential of these compounds was evaluated based on DENV 2 caused cytopathic effect (CPE) reduction-based in vitro evaluation. The compounds were validated through secondary foci forming unit reduction assay (FFURA). Compounds were also tested for their cytotoxicity against Vero cells. The compounds showed variable degrees of antiviral activity with the camphor compounds displaying the highest antidengue potential. The enantiomers of the compounds behaved almost similarly during the antiviral evaluation.

in the data set. The molecular structures and their packing diagrams were generated using ORTEP [22], and DIAMOND [23] respectively while the crystal data were analyzed by PLATON [24]. Details of unit cell data, X-ray data collection, and structure refinement are given in Table 1.

General procedure for the preparation of compounds 5-6 and 11-12
The synthesis was similar to the rest of the compounds except that 1 mL of diluted HCl (1 M) was used and the reaction was conducted at room temperature.
Ribavirin tablets (Copegus, 200 mg, Genentech, Inc., South San Francisco, CA, USA) were purchased and stock solutions were stored. Dilutions were prepared in 2% FBS (maintenance media) at the time of assay. TCID50 assay was employed to measure the titer of the virus stocks for antiviral evaluation following standard methods [25]. The viral CPE was observed on the third and fourth days and observations were recorded based on the Reed-Muench method.

MTT cytotoxicity assay
The 96-well plates were seeded with 100 µL of Vero cell suspension at a density of 1 × 105 cells / mL and the next day were treated with different dilutions of the synthesized compounds. Control wells were treated with only the vehicle media used to prepare the antiviral dilutions. Plates were incubated for four days at 37 °C. Following the addition of 15 μL of MTT (Sigma-Aldrich Corp., St. Louis, MO, USA) solution into each well, the plates were incubated at 37 o C for another 4 h. Subsequently, the medium was carefully aspirated from all the wells and 100 μL of DMSO was added to each well [26]. The optical density (OD) of the wells was measured at 570 nm using a 96-well plate reader (Tecan Group AG, Männedorf, Switzerland). The percentage cell viability was calculated against the untreated controls. Dose-response curves were plotted. Half maximal cytotoxic concentration (CC50) was calculated using Graph Pad Prism for Windows, Version 5 (GraphPad Software Inc., San Diego, CA, USA).

CPE reduction-based antiviral assay
Plates were seeded with Vero cell suspension (100 µL) at a density of 1 × 105 cells/mL one day before the wells were treated with TCID50 (100 µL) dilution of the DENV 2 stock and untreated control wells were maintained in parallel. After 1 h of viral adsorption, unbound viruses were washed and different antiviral dilutions were added. These plates were further incubated for 4 more days and were observed daily [26,27]. Observations were recorded on the third and fourth days. The wells were marked for CPE reduction according to the grading system defined by Kudi and Myint [28].

Foci forming unit reduction assay (FFURA)
Antiviral activity was quantitatively validated by measuring the reduction in the number of DENV infectious foci after 4 days of treatment. DENV 2 infected Vero cells were treated with different antiviral dilutions supplemented with 2% FBS and 1.5% carboxymethyl cellulose (CMC). Virus foci were stained and visualized according to the published protocol [29]. The number of foci was counted through stereomicroscope and virus titer was stated as foci forming unit (FFU). To calculate the antiviral activity of antidengue compounds, the percentage reduction in foci (%RF) was calculated between the treated, and untreated wells maintained in parallel. The assay was repeated three times, dose-response curves were plotted and half minimal inhibitory concentration (IC50) and selectivity index (SI) were calculated. 3.1. Chemistry  3.1.1. Synthesis and spectroscopic characterization of compounds 1-4, and 7-10 Conventionally, the formation of Schiff bases involving SMDTC or SBDTC is rather facile without much reliance on factors such as temperature, pH, and catalytic reagents [30]. In this study, however, the condensation reaction did not take place even after prolonged reflux at elevated temperature. A plausible reason for the failure of the reaction is that the structurally rigid bicyclic ring of the monoterpenes (camphor and camphorquinone in particular) coupled with the weakened dipole moment and reduced electrophilicity of the carbonyl carbon might have inhibited the reactivity by preventing nucleophilic attack from the electronegative S-substituted dithiocarbazates. One possible way to circumvent the inertness of the reaction is through capitalization of the electro negativity of the carbonyl oxygen by introducing a Lewis acid such as hydrogen ion to activate the nucleophilic oxygen.

Results and discussion
IR analysis shows the presence of some essential peaks at about 3150-3250 cm -1 and 1605-1660 cm -1 which can be attributed to secondary νN-H and νC=N peaks. The difference between the νN-H peaks and the two distinctive peaks characteristic of asymmetric and symmetric stretching of primary amines as observed for SM-and SBDTC, as well as the absence of νC = N in the spectra of the precursors imply that condensation has indeed taken place between the dithiocarbazates and monoterpenes. While the disappearance of νC=O frequency in SMRCM, SMSCM, SBRCM, and SBSCM is further evidence which confirms the formation of the Schiff bases, the peak is still observed at 1706-1713 cm -1 in the corresponding spectra of SMRCQ, SMSCQ, SBRCQ, and SBSCQ and it is more resolved compared to the broad asymmetric C=O peak of the starting precursor (i.e. camphorquinone) at ca. 1745 cm -1 . This suggests that only one of the C = O groups reacted with the substituted dithiocarbazates, while the other C=O which is likely the one adjacent to the methyl group, remains intact due to steric hindrance. νC=S at 1030-1100 cm -1 together with the absence of νS-H (2600-2700 cm -1 ) are clear evidence that the solid Schiff bases are predominantly in their thione tautomeric form [31]. 1 H NMR spectra in DMSO-d 6 show thionamide N-H peaks for SMRCM, SMSCM, SBRCM, and SBSCM at 11.98-12.04 ppm, which is within the typical range for E-configuration along the C=N imine bond [32]. SMRCQ, SMSCQ, SBRCQ, and SBSCQ Schiff bases exhibit a resonance for the thionamide proton between 12.71-12.81 ppm indicating Z-configuration (Figure 1a). Such configuration enables an intramolecular interaction between thionamide, and unreacted carbonyl groups resulting in the relatively high field chemical shift for the thionamide proton ( Figure 1b). As expected, all Schiff bases remain in the thione tautomeric form even in a polar solvent like DMSO with no signal assignable to thiol proton being seen near 4 ppm [32]. The thione form of a dithiocarbazate Schiff base has been found to be more stable than its thiol counterpart by 14.5 kJ/mol through gas phase DFT calculation [33]. For the rest of the protons and carbon nuclei, their resonances appear in the typical range, and assignment has been unambiguously determined upon comparison with the 1 H and 13 C{ 1 H} NMR spectra of the corresponding precursors (cf. Section 2.3. and 2.4.).

Synthesis and spectroscopic characterization of compounds 5-6, and 11-12
Since the enantiomeric carvones did not react with the substituted dithiocarbazates under the heat-and-reflux routine commonly used, acid was used to catalyze the reaction. However, the additional alkene functional group in carvone makes it susceptible to acid addition leading to undesired stereoisomerization or racemization of the enantiomerically pure reagents (see Scheme 2).
To reduce the possibility of the occurrence of the undesired side reactions, a catalytic amount of dilute acid was used and, in addition, the synthesis was conducted at room temperature even though the condensation reaction, being an endothermic process, would have been favored by higher temperatures [34]. IR spectra of the carvone-DTC Schiff bases possess a profile similar to that of their camphor-and camphorquinone-DTC counterparts implying that the Schiff bases were successfully synthesized through acid-catalyzed condensation. More importantly, the presence of ν(C=C) at ca. 1582 cm -1 together with the absence of an intense ν(C-Cl) peak (~700 cm -1 ) prove that the acid addition reaction did not take place.
The thionamide protons in these compounds exhibit a resonance between 12.38 and 12.43 ppm which is close to those observed for the camphor-DTC Schiff bases indicating that they too adopt an E-configuration (Figure 2) along the C=N bond [32]. However, given that the substituted dithiocarbazate fragment is sterically less hindered about the N-N single bond compared to the bulkier camphor-DTC Schiff bases, the possibility of the formation of the Z-configuration, generated by flipping towards either side of the monocyclic ring of carvone, cannot be totally omitted. These compounds also display no sign of thione-thiol tautomerism in DMSO solution. The intactness of the alkene groups in the Schiff bases is confirmed by the 13 Figure 3 shows the ORTEP diagrams of SMSCM, SMSCQ, and SMRCV in which SMSCM (Figure 3a  the appending methyl, and isocyclic ring. The central dithiocarbazate fragment for all structures is essentially planar as shown through the corresponding least-square plane fitting with a r.m.s. deviation of 0.0064 Å and 0.0123 Å for molecule I [S2-C11(S1)-N2-N1] and molecule II [S4-C23(S3)-N4-N3] in SMSCM, 0.0447 Å for SMSCQ as well as 0.0154 Å and 0.0347 Å for molecule I and II in SMRCV. Molecules I and II in SMSCM are slightly differed from each other in that the thiomethyl plane (S1-C11-S2-C12) of the former is relatively less twisted from the central dithiocarbazate plane [S2-C11(S1)-N2-N1] with a dihedral angle of 3.6(1)° as compared to 5.8(1)° for the equivalent planes [S3-C23-S4-C24// S4-C23(S3)-N4-N3] in molecule II. The superimposition of the two molecules results in a r.m.s. deviation of merely 0.073 Å. As for SMRCV, molecules I and II are relatively more deviated from each other with a r.m.s. deviation of the superimposed structures being 0.796 Å. Overall, the least-square plane of thiomethyl (S1-C11-S2-C12 or S3-C23-S4-C24) and isocylic ring (C1-C2-C3-C4-C5-C6 or C13-C14-C15-C16-C17-C18) are twisted from the central dithiocarbazate plane by 14.7(1)° for molecule I and 10.7(1)° for molecule II, respectively.
In the terms of supramolecular features, the crystal packing of SMSCM and SMRCV are mainly governed by pairwise N-H···S interactions between the molecules in the asymmetric unit leading to the formation of an eight-membered {···NHCS} 2 homosynthon which is replicated by 2 1 symmetry along all crystallographic directions for the former and along the b-plane for the latter, with both showing no directional interactions between those pairs of molecules ( Figure  4a). In contrast, SMSCQ is sustained by C-H···O intermolecular interaction between C12-H12A···O1 together with an intramolecular interaction between N2-H2N···O1 leading to CH···O···HN heterosynthon that connects the molecules in a zigzag array, and extends along the b-direction (Figure 4b). The molecular packing of SMSCQ were governed by intermolecular C-H···O and intramolecular N-H···O interactions to form a CH···O···HN heterosynthon arranged in a zigzag array along the b-direction (Figure 4c). The molecular packing of SMRCV were associated by eight-membered {···NHCS} 2 homosynthon replicated by 2 1 symmetry along the crystallographic a direction (Figure 4d). The stark difference observed in the crystal packing of SMSCQ as compared to SMSCM, or SMRCV is presumably owing to its arrangement in Z-configuration which is stabilized by intramolecular N-H···O interaction. It is noteworthy that N-H···O possesses greater interaction energy than N-H···S interaction and hence it might have prevented the formation of the typical eightmembered {···NHCS} 2 homosynthon as observed in SMSCM, and SMRCV [36]. The geometric parameters characterizing the interactions for the corresponding structures are presented in Table 3.

Biological assays 3.2.1. MTT-based cytotoxicity and CPE reduction-based primary antiviral evaluation
The compounds did not display strong cytotoxicity. Toxicity decreased in the order carvone (CV) > camphorquinone (CQ) > camphor (CM). All twelve synthetic compounds showed a gradually decreasing trend in CC50 doses (Table 4).  (17) The chirality in the form of R and S enantiomers did not show any differential effects on either the cytotoxic dose, or the antiviral activity. The two enantiomers of the compounds showed parallel antidengue potential. Therefore, they may have same pharmacodynamics and can be equally potent. If this is so, the compounds could be used in their enantiomerically pure forms or as a racemic mixture [37]. However, pharmacokinetics of these enantiomers requires further exploration.  The CM compounds showed higher antidengue activity than the other substituted Schiff bases whereas; the CQ and CV compounds were rather more toxic than the CM compounds ( Table 2). The DTC-derivatives showed antibiotic potential in a study where some preexisting antibacterial molecules demonstrated greatly enhanced biological activity when conjugated with dithiocarbazate (DTC) [38]. During the present study, camphor compounds showed strong antidengue potential. Camphor is a terpenoid class of biologically active compounds found in the majority of natural products [39]. Camphor is a major component of many essential oils and the biological activities of essential oils are evident [40]. See supplementary materials Figure S34 for dose-response curves and Figures S35a-S35c for CPE-reduction during primary antiviral evaluation.

Foci forming unit reduction (FFUR) based secondary antiviral evaluation
The treatment of Vero cells infected with DENV 2 with three different dilutions of the synthetic compounds revealed reduction in the number of foci appearance for only six of the synthesized compounds, SMRCM, SMSCM, SBRCM, SBSCM, SMRCQ, and SMSCQ. Similarly, the CC50 doses of these compounds (Table 4) were also lower than the six compounds showing viral inhibition during secondary evaluation. This may be due to the cytotoxicity of the compounds during viral foci development that may damage the cells before viral infection [41]. The treatments were compared to the nontreated virus infections incubated parallel. The six potential antiviral compounds were inhibiting the virus in almost a close range of concentration, around 20 µM of the compounds were able to inhibit the virus just about 100%.
The secondary evaluation provided half minimal inhibitory concentrations (IC50) for the six potential antiviral compounds between 6.2 µM and 7.8 µM. Selectivity index (SI) identified the window between the cytotoxic and inhibitory concentrations ( Table 5). The SI values were within the range of 8.3 to 14.3, which are very close to each other, therefore, compounds can be considered equally potent. These six compounds were enantiomers of each other and the pattern of inhibition was very similar for each enantiomer of these compounds demonstrating that the antidengue activity might not be the virtue of the chirality of the compounds. See supplementary material, Figure S36 for antidengue effects of ribavirin as a positive control while, and Figures S37a and S37b for the antiviral effects of compounds during secondary evaluation.

Conclusion
A series of S-substituted dithiocarbazate Schiff bases containing enantiomeric monoterpenes have been synthesized and their molecular structures were proposed on the basis of elemental analysis and various spectroscopic techniques. To the best of our knowledge, this is the first test of this class of compounds against clinically isolated DENV 2. Compounds have demonstrated strong antiviral potential of varying degree. Secondary evaluation shows that camphor containing compounds may have enhanced potential and should warrant further study.