New HSV-1 Anti-Viral 1′-Homocarbocyclic Nucleoside Analogs with an Optically Active Substituted Bicyclo[2.2.1]Heptane Fragment as a Glycoside Moiety

New 1′-homocarbanucleoside analogs with an optically active substituted bicyclo[2.2.1]heptane skeleton as sugar moiety were synthesized. The pyrimidine analogs with uracil, 5-fluorouracil, thymine and cytosine and key intermediate with 6-chloropurine (5) as nucleobases were synthesized by a selective Mitsunobu reaction on the primary hydroxymethyl group in the presence of 5-endo-hydroxyl group. Adenine and 6-substituted adenine homonucleosides were obtained by the substitution of the 6-chlorine atom of the key intermediate 5 with ammonia and selected amines, and 6-methoxy- and 6-ethoxy substituted purine homonucleosides by reaction with the corresponding alkoxides. No derivatives appeared active against entero, yellow fever, chikungunya, and adeno type 1viruses. Two compounds (6j and 6d) had lower IC50 (15 ± 2 and 21 ± 4 µM) and compound 6f had an identical value of IC50 (28 ± 4 µM) to that of acyclovir, suggesting that the bicyclo[2.2.1]heptane skeleton could be further studied to find a candidate for sugar moiety of the nucleosides.


Introduction
Nucleosides are a chemical class of active substances used as efficient antitumor [1][2][3] or antiviral drugs [4][5][6][7]. The resistance which appears after prolonged use and their toxicity are two principal secondary effects that request the development of newer, more safe candidates. An impressive number of modifications of the sugar moiety or replacements with carbocyclic fragment, replacement of the base or both have been done and some of the corresponding molecules had antiviral or anticancer activity and became useful drugs [2,3,6,7].
Another direction pursued in search for new active nucleosides was to introduce a methylene group between the heterocyclic base and the sugar moiety. The new compounds are, in fact, acyclic nucleosides, but in the literature, their most common name is "1 -homonucleosides". The insertion of the methylene group alters the chemical and biological characteristics of the molecules, as follows: (1) The anomeric carbon atom C1 in a tetrahydrofuran moiety has no more reactivity and the compounds are resistant to the enzyme hydrolysis of C1 -CH 2 Base bond, (2) the methylene group the methylene group alters the chemical and biological characteristics of the molecules, as follows: (1) The anomeric carbon atom C1′ in a tetrahydrofuran moiety has no more reactivity and the compounds are resistant to the enzyme hydrolysis of C1′-CH2Base bond, (2) the methylene group allows free rotation between the base and sugar moiety, increasing the flexibility of the molecule in the enzyme pocket(s), (3) electronic and steric interactions between the base and sugar moiety are decreased, (4) the lipophilicity of the compounds is slightly increased, a factor sometimes important for the transport of the molecule across the cell wall, (5) the separation between HO-C5′ of the (un)modified sugar and the N1 or N9 nitrogen atoms of the heterocyclic bases is slightly increased and many times this is the cause of the biological activity or of the lack of activity of the analogs [8].
In the 1′-homonucleoside, the biological activity mainly decreased compared with that of the parent nucleoside [9,10]. The substitution of oxygen with nitrogen [11] or sulfur [12] did not lead to active compounds. However, some changes in the sugar moiety have led to carbocyclic compounds with potent antiviral or anticancer activity.

Molecular Design
The docking studies were performed using CLC Drug Discovery Workbench Software (professional software). The score and hydrogen bonds formed with the amino acids from the group interaction atoms were used to predict the binding modes, the binding affinities and the orientation of the docked ligands in the active site of the protein receptor. The protein-ligand complex was realized based on the X-ray structure of herpes simplex type-1 thymidine kinase (TK) in complex with acyclovir (AC2), who was downloaded from the Protein Data Bank (PDB ID: 2KI5, for a 1.9Å resolution) [39]. The TK enzyme was studied for binding capabilities in the development of new, more effective HSV-1 inhibitors in many papers (for example [40]). The binding site and binding pockets, used to guide the molecular docking of the ligands, were established, and the search was carried out inside the binding site volume (the green sphere, Figure S1a). In our previous papers [33][34][35][36][37][38] we used an optically active bicyclo[2.2.1]heptane scaffold to obtain a library of L-type carbocyclic nucleosides X and tested them for antiviral and anticancer activity ( Figure 3). Compounds Xa and Xb were "the most prospective for their antiviral activity against influenza virus due to their low toxicity and high activity" [35]. Compound Xc was very active against coxsackievirus B4, with EC 50 = 0.6 µg/mL and selective index of 141, compound Xb being also promising. The compounds had low anticancer activity.
Molecules 2019, 24, x 3 of 20 present significant anticancer activity on L1210/0, Molt4/C8 and CEM cell lines (IC50 = 3.2-11 µg/mL) [8,31,32]. In our previous papers [33][34][35][36][37][38] we used an optically active bicyclo[2.2.1]heptane scaffold to obtain a library of L-type carbocyclic nucleosides X and tested them for antiviral and anticancer activity ( Figure 3). Compounds Xa and Xb were "the most prospective for their antiviral activity against influenza virus due to their low toxicity and high activity" [35]. Compound Xc was very active against coxsackievirus B4, with EC50 = 0.6 µg/mL and selective index of 141, compound Xb being also promising. The compounds had low anticancer activity. The promising results with the specified substituted constrained bicyclo[2.2.1]heptane scaffold, gave us the idea to use it for obtaining 1′-homocarbanucleosides, linking the nucleobase to the hydroxymethyl group and keeping the 5-endo-OH group free, in search to obtain potentially antiviral compounds. The works in this direction are presented below.

Molecular Design
The docking studies were performed using CLC Drug Discovery Workbench Software (professional software). The score and hydrogen bonds formed with the amino acids from the group interaction atoms were used to predict the binding modes, the binding affinities and the orientation of the docked ligands in the active site of the protein receptor. The protein-ligand complex was realized based on the X-ray structure of herpes simplex type-1 thymidine kinase (TK) in complex with acyclovir (AC2), who was downloaded from the Protein Data Bank (PDB ID: 2KI5, for a 1.9Å resolution) [39]. The TK enzyme was studied for binding capabilities in the development of new, more effective HSV-1 inhibitors in many papers (for example [40]). The binding site and binding pockets, used to guide the molecular docking of the ligands, were established, and the search was carried out inside the binding site volume (the green sphere, Figure S1a). The promising results with the specified substituted constrained bicyclo[2.2.1]heptane scaffold, gave us the idea to use it for obtaining 1 -homocarbanucleosides, linking the nucleobase to the hydroxymethyl group and keeping the 5-endo-OH group free, in search to obtain potentially antiviral compounds. The works in this direction are presented below.

Molecular Design
The docking studies were performed using CLC Drug Discovery Workbench Software (professional software). The score and hydrogen bonds formed with the amino acids from the group interaction atoms were used to predict the binding modes, the binding affinities and the orientation of the docked ligands in the active site of the protein receptor. The protein-ligand complex was realized based on the X-ray structure of herpes simplex type-1 thymidine kinase (TK) in complex with acyclovir (AC2), who was downloaded from the Protein Data Bank (PDB ID: 2KI5, for a 1.9Å resolution) [39]. The TK enzyme was studied for binding capabilities in the development of new, more effective HSV-1 inhibitors in many papers (for example [40]). The binding site and binding pockets, used to guide the molecular docking of the ligands, were established, and the search was carried out inside the binding site volume (the green sphere, Figure S1a).
Firstly, the binding site and docking pose of the co-crystallized AC2 interacting with amino acids residues are shown in Figure S1a-c. Then hydrogen bonds between amino acids residues of thymidine kinase and co-crystallized AC2 and docking validation (for the co-crystallized) were done (see Figure S2). The new 1 -homocarbacyclonucleosides were finally docked, and the results of the calculated properties are presented in Table 1 (flexible bonds, Lipinski violations, the number of hydrogen bond donors, the number of hydrogen bond acceptors and log P). These parameters can predict if a molecule possesses properties that might turn it into an active drug, according to the Lipinski's rule of five. The number of violations of the Lipinski rules allows to evaluate drug-likeness for a molecule. According to the data presented in Table 1, all of the compounds comply with the Lipinski rules (Lipinski violation is 0) [41]. In Table 1, the results of the docking study (docking score, RMSD < 2Å) are also presented. The docking score (PLANT PLP score) is a function described in Korb et al. [42]. For a strong binding, the score has a negative value, for weak or non-existing binding the score has a less negative or even positive value.
Additionally, group interaction, hydrogen bonds of ligands with amino acid residues were determined and hydrogen bond length was calculated. These are presented in the Table S1.
The data presented in the Table 1 and Table S1 and Figure 4 show that compounds 4a-4d, 5, 6a-6b, 6e, 6i-6k, 7a-7b have a docking score greater than that of acyclovir (−49.29, RMSD 0.71).  Compounds 6i and 6k presented the best scores, −70.21, respectively −70.07. The docking pose of the two ligands interacting with the amino acids residues is presented in Figure S3. In fact, all new ligands (1′-homocarbanucleosides) were found to have the same orientation as AC2 ( Figure S4a), as it can be observed in Table S1 for linking to a group of interaction with the amino acids.
The orientation and the docking pose of AC2 with the pyrimidine ligands 4a-4d, 6a-6k N 6substituted adenine ligands and N 6 -alkoxy-purine 7a-7b ligands are presented in Figures S4b,   Compounds 6i and 6k presented the best scores, −70.21, respectively −70.07. The docking pose of the two ligands interacting with the amino acids residues is presented in Figure S3. In fact, all new ligands (1 -homocarbanucleosides) were found to have the same orientation as AC2 ( Figure S4a), as it can be observed in Table S1 for linking to a group of interaction with the amino acids.
The orientation and the docking pose of AC2 with the pyrimidine ligands 4a-4d, 6a-6k N 6 -substituted adenine ligands and N 6 -alkoxy-purine 7a-7b ligands are presented in Figures S4b, and S5a,b, and show a good overlay.
The docking score was correlated with the experimental HSV-1 antiviral results (IC 50 ), and the best matching is for compound 6j (score −62.08, RMSD 1.57, IC 50 = 15 ± 2 µM) ( Figure 5). The second compound (6d) with lower IC 50 (21 ± 4 µM) than acyclovir (28 ± 4 µM) had a little lower value for the docking score (−39.94, RMSD 0.13) than that for acyclovir (−49.29, RMSD 0.71). Compounds 6i and 6k presented the best scores, −70.21, respectively −70.07. The docking pose of the two ligands interacting with the amino acids residues is presented in Figure S3. In fact, all new ligands (1′-homocarbanucleosides) were found to have the same orientation as AC2 ( Figure S4a), as it can be observed in Table S1 for linking to a group of interaction with the amino acids.
The orientation and the docking pose of AC2 with the pyrimidine ligands 4a-4d, 6a-6k N 6substituted adenine ligands and N 6 -alkoxy-purine 7a-7b ligands are presented in Figures S4b, and S5a,b, and show a good overlay.

Chemistry
The synthesis of these new 1 homocarbanucleosides started from the diol 2, obtained as a major isomer by sodium borohydride reduction of the optically active keto-alcohol intermediate 1 by our previous procedure [36]. The bulk of 5-endo-OH isomer 2 was obtained in pure form by crystallization, in yield greater than 77%. By low pressure chromatography (LPC) purification of the mother liquors, the total yield was increased to 91% (Scheme 1).
We chose to use a Mitsunobu reaction to alkylate the pyrimidine bases with the primary hydroxyl group of the diol 2, as the most direct chemical option to obtain the 1′-homocarbapyrimidine nucleosides 4 (Scheme 1). For this goal, we relied on the fact that in the Mitsunobu reaction, the primary hydroxyl group has greater reactivity than the secondary hydroxyl group, linked endo to C5 carbon atom. Certainly, the protection of the secondary hydroxyl group would lead to greater yields of the final products 4, but this would make the sequence of reactions longer, and at that time the yield was not considered so important. Compounds 4a, 4b, and 4c were obtained in 20.3, 34.5 and respectively 36.6% yields. Compound 4d was obtained in 23.5% yield by the Mitsunobu reaction of diol 2 with N 4 -Cytosine benzoate, followed by the deprotection of the benzoate group on crude alkylation reaction product by transesterification (MeONa/MeOH). The isolation of the pure compound 4d was realized by LPC. The reduced yield for the pyrimidine analogs is attributed to the formation of O 2 ,O 4 -and N 1 ,O 4 -bis-akylated secondary compounds, as we observed previously [43] for the alkylation of the diol 2 protected at the primary hydroxyl as benzoate. In this case, alkylated compounds to both hydroxyls of the diol 2 could also be formed. All secondary compounds had higher mobility on TLC, in the domain of triphenylphosphine oxide as Rf, and were not isolated pure for characterization.
The second part of the paper is focused on obtaining 1′-homocarbanucleoside analogs with a purine base. In this case, we also chose the most direct synthesis, which implies two steps. In the first step, we obtained the key optically active intermediate 5 in 67.6% yield by a slight modification of the Mitsunobu reaction of 6-chloropurine [44] with our diol 2 (Scheme 2). In the second step, the 6chlorine atom of the key intermediate 5 was substituted with ammonia to obtain the adenine analog 6a, and with selected amines, used previously in our papers, where 6-chloropurine was linked exo to the C5 carbon atom [33,34], as in X (Figure 3, R = Cl), to 1′-homocarba-N 6 -substituted adenine We chose to use a Mitsunobu reaction to alkylate the pyrimidine bases with the primary hydroxyl group of the diol 2, as the most direct chemical option to obtain the 1 -homocarba-pyrimidine nucleosides 4 (Scheme 1). For this goal, we relied on the fact that in the Mitsunobu reaction, the primary hydroxyl group has greater reactivity than the secondary hydroxyl group, linked endo to C 5 carbon atom. Certainly, the protection of the secondary hydroxyl group would lead to greater yields of the final products 4, but this would make the sequence of reactions longer, and at that time the yield was not considered so important. Compounds 4a, 4b, and 4c were obtained in 20.3, 34.5 and respectively 36.6% yields. Compound 4d was obtained in 23.5% yield by the Mitsunobu reaction of diol 2 with N 4 -Cytosine benzoate, followed by the deprotection of the benzoate group on crude alkylation reaction product by transesterification (MeONa/MeOH). The isolation of the pure compound 4d was realized by LPC. The reduced yield for the pyrimidine analogs is attributed to the formation of O 2 ,O 4 -and N 1 ,O 4 -bis-akylated secondary compounds, as we observed previously [43] for the alkylation of the diol 2 protected at the primary hydroxyl as benzoate. In this case, alkylated compounds to both hydroxyls of the diol 2 could also be formed. All secondary compounds had higher mobility on TLC, in the domain of triphenylphosphine oxide as R f , and were not isolated pure for characterization.
The second part of the paper is focused on obtaining 1 -homocarbanucleoside analogs with a purine base. In this case, we also chose the most direct synthesis, which implies two steps. In the first step, we obtained the key optically active intermediate 5 in 67.6% yield by a slight modification of the Mitsunobu reaction of 6-chloropurine [44] with our diol 2 (Scheme 2). In the second step, the 6-chlorine atom of the key intermediate 5 was substituted with ammonia to obtain the adenine analog 6a, and with selected amines, used previously in our papers, where 6-chloropurine was linked exo to the C 5 carbon atom [33,34], as in X (Figure 3, R = Cl), to 1 -homocarba-N 6 -substituted adenine nucleosides 6a-6k, in good yields (from 77.9% for 6k to 95.1% for 6d). Only the compound 6g was obtained in a moderate yield of 49.2%. The chlorine atom was also substituted by a methoxy (7a) or ethoxy group (7b), by the reaction of the key intermediate 5 with sodium methoxide in methanol or sodium ethoxide in ethanol in good yield (70.9%, respectively 83.6%).
The pure compounds were fully characterized and have been screened for antiviral activity.

1 H, 13 C-NMR, MS, Elemental Analysis and Optical Rotation
All new compounds were purified by pressure chromatography and analyzed by optical rotation, IR, 1 H-, 13 C-NMR, and 2D-NMR spectra, presented at the experimental part. The analytical data were in full agreement with the proposed structures. In 1 H-NMR spectra, the assignment was performed on the basis of chemical shifts, signal intensity and multiplicity of H-H coupling constants. 13 C-NMR and complementary 2D-NMR and decoupling spectra gave the correct signal for each proton and carbon atom in the molecules. The 1 H-, 13

Antiviral Activity of the Compounds
The new 1′-homocarbanucleosides analogs were tested on different viruses: The pure compounds were fully characterized and have been screened for antiviral activity.

1 H, 13 C-NMR, MS, Elemental Analysis and Optical Rotation
All new compounds were purified by pressure chromatography and analyzed by optical rotation, IR, 1 H-, 13 C-NMR, and 2D-NMR spectra, presented at the experimental part. The analytical data were in full agreement with the proposed structures. In 1 H-NMR spectra, the assignment was performed on the basis of chemical shifts, signal intensity and multiplicity of H-H coupling constants. 13 C-NMR and complementary 2D-NMR and decoupling spectra gave the correct signal for each proton and carbon atom in the molecules. The 1 H-, 13

Antiviral Activity of the Compounds
The new 1 -homocarbanucleosides analogs were tested on different viruses: The results of anti-viral testing of novel 1 -homocarbanucleosides analogs against adeno-, influenza, and herpesvirus type 1are summarized in Table 2.
As can be seen from the data presented, no derivatives appeared active against adenovirus. Only one compound demonstrated moderate inhibiting activity against influenza virus while 5 out of 18 compounds were active against the herpes virus. Importantly, there was no correlation between activity against influenza virus and herpes virus. Indeed, the cytosine compound 4d, the only derivative active against influenza virus, was ineffective against HSV-1. On another hand, the compounds active against herpes (5-chloro-purine 5, and 6 alkyl-substituted 6c, 6d, 6f, 6j) did not show the same activity against influenza virus. These differences may reflect the specificity in their targets. With high probability, the targets for nucleoside analogs should be the enzymes participating in the nucleic acids metabolism, in particular, viral polymerases. Herpes virus and influenza virus have a different organization of their genomes. The former has double-stranded DNA genome while the latter-single strand RNA genome of negative polarity. Due to differences in genomes of influenza virus and herpes virus, one can suggest that different inhibitors, although belonging to one chemical class, are needed to interfere with the activities of two different enzymes.
It is known that the best-known anti-herpes compound acyclovir is phosphorylated in infected cells by viral thymidine kinase following by incorporation of its triphosphate form into the DNA resulting in chain termination and death of infected virus cells [45]. Probably, the compounds we studied have a similar mechanism of activity, and for this reason, are ineffective against influenza virus who does not encode its own kinase. Further experiments are needed for deciphering the mechanism of activity and identifying the target(s) of lead compounds, including but not limited to time-of-addition experiments, virus yield reduction assay as well as tests for specific viral enzymes and selection and analysis of resistant mutants.
By analyzing the structures of active compounds, one can infer that even minor differences in the substituents R result in a dramatic change in the anti-viral properties. Indeed, the similar compounds 6f and 6g differing in one nitrogen bridge atom demonstrate almost a 30-fold difference in the virus-inhibiting activity against HSV-1. The introduction of an amino group between 4-methylpyperazine and N 6 nitrogen atom resulted in the complete loss of virus inhibition in 6g compared to 6f. Change of the hexane group to unsaturated phenyl group (6d to 6k) resulted in the decrease of virus inhibition (IC 50 from 21 to 48 µM), while the addition of the metoxy group (6j) restored the activity (IC 50 15). Even the change of cyclohexyl for cyclo-pentyl group (6d to 6c) decreased the virus inhibition (IC 50 from 21 to 47 µM). Chorine-and amino-substituted derivatives 5 and 6a did not differ significantly in their properties. Taking into account only the IC 50 of the new 1 -homocarbacyclonucleosides in comparison with that of acyclovir (28 ± 4 µM), compounds 6j and 6d have a lower IC 50 (15 ± 2 and 21 ± 4 µM), compound 6f-an identic value (28 ± 4 µM) and another three compounds had the values of IC 50 about twice higher. Five of them have also a SI greater than 10 (See Table 3). Importantly, compounds 6f, 6d, and 6j showed values of IC 50 against herpes virus similar to that of acyclovir, but their cytotoxicity was higher. This suggests that further optimization of the structure directing to decrease the toxicity could provide good pharmacologic characteristics of the resulting compounds.
Taken together, our findings suggest that the bicyclo[2.2.1]heptane skeleton could be a candidate for sugar moiety of the nucleosides with potential virus-inhibiting properties.

Experimental-Chemistry
Melting points (uncorrected) were determined in open capillary on an OptiMelt melting point apparatus (MPA 100, Stanford Research System, Inc., Sunnyvale, CA, USA). The progress of the reaction was monitored by TLC on silica gel 60 or 60F 254 plates (Merck, Darmstadt, Germany) eluted with the solvent systems: I dichloromethane-methanol, 9:1, II dichloromethane-methanol, 95:5, III dichloromethane-methanol, 4:1. Spots developed in UV and with 15% H 2 SO 4 in MeOH (heating at 110 • C, 10 min). IR spectra were recorded on FT-IR-100 Perkin Elmer spectrometer (Perkin Elmer, Shelton, CT, USA), in solid phase by ATR and frequencies were expressed in cm −1 , with the following abbreviations: w = weak, m = medium, s = strong, v = very, br = broad. 1 H-NMR and 13 C-NMR spectra are recorded on Bruker Fourier 300 MHz spectrometer (300 MHz for 1 H and 75 MHz for 13 C, Karlsruhe, Germany), or Bruker Avance III 500 MHz spectrometer (500 MHz for 1 H and 125 MHz for 13 C), spectrometer chemical shifts are given in ppm relative to TMS as internal standard. Complementary spectra: 2D-NMR and decoupling were done for the correct assignment of NMR signals. The numbering of the atoms in the compounds is presented in schemes. Diol 2 was obtained by sodium borohydride reduction of the keto group of compound 1, as previously presented [36], having mp 116-117. Triphenylphosphine (Ph 3 P) (5.246 g, 20 mmol) and 5-fluorouracil (2.62 g, 20 mmol) were suspended in 120 mL anh tetrahydrofuran and stirred at rt for 20 min under anh. argon atmosphere, then cooled to 0 • C with an ice-water bath. DIAD (4.16 mL, 20 mmol) was added dropwise. After one hour, a solution of diol 2 (1.766 g, 10 mmol) in tetrahydrofuran (80 mL) was added during 50 min and stirred overnight monitoring the end of the reaction by TLC (I, R f 2 = 0.43, R f = 0.55). A secondary compound with R f = 0.63 was also formed in the reaction and was not isolated pure. Solvent was distilled under reduced pressure, the residue was purified by low pressure chromatography (LPC) (solvent system, I), resulting a pure fraction of 587 mg (20.3%) 4a, mp 212.3-213.7 • C (EtOH), [α] D = 39.4 • (1% EtOH), IR: 3377s, 3170w, 3045w, 2962w, 2878w, 2819w, 1762w, 1703s, 1663vs, 1475w, 1434w, 1375m, 1339m, 1237s, 1113m, 1077m, 1006m, 900w, 817m, 1 H-NMR (DMSO-d 6  General Procedure for Synthesis of Carbocyclic 1 -Homonucleosides 6a-7k The key intermediate 5 (0.4 or 0.8 mmol) was stirred with the specified amount of amine, without or in the presence of ethanol (3.5 or 7 mL) as solvent, without or in the presence of triethylamine, for the time frame mentioned in each example. The solvent (ethanol) and the volatile amines were removed under reduced pressure, the residue was taken in CH 2 Cl 2 , washed with water, organic phase dried (Na 2 SO 4 ) and concentrated to dryness (the aqueous phases were extracted with CH 2 Cl 2 , and unified with the crude product). The crude product was purified by LPC and the pure compound was obtained as foam or waxy. Some of the compounds were crystallized in the solvent(s) specified in each case.

Virus Titration.
The compounds in appropriate concentrations were dissolved in MEM with 1 µg/mL trypsin for Flu A and without additives for AdV5 and HSV-1, and incubated with cells for 1 h at 36 • C. The cell culture was then infected with either virus (moi 0.01). The plates were incubated for 48 h (Flu A) or 72 h (AdV5 and HSV-1) at 36 • C in the presence of 5% CO 2 . The activity of the viruses was evaluated by the MTT test, as described above, as its ability to decrease cell viability. Each concentration of the compounds was tested in triplicate. The antiviral activity of the compounds was estimated by the protection of cells from virus-induced death as compared with the control. The 50% inhibiting concentration (IC 50 ) of the drug, that is, the concentration at which 50% of cells were protected, compared to control wells, and the selectivity index (the ratio of CC 50 to IC 50 ) were calculated from the data obtained.
These suggest that the bicyclo[2.2.1]heptane fragment can be effectively used instead of ribose or deoxyribose moiety to give nucleoside analogs with virus-inhibiting activity. The exact mechanism(s) of their anti-viral activity should be further studied in order to optimize the structure to create derivatives with low toxicity and high selectivity against specific viruses and with broad range of activity.

Patents
A patent pending dealing with the synthesis of the compounds was also filled.
Author Contributions: C.I.T. and L.P. conceived and designed the experiments, performed the experiments, and analyzed the data; C.I.T., V.V.Z. and C.D. wrote the paper; A.H. and M.M. did the physico-chemical experiments of the synthesized compounds, A.V., E.S., A.V.S., D.J. for anti-viral screening, V.Z. and J.N. supervised anti-viral screening.