key: cord-0004071-90pat7ij authors: Khandaker, Tashrique A.; Hess, Jessica D.; Aguilera, Renato; Andrei, Graciela; Snoeck, Robert; Schols, Dominique; Pradhan, Padmanava; Lakshman, Mahesh K. title: Synthesis and Evaluations of “1,4‐Triazolyl Combretacoumarins” and Desmethoxy Analogs date: 2019-08-07 journal: European J Org Chem DOI: 10.1002/ejoc.201900569 sha: 0926ed33799458d78b8cc21fa20a8956675a3d71 doc_id: 4071 cord_uid: 90pat7ij 1,4‐Triazolyl combretacoumarins have been prepared by linking the trimethoxyarene unit of combretastatin A4 with coumarins, via a 1,2,3‐triazole. For this, 4‐azidocoumarins were accessed by a sequential two‐step, one‐pot reaction of 4‐hydroxycoumarins with (benzotriazol‐1‐yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP), followed by reaction with NaN(3). In the reaction with BOP, a coumarin‐derived phosphonium ion intermediate seems to form, leading to an O (4)‐(benzotriazolyl)coumarin derivative. For the CuAAC reaction of azidocoumarins with 5‐ethynyl‐1,2,3‐trimethoxybenzene, catalytic [(MeCN)(4)Cu]PF(6) in CH(2)Cl(2)/MeOH with 2,6‐lutidine, at 50 (o)C, was suitable. The 4‐azidocoumarins were less reactive as compared to PhN(3) and the NBO coefficients of the azido groups were compared by DFT analysis. Compound solubility was a problem in biological assays. On the basis of the biological and solubility data of one 1,4‐triazolyl combretacoumarin, four analogs lacking one or two methoxy groups were synthesized. Reactivity differences among the phenylacetylenes were noted and the NBO coefficients of the alkynes were compared by DFT analysis. In cytotoxicity assays, 1‐phenyl‐4‐(3,4,5‐trimethoxyphenyl)‐1H‐1,2,3‐triazole showed activity in CEM and MDA‐MB‐231 cell lines by apoptosis. The desmethoxy 6‐bromo‐4‐(4‐(4‐methoxyphenyl)‐1H‐1,2,3‐triazol‐1‐yl)‐2H‐chromen‐2‐one also showed cytotoxicity against the two cell lines, but this did not appear to be consistent with apoptosis. The antiviral activity of the compounds was unremarkable. Coumarins (Figure 1 ), a family of benzopyrones, are known to have diverse bioactivities. [1] [2] [3] [4] [5] [6] [7] They have attracted attention as a potential structural motif for the development of new anticancer agents. [8] [9] [10] [11] Derivatives of α-pyrones and coumarins have also been subjects for the development of antiviral agents, for example, against HIV1 and 2, HCV, RSV, HSV1, and A/PR8/ H1N1. [12] [13] [14] [15] [16] [17] [18] Combretastatin A4 (CA4, Figure 1 ) is a microtubulebinding natural product that targets the colchicine-binding site and is a vascular-disrupting agent. [19] [20] [21] [22] [23] [24] In a recent mechanistic study, [25] CA4 was shown to suppress VEGF signaling and cell proliferation, which may be related to its anti-angiogenic potency. Further, the crystal structure of CA4-tubulin complex has azido groups were compared by DFT analysis. Compound solubility was a problem in biological assays. On the basis of the biological and solubility data of one 1,4-triazolyl combretacoumarin, four analogs lacking one or two methoxy groups were synthesized. Reactivity differences among the phenylacetylenes were noted and the NBO coefficients of the alkynes were compared by DFT analysis. In cytotoxicity assays, 1-phenyl-4-(3,4,5-trimethoxyphenyl)-1H-1,2,3-triazole showed activity in CEM and MDA-MB-231 cell lines by apoptosis. The desmethoxy 6-bromo-4-(4-(4-methoxyphenyl)-1H-1,2,3-triazol-1-yl)-2Hchromen-2-one also showed cytotoxicity against the two cell lines, but this did not appear to be consistent with apoptosis. The antiviral activity of the compounds was unremarkable. been determined and this indicates CA4 to have common as well as different interactions to colchicine within the binding site. [26] In this analysis, [26] the trimethoxy aryl ring was found to reside in a hydrophobic region containing Ala250, Cys241, Leu255, Asp354, and Ala316, and the hydroxyl group was hydrogen-bonded to Val181 and Thr179. By comparison to colchicine, CA4 was deeper in the hydrophobic pocket and the hydroxyl group of colchicine was bonded to only Val181. Further, binding of CA4 did not significantly alter the structure of tubulin. [27] [28] [29] [30] [31] [32] ). Spurred by the importance of both coumarins and CA4, we decided to explore the synthesis and bioactivities of a class of compounds we dub 1,4-triazolyl combretacoumarins. There are, of course, isomeric 1,4-and 1,5-combretacoumarins ( Figure 2 ). Among the two, the latter class, which would be equivalent to (Z)-CA4, could be quite interesting due to the proximal disposition of two aryl rings and, in this context, 1,5triazolyl combretastatin analogs have been prepared. [33] In the present case, synthesis of this regioisomer would require reaction of a magnesiated alkyne with an azide, [34] but the organometallic reagent would be incompatible with the coumarin. Use of Ru-catalysis is known to yield 1,5-disubstituted 1,2,3-triazoles, [35] but this was unproductive in the present study (see the Supporting Information). On the basis of these considerations, we describe the synthesis and evaluations of 1,4-combretacoumarins, which are structurally closer to (E)-CA4. Our studies commenced with consideration of developing a rapid method for the synthesis of 4-azidocoumarins, one of the components needed to access the triazolyl combretacoumarins. Typically, 4-azidocoumarins are prepared from 4-hydroxycoumarins, by conversion to either the 4-chloro or 4-tosylate derivatives, followed by displacement with azide ion. [36, 37] In our research on uncatalyzed nucleoside modification methods, we have developed facile approaches for C6 and C4 modification of purine and pyrimidine nucleosides. [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] That chemistry involved the reaction of the oxygen atoms of the amide linkages in the nucleobases with (benzotriazol-1yloxy)tris-(dimethylamino)phosphonium hexafluorophosphate (BOP), a cheap and commercially available peptide-coupling agent. On the basis of those results, we were curious to assess whether 4hydroxycoumarin, essentially the tautomeric form of a -keto lactone, will react with BOP under appropriate conditions. Also in the current context, we have previously demonstrated a general synthesis of 6-azidopurine nucleosides, where O 6 -(benzotriazol-1-yl)inosine derivatives were isolated and then converted into the azido analogs. [43] On account of these combined observations, we were also interested in evaluating a one-pot synthesis of 4-azidocoumarins via reactions with BOP (Scheme 1). With these considerations and on the basis of previous experience, we elected to study the reaction of 4-hydroxycoumarin (1) with BOP, in MeCN as a solvent, and 1,8-diazabicyclo- [5.4 .0]undec-7-ene (DBU) as a base. The reaction leading to the formation of the C4 benzotriazolyl derivative 2 was rapid, reach-ing completion within 1 h at room temperature (other solvents tested were THF and DME [48] ). From the entire set of optimizations (see Table 1 in the Supporting Information), the use of MeCN and DBU for the formation of 2, and 18-crown-6 and NaN 3 for the displacement step, proved optimal for a two-step, one-pot conversion of 4-hydroxycoumarin (1) to 4-azidocoumarin (3) . Scale-up of reactions to 1 mmol was considered for the synthesis of various 4-azidocoumarins. However, in the scale-up, the second step of the conversion did not reach completion under the initially optimized conditions. A slight modification of the reaction conditions for this step, by increasing the amount of 18-crown-6 to 20 mol-% and NaN 3 to 3 equiv., led to full conversion within 2 h. Using these further modified conditions, seven 4-azidocoumarins were synthesized on the 1 mmol scale (Figure 3 ). Having shown the generality of this scalable, two-step, onepot synthesis of 4-azidocoumarins, we then assessed the mechanism of this conversion. In our previous work, we have observed the appearance of phosphonium ion intermediates in the reactions of the amide groups of hypoxanthine [38] and guanine nucleosides, [42] with BOP. In marked contrast to this, a corresponding phosphonium ion species was not observable in the reaction of a pyrimidine nucleoside. [46] Thus, we wanted to assess how the reaction of 4-hydroxycoumarin would compare to these previous results. Plausible mechanistic pathways are shown in Scheme 2. Deprotonation of the C4 hydroxyl in 1 can be followed by: (a) reaction at the phosphorus atom of BOP (in red), or (b) reaction at the N1 atom of BOP (in blue), or (c) an S N 2′-like reaction at the N3 atom of BOP (in blue). The reaction at the phosphorus atom could lead to a phosphonium intermediate from 4-hydroxycoumarin. In this case, the BtOthat is expelled in the process can then displace HMPA from the C4 position of the coumarin, resulting in 4-((1H-benzo[d] [1, 2, 3] triazol-1-yl)oxy)-2Hchromen-2-one (2) . On the other hand, reaction at the N1 or N3 atom would directly result in 2 with the expulsion of HMPA. To evaluate this question, a reaction of 1 with BOP (1.2 equiv.) and DBU (1.2 equiv.), in MeCN (0.5 mL) was moni- Figure 4 . Monitoring the reaction of 4-hydroxycoumarin (1) with BOP and DBU in MeCN by 31 ). Then compound 1 was added and a spectrum was acquired, but it showed no new resonance. Finally, DBU was added and spectra were acquired over 99 minutes. At 5 minutes, two additional resonances were observed, one at δ = 23.9 ppm corresponding to HMPA and another at δ = 33.9 ppm. The resonance at δ = 33.9 ppm is similar in the chemical shift to those observed for nucleoside phosphonium salts shown in Scheme 2. As the reaction proceeded, a gradual decrease of phosphonium ion and BOP occurred while the formation of HMPA increased. The NMR experiment also indicated that the phosphonium ion was almost completely consumed within 24 min. Next, the effort was directed to the conversion of 3,4,5-trimethoxybenzaldehyde (10) to 5-ethynyl-1,2,3-trimethoxybenzene (12) . Direct conversion of 10 by the Ohira-Bestmann protocol [49, 50] returned a product that remained impure after purification. Gem-dibromide 11 was then prepared by reaction of 10 with CBr 4 /PPh 3 in CH 2 Cl 2 . [51] In our hands, a DBU-based conversion of 11 to 12 [52] did not result in an adequately pure product (12, a while solid, has been reported as a colorless oil in the multiple procedures in the original report [52] ). By contrast, use of nBuLi, [53] led to the desired alkyne 12 (Scheme 3), isolated as a white solid. [51] Scheme 3. Synthesis of 5-ethynyl-1,2,3-trimethoxybenzene (12) and the CuAAC reaction. With the coupling partners at hand, we then proceeded to assess conditions for the copper-catalyzed azide-alkyne cycloaddition (CuAAC). [54, 55] We selected the following copper-based catalytic systems: A = 10 mol-% of CuSO 4 ·5H 2 O/20 mol-% of Na-ascorbate, B = 1.4 equiv. of CuSO 4 ·5H 2 O/0.8 equiv. of Cu, C = 5 mol-% of CuSO 4 ·5H 2 O/20 mol-% of Na-ascorbate, D = 10 mol-% of Cu-2-thiophene carboxylate, and E = 10 mol-% of [(MeCN) 4 Cu]PF 6 . The ligation reaction proved to be nontrivial and several experiments were necessary to find appropriate conditions. Results from the screening are shown in Table 1 . In both tBuOH/H 2 O and in CH 2 Cl 2 /H 2 O, reactions using CuSO 4 ·5H 2 O/Na-ascorbate showed low productivity at room and elevated temperatures (entries 1-5). Previously reported conditions for CuAAC reactions of azido coumarins were also tested, [28, 56] but these gave very little conversion in the present case, as did Cu-2-thiophene carboxylate (entries 6-9). Use of [(MeCN) 4 Cu]PF 6 in CH 2 Cl 2 /MeOH proved to be more promising (entries 10 and 11), and a reaction at 50 o C was successful although the product contained a contaminant. The contaminant could be removed by washing the product but this diminished the yield. In CH 2 Cl 2 and in the absence of MeOH, there was no reaction at room temperature (entry 12), but at 50 and 80 o C reaction occurred (entries 13 and 14) with results similar to that in entry 11. Outcomes from reactions in ClCH 2 CH 2 Cl and ClCH 2 CH 2 Cl/MeOH were inferior to those in CH 2 Cl 2 (compare entries 15 to 13 as well as 16 to 11). Because amine bases can be employed in reactions involving CuI and [(MeCN) 4 Cu]-PF 6 , [54, 55] we considered the addition of 2,6-lutidine to the reaction mixtures. This base minimizes by-product formation when Cu I catalysts are used for CuAAC reactions and could augment proton transfer to the cuprated triazolide. [55, 57] With the modified conditions, incomplete reactions were observed in CH 2 Cl 2 / MeOH at room temperature and in neat CH 2 Cl 2 at 50 o C (entries 17 and 18), but when both MeOH and 2,6-lutidine were present, a complete reaction was attained at 50 o C within 8 h, and a good product yield, as well as purity, was attained (entry 19) . Therefore, conditions in entry 19 were used to prepare the seven combretacoumarins shown in Figure 5 . In order to confirm the 1,4-disubstitution pattern, the NOESY spectrum of compound 13 (in [D 6 ]DMSO) was analyzed. This showed a distinct correlation between the C5 triazolyl proton (δ = 9.35 ppm, singlet) and the chemically equivalent aryl protons (δ = 7.31 ppm, singlet) as well as the vinyl proton of the coumarin (δ = 7.00 ppm, singlet). Despite the reported exothermicity and extremely rapid rates of many CuAAC reactions, [57] the azido coumarins used herein appeared to be slow reacting. On account of this unanticipated reactivity, we assessed the reactivity of 12 with phenylazide (PhN 3 ) using [(MeCN) 4 Cu]PF 6 and 2,6-lutidine, in 3:1 CH 2 Cl 2 /MeOH. In contrast to the reactions of the azido coumarins, the reaction of PhN 3 was complete within 3 h at room temperature, giving product 20 in 85 % yield. This not just indicated a reactivity difference between azido coumarins and PhN 3 but it also indicated that the potentially turnover-limiting alkyne cupration was likely not the cause. Therefore, after initial geometry optimization, the natural bond order (NBO) analyzed natural charges on the nitrogen atoms of the azido functionality in the azido coumarins and PhN 3 were evaluated by B3LYP Density Functional Theory (DFT) computations at the 6-311++g (2d, 2p) level. These data are shown in Table 2 . In CuAAC reactions the internal nitrogen atom of the azide moiety (N1 in the structures above Table 2) is nucleophilic and interacts with the copper ion, whereas the terminal nitrogen atom (N3 in the structures) is the electrophile. From the data in Table 2 , the NBO charges on the N1 atom of the coumarins did not show marked variations and are generally similar to that of PhN 3 . No major variations were seen with the NBO charges on the N2 atom as well in all cases. With the exception of the 7-OMe derivative that showed some negative charge on the N3 atom, all other coumarin derivatives show electron deficiency at this site. This is in marked contrast to the N3 atom of PhN 3 that by comparison, shows a greater negative charge at the N3 atom. In the coumarins, the negative charge can be delocalized onto the carbonyl groups. This could influence the electrophilicity of the N3 atom, but PhN 3 reacted more readily than the coumarins. The mesomeric effect of a 7-OMe group on the coumarin in 9 leads to restoration of some negative charge at the N3 center. Notably, 9 was very slow reacting in the CuAAC even as compared to the other azido coumarins, with the reaction remaining incomplete at 8 h (the reaction was complete within 24 h). These results seem to indicate that reasons beyond electron density considerations at the reacting atoms may be responsible for the reactivities of the azido coumarins. After these analyses on the initial 1,4-combretacoumarin series were completed, we undertook the preparation of additional analogs to assess if compounds with better solubility properties could be prepared. From the initial biological data (vide infra), we selected the relatively soluble bromo analog 16, which also showed activity, for additional modifications. We decided to remove one or two methoxy groups on the combretastatin portion of the analogs and therefore, CuAAC reactions of mono and dimethoxyphenyl acetylenes were conducted with 4-azido-6-bromo-2H-chromen-2-one (6). These desmethoxy products are shown in Figure 6 . Some additional reactivity factors became evident in these CuAAC reactions where the structure of the alkyne appeared to have an influence. Reactions leading to 21 and 24 were complete within 8 h, that leading to 22 was complete within 24 h, but the reaction towards 23 was the most difficult, remaining incomplete even at 24 h. In the reaction with p-methoxyphenylacetylene two modifications were attempted to assess whether the reaction could be improved: (a) addition of Figure 6 . Desmethoxy "1,4-combretacoumarin" analogs prepared from 4-azido-6-bromo-2H-chromen-2-one (6). 1.2 equiv. of the alkyne at the start of the reaction followed by addition of another 1.2 equiv. after 8 h, and a total reaction time of 24 h, and (b) addition of [(MeCN) 4 Cu]PF 6 in two aliquots (0.1 equiv. at the start of reaction and 0.1 equiv. after 8 h) and a total reaction time was 24 h. However, neither modification led to any improvement. The strong mesomeric effect of a pmethoxy group leading to increased electron density at the terminal alkynyl carbon atom could be a factor for this outcome. This effect could also influence the cycloaddition itself. Therefore, we assessed the natural charges on the alkynyl carbon atoms of the five alkynes used in this study, as well as phenylacetylene, by B3LYP Density Functional Theory (DFT) computations at the 6-311++g (2d, 2p) level (Table 3 ). In comparing the coefficients of the alkynyl C1 carbon atoms, the coefficient of p-methoxyphenylacetylene is the lowest in the set, and reactions with this alkyne remained incomplete under a variety of conditions. The NBO coefficients of the C2 alkynyl carbon atoms in the m-methoxy and m,m-dimethoxyphenylacetylene are similar, lower than those of the remaining alkynes, and comparable to that of phenylacetylene. As would be anticipated, the presence of a p-methoxy group increases electron density at the alkynyl C2 carbon atoms. On the basis of these data, one must surmise that these electronic differences contribute to the subtle reactivity differences observed. To evaluate the cytotoxic potential of compounds 13-20, a live cell imaging-based differential nuclear staining (DNS) assay was performed against two human cancer cell lines. The DNS assay utilizes two fluorescent nuclear stains, Hoechst 33342 and propidium iodide (PI), to selectively label living and dead cells. [58] An initial screen for cytotoxicity was conducted using the CEM leukemia cell line. Stock solutions of each compound were prepared at their highest soluble concentration in DMSO (Table 4 ). Stock and serially diluted solutions of the compounds were directly added to wells containing cells suspended in complete media at a final concentration of 1 % v/v DMSO and assessed in quadruplicate after 48 hours. The CC 50 is defined as the concentration of compound that causes loss of membrane integrity to 50% of the cell population as compared to untreated cells. CC 50 values were determined by linear interpolation of the two concentrations with cell death nearest the 50% mark. [a] NA = Not available, unable to obtain at highest soluble concentration. Analysis of 13, 18, and 19 treated cells revealed significant compound crystallization and did not induce quantifiable CEM cell death. Similarly, treatment with 14, 16, and 17 did not induce significant CEM cell death at the prepared concentrations. Compound 15, one of the least soluble compounds in the series, demonstrated cytotoxicity at its highest soluble concentration. A CC 50 value of 37.2 ± 0.6114 μM was calculated for 15 (Table 4) . Phenyl derivative 20, the most readily soluble of the compounds, induced cell death at a CC 50 of 34.6 ± 2.5939 μM (Table 4) . Accordingly, compound 20 was isolated for use in a secondary DNS assay conducted on the MDA-MB-231 (MB-231) triple-negative breast cancer cell line. In this DNS assay, MB-231 cells were treated with serial dilutions of 20 starting at 100 μM. Treatments were directly added to cell-containing wells (as indicated previously) and assessed in quadruplicate after 72 hours. Treatment with 20 revealed a CC 50 value of 55.4 ± 5.0951 μM against this cell line ( Table 4 ). The four desmethoxy compounds (21) (22) (23) (24) were also assessed for cytotoxicity against CEM and MB-231 cell lines. In addition, combretastatin A4 (CA4), a known tubulin-depolymerizing agent, was analyzed as a standard for apoptosis. [59] CEM and MB-231 cells were treated with compounds 21-24 at concentrations ranging from 100 μM to 0.5 μM, using the protocols previously indicated. Compounds 22, 23, and 24 all demonstrated similar cytotoxic activity and induced significant death in CEM cells at approximately 10 μM. This pattern was not observed upon subsequent analysis of MB-231 cells, which appeared selective to compound 23. These data are shown in Table 5 . Figure 7 shows the dose-response curves for compounds 20 and 23 against CEM and MB-231 cell lines. In all the aforementioned DNS assays, three experimental controls were included in each 96-well plate and assessed in quadruplicate. Untreated and DMSO-treated (1 % v/v) cells were used as negative controls to establish basal levels of cell death due to cell manipulation or culture conditions and to account for solvent-induced cell death. As a positive control for cytotoxicity, cells were treated with 1.62 mM H 2 O 2 . Externalization of phosphatidylserine (PS) to the cell surface is a well-established marker of cells undergoing early apoptosis. Annexin V is an intracellular protein with high affinity for PS which is routinely fluorochrome-labeled with FITC and used as a probe for PS translocation via flow cytometry. [60] In early-stage apoptosis, the cell membrane is still intact and is not yet permeable to propidium iodide (PI). To determine whether cell death was occurring via apoptosis or necrosis, CEM cells were treated using 100 μM, 30 μM, and 60 μM concentrations of 20 for 24 h, followed by annexin-FITC and PI staining. Non-treated and DMSO-treated cells (1 % v/v) served as negative controls and H 2 O 2 (1.62 mM) as a positive control for apoptosis/necrosis. Apoptotic cell populations are expressed as the sum of annexin-FITC positive stained cells, whereas the necrotic populations are cells that stained with PI but negative for annexin-FITC. In this assay, compound 20 was found to induce significant PS externalization in CEM cells at its highest soluble concentration. These data demonstrate a concentration-dependent increase in apoptosis/necrosis. However, they suggest that compound 20 preferentially induces cell death via apoptosis (Figure 8, Panel A) . In a subsequent assay, CEM cells treated with compound 23 were analyzed using concentrations ranging from 9.8 μM to 49 μM using the method previously described. Compound 23 did not demonstrate appreciable PS externalization when compared to the vehicle control (Figure 8, Panel B) . Thus, the modes of action of these two compounds are possibly quite different. In summary, in this work, we have demonstrated a simple and facile two-step, one-pot conversion of 4-hydroxycoumarins to the corresponding 4-azido derivatives. This involves activation of the 4-hydroxyl group with BOP and DBU in MeCN. The intermediate O 4 -(benzotriazolyl)coumarin derivatives formed in situ, possibly via the intermediacy of a phosphonium ion, can be reacted with NaN 3 and 18-crown-6 to give the 4-azidocoumarins. These products were subjected to CuAAC reactions with 5ethynyl-1,2,3-trimethoxybenzene using [(MeCN) 4 Cu]PF 6 in CH 2 Cl 2 /MeOH, and 2,6-lutidine as additive, at 50 o C. Critical to the success of these reactions was the need for both MeOH and 2,6-lutidine. Using these conditions, seven 1,4-combretacoumarins and one phenyl derivative, 1-phenyl-4-(3,4,5-trimethoxyphenyl)-1H-1,2,3-triazole, were synthesized. The latter was prepared in order to compare the reactivities of the 4-azidocoumarins with PhN 3 . Interestingly, although the azidocoumarins required an elevated temperature to attain complete reactions, PhN 3 reacted at room temperature and within a shorter reaction time. To understand possible reasons for this difference in reactivities, DFT computations were utilized to evaluate the NBO coefficients of the azido groups in the 4azidocoumarins but a link between the NBO coefficients on the nitrogen atoms of the azido groups and reactivity was not forthcoming. However, a difference was observed between the azido coumarins and PhN 3 . Four desmethoxy 1,4-combretacoumarins were prepared from reactions of 4-azido-6-bromo-2H-chromen-2-one (6) with two di-and two mono-methoxy phenylacetylenes. In these reactions, we found that the structures of the phenylacetylenes also influence the CuAAC reactions, and DFT computations were used to assess the NBO coefficients of the alkynyl carbon atoms. As anticipated, the terminal alkynyl carbon atom has a greater negative charge when a p-methoxy group is present. The varying electronic properties of the alkynes are likely also linked to the subtle reactivity differences observed. Antiproliferative and antiviral assays were undertaken. Unfortunately, the 1,4-combretacoumarins suffered from solubility problems. Among the various compounds, 1-phenyl-4-(3,4,5-trimethoxyphenyl)-1H-1,2,3triazole (20) was relatively soluble and showed antiproliferative effect, appearing to operate via an apoptotic pathway due to significant phosphotidylserine (PS) externalization. Three of the four desmethoxy compounds showed comparable cytotoxicity against CEM cells. Of these, 6-bromo-4-(4-(4-methoxyphenyl)-1H-1,2,3-triazol-1-yl)-2H-chromen-2-one (23) was also selective for MB-231 cells. However, this compound did not demonstrate appreciable PS externalization, in contrast to 1-phenyl-4-(3,4,5trimethoxyphenyl)-1H-1,2,3-triazole (20) . None of the compounds synthesized showed significant antiviral activity against a broad spectrum of DNA and RNA viruses. In an oven-dried 25 mL round bottomed flask equipped with a stir bar was placed 4-hydroxycoumarin (1, 162 mg, 1.00 mmol) in dry MeCN (10 mL). DBU (179 μL, 1.20 mmol, 1.20 equiv.) was added and the resulting homogeneous, light-yellow solution was stirred at room temperature for 10 min. BOP (531 mg, 1.20 mmol, 1.20 equiv.) was added and the mixture was stirred for 1 h at r.t. To the white mixture, NaN 3 (195 mg, 3.00 mmol, 3.00 equiv.) and 18crown-6 (52.0 mg, 0.20 mmol, 0.20 equiv.) were added sequentially. The mixture was stirred at room temperature for 1 h at which time TLC (SiO 2 /25 % EtOAc in hexanes) indicated the reaction to be complete. The mixture was transferred to a separatory funnel and partitioned between EtOAc and brine. The aqueous layer was separated and back extracted with EtOAc (3 ×). The combined organic layer was dried with anhydrous Na 2 SO 4 and concentrated under reduced pressure. Purification was performed by chromatography on a silica gel column packed in hexanes, and eluted with hexanes followed by 25 % EtOAc in hexanes. All azidocoumarins were synthesized and purified using this procedure (see individual compound headings for details). In an oven-dried 25 mL round bottomed flask equipped with a stir bar, was placed 4-hydroxycoumarin ( Reaction time for displacement with azide was 4.5 h. Pale [28] Pale-yellow solid (243 mg, 91 % yield Pale-yellow solid (205 mg, 80 % yield [51] 3,4,5-Trimethoxybenzaldehyde (10, 1.96 g, 10.0 mmol) was placed in an oven-dried 100 mL round bottomed flask, equipped with a stir bar. CH 2 Cl 2 (50 mL) and CBr 4 (3.98 g, 12.0 mmol, 1.20 equiv.) were added sequentially. The solution was cooled to 0 o C with stirring, and PPh 3 (5.25 g, 20.0 mmol, 2.00 equiv.) was added in portions. The mixture was stirred at 0 o C for 0.5 h, warmed to room temperature, and stirred for an additional 3 h. TLC (SiO 2 /10 % EtOAc in hexanes) indicated the reaction to be complete. The mixture was evaporated under reduced pressure and EtOAc was added. The crude mixture was filtered through a short pad of silica gel and the filtrate was concentrated under reduced pressure. Chromatography of the crude material on a silica gel column packed in hexanes and eluted with 5 % followed by 10 % EtOAc in hexanes, gave 3. In an oven-dried, 3-neck 25 mL round-bottomed flask equipped a stir bar, a solution of 5-(2,2-dibromovinyl)-1,2,3-trimethoxybenzene (11, 1.00 g, 2.84 mmol) in dry THF (7.1 mL) was cooled to -78 o C, under a nitrogen atmosphere. To the stirring mixture 1.6 M nBuLi in hexanes (4.44 mL, 2.50 equiv.) was added dropwise. The black mixture was stirred for 4 h at -78°C and for 1 h at room temperature. TLC (SiO 2 /10 % EtOAc in hexanes) indicated the reaction to be complete. The reaction was quenched with saturated aq. NH 4 Cl and transferred to a separatory funnel using EtOAc. The mixture was partitioned between EtOAc and water. The aqueous layer was separated and back extracted with EtOAc (3 ×). The combined organic layer was washed with brine, dried with anhydrous Na 2 SO 4 , and evaporated under reduced pressure. Chromatography on a silica gel column packed in hexanes and eluted with hexanes followed by 10 % EtOAc in hexanes, gave 455 mg (83 % yield) of 12 as a white solid. R f (SiO 2 /10 % EtOAc in hexanes) = 0. 28 In an oven-dried 12 mL vial equipped with a stir bar was placed 4azidocoumarin (3, 94.0 mg, 0.50 mmol) in dry CH 2 Cl 2 (3.75 mL), and 5-ethynyl-1,2,3-trimethoxybenzene ( 4 Cu]PF 6 (19.0 mg, 0.05 mmol, 0.10 equiv.) and dry MeOH (1.25 mL) were added to the resulting homogeneous, light-yellow solution. Next, 2,6-lutidine (58.0 μL, 0.50 mmol, 1.00 equiv.) was added, the mixture was flushed with nitrogen gas, and the resulting mixture was stirred for 8 h at 50 o C. TLC (SiO 2 /25 % EtOAc in hexanes) indicated the reaction to be complete. The yellow mixture was transferred to a separatory funnel using CH 2 Cl 2 , and partitioned between CH 2 Cl 2 and brine. The aqueous layer was separated and back extracted with CH 2 Cl 2 (3 ×). The combined organic layer was dried with anhydrous Na 2 SO 4 and evaporated under reduced pressure. Purification was performed by column chromatography on a silica gel column (see individual compound headings for details). Synthesized from azidocoumarin 4 (101 mg) and 5-ethynyl-1,2,3trimethoxybenzene (12, 115 mg Synthesized from azidocoumarin 6 (133 mg) and 5-ethynyl-1,2,3trimethoxybenzene (12, 115 mg Synthesized from azidocoumarin 7 (128 mg) and 5-ethynyl-1,2,3trimethoxybenzene (12, 115 mg Synthesized from azidocoumarin 8 (101 mg) and 5-ethynyl-1,2,3trimethoxybenzene (12, 115 mg Synthesized from azidocoumarin 9 (109 mg) and 5-ethynyl-1,2,3trimethoxybenzene (12, 115 mg Azidobenzene [61] In an oven-dried 25 mL round-bottomed flask equipped with a stir bar was placed PhB(OH) 2 (488 mg, 4.00 mmol) in MeOH (12 mL In an oven-dried 12 mL vial equipped with a stir bar was placed PhN 3 (59.0 mg, 0.50 mmol) in dry CH 2 Cl 2 (3.75 mL), and 5-ethynyl-1,2,3-trimethoxybenzene ( In an oven-dried 12 mL vial equipped with a stir bar was placed 4azido-6-bromocoumarin (6, 133 mg, 0.5 mmol) in dry CH 2 Cl 2 (3.75 mL), and 1-ethynyl-3,5-dimethoxybenzene (0.60 mmol, 1.20 equiv.) was added. [(MeCN) 4 Cu]PF 6 (19.0 mg, 0.05 mmol, 0.10 equiv.) and dry MeOH (1.25 mL) were added to the resulting solution. Next, 2,6-lutidine (58.0 μL, 0.50 mmol, 1.00 equiv.) was added, the mixture was flushed with nitrogen gas, and the resulting mixture was stirred for 8-24 h at 50°C. After an appropriate period of time, the reaction was checked by TLC for completion. The mixture was transferred to a separatory funnel using CH 2 Cl 2 and partitioned between CH 2 Cl 2 and brine. The aqueous layer was separated and back-extracted with CH 2 Cl 2 (3 ×). The combined organic layer was dried with anhydrous Na 2 SO 4 and evaporated under reduced pressure. See the individual compound headings for specific reaction times and purification details. Note: the synthesis of compound 23 was modified (see Results and Discussion for details). In this case, at the start of the reaction, one aliquot of 1-ethynyl-4-methoxybenzene (79 mg, 0.60 mmol, 1.20 equiv.) was added and the volume of CH 2 Cl 2 was 3 mL. Quantities of all other reagents were as stated above and the reaction temperature was 50 o C. After 8 h, a second aliquot of 1-ethynyl-4methoxybenzene (79 mg, 0.60 mmol, 1.20 equiv.) was added along with 0.75 mL of CH 2 Cl 2 , and the reaction was allowed to proceed for an additional 16 h at 50 o C. Synthesized from azidocoumarin 6 and 1-ethynyl-3,5-dimethoxybenzene (97 mg, 0.60 mmol, 1.20 equiv.). TLC (SiO 2 /25 % EtOAc in hexanes) after 24 h indicated the reaction to be complete. Chromatography was precluded by the low solubility of the product. Therefore, the product was initially sonicated in 25 % CH 2 Cl 2 in hexanes (8 mL) and filtered. Next, the precipitate was heated in PhH (3 mL Synthesized from azidocoumarin 6 with two aliquot additions of 1-ethynyl-4-methoxybenzene (79 mg, 0.60 mmol, 1.20 equiv.), see the note above. TLC (SiO 2 /25 % EtOAc in hexanes) after 24 h indicated the reaction was incomplete. Chromatography was precluded by the low solubility of the product. Therefore, the product was initially washed with 25 % CH 2 Cl 2 in hexanes (2 × 4 mL) and the supernatant was removed. Next, the precipitate was heated in PhH (4 mL), cooled, and the supernatant was removed. Chromatography was precluded by the low solubility of the product. Therefore, the product was initially sonicated in 25 % CH 2 Cl 2 in hexanes (4 mL) and the supernatant was removed. Next, the precipitate was heated in PhH (4 mL), cooled, and the supernatant was removed. The precipitate was dried under vacuum to give 192 mg (90 % yield) of 21 as a pale-yellow solid. R f (SiO 2 /25 % EtOAc in hexanes) = 0.13 The precipitate was dried under vacuum to give 135 mg (68 % yield) of 24 as an off-white solid. R f (SiO 2 /25 % EtOAc in hexanes) = 0.32. Mp: 240-242°C (dec., and associated with a color change to dark brown with no phase change). 1 H NMR (500 MHz Advances in Structure and Activity Relationship of Coumarin Derivatives Drug Resist. Updates Synthesis of 4-Azidocoumarins and Their Use in Copper-Catalyzed Azide-Alkyne Cycloaddition Reactions This work was supported by grant CHE-1265687 from the National Science Foundation to MKL. Infrastructural support at CCNY was provided by National Institutes of Health grant G12MD007603 from the National Institute of Minority Health and Health Disparities (NIMHD). The authors thank the personnel of the Cytometry, Screening and Imaging Core Facilities, which is supported by a Research Centers in Minority Institutions NIMHD grant 5G12MD007592 to the Border Biomedical Research Center at the University of Texas at El Paso. The authors express their gratitude to Mrs. Leentje Persoons, Mrs. Ellen De Waegenaere, Mrs. Bianca Stals, Mrs. Kirsten Lepaige, and Mrs. Nathalie Van Winkel for excellent technical assistance in the antiviral assays. Indofine Chemical Company is thanked for samples of some of the 4-hydroxycoumarins that have been utilized in this work.Keywords: Alkynes · Click chemistry · Coumarin · Copper · Coumarin · Triazole