Organo NHC catalyzed aqueous synthesis of 4β-isoxazole-podophyllotoxins: in vitro anticancer, caspase activation, tubulin polymerization inhibition and molecular docking studies

Abstract

We present, for the first time, the organo-N-heterocyclic carbene (NHC) catalyzed 1,3-dipolar cycloaddition of 4β-O-propargyl podophyllotoxin (1) with in situ aromatic nitrile oxides to afford regioselective 4β-isoxazolepodophyllotoxin hybrids (6a–n) in benign aqueous-organic media. Preliminary anticancer activity results showed that compound 6e displayed superior activity against MCF-7, HeLa and MIA PaCa2 human cell lines compared with podophyllotoxin. Compounds 6j and 6n showed greater activity against the MCF-7 cell line than the positive control. Caspase activation studies revealed that compound 6e at 20 μg ml⁻¹ concentration had greater caspase 3/7 activation in MCF-7 and MIAPaCa2 cells than podophyllotoxin. Furthermore, in vitro tubulin polymerization inhibition studies revealed that compound 6e showed comparable activity with podophyllotoxin. Finally, in silico molecular docking studies of compounds 6e, 6j, 6n and podophyllotoxin on α,β-tubulin (pdb id 1SA0) revealed that compound 6n showed excellent binding energies and inhibition constants compared with podophyllotoxin.

We present, for the first time, the organo-N-heterocyclic carbene (NHC) catalyzed 1,3-dipolar cycloaddition of 4β-O-propargyl podophyllotoxin (1) with nitrile oxides to afford 4β-isoxazolepodophyllotoxin hybrids (6a–n) in aqueous-organic media.

Introduction

Nature has consistently offered us a wide variety of bioactive products to treat serious diseases such as cancer, immune system diseases, neurological conditions, and infections.¹ Among these products is podophyllotoxin (1), one of the most prevalent naturally occurring cyclolignans isolated from Podophyllum peltatum L. and Podophyllum hexandrum,² which has been broadly used in clinical studies on diverse malignancies. The mechanisms of action of semi-synthetic derivatives etoposide (2) and teniposide (3) are significantly different from those of the parent podophyllotoxin.^3,4 These two semi-synthetic derivatives inhibit DNA topoisomerase II, whereas the parent podophyllotoxin inhibits the assembly in the microtubulin.^5,6 Even yet, there have been reports on etoposide's toxicity and limitations, including its moderate efficacy, low solubility in water, potential for drug resistance, metabolic inactivation, and other adverse consequences.^7,8

These findings have prompted numerous investigations into the structural modification of etoposide, which includes etopophos (4), that addresses the bioavailability aspect. The substitution at the 4β-position resulting in strong inhibition of topoisomerase II was the most significant change. After a study^9,10 on the substitution of heterocycles for etoposide's C-4 sugar unit, MacDonald and colleagues¹¹ produced a composite pharmacophore model that identified the C-4 molecular area of podophyllotoxin (1) as a potentially variable position for further research. Bulky substituents at C-4 may be advantageous for DNA topo-II inhibition, as further evidenced by comparative molecular field analysis (CoMFA) models revealed by Lee and the group.^11,12 These hypotheses align with the outstanding activity profiles of GL-331 (5), TOP-53 (6), and NK 611 (7).¹³ Interestingly, drug-resistance profiles of GL-331 and TOP-53 differed significantly from that of the parent compound 1, and both compounds demonstrated increased DNA topo-II inhibition and antitumor potential. The activity profiles of these classes of compounds suggest that substitution at position C-4 plays a crucial role and rational modifications at the C-4 position are feasible.¹⁴

One of the main challenges in medicinal chemistry is the development of anticancer agents, as evidenced by severe side effects of current chemotherapeutics, which include non-specific targeting, low solubility and incapability to enter tumour cells.¹⁵ These factors suggest the need for ongoing efforts to develop desired anti-cancer drug-like candidates with minimal side effects.^16,17 As a result, current research is concentrated on the development of novel, safer therapeutic agents that are crucial for clinical use.^18,19 Remarkably, because of their numerous biological uses, N-heterocycles with an oxygen atom are regarded as an important class of compounds in medicinal chemistry.²⁰ In view of this, the selection of an isoxazole ring in the design and synthesis of biologically active compounds is considered to be a better choice, since in isoxazole both O and N atoms are present in adjacent positions and have low bond dissociation energy.²¹ Also, isoxazole consists weak basic character and due to a weak N–O bond, this ring breaks simply by photolysis and thermolysis. Predominantly, deprotonation of isoxazole leads to ring-opening and additional substitutions would lead to well therapeutic activity.²² Many research groups have been working on the development of new isoxazole-based anticancer active compounds (Fig. 1).²³

Fig. 1. Structures of a few anticancer-active podophyllotoxin derivatives.

Synthesis of isoxazole derivatives is extensively carried out through²⁴ cyclomerization, cycloaddition, condensation and functionalization etc. In particular, 1,3-dipolar cycloaddition of nitrile oxides with alkynes has good synthetic value, since it produces biologically useful isoxazoles.^22,23 Thus, Cu(i) and Ru(ii) catalyzed 1,3-dipolar cycloaddition of nitrile oxides with terminal alkynes to access regioselective 3,5-di- and 3,4-disubstituted isoxazoles correspondingly have been well-developed.^24,25 However, considering the probable disproportionation in metal catalysis and advantages observed with organo-catalysis in homogeneous catalysis,²⁶ in 2011, our group developed an organo-N-heterocycliccarbenes (NHCs) catalyzed 1,3-dipolar cycloaddition of nitrile oxides with alkynes to obtain regioselective 3,5-di- and 3,4,5-trisubstituted isoxazoles in DCM solvent,²⁷ since, NHCs are distinct Lewis base (nucleophilic) organocatalysts, they have both σ-basicity and π-acidity properties.²⁸ Starting from initial studies on thiamine-derived NHCs in benzoin,^28a and Stetter reactions,^28b the mechanistic variety of NHCs contingent on their properties has led to the progress of numerous extraordinary C–C and C–X (X = heteroatom) bond formations. Certainly, the isolation of the first stable NHC by Arduengo in 1991 (ref. 29) from imidazolium salts revealed their tunable steric and electronic properties by changing N-substituents on the imidazole ring. In addition, the imidazolium salts, precursors to NHCs, remain stable and easy to handle.

Conversely, organocatalytic reactions have appeared as a substitute synthetic strategy that may eventually lead to large-scale pharmaceutical synthesis applications.³⁰ The development of organocatalytic reactions in aqueous media is especially promising because, generally speaking, organocatalysts are stable in the presence of aqueous media. The use of water as the solvent in N-heterocyclic carbene (NHC)-catalyzed polarity inversion reactions is restricted when compared to enamine catalysis.^28c,31 In 2004, Bode revealed the use of aq. organic (THF–H₂O, 10 : 1) in organo-NHC-catalyzed addition of enals and aldehydes.^28p,32 The similar group then used a stoichiometric amount of water as a reagent or co-solvent in the organo-NHC catalyzed reaction of formylcyclopropanes and α-chloroaldehyde bisulfite salts.³³ In 2010, Rovis and group developed an organo-NHC catalyzed reaction of α-chloro aldehydes in toluene–water solvent media comprising around 1.0 equivalent of water.³⁴

In another study, Hoveyda and group developed an enantioselective addition between dimethylphenylsilyls and to α,β-unsaturated carbonyls in the aqueous medium using organo-NHC-catalysis and results displayed that water is comfortable to organo-NHC catalysis.³⁵ Certainly, traces of water are usually presumed to exist even when anhydrous organic solvents are utilized in organo-NHC-catalyzed reactions. Also, results of the experimental and theoretical studies of Amyes, Diver, Gudat, and Nyulaszi groups show that NHCs are reasonably stable in aqueous environment.³⁶

Considering that thiamine, an NHC precursor, is indispensable for several biological processes that arise mainly in an aqueous atmosphere.³⁷ In 2013, Y. R. Chi and co-workers first time utilized catalytic efficacy of organo-NHC to promote the reaction of enals with enones in pure water.³⁸ Following nature's lead, the use of aqueous solvent media would be an appropriate choice in NHC-catalyzed reactions.

Based on all the above findings and in our ongoing effort to develop aqueous organic synthesis^39a–d and anticancer agents,^39e–k we report an organo-NHC catalyzed 1,3-dipolar cycloaddition between in situ nitrile oxides and 4β-O-propargyl podophyllotoxin to construct new C4-modified isoxazole linked podophyllotoxins in aq. MecCN media. As well, we also investigated the in vitro anticancer activity, caspases activation and in vitro tubulin polymerization of newly synthesized isoxazole-linked podophyllotoxins. Finally, molecular docking studies were carried out for the most potent compounds found in in vitro activity studies. To the best of our knowledge, this is the first report concerning the organo-NHC catalyzed 1,3-dipolar alkyne-nitrile oxide cycloaddition in aqueous organic media to obtain new biologically active isoxazole compounds.

Results and discussion

At first, the synthesis of the key starting materials of the current work such as 4β-O-propargyl podophyllotoxin (1)^39f,40 and chlorooximes (hydroximoyl chlorides) (6a–n)⁴¹ was achieved according to literature procedures.

Later, we concentrated on the development of the optimization reaction conditions from 1,3-dipolar cycloaddition of terminal alkyne 1 with N-hydroxybenzimidoyl chloride (6a) as a model reaction using K₂CO₃ and diverse 5 mol% NHC pre-catalysts (A–G) at RT in (7 : 3) H₂O/MeCN solvent media.

The results revealed that the NHC pre-catalysts (imidazolium salts) (A–G) (Fig. 2) used in this reaction gave desired cycloaddition product (5R,8aR,9S)-9-((3-phenylisoxazol-5-yl)methoxy)-5-(3,4,5-trimethoxyphenyl)-5,8,8a,9-tetrahydrofuro[3′,4′:6,7]naphtha [2,3-d][1,3]dioxol-6(5aH)-one (6a) in 22–76% yields (Table 1, entries 1–7). Out of all, NHC pre-catalyst B gave the best yield (76%) of 6a (Table 1, entry 2). The molecular structure of NHC pre-catalysts had a robust effect on the catalytic efficacy. The NHC pre-catalysts comprising bulky aromatic groups provided good yields of 6a (Table 1, entries 1–2), while, NHC pre-catalysts containing aliphatic groups gave low to moderate yields of 6a (Table 1, entries 3–7). Moreover, lowering the NHC pre-catalyst B concentration from 5 mol% led to a lower yield of 6a (Table 1, entry 8), whereas, increasing NHC pre-catalyst B loading from 5 mol% product did not show any effect on the yield of the product 6a (Table 1, entry 9). Using the optimal NHC pre-catalyst B (Table 1, entry 2), the effect of the bases on product 6a yield was also examined. The results revealed that KHCO₃, DIEA and Et₃N provided low yields (38–52%) of 6a (Table 1, entries 10–12), while, DBU gave a moderate (68%) yield of 6a (Table 1, entry 13). Interestingly, K₂CO₃ was found to be suitable, as it provided a good yield (76%) of 6a (Table 1, entry 2).

Fig. 2. Structures of NHC pre-catalysts used in the present study.

Optimization of reaction conditions^a.

Entry	Catalyst precursor (n mmol)	Base	Solvent	Time (h)	Yieldb (%)
1	A (5)	K2CO3	H2O/MeCN	2	71
2	B (5)	K2CO3	H2O/MeCN	2	76
3	C (5)	K2CO3	H2O/MeCN	3	46
4	D (5)	K2CO3	H2O/MeCN	3	40
5	E (5)	K2CO3	H2O/MeCN	5	31
6	F (5)	K2CO3	H2O/MeCN	5	22
7	G (5)	K2CO3	H2O/MeCN	5	27
8	B (3)	K2CO3	H2O/MeCN	5	35
9	B (7)	K2CO3	H2O/MeCN	2	76
10	B (5)	KHCO3	H2O/MeCN	3	38
11	B (5)	DIEA	H2O/MeCN	2	56
12	B (5)	Et3N	H2O/MeCN	3	52
13	B (5)	DBU	H2O/MeCN	2	68
14	B (5)	K2CO3	H2O/DCM	6	Trace
15	B (5)	K2CO3	H2O/toluene	6	Trace
16	B (5)	K2CO3	H2O/MeOH	2	62
17	B (5)	K2CO3	H2O/EtOH	2	65
18	B (5)	K2CO3	H2O	3	31

^aReaction conditions: (i) intermediate 1 (1 mmol), NHC pre-catalysts A–G (0.05 mmol), base (0.15 mmol), (7 : 3) aqueous/organic solvent media RT and 20 min. (ii) 5a (1 mmol), base (1.5 mmol) and remaining time according to Table 1.

^bIsolated yields after column chromatography.

Finally, we investigated the effect of the solvent system on the yields of 6a, under optimized reaction conditions. The trace amount of 6a was isolated using bi-phasic systems such as aq. DCM (Table 1, entry 14) and aq. toluene (Table 1, entry 15). The yield of 6a was slightly good in the case of aq. MeOH (Table 1, entry 16) and aq. EtOH (Table 1, entry 17). However, the use of pure water gave a low yield (31%) of the product 6a (Table 1, entry 18).

With the above optimal conditions (5 mol% NHC pre-catalyst B, K₂CO₃ and 7 : 3 water/MeCN) in our hand, the current approach was extended to a variety of aromatic aldehydes. In general, aromatic aldehydes containing electron-withdrawing groups provided better yields of the corresponding isoxazoles (Scheme 1, entries 6f–n) than the remaining aldehydes (Scheme 1, entries, 6a–g).

Scheme 1. Synthesis of 4β-isoxazolepodophyllotoxin hybrids (6a–n).

Based on the literature survey,^27,42 we propose a plausible mechanism as shown in Scheme 2. At first, an in situ organo-NHC catalyst (obtained from pre-catalyst B and K₂CO₃) would react with alkyne (1) to form the zwitterion intermediate (H). Later this reactive intermediate interacts with in situ nitrile oxides (obtained from the reaction between chlorooximes (5a–n) and K₂CO₃) through a nucleophilic attack to give additional zwitterion intermediate (I), which, finally undergo C–O hetero-cyclization to give regioselective 3,5-di-substituted 4β-isoxazolepodophyllotoxin hybrids (6a–n).

Scheme 2. A possible mechanism.

In vitro anticancer activity

All the above 4β-isoxazolepodophyllotoxin hybrids (6a–n) were examined for their in vitro anticancer activity against three human cancer cell lines (MCF-7, HeLa and MIAPACA) using podophyllotoxin as a positive control. The results (Table 2) revealed that compound 6e (GI₅₀ = 0.18–0.32 μM) showed greater activity than the podophyllotoxin (GI₅₀ = 0.31–0.51 μM) against three cell lines. As well, compounds 6j and 6n exhibited greater activity (GI₅₀ values 0.23 and 0.12 μM, respectively) than the positive control (GI₅₀ = 0.31 μM) against MCF-7, while, the same compounds (GI₅₀ = 1.03–1.54 μM) had promising activity in comparison to podophyllotoxin (GI₅₀ = 0.51–0.74 μM) against HeLa and MIAPACA. Furthermore, compounds 6i (GI₅₀ = 0.42 μM) and 6m (GI₅₀ = 0.36 μM) had almost similar activity with the podophyllotoxin (GI₅₀ = 0.31 μM) against MCF-7.

In vitro anticancer activity of newly developed podophyllotoxin compounds (6a–n) with GI₅₀ in μM^a.

Compound	R	MCF-7b	HeLac	MIAPACAd
6a	H	2.19 ± 0.33	4.45 ± 1.17	3.98 ± 1.26
6b	4-CH3	3.08 ± 0.54	3.92 ± 1.18	3.22 ± 1.02
6c	3-CH3	3.92 ± 0.82	3.83 ± 1.54	4.67 ± 1.93
6d	3,5-diCH3	1.34 ± 0.23	2.74 ± 0.52	2.28 ± 0.29
6e	4-OCH 3	0.18 ± 0.07	0.32 ± 0.11	0.17 ± 0.05
6f	4-Cl	2.13 ± 0.37	3.05 ± 1.02	3.14 ± 0.32
6g	4-Br	4.32 ± 0.83	7.48 ± 0.69	6.84 ± 2.01
6h	4-F	4.87 ± 0.79	6.12 ± 1.08	13.13 ± 3.27
6i	4-CN	0.42 ± 0.15	3.64 ± 0.93	4.02 ± 0.79
6j	4-NO2	0.23 ± 0.12	1.11 ± 0.29	1.54 ± 0.18
6k	3-Cl	3.15 ± 0.43	9.68 ± 3.12	18.32 ± 2.87
6l	3-CN	2.52 ± 0.91	12.25 ± 2.24	10.22 ± 3.06
6m	3,5-diCl	0.36 ± 0.17	2.19 ± 0.84	2.33 ± 0.18
6n	3,5-diCN	0.12 ± 0.08	1.28 ± 0.25	1.03 ± 0.19
Podophyllotoxin		0.31 ± 0.19	0.74 ± 0.23	0.51 ± 0.12

^a50% growth inhibition and values are the mean of three independent experiments.

^bBreast cancer.

^cCervical cancer.

^dPancreatic cancer.

The nature of the phenyl ring that is attached to isoxazole moiety, affecting in vitro anticancer activities was revealed using structure–activity relationship (SAR) studies. Introducing a strong electron-releasing group (OCH₃) at the 4th position resulted in compound 6e displayed enhanced activity. The compound 6d containing a weak electron-donating group at 3rd and 5th positions (3,5-diCH₃) was ranked second in this category. However, simple phenyl ring compound 6a or compounds 6b and 6c bearing mono-CH₃ groups irrespective of their positions had reduced potency as compared to compounds 6d and 6e (Fig. 3).

Fig. 3. GI₅₀ curves of compounds 6e, 6i, 6j, 6m and 6n and podophyllotoxin.

With respect to the electron-withdrawing group series, compounds 6j and 6n bearing 4-NO₂ and 3,5-diCN groups, respectively, displayed improved activity. The next better activity was shown by compound 6m having the 3,5-diCl group. Compounds 6f and 6i with 4-Cl and 4-CN groups, respectively, had slightly weaker activity than the compound 6m. However, compounds 6g, 6h and 6k containing mono-halogen groups such as 4-Br, 4-F and 3-Cl, respectively, or compound 6l with 3-CN group displayed poorer activity than the remaining compounds in this series (Fig. 4).

Fig. 4. SAR summary.

Caspases activation study

The caspase protease family of enzymes is essential for both the start and completion of apoptosis. When they become active, they cleave many regulatory and structural proteins, which causes the cell to break down internally.⁴³ The traditional indicators of apoptosis, including nuclear condensation, DNA fragmentation, and plasma membrane blebbing, are caused by these proteolytic processes.⁴⁴ The elimination of damaged cells and preservation of cellular homeostasis are two processes in which apoptosis is essential.⁴⁵ The terms “intrinsic pathway” and “extrinsic pathway” refer to the two primary caspase activation mechanisms that lead to apoptosis; they are so named because pressures from the outside or inside of the cell, respectively, often activate them. Cellular stressors such as end plasmicreticulum stress, metabolic stress, and DNA damage are the main causes of the intrinsic pathway's activation. This route is activated by several chemotherapy medications that are used to treat different types of cancer. When these stimuli come together, the mitochondria experience outer membrane permeabilization (MOMP), which releases cytochrome c from the intermembrane space of the mitochondria into the cytosol.

Hence the podophyllotoxin and most potent compounds 6e, 6i, 6j, 6m and 6n observed in the above in vitro anticancer activities were further tested for their caspase activation and results are shown in Table S1† (Fig. 5–9). Out of all, compound 6e showed paramount caspase activation. In accordance with the results, we noted that the activation of caspases by the compounds was found to be concentration-dependent. Among all, compound 6e highly activated caspase 3/7 in MCF-7 cells with 94.5%, which is even higher than the percentage of the podophyllotoxin 91.9% recorded at 20 μg ml⁻¹. Subsequently, the 3/7 caspase activation by 6e was also found significant in MIA PaCa2 cells with 92.3%, which is higher than the percentage value of podophyllotoxin 88.6% recorded at 20 μg ml⁻¹. On the other hand, the caspase 3/7 activation by 6e in HeLa cells was found slightly lower at 81.3% compared to podophyllotoxin at 86.9% recorded at 20 μg ml⁻¹. These results suggest that compound 6e is a good selection for the activation of 3/7 caspase. Next to 6e, compounds 6j and 6i also highly induced caspase 3/7 activation in MCF-7 cells with 90.1% and 89.3%, respectively, recorded at 20 μg ml⁻¹. The results are compared with 91.9% recorded with podophyllotoxin at 20 μg ml⁻¹. While the different caspase action by the remaining compounds was found average or least. Compound 6m exhibited the least caspase 3/7, 8, and 9 activations with 44.8%, 53.4, 65.0%, 63.0%, 41.9% and 51.7%, respectively, in HeLa, MCF-7 and MIA PaCa2 cancer cells.

Fig. 5. Effect of compound 6e against the activation of different caspases in selected cancer cell lines.

Fig. 6. Effect of compound 6i against the activation of different caspases in the selected cancer cell lines.

Fig. 7. Effect of compound 6j against the activation of different caspases in selected cancer cell lines.

Fig. 8. Effect of compound 6m against the activation of different caspases in selected cancer cell lines.

Fig. 9. Effect of compound 6n against the activation of different caspases in selected cancer cell lines.

Tubulin polymerization inhibition

Targeting the microtubules in the development of anticancer drugs was considered to be an important goal in current medicinal chemistry,⁴⁶ as they are involved in a few cellular functions, such as cell division, organelle transport, motility and keeping signal transduction.⁴⁷ A few antimitotic agents such as Vinblastine, Colchicine, and Vincristine inhibit tubulin polymerization via their binding to the colchicine or vinca-binding sites of tubulins.⁴⁸ Curiously, for the past few years, few anti-mitotic agents such as taxanes and vinca alkaloids were used for clinical trials of varied cancer patients.^46,49 However, little solubility and oral bioavailability and slightly high toxicity make these anti-mitotic agents less in clinical trials of cancer,^46–50 which were made a crucial need to develop novel anti-mitotic agents. Interestingly, a few reports of C4-ring-modified podophyllotoxin⁵¹ and a few isoxazole⁵² derivatives have been recognised to inhibit tubulin polymerization.

Hence, we studied the in vitro tubulin polymerization inhibition studies of compounds 6e, 6i, 6j, 6m and 6n using a tubulin assembly assay. The results revealed that compound 6e (IC₅₀ = 0.88 μM) showed comparable activity with the podophyllotoxin (IC₅₀ = 0.84 μM). As well, compounds 6j (IC₅₀ = 1.32 μM) and 6n (IC₅₀ = 1.43 μM) have shown promising potency in comparison to positive control. However, compounds 6i (IC₅₀ = 4.85 μM) and 6m (IC₅₀ = 3.79 μM) had moderate activity as compared to the positive control (Table 3).

Tubulin polymerization inhibition studies of compounds 6e, 6i, 6j, 6m and 6n.

Compound	IC50 (μM)
6e	0.88
6i	4.85
6j	1.32
6m	3.79
6n	1.43
Podophyllotoxin	0.84

Molecular docking studies

The molecular docking studies of most potent compounds 6e, 6j and 6n found in the above in vitro studies and positive control podophyllotoxin were carried out by taking α,β-tubulin (pdb id 1SA0) as target⁵³ and results are shown in Table 4. It was found that compound 6n displayed greater binding energy (−9.39 kcal mol⁻¹) and showed 130.79 nM inhibition constant. Compound 6e was ranked second in this study, with its binding energy as −7.99 kcal mol⁻¹ and 1.38 μM as the inhibition constant. Compound 6j displayed binding energy (−7.84 kcal mol⁻¹) and showed a 1.84 μM as the inhibition constant. On the other aspect, positive control podophyllotoxin showed binding energy = −9.22 kcal mol⁻¹ and 173.47 nM as the inhibition constant.

Molecular docking results of 6e, 6j and 6n and podophyllotoxin.

Entry	Binding energy (kcal mol−1)	No. of H-bonds	Inhibition constant
6e	−7.99	—	1.38 μM
6j	−7.84	—	1.80 μM
6n	−9.39	—	130.79 nM
Podophyllotoxin	−9.22	—	173.47 nM

With respect to binding interaction, podophyllotoxin showed π–π stacking with TYR224 residue and compound 6j formed salt bridge with ARG2 residue.

Overall results revealed that the compound 6n showed an encouraging binding energy and inhibition constant than the podophyllotoxin (Fig. 10 and 11).

Fig. 10. Effect of podophyllotoxin against the activation of different caspases in the selected cancer cell lines.

Fig. 11. Tubulin polymerization inhibition studies.

Conclusions

Herein, we, first time developed an aqueous organo-NHC catalyzed 1,3-dipolar cycloaddition between 4β-O-propargyl podophyllotoxin (1) and in situ nitrile oxides to afford regioselective 3,5-di-substituted 4β-isoxazolepodophyllotoxin hybrids (6a–n). Compound 6e was the most potent analogue showing greater anti-cancer activity against MCF-7, HeLa and MIAPACA cell lines with GI₅₀ values of 0.18, 0.32 and 0.17 μM, respectively, compared to the standard drug podophyllotoxin (GI₅₀; 0.31, 0.74 and 0.51 μM, respectively). Compounds 6j and 6n showed greater activity (GI₅₀ = 0.23 and 0.12 μM, respectively) than the positive control (GI₅₀ = 0.31 μM) against the MCF-7 cell line. Furthermore, compounds 6i (GI₅₀ = 0.42 μM) and 6m (GI₅₀ = 0.36 μM) had closer activity to the podophyllotoxin (GI₅₀ = 0.31 μM) against the MCF-7 cell line. The results of in vitro tubulin polymerization inhibition revealed that compound 6e (IC₅₀ = 0.88 μM) displayed almost similar activity with podophyllotoxin (IC₅₀ = 0.84 μM). The molecular docking studies of potent compounds 6e, 6j, 6n and podophyllotoxin as tubulin polymerization inhibitors were found to be supportive of the corresponding in vitro activity data.

Data availability

Data will be made available on the reader's request.

Conflicts of interest

There are no conflicts to declare.