Isoflavone Derivatives as Potential Anticancer Agents: Synthesis and Bioactivity Studies

Abstract

Isoflavones are phenolic natural compounds with a C₆C₃C₆ framework. They possess a plethora of biological activities that are associated with putative benefits to human health. In particular, the cancer chemopreventive and chemotherapeutic potential of isoflavones has attracted the interest of researchers. Several isoflavone derivatives have been synthesised and probed for their anticancer activities. The isoflavone analogues are mainly synthesised by molecular hybridisation and other strategies that enable diversification through early or late‐stage functionalisation of A‐, B‐ and C‐rings of the isoflavones. This has resulted in the discovery of isoflavone analogues with improved antiproliferative activities against several cancer cells and different mechanisms of action. In this review, the synthesis of isoflavone derivatives and their anticancer activity studies are discussed.

Several isoflavone analogues with potential anticancer activities have been discovered. They include tubulin polymerase, kinases (EGFR, PI3Kδ, PI3 Kγ) HDACs and angiogenesis inhibitors. They also block signalling pathways such as Hedgehog, PI3 K/AKT/mTOR, MAPK/Wnt, EGFR/PI3 K/Akt/Bad, EGFR/ERK and EGFR/PI3 K/Akt/b‐catenin.

1. Introduction

Isoflavones are phenolic secondary metabolites commonly found in the Leguminosae family.[ ¹ , ² ] Isoflavone compounds consist of two aromatic systems, the A‐ring and the B‐ring adjoined to the heterocyclic pyran‐4‐one moiety (C‐ring). They differ from other classes of flavonoids by the position of the B‐ring, which is at C‐3 (Figure 1). Isoflavones have been reported to exhibit a myriad of biological activities that include anti‐inflammatory, anti‐microbial, vasorelaxation, cardioprotective, neuroprotective, chemoprotective and antiproliferative activities.[ ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ ] The anticancer activities of isoflavones have been of great interest to researchers. In vivo and in vitro antiproliferative activity studies show that isoflavones attenuate the growth of different human cancer cells including breast, prostate, colon, pancreatic, lung and others.[ ¹⁰ , ¹¹ , ¹² , ¹³ ] Furthermore, isoflavones have shown the potential to block tumour invasion and metastasis and suppress cancer stem cells.[ ¹⁰ , ¹¹ , ¹⁴ , ¹⁵ ] The anti‐tumour activities of isoflavones are attributed to different mechanisms of action and their ability to inhibit multiple signalling pathways that are involved in tumourigenesis, invasion and migration.[ ¹¹ , ¹⁵ , ¹⁶ ]

Figure 1.

Isoflavone core structure and isoflavone analogues with potential anticancer activity.

Although natural isoflavones offer promise for drug development, their progression in the drug discovery pipeline is hampered by several factors that include poor solubility, bioavailability and selectivity.[ ¹⁷ , ¹⁸ ] Therefore, isoflavone derivatives have been prepared to improve their physicochemical properties and efficacy,[ ¹⁹ , ²⁰ ] as well as chemical accessibility ^[21] by reducing the complexity of some of the isoflavone structures, while retaining the activity. ^[22] In some instances, isoflavone derivatives were synthesised to confer favourable properties and reduce the toxicity of existing chemotherapeutic drugs and other potential anticancer compounds.[ ²³ , ²⁴ , ²⁵ ] The isoflavone derivatives have mainly been prepared by molecular hybridisation,[ ²⁶ , ²⁷ , ²⁸ ] whereby natural or unnatural isoflavone compounds were amalgamated with other privileged pharmacophores, other natural compounds or chemotherapeutic drugs.[ ¹⁹ , ²⁹ , ³⁰ , ³¹ , ³² ] Chemical space expansion could also be attained by the replacement of the B‐ring with other heterocyclic ring structures or by key scaffolds that mimic other bioactive compounds.[ ³³ , ³⁴ , ³⁵ ] Other strategies included diversity‐oriented synthesis (DOS)[ ³⁶ , ³⁷ ] involving divergent transformation and elaboration of functional groups and substituents in isoflavones.[ ³⁸ , ³⁹ ] Computer‐aided drug design (CADD) including scaffold hoping has also been employed to guide the synthesis of isoflavone analogues with improved properties.[ ⁴⁰ , ⁴¹ , ⁴² ] These efforts have resulted in the discovery of isoflavone derivatives that exhibit enhanced antiproliferative activities than the parent isoflavones,[ ⁴⁰ , ⁴³ , ⁴⁴ ] inhibit proliferation of resistant cancer cells, ^[34] and sensitise cells to chemotherapeutic drugs and radiation.[ ⁴⁵ , ⁴⁶ ] The isoflavone derivatives elicit their anticancer activities as mitotic,[ ³⁴ , ⁴⁷ , ⁴⁸ ] kinases (EGFR, PI3Kδ, PI3 Kγ),[ ⁴⁰ , ⁴² , ⁴⁹ , ⁵⁰ , ⁵¹ , ⁵² ] SIRT1, ^[53] and angiogenesis inhibitors. ^[44] Moreover, analogues that supress signalling pathways that include Hedgehog, ^[41] PI3 K/AKT/mTOR, ^[45] MAPK/Wnt, ^[19] EGFR/PI3 K/Akt/Bad, EGFR/ERK and EGFR/PI3 K/Akt/b‐catenin ^[40] have been discovered. Examples of the antiproliferative isoflavone analogues are shown in Figure 1.[ ³⁴ , ⁴⁰ , ⁴¹ , ⁴⁴ , ⁵¹ , ⁵³ ]

This review discusses the synthesis and anticancer activity studies of isoflavone derivatives optimised through different strategies including molecular hybridisation, CADD, DOS and functional group manipulation. They consist of analogues of natural isoflavones such as genistein,[ ²³ , ²⁹ ] formononetin,[ ¹⁹ , ²⁴ , ²⁵ , ⁴⁰ ] glaziovianin A[ ¹² , ⁴⁷ ] and prenylated isoflavone derivatives, including barbigerone,[ ⁴⁴ , ⁴⁸ ] 4’‐O‐methylgrynullarin[ ¹³ , ²² ] and glabrescione B derivatives. ^[41] Other non‐natural isoflavone analogues with structural modification at different positions on the A‐, B‐ and C‐rings are also discussed. They comprise B‐ring replaced derivatives,[ ³³ , ³⁴ ] including antimitotic benzo[b]thiophene (BT) analogues. ^[34] Others encompass analogues inspired by isoflavone‐bearing kinase inhibitors such as tenalisib, which is undergoing phaseI/II clinical trials and umbralisib, which was approved for treatment of different types of lymphoma and later withdrawn.[ ⁴⁹ , ⁵¹ ] The final set of compounds under discussion are the isoflavone‐inspired preclinical and clinical super‐benzopyran analogues.[ ⁵⁴ , ⁵⁵ , ⁵⁶ ] The focus is on reports published from 2012 to March 2024.

2. Genistein Derivatives

Genistein (1) is the main metabolite of Glycine max (soy). ^[57] Genistein (1) and other soy isoflavones are regarded as phytoestrogens, due to their ability to modulate estrogenic effects. ^[58] Apart from their potential role as phytoestrogens, soy isoflavones have also been reported to exhibit several biological activities including chemopreventive and anticancer activity.[ ⁵⁹ , ⁶⁰ ] Clinical studies have been conducted for genistein (1) and other soy isoflavones, daidzein and glycitein in patients with prostate and urothelial bladder cancer.[ ¹⁰ , ¹⁶ ]

Several derivatives that link genistein (1) with synthetic or natural scaffolds have been prepared and evaluated for potential anticancer activities.[ ¹⁸ , ²⁰ , ⁶¹ , ⁶² ] Some of the synthesised derivatives exhibited improved activity than genistein (1).[ ¹⁸ , ⁶¹ ] A detailed review of the anticancer potential of genistein (1) and its derivatives was written by Gupta and colleagues in 2022. ^[62] Recent studies involved the synthesis of 5‐fluorouracil‐genistein hybrids, ^[23] triazine‐genistein derivatives ^[29] and fluorinated genistein analogues. ^[63]

5‐Fluorouracil‐genistein analogues were prepared and evaluated for antiproliferative activity against human colon adenocarcinoma cells (SW480 and SW620) and non‐malignant cell lines (HaCaT and CHO−K1). ^[23] Some of the synthesised derivatives showed enhanced antiproliferative activity than that of 5‐fluorouracil. The synthesis commenced with the preparation of the main precursors, N‐propargylated 5‐fluorouracil 5 and genisteinalkylazides 3 a–h (Scheme 1). Compound 5 was prepared by N‐alkylation of 5‐fluorouracil (4) with propargyl bromide, while genisteinalkylazides 3 a–h were prepared by 7‐O‐etherification of genistein (1) with dibromoalkanes of different chain lengths, followed by ultrasound (US)‐mediated nucleophilic substitution with sodium azide. Click reaction of genisteinalkylazides 3 a–h and propargyl 5 rendered the hybridised genistein analogues 6 a‐h. Of the synthesised compounds, 6 a showed antiproliferative activity with IC₅₀ values of 62.73±7.26 and 50.58±1.33 μM for SW480 and SW620 cells, respectively. This activity was greater than that of 5‐fluorouracil (174.3±19.10 and 180.90±18.80 μM, for SW480 and SW620 cells, respectively) and that of genistein (1) (75.84±5.83 μM) against SW620. ^[23]

Scheme 1.

Synthesis of 5‐Fluorouracil‐genistein analogues. Reagents and conditions: a) DIPEA, DMF, 61–71 %; b) NaN₃, DMF, 40 °C, US, 78–89 %; c) DIPEA, KI, DMF, US, 40 %; d) Ascorbic acid, Cu(OAc)₂, DMF‐H₂O, 40 °C, US, 43–93 %.

Zou and colleagues synthesised 1,3,5‐triazine analogues of genistein. ^[29] Firstly, they reacted 2,4,6‐trichloro‐1,3,5‐triazine (7) with secondary amines to give aminotriazine derivatives 8 a‐l (Scheme 2). The reaction of 8 a‐l with genistein (1) gave the 7‐O and 4’‐O‐disubstituted analogues (9 a‐l). The triazine‐genistein derivatives were assessed for antiproliferative activity against MDA‐MB‐231, HeLa, HCT‐116 and Huh‐7 cancer cell lines. Most of the synthesised compounds exhibited superior antiproliferative activity than the parent compound genistein (1). Compound 9 i exhibited the most potent activity against MDA‐MB‐231 and HeLa cell lines with IC₅₀ values of 23.13±1.29 and 39.13±0.89 μM, respectively. Compound 9 a exhibited the best activity against HCT‐116 cell line (IC₅₀=18.40±3.41 μM) and compound 9 b showed better activity against Huh‐7 cell line (37.56±1.92 μM). Compound 9 i was subjected to further studies on cell migration, invasion and adhesion, as well as in vivo studies in xenograft models of MDA‐MB‐231 cells. The results showed that 9 i could inhibit the migration, invasion and adhesion of MDA‐MB‐231 cells, and also inhibit the proliferation of MDA‐MB‐231 tumour in xenografts. ^[29]

Scheme 2.

Synthesis of 1,3,5‐triazine analogues of genistein. Reagents and conditions: a) Secondary amine, Acetone, K₂CO₃, −20 °C; b) Acetone, K₂CO₃, rt, 69–81 %.

3. Formononetin Derivatives

Formononetin (10) is mainly found in red clover plant and is widespread in other plants of the Leguminosae family. ^[64] Formononetin has been reported to exhibit many biological activities that include antioxidant, antitumor, neuroprotective, antihypertensive, antibacterial and antiviral effects.[ ⁶⁴ , ⁶⁵ , ⁶⁶ ] Several derivatives of formononetin have been synthesised and evaluated for their potential anticancer activities.[ ¹⁹ , ²⁴ , ²⁵ , ⁴⁰ ]

Formononetin N‐mustard derivatives were synthesised and evaluated for cytotoxicity against a panel of cancer cell lines. ^[24] The synthesis is outlined in Scheme 3. B‐ring nitration of formononetin (10) gave 3’‐nitroformononetin (11) in a 75 % yield. Alkylation of the 7‐hydroxy group with different alkyl bromides or by Mitsunobu reaction yielded 7‐O‐alkylated derivatives 12 a‐o. The zinc‐mediated reduction of the nitro group gave 3’‐aminoisoflavones 13 a‐o in 78–85 % yields, which were N‐alkylated with ethylene oxide in the presence of AcOH to render compounds 14 a‐o. Treatment of 14a‐o with SOCl₂ in dichloromethane gave the final compounds 15 a‐o.

Scheme 3.

Synthesis of formononetin nitrogen mustard derivatives. Reagents and conditions: a) concentrated HNO₃, concentrated H₂SO₄, AcOH, 50 °C, 12 h, 75 %; b) Method A: K₂CO₃, R−Br, acetone, reflux, 3–8 h, 79–87 % (for 12 a‐l); Method B: DIAD, PPh₃, THF, R‐OH, 0 °C to rt, overnight, 65–77 % (for 12 m‐o); c) Zn, EtOH, HOAc, reflux, 1–2 h, 78–85 %, d) ethylene oxide, HOAc, rt, overnight, 70–82 %; e) SOCl₂, DCM, reflux, 1.5 h, 66–75 %.

Another targeted compound 17 with a free hydroxy group at the 7 position was prepared from 14 j as shown in Scheme 4. Benzyl deprotection of 14 j and subsequent treatment of the resulting 7‐hydroxyisoflavone 16 with SOCl₂ in dichloromethane gave the targeted compound 17. ^[24]

Scheme 4.

Synthesis of formononetin nitrogen mustard derivative 17. Reagents and conditions: a) H₂, 10 % Pd/C, MeOH, rt, 1 h, 90 %; b) SOCl₂, DCM, reflux, 1.5 h, 70 %.

The synthesised formononetin N‐mustard hybrids were evaluated for cytotoxicity against cancer cell lines that included SH‐SY5Y, HCT‐116, DU‐145, Hela and SGC‐901. ^[24] All synthetic derivatives showed better antiproliferative activity against HCT‐116, SH‐SY5Y and DU‐145 cell lines compared to formononetin (10). Compound 15 o exhibited activity against SH‐SY5Y cell line with IC₅₀ value of 2.08 μM and compound 15 n exhibited activity against Hela cell line with IC₅₀ value of 8.29 μM. Compounds 15 d and 17 showed almost equipotent activity against HCT‐116, SH‐SY5Y and SGC‐7901 cell lines with IC₅₀ values of 3.8, 2.17 and 9.21 μM, respectively for 15 d; and IC₅₀ values of 4.1, 2.7 and 8.2 μM, respectively for 17. Further studies demonstrated that compounds 15 d and 15 n could induce cell cycle arrest at G2/M phase and cell apoptosis. ^[24]

Several formononetin‐dithiocarbamate conjugates were synthesised and evaluated for antiproliferative activity against MGC‐803, EC‐109 and PC‐3 cell lines. All the synthesised derivatives were more active than formononetin. Of the synthesised compounds, 19 was identified as the most potent compound against MGC‐803, EC‐109 and PC‐3 cells (IC₅₀=6.07±0.88, 3.54±1.47 and 1.97±0.01 μM, respectively). In addition, compound 19 showed superior activity compared to 5‐FU on all cell lines. Mechanistic studies showed that 19 induced cell cycle arrests in G1 phase and exhibited concentration‐dependent downregulation of cyclin D1 and CDK4 proteins. Further studies demonstrated that 19 could reduce PC‐3 cell growth and migration via MAPK/Wnt signalling pathways. ^[19]

Compound 19 was prepared by alkylation of the 7‐hydroxy group of formononetin (10) with 1,3‐dibromopropane to give compound 18. The reaction of tert‐butyl piperazine‐1‐carboxylate with CS₂ followed by coupling with 18 in the presence of Na₃PO₄.12H₂O gave 19 (Scheme 5). ^[19]

Scheme 5.

Synthesis of formononetin‐dithiocarbamate conjugate 19. Reagents and conditions: a) 1,3‐dibromopropane, K₂CO₃, THF, reflux, 70–83 % yield; (b) CS₂, tert‐butyl piperazine‐1‐carboxylate, Na₃PO₄⋅12H₂O, acetone, rt, 78–85 % yield.

7‐O Derivatives of formononetin that target human epidermal growth factor receptor (EGFR) were designed based on CADD method and synthesised by Lin and colleagues. ^[40] The synthesised compounds were first assessed for inhibition of EGFR. All synthesised compounds showed greater potency at 150 nm than formononetin (10). The most active compound, 22 exhibited 87.3 % inhibition and IC₅₀ value of 14.5 nm against EGFR. In vitro antiproliferative study on four cancer cells, MCF‐7, MDA‐MB‐231, H460 and H1650, and non‐cancer cells, L02 and VERO revealed 22 to be the most active compound. It showed cell growth inhibition with IC₅₀ values of 11.5±1.52, 5.44±1.28, 6.36±1.55 and 7.26±1.02 μM against MCF‐7, MDA‐MB‐231, H460 and H1650, respectively. It was less active on the non‐cancer cells L02 and VERO (IC₅₀>100 and 95.2±4.72 μM, respectively). Studies on the mode of action of 22 for the observed activity in MDA‐MB‐231 cells showed that it induced apoptosis by regulating EGFR/PI3 K/Akt/Bad pathway, inhibited cell growth by blocking EGFR downstream Ras/Raf/MEK/ERK signalling pathway and migration by targeting EGFR/PI3 K/Akt/b‐catenin pathway. Compound 22 did not induce cell cycle arrest and had no significant effect on Cyclin A, Cyclin D1, CDK4 protein expression. In vivo anti‐tumour study using female nude mice showed that 22 greatly inhibited tumour growth with effectiveness similar to that of the positive control, lapatinib. ^[40]

The synthesis of 22 is depicted in Scheme 6. The reaction of formononetin (10) with ethyl bromoacetate gave compound 20, which was converted into hydrazide 21. The reaction of hydrazide 21 with 4‐benzyloxybenzaldehyde gave the most active formononetin hydrazide derivative 22. ^[40]

Scheme 6.

Synthesis of formononetin hydrazide derivative 22. Reagents and conditions: a) ethyl bromoacetate, acetone, K₂CO₃, reflux, 10 h, 95 %; b) hydrazine hydrate, ethanol, reflux, 10 h, 87 %; c) 4‐benzyloxybenzaldehyde, glacial acetic acid, ethanol, reflux, 6 h, 69 %.

Three novel compounds were prepared by hybridisation of an anticancer lignan, podophyllotoxin (23) and isoflavone, formononetin (10). The target compounds were obtained by esterification of podophyllotoxin (23) with chloroacyl chloride and subsequent nucleophilic substitution reaction with formononetin (10) (Scheme 7). ^[25] The conjugates 24 a‐c were evaluated for cytotoxicity against SKOV3, MCF7, HepG2, HeLa, A549, CT26, B16f10 and HUVEC cancer cell lines. The three synthesised compounds exhibited potent cytotoxic activity than formononetin (10) in all cancer cell lines. Compound 24 a exhibited improved activity against A549 human lung carcinoma cell line (IC₅₀=0.753±0.173 μM) than podophyllotoxin (10) (IC₅₀=1.934±0.089 μM). ^[25] Mechanistic studies demonstrated that 24 a could reduce caspase‐8 expression, induce apoptosis and disrupt microtubule network in A549 cells. Furthermore, 24 a showed potential to inhibit migration and invasion of A549 cells.

Scheme 7.

Synthesis of podophyllotoxin‐formononetin hybrid. Reagents and conditions: a) Et₃N, DCM; b) 10, K₂CO₃, KI, DMF, 64–85 % over two steps.

Coumarin‐formononetin hybrid 29 was prepared by Yao and colleagues and evaluated for antiproliferative activity against three gastric cancer cell lines (SGC7901, MKN45 and MGC803). ^[67] The synthesis commenced with the preparation of two main precursors, 7‐O‐propargylformononetin (25) and 7‐O‐alkylazide coumarin 28 (Scheme 8). Compound 25 was prepared by alkylation of 10 with propargyl bromide using NaOH as a base. Compound 28 was prepared by the reaction of coumarin (26) with dibromopropane to give bromoalkylcoumarin 27 and subsequent displacement of bromine with NaN₃. Click reaction of compounds 25 and 28 in the presence of CuSO₄.5H₂O gave the requisite triazole bridged coumarin‐formononetin hybrid 29. The antiproliferative results of the three gastric cancer cell lines showed that 29 potently inhibited SGC7901 with IC₅₀ value of 1.07 μM. Compound 29 was also evaluated for inhibition of SIRT1 activity and it exhibited inhibitory activity with IC₅₀ value of 2.52 μM. Furthermore, compound 10 was reported to inhibit SGC7901 growth and migration by Wnt/β‐Catenin and AKT/mTOR signalling pathways and show in vivo antitumor activity. ^[67]

Scheme 8.

Synthesis of the formononetin‐coumarin hybrid 29. Reagents and conditions: a) propargyl bromide, NaOH, acetone, reflux; b) 1,3‐dibromopropane, K₂CO₃, DCM, reflux, 50.7 %; c) NaN₃, CH₃CN, reflux, 41.3 %; d) intermediate 25, CuSO₄.5H₂O, sodium ascorbate, DMSO:H₂O (1 : 1), rt, 32.9 %.

4. Glaziovianin A Derivatives

Glaziovianin A (30) was isolated from the leaves of a Brazilian tree, Ateleia glazioviana Baillon (Leguminosae) through bioassay‐guided fractionation against HL‐60 leukemia cells. ^[12] It was determined to exhibit cytotoxic activity against HL‐60 cell line with IC₅₀ value of 0.29 μM. ^[12] It also exhibited differential cytotoxicity in a panel of 39 cell lines from the Japanese Foundation for Cancer Research ^[12] and was determined to be a microtubule dynamics inhibitor. ^[68] Owing to its biological activities, different research groups have synthesised glaziovianin A (30)[ ⁶⁹ , ⁷⁰ , ⁷¹ ] and its derivatives.[ ⁴⁷ , ⁷⁰ , ⁷² , ⁷³ , ⁷⁴ ] Hayakawa and colleagues prepared several derivatives of glaziovianin A with modifications on the A‐ and B‐rings.[ ⁴⁷ , ⁷³ , ⁷⁴ ] Some of the synthesised compounds showed improved cytotoxic activity against HeLa S3 cells than glaziovianin A (30). ^[47] Interestingly, the 6‐O‐benzyl derivative 31 was discovered to be an α,β‐tubulin inhibitor, ^[47] while the 7‐O‐benzyl derivative, gatastatin (32), was determined to be a specific γ‐tubulin inhibitor (Figure 2). ^[75] However, gatastatin (32) showed less potent activity against HeLa S3 cells. ^[47] Recently, more active derivatives of 32 were synthesised by modification at C‐6. ^[74] These included the 6‐O‐propargyl derivative, gatastatin G2 (33), which showed improved activity against HeLa S3 cells and γ‐tubulin, than gatastatin (32).

Figure 2.

Glaziovianin A and derivatives.

The synthesis of glaziovianin A derivatives was based on the Suzuki‐Miyaura coupling reaction of halochromones with boronic esters.[ ⁷⁶ , ⁷⁷ ] As shown in Scheme 9, 6‐O‐benzylglaziovianin A (31) was prepared by coupling of iodochromone 37 with boronate ester 38. ^[47] The iodochromone 37 was in turn synthesised by a sequence of steps that involved the conversion of sesamol (34) into an appropriately substituted acetophenone 35,[ ⁴⁷ , ⁷⁸ ] followed by condensation with DMF‐DMA to give enaminone 36, and finally iodine‐mediated cyclisation by modified Gammill procedure. ^[79]

Scheme 9.

Synthesis of 6‐O‐benzylglaziovianin A (31). Reagents and conditions: a) DMF‐DMA, 95 °C; b) I₂, py, CHCl₃, rt, 86 % in two steps; c) 38, PdCl₂(dppf).DCM, 1 M Na₂CO₃ aq, 1,4‐dioxane, rt, 76 %.

Recently the synthesis of gatastatin (32) and other modified A‐ and B‐ring analogues led to the discovery of the more potent specific γ‐tubulin gatastatin G2 (33). ^[74] The synthesis of gatastatin analogues was initiated from isovanillin (39), from which the main precursors, 3‐iodochromones 40 and 41 were prepared (Scheme 10). The Suzuki–Miyaura reaction of iodochromones 40 and 41 with different boronate esters and subsequent derivatisation at C‐6 yielded several gatastatin analogues. Specifically, gatastatin (32) was synthesised by the Suzuki coupling of iodochromone 40 with boronate ester 38, while gatastatin G2 (33) was synthesised by coupling iodochromone 41 with 38, followed by deprotection of the resulting isoflavone 42 and alkylation of 43 with propargyl bromide (Scheme 10). ^[74]

Scheme 10.

Synthesis of gatastatin analogues. Reagents and conditions: a) 38, PdCl₂(dppf)⋅DCM, 1 M Na₂CO₃ aq, 1,4‐dioxane, rt, 61 % for 32 and 97 % for 42; b) p‐TSOH⋅H₂O, CHCl₃, MeOH, rt, 75 % c) propargyl bromide, K₂CO₃, acetone, reflux, 97 %.

5. Derivatives of Prenylated Isoflavones

Prenylated isoflavones constitute the largest group of isoflavones. ^[2] They refer to isoflavones bearing a C5‐isoprenoid unit and other long‐chain units (geranyl, farnesyl, etc). ^[80] Modification of the prenyl substituents can involve reduction, hydroxylation and epoxidation of the double bond or cyclisation of prenyl chain with adjacent hydroxy groups of phenols resulting in pyrano‐ and furanoisoflavones. ^[80] Prenylated isoflavones exhibit a myriad of biological activities, including anticancer activity.[ ¹³ , ²² , ⁴⁴ , ⁸¹ , ⁸² ] Owing to their interesting structures and bioactivities, several research groups have embarked on the synthesis of prenylated isoflavones and their derivatives.[ ²² , ⁴⁴ , ⁸³ , ⁸⁴ , ⁸⁵ , ⁸⁶ , ⁸⁷ ]

5.1. Barbigerone Derivatives

Barbigerone (44), an angular pyranoisoflavone first isolated from the seeds of Tephrosia barbigera ^[88] was reported to exhibit cytotoxicity against several cancer cell lines that include HepG2, C26, LL2 and B16. ^[81] In addition, it exhibited apoptotic‐inducing effects and sensitised adriamycin (ADR)‐resistant human breast carcinoma (MCF‐7/ADR) cells.[ ⁸¹ , ⁸² ] Barbigerone derivatives with modifications on the A‐ and B‐rings were synthesised by Wang and colleagues. ^[44] To keep the A‐ring intact, 3‐iododimethylpyranochromone 47 was coupled to various boronic acids leading to barbigerone derivatives 48 a‐w with different substituents on the B‐ring (Scheme 11). The dimethylpyranochromone 47 was prepared by the reaction of 3‐iodo‐7‐hydroxychromone (46) with 1,1‐diethoxy‐3‐methyl‐2‐butene in the presence of 3‐picoline. The chromone 46 was derived from 2,4‐dihydroxyacetophenone (45) (Scheme 11). ^[44]

Scheme 11.

Synthesis of B‐ring modified barbigerone analogues. Reagents and conditions: a) 1,1‐diethoxy‐3‐methyl‐2‐butene, 3‐picoline, xylene, reflux, 24 h (48.4 %); b) ArB(OH)₂, 10 % Pd/C, Na₂CO₃, H₂O, DME, 45 °C, 1 h (49.6–89.7 %).

The synthesis of the A‐ring modified analogues was initiated by preparing the boronic acid 51 by bromination of trimethoxybenzene 49 and subsequent treatment of the resulting bromobenzene 50 with n‐BuLi and trimethyl borate. Suzuki‐Miyaura coupling of boronic acid 51 with iodochromone 52, followed by deprotection gave an isoflavone precursor 54. Alkylation or esterification gave 7‐O derivatives 55 a‐x (Scheme 12). ^[44]

Scheme 12.

Synthesis of A‐ring modified barbigerone analogues. Reagents and conditions: a) Br₂, CH₂Cl₂, 0 °C (92.3 %); b) n‐BuLi, trimethyl borate, THF, −78 °C (37.7 %); c) 10 % Pd/C, Na₂CO₃, H₂O, DME, 45 °C, 1 h (61.4 %); d) p‐TsOH, CH₃OH, THF, 60 °C, 1 h (87.1 %); e) RX, K₂CO₃, acetone, rt, overnight, or RCOOH, DCC, DMAP, DCM, rt, overnight (18.7–99.4 %).

The barbigerone analogues 48 a–w and 55 a–x were evaluated for antiproliferative activity against cancer cell lines that included HepG2, A375, U251, B16, HCT116 and HUVEC. Most derivatives bearing the dimethylpyran scaffold in the A‐ring showed reduced cytotoxic activity against all cancer cell lines compared to barbigerone (44). The exception was 48 o, which showed improved activity against U251 cell line (2.50 μM) than barbigerone (4.10 μM). Several 7‐O derivatives exhibited antiproliferative activity with IC₅₀ below 10 μM. The most active derivative 55 a, exhibited cytotoxic activity with IC₅₀ values of 0.28, 1.58, 3.50, 1.09, 0.68 and 3.80 μM, against HepG2, A375, U251, B16, HCT116 and HUVEC cells, respectively. Furthermore, compound 55 a showed anti‐angiogenic activities.

Following the discovery that barbigerone (44) inhibits tubulin polymerisation and the determination by X‐ray crystal structure that it binds to colchicine‐binding site, Yang and Chen's group designed and synthesised a more active barbigerone analogue 59. ^[48] The compound 59 was synthesised by coupling boronic acid 51 with 3‐iodochromone 57, and subsequent reaction of resulting isoflavone 58 with bromobenzylbromide (Scheme 13). Compound 59 exhibited antiproliferative activity against H460, Ramos, HeLa and HCT116 with IC₅₀ values of 0.46±0.14, 0.62±0.09, 0.17±0.11 and 0.12±0.07 μM, respectively. Furthermore, both 44 and 59 inhibited tubulin polymerisation, induced G2/M phase cell cycle arrest and apoptosis and exhibited in vivo anticancer activity in H460 xenograft model. ^[48] The results from the two studies by Chen's group indicate that the presence of the dimethylpyran ring is not significant for the anticancer activity of barbigerone derivatives, but keeping the substitution pattern of the B‐ring intact is important for the activity of the analogues.

Scheme 13.

Synthesis of barbigerone analogue 59. Reagents and Conditions: a) 51, 10 % Pd/C, Na₂CO₃, H₂O, DME, 45 °C, 65.8 %; b) 1‐bromo‐2‐(bromomethyl)benzene, K₂CO₃, MeCN, reflux, 92.2 %.

5.2. 4’‐O‐Methylgrynullarin Derivatives

A novel diprenylated isoflavone, 4’‐O‐methylgrynullarin (60) and other known prenylated isoflavones 61–62 with potent preferential cytotoxic activity against pancreatic cancer cell lines were isolated from Derris scandens flowers by Awale's group (Figure 3). ^[13]

Figure 3.

Natural diprenylated isoflavones with preferential cytotoxic activity against pancreatic cancer cell line.

Simpler derivatives of these compounds bearing one prenyl group in the A‐ring were synthesised for structure‐activity relationship study. ^[22] Most of the synthesised derivatives showed potent cytotoxic activity against PANC‐1 cell line under nutrient‐deprived conditions. The most active compounds were 69 a, 69 b, 74 b, 74 f and 74 h with PC₅₀ values of 1.5, 1.6, 0.8, 1.6 and 1.3 μM, respectively, against PANC‐1 cell lines in nutrient‐deprived medium (NDM). The 8‐prenyl isoflavone 78 was less active (PC₅₀=5.7 μM) than its 6‐prenyl counterpart 69 a. These results confirmed that the prenyl group at the 6 position was important for the cytotoxic activity and that the absence of the prenyl substituents in the B‐ring did not negatively impact the activity of the molecules. The initial synthesis employed the deoxybenzoin intermediates 65 a‐b for the construction of the isoflavones 66 a‐b, and Claisen rearrangement of allyl ethers 67 a‐b followed by cross‐coupling metathesis reaction for C‐6 prenylation (Scheme 14). The deoxybenzoins 65 a‐b were prepared by acylation of phloroglucinol (63) with phenyl acetyl chlorides 64 a‐b. ^[22]

Scheme 14.

Synthesis of prenylated isoflavone derivatives 69 a‐b. Reagents and Conditions: a) MeSO₃H, 60 °C, 31 % for 65 a and 65 b; b) MsCl, BF₃⋅OEt₂, DMF, 80 °C, 77 % for 66 a and 69 % for 66 b; c) i: MOMCl, NaH, THF; ii) AllylBr, NaH, THF, reflux, 88 % for 67 a and 51 % for 67 b; d) Eu(fod)₃, ClCH₂Cl, 100 °C, sealed tube, 90 % for 68 a and 83 % for 68 b; e) i: isobutene, G2, benzene, 100 °C, sealed tube; ii: conc. HCl, MeOH, 64 % for 69 a and 51 % for 69 b.

The low yields obtained from the acylation of phloroglucinol and limited phenylacetic acid starting materials prompted the group to synthesise additional derivatives by the Suzuki‐Miyaura reaction (Scheme 15). The 6‐prenylisoflavones 74 a‐h were prepared by late‐stage prenylation of 73 a‐h following the established procedure. The isoflavones 73 a‐h were in turn synthesised by the Pd‐catalysed coupling of 3‐iodochromones 71 a‐b with various phenylboronic acids followed by selective deprotection of the resulting MOM‐protected isoflavones 72 a‐h. The iodochromones 71 a‐b were obtained from acetophenones 70 a‐b (Scheme 15). ^[22]

Scheme 15.

Synthesis of prenylated isoflavone derivatives 74 a‐h. Reagents and Conditions: a) (PhCN)₂PdCl₂, dppb, Na₂CO₃, toluene/EtOH/H₂O, 70 °C, 35–98 %; b) I₂, MeOH, 72–95 %.

To evaluate the effect of the prenyl side chain at the 8 position, 8‐prenylisoflavone 78 was synthesised by conversion of MOM‐protected acetophenone 75 into 8‐prenyliodochromone 76, ^[84] followed by the Suzuki reaction and deprotection of the isoflavone 77 (Scheme 16). ^[22]

Scheme 16.

Synthesis of 8‐prenylated isoflavone derivative 77. Reagents and Conditions: a) (PhCN)₂PdCl₂, dppb, Na₂CO₃, toluene/EtOH/H₂O, 70 °C, 57 %; b) 10 % HCl, MeOH, 40 °C, 40 %.

5.3. Glabrescione B derivatives

Berardozzi and colleagues designed and synthesised Hedgehog pathway inhibitors based on the prenylated isoflavone glabrescione B (79). ^[41] Glabrescione B (79) inhibited Gli1/DNA interaction and exhibited in vitro and in vivo activity against Hedgehog‐dependent human and murine basal cell carcinoma (BCC) and medulloblastoma (MB) cells. ^[89] The glabrescione B derivatives were tested for Hedgehog pathway inhibitory activity using luciferase reporter assays. The experimental results together with computational studies led to the identification of compounds 80 and 81 as potential Gli1 inhibitors, while compounds 82 and 83 were identified as Smo antagonists (Figure 4).

Figure 4.

Glabrescione B (79) and derivatives.

The Gli1 and Smo inhibitors, together with glabrescione B (79) were also evaluated for antiproliferative activity against MB cells, individually and in combination. The combination treatments showed synergistic effects between the isoflavones analogues acting as Gli1 and Smo antagonists as well as glabrescione B (79).The active compounds and other analogues were synthesised by BF₃⋅OEt₂‐catalysed Friedel‐Crafts acylation of dimethoxyphenol 84 with differently substituted phenylacetic acids 85 a‐d to give the benzylketones 86 a‐d. Formylation and subsequent O‐cyclisation gave isoflavones 87 a‐d, which were alkylated by benzyl, prenyl and geranyl bromides to give several derivatives including 80–83 (Scheme 17).

Scheme 17.

Synthesis of glabrescione B derivatives. Reagents and Conditions: a) BF₃⋅OEt₂, 90 °C, 90 min 50–60 %; b) BF₃⋅OEt₂, (chloromethylene)dimethyliminium chloride, DMF, rt, 2 h, 28–32 %; c) alkyl‐ or benzylbromide, K₂CO₃, acetone, 45 °C, 17 h, 88–95 %.

6. Other Isoflavone Derivatives

6.1. Alkaloid‐Isoflavone Hybrids

Frasinyuk and colleagues synthesised several analogues of isoflavones hybridised with the alkaloid, cytisine.[ ³⁰ , ⁹⁰ ] These included the 7‐O cytisine‐linked derivatives, ^[30] as well as derivatives with cytisine at the 6 and 8 positions. ^[90] The 7‐O derivatives were prepared as shown in Scheme 18. Firstly, the isoflavones 89 a‐d were constructed by the deoxybenzoin route starting from substituted phenylacetic acids 88 a‐c. Alkylation of OH‐7 of isoflavones 89 a‐d with dibromoethane gave compounds 90 a‐d. The reaction of 90 a‐d with piperazine analogues gave compounds 91 c and 92 a‐c. The cytisine‐linked isoflavones 93 b‐d were prepared by reacting the isoflavones 90 b‐d with cytisine using NaI and diisopropylamine in DMF. ^[30]

Scheme 18.

Synthesis of cytisine and piperazine isoflavone analogues. Reagents and Conditions: a) K₂CO₃, BrCH₂CH₂Br (62–88 %); b) piperazine, NaI, K₂CO₃, DMF (79 % for 91c); or N‐(2‐hydroxyethyl)piperazine, NaI, K₂CO₃, DMF (60–74 %); c) cytisine, NaI, iPr₂NH, DMF (62–76 %).

The synthesised isoflavones 89 a‐d, 91 c, 92 a‐c and 93 b‐d were evaluated for antiproliferative effects on PC‐3 prostate cancer cells. Compounds 91 c, 92 b‐c, and 93 b‐d inhibited the proliferation of PC‐3 cells significantly at 10 μM. The most active compound was 93 c, which also inhibited the growth of LS174T colon cancer cells and weakly inhibited normal cell lines, BEAS‐2B and BCL‐299. To identify the cellular targets of 93 c, an almost equipotent biotinylated derivative 94 d was synthesised and subjected to a pull‐down assay using streptavidin beads and subjected to enzymatic studies. The results showed that cytisine‐linked isoflavones specifically bound to hydroxysteroid 17β‐dehydrogenase‐4 (HSD17B4) and were selective inhibitors of the enoyl CoA hydratase.

6.2. Isoflavone‐Anchored Aminopyrimidine Hybrids

Several purine and pyrazolo[3,4‐d]pyrimidine derivatives have been developed for cancer treatment and progressed to clinical trials.[ ⁴⁹ , ⁹¹ ] Some of the developed compounds include the isoflavone‐anchored kinase inhibitors, umbralisib and tenalisib.[ ⁴⁹ , ⁵¹ , ⁹¹ ] Umbralisib is a dual inhibitor of PI3Kδ and CK1ϵ.[ ⁹¹ , ⁹² , ⁹³ ] It was approved for the treatment of marginal zone lymphoma and follicular lymphoma and later withdrawn. ^[49] Tenalisib is a dual PI3 Kγ/PI3Kδ inhibitor, which is currently undergoing clinical trials.[ ⁴⁹ , ⁵¹ , ⁹⁴ , ⁹⁵ ] The syntheses of umbralisib and tenalisib have been patented and discussed in several reviews.[ ⁹³ , ⁹⁶ , ⁹⁷ ]

To discover new PI3Kδ inhibitors with improved water solubility, Lei and co‐workers synthesised racemic derivatives of pyrazolo[3,4‐d]pyrimidine and purinone attached to isoflavone or benzopyrimidinone skeletons. ^[42] The isoflavone‐linked derivatives 97, 98 and 99 were synthesised from the main precursor 100 (Scheme 19). The coupling of C‐2 substituted isoflavone 100 with 3‐iodo‐1H‐pyrazolo[3,4‐d]pyrimidin‐4‐amine (101), followed by the Suzuki‐Miyaura coupling of 102 with boronic acids 103 and 104 rendered compounds 97 and 98, respectively. Compound 99 was synthesised by N‐alkylation of intermediate 106 with bromoethylisoflavone 100. The intermediate 106 was prepared in a sequence of steps starting from 4,6‐dichloropyrimidin‐5‐amine (105) (Scheme 20). Of the synthesised isoflavone analogues, 97 and 98 showed inhibitory effects against PI3Kδ with IC₅₀ values of 163 and 176 nM, respectively. However, compound 99 lost activity (IC₅₀>1000 nM).

Scheme 19.

Synthesis of isoflavone linked pyrazolo[3,4‐d]pyrimidine derivatives. Reagents and Conditions: a) K₂CO₃, DMF, rt, 83 %; b) Pd(PPh₃)₄, Na₂CO₃, DMF, 45.3 % for 1 and 86.9 % for 2.

Scheme 20.

Synthesis of isoflavone linked purinone derivative 99. Reagents and Conditions: a) 100, K₂CO₃, DMF, 47.7 %.

The success of this study was in the discovery of 110, a pyrazolo[3,4‐d]pyrimidine analogue anchored by the 3‐phenylquinazolinone scaffold. Compound 110 was prepared by coupling of 101 with 107 and subsequent Suzuki cross‐coupling reaction of 108 with boronic acid 109 (Scheme 21). It inhibited PI3Kδ with IC₅₀ value of 72 nm and exhibited cytotoxicity against jeko‐1 cancer cell line overexpressing PI3Kδ. In addition, it showed improved solubility (15 times higher than that of TGR1202/umbralisib), and diminished cytotoxicity against normal human cell lines at high concentrations. The design of compound 110 and related compounds was inspired by the realisation that the quinazolinone and the chromone rings in idelalisib and umbralisib overlapped when the two structures were superimposed. ^[42] Indeed, 3‐phenylquinazolinone and isoflavone scaffolds possess similar frameworks, except that oxygen and carbon atoms at the 1 and 3 positions of the isoflavone nucleus are replaced by nitrogen atoms in the 3‐phenylquinazolinone unit.

Scheme 21.

Synthesis of 3‐phenylquinazolinone linked pyrazolo[3,4‐d]pyrimidine derivative 110. Reagents and Conditions: a) K₂CO₃, DMF, rt, 72.6 %; b) Pd(PPh₃)₄, Na₂CO₃, DMF, 49.7 %.

6.3. B‐Ring Modified Analogues

Several analogues have been synthesised by replacement of the isoflavone B‐ring with other heterocyclic ring structures or by pharmacophores that mimic other bioactive compounds.[ ³³ , ³⁴ , ³⁵ ] Inspired by the potential anticancer properties and synergistic effects of genistein and curcumin,[ ⁹⁸ , ⁹⁹ ] Chen and colleagues designed and synthesised analogues with a chromone core attached to a conjugated system at the 3 position. ^[33] The chromone component was inspired by isoflavone and flavone natural compounds, while the conjugated moieties with heterocyclic scaffolds mimicked curcumin. Ten analogues 115 a‐j were synthesised by aldol condensation of 3‐formylchromone (114) and imidazolylbutenones 113 a‐j (Scheme 22). The (E)‐4‐(1‐alkyl‐1H‐imidazol‐2‐yl)but‐3‐en‐2‐ones (113 a–j) were prepared by the Wittig reaction of aldehydes 112 a–j with 1‐(triphenyl‐phosphanylidene) propan‐2‐one (111) (Scheme 22).

Scheme 22.

Synthesis of genistein‐curcumin inspired hybrids. Reagents and Conditions: a) Toluene, 70 °C, 5 h, 75–99 %; b) Toluene, PTSA, 100 °C, overnight, 11–58 %.

The synthesised compounds together with curcumin, genistein and quercetin were evaluated for antiproliferative activity against prostate cancer cell lines, PC‐3, DU‐145 and LNCaP. All the synthesised derivatives inhibited the proliferation of the cancer cells at concentrations that were significantly lower than those of curcumin, genistein, and quercetin. Compounds 115 b and 115 j were the most active compounds against the PC‐3 cell line, with IC₅₀ values of 1.8±0.3 and 1.8±0.4 μM, respectively. Compound 115 b together with 115 e, 115 f and 115 j exhibited good inhibitory activity against LNCaP with IC₅₀ values of 1.0±0.2, 1.8±0.9, 1.2±0.6 and 1.3±0.2 μM, respectively. The most active compound against DU‐145 cell line was 115 i (IC₅₀=1.4±0.3 μM). ^[33]

In another example, Hirazawa and colleagues synthesised flavonoid‐based derivatives with the B‐ring substituted mainly by the benzo[b]thiophene (BT) system. ^[34] These included analogues of chalcones, isoflavones, and aurones. Of the synthesised isoflavone derivatives, compound 119 a potently inhibited the proliferation of A549, MDA‐MB‐231, MCF‐7, KB and KB‐VN (the P‐gp‐overexpressing MDR subline of KB) with IC₅₀ values of 0.64, 0.82, 0.72, 0.82 and 0.51 μM, respectively. Other active isoflavone analogues were 119 b and 119 c. The compounds exhibited antiproliferative activity with IC₅₀ values of 2.6, 1.0, 3.1, 4.2 and 0.67 μM (119 b) and 4.8, 5.2, 4.0, 4.0 and 0.84 μM against A549, MDA‐MB‐231, MCF‐7, KB and KB‐VN (119 c), respectively. The selectivity of compounds 119 b and 119 c against the resistant cell line, KB‐VN is worth noting. Compound 119 c and other active derivatives were further determined to cause cell cycle arrest and induce multipolar spindle formation in the prometaphase. ^[34]

The isoflavone derivatives were synthesised by the Suzuki‐Miyaura coupling of 3‐iodochromones with differently substituted phenyl, naphthyl and BT boronic acids. The synthesis of active isoflavones 119 a‐c was achieved from acetophenones 116 a‐c, which were converted into 3‐iodochromones 117 a‐c by condensation with DMF‐DMA, followed by iodine‐mediated cyclisation of the resulting enaminone. The Suzuki‐Miyaura coupling reaction of 3‐iodochromone 117 a‐c with BT boronic (118) rendered the isoflavone 119 a‐c (Scheme 23). ^[34]

Scheme 23.

Synthesis of the most active BT‐isoflavone analogues. Reagents and conditions: a) i: DMF−DMA, xylene, 150 °C, 99, 87, 87 % for anaminone a, b and c, respectively; ii) I₂, CHCl₃, rt, 75, 78, 91 % for 117 a, b and c, respectively; (b) Pd(PPh₃)₄, 2 M Na₂CO₃(aq), PhH, reflux, 67, 82, and 88 % for 119 a, b and c, respectively.

In 2019, Selepe's group reported the unexpected conversion of methoxybenzoylbenzofurans into isoflavone derivatives through a cascade of processes that involved demethylation and oxa‐Michael‐type cycloaddition, leading to furan ring deconstruction and chromone reconstruction. ^[71] Treatment of a 2’‐methoxybenzoylbenzofuran 120 with different demethylating reagents rendered either an isoflavone derivative 121 or the 2’‐hydroxybenzoylbenzofuran 122 or both products depending on the reaction conditions (Table 1).

Table 1.

Demethylation of 2’‐methoxybenzoylbenzofuran. ^[71]

Entry	T (°C)	Time (h)	Solvent	Demethylating Reagent	Mol Equiv	Yield (%)[a] 121 122
1	−78	6	CH2Cl2	BBr3	3	11	18
2	−78 −0	3	CH2Cl2	BBr3	3	28	33
3	0‐rt	4	CH2Cl2	BBr3	3	78	‐
4	0‐rt	18	CH2Cl2	AlCl3	6	‐	91
5	0‐rt	12	CH2Cl2	AlCl3/EtSH	3	[b]	[b]
6	80	48	DMF	TMSI[c]	8	[d]	[d]
7	80	48	DMF	TMSI[e]	6	68	4
8	80	48	DMF	NaI/TMSCl	6	56	9

[a] Isolated yields. [b] Uncharacterized products. [c] 1 M TMSI was used. [d] No reaction. [e] 98 % TMSI was used.

The substrate scope of the unexpected transformation using BBr₃ at 0 °C to rt was evaluated using differently substituted benzoylbenzofuran intermediates. This led to the correction of the structures of benzoylbenzofuran SIRT1 inhibitors to isoflavones. ^[71] A follow‐up study on antiproliferative activity and SIRT1 inhibitory activity of the synthesised benzoylbenzofurans and isoflavone analogues revealed that isoflavone analogues were potent SIRT1 inhibitors and three isoflavonequinones 123 a, 123 b, and 123 c exhibited significant SIRT1 inhibitory activity with IC₅₀ values of 5.58±0.373, 1.62±0.0720, and 7.24±0.823 μM, respectively. ^[53] The most active compound, 123 b displayed SIRT1 inhibitory activity comparable to that of suramin. The antiproliferative effects of the SIRT1 inhibitors 123 a‐c and other compounds against the MDA‐MB‐231 cell line showed that both benzoylbenzofurans and isoflavone analogous potently inhibited the proliferation of the MDA‐MB‐231 cells. ^[53]

The synthesis of the active isoflavonequinones 123 a‐c is shown in Scheme 24. ^[71] The acetophenones 124 a‐c were converted into enaminones, which were reacted with benzoquinones in acetic acid to give methoxybenzoylbenzofurans 125 a‐c. Treatment of the benzoylbenzofurans 125 a‐c with BBr₃ rendered isoflavonequinones 123 a‐c.

Scheme 24.

Synthesis of isoflavonequinone SIRT1 inhibitors. Reagents and conditions: (a) i. DMF−DMA, ii. 2,5‐dimethylbenzoquinone or 3,5‐dimethylbenzoquinone, HOAc; (b) BBr₃ (3 equiv. per methoxy), CH₂Cl₂, 0−rt, 66, 42, 11 % for 120 a‐c, respectively.

6.4. A‐ring Derivatives‐ Mannich Bases Derived C‐6 and C‐8 Analogues

Frasinyuk and colleagues synthesised an array of C‐6 and C‐8 substituted isoflavone derivatives with antiproliferative activities from aminomethylated intermediates.[ ³⁹ , ¹⁰⁰ ] These included polycyclic isoflavone analogues derived from Diels–Alder reaction of ortho‐quinone methides with dienophiles. ^[39] The synthesis commenced with regioselective aminomethylation of isoflavones 126 a‐i, 128 a and 128 g to give aminomethylated compounds 127 a‐i, 129 a and 129 g (Scheme 25). Transformation of aminomethylated isoflavones by in situ generation of ortho‐quinone methides 130 and 131 and subsequent trapping with various electron‐rich dienophiles rendered a diverse library of C‐8 and C‐6 polycyclic isoflavone derivatives 133–139 (Scheme 26). The isoflavone derivatives were evaluated for antiproliferative activity against PC‐3 cell line. Some of the most active derivatives were 135 b, 137 g and 138 a, which attenuated cell proliferation by 86, 78 and 78 %, respectively at 10 μM.

Scheme 25.

Regioselective modification of 7‐hydroxyisoflavones 126 and 7‐hydroxy‐8‐methylisoflavones 128. Reagents and conditions: a) CH₂(NMe)₂, ⁱ PrOH, 80 °C, 2–4 h, 68–91 %; b) CH₂(NMe)₂, dioxane, 100 °C, 16 h, 72–99 %.

Scheme 26.

Diels–Alder reaction of N,N‐dimethylaminoisoflavones 127 and 129 with dienophiles. Reagents and conditions: a) 2,3‐dihydrofuran, DMF, reflux, 24–40 h, 27–75 %; b) 3,4‐dihydro‐2H‐pyran, DMF, reflux, 36–40 h, 15–55 %; c) 3‐(dimethylamino)‐5,5‐dimethylcyclohex‐2‐en‐1‐one, DMF, reflux; 4 h, 68–92 %; d) 4‐cyclopent‐1‐en‐1‐yl morpholine, DMF, reflux, 4 h, 53–58 %; e) 4‐cyclohex‐1‐en‐1‐yl morpholine, DMF, reflux, 4 h, 51–91 %.

6.5. Functionalised Benzopyrans

Three generations of benzopyran derivatives with anticancer activities have been developed based on the isoflavone framework.[ ⁵⁵ , ⁵⁶ ] The compounds exhibit broad anticancer activity and inhibit the proliferation of resistant cancer cells and stem cells.[ ⁵⁴ , ¹⁰¹ , ¹⁰² ] Examples of second and third generation super‐benzopyran analogues are Me‐344 (140), Cantrixil/TRX−E‐002‐1 (141) and Trilexium/TRX−E‐009‐1 (142) (Figure 5).[ ⁵⁵ , ⁵⁶ , ¹⁰¹ ] Me‐344 (140) underwent Phase 1 clinical trials in patients with refractory solid tumours that included colorectal, non‐small cell lung, ovarian cancers and others,[ ¹⁰² , ¹⁰³ ] while TRX−E‐002‐1 (141) was investigated for resistant or recurrent ovarian, fallopian tube and primary peritoneal cancers,[ ¹⁰⁴ , ¹⁰⁵ ] TRX−E‐009‐1 (142) is a preclinical candidate and it has been determined to be tubulin polymerisation inhibitor. ^[101]

Figure 5.

Second and third generation super‐benzopyran analogues.

A recent study investigated the effects of TRX−E‐009–1 against (diffuse intrinsic pontine gliomas) DIPG neurosphere cultures as a single agent and in combination with histone deacetylase inhibitor, SAHA and radiation. TRX−E‐009‐1 exhibited tumour‐specific activity against DIPG neurosphere cultures. A triple combination treatment of TRX−E‐009‐1 with SAHA and radiation significantly enhanced survival in DIPG models. ^[106]

The synthesis of the benzopyran analogues was reported by Heaton and colleagues in 2015. ^[54] Using TRX−E‐002–1 and TRX−E‐009‐1 as examples, the synthesis was initiated by acylation of 2‐methylresorcinol (143) with benzoic acids 144 a‐b to give benzophenones 145 a‐b (Scheme 27). Condensation of benzophenones 145 a‐b with phenylacetic acid 146 rendered 3,4‐diphenylcoumarins 147 a‐b, which underwent successive reduction to give 141 and 142 as racemic mixtures. The requisite enantiomers were obtained by chiral resolution.

Scheme 27.

Synthesis of third generation super‐benzopyrans. Reagents and conditions: a) ZnCl₂, POCl₃, 70 °C, 2 h, 87 % (145 a) and 41 %(145 b); b) DiPEA, Ac₂O, 135 °C, 18 h, 76 % (147 a) and 68 % (147 b); c) THF, BH₃⋅Me₂S in THF, 35 °C, 18 h, 53 % from 147 a and 53 % from 147 b; d) H₂, Pd/C, EtOH, 3 bar, 40 °C, 18 h, 88 % (141) and 60 % (142).

7. Conclusions

Several isoflavone derivatives have been synthesised and evaluated for their potential anticancer activities. This has led to the discovery of analogues with improved in vitro and/or in vivo antiproliferative activities, selectivity, specificity and physicochemical properties. Further studies revealed that the isoflavone derivatives induce cell cycle arrest and apoptosis and act as mitotic, kinases, sirtuins and angiogenesis inhibitors. Moreover, some of the analogues were determined to block signalling pathways involved in cancer initiation and progression.

Strategies that were employed for optimisation included molecular hybridisation, CADD, DOS and functional group manipulation. Molecular hybridisation was employed in several instances and its combination with other techniques such CADD was determined to be beneficial for the development of compounds with high target binding affinity and physicochemical properties. An example includes the discovery of the formononetin derivative 22, which potently inhibited EGFR and also suppressed MDA‐MB‐231 tumour growth in vivo. ^[40] In another example, optimisation of isoflavone‐pyrazolo[3,4‐d]pyrimidine hybrids through scaffold hopping approach led to the discovery of a 3‐phenylquinazolinone‐pyrazolo[3,4‐d]pyrimidine hybrid 110 with enhanced affinity for PI3Kδ than other derivatives, low toxicity and 15 times improved solubility than umbralisib (96). ^[42] On the other hand, hybrids that involved replacement of the phenyl B‐ring with electron rich heterocyclic ring resulted in improved activity and selectivity. ^[34] For instance, the BT‐isoflavone analogues, showed improved activity compared to 3‐phenylchromones and two analogues, 119 b and 119 c displayed selectivity for the P‐gp‐overexpressing MDR subline of KB (KB‐VN). ^[34]

Derivatisation of the isoflavones by functional group manipulation also yielded more active derivatives, improved specificity and facilitated chemical accessibility of some of the complex isoflavone compounds. An Intriguing observation was in the development of GVA derivatives, whereby the 6‐O‐benzyl derivative 31 was discovered to be an α,β‐tubulin inhibitor, ^[47] while the 7‐O‐benzyl derivative, gatastatin (32) and its analogue gatastatin G2 (33) were determined to be specific γ‐tubulin inhibitors.[ ⁷⁴ , ⁷⁵ ] This showed that the position of the benzyl group influenced the specificity of the GVA derivatives. Other analogues that demonstrated specificity included glabrescione B derivatives. ^[41] Compounds 80 and 81 were identified as potential Gli1 inhibitors, while compounds 82 and 83 were identified as Smo antagonists. ^[41]

These studies affirm the importance of the isoflavone framework in the discovery and development of novel anticancer agents. Future studies could leverage a combination of optimisation techniques to facilitate efficient discovery of more active analogues with improved properties.

Conflict of Interests

The author declares no conflicts of interest.

Biographical Information

Mamoalosi Selepe completed a Bachelor of Chemical Technology from the National University of Lesotho. She received BSc (honours) and PhD degrees from the University of KwaZulu‐Natal (South Africa). Upon completion of PhD study, she worked for iThemba Pharmaceuticals as a scientist. Thereafter, she was appointed to the position of Chief Researcher (NMR) at the University of Pretoria. Her research interests are on isolation of compounds from natural sources and synthesis of natural compounds and derivatives.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Selepe M. A., ChemMedChem 2024, 19, e202400420. 10.1002/cmdc.202400420