Quinones and hydroquinones can be found in the cells of all respiring organisms. Some familiar quinones are phylloquinone (1) and plastoquinone (2), both quinones needed in photosynthesis, and ubiquinone (3), also known as coenzyme Q10, which participates in the aerobic cellular respiration (Figure 1). In organisms in general, quinones can have two major biologically relevant chemical properties. One is that they can be reactive oxygen generator. Quinones generate reactive oxygen species (ROS), such as superoxide ( O 2 · − ) and subsequently hydrogen peroxide (H₂O₂). Because of this, they can induce a variety of hazardous effects in vivo [1] [2] . The second property is their ability to add nucleophiles, including nucleophilic biomolecules. This can lead to cellular damage through alkylation of crucial cellular proteins and/or DNA [2] , the first of which

explains some of the irreversible binding of estrogens to proteins. Nevertheless, many natural and synthetic compounds that possess quinoid structures have antitumour and antibiotic activity. Typical examples are mitomycin C (4) isolated from Streptomyces caespitosus [3] and Streptomyces lavendulae [4] , avarone (5) found in the marine sponges Pleraplysilla spinifera [5] and Dysidea avara [6] and the synthetic mitoxantrone (6) (Figure 2). That the anti-tumour activity of quinones such as these can be the cause of mechanisms other than the two basic chemical reactive properties of quinones named above can be seen in the case of mytomycin C (MMC). MMC is used in the treatment of gastro-intestinal, anal and breast cancers [7] [8] . MMC itself is inert to nucleophiles, but is reduced to a bisphenol which itself is a very reactive bis-electrophile [9] . At that point, alkylation of DNA becomes the most favored mechanism of action as does the inactivation of thioredoxin reductase (TrxR) by addition of the reduced MMC to the peptide. In both processes, the nucleophiles are not directly added to the quinone moiety. Other modes of action, such as redox cycling are also in evidence. Avarone (5) is a cytostatic agent which has potent antileukemic activity both in vitro and in vivo (mice) [10] . Also, it displays antibacterial and antifungal activities. While mitoxantrone (6) is used for the treatment of secondary progressive, progressive relapsing, or worsening relapsing-remitting multiple sclerosis, it is also used to treat advanced prostate cancer and acute non-lymphocytic leukemia. Menadione (2-methylnaphtho-1,4-quinone, 7), which is sold as a nutritional supplement as a vitamin K mimic, has also been viewed as a potential drug for prostate cancer treatment [11] . In 1991, quinones constituted the second largest group of cytotoxins used in chemical cancer therapy, after specifically alkylating agents such as mustards [12] , with about 1500 quinones already tested in 1974 [13] .

Triterpenoidal quinones, secosteroids with quinoid subunits and steroidal quinones of natural origin are less commonly isolated structures. Nevertheless, catechol estrogen quinones, 4 and 5, present in mammals (Figure 3), are implicated as initiators in the development of breast and other human cancers. On the other hand, there has been a significant effort in synthesizing steroidal and triterpenoidal quinones as potential cancer-active substances. These molecules are within the scope of this review. However, triterpenoid quinonemethides, which

merit a review of their own (for a previous review, see: [14] ), are not part of this review. This includes those structures which are composed of a hydroxyl-quinonemethide system such as celasterol [15] , although in certain cases they can be regarded as the enol form of ortho quinoids. Steroidal and triterpenoidal anthraquinones (anthracenediones) and naphthoquinones, which some readers may not see as true quinones, have been included in the review.

Marine sponges have been noted as sources of rare terpenoids of significant complexity, including of terpene quinoids [16] . Two secosteroids with a trans-fused quinoid structure, 10 and 11, were found in a Korean marine sponge Ircinia sp. [17] [18] (Figure 4). 11 was found to be active against the bacteria Micrococces lutes (3.1 μg/L), Staphylococcus epidermidis (MIC 25 μg/L) and Bacillus subtilis (MIC 25 μg/L).

Aplysiasecosterols B and C, 12 and 13, are two 9,11-secosteroids with a cis-fused 1,4-quinone unit. They have been isolated from the sea hare Aplysia kurodai [19] (Figure 5). Aplysiasecosterol B was thought to be the biosynthetic precursor of aplysiasecosterol A (14), also isolated from Aplysia kurodai [20] and subsequently synthesized [21] [22] [23] . A possible biosynthetic pathway for 14 was suggested to have cholest-7-en-3S,5R,6R-triol as precursor due to the structural similarity between both the cyclopentane ring and the side-chain of aplysiasecosterol A (14) and those of known 9,11-secosteroids [20] [24] .

However, while aplysiasectosterol A (14) exhibits an appreciable cytotoxicity against human myelomonocytic leukemia cell line HL-60 (IC₅₀ = 16 μM), secosterols B (12) and C (13) were found not to exhibit any appreciable cytotoxicity, showing that the tricyclic structure of 14 is important for its cytoxicity [16] . The stereochemistry of secosteroidal quinones 12 and 13 was established by extensive ROESY spectroscopy as well as wet-chemical derivatization.

In 2023, also the total synthesis of aplysiasecosterol B (12) was published [23] . In the synthesis, the building blocks 15 and 16 were joined by Suzuki-Miyaura C-C-cross-coupling reaction at a late stage of the sequence [23] . This constitutes a de novo synthesis, differing from other synthetic strategies which commence with a steroidal structure as the starting material [24] . 15 was prepared from (+)-Wieland-Miescher diketone monoacetal (17) [25] [26] , which was reduced via its dienol silyl ether. Reacting 18 with oxone^® in the presence of Shi’s reagent [27] gave epoxide 19 in good diastereoselectivity. Then, epoxide 19 was hydrolyzed to triol 20, with a careful hydrolytic deacetalization giving 21. Protection of both the 1,2-diol moiety at C4/C4a and the hydroxyl group at C6 in 21, as acetonide and as silyl ether, respectively, led to 22. Ito-Saegusa oxidation [28] of 22 gave an enone, which was iodinated to 15 as the first of the coupling partners in the subsequent Suzuki cross coupling reaction (Scheme 1).

For the other coupling partner, T. Ohyoshi et al. [22] started out with a reduction of carbonyl group (C1) of the (−)-Hajos Parrish diketone (23) [29] . The resulting alcohol was protected as a silyl ether. The resulting product was oxidized at C6 with Davis reagent [30] to provide the hydroxyketone 24. Oxidative cleavage of the enone moiety in 24 with Pb(OAc)₄ to a carbaldehyde, which was then reduced with NaBH₄ to alcohol 25. Crabtree’s reagent [31] made possible the

stereoselective reduction of the ene-ester moiety in 25 to provide 26. The primary alcohol in 26 was protected as MOM ether, the ester group was reduced with DIBAH and the resulting alcohol was converted to a tosylate, which allowed the introduction of a homoallyl-group in 27 in a nucleophilic substitution reaction with allylmagnesium bromide as reagent. Cross metathesis with 2-methyl-2-butene gave prenylated compound 28. This was followed by Sharpless dihydroxylation of the double bond in 28, where the α,β-dihydroxy motif was protected as acetonide 29. Deprotection of the siloxy group in 28 with tetrabutylammonium fluoride (TBAF) was followed oxidation of the ensuing alcohol with tetrapropylammonium perruthenate/N-methylmorpholine N-oxide (TPAP/NMO). Enol triflation to 29 and borylation finally provided the second coupling partner 16 (Scheme 2). The coupling of 15 and 16 itself was affected using Pd(dba)₂, Ph₃As, and Ag₂O in 78% yield [22] . Thereafter, the MOM-protective group as well as the acetonide in 27 were hydrolyzed to give 32. The allylic hydroxy group in 32 was oxidized MnO₂ to the quinoid structure in 33. Reduction of the 14,15-ene in 33 with Crabtree’s catalyst, and subsequent removal of the acetonide protective group in 34 gave the desired aplysiasecosterol B (12) in an overall yield of 9.2% [22] (Scheme 3).

The Formosan gorgonian coral Pinnigorgia sp. (Gorgoniidae) yielded seven new secosteroids pinnisterols D-J, all with a trans-fused quinone system, along with the known secosteroid 11 [32] (Figure 6).

Pinnisterol D (35a) inhibits cell viability in the hepatic stellate cell line HSC-T6 (IC₅₀ 3.93 μM, i.e. at 10 μM of 35a, viability of HSC-T6 decreases to 16.8%). At the same concentrations, pinnasterol F (35c) and pinnasterol H (35e) decrease the viability of HSC-T6 cells to 56.9% and 37.1%, respectively. In anti-inflammatory testing, pinnisterol E (35b, IC₅₀ 2.33 μM), pinnisterol H (35e, IC₅₀ 2.59 μM), and pinnisterol J (35g, IC₅₀ 3.89 μM) reduce elastase enzyme release. Pinnisterol F (35c, IC₅₀ 5.52 μM), pinnisterol H (35e, IC₅₀ 3.26 μM), and pinnisterol J (31g, IC₅₀ 3.71 μM) lower the production of superoxide anions from human neutrophils [32] .

Miao et al. [33] developed a multi-gram synthesis of pinnesterol E (35b) from 7(11)-dehydroergosterol 36 (Scheme 4), which itself is obtained easily from ergosterol in 1 step [33] . The O-silylated 7-dehydroergosterol 36-OTBDPS undergoes a cycloaddition with singlet oxygen to give endoperoxide 37, which subsequently is cleaved reductively to give 5α,8α-dihydroxysterol 38. Secosteroid 39 is obtained upon subjecting 38 to ozonolysis with reductive work-up. Burgess reagent as a mild dehydration reagent of the 5a,8a-dihydroxysteroid gives cyclohexadienone 40. Lastly, 40 is transformed by epoxidation and oxidative ring opening to pinnisterol E (35b). An X-ray single crystal structural analysis of 35b was carried out [33] . Miao et al. have also devised a synthesis to 6-ketoaplidiasterol B (11-O-acetylpinnesterol H, 11-Ac) (see above) [17] from cholesterol [33] , again involving the creation in steroidal ring B of a cyclohexadiene-structure and an unsaturation at 9(11) by dehydrogenation using Hg(OAc)₂ [34] , a cycloaddition of singlet oxygen across the diene system 36, the reductive ring opening of the formed endo-peroxide 37 with concomitant reduction of the 6(7)-ene system. Subsequently, cleavage of the 9(11)-ene in 44 by ozonolysis with reductive work-up, dehydration of the 5a,8a-dihydroxysteroid 45 to cyclohexadienone 45, epoxidation of the 5(6)-ene system with oxidative ring opening led after deprotection of 11-Ac to 6-keto-aplidiasterol (11) [33] (Scheme 5).

From the endophytic fungus Talaromyces sp. HYZX-1, isolated from the healthy leaves of the marine mangrove Kandelia obovate, cyclosecosterol 47 (Figure 7) was obtained [35] . An X-ray single crystal structural determination was carried out. Structurally, the compound is the first reported 9,11-secosterol with a novel lactone ring structure fused with the D ring of a steroid. 47 was evaluated for its inhibitory activity against acetylcholinesterase (AChE) in vitro, but showed only moderate activity with an IC₅₀ of 46 μM in comparison with tacrine A (IC₅₀ 0.4 μM) [35] .

Mycoleptodiscin A (48) and its 9α-hydroxylated derivative mycoleptodiscin B (49) are two indolosesquiterpenoids isolated from liquid cultures of the endophytic fungus Mycoleptodiscus sp. [36] isolated from the flowering plant Desmotes incomparabilis in Panama (Figure 8). In a cytotoxicity assay with cancer cell lines, mycoleptodiscin B (49) was found to have an IC₅₀ of 0.660 μM against H460, 0.780 μM against A2058, 0.630 μM against H522-T1, 0.600 μM against PC-3, and 0.41 μM against IMR-90 cell line [36] . Significant synthetic efforts have been devoted to the synthesis of the structures [37] [38] [39] .

Nagaraju et al. [37] started their synthesis of mycoleptodiscin A (48) with the addition of lithiated veratrole, prepared from bromoveratrole (51) and n-BuLi to the lactone unit of the sesquiterpene (+)-sclareolide (50), a natural product found in different plants, including in clary sage (Salvia sclarea). 53 was produced (Scheme 6). The authors tried to cyclize 52 directly to 55 by intramolecular electrophilic substitution, but the reaction did not proceed, irrespective of the Lewis and protonic acid (SnCl₄, BF₃^.Et₂O, TFA, MeSO₃H, or AlCl₃) used [37] . This necessitated prior reduction of the benzylic keto group in 52, which was achieved with Pd/C, H₂ in conc. HCl/PEG-400. It was possible to cyclize 53 to 54 via electrophilic substitution reaction at low temperature (−78˚C) using SnCl₄ as Lewis acid, taking advantage of the more electron-rich aromatic ring. To the

cyclized product 54 the benzylic keto group was reintroduced. Thereafter, 55 was subjected to electrophilic nitration, where interestingly the nitro group could be introduced into the aromatic ring ortho to the keto group. The keto group in 55 was reduced with NaBH₄, and the ensuing alcohol 56 was subjected to a nucleophilic substitution reaction, utilizing Me₃SiCN, giving cyanide 57 with inversion of configuration. X-ray single crystal structures of both compounds, 56 and 57, were carried out. The synthesis of 52 from 56 had an overall yield of 60%. Reaction of 57 with BBr₃ led via mono-demethylation to 58. 58 could then be ring-closed reductively with HCO₂NH₄ in the presence of Pd/C. Spontaneous aerobic oxidation of the ring-closed product gave the desired indolo-ortho-quinone 48 (Scheme 7). 59 is a further interesting intermediate, which leads to substituted mycoleptodiscin A analogs and can be prepared from 52 in 62%. This route lends itself to scale-up of the synthesis as the purification of 59 involves only one recrystallization. Specifically the two routes from 52 to 56 and from 52 to 59 differ only in the last step of the five step sequence where the nitro group is reduced with H₂ in the presence of Pd/C in toluene [37] . The amino group of 59 can be acylated such as benzoylated to give 60. McMurry type intramolecular cyclization [40] utilizing low valent titanium (TiCl₄, Zn) leads to dimethoxybenzoindole 61. Bis-O-demethylation of 61 with BBr₃ at −78˚C yielded a dihydroxyderivative, which spontaneously oxidized under aerobic conditions to 62, which can be seen as a mycoleptodiscin A analog [37] (Scheme 8).

D. Dethe et al. [39] based their synthesis of (−)-mycoleptodiscin A (48) on a Lewis acid catalyzed cyclization of 70b, prepared by the alkylation of 7-methoxyindole (69) with alcohol 68 (Scheme 9). Alcohol 68 itself was synthesized from Wieland-Miescher diketone derivative 63. 63 was acetalized selectively to 64. This was followed by a reductive methylation utilizing iodomethane in lithium-liquid ammonia. Thereafter, the remaining keto group in the resulting trans-decalone was subjected to a Wolff-Kishner reduction, followed by deacetalization to 65. Methylation α to keto group of 65 with subsequent epimerization of the newly formed stereocentre led to 66. 66 was transformed to α,β-unsaturated aldehyde 67 using dichloromethyllithium, generated in situ from CH₂Cl₂ and LDA at −100˚C, followed by treatment with HMPA, LiClO₄, and CaCO₃ at 140˚C [41] . Finally, the aldehyde was reduced to alcohol 68. The alkylation of 7-methoxyindole (69) with alcohol 68 was achieved with BF₃∙Et₂O as Lewis acid at rt. Thereafter, the nitrogen of the indole unit was sulfonated in order for the N-heterocycle of the indole unit not to compete in the subsequent electrophilic ring closure reaction. Cyclization of 70 by intramolecular electrophilic substitution was affected with TMSOTf. N-desulfonation of 71a was carried out with Na/Hg in MeOH. Subsequent O-demethylation yielded a dihydroxyindole derivative which was immediately treated with Fremy’s salt [42] to yield desired 48.

The first total asymmetric synthesis of mycoleptidsicin A, published in 2015, was achieved by Zhou et al. [38] (Scheme 10). The synthesis started with farnesyl acetate (72), which was transformed in 3 steps to 73 [43] . Hydroboration and oxidative cleavage with H₂O₂ gave 74. Dess-Martin oxidation converted alcohol 74 to the corresponding carbaldehyde which was subjected to addition with vinylmagnesium bromide to give a secondary allylic alcohol which was protected with a silyl function. Deacetylation provided 75 which, converted to the mesylate, underwent a nucleophilic bromination. The bromide was reacted with the anion of sulfone 76. The adduct was desulfonated using Na/Hg, and the silyl protective group was removed utilizing HF^.py (77a → 77b). The tetracyclic system 77b was prepared by an enantioselective polyene cyclization catalyzed by using an iridium catalyst prepared in situ from [Ir(cod)(Cl)₂] and chiral ligand 78. Desired cyclization product 79 was isolated in 21%, initially, however, separate subsequent treatment of mixture of side products with BF₃∙Et₂O increased the yield to overall 71%. 79 was subjected to a one-pot dihydroxylation/cleavage with the resultant aldehyde methylated at the α-position, utilizing KOBu^t and CH₃I. The aldehyde function in 80 was converted to a methyl group by Wolff-Kishner-Huang reduction. Next, a keto group was introduced at the benzylic position, subsequently reduced, Finally, the O,O-dimethyl catechol unit was converted into the ortho-quinone structure of 48 by BBr₃ driven O-demethylation and aerobic oxidation of the ensuing catechol in the presence of Mg/NH₄Cl [38] .

Hydroxylation of estrone (87) and estradiol (86) leads to catechol estrogens. These include 2-hydroxyestrone (89) as the main metabolite found in human urine and serum, 2-hydroxyestradiol (90), and 2-hydroxyestriol (92), along with the less abundant, but potently estrogenic 4-hydroxyestrone (88) (Scheme 11). These catechol estrogens are formed from estradiol (86) and estrone (87) by cytochrome P450 enzymes such as CYP1A2 and CYP3A4, predominantly in the liver but also in extrahepatic tissues [44] . 2-Hydroxyestradiol (90) has been detected in the breast, uterus, kidney, rain and pituitary gland. There is significant production of especially 2-hydroxyestradiol (90) in the body, but it is converted

rapidly to downstream products by O-methylation, glucuronidation and sulfonation, so that at every one time, 2-hydroxyestradiol (90) can only be found at low concentrations in the body. Also, steroidal catechols can be transformed further to estrogen quinones by oxidative enzymes such as by phenol oxidase [45] , prostaglandin H synthetase [46] , and again by cytochrome P-450 oxidase [47] . In the latter, cytochrome P450 oxidoreductase (NADPH-POR) oxidizes the catechol estrogens via one-electron transfer steps to the semiquinone (SQ, 94 and 95) and quinone metabolites (Q) leading to a redox-cycling which results in the production of reactive oxygen species (ROS) [48] (see below, Scheme 13). In the lab, 3,4-estrone ortho-quinone (estra-1,5(10)-diene-3,4,17-trione, 93) can easily be prepared by oxidation of 4-hydroxyestrone (88) with MnO₂ in CHCl₃ at rt [49] and in AcCN at 0˚C [50] (Scheme 12).

Estrogen quinones are seen as cardiovascular protecting agents. However, the compounds also have the potential to be cytotoxic and genotoxic. They can undergo one electron redox cycling, resulting in the formation of semiquinone, superoxide anion, and hydroxyl radical (Scheme 13). Furthermore, they are Michael acceptors and may alkylate cellular nucleophiles as well as macromolecules [51] . The reactivity of steroidal, estradiol derived ortho-quinones such as 8 and 9 has been studied extensively in vitro. There are a number of Michael acceptor positions in the quinones due to the formation of tautomers (Scheme 14), that leads to a plethora of 1,4- and 1,6-addition reactions [50] . Typical examples are the reaction with n-propylamine (100) [51] (Scheme 15), cysteine (108a) [52] , glutathione (108c) [52] (Scheme 16), and with deoxyribonucleosides such as 113 and 115 [53] (Scheme 17). Here, a number of adducts are formed, albeit in poor yield [51] (Schemes 15-17). The adducts were characterized by ¹H-NMR and mass spectrometry (FAB MS/MS). Both quinones 93 and 103 did not react with deoxycytidine or thymidine [53] , 93 did also not react with deoxyadenosine [53] . The Michael addition reaction of the estrogen quinones with amines and S-containing amino acids can also be studied in vivo where the quinones also form covalent bonds with cysteine (and in part histidine/lysine) residues in neuroglobin and serum proteins [54] [55] [56]

In vivo, it has been noted that 3,4-estrone ortho-quinone (3,4-EQ) is capable of inducing specific DNA damage in a human breast cancer cell line. Elevation of 3,4-estra-3,17β-diol quinone (3,4-E2Q) to 2,3-estra-3,17β-diol quinone (2,3-E2Q) ratio is thought to be an important indicator of estrogen-induced carcinogenesis [57] . Above, it was stated that 2,3-EQ, 3,4-EQ and derivatives can form Michael adducts with nucleophilic residues of neuroglobin and serum proteins. These adducts can be used as biomarkers. Thus, it was found in a study to evaluate the treatment-related effects on breast cancer survivors that the mean levels of 2,3-E2Q adducts with albumin (2,3-E2Q-4-S-Alb) and 3,4-E2Q-2-S-Alb adducts in the serum of 5-year survivors were by 60% - 70% lower, when compared to those in the breast cancer patients with less than one year of diagnosis/preoperative treatment. This finding adds support to the theme that hormonal therapy including aromatase inhibitors and tamoxifen^® may dramatically reduce the burden of estrogen quinones [58] . The authors hypothesized that a combination of treatment-related effects and environmental factors may modulate estrogen homeostasis and diminish the production of estrogen quinones in breast cancer patients [58] .

When steroidal 1,4-quinone is reacted with osmium tetroxide (OsO₄) in the presence of N-morpholine oxide (NMO), dihydroxylated steroid 136 is formed (Scheme 22), albeit in low yield [67] . The stereochemistry of the hydroxyl groups at C2 and C3 could not be established unequivocally. The compound was tested against a number of cancer cell lines, but was found to be largely inactive [67] .

An interesting approach to steroid-like quinone derivative 141 was presented by Chenera et al., where ring A was joined to rings C/D by a Diels-Alder reaction of p-benzoquinone (137) and 4-ethenyl tetrahydroindanone 138. Subsequent keto-enol tautomerisation led to hydroquinone 140, which was converted by oxidative dehydrogenation to 141 [68] (Scheme 23).

Kaliappan and Ravikumar [69] used the Diels-Alder strategy discussed above to prepare the sugar-oxasteroid-quinone hybrid 142 (Figure 10).

De Riccardis et al. [70] have prepared steroid-anthraquinone hybrids 151 - 154 using the A + CD → ABCD Diels-Alder approach [71] [72] for the construction of the cholestane framework, an approach that was forwarded previously by Akhrem and Titov [73] and Chenera et al. [68] (Scheme 24). For this naphthoquinones 143 - 145 were reacted with dienes 146a/b in toluene at 100˚C. The cycloaddition with 5-hydroxynaphthoquinone (juglone, 144) delivers two regioisomeric pairs (148a-b/149a-b). An X-ray structure of 147a was carried out to ascertain the stereochemistry of the product. Oxidative dehydrogenation of 147a/b - 150a/b to 151a/b - 154 a/b was accomplished by air oxygen in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) [70] (Scheme 24).

Compounds 151a/b - 154a/b were tested against the 4 cancer cell lines J774 (murine monocyte/macrophage), GM7373 (bovine aortic endothelial), IGR-1 (human melanoma), and P388 (murine leukemia), using doxorubicine as comparison. After 48 h, 153b exhibited an IC₅₀ of 10.1±1.1 µg/mL, 154b had an IC₅₀ of 10.9 ± 1.3 µg/mL (doxorubicine IC₅₀ = 54.2 ± 7.1 µg/mL) against cell line J774. After 48 h, 154b exhibited an IC₅₀ of 16.3 ± 2.1 µg/mL (doxorubicine IC₅₀ = 17.6 ± 3.5 µg/mL) against cell line GM7373 [67] . 154b showed an IC₅₀ of 10.9 ± 2.3 µg/mL against cell line IGR-1 (doxorubicine IC₅₀ = 11.5 ± 1.4 µg/mL) after 48 h, and 153b had an IC₅₀ of 43.4 ± 3.4 µg/mL against cell line P388 after 24h (doxorubicine IC₅₀ = 52.3 ± 7.3 µg/mL) [70] . With 155 and 156, Kaliappan and Ravikumar [69] [71] prepared sugar-oxasteroid-quinone analogs of the steroid-anthraquinone hybrids discussed above (Figure 11).

De Riccardis et al. [74] synthesized estrarubicin (163-OH) (Scheme 25) as a representative of a novel class of estrogen-anthacenedione hybrids by fusing the anti-tumour pharmacore dihydroxyanthracenedione at C16α,17α positions to an estrane framework with the idea of using the estrogen as a vector for the cytotoxic moiety. First, steroidal diene 157 was reacted in a Diels Alder reaction with the epoxytetrone 158. In refluxing toluene, cycloadduct 159 was obtained in moderate yield (24% - 42%). It is interesting that running the reaction in diethyl ether in presence of LiClO₄ [75] gives a diastereomeric cycloadduct of 159 in quant. yield, where the epoxy function positioned at the β-face. Upon reacting with 10 eq. Zn in AcOH, 159 and its diastereoisomer are deoxygenated and are transformed to dihydro-dihydroxyanthraquinone hybrid 160, which is oxidized to tetrone 161 with Pb(OAc)₄ in AcOH. Tautomerisation under basic conditions leads to dihydroxyanthaquinone hybrid 162. 162 is converted stereoselectively to α-epoxide 163-OBn, which is debenzylated (H₂, Pd/C) to estrarubicin (163-OH) in 52% (Scheme 25) [74] . The stereochemistry of products was determined by extensive NMR spectroscopy, including by ROESY experiments [74] .

Also G. Ribeiro Morais et al. annelated a quinoid moiety at C16-C17 of an estrane structure through a Diels Alder reaction [76] [77] . For this, steroidal dienes 164 and 166 were reacted with quinones such as p-naphthoquinone (143). Exposure of 167 to K₂CO₃ in CH₃OH/H₂O gave in addition to the deacetylated products methylanthraquinoestrane derivative 168 in 22% [76] [77] (Scheme 26).

A number of papers have dealt with the tethering of quinone moieties to triterpenoids and steroids [78] . Thus, recently, with compounds 169 - 177 (Scheme 27 and Figure 12), Li et al. [79] have reported on chloro-substituted 1,4-naphthoquinones and chloro-substituted quinoline 5,8-diones linked through an ether bridge to C3 of the sapogenin diosgenin which is abundantly found in tubers of the family Dioscoreaceae (D. villosa, D. mexicana, D. composita and D. tokoro) [80] . The substances were evaluated against three cancer cell lines (MCF-7, HepG2 and HeLa) [79] . The cytotoxicity of the compounds was found to be related to the quinone unit. Hybrid 176 showed an IC₅₀ of 1.24 µM against HepG2 cells, which is 35-fold the toxicity that diosgenin itself exerts [79] . It is stated that hybrid 176 activates the mitochondrial apoptosis pathway in HepG2 cells [79] .

1,4-Naphthoquinone and quinoline-5,8-dione were also tethered to betulin, a triterpenoid isolated from the bark of the birch tree, at C3-OH, C-28-OH, and directly to C30 [81] to give the betulin-quinone hybrids 179 - 186, shown in Scheme 28. The molecules were screened against 7 cell lines—brain tumor cell line SNB-19, colon cancer cell line Colo-829, melanoma cell line C-32, estrogen

positive breast cancer cell line MCF-7, breast cancer cell line T47D, triple negative breast cancer cell line MDA-MB-231, and lung cancer cell line A549. The compounds showed appreciable anticancer activity (Table 1). The cytotoxic activity of the botulin-quinone hybrids increased with the NADPH-quinone reductase protein (NQO1) protein level in the cancer cell line. The authors suggest a mitochondrial apoptosis pathway in A549 and MCF-7 cells. Docking studies with the human NADPH-quinone reductase protein showed an interaction of the botulin-quinone hybrids with the active center of the enzyme, which depended on the type of 1,4-quinone moiety [81] .

Kadela-Tomanek et al. have also looked at the cytotoxicity of betuline-quinone hybrids 187 carrying a 1H-1,2,3-triazolecarboxylate unit in the linker (Figure 13) [82] [83] . These compounds were screened against the 8 cancer cell lines Colo-829, MDA-MB-231, T47D, MCF-7, A549, human adenocarcinoma cell line Caco-2, human ovarian cancer cell line SK-OV-3 and human foreskin fibroblast cells HFF-1 [83] . Again, it was seen that the betulin-quinone hybrids interacted with the NQO1 protein. Table 2 shows the most promising anticancer activity by compounds 187 as reported by Kadela-Tomanek et al. [83] .

Quinone structures can link easily to steroids through an ester linkage. Typical examples are shown in Scheme 29, where 2,5-dimethoxybenzoic acid (188) is linked to cholesterol (189). Thereafter, the dimethoxyphenyl group is oxidized with cerium ammonium nitrate (CAN) to give the corresponding p-quinone 191. Similarly, a 3-hydroxy-4-methoxyphenyl group as in 192 can be oxidized to a methoxy-substituted p-quinone such as 193 (Scheme 29) [84] [85] .

Quinones can also be joined to an estrane structure. Here, the estrane such as estradiol can be used to find estrogen receptor positive cancer cells that would then be exposed to the cytotoxic effect of the quinoid unit. In addition, depending on the structure of the estradiol derivative used, the steroid can act as an antiestrogen and or as a selective estrogen receptor modulator and would help to suppress the proliferation of the cancer cells.

In 2011, Fujiwara et al. published a reaction methodology coupling a quinone directly to an arylboronic acid. This methodology was used to prepare estrone-quinone hybrid 196 (Scheme 30) [86] . Estroneboronic acid 194 can be synthesized from estrone 3-O-triflate and pinacolborane in the presence of a palladium catalyst to give a boronate ester which is hydrolyzed subsequently with NaIO₄ in NH₄Ac-water [87] .

Estradiol doxorubicin conjugate 197 (Figure 14) was evaluated for selective uptake and cytotoxicity in MCF-7 and triple negative MDA-MB-231 breast cancer cell lines. The results showed that the anti-estrogenic component in the hybrid

compound was critical for selectivity and cytotoxicity in MCF-7 cells, where the hybrid was ~70-fold more potent than doxorubicin in inhibition of cell proliferation and promoting cell death [88] .

Finally, in Danishefsky’s lab, Kuduk et al. [89] have developed estradiol-geldanamycin conjugates. Geldanamycin is an ansamycin antibiotic, which was first isolated from Streptomyces hygroscopicus. It binds to the Hsp90 chaperone protein and causes the degradation of several important signaling proteins, to the human epidermal growth factor receptor 2 (HER2) and the estrogen receptor (ER). With binding geldanamycin to estradiol, it was hoped that estradiol could be used as a delivery system for geldanamycin to the estrogen receptor (ER) [90] and help degrade the receptor in estrogen receptor positive cancer cells. It was found that conjugates 198a/b (Figure 15) are active and more selective than geldanamycin itself, causing degradation of ER and HER2, but not of other geldanamycin targets.

Over the last three decades a larger number of quinoids have been isolated as natural products. Especially worth mentioning are secosteroids with a cis-fused quinone unit. Many of the quinones exhibit cytotoxicity. Here, the catechol estrogens as endogenous metabolites of estrone and estradiol need to be named, which can undergo oxidation to quinones and semiquinones. These quinones and especially the semiquinones have been thought to contribute to estrogen-induced carcinogenesis. The mechanisms by which quinones exert their toxic effects are complex. Nevertheless, two processes appear to be important: the reaction of N- and S-nucleophiles with the quinoid structure, and the generation of active oxygen species via redox cycling. Quite a few quinones have been found to be mutagenic. Nevertheless, quinones are also effective anticancer agents. This has led to the development of synthetic quinones with a steroidal framework and of steroid-quinone and triterpene-quinone hybrids. In some cases the steroidal or triterpenoidal structure is used as a drug-delivery system, and in other cases the terpenoid structure itself exerts added cytotoxicity. Where estradiol-based steroid-quinone is used on estrogen-positive cancer cells, the steroidal moiety can act as an estrogen receptor inhibitor or a selective estrogen receptor modulator. Further research in the development of steroid quinone and triterpenoid quinone structures as anti-tumor agents is needed in order to find compounds with a higher specificity, to distinguish between cancerous and normal cells.

Support by grant UAEU-G2861 is gratefully acknowledged.

The authors declare no conflicts of interest regarding the publication of this paper.

Aghaei, M.S.M. and Thiemann, T. (2024) Steroids, Tri- and Meroterpenoids with a Quinone Structure—A Review. International Journal of Organic Chemistry, 14, 32-68. https://doi.org/10.4236/ijoc.2024.141003

A synthetic entry to a quinoid structure in the A-ring of steroids was achieved with the possibility of positioning a hydroxyl function at C10. Already in 1958, A. M. Gold and E. Schwenk had shown that the action of lead tetraacetate on estrone (87) in glacial acetic acid leads to 10ξ-acetoxy-1,4-estradiene-3,17-dione (118), albeit only in 20% yield [59] (Scheme 18). Y. Yamada et al. reacted estrone with thallium perchlorate at 25˚C in a mixture of CH₂Cl₂ and perchloric acid [60] . However, here the 10ξ-hydroxy-1,4-estradiene-3,11,17-trione (120, 40%) accompanied 121 (20%) as the major product (Scheme 18). B. A. Šolaja et al. found that the photoirradiation (60W tungsten lamp) of estrone (87) and estradiol 17-acetate (128) in the presence of meta-chloroperxoybenzoic acid (m-CPBA) - benzoylperoxide [(BzO)₂] (solvent system: CH₂Cl₂/acetone 4:1; 3.5 and 36 h) results in p-quinols 119 and 129 in 57 and 50%, respectively [61] . The C10β-stereochemistry of the hydroxyl group in 119 has been confirmed by NOE experiments [61] . Interestingly, an exchange of a 60W lamp in the photoirradiation of 87 to a 250W lamp changed the outcome of the reaction, with 122 (Figure 9) as the main compound. p-Quinol 119 can be converted to steroidal quinone 121 by reaction of 119 with HCl in AcOH under reflux to give the hydroquinone 123 (Figure 9), which is oxidized in situ with Ag₂O to 121 [62] [63] [64] . Long reaction times transfer 121 further to epoxide 124 (Figure 9), which could be isolated in quantitative yield, when running the reaction for a number of days, and for which also an X-ray single crystal structure was obtained [64] . The underlying reaction of the transformation of p-quinol 119 to quinone 121 is a dienone-phenol rearrangement [65] , which in this case leads to a hydroquinone 123, which then is oxidized with silver oxide to quinone 121. Quinones such as 121 and its 17β-acetate derivative can be derivatized further to 2-substituted quinones 126 - 130 and 133 with O- and N-nucleophiles [62] [63] (Scheme 19 and Scheme 20). Here, the aziridine- and (S)-alanine adducts 126 and 127 were found to be very instable and their regioselectively has not been established firmly, yet [62] [63] , but it is believed that the substituents are at C2 as also shown in Scheme 19. Interestingly, from 1973, there is also a report that 3-O-methylestadiol 17β-acetate (134) can be converted directly to 3-methoxy substituted estradiol quinone 135 by the action of H₂O₂/CF₃CO₂H [66] (Scheme 21).