Advertisement

Concise chemoenzymatic total synthesis of (−)-rubroskyrin, (−)-deoxyrubroskyrin (−)-luteoskyrin, and (−)-deoxyluteoskyrin

  • Amit Mondal
    Affiliations
    Department of Biological and Synthetic Chemistry, Centre of Biomedical Research, Sanjay Gandhi Postgraduate Institute of Medical Sciences Campus, Raebareli Road, Lucknow, 226014, India
    Search for articles by this author
  • Nirmal Saha
    Affiliations
    Department of Biological and Synthetic Chemistry, Centre of Biomedical Research, Sanjay Gandhi Postgraduate Institute of Medical Sciences Campus, Raebareli Road, Lucknow, 226014, India
    Search for articles by this author
  • Syed Masood Husain
    Correspondence
    Corresponding author.
    Affiliations
    Department of Biological and Synthetic Chemistry, Centre of Biomedical Research, Sanjay Gandhi Postgraduate Institute of Medical Sciences Campus, Raebareli Road, Lucknow, 226014, India
    Search for articles by this author
Open AccessPublished:October 03, 2022DOI:https://doi.org/10.1016/j.tchem.2022.100030

      Abstract

      Synthesis of complex dimeric natural products (−)-luteoskyrin and (−)-deoxyluteoskyrin isolated from P. islandicum Sopp nearly 70 years ago, remained a challenge until now. Their biosynthesis had been proposed to involve dimerization using a putative intermediate dihydrocatenarin as a key step. In the current work, we employed a chemoenzymatic strategy to synthesize (R)-dihydrocatenarin using an anthrol reductase of T. islandicus. Its homodimerization in the presence of molecular oxygen gave (−)-rubroskyrin, which on Michael reaction led to the first total synthesis of (−)-luteoskyrin in an overall yield of 21%. In contrast, the heterodimerization between (R)-dihydrocatenarin and (R)-dihydroemodin led to non-natural, (−)-deoxyrubroskyrin analogue, while the use of molecular oxygen gave natural (−)-deoxyrubroskyrin. Both (−)-deoxyrubroskyrin and its analogue on treatment with pyridine gave (−)-deoxyluteoskyrin with an overall yield of up to 10%. The presence of dihydrocatenarin in P. islandicum NRRL 1036 culture is verified through mass spectrometry, which implied a similar biosynthetic pathway.

      Graphical abstract

      Keywords

      1. Introduction

      Penicillium islandicum is known to produce a large number of bioactive aromatic polyketides [
      • Takeda N.
      • Seo S.
      • Ogihara Y.
      • Sankawa U.
      • Iitaka I.
      • Kitagawa I.
      • Shibata S.
      ]. Investigations by Raistrick and Shibata led to the isolation of more than twenty coloured metabolites from P. islandicum Sopp and related fungus which includes modified bisanthraquinones (14, 7, 8, 11) (Fig. 1) [
      • Takeda N.
      • Seo S.
      • Ogihara Y.
      • Sankawa U.
      • Iitaka I.
      • Kitagawa I.
      • Shibata S.
      ,
      • Howard B.H.
      • Raistrick H.
      ,
      • Howard B.H.
      • Raistrick H.
      ,
      • Breen J.
      • Dacre J.C.
      • Raistrick H.
      • Smith G.
      ,
      • Shibata S.
      ]. The homodimeric, (−)-rubroskyrin (1), (−)-luteoskyrin (3), and heterodimeric, (−)-deoxyrubroskyrin (2), (−)-deoxyluteoskyrin (4) were isolated from P. islandicum Sopp NRRL 1036 [
      • Takeda N.
      • Seo S.
      • Ogihara Y.
      • Sankawa U.
      • Iitaka I.
      • Kitagawa I.
      • Shibata S.
      ,
      • Shibata S.
      ]. The hepatoxic studies of rice infected with P. islandicum identified (−)-luteoskyrin (3) as the mycotoxin, which can lead to liver cirrhosis, and sometimes liver carcinoma [
      • Shibata S.
      ,
      • Umeda M.
      ]. The (−)-luteoskyrin (3) was also found to inhibit RNA polymerase by the formation of a complex with DNA using Mg2+ ions [
      • Ueno Y.
      • Habano W.
      • Yamaguchi H.
      • Masuda T.
      • Morimura S.
      • Nemoto K.
      • Kojima S.
      • Tashiro F.
      ]. In addition, it has been shown to display antimalarial, antitubercular, antibacterial and antifungal activities in recent studies [
      • Sadorn K.
      • Saepua S.
      • Boonyuen N.
      • Komwijit S.
      • Rachtawee P.
      • Pittayakhajonwut P.
      ]. Despite, the isolation of bisanthraquinones 14 almost 70 years ago, no synthesis has been reported so far for these natural products. The biosynthetic studies suggest (−)-rubroskyrin (1) and (−)-deoxyrubroskyrin (2) as precursors for the formation of (−)-luteoskyrin (3), and (−)-deoxyluteoskyrin (4), respectively (Fig. 1A) [
      • Sankawa U.
      ]. In addition, Bu’Lock and Smith isolated a quinone from the cultures of Penicillium islandicum NRLL 1036, which was later assigned a putative structure dihydrocatenarin 5a based on the chemical analysis and mass (Fig. 1A) [
      • Takeda N.
      • Seo S.
      • Ogihara Y.
      • Sankawa U.
      • Iitaka I.
      • Kitagawa I.
      • Shibata S.
      ,
      • Bu'Lock J.D.
      • Smith J.R.
      ]. This led Shibata to propose 5a as a putative biosynthetic precursor required for the formation of dimeric intermediate, (−)-rubroskyrin (1) [
      • Takeda N.
      • Seo S.
      • Ogihara Y.
      • Sankawa U.
      • Iitaka I.
      • Kitagawa I.
      • Shibata S.
      ,
      • Ebizuka Yutaka
      • Sankawa Ushio
      • Shibata Shoji
      ]. However, the formation of dihydrocatenarin 5a and the mechanism of its conversion to (−)-rubroskyrin (1) remained elusive. Therefore, in the current work, we aimed to synthesize dihydrocatenarin and explore its dimerization to (−)-rubroskyrin (1), which can be used to develop a concise biomimetic synthesis of (−)-luteoskyrin (3). A Similar strategy can be used to synthesize heterodimeric (−)-deoxyrubroskyrin (2), (−)-deoxyluteoskyrin (4). This may provide vital clues about the putative biosynthetic intermediates involved, the mechanism of dimerization and, the formation of bisanthraquinones 14 in nature.
      Fig. 1
      Fig. 1A. Secondary metabolites isolated from Penicillium islandicum Sopp. B. Chemoenzymatic synthesis of (−)-rugulosin (8). C. Proposed biosynthesis of (+)-rugulosin (11).
      The early isolation of a related bisanthraquinone, (−)-rugulosin (8) from P. islandicum Sopp NRRL 1175, along with a dimer of unusual architecture, (−)-flavoskyrin (7) led Shibata to investigate their biosynthesis (Fig. 1B) [
      • Takeda N.
      • Seo S.
      • Ogihara Y.
      • Sankawa U.
      • Iitaka I.
      • Kitagawa I.
      • Shibata S.
      ]. They showed that it is possible to convert (−)-flavoskyrin (7) to (−)-rugulosin (8) in the presence of a base following a cascade [
      • Seo S.
      • Sankawa U.
      • Ogihara Y.
      • Iitaka Y.
      • Shibata S.
      ]. This result supported flavoskyrin-type compounds as an intermediate in the biosynthesis of bisanthraquinones [
      • Sankawa U.
      ]. However, since (−)-flavoskyrin (7) is the only compound of its type that has been isolated from any natural sources thus far, it raises the question of it being a true biosynthetic intermediate. Nevertheless, we reported the first synthesis of (−)-flavoskyrin (7), by the dimerization of chemoenzymatically synthesized dihydroemodin (6a) through inverse electron demand Diels‒Alder reaction [
      • Saha N.
      • Mondal A.
      • Witte K.
      • Singh S.K.
      • Müller M.
      • Husain S.M.
      ]. This allowed us to synthesize, (−)-rugulosin (8) in just three steps starting from emodin (9) (Fig. 1B) [
      • Saha N.
      • Mondal A.
      • Witte K.
      • Singh S.K.
      • Müller M.
      • Husain S.M.
      ]. In contrast, the (+)-rugulosin (11), initially isolated from P. islandicum Thom and later from dozens of fungi [
      • Takeda N.
      • Seo S.
      • Ogihara Y.
      • Sankawa U.
      • Iitaka I.
      • Kitagawa I.
      • Shibata S.
      ], was shown to follow different biosynthetic pathways in recent studies by Tan and co-workers [
      • Bin Han Y.
      • Bai W.
      • Ding C.X.
      • Liang J.
      • Wu S.-H.
      • Tan R.X.
      ]. Their investigation of the biosynthetic gene cluster of Talaromyces sp. YE3016 revealed, RugG as a cytochrome P450 monooxygenase that catalyzes the dimerization of emodin radicals to produce skyrin, which on reduction by RugH produces intermediate 10, that undergoes spontaneous Michael addition to form (+)-rugulosin (11) (Fig. 1C) [
      • Bin Han Y.
      • Bai W.
      • Ding C.X.
      • Liang J.
      • Wu S.-H.
      • Tan R.X.
      ].These studies suggest that possibly the biosynthesis of modified bisanthraquinones is taking place following two different pathways. One that involves dimerization through hetero-Diels–Alder reaction and another through a radical coupling. Hence, we speculated that the (−)-rubroskyrin (1) might have been formed either through flavoskyrin type intermediate like 12 or through dimeric intermediate like 13 (Scheme 1).
      Scheme 1
      Scheme 1Two possible pathways for the dimerization of dihydrocatenarin to (−)-rubroskyrin (1).

      2. Results and discussion

      2.1 Chemoenzymatic reduction of catenarin

      To investigate the process of dimerization, we first required dihydrocatenarin (5a). However, the synthesis of such a monomeric compound is quite challenging, due to the presence of multiple hydroxy groups and a chemically sensitive β-hydroxy ketone group. Considering the isolation of dihydrocatenarin 5a as well as catenarin (14) from Penicillium islandicum NRLL 1036 [
      • Bu'Lock J.D.
      • Smith J.R.
      ], we hypothesized that dihydrocatenarin (5a) may be formed by the stereo- and regioselective reduction of catenarin (14) by the enzyme(s) present in the fungus, followed by tautomerism. To test our hypothesis, we planned to use recently identified NADPH-dependent anthrol reductase of T. islandicus WF-38-12 (previously called Penicillium islandicum) for the reduction [
      • Singh S.K.
      • Mondal A.
      • Saha N.
      • Husain S.M.
      ]. The same fungus is known to produce a large number of anthraquinones and bisanthraquinones [
      • Schafhauser T.
      • Wibberg D.
      • Rueckert C.
      • Winkler A.
      • Flor L.
      • van Pée K.H.
      • Fewer D.P.
      • Sivonen K.
      • Jahn L.
      • Ludwig-Müller J.
      • Caradec T.
      • Jacques P.
      • Huijbers M.M.E.
      • van Berkel W.J.H.
      • Weber T.
      • Wohlleben W.
      ], and therefore supports the use of anthrol reductase of T. islandicus (ARti) for the reduction of catenarin. To test the reduction, ARti was obtained by the expression of ARti_his gene in E. coli BL21 (DE3) following a reported procedure (see Supporting Information) [
      • Singh S.K.
      • Mondal A.
      • Saha N.
      • Husain S.M.
      ]. The desired substrate, catenarin (14) was synthesized by the oxidation of emodin (9) using oleum (65%) and boric acid for 3 days in 54% yield [
      • Morooka N.
      • Nakano S.
      • Itoi N.
      • Veno Y.
      ]. This is followed by incubation of 14 with ARti in KPi buffer (50 ​mM, pH 7.0) and NADPH (regenerated using glucose/glucose dehydrogenase). However, no product was formed under those conditions. Therefore, catenarin (14) was incubated with ARti and NADPH in the buffer in the presence of 20 equiv. of Na2S2O4 under anoxic conditions. Sodium dithionite is used for in situ reduction of catenarin to 15a/15b, which may act as a substrate for enzymatic reduction. The formation of 15a/15b, which are highly prone to oxidation was confirmed by quick extraction in acetonitrile-d3 and their characterization using NMR spectroscopy and mass spectrometry (Figure S1, Supporting Information).
      After 14 ​h, the reaction showed the formation of several products on TLC. The purification by oxalic acid impregnated silica gel, resulted in the isolation of a small amount of 3,5,8,9,10-pentahydroxy-6-methyl-3,4-dihydroanthracen-1(2H)-one (16), islandicin (17) and two unexpected side products, 18 and 19 (Scheme 2). To investigate the formation of the unexpected side products 18 and 19, catenarin (14) was incubated just in the presence of Na2S2O4 in the same buffer. This also resulted in the formation of side products (18 & 19), which suggests that these side products are formed non-enzymatically by the reduction of hydroanthraquinones 15a/15b or their tautomers similar to previously reported reductions of anthraquinones [
      • Prinz H.
      • Wiegrebe W.
      • Müller K.
      ]. Attempts to prevent the formation of side products by replacing Na2S2O4 with other reducing agents (SnCl2, H2S) failed. Therefore, we optimized the reaction conditions to improve the yield of the desired compound 16 (Table S1, Supporting Information). We found that the use of 2 equivalents of NADP+ for regeneration of NADPH by glucose/GDH system and lowering of temperature gave the best conversion of 69% to 16, with 45% isolated yield (Scheme 2).
      Scheme 2
      Scheme 2Chemoenzymatic reduction of catenarin using ARti and NADPH in the presence of Na2S2O4.
      Furthermore, the enantiomeric purity was determined using racemic 16, which was synthesized by the reduction of catenarin (14) using NaBH4 in the presence of Na2S2O4 in water. The chiral HPLC shows >99% ee for 16. Finally, the (R)-configuration was assigned to 16 by comparing the CD spectra of 16 with that of (R)-26 obtained by the chemoenzymatic reduction of emodin (9) using ARti and NADPH reported previously (see Supporting Information) [
      • Singh S.K.
      • Mondal A.
      • Saha N.
      • Husain S.M.
      ].

      2.2 Synthesis of (R)-dihydrocatenarin and its homodimerization to (−)-rubroskyrin

      To obtain dihydrocatenarin, we aimed to oxidize (R)-16 to its anthraquinone. To achieve this, (R)-16 was stirred in AcOH with Pb(OAc)4 ​at 0 ​°C temperature. After 30 ​min, the 1H NMR of the crude reaction mixture showed the formation of dihydrocatenarin 5a as a major product along with islandicin (17), which might have been formed by dehydration during the extraction process (Scheme 3). This is by far the first known synthesis of this key biosynthetic intermediate 5a. Interestingly, when the reaction mixture was subjected to purification using oxalic acid impregnated silica gel, it resulted in the isolation of dimerized product, (−)-rubroskyrin (1) and islandicin (17), which might have been formed on the column. Therefore, unpurified dihydrocatenarin 5a was treated directly with silica gel, and the products were extracted in MeOH–CHCl3 (20:80) mixture. The purification gave (−)-rubroskyrin (1) in 38% yield, which was fully characterized using NMR spectroscopy and mass spectrometry (see Supporting Information). In the light of earlier studies on dimerization, (−)-rubroskyrin (1) is expected to be formed via a flavoskyrin type intermediate 13, which is supposed to be the result of a hetero-Diels‒Alder reaction between the enone tautomer 5c of the dihydrocatenarin 5a (Scheme 3). However, no intermediate like 13 was isolated from the reaction mixture. This might be due to the presence of a phenolic group in C-5 position of 5c, which is involved in hydrogen bonding, and therefore could not act as a diene for the expected hetero-Diels‒Alder cycloaddition reaction. This suggests that there might be an alternate mechanism operational for the formation of (−)-rubroskyrin (1) from dihydrocatenarin 5a. The hint came from the observation, that (R)-16 oxidizes to dihydrocatenarin 5a in an NMR tube over time, possibly by aerial oxidation (Figure S2, Supporting Information). Therefore, we aimed to test the oxidation of (R)-16 by molecular oxygen under the optimized condition reported previously [
      • Mondal A.
      • De A.
      • Husain S.M.
      ]. For this, the KPi buffer (50 ​mM, pH 6) containing (R)-16 dissolved in acetonitrile (30% v/v) was bubbled with the molecular oxygen at a low temperature (10–15 0C) for 6 ​h. The low temperature was used to prevent the formation of islandicin (17) through dehydration.
      Scheme 3
      Scheme 3Homodimerization to (−)-rubroskyrin (1) through oxidation of (R)-16 using A) Pb(OAc)4 and B) molecular oxygen as oxidants.
      Interestingly, the 1H NMR of the crude reaction mixture showed the formation of (−)-rubroskyrin (1) along with other side products. Purification of the reaction mixture with oxalic acid impregnated silica gel allowed us to isolate (−)-rubroskyrin (1) in 55% yield, along with islandicin (17) in 25% yield and a small amount of a deep red coloured compound. The characterization by 1H, 1D TOCSY NMR experiments and mass spectrometry confirmed it to be a single C–C bonded dimeric intermediate 23 (Figure S3, Supporting Information). We proposed that it is intermediate 23 which on intramolecular Michael addition forms (−)-rubroskyrin (1). We believe that 23 might be the result of the tautomerization of intermediate 22, which was formed by the coupling of two molecules of radical 21 (Scheme 3). The formation of such an intermediate is further supported by the recent studies on the biosynthesis of (+)-rugulosin (11), in which Tan and co-workers showed the dimerization of emodin radicals involving RugG enzyme [
      • Bin Han Y.
      • Bai W.
      • Ding C.X.
      • Liang J.
      • Wu S.-H.
      • Tan R.X.
      ]. Similar dimerization took place in our case also but through a non-enzymatic route. We believe that monomeric radicals 20/21 were formed due to the ROS generated through autooxidation of (R)-16 to dihydrocatenarin 5a in the presence of molecular oxygen, which is commonly observed in hydroanthraquinone-anthraquinone redox system [
      • Szwalbe J.
      • Williams K.
      • Song Z.
      • De Mattos-Shipley K.
      • Vincent J.L.
      • Bailey A.M.
      • Willis C.L.
      • Cox R.J.
      • Simpson T.J.
      ,
      • Nishimi T.
      • Kamachi T.
      • Kato K.
      • Kato T.
      • Yoshizawa K.
      ]. Hence, the dimerization of putative biosynthetic intermediate dihydrocatenarin (5a) to (−)-rubroskyrin (1) took place through radical coupling in the presence of molecular oxygen, without the involvement of a flavoskyrin type intermediate. This is a crucial advancement in our understanding of the non-enzymatic dimerization process that might be operative in the biosynthesis of bisanthraquinones.

      2.3 Conversion of (−)-rubroskyrin to (−)-luteoskyrin

      Next, we planned to convert (−)-rubroskyrin (1) to (−)-luteoskyrin (3), which required the formation of a third C–C bond and could be achieved by the base catalysed Michael reaction. Therefore, 1 was incubated in pyridine under the oxygen atmosphere. This led us to obtain (−)-luteoskyrin (3) with quantitative conversion and 87% isolated yield (Scheme 4).
      Scheme 4
      Scheme 4Cascade conversion of (−)-rubroskyrin (1) to (−)-luteoskyrin (3) via intermediate 24 in pyridine and synthesis of (+)-iridoskyrin (25).
      The comparison of NMR and CD data of the synthesized and natural (−)-luteoskyrin (3) showed it to be the same compound (see ESI) [
      • Shibata S.
      ,
      • Bouhet J.C.
      • Pham Van Chuong P.
      • Toma F.
      • Kirszenbaum M.
      • Fromageot P.
      ]. For the formation of (−)-luteoskyrin (3), we proposed that the tautomer 24 of 1 undergoes Michael addition in the presence of a base, to form 3 (Scheme 4). In addition, the treatment of (−)-rubroskyrin (1) with conc. H2SO4 gave the dehydrated bisanthraquinone, (+)-iridoskyrin (25) in 71% yield (Scheme 4). Hence, despite the complexity, the homodimeric (−)-luteoskyrin (3) was synthesized in just three steps from catenarin (14), for the first time, following a simple, biomimetic, chemoenzymatic strategy.

      2.4 Oxidation of dihydroemodin and dihydrocatenarin by Pb(OAc)4

      This motivated us to synthesize the heterodimeric natural product (−)-deoxyrubroskyrin (2) and (−)-deoxyluteoskyrin (4). Considering their isolation from the same fungus that is P. isilandicum and looking at the structural similarity, we speculated that (−)-deoxyrubroskyrin (2) might act as a precursor for the formation of deoxyluteoskyrin (4). Hence, to obtain (−)-deoxyrubroskyrin (2), we proposed heterodimerization between monomeric dihydroemodin (R)-6a and dihydrocatenarin (R)-5a. Therefore, (R)-3,8,9,10-tetrahydroxy-6-methyl-3,4-dihydroanthracen-1(2H)-one (26) was synthesized by the chemoenzymatic reduction of emodin (9) using ARti in the presence of Na2S2O4 and NADPH (regenerated using glucose/GDH system) (see Supporting Information).
      Then, the 1:1 mixture of (R)-16 and (R)-26 was subjected to oxidation first using Pb(OAc)4 in acetic acid. The analysis of 1H NMR of the crude reaction mixture showed the formation of a new heterodimeric, (−)-flavoskyrin analogue 28 in a small amount along with the homodimeric bisanthraquinones, (−)-flavoskyrin (7), (−)-rubroskyrin (1), and deoxyanthraquinones, chrysophanol (30) and islandicin (17). The attempts to purify intermediate 28 were not successful. However, its presence in a mixture with (−)-rubroskyrin (1) was confirmed by NMR spectroscopy and mass spectrometry (Figure S5, S6, Supporting Information). The formation of 28 can be explained by the possible hetero-Diels–Alder reaction between 6b and 5c. This is because, the dienol tautomer 6b of dihydroemodin, could act as a diene and 5c function as a dienophile during the hetero-Diels‒Alder reaction, which resulted in the formation of flavoskyrin type intermediate 28 (Scheme 5). Since 5c could not act as diene during hetero-Diels‒Alder reaction, due to the hydrogen bonding, the probable intermediate 27 could not be formed as shown in arrangement B (Scheme 5). Considering the reported conversion of (−)-flavoskyrin (7) to (−)-rugulosin (8) [
      • Seo S.
      • Sankawa U.
      • Ogihara Y.
      • Iitaka Y.
      • Shibata S.
      ], partially purified 28 was incubated in pyridine under an argon atmosphere. This resulted in the formation of a non-natural (−)-deoxyrubroskyrin analogue 2’ which was purified by oxalic acid impregnated silica gel in 67% yield and characterized using NMR spectroscopy and mass spectrometry (Scheme 5). The overall yield of the reaction was determined to be 34%.
      Scheme 5
      Scheme 5Heterodimerization of (R)-16 and (R)-26 using Pb(OAc)4 as an oxidizing agent and synthesis of (−)-deoxyrubroskyrin analogue 2’.

      2.5 Oxidation of dihydroemodin and dihydrocatenarin by molecular oxygen

      To investigate the heterodimerization in the presence of molecular oxygen, the 1:1 mixture of (R)-16 and (R)-26 was oxidized in potassium phosphate buffer (pH 6.0) by bubbling molecular oxygen. This also resulted in the formation of the expected homo-as well as heterodimeric products. Purification of the reaction mixture using oxalic acid-impregnated silica gel allowed us to isolate a new deep red-coloured compound in 28% yield. The characterization by NMR spectroscopy and mass spectrometry confirmed it to be (−)-deoxyrubroskyrin (2). Hence, the complex, heterodimeric, (−)-deoxyrubroskyrin (2) was synthesized for the first time in just three steps starting from emodin. In addition, homodimeric natural products, (−)-flavoskyrin (7), (−)-rubroskyrin (1) and deoxyanthraquinones, islandicin (17), chrysophanol (30) (Scheme S6, Supporting Information) were isolated as side products in 5–10% yield. Interestingly, we also obtained (+)-roseoskyrin (34) in a 5% yield (Scheme 6). The absence of heterodimeric (−)-flavoskyrin (27 or 28) under these conditions, suggests that the dimerization took place through radical coupling between the two monomeric anthraquinone radicals (20 & 31) (Scheme 6).
      Scheme 6
      Scheme 6Synthesis of deoxyrubroskyrin (2) by the oxidation of a mixture of (R)-16 and (R)-26 in the presence of molecular oxygen.
      We believe that such radicals are formed due to ROS formed during the oxidation of (R)-16 and (R)-26 in the presence of molecular oxygen, giving a single C–C bonded intermediate 32, which on dehydration forms a natural product, (+)-roseoskyrin (34). Although the putative intermediate 32 was not isolated from the reaction mixture, the isolation of a single C–C bonded natural product 34 supports the formation of intermediate 32.32 is expected to undergo tautomerization to form intermediate 33, which on intramolecular Michael reaction gives natural (−)-deoxyrubroskyrin (2) and supports a similar type of non-enzymatic coupling during the biosynthesis (Scheme 6).

      2.6 Cascade synthesis of (−)-deoxyluteoskyrin

      Next, we focussed on converting (−)-deoxyrubroskyrin (2) to (−)-deoxyluteoskyrin (4), through Michael addition. Therefore, (−)-deoxyrubroskyrin (2) was incubated in pyridine under an oxygen atmosphere overnight. This resulted in quantitative conversion to (−)-deoxyluteoskyrin (4), which on purification gave 82% yield (Scheme 7). We believe that (−)-deoxyrubroskyrin 2 is formed through tautomer 35, which on Michael addition in presence of pyridine gives (−)-deoxyluteoskyrin (4) (Scheme 7).
      Scheme 7
      Scheme 7Cascade conversion of (−)-deoxyrubroskyrin (2) and (−)-deoxyrubroskyrin analogue (2′) to (−)-deoxyluteoskyrin (4).
      Interestingly, when (−)-deoxyrubroskyrin analogue 2′ was incubated in pyridine in the presence of molecular oxygen, it also resulted in quantitative conversion to (−)-deoxyluteoskyrin (4), which was purified using oxalic acid impregnated silica gel to obtain 4 in 80% yield (Scheme 7). To investigate the cascade conversion of (−)-deoxyrubroskyrin analogue 2′ to the (−)-deoxyluteoskyrin (4), a reaction was performed in pyridine-d5 and changes were monitored through 1H NMR (Figure S7, Supporting Information). We found that (−)-deoxyrubroskyrin analogue 2’ undergoes oxidation to form intermediate 36, which on Michael addition forms (−)-deoxyluteoskyrin (4). It is the formation of the third C–C bond via intermediate 35 or 36, that brings symmetry to the molecule and hence resulted in the formation of the same bisanthraquinone, (−)-deoxyluteoskyrin (4) as observed (Scheme 7). In comparison to the 16-steps total synthesis of (+)-rugulosin, which required extensive protection [
      • Nicolaou K.C.
      • Yee H.L.
      • Piper J.L.
      • Papageorgiou C.D.
      ], our concise chemoenzymatic approach gives access to complex bisanthraquinones in just 3–4 steps.
      To validate the biomimetic nature of our synthesis, we aimed to confirm the presence of the putative biosynthetic intermediates synthesized in this study, in the culture of the fungus Penicillium islandicum NRLL 1036. Hence, the fungus was grown in Czapek medium on agar plates (see Supporting Information). The extraction of the dried mycelium after 15 days of culture using ethyl acetate and analysis through NMR spectroscopy and mass spectrometry confirmed the presence of known metabolites including (−)-rubroskyrin (1) and (−)-luteoskyrin (3). Interestingly, the HRMS of the crude extract showed the presence of a molecular ion peak corresponding to dihydrocatenarin 5a (Figure S7, see Supporting Information), which has been synthesized by us using a chemoenzymatic approach. This confirms that dihydrocatenarin claimed to be isolated by Bu’Lock and Smith does exist in the culture of P. islandicum NRRL 1036, and is more likely to be involved in the biosynthesis of (−)-luteoskyrin (3) and (−)-deoxyluteoskyrin (4) similar to our chemoenzymatic and biomimetic approach.

      3. Conclusions

      Here, we report the first synthesis of (−)-luteoskyrin (3) and (−)-deoxyluteoskyrin (4) and their biosynthetic intermediates (−)-rubroskyrin (1) and (−)-deoxyrubroskyrin (2) using a chemoenzymatic, biomimetic approach in just 3–4 steps. The key biosynthetic precursor, dihydrocatenarin 5a was synthesized chemoenzymatically from catenarin using an anthrol reductase. The dihydrocatenarin (R)-5a is being used for homo- and heterodimerization in the presence of molecular oxygen to obtain (−)-rubroskyrin (1) and (−)-deoxyrubroskyrin (2) in 55% and 28% yield, respectively. In addition, the heterodimerization using Pb(OAc)4 gave the (−)-deoxyrubroskyrin analogue 2’. This suggests the probable involvement of a non-enzymatic radical cross-coupling for dimerization to (−)-rubroskyrin (1) and (−)-deoxyrubroskyrin (2) using molecular oxygen during biosynthesis. Finally, these dimeric intermediates were converted to (−)-luteoskyrin (3) and (−)-deoxyluteoskyrin (4) on treatments with pyridine following a cascade. In addition, the presence of dihydrocatenarin 5a in the culture of P. islandicum NRRL 1036 was confirmed through HRMS, which implies its role in the biosynthesis of bisanthraquinones 14 and supports the biomimetic nature of our chemoenzymatic strategy.

      4. Experimental procedures

      Protein expression and purification, experimental details, characterization data, copies of NMR spectra, CD-spectra, and HPLC spectra (PDF) is available in Supporting Information.

      Author contributions

      AM, NS and SMH designed the experiments. AM performed the major work with assistance from NS. AM and SMH wrote the manuscript.

      Data availability

      All data is available for free online as Supporting Information

      Declaration of competing interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Acknowledgement

      We are grateful to the Council of Scientific and Industrial Research, New Delhi (Project no. 02(0448)/21/EMR-II) and SERB(CRG/2018/002682) for funding and CSIR, New Delhi for funding fellowships to A. M. and N. S. We are also thankful to the Department of Medical Education, Uttar Pradesh and the Director of CBMR for research facilities.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Takeda N.
        • Seo S.
        • Ogihara Y.
        • Sankawa U.
        • Iitaka I.
        • Kitagawa I.
        • Shibata S.
        Tetrahedron. 1973; 29: 3703-3719
        • Howard B.H.
        • Raistrick H.
        Biochem. J. 1949; 44: 227-233
        • Howard B.H.
        • Raistrick H.
        Biochem. J. 1954; 57: 212-222
        • Breen J.
        • Dacre J.C.
        • Raistrick H.
        • Smith G.
        Biochem. J. 1955; 60: 618-626
        • Shibata S.
        Pure Appl. Chem. 1973; 33: 109-128
        • Umeda M.
        Pathol. Int. 1964; 14: 373-394
        • Ueno Y.
        • Habano W.
        • Yamaguchi H.
        • Masuda T.
        • Morimura S.
        • Nemoto K.
        • Kojima S.
        • Tashiro F.
        Food Chem. Toxicol. 1991; 29: 607-613
        • Sadorn K.
        • Saepua S.
        • Boonyuen N.
        • Komwijit S.
        • Rachtawee P.
        • Pittayakhajonwut P.
        Tetrahedron. 2019; 75: 3463-3471
        • Sankawa U.
        Steyn P.S. The Biosynthesis of Mycotoxins. Academic Press, New York1980: 357-394
        • Bu'Lock J.D.
        • Smith J.R.
        J. Chem. Soc. C Org. 1968; : 1941-1943
        • Ebizuka Yutaka
        • Sankawa Ushio
        • Shibata Shoji
        Symp. Chem. Nat. Prod., Symposium on the Chemistry of Natural Products Steering Committee. 1973: 328-335
        • Seo S.
        • Sankawa U.
        • Ogihara Y.
        • Iitaka Y.
        • Shibata S.
        Tetrahedron. 1973; 29: 3721-3726
        • Saha N.
        • Mondal A.
        • Witte K.
        • Singh S.K.
        • Müller M.
        • Husain S.M.
        Chem. Eur J. 2018; 24: 1283-1286
        • Bin Han Y.
        • Bai W.
        • Ding C.X.
        • Liang J.
        • Wu S.-H.
        • Tan R.X.
        J. Am. Chem. Soc. 2021; 143: 14218-14226
        • Singh S.K.
        • Mondal A.
        • Saha N.
        • Husain S.M.
        Green Chem. 2019; 21: 6594-6599
        • Schafhauser T.
        • Wibberg D.
        • Rueckert C.
        • Winkler A.
        • Flor L.
        • van Pée K.H.
        • Fewer D.P.
        • Sivonen K.
        • Jahn L.
        • Ludwig-Müller J.
        • Caradec T.
        • Jacques P.
        • Huijbers M.M.E.
        • van Berkel W.J.H.
        • Weber T.
        • Wohlleben W.
        J. Kalinowsky J. of Biotechnol. 2015; 211: 101-102
        • Morooka N.
        • Nakano S.
        • Itoi N.
        • Veno Y.
        Agric. Biol. Chem. 1990; 54: 1247-1252
        • Prinz H.
        • Wiegrebe W.
        • Müller K.
        J. Org. Chem. 1996; 61: 2853-2856
        • Mondal A.
        • De A.
        • Husain S.M.
        Org. Lett. 2020; 22: 8511-8515
        • Szwalbe J.
        • Williams K.
        • Song Z.
        • De Mattos-Shipley K.
        • Vincent J.L.
        • Bailey A.M.
        • Willis C.L.
        • Cox R.J.
        • Simpson T.J.
        Chem. Sci. 2019; 10: 233-238
        • Nishimi T.
        • Kamachi T.
        • Kato K.
        • Kato T.
        • Yoshizawa K.
        Eur. J. Org. Chem. 2011; : 4113-4120
        • Bouhet J.C.
        • Pham Van Chuong P.
        • Toma F.
        • Kirszenbaum M.
        • Fromageot P.
        J. Agric. Food Chem. 1976; 24: 964-972
        • Nicolaou K.C.
        • Yee H.L.
        • Piper J.L.
        • Papageorgiou C.D.
        J. Am. Chem. Soc. 2007; 129: 4001-4013