Synthesis of the 8,19-Epoxysteroid Eurysterol A.

Abstract We report the first chemical synthesis of eurysterol A, a cytotoxic and antifungal marine steroidal sulfate with a unique C8−C19 oxy‐bridged cholestane skeleton. After C19 hydroxylation of cholesteryl acetate, used as an inexpensive commercial starting material, the challenging oxidative functionalization of ring B was achieved by two different routes to set up a 5α‐hydroxy‐7‐en‐6‐one moiety. As a key step, an intramolecular oxa‐Michael addition was exploited to close the oxy‐bridge (8β,19‐epoxy unit). DFT calculations show this reversible transformation being exergonic by about −30 kJ mol−1. Along the optimized (scalable) synthetic sequence, the target natural product was obtained in only 11 steps in 5 % overall yield. In addition, an access to (isomeric) 7β,19‐epoxy steroids with a previously unknown pentacyclic ring system was discovered.

here disclose the resultso fastudy which has culminated in the elaboration of af irst (and even scalable) synthesis of eurysterol A.
Our retrosynthetic analysis (Scheme1)s tartedw ith the consideration that the sodium sulfateg roup should be installed at al ate stage of the synthesis since it renders the molecule water-soluble.W et hus selected the dihydroxyketone 4 as a pre-target molecule which could be selectively sulfated at the secondary OH group followed by diastereoselective reduction of the keto function. As ak ey step,w ee nvisioned to exploit an oxa-Michael addition [5] to close the oxy-bridge between C8 and C19.
It appearedf easible to us to preparet he required enone of type 5 by semi-synthesis from commercial cholesteryl acetate 9 through the known 19-hydroxylated derivative 8. [6] However, a crucial aspecto ft he synthetic plan was the oxidativef unctionalizationo fr ing B, that is, the conversion of 8 into 5.T his task oughtt ob ea chieved by two different approaches.A safirst option( route A), we considered ar egioselective oxidation of the D 5 -double bond of a7 ,8-dehydro-steroid of type 6.A lternatively( route B), the double bond in 9 could first be oxidized to give ak etol intermediate of type 7 which then would have to be converted into the enone 5 by a,b-dehydrogenation of the ketone.
As af irst task, we converted cholesteryl acetate 9 into the 19-hydroxyd erivative 8 following the method of Heusler and Kalvoda (Scheme 2). [6,7] This method exploits the 1,3-diaxial vicinity of the 6b-OH group to the angular C19-methyl group in the bromohydrin intermediate 10 to achievear emote functionalizationb yr adical hydrogen atom transfer.I nc ontrast to the original protocol, we performed the (photo-mediated) hy-poiodite reactione mploying (diacetoxyiodo)benzene (DIB) in cyclohexane [8] insteado ft oxic lead tetraacetate in benzene. After treatment of the resulting 6,19-epoxy compound 11 with zinc in AcOH the desired alcohol 8 was obtained in 43 %o verall yield on a2 5gram scale and with as ingle chromatographic purification at the very end of the sequence.
According to strategy A( Scheme 1), we next investigated the preparation and oxidation of cholesta-5,7-dien-3,19-diol derivatives of type 6.A fter protecting the free alcohol group of 8 as aT BS ether (12)t he O-acetyl group at C3 was replaced by aM OM group to give 13 (Scheme 3). This was necessary to insure compatibility with the conditions of the later Bamford-Stevens elimination. [9] Originally,t he allylic oxidationo ft he alkene 13 to the enone 14 was performed using the Collins-Ratcliffe reagent (CrO 3 ·2p y, 77 %; see the Supporting Informa-tion for details). [10] However,t oa void the use of stoichiometric amountso ft oxic chromium(VI) we also tested other methods and found that the transformationo f13 to 14 coulda lso be achieved in comparabley ield (71 %) with tert-butyl hydroperoxide as the main oxidant in the presence of catalytic amounts (0.7 mol %) of RuCl 3 . [11] The keto group of 14 was then converted into the corresponding tosylhydrazone (syn/anti mixture) from which the D 5, 7 -diene 15 was obtained in high yield upon treatment with LiH in refluxing toluene. [12] Initial attempts to regioselectively oxidizet he D 5 -doubleb ond of 15 employing different Cr VI reagents [13] only gave low yields. Moreover,t he desired ketol product obtained from 15 using in situ generated RuO 4 [14] did not yield any of the desired 8,19-epoxy product 18 upon TBAF-mediated deprotection of the TBS ether (see Supporting Information for details). [15] Therefore, we replaced the TBS by an acetyl protecting group and examined the oxidation of the resulting diene 16 which provedt ob ep articularly difficult. All attempts to achieve this reaction by OsO 4 -catalyzed dihydroxylation [16] or by methyltrioxorhenium-catalyzed reactionw ith urea-H 2 O 2 [17] failed. Only the protocol of Plietker (using NaIO 4 in the presence of CeCl 3 andc atalytic amounts of Scheme1.Retrosynthetic analysis of eurysterol A( 1).
RuCl 3 ) [18] afforded the desired a-diol, albeit in only 21 %y ield (61 %b ased on recovered 16). Nevertheless,o xidation of the allylic OH group at C6 with MnO 2 afforded the desired ketol 17 which could be used to study the planned key step of the synthesis, that is, the construction of the 8,19-epoxyb ridge throughi ntramolecular oxa-Michaeladdition.
Much to our satisfaction, the desired cyclization product 18 was indeed formed upon treatment of 17 with K 2 CO 3 in MeOH. Subsequentr emoval of the MOM protecting group under mild conditions [19] finally afforded the anticipated pentacyclic pretarget compound 4,the structure of which was unambiguously confirmed by X-ray crystallography (Figure 2). [20] Having thus demonstrated the general feasibility of our synthetic strategy,t he unsatisfying efficiency of the developed sequence( Scheme 3) prompted us to also investigate route B (compare Scheme 1). After considerable experimentation (testing different combinations of protecting groups), we came up with the improved sequence outlined in Scheme 4. In this case, the OH group of 8 was protected by acetylation and the resultingd iacetate 19 wasc onverted to the ketol 20 by Os-catalyzed dihydroxylation of the D 5 -doubleb ond in presence of citric acid [21] and subsequent Dess-Martin oxidation. [22] The installation of the D 7 -doubleb ond by direct dehydrogenation of 20 could not be achieved under the conditions (IBX in DMSO) of Nicolaou. [23] However,w es ucceeded in achieving the desired a,b-dehydrogenation through an a-bromination/elimination sequence. [24] Thus, reaction of ketol 20 with bromine in the presence of acatalytic amount of HBr in acetic acid yielded the a-brominated intermediate which upon treatment with LiBr and Li 2 CO 3 in refluxing DMF afforded the enone 21 in satisfyingy ield.
Reaction of enone 21 with K 2 CO 3 in methanola tr oom temperaturef or 14 hours not only resulted in the cleavage of both acetoxyg roups but also (again)i naspontaneous cyclization (intramolecular oxa-Michael reaction) to furnish the 8b,19epoxy steroid 4 in 30 %i solated yield. Notably,d espite the full conversion of the startingm aterial 21,w ew ere unable to isolate any side product.
To shed some light on the thermodynamicso ft he (reversible) oxa-Michael reaction 21!4 (Scheme 3), we calculated the relative Gibbs free energies for the model systemsd epictedi n Scheme 5a tt he DLPNO-CCSD(T)/def2-TZVPPD/SMD(MeOH)// TPSS-D3BJ/6-31 + G(d,p)/SMD(MeOH) level of theory. [25] Based on our calculations, we can conclude that the intramolecular cyclizationso fb otht he neutral system A as well as the anionic system A' are exergonic reactions und clearly favor the cyclized products B and B'b yapprox.30kJmol À1 .
The comparison of the spectroscopicd ata of our synthetic product with those reportedf or naturale urysterol A( 1)c onfirmed the identityo fb oth samples (see Supporting Information). Moreover,w es ucceeded in growing crystalso fe urysterol A( 1)w hat allowed us to determine its precise structure by means of X-ray crystallography.W hile the constitutional and configurational assignments were confirmed, the structure also revealed an intramolecular hydrogen bridge between the axial OH group at C6 and the epoxy bridge ( Figure 3).
As an additional outcome of the synthetic endeavor described herein, we by chance also discovered as ynthetic access to iso-4,w hich is ac onstitutional isomer of 4 and af irst representative of the so far completely undescribed class of 7,19-epoxy steroids. As shown in Scheme 7, we oxidized the double bond of MOM-protected 19-hydroxy-cholesteryl acetate to the corresponding ketol related to 20.H owever,i nt his case, the a-bromination of the ketone went along with the cleavage of the MOM group to give the hemiacetal 23.T oo ur surprise, the envisaged elimination then did not take place upon heating of 23 with Li 2 CO 3 /LiBr in DMF.I nstead,t he 7,19-epoxy bridge was formed, probably by S N 2r eactiono ft he anionic intermediate 23' to give iso-4 after methanolytic cleavage of the acetate protecting group in high yield (Scheme7).
In conclusion, we have elaborated an efficient semi-synthesis of eurysterol A( 1)s tarting from inexpensive cholesteryl acetate. The synthetic sequence (11steps;5%o verall yield), which opensa ne ntry into the class of 8,19-epoxy steroids fort he first time, is scalable, requires only as ingle protection step, and exploits an intramolecular oxa-Michael addition as ak ey step to close the oxy-bridge betweenC 8a nd C19. Importantly, an ovel, practical and highly efficient protocol for the final sulfation step was introduced as well. The developed route allows the production of substantial amounts of the target sterol (150 mg prepared). In addition, we also discovereda ne fficient entry towards 7b,19-epoxy steroids, ap reviously unknown class of compounds with as lightly different (isomeric) pentacyclic ring system.
Thus, this work paves the way for the future exploration of the eurysterols and related epoxy steroidsa sp otentialb ioactive compounds. Considering the ongoing interesti nt he synthesis of steroids with unusual oxidationa nd ring patterns [26] we are convinced that the developedp rotocols for the B-ring functionalization of 19-oxygenated steroids will prove of value also for other researchers in the future.