Short, Tin‐Free Synthesis of All Three Inthomycins

Abstract The inthomycins are a family of structurally and biologically rich natural products isolated from Streptomyces species. Herein the implementation of a modular synthetic route is reported that has enabled the enantioselective synthesis of all three inthomycins. Key steps include Suzuki and Sonogashira cross‐couplings and an enantioselective Kiyooka aldol reaction.

The inthomycins A-C (1-3,F igure 1), also known as the phthoxazolins, are af amily of oxazole triene natural products isolated from Streptomyces culture that displayboth interesting structuresa nd aw ide range of biological activities. The isolation of inthomycin Aw as first reported by Ō mura in 1990 [1] who subsequently reported the isolation of inthomycinsBand Ci n1 995. [2] Between thesed ates, Zeek had reported the reisolation of inthomycin Aa nd the first isolation of inthomycins B and C. [3] InthomycinAwas discovered in as creen for inhibitors of cellulose biosynthesis, however,n ot only does inthomycin A inhibit cellulose biosynthesis, [1,4] but it also showsb oth herbici-dal, [4,5] and antifungal [4] activity,a nd both inthomycin A [6] and inthomycin B [6b] inhibitp rostatec ancer cell growth. Very recentlyt he cytotoxicity of inthomycin Ca gainst ar ange of human cancerc ell lines has been investigated but the natural product showed little biological activity.H owever,aclose analogue (23)w as foundt oh ave proteasome inhibition activity. [7] Apartf rom their biological significance,t he structures of the inthomycins are particularly strikingin that they contain a methylene interrupted oxazolyl-trienem oiety including at risubstituted alkene and ac hiral allylic b-hydroxy carbonyl moiety.M oreover,t he full structuralm otif of the inthomycinsi s found within an umber of more complex natural products including the oxazolomycins, 16-methyloxazolomycin,c urromycin Aand B, and KSM 2690.
Given their wide ranging biological activities and interesting structures,t he inthomycins have attracted considerable attention from the synthetic community,a lthought heir deceptively simple structures belie the challenge associated with their synthesis. To date, synthetic effort has primarily focused on intho-mycinCwith only one reporto nt he enantioselectives ynthesis of inthomycinA ando nly two on the enantioselective synthesis of inthomycin B. The first synthesis of inthomycin A( 1)i n racemic form was disclosedb yW hitingi n1 999 [8] with the only enantioselective synthesis of 1 being reported by Hatakeyama in 2012 [9] who disclosed the enantioselectives ynthesis of inthomycins B 2 andC3 in the same publication. In 2006 Ta ylor reported the first enantioselective synthesis of inthomycin B [10] 2 followed by ar eport of the enantioselective synthesis of inthomycin C 3 and of racemic inthomycin A 1 in 2008. [11] In 2010,R yu reported an enantioselective synthesis of inthomycin C 3 [12] which was followed by Hale's reports on the enantioselectives ynthesis of 3.
[13] Very recently the Donohoe group published an enantioselectives ynthesis of inthomycin C 3. [7] All of the syntheses of the inthomycins bar one, [7] feature aS tille cross-couplinga sakey step with the inherent problems associatedw ith the toxicitya nd disposal of stoichiometrico rganotin waste. Herein we report short (9/10 steps,l ongest linear sequence), tin-free syntheses of all three inthomycins using Suzuki or Sonogashira couplings as key steps and aK iyooka aldol to set the necessary asymmetry.
Our synthesis commenced with the preparation of the enyne oxazole 6 (Scheme 1). Commercially available( E)-pent-2en-4-yn-1-ol 7 [15] was readily converted into the knownb romide 12 [16] in two simple steps. After careful optimization we found that lithiation of the known silyl-protected oxazole 13 [17] with n-butyllithium followed by addition to CuCN·LiCl and addition of the bromide 12 gave the desired coupled product 14 in 81 %y ield. [18] The next stage in the synthesis involved the seemingly simple selective silyl group removal from 14 which provedu nexpectedly challenging. Commonly used basic conditions for trimethylsilyl group deprotection failed to give the desired product with allene formationb eing the major reaction pathway. [19] The use of silver(I)s alts to promote acetylene deprotection resulted in the formation of mixtures of starting material 14,t he desired product 6,a nd fully desilylated material. Ultimately, we found that modifying Basak's procedure [20] by using sodium sulfide in am ixture of THF and water,g ave the desired mono-desilylated product 6 without allene formation although the reactiond id not reachc ompletion;t he product 6 could be obtained in 85 %y ield after one recycle. Zirconium catalyzed hydroboration of the terminal acetylene in 6 gave the desired (E,E)-dienylboronic ester 15 in good yield andw ith complete stereocontrol. [21] The necessary Suzuki coupling partners for the dienylboronic ester 15 were prepared from the known (Z)-and (E)-iodoalkenes (Z)-16 and (E)-16 (Scheme 2). Thus, propargyl alcohol 11 was readilyc onverted into the (Z)-and (E)-alkenyl iodides (Z)-16 [8,22] and (E)-16 [23] using Negishi's protocols. The (Z)-and (E)alkenyl iodides (Z)-16 and (E)-16 were individually oxidized with manganese dioxide to the corresponding aldehydes 10 and subjectedt ot he enantioselective Mukaiyama aldol reaction developed by Kiyooka [14] using the ketene acetal 9 in the presence of l-N-tosylvaline ( 18). [24] This gave the corresponding aldols (Z)-5 (68 %y ield, 94.5:5. Having established reliable routes to both the dienylboronic ester 15 and the two alkenyl iodides (Z)-17 and (E)-17,w en ext addressed the key Suzukic oupling reaction (Scheme 3). [25] After extensive experimentation we found that the use of palladium(II)a cetate and triphenylphosphine in the presenceo f 1 m aqueouss odium bicarbonate allowed the union of the dienylboronic ester 15 with the (Z)-alkenyl iodide (Z)-17 to proceed with complete stereochemical fidelity to give the correspondingc oupled product 19 in 64 %y ield. Double deprotection of the triene 19 with HF in acetonitrile [9] gave the alcohol 20 which wasc onverted into inthomycin B 2 via aminolysis of the corresponding pentaflurorophenyl ester 21.O ur synthetic For the synthesis of inthomycin C 3,t he Suzukic oupling between the (Z,Z)-dienyl boronate 15 and the (E)-alkenyl iodide (E)-17 required furthero ptimization. Ultimately,w ef ound that the concentration of aqueous base proved crucial with the use of 0.25 m sodium bicarbonate giving the coupled product 22 in 65 %y ield. In as imilarm anner to the synthesis of inthomycin B 2,i nthomycin C 3 wasp repared from the triene 22 by the same reaction sequence. Our synthetic inthomycin C 3 had spectroscopic properties in accord with that of both natural and synthetic inthomycin C 3.
Having successfully synthesized inthomycins B 2 and C 3 we turned our attention to inthomycin A 1 (Scheme 4). We had originally aimed to preparei nthomycin A 1 by the same strategy namely Suzukic ross-coupling of a( Z,E)-dienylboronic acid (Z,E)-15,h owever,r hodium(I) catalyzed anti-selectiveh ydroboration [26] of the enyne 6 gave the corresponding (Z,E)-dienylboronic ester (Z,E)-15 in low yields (< 40 %) under an umber of conditions. We therefore alteredo ur synthetic strategy and in-vestigatedaSonogashira/semi-hydrogenation sequence. Pleasingly,t he Sonogashira reaction of the alkenyl iodide (Z)-17 with the enyne 6 proceeded smoothly under standard conditions to give the coupled product 25 in 62 %y ield. The next challenge was the semi-hydrogenation of the alkyne to give the (Z,Z,E)-triener equired for completion of the synthesis of inthomycin A. Semi-hydrogenation of 25 under av ariety of con-ditions [Pd, CaCO 3 ,q uinoline;P d, CaCO 3 ;P d, BaSO 4 ; nickel boride;Z n( Cu/Ag)] gave mixtures of the desired product, over reduced products and starting materiala nd we were unable to isolate the desired triene in synthetically useful yields. We therefore investigated the semi-hydrogenation of the alcohol 26 formed by double deprotection of 25.P leasingly the use of Zn(Cu/Ag) couple in methanola ta bove room temperature gave the desired (Z,Z,E)-triene 27 in 80 %y ield. [27] As before, the methyl ester 27 was readily transformed into the corresponding amide target inthomycin A 1 via the pentafluorophenyl ester 28.C areful analysiso ft he 1 Ha nd 13 CNMR spectra of 1 indicated that it was contaminated with a small amount (< 10 %) of inthomycin B 2 which appears to arise during conversion of the ester 27 into inthomycin A 1.
All of our synthesized inthomycins had spectroscopicp roperties in accord with the natural and previously synthesized compounds. Importantly,t he absolute configuration of inthomycin C 1 has been the subjecto fm uch confusion and debate in the literature. However,r ecently these ambiguities have been laid to rest by Hale and Hatakeyama [13b] with the absolute configurationo fi nthomycin C 1 being firmly established as (R)c onfirming the original assignment by Henkela nd Zeek. [3] We had assigned the absolute configuration of the alkenyli odides (Z)-5 and (E)-5 as (S)u sing Kakisawa's extension of Mosher's method [28] which translates into the absolute configuration of all of the inthomycins being (R), and our optical rotationf or 3 was in agreement with the recently remeasured values. [13b] In summary,wehave developed efficient modularenantioselectivet otal syntheses of all three inthomycins, which proceeds in only 9/10 steps from commercially availablem aterials. The key steps include Suzuki and Sonogashira cross-couplings, and an enantioselective Kiyooka aldol reaction. Our modularr oute has allowedt he efficient syntheses of these biologically active natural products and we will use this synthetic sequence in our assault on the synthesis of the oxazolomycins.###