DOI: 10.1002/chem.201405790

Full Paper

& Natural Products | Hot Paper|

Concise Total Syntheses of Amphidinolides C and F Galle Valot, Damien Mailhol, Christopher S. Regens, Daniel P. O’Malley, Edouard Godineau, Hiroshi Takikawa, Petra Philipps, and Alois Frstner*[a]

blocks 47 and 65 carrying the different side chains of the two target macrolides. These fragments derive from a common aldehyde precursor 46 formed by an exquisitely alkene-selective cobalt-catalyzed oxidative cyclization of the diunsaturated alcohol 44, which left an adjacent acetylene group untouched. The northern sector 29 was prepared by a two-directional Marshall propargylation strategy, whereas the highly adorned acid subunit 41 derives from d-glutamic acid by an intramolecular oxa-Michael addition and a proline-mediated hydroxyacetone aldol reaction as the key steps; the necessary Me3Sn-group on the terminus of 41 for use in the Stille coupling was installed via enol triflate 39, which was obtained by selective deprotonation/triflation of the ketone site of the precursor 38 without competing enolization of the ester also present in this particular substrate.

Abstract: The marine natural products amphidinolide C (1) and F (4) differ in their side chains but share a common macrolide core with a signature 1,4-diketone substructure. This particular motif inspired a synthesis plan predicating a late-stage formation of this non-consonant (“umpoled”) pattern by a platinum-catalyzed transannular hydroalkoxylation of a cycloalkyne precursor. This key intermediate was assembled from three building blocks (29, 41 and 47 (or 65)) by Yamaguchi esterification, Stille cross-coupling and a macrocyclization by ring-closing alkyne metathesis (RCAM). This approach illustrates the exquisite alkynophilicity of the catalysts chosen for the RCAM and alkyne hydroalkoxylation steps, which activate triple bonds with remarkable ease but left up to five other p-systems in the respective substrates intact. Interestingly, the inverse chemoselectivity pattern was exploited for the preparation of the tetrahydrofuran building

Introduction

The attraction of these targets is amplified by the dense decoration of the carbon skeleton with 11(12) stereocenters, two trans-configured THF-rings and two 1,3-diene units, one of which is in a “non-thermodynamic” exo/endo-orientation. Despite the high degree of structural homology between the individual family members, 1–4 exhibit strikingly different biological activities. The few available data suggest that 1 is exceptionally cytotoxic, with IC50 values in the single-digit

Dinoflagellates of the genus Amphidinium are a prolific source of highly bioactive and structurally complex secondary metabolites. The different strains gave rise, inter alia, to over thirty macrolides, which are generally subsumed under the family name “amphidinolides”.[1] Although certainly of polyketide origin, these compounds incorporate substructures that indicate a rather unorthodox biosynthetic machinery. This is evident, for example, from the amphidinolides C (1),[2] C2 (2),[3] C3(3)[4] and F (4),[5] which differ only in the length and/or functionalization of the individual side chains pending off a common 25-membered macrolactone ring; such odd-numbered cycles are rare and their biosynthetic origin is non-obvious.[6] The same holds true for the “umpoled” 1,4-diketone subunit at C.15/C.18 and the vicinal one-carbon branches residing at C.11/C.12; again, the formation of neither element is readily explained by the conventional logic of polyketide biosynthesis.[1, 6]

[a] Dr. G. Valot, Dr. D. Mailhol, Dr. C. S. Regens, Dr. D. P. O’Malley, Dr. E. Godineau, Dr. H. Takikawa, P. Philipps, Prof. A. Frstner Max-Planck-Institut fr Kohlenforschung 45470 Mlheim/Ruhr (Germany) E-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201405790. Chem. Eur. J. 2014, 20, 1 – 12

These are not the final page numbers! ÞÞ

1

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper ng mL1 range,[2] whereas its siblings 2–4 are up to three orders of magnitude less potent against the very same cell lines.[3–5] Such dramatic changes upon a seemingly minor modification of a single site on a non-trivial scaffold are rather unusual. For these intriguing structural attributes and their functional implications, the amphidinolides C and F drew much attention from the synthetic community;[7, 8] yet, it took more than two decades after the initial disclosure until 1 and 4 finally succumbed to total synthesis.[9, 10] Outlined below is a full account on our work in this field, including the up-scaling of our original route to amphidinolide F (4).[10, 11] This approach was found robust and flexible to the extent that the modification of a single building block brought amphidinolide C (1) into reach without the need to amend the overall strategy.

chemistry has met with great success during the last decade, late-stage applications to natural product synthesis are underrepresented in the literature.[14, 15] Our group is committed to changing this situation[16, 17] and the amphidinolides C (1) and F (4) provided an excellent opportunity to show our trust. The choice of 4 as the first target was based on the expectation that the side chain devoid of the extra doubly allylic alcohol group, which decorates the tail of the sister compound 1, is likely more robust. If successful, however, it was also clear that the project could eventually be extended to encompass the supposedly much more bioactive parent macrolide 1 as well. Of the various conceivable incarnations of such an alkyne hydration strategy, we opted for the C18 carbonyl group to be encoded as the triple bond. It seemed likely that the projected hydration would proceed regioselectively, provided it can be performed with an oxygen substituent at the transannular C15 position (path I). Upon activation of the alkyne, this O-nucleophile should lead to a clean 5-endo-dig cyclization (C!B!A), since the alternative 4-exo-dig pathway is higher in energy and therefore unlikely to compete. This plan tacitly assumes that the macrocyclic frame is sufficiently flexible to allow for the necessary quasi-linear alignment of the incoming O-nucleophile and the catalytically active metal fragment on the opposite sites of the triple bond.[12, 13] For a molecule as complex as the envisaged intermediate C, it remains difficult to predict if this stereoelectronic

Results and Discussion Strategic considerations In an attempt to reflect the peculiarities of the targets, our synthesis blueprint centered on the non-consonant (umpoled) 1,4diketone pattern at the C15 and C18 positions (Scheme 1). This peculiar motif might be formed by a directed transannular alkyne hydration in the presence of a suitable carbophilic catalyst, preferably based on platinum or gold.[12, 13] While this

Scheme 1. Retrosynthetic analysis of amphidinolide F (4); because of the convergent pathway, however, replacement of fragment G (or G’) by the analogous fragment G’’ would also bring amphidinolide C (1) into reach.

&

&

Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

2

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper condition can be met. Yet, hand-held models made us confiadvantageous attribute of this blueprint. Finally, it should be dent that such an outer-sphere process is feasible, although mentioned that the convergent nature renders this plan, if suca non-negligible risk remained for this very late-stage strategic cessfully reduced to practice, inherently flexible. Thus, replacemaneuver. ment of the tetrahydrofuran segment G by the chain-extended The proper choice of the steering oxygen nucleophile at analogue G’’ allows the sequence to be reprogrammed to C15 is equally decisive. As the targets contain a ketone at this a synthesis of amphidinolide C (1); moreover, the preparation site, recourse to a carbonyl seemed tempting. Yet, this option of non-natural analogues and derivatives should be possible.[24] was not pursued because the p-acid catalyzed addition of an Therefore it was of prime importance to ensure that the routes oxo-substituent onto an alkyne bears the risk of forming to all building blocks are scalable and the individual fragment a furan as the primary product.[18] Although furan hydrolysis coupling reactions chemically robust. with release of the signature 1,4-diketone should still be possible, epimerization of the methyl branch at the a-position (C16) Path-finding studies would necessarily ensue. Therefore we chose a hydroxyl group at C15 as the intramolecular nucleophile (see compound A); to As outlined above, two different scenarios for the implementaopt for the R-configuration at this site was an informed guess tion of the transannular alkyne hydration were conceived based on the inspection of the hand-held models, which sug(Scheme 1). However, the tantalizing outlook of linking this gested that this diastereomer might have a better chance to transformation to the formation of the flanking tetrahydrofurpopulate the trajectory necessary for alkyne hydration referred an ring (path II) was abandoned at an early stage. This decision to above. reflected the outcome of our model studies with substrate 5, The unraveling of the C18 carbonyl group by such an intrawhich was cyclized to product 6 as an unadorned mimic of the molecular alkyne hydroalkoxylation was envisaged for a commacrocyclic core of 4 (Scheme 2). Since only few RCAM pound of type C with the flanking trans-disubstituted tetrahyreactions of enynes had been reported at the outset of this drofuran ring already in place (path I). However, one might conceive an even more involved process: if alkyne hydration were to be carried out with an enyne precursor of type I, the resulting enone H would be set up for a (spontaneous) oxa-Michael addition of the C23-OH group to furnish A (path II). Although highly risky, it was planned to explore this scenario too which might allow alkyne hydration and tetrahydrofuran formation to be combined to an attractive transannular cascade.[19] Either pathway to A is contingent on the ability to provide good amounts of the required cycloalkyne precursors of type C or I by ring-closing alkyne meta- Scheme 2. Model study to evaluate the feasibility of path II as a possible gateway to amphidinolide F; a) see Table 1; b) TBAF, THF, 74 %; TBAF = tetra-n-butylammonium fluoride. thesis (RCAM).[20, 21] While this reaction is currently gaining moproject,[25, 26] it was gratifying to see that this transformation mentum, applications to substrates of this level of complexity are still scarce. The excellent application profile of the latest proceeded well, although the available catalysts showed markgeneration of catalysts, however, made us confident that the edly different performances; it is emphasized, however, that projected macrocyclizations would proceed without any major the results compiled in Table 1 were not fully optimized. The complications. Most importantly, the available evidence sugbest results were obtained with the molybdenum alkylidyne 9 gests that the catalysts would rigorously select for the alkyne endowed with silanolate ligands, which required only a low units to be engaged in ring closure, while leaving all other ploading and was fully operative at ambient temperature when systems in the substrates unaffected.[22, 23] Only under this the reaction was performed in the presence of molecular sieves to trap the released 2-butyne.[27] A similarly favorable premise can RCAM be performed with the fully elaborate carbon framework which is to be assembled by esterification yield was secured with the help of complex 10, activated on and cross coupling of three building blocks E, F and G (or G’). treatment with CH2Cl2 as previously described by our group.[28] The latent symmetry of fragment F was considered yet another However, the sensitivity of this catalyst system and the necesChem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

These are not the final page numbers! ÞÞ

3

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper tion appeared to be quantitative and the crude product was very clean (Scheme 3). PdCl2 and [AuMe(PPh3)] were also effective but required gentle heating.[36, 37] As expected, 18 is in equilibrium with the corresponding open-chain hydroxyketone; treatment of this mixture with PDC furnished the desired 1,4-diketone 19.[38] Collectively, these observations argued against the more involved approach to amphidinolide F based on a hydration/Michael addition cascade (path II), whereas the envisaged path I should have a fair chance to be ultimately successful (Scheme 1).

Table 1. RCAM reaction of the model substrate 5; preliminary catalyst screening.[a] Entry 1 2 3 4

Cat. [b]

9 10[c] 11 12

Loading [mol %)

T [8C]

t [h]

Yield [%]

5 40 30 30

20 80 80 80

1 40 40 16

90 91 76 55

[a] All reactions were carried out in toluene. [b] In the presence of MS 5 . [c] Activated with CH2Cl2.

sary elevated reaction temperature had to be accounted for by a fairly high loading. Schrock’s classical tungsten alkylidyne 11[29] also gave a respectable outcome, whereas the modified variant 12[30] thereof was distinctly less productive. Subsequent cleavage of the TBS-group furnished the macrocyclic enyne 7 that allowed a directed transannular enyne hydration to be tested.[31] Unfortunately, all attempts to form the desired enone 8 basically met with failure, although a host of palladium, platinum, mercury and gold catalysts were screened under a variety of experimental conditions.[32] Although the substrate was consumed in all but a few cases, the NMR spectra of the resulting crude materials were featureless and could not be attributed to a defined monomeric product. Additional studies with the very simple acyclic model 13 were not encouraging either (Scheme 3). Using Zeise’s dimer

The northern segment For the hidden symmetry of fragment F, we pursued a two-directional approach based on the reliable anti-propargylation reactions developed by Marshall and co-workers (Scheme 4).[39] To this end, cheap propane-1,3-diol (20) was converted on a multigram scale to aldehyde 21 by selective silylation fol-

Scheme 3. a) [PtCl2(C2H4)]2 (1 mol %), aq. [D8]THF, see text; b) and c) see text; d) PDC, DMF, 95 %. PDC = pyridinium dichromate.

Scheme 4. a) TBSCl, Et3N, CH2Cl2, 0 8C ! RT, 86 %; b) TEMPO (10 mol %), KBr, NaOCl, pH 8.6 buffer, CH2Cl2, 0 8C, 86 %; c) 22, Pd(OAc)2 (5 mol %), PPh3 (5 mol %), Et2Zn, THF, 78 8C ! 20 8C, 87 %; d) HCl (1 %, v/v) in EtOH, 91 %; e) TESOTf, 2,6-lutidine, CH2Cl2, 0 8C, quant.; f) PPTS (10 mol %), MeOH, CH2Cl2, 50 8C, 80 %; g) SO3·pyridine, iPr2NEt, DMSO, CH2Cl2, 30 8C, 93 %; h) 25, InI, PdCl2(dppf) (5 mol %), THF/HMPA, 73 %; i) TBSOTf, 2,6-lutidine, CH2Cl2, 0 8C, 91–96 %; j) PhMe2SiLi, CuCN, MeI, THF, 0 8C, 77–90 %; k) K2CO3, MeOH, 40 8C, 84–88 %; l) nBuLi, MeI, THF, 78 8C!RT, 97–99 %; m) NIS, MeCN, benzene, 0 8C, 84–88 %; n) i) CSA cat., MeOH, CH2Cl2 ; ii) 2,2-dimethoxypropane, CSA cat., CH2Cl2, 57 % (over both steps); CSA = ()-camphor-10-sulfonic acid; dppf = 1,1’-bis(diphenylphosphino)ferrocene; Ms = methanesulfonyl; NIS = Niodosuccinimide; PPTS = pyridinium p-toluenesulfonate; TBS = tert-butyldimethylsilyl; TES = triethylsilyl; TEMPO = 2,2,6,6-tetramethyl-1-piperidinyloxy radical; Tf = trifluoromethansulfonyl; TMS = trimethylsilyl.

as the catalyst under buffered conditions,[33] it was possible to confirm the rapid initial formation of the corresponding enol ether 14 by NMR spectroscopy, although the product decomposed with time. Attempted hydrolysis of 14 was to no avail and caused rapid degradation. In striking contrast, the hydration of the regular alkyne 16 worked remarkably well using either [Pd(OCOCF3)2], HgOTf2,[34] PtCl2 or Zeise’s dimer as the catalyst;[33, 35] in all cases, the reac&

&

Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

4

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper lowed by a TEMPO-catalyzed oxidation using sodium hypochlorite as the ultimate oxidant.[40] A palladium-catalyzed, zincinduced Marshall propargylation furnished sizeable amounts of product 23 with respectable selectivity (d.r. = 90:10).[41] After elaboration into aldehyde 24 by standard protecting group and oxidation-state management, a second propargylation reaction extended the chain in the other direction while installing the two remaining stereocenters. This transformation was best achieved by the indium variant of the Marshall coupling using mesylate 25 devoid of any substituent on the alkyne[42] to ensure that the termini of the resulting product 26 can be selectively addressed later on. The spectral fingerprints of the isopropylidene acetal 30 derived from 26 confirmed the 1,3syn configuration of the major isomer.[43] Moreover, the minor diastereomers could be largely removed at this stage by conventional flash chromatography. After protection of the C13 hydroxyl group in 26 as TBSether to ensure orthogonality to the TES-ether at C15, the terminal alkyne was subjected to an exquisitely selective silylcupration reaction terminated by a MeI-quench.[44] Treatment of the resulting product 27 with basic methanol led to the selective removal of the alkynyl TMS group while leaving the alkenylsilane as well as the two different silyl ethers untouched. The released acetylene moiety was C-alkylated with MeI prior to subjecting 28 to an iodine-for-silicon exchange to form the alkenyl iodide 29 as a fully functional surrogate of synthon F. However, this product is light sensitive and was therefore prepared only on demand. This practical issue notwithstanding, the entire sequence proved scalable; overall, more than 3 grams of alkenylsilane 28 as the storable endpoint were prepared during this endeavor.

Scheme 5. a) TrCl, pyridine, 91 %; b) i) LDA, MeI, THF, 78 8C!30 8C; ii) LDA, THF, 78 8C, aq. workup, 93 % (over both steps); c) Dibal-H, CH2Cl2, 78 8C; d) Ph3P=CHCOOEt, toluene, 80 8C, 80–85 % (over both steps); e) TBAF·3 H2O, THF, 0 8C, 82–87 %; f) TFA, EtOH, CH2Cl2, 0 8C, 75–93 %; g) SO3·pyridine, iPr2NEt, DMSO, CH2Cl2, 0 8C; h) 36, l-proline (50 mol %), DMF, 63–66 % (over both steps); i) TBSOTf, pyridine, CH2Cl2, 0 8C!RT, 91–95 %; j) KHMDS, PhNTf2, THF, 78 8C, 71–73 %; k) [Pd2(dba)3] (15 mol %), P(2-furyl)3 (60 mol %), Me6Sn2, LiCl, THF, 80 %; l) KOH, THF, EtOH, H2O, 45 8C, 98 %; m) i) HF·pyridine, THF/pyridine; ii) 2,2-dimethoxypropane, TsOH, 67 % (over both operations); dba = dibenzylideneacetone; Dibal-H = diisobutylaluminum hydride; KHMDS = potassium hexamethyldisilazide; LDA = lithium diisopropylamide; TFA = trifluoroacetic acid; Tr = trityl; Ts = p-toluenesulfonyl.

To this end, ketoester 38 was subjected to a chemo- and regioselective deprotonation followed by trapping of the resulting enolate with PhNTf2. Obviously, the bulk of the chosen KHMDS base and the differential in the acidity of the two different carbonyl groups synergized in that deprotonation of 38 at the methyl terminus was largely favored. After some optimization, this delicate transformation was well reproducible and furnished alkenyl triflate 39 in a respectable 71 % yield on a 1.2 gram scale (single largest batch).[51] The subsequent conversion into the corresponding stannane 40 was best performed with a catalyst formed in situ from [Pd2(dba)3] and tris-2-furylphosphine.[52] Final saponification of the ester completed the preparation of the required acid building block 41. Although the 13 steps of this approach open the longest linear sequence, they met our needs and allowed about 5 grams of 41 to be prepared during the course of the project.

The acid sector Elaboration of the commercial lactone 31[45] into compound 34 was carried out on a multigram scale by adapting a literature route (Scheme 5).[2c, 46] Key to success was a TBAF-mediated oxa-Michael addition to form the 1,4-trans-configured tetrahydrofuran ring.[47] Alcohol 34 was then converted to the corresponding aldehyde 35, which proved unexpectedly sensitive and therefore difficult to handle. However, reliable results were obtained under Parikh–Doering conditions,[48] followed by chromatographic purification of the crude material over silica deactivated with triethylamine. Aldehyde 35 was subjected to a proline catalyzed reaction with the hydroxyacetone derivative 36,[49] which necessitated a high loading but furnished the required anti-aldol product 37 in acceptable yield, along with some minor but removable byproducts. Mosher-ester analysis[50] showed that the stereochemistry of the newly formed alcohol center in 37 matched that of the corresponding site in the amphidinolide target. Moreover, the cyclic acetal 42 was prepared to corroborate the 1,2-anti diol relationship at C.7 and C.8 by NOE studies. However, this acetal turned out to be labile; therefore, it seemed prudent to use the corresponding bisTBS-ether 38 for the further elaboration of the building block. Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

These are not the final page numbers! ÞÞ

The side chain sector An oxidative Mukaiyama cyclization reaction opened a concise entry into the tetrahydrofuran sector G.[53, 54] Our foray is distinguished by the fact that the chosen cyclization precursor 44, which is readily available from the enantiopure epoxide 43,[55] comprises two different p-systems amenable to oxidation (Scheme 6). Yet, the cobalt-catalyzed reaction proved exquisitely chemoselective in that only the alkene site of 44 reacted 5

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper promotes the reaction by transmetallation of the stannane donor to an arguably more reactive organocopper species, the phosphinite scavenges the released tin byproducts and hence renders this transmetallation irreversible. As both reagents are hardly basic and the conditions are fluoride free, this method has been widely used by us and others for the preparation of polyfunctionalized and/or polyunsaturated compounds.[61, 64, 65] It has also outperformed the venerable Suzuki–Miyaura coupling on more than one occasion.[66, 67] Despite its good track record, this procedure allowed the exigent Stille coupling of 29 with 49 to be accomplished with only 56 % yield. Efforts to improve on this outcome by varying the loadings and the ratio of the individual components were to no avail. This optimization exercise was rendered difficult by the fact that the crude product always seemed fairly clean by TLC analysis and NMR spectroscopy. In any case, the result itself was well reproducible in over 10 runs on different scales, thus furnishing close to 1 gram of diyne 50 in readiness for ring closure. The crucial macrocyclization was achieved on exposure of 50 to 10/CH2Cl2[28] in toluene at elevated temperature, although the yield was somewhat variable (50–73 %) depending on the scale of the reaction. In contrast, the molybdenum alkylidyne 9[27] was less effective because it gave substantial amounts of an acyclic dimer as a byproduct. This outcome was unexpected because 9 had proven superior to 10/CH2Cl2 in several challenging cases before (see also the model study shown above in Scheme 2).[21a] However, this particular catalyst seems to find its limitations with sterically hindered substrates that might not be able to bind to the operative molybdenum alkylidyne unit surrounded by three bulky triphenylsilanolate ligands.[67] To probe this aspect, the TES-group flanking one of the alkynes in 50 was removed and the resulting slimmer diyne 52 subjected to ring closure. In line with our expectations, complex 9 now allowed product 53 to be formed in up to 70 % yield, even though a high temperature was necessary for the cyclization to proceed. From the organometallic viewpoint, this outcome is remarkable: all alkyne metathesis catalysts known to date are high-valent Schrock alkylidynes of early transition metals and as such inherently nucleophilic at the a-carbon atom.[68] The compatibility of 9 with a protic functional group is therefore noteworthy and has hardly any precedent in the literature, except for a few cases from our laboratory.[67, 69, 70] The NMR spectra of both cycloalkynes 51 and 53 feature massive line broadening over the entire temperature range accessible to our spectrometer. At 20 8C or below, (at least) two slowly interconverting conformers seem to be populated, but no details could be deduced. Therefore it remained open whether or not the C15OH group and the alkyne moiety would be appropriately disposed for the projected transannular hydration reaction. It required a short screening to reduce this key transformation to practice. Although Hg(OTf)2 and PdCl2 had been successful in our early model studies (see above), the application of these catalysts to 53 resulted in decomposition of the precious material. In contrast, PtCl2 in Et2O/H2O seemed promising

Scheme 6. a) Propyne, nBuLi, BF3·Et2O, THF, 78 8C, 87 %; b) (nmp)2Co (10 mol %), tBuOOH (10 mol %), O2 (1 atm), iPrOH, 55 8C, 84 %; c) SO3·pyridine, iPr2NEt, DMSO, CH2Cl2, 0 8C, 86–89 %; d) (i) 48, tBuLi, Et2O, 78 8C, then ZnBr2, Et2O, 35 8C ! 0 8C; (ii) ()-N-methylephedrine/nBuLi, toluene, 0 8C; (iii) 46, Et2O, 20 8C, 80–85 % (dr = 95:5).

while the alkyne remained untouched, thus furnishing the trans-disubstituted tetrahydrofuran 45 as the only detectable isomer in consistently high yield. With regard to the catalyst, best results were achieved with the “second generation” nmp ligand for cobalt introduced by Pagenkopf and co-workers.[8e,l] With ample 45 in hand, the fragment synthesis was completed by oxidation to the corresponding aldehyde followed by an N-methylephedrine-assisted[56] addition of the dienylzinc derivative derived from 48.[57] However, this bromide turned out to be surprisingly unstable and had to be prepared immediately prior to use. Except for this caveat, the route to alcohol 47 is deemed satisfactory in that it is concise, selective and high yielding. Moreover, addition of nucleophiles other than the zinc reagent derived from 48 allow the side chain to be modified with ease, which seems to account for much of the bioactivity of the amphidinolides of this particular series (see the Introduction section). This aspect is underlined by our approach to amphidinolide C (1) described below. The assembly phase and completion of the total synthesis of amphidinolide F Although only few steps remained to be accomplished at this point, all strategic maneuvers were still ahead of us, including the challenging ring-closing alkyne metathesis (RCAM) and the largely unprecedented transannular alkyne hydration as the cornerstones of the synthesis plan. The critical assembly phase started with a Yamaguchi esterification of acid 41 and alcohol 47 to give product 49 without incident (Scheme 7).[58] Next, a Stille coupling with iodide 29 was planned to set the conspicuous exo/endo diene motif.[59] We were apprehensive that this subunit might be fragile and isomerization-prone; yet, our group had previously developed a modified Stille protocol for particularly sensitive compounds that was thought to meet the challenge.[60, 61] In addition to the mandatory palladium catalyst, this method uses a combination of copper thiophene-2-carboxylate (CuTC)[62] and [Ph2PO2][NBu4][63] as additives; while the former &

&

Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

6

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper

Scheme 7. a) 2,4,6-Trichlorobenzoyl chloride, Et3N, DMAP, toluene, 72–80 %; b) 29, Pd(PPh3)4 (30 mol %), Ph2PO2NBu4, CuTC, DMF, 56 %; c) 10 (30 mol %), toluene, CH2Cl2, 60 8C, 50–73 %; d) PPTS (30 mol %), MeOH, CH2Cl2, 0 8C, 87–92 % (53), 89–95 % (52); e) 9 (25 mol %), toluene, 100 8C, 60–70 %; f) [(C2H4)PtCl2]2 (0.2 mol %), Et2O, quant.; g) PPTS, benzene, 98 %; h) TPAP, CH2Cl2, 70 %; i) HF·NEt3, Et3N, MeCN, 40 8C, 60 %; CuTC = copper thiophene-2-carboxylate; DMAP = 4dimethylaminopyridine; TPAP = tetra-n-propylammonium perruthenate.

exo-dig pathway were noticed at the enol or the ketone stage. In any case, the remarkable ease and efficiency of this reaction imply that the conformation of 53 is actually supportive of the transannular hydroalkoxylation event. Moreover, we suppose that this transformation owes its success to the pronounced alkynophilicity of PtII, which leaves the two different 1,3-dienes unaffected; in constrast, HgII as well as PdII are less discriminative and have, therefore, failed with this elaborate polyunsaturated compound. In our model study we had used PDC to oxidize the equilibrating hydroxyketone/hemiacetal mixture to the corresponding 1,4-diketone (see Scheme 3). With the real substrate 55, however, the basic character of this reagent led to partial epimerization, most likely of the a-methyl group at C16. This complication was circumvented when TPAP was used instead; we opted to employ this oxidant as a stoichiometric reagent to avoid any hassle.[72, 73] The final deprotection of diketone 56 turned out to be more difficult than anticipated. While a pre-screening showed that the standard reagents commonly used for the cleavage of TBS-

in that the desired ketone 55 could be detected, together with a byproduct of unknown structure. Since bare PtCl2 is oligomeric and only partly soluble in the medium, it was replaced by Zeise’s dimer [PtCl2(C2H4)]2.[33] Moreover, we soon recognized that it was advantageous to decouple the transannular hydroalkoxylation from the subsequent hydrolysis of the resulting enol ether, which turned out to be surprisingly delicate and proceeded without extensive decomposition only in wet benzene in the presence of pyridinium p-toluenesulfonate (PPTS).[71] Under this proviso, it was possible to harness the potential of the transannular functionalization process. With dry ether as the solvent and a loading of [(C2H4)PtCl2]2 as low as 0.2 mol %, the conversion of 53 into the corresponding enol ether 54 proceeded quantitatively within minutes at ambient temperature. Evaporation of the solvent followed by hydrolysis of crude 54 with PPTS in wet benzene furnished the desired product 55 (in equilibrium with the corresponding hemiacetal); in all runs the yield was > 95 % over both operations, with the single largest batch furnishing 235 mg of product. As expected, no signs of another regioisomer formed by a competing 4Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

These are not the final page numbers! ÞÞ

7

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper ethers endangered the integrity of the product, the commercial adduct [3 HF·Et3N] in MeCN, supplemented by extra Et3N, was sufficiently mild but unbearably slow; up to three weeks were necessary to achieve complete conversion and the isolated yield of only approximately 50 % is indicative of some degradation of the product occurring in the background.[74] A good compromise was found by heating the mixture to 40 8C. Under these conditions, the global deprotection was complete in 7 days and the yield of amphidinolide F (4) eventually reached 60 %.[75] The total synthesis of amphidinolide F (4) outlined above comprises a longest linear sequence of 21 steps (39 steps overall), which is deemed appropriate for a target of this size and complexity. It compares favorably to the only other total synthesis reported in the literature, which proceeds over a longest linear sequence of 36 steps and has a total step count of 62.[9]. Our route allowed us to obtain approximately 200 mg of the penultimate product 56 as our preferred store option, from which we chose Scheme 8. a) (i) nBuLi, Et2O, 78 8C; (ii) ZnBr2 (0.5 equiv), then distillation; b) 59, ()-MIB to release approximately 50 mg of 4. With all steps (10 mol %), toluene, 30 8C, 97 %, er = 92.5:7.5; c) TBSCl, imidazole, DMAP (20 mol %), CH2Cl2, 0 8C ! RT, 98 %; d) Dibal-H, THF, 78 8C, 95 %; e) (i) NBS, PPh3, THF, 20 8C ! now being optimized, a further up-scaling of the se- 0 8C; (ii) PhSO Na, TBAI, THF, 0 8C ! RT, 91 %; f) (i) LDA, CH Br , THF, 95 8C; (ii) 62, nBuLi, 2 2 2 quence is certainly possible without undue efforts. THF, 78 8C, 90 % (E:Z = 4:1), 35 % (pure E, after HPLC, see Text); g) (i) tBuLi, Et2O, 78 8C, Perhaps more interesting than these metrics is the then ZnBr2, 35 8C; (ii) ()-N-methylephedrine/nBuLi, toluene, 0 8C; (iii) 46, Et2O, 20 8C, conclusion that the key transformations (RCAM, 67 % (dr = 4.2:1); NBS = N-bromosuccinimide; TBAI = tetra-n-butylammonium iodide. transannular alkyne hydration), which seemed risky at the outset of the project, proceeded exceedingly well, whereas the ostensibly routine Stille cross-coupling perative to use a strictly salt-free zinc reagent,[78] which was enturned out to be challenging. A number of arguably similar resured by distillation under high vacuum. Reduction of 60 folactions had previously worked without incident;[61] the fact lowed by one-pot conversion of the resulting primary alcohol 61 into sulfone 62 set the stage for the elaboration of the 1that we are not sure which subtle factors make all the differbromo-1,3-diene terminus in 63. To this end, the organolithium ence reminds us of our still limited ability to extrapolate from species derived from 62 was reacted with the carbenoid genera recorded to a projected case, even if a transformation is ated in situ from CH2Br2 and LDA.[57] Provided that the tempermechanistically as well understood as the Stille coupling.[59] This notion is highlighted by the somewhat counterintuitive ature was carefully controlled, this reaction was high yielding result that the analogous step en route to the sister compound but product 63 was formed as a mixture of E/Z-isomers. Deamphidinolide C (1) was considerably higher yielding despite spite of the doubly allylic OTBS substituent this compound the use of an arguably more delicate coupling partner (see was sufficiently stable for HPLC separation, whereas the analobelow). gous bromo-diene 48 used en route to 4 had been too fragile (see Scheme 6); however, the isolation of pure E,E-63 still resulted in some material loss. Total synthesis of amphidinolide C The subsequent metalation was challenging because of posThe brevity of our convergent route to amphidinolide F (4) ensible elimination of the vinylogous OTBS in the intermediate couraged us to embark on a foray towards amphidinolide C (1) of type 64. Strict temperature control during metal/halogen as the parent compound of this panel of marine macrolides.[2] exchange with tBuLi and the subsequent transmetalation, however, afforded the appropriate organozinc species for All it takes is to supplant fragment G by the analogous buildasymmetric addition to aldehyde 46, which proceeded with reing block G’’ endowed with the appropriately extended side spectable yield and diastereoselectivity in favor of the desired chain (see Scheme 1). Because it is decorated with an alcohol product 65. in a doubly allylic position, it remained to be seen whether The assembly phase commenced with an uneventful Yamathis group withstands all conditions passed through during guchi esterification with acid 41 followed by the critical Stille the assembly of the target. reaction with iodide 29 (Scheme 9). In view of the rather The required building block was prepared by asymmetric admodest yield of 56 % of the analogous step en route to amphidition of the bis-alkenylzinc reagent 58 derived from iodide dinolide F (4), we were concerned that the allylic alcohol motif 57[76] to the commercial aldehyde 59, which occurred with on the side chain of the new coupling partner 66 might furgood yield and respectable induction under the aegis of ()ther complicate matters. Gratifyingly though, the opposite was MIB as the chiral ligand (Scheme 8).[77] As expected, it was im&

&

Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

8

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper supply. Ongoing work in this laboratory intends to further leverage the merits of catalytic alkyne chemistry.

Experimental Section All experimental details can be found in the Supporting Information. The material includes compound characterization and copies of spectra of new compounds.

Acknowledgements Scheme 9. a) 2,4,6-Trichlorobenzoyl chloride, Et3N, DMAP, toluene, 91 %; b) 29, Pd(PPh3)4 (30 mol %), Ph2PO2NBu4, CuTC, DMF, 77 %; c) 10 (30 mol %), toluene, CH2Cl2, 60 8C, 70 %; d) PPTS (20 mol %), MeOH, CH2Cl2, 0 8C, 88 %; e) [(C2H4)PtCl2]2 (0.2 mol %), Et2O; f) PPTS, benzene, 91 % (over both steps); g) TPAP, CH2Cl2, 65 %; h) HF·NEt3, Et3N, MeCN, 40 8C, 66 %.

Generous financial support by the MPG, the Alexander-vonHumboldt Foundation (fellowship to D.P.O’M.), JSPS (fellowship to H. T.) and the Fonds der Chemischen Industrie is gratefully acknowledged. We thank the analytical departments of our Institute for invaluable support.

true: the coupling was surprisingly uneventful, furnishing product 67 in well-reproducible 77 % yield. The macrocyclization of this diyne and the subsequent selective cleavage of the TESether in the resulting cycloalkyne 68 were also straightforward and productive. As in the previous case, line broadening rendered the NMR spectra unhandsome, although the integrity of the compound could be deduced beyond doubt. Most important was the fact that the transannular alkyne hydration catalyzed by as little as 0.2 mol % of Zeise’s dimer in Et2O once again worked with exceptional ease, requiring no more than 15 min reaction time for complete conversion. After hydrolysis of the resulting enol ether under the previously optimized conditions, the required hydroxy-ketone 69 (in equilibrium with the hemiketal) was isolated in 91 % yield. In view of the six different p-bonds present in substrate 68, it is the chemoselectivity of this platinum-catalyzed transformation which is particularly noteworthy. Oxidation of 69 with TPAP followed by global deprotection with Et3N·HF as described above completed the total synthesis of amphidinolide C (1).

Keywords: alkyne metathesis · molybdenum products · platinum · total synthesis

Before the turn of the millennium, neither p-acid catalysis nor alkyne metathesis played any significant role in advanced organic chemistry and natural product synthesis. This situation is currently changing; the conquests of amphidinolide C (1) and F (4) outlined above bear witness for the fact that both methods have reached a level of maturity that allows them to be implemented with confidence even into the late stages of rather challenging endeavors.[14, 21] Moreover, it becomes increasingly clear that ample synergy can be drawn from these orthogonal ways of making and/or manipulating triple bonds, when properly synchronized.[17, 79] This strategic dividend rendered our syntheses of 1 and 4 very competitive, not least in terms of an auspicious step count and a generous material www.chemeurj.org

These are not the final page numbers! ÞÞ

natural

[1] a) J. Kobayashi, T. Kubota, J. Nat. Prod. 2007, 70, 451 – 460; b) J. Kobayashi, M. Tsuda, Nat. Prod. Rep. 2004, 21, 77 – 93. [2] a) J. Kobayashi, M. Ishibashi, M. R. Wlchli, H. Nakamura, Y. Hirata, T. Sasaki, Y. Ohizumi, J. Am. Chem. Soc. 1988, 110, 490 – 494; b) T. Kubota, M. Tsuda, J. Kobayashi, Org. Lett. 2001, 3, 1363 – 1366; c) T. Kubota, M. Tsuda, J. Kobayashi, Tetrahedron 2003, 59, 1613 – 1625. [3] T. Kubota, Y. Sakuma, M. Tsuda, J. Kobayashi, Mar. Drugs 2004, 2, 83 – 87. [4] T. Kubota, A. Suzuki, M. Yamada, S. Baba, J. Kobayashi, Heterocycles 2010, 82, 333 – 338. [5] J. Kobayashi, M. Tsuda, M. Ishibashi, H. Shigemori, T. Yamasu, H. Hirota, T. Sasaki, J. Antibiot. 1991, 44, 1259 – 1261. [6] Macrolide Antibiotics. Chemistry, Biology, and Practice (Ed.: S. O¯mura), 2nd ed., Academic Press, San Diego, 2002. [7] A. Frstner, Isr. J. Chem. 2011, 51, 329 – 345. [8] For studies towards 1–4, see: a) H. Ishiyama, M. Ishibashi, J. Kobayashi, Chem. Pharm. Bull. 1996, 44, 1819 – 1822; b) J. B. Shotwell, W. R. Roush, Org. Lett. 2004, 6, 3865 – 3868; c) D. K. Mohapatra, H. Rahaman, M. S. Chorghade, M. K. Gurjar, Synlett 2007, 0567 – 0570; d) R. H. Bates, J. B. Shotwell, W. R. Roush, Org. Lett. 2008, 10, 4343 – 4346; e) C. Palmer, N. A. Morra, A. C. Stevens, B. Bajtos, B. P. Machin, B. L. Pagenkopf, Org. Lett. 2009, 11, 5614 – 5617; f) A. Armstrong, C. Pyrkotis, Tetrahedron Lett. 2009, 50, 3325 – 3328; g) D. K. Mohapatra, P. Dasari, H. Rahaman, R. Pal, Tetrahedron Lett. 2009, 50, 6276 – 6279; h) S. Mahapatra, R. G. Carter, Org. Biomol. Chem. 2009, 7, 4582 – 4585; i) M. P. Paudyal, N. P. Rath, C. D. Spilling, Org. Lett. 2010, 12, 2954 – 2957; j) L. Ferri, B. Figadre, Org. Lett. 2010, 12, 4976 – 4979; k) S. Roy, C. D. Spilling, Org. Lett. 2010, 12, 5326 – 5329; l) N. A. Morra, B. L. Pagenkopf, Eur. J. Org. Chem. 2013, 756 – 760; m) J. S. Clark, G. Yang, A. P. Osnowski, Org. Lett. 2013, 15, 1460 – 1463; n) J. S. Clark, G. Yang, A. P. Osnowski, Org. Lett. 2013, 15, 1464 – 1467; o) N. A. Morra, B. L. Pagenkopf, Tetrahedron 2013, 69, 8632 – 8644; p) D. Wu, C. J. Forsyth, Org. Lett. 2013, 15, 1178 – 1181. [9] a) S. Mahapatra, R. G. Carter, J. Am. Chem. Soc. 2013, 135, 10792 – 10803; b) S. Mahapatra, R. G. Carter, Angew. Chem. Int. Ed. 2012, 51, 7948 – 7951; Angew. Chem. 2012, 124, 8072 – 8075. [10] G. Valot, C. S. Regens, D. P. O’Malley, E. Godineau, H. Takikawa, A. Frstner, Angew. Chem. Int. Ed. 2013, 52, 9534 – 9538; Angew. Chem. 2013, 125, 9713 – 9717.

Conclusion

Chem. Eur. J. 2014, 20, 1 – 12

·

9

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper [11] For total syntheses of other amphidinolides from our laboratory, see ref. [61] and the following: a) A. Frstner, C. Aı¨ssa, R. Riveiros, J. Ragot, Angew. Chem. Int. Ed. 2002, 41, 4763 – 4766; Angew. Chem. 2002, 114, 4958 – 4960; b) C. Aı¨ssa, R. Riveiros, J. Ragot, A. Frstner, J. Am. Chem. Soc. 2003, 125, 15512 – 15520; c) O. Lepage, E. Kattnig, A. Frstner, J. Am. Chem. Soc. 2004, 126, 15970 – 15971; d) A. Frstner, E. Kattnig, O. Lepage, J. Am. Chem. Soc. 2006, 128, 9194 – 9204; e) A. Frstner, O. Larionov, S. Flgge, Angew. Chem. Int. Ed. 2007, 46, 5545 – 5548; Angew. Chem. 2007, 119, 5641 – 5644. [12] a) D. J. Gorin, F. D. Toste, Nature 2007, 446, 395 – 403; b) E. JimnezNfflÇez, A. M. Echavarren, Chem. Rev. 2008, 108, 3326 – 3350; c) A. S. K. Hashmi, Chem. Rev. 2007, 107, 3180 – 3211. [13] a) A. Frstner, P. W. Davies, Angew. Chem. Int. Ed. 2007, 46, 3410 – 3449; Angew. Chem. 2007, 119, 3478 – 3519; b) A. Frstner, Chem. Soc. Rev. 2009, 38, 3208 – 3221. [14] A. Frstner, Acc. Chem. Res. 2014, 47, 925 – 938. [15] Y. Zhang, T. Luo, Z. Yang, Nat. Prod. Rep. 2014, 31, 489 – 503. [16] For early examples, see: a) A. Frstner, H. Szillat, B. Gabor, R. Mynott, J. Am. Chem. Soc. 1998, 120, 8305 – 8314; b) V. Mamane, P. Hannen, A. Frstner, Chem. Eur. J. 2004, 10, 4556 – 4575; c) A. Frstner, P. Hannen, Chem. Eur. J. 2006, 12, 3006 – 3019; d) A. Frstner, A. Schlecker, Chem. Eur. J. 2008, 14, 9181 – 9191. [17] a) W. Chaładaj, M. Corbet, A. Frstner, Angew. Chem. Int. Ed. 2012, 51, 6929 – 6933; Angew. Chem. 2012, 124, 7035 – 7039; b) S. Benson, M.-P. Collin, A. Arlt, B. Gabor, R. Goddard, A. Frstner, Angew. Chem. Int. Ed. 2011, 50, 8739 – 8744; Angew. Chem. 2011, 123, 8898 – 8903; c) L. Hoffmeister, P. Persich, A. Frstner, Chem. Eur. J. 2014, 20, 4396 – 4402; d) L. Brewitz, J. Llaveria, A. Yada, A. Frstner, Chem. Eur. J. 2013, 19, 4532 – 4537. [18] M. Rudolph, A. S. K. Hashmi, Chem. Commun. 2011, 47, 6536 – 6544. [19] For a related scenario, see: H. H. Jung, P. E. Floreancig, Org. Lett. 2006, 8, 1949 – 1951. [20] a) A. Frstner, G. Seidel, Angew. Chem. Int. Ed. 1998, 37, 1734 – 1736; Angew. Chem. 1998, 110, 1758 – 1760; b) A. Frstner, O. Guth, A. Rumbo, G. Seidel, J. Am. Chem. Soc. 1999, 121, 11108 – 11113. [21] a) A. Frstner, Angew. Chem. Int. Ed. 2013, 52, 2794 – 2819; Angew. Chem. 2013, 125, 2860 – 2887; b) A. Frstner, P. W. Davies, Chem. Commun. 2005, 2307 – 2320. [22] See ref. [17a] and the following: a) K. Lehr, R. Mariz, L. Leseurre, B. Gabor, A. Frstner, Angew. Chem. Int. Ed. 2011, 50, 11373 – 11377; Angew. Chem. 2011, 123, 11575 – 11579; b) V. Hickmann, A. Kondoh, B. Gabor, M. Alcarazo, A. Frstner, J. Am. Chem. Soc. 2011, 133, 13471 – 13480; c) A. Guy, C. Oger, J. Heppekausen, C. Signorini, C. De Felice, A. Frstner, T. Durand, J.-M. Galano, Chem. Eur. J. 2014, 20, 6374 – 6380; J.M. Galano, Chem. Eur. J. 2014, 20, 6374 – 6380. [23] a) A. Frstner, K. Grela, C. Mathes, C. W. Lehmann, J. Am. Chem. Soc. 2000, 122, 11799 – 11805; b) A. Frstner, T. Dierkes, Org. Lett. 2000, 2, 2463 – 2465; c) A. Frstner, C. Mathes, K. Grela, Chem. Commun. 2001, 1057 – 1059. [24] For synthesis-driven explorations of the chemical space surrounding select members of the amphidinolide family, see: a) A. Frstner, E. Kattnig, G. Kelter, H.-H. Fiebig, Chem. Eur. J. 2009, 15, 4030 – 4043; b) A. Frstner, S. Flgge, O. Larionov, Y. Takahashi, T. Kubota, J. Kobayashi, Chem. Eur. J. 2009, 15, 4011 – 4029. [25] a) A. Frstner, L. Turet, Angew. Chem. Int. Ed. 2005, 44, 3462 – 3466; Angew. Chem. 2005, 117, 3528 – 3532; b) A. Frstner, D. De Souza, L. Turet, M. D. B. Fenster, L. Parra-Rapado, C. Wirtz, R. Mynott, C. W. Lehmann, Chem. Eur. J. 2007, 13, 115 – 134. [26] For recent examples, see: a) J. Willwacher, A. Frstner, Angew. Chem. Int. Ed. 2014, 53, 4217 – 4221; Angew. Chem. 2014, 126, 4301 – 4305; b) K. Micoine, A. Frstner, J. Am. Chem. Soc. 2010, 132, 14064 – 14066; c) K. Micoine, P. Persich, J. Llaveria, M.-H. Lam, A. Maderna, F. Loganzo, A. Frstner, Chem. Eur. J. 2013, 19, 7370 – 7383. [27] a) J. Heppekausen, R. Stade, A. Kondoh, G. Seidel, R. Goddard, A. Frstner, Chem. Eur. J. 2012, 18, 10281 – 10299; b) J. Heppekausen, R. Stade, R. Goddard, A. Frstner, J. Am. Chem. Soc. 2010, 132, 11045 – 11057. [28] a) A. Frstner, C. Mathes, C. W. Lehmann, J. Am. Chem. Soc. 1999, 121, 9453 – 9454; b) A. Frstner, C. Mathes, C. W. Lehmann, Chem. Eur. J. 2001, 7, 5299 – 5317; c) A. Frstner, K. Radkowski, J. Grabowski, C. Wirtz, R. Mynott, J. Org. Chem. 2000, 65, 8758 – 8762.

&

&

Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

[29] a) J. H. Wengrovius, J. Sancho, R. R. Schrock, J. Am. Chem. Soc. 1981, 103, 3932 – 3934; b) M. L. Listemann, R. R. Schrock, Organometallics 1985, 4, 74 – 83. [30] a) S. Beer, K. Brandhorst, C. G. Hrib, X. Wu, B. Haberlag, J. Grunenberg, P. G. Jones, M. Tamm, Organometallics 2009, 28, 1534 – 1545; b) S. Beer, C. G. Hrib, P. G. Jones, K. Brandhorst, J. Grunenberg, M. Tamm, Angew. Chem. Int. Ed. 2007, 46, 8890 – 8894; Angew. Chem. 2007, 119, 9047 – 9051. [31] For a pertinent case, see: G. Jalce, X. Franck, B. Seon-Meniel, R. Hocquemiller, B. Figadre, Tetrahedron Lett. 2006, 47, 5905 – 5908. [32] For successful enone synthesis via palladium catalyzed hydration of enynes, see: A. Arcadi, S. Cacchi, F. Marinelli, Tetrahedron 1993, 49, 4955 – 4964. [33] B. Liu, J. K. De Brabander, Org. Lett. 2006, 8, 4907 – 4910. [34] a) M. Nishizawa, T. Takemoto, I. Sasaki, M. Nakano, E. Ho, K. Namba, H. Yamamoto, H. Imagawa, Synlett 2009, 1175 – 1179; b) K. Ravindar, M. S. Reddy, P. Deslongchamps, Org. Lett. 2011, 13, 3178 – 3181; c) see also: G. Stork, R. Borch, J. Am. Chem. Soc. 1964, 86, 935 – 936. [35] a) P. W. Jennings, J. W. Hartman, W. C. Hiscox, Inorg. Chim. Acta 1994, 222, 317 – 322; b) J. W. Hartman, L. Sperry, Tetrahedron Lett. 2004, 45, 3787 – 3788; c) L. W. Francisco, D. A. Moreno, J. D. Atwood, Organometallics 2001, 20, 4237 – 4245. [36] a) K. Imi, K. Imai, K. Utimoto, Tetrahedron Lett. 1987, 28, 3127 – 3130; b) K. Utimoto, Pure Appl. Chem. 1983, 55, 1845 – 1852. [37] a) E. Mizushima, K. Sato, T. Hayashi, M. Tanaka, Angew. Chem. Int. Ed. 2002, 41, 4563 – 4565; Angew. Chem. 2002, 114, 4745 – 4747; see also: b) Y. Fukuda, K. Utimoto, J. Org. Chem. 1991, 56, 3729 – 3731; c) Y. Fukuda, K. Utimoto, Bull. Chem. Soc. Jpn. 1991, 64, 2013 – 2015; d) J. H. Teles, S. Brode, M. Chabanas, Angew. Chem. Int. Ed. 1998, 37, 1415 – 1418; Angew. Chem. 1998, 110, 1475 – 1478; e) N. Marion, R. S. Ram n, S. P. Nolan, J. Am. Chem. Soc. 2009, 131, 448 – 449; f) A. Leyva, A. Corma, J. Org. Chem. 2009, 74, 2067 – 2074. [38] Y.-C. Lin, C.-H. Lin, C.-Y. Chen, S.-S. Sun, B. Pal, Org. Biomol. Chem. 2011, 9, 4507 – 4517. [39] J. A. Marshall, J. Org. Chem. 2007, 72, 8153 – 8166. [40] a) P. Lucio Anelli, C. Biffi, F. Montanari, S. Quici, J. Org. Chem. 1987, 52, 2559 – 2562; b) A. E. J. de Nooy, A. C. Besemer, H. van Bekkum, Synthesis 1996, 1153 – 1174. [41] a) J. A. Marshall, N. D. Adams, J. Org. Chem. 1999, 64, 5201 – 5204; b) J. A. Marshall, P. Eidam, H. S. Eidam, J. Org. Chem. 2006, 71, 4840 – 4844. [42] a) J. A. Marshall, C. M. Grant, J. Org. Chem. 1999, 64, 696 – 697; b) B. A. Johns, C. M. Grant, J. A. Marshall, Org. Synth. 2002, 79, 59 – 71. [43] a) S. D. Rychnovsky, D. J. Skalitzky, Tetrahedron Lett. 1990, 31, 945 – 948; b) S. D. Rychnovsky, B. N. Rogers, T. I. Richardson, Acc. Chem. Res. 1998, 31, 9 – 17. [44] I. Fleming, T. W. Newton, F. Roessler, J. Chem. Soc. Perkin Trans. 1 1981, 2527 – 2532. [45] Compound 31 can be prepared on large scale from d-glutamic acid, c.f.: X. Cai, M. S. Chorghade, A. Fura, G. S. Grewal, K. A. Jauregui, H. A. Lounsbury, R. T. Scannell, C. G. Yeh, M. A. Young, S. Yu, L. Guo, R. M. Moriarty, R. Penmasta, M. S. Rao, R. K. Singhal, Z. Song, J. P. Staszewski, S. M. Tuladhar, S. Yang, Org. Process Res. Dev. 1998, 2, 73 – 76. [46] A benzyl ether had previously been used to protect the primary OH; however, a trityl group led to better results in the alkylation/epimerization step introducing the methyl branch adjacent to the lactone, see: a) Y. K. Nishida, M. Konno, Y. Fukushima, H. Ohrui, H. Meguro, Agric. Biol. Chem. 1986, 50, 191 – 193; b) Y. Shiro, K. Kato, M. Fujii, Y. Ida, H. Akita, Tetrahedron 2006, 62, 8687 – 8695. [47] The X-ray structure of this compound is contained in the Supporting Information as well as in our Communication (ref. [10]); CCDC 815411 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif. [48] J. Parikh, W. von E. Doering, J. Am. Chem. Soc. 1967, 89, 5505 – 5507. [49] a) W. Notz, B. List, J. Am. Chem. Soc. 2000, 122, 7386 – 7387; b) S. Mukherjee, J. W. Yang, S. Hoffmann, B. List, Chem. Rev. 2007, 107, 5471 – 5569. [50] T. R. Hoye, C. S. Jeffrey, F. Shao, Nat. Protoc. 2007, 2, 2451 – 2458.

10

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ÝÝ These are not the final page numbers!

Full Paper [51] For a related maneuver with an even more complex keto-lactone as the substrate, see: A. Arlt, S. Benson, S. Schulthoff, B. Gabor, A. Frstner, Chem. Eur. J. 2013, 19, 3596 – 3608. [52] V. Farina, B. Krishnan, J. Am. Chem. Soc. 1991, 113, 9585 – 9595. [53] S. Inoki, T. Mukaiyama, Chem. Lett. 1990, 67 – 70. [54] B. Menndez Prez, D. Schuch, J. Hartung, Org. Biomol. Chem. 2008, 6, 3532 – 3541. [55] a) M. Tokunaga, J. F. Larrow, F. Kakiuchi, E. N. Jacobsen, Science 1997, 277, 936 – 938; b) J. M. Keith, J. F. Larrow, E. N. Jacobsen, Adv. Synth. Catal. 2001, 343, 5 – 26. [56] W. Oppolzer, R. N. Radinov, Tetrahedron Lett. 1991, 32, 5777 – 5780. [57] P. Charreau, M. Julia, J.-N. Verpeaux, J. Organomet. Chem. 1989, 379, 201 – 210. [58] a) J. Inanaga, K. Hirata, H. Saeki, T. Katsuki, M. Yamaguchi, Bull. Chem. Soc. Jpn. 1979, 52, 1989 – 1993; b) A. Parenty, X. Moreau, J.-M. Campagne, Chem. Rev. 2006, 106, 911 – 939; update 1: A. Parenty, X. Moreau, G. Niel, J. M. Campagne, Chem. Rev. 2013, 113, PR1 – PR40. [59] a) J. K. Stille, Angew. Chem. Int. Ed. Engl. 1986, 25, 508 – 524; Angew. Chem. 1986, 98, 504 – 519; b) V. Farina, V. Krishnamurty, W. J. Scott, Org. React. 1997, 50, 1 – 652. [60] A. Frstner, J.-A. Funel, M. Tremblay, L. C. Bouchez, C. Nevado, M. Waser, J. Ackerstaff, C. C. Stimson, Chem. Commun. 2008, 2873 – 2875. [61] This method was instrumental for the total syntheses of several members of the amphidinolides B, H and G series, which contain a similar exo/endo diene substructure, see: a) A. Frstner, L. C. Bouchez, J.-A. Funel, V. Liepins, F.-H. Pore, R. Gilmour, F. Beaufils, D. Laurich, M. Tamiya, Angew. Chem. Int. Ed. 2007, 46, 9265 – 9270; Angew. Chem. 2007, 119, 9425 – 9430; b) A. Frstner, L. C. Bouchez, L. Morency, J.-A. Funel, V. Liepins, F.-H. Pore, R. Gilmour, D. Laurich, F. Beaufils, M. Tamiya, Chem. Eur. J. 2009, 15, 3983 – 4010. [62] G. D. Allred, L. S. Liebeskind, J. Am. Chem. Soc. 1996, 118, 2748 – 2749. [63] J. Srogl, G. D. Allred, L. S. Liebeskind, J. Am. Chem. Soc. 1997, 119, 12376 – 12377. [64] a) A. Frstner, J. Ackerstaff, Chem. Commun. 2008, 2870 – 2872; b) A. Larive, J. B. Unger, M. Thomas, C. Wirtz, C. Dubost, S. Handa, A. Frstner, Angew. Chem. Int. Ed. 2011, 50, 304 – 309; Angew. Chem. 2011, 123, 318 – 323. [65] a) A. Francais, A. Leyva, G. Etxebarria-Jardi, S. V. Ley, Org. Lett. 2010, 12, 340 – 343; b) R. S. Paley, K. E. Berry, J. M. Liu, T. T. Sanan, J. Org. Chem. 2009, 74, 1611 – 1620; c) I. Paterson, S. B. J. Kan, L. J. Gibson, Org. Lett. 2010, 12, 3724 – 3727; d) I. Paterson, L. J. Gibson, S. B. J. Kan, Org. Lett. 2010, 12, 5530 – 5533; e) I. Paterson, T. Paquet, S. M. Dalby, Org. Lett. 2011, 13, 4398 – 4401; f) A. FranÅais, A. Leyva-Prez, G. Etxebarria-Jardi, J. PeÇa, S. V. Ley, Chem. Eur. J. 2011, 17, 329 – 343; g) S. Mller, T. Mayer, F. Sasse, M. E. Maier, Org. Lett. 2011, 13, 3940 – 3943; h) C. Rink, V. Navickas, M. E. Maier, Org. Lett. 2011, 13, 2334 – 2337; i) R. S. Paley, M. C.

Chem. Eur. J. 2014, 20, 1 – 12

www.chemeurj.org

These are not the final page numbers! ÞÞ

[66]

[67]

[68] [69] [70]

[71] [72] [73]

[74] [75]

[76] [77]

[78] [79]

Laupheimer, N. A. K. Erskine, P. R. Rablen, R. D. Pike, J. S. Jones, Org. Lett. 2011, 13, 58 – 61; j) M. Domnguez, S. lvarez, R. lvarez, . R. de Lera, Tetrahedron 2012, 68, 1756 – 1761; k) L. Otero, B. Vaz, R. lvarez, . R. de Lera, Chem. Commun. 2013, 49, 5043 – 5045; l) H. Renata, Q. Zhou, P. S. Baran, Science 2013, 339, 59 – 63; m) I. Paterson, K. K.-H. Ng, S. Williams, D. C. Millican, S. M. Dalby, Angew. Chem. 2014, 126, 2730 – 2733; Angew. Chem. Int. Ed. 2014, 53, 2692 – 2995. a) J. Gagnepain, E. Moulin, A. Frstner, Chem. Eur. J. 2011, 17, 6964 – 6972; b) J. Gagnepain, E. Moulin, C. Nevado, M. Waser, A. Maier, G. Kelter, H.-H. Fiebig, A. Frstner, Chem. Eur. J. 2011, 17, 6973 – 6984. D. Mailhol, J. Willwacher, N. Kausch-Busies, E. E. Rubitski, Z. Sobol, M. Schuler, M.-H. Lam, S. Musto, F. Loganzo, A. Maderna, A. Frstner, J. Am. Chem. Soc. 2014, 136, 15719 – 15729. R. R. Schrock, Chem. Rev. 2002, 102, 145 – 179. K. Gebauer, A. Frstner, Angew. Chem. 2014, 126, 6511 – 6514; Angew. Chem. Int. Ed. 2014, 53, 6393 – 6396. a) P. Persich, J. Llaveria, R. Lhermet, T. de Haro, R. Stade, A. Kondoh, A. Frstner, Chem. Eur. J. 2013, 19, 13047 – 13058; b) R. Lhermet, A. Frstner, Chem. Eur. J. 2014, 20, 13188 – 13193. K. Bannai, T. Toru, T. O¯ba, T. Tanaka, N. Okamura, K. Watanabe, A. Hazato, S. Kurozumi, Tetrahedron 1986, 42, 6735 – 6746. S. V. Ley, J. Norman, W. P. Griffith, S. P. Marsden, Synthesis 1994, 639 – 666. Catalytic TPAP is used in combination with N-methyl-morpholine-Noxide which, upon oxygen-transfer, releases the parent tertiary amine and hence renders the medium basic. Similar observations were reported by Mahapatra and Carter, see ref. [9]. Occasionally, traces of epimers were detected, which had to be removed by preparative HPLC, obviously at the expense of the isolated yield. S. Hara, H. Dojo, S. Takinami, A. Suzuki, Tetrahedron Lett. 1983, 24, 731 – 734. W. A. Nugent, Chem. Commun. 1999, 1369 – 1370; b) Y. K. Chen, S.-J. Jeon, P. J. Walsh, W. A. Nugent, Org. Synth. 2005, 82, 87 – 92; c) J. G. Kim, P. J. Walsh, Angew. Chem. Int. Ed. 2006, 45, 4175 – 4178; Angew. Chem. 2006, 118, 4281 – 4284. R. Noyori, Asymmetric Catalysis in Organic Synthesis Wiley, New York, 1994. A. Frstner, Angew. Chem. 2014, 126, 8728 – 8740; Angew. Chem. Int. Ed. 2014, 53, 8587 – 8598.

Received: October 23, 2014 Published online on && &&, 0000

11

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper

FULL PAPER & Natural Products G. Valot, D. Mailhol, C. S. Regens, D. P. O’Malley, E. Godineau, H. Takikawa, P. Philipps, A. Frstner* && – && Concise Total Syntheses of Amphidinolides C and F

&

&

Chem. Eur. J. 2014, 20, 1 – 12

Non-Canonical: Amphidinolides C and F are unusual in that these macrolides of polyketide origin comprise an “umpoled” 1,4-diketone motif. This pattern served as the cornerstone of a uniform

blueprint based on a late-stage ringclosing alkyne metathesis followed by a platinum-catalyzed transannular hydroalkoxylation reaction (see scheme).

12

 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

ÝÝ These are not the final page numbers!