摘要:

1976 1485Microbial Metabolites. Part XI2 Total Synthesis and Absolute Con-figuration of (S)-Carlosic Acid (4-Butyryl-2,5-dihydr0-3-hydroxy-5-0~0-furan-2-acetic Acid) and Conversion of (R)-5-Methyltetronic Acid into( R ) -Carolic Acid (3,4-Di hydro-8-methylf uro [3,4-b]oxepin-5,6(2H,8H)-di-one)By James L. Bloomer and Ffancis E. Kappler, Department of Chemistry, Temple University of the Common-Syntheses of (S)-2,5-dihydro-3-hydroxy-5-oxofuran-2-acetic acid (5) and its methyl ester are described. Bio-mimetic acylations of the ester and of (R)-3-hydroxy-2-methylfuran-2(5H)-one[(R)-~-methyltetronic acid] (1)to produce (S)-carlosic acid (3) and (R)-carolic acid (2). respectively, are described. The absolute configurationof carlosic acid from Penicillium charlesii NRRL 1887 has been established as S.wealth System of Higher Education, Philadelphia, Pennsylvania 1 91 22, U.S.A.THE considerable recent interest in the synthesis ofracemic tetronic acids of natural typez-9 prompts us toreport our own studies in the optically active series.One of the primary objectives in our studies on mouldtetronic acids has been the identification of the biogeneticintermediates involved and the development of methodsfor their synthesis, especially methods adaptable to iso-topic labelling studies.At the inception of these studies the absolute configur-ations of carolic and carlosic acids were not known, anda number of fundamental questions regarding the bio-genetic relationship of y-methyltetronic acid (1) tocarolic acid (2) and carlosic acid (3) had not been answered.0 t her workers subsequently established the configurationof the acid (1) to be R as shown.1°The incorporation of both y-methyltetronic acid (1)and carlosic acid (3) into carolic acid (2) in Penicilliumcharlesii NRRL 1887 was demonstrated by isotopecompetition experiments against [U-14C]glucose, andsubsequently by feeding [U-14C]-(1) and [U-14C]-(3) tothe organism.The latter compounds were prepared asfollows. Feeding [U-14C]glucose to the organism gavecarolic acid (2) and carlosic acid (3), both uniformlylabelled. The former was converted via the bromide (4)into [U-14C]-y-methyltetronic acid (1) by previousmethods.ll Both the uniformly labelled precursors wereincorporated in good yield.From these biosynthetic studies l2 it was apparentthat dual metabolic pathways to (2) existed and thatdevelopment of a procedure for the acylation of tetronicacids in the 3- (or a-) position would be not only bio-mimetic but also potentially applicable to the synthesisof all known mould tetronic acids from two simple a-(a) J.L. Bloomer, S. M. H. Zaidi, J. T. Strupczewski, C. s.Brosz, and L. A. Gudzyk, J . Org. Chem., 1974, 39, 3615 shouldbe considered as Part X. For preliminary communications seeJ. L. Bloomer and F. E. Kappler, (b) Tetrahedron Letters, 1973,163; (c) J . Ovg. Chem., 1974, 39, 113. For previous papers inthe series see ( d ) J. L. Bloomer, W. R. Eder, and W. F. Hoffman,Chewz. Cornin., 1968, 354; J. L. Bloomer and W.F. Hoffman,Tetvahedvon Letters, 1969, 4339; J. L. Bloomer, M. A. Gross,F. E. Kappler, and G. N. Pandey, Chem. Comm., 1970, 1030;J. L. BIoomer, W. R. Eder, and W. F. Hoffman, J . Chem. SOC.( C ) , 1970, 1848; J. L. Bloomer, W. R. Eder, and W. F. Hoffman,Bryologist, 1970, 73, 586; J. L. Bloomer, F. E. Kappler, andG. N. Pandey, Chem. Comin., 1972, 243; J. L. Bloomer andF. E. Kappler, ibid., p. 1047.F. H. Andresen, A. Svendsen, and P. M. Boll, Actu Chem.Scand., 1974, %B, 130.S. Gelin and A. Galliaud, Comfit. rend., 1972, 275C, 897.unsubstituted precursors, viz. (1) and (5) [didehydro-carolic acid (6) being an exception owing to the antici-pated lability of the vinyl ether group].HM e s o H(41HOAs a model for our acylation studies we selected (RS)-(l), readily prepared by two general methods of tetronicacid synthesis, viz.the closure of y-bromo-esters of type(8) 1 3 9 1 4 and the cyclisation of acetoacetates of lacticesters, described below.16 The former synthesis, though,has thus far been limited to racemic materials.S. Gelin and P. Pollet, Compt. rend., 1974, 279C, 346.J. V. Greenhill and T. Tomassini, Tetrahedron Letters, 1974,T. P. C. Mulholland, R. Foster, and D. B. Haydock, J.C.S.A. Svendsen and P. M. Boll, Tetrahedyon, 1973,29, 4261.A. Svendsen and P. M. Boll, J . Org. Chem., 1976, 40, 1927.M. Omo and N. Kawabe, Japan Kokai, 1974, 69, 659 (Chem.P. M. Boll, E. Sorensen. and K. Balieu, Acta Chem. Scand.,l1 P. W. Clutterbuck, H. Raistrick, and F. Reuter, Biochem.J.,12 J. L. Bloomer, F. E. Kappler, and G. N. Pandey, J.C.S.la M. E. DeMarcay, Compt. rend., 1879, 88, 126.l4 L. Wolff, Annalen, 1896, 291, 226.l6 R. N. Lacep, J. Chem. Soc., 1954, 832.2683.Perkin I , 1972, 1225.Abs., 1974, 71, 1 5 , 1 9 7 6 ~ ) .1968, 22, 3251.1935, 29, 300.Chem. Comm., 1972, 2431486 J.C.S. Perkin IAs models for a-acylation, tetronic acids were preparedby the classical methods,13*14 whereby p-oxo-esters oftype (7) were brominated and rearranged to y-bromo-p-oxo-esters (€9, which were then cyclised to the racemicBr( 8 )HO xY &*tetronic acids (9) (in the original studies R = Et,X = Me, Y = H13 and R = Et, X = Br, Y = H 14).With the y-bromoacetoacetic ester (8; R = Et, X = Y= H) closure to the a-unsubstituted parent tetronic acid(9; X = Y = H) could not be effected, and in order toprepare the required a-unsubstituted tetronic acids (9 ;X = H) it was necessary to reduce the a-bromotetronicacids (9; X = Br) or to prepare the a-ethoxycarbonyl-tetronic acids (9; X = C0,Et) via cyclisation of y-bromoacylmalonates.For example, acylation of mal-onic ester with a-bromopropionyl chloride gave the y-bromo-ester (8; R = Et, X = CO,Et, Y = Me), cyclis-ation of which afforded the a-ethoxycarbonyltetronicacid (9; X = CO,Et, Y = Me); l6 this could be de-ethoxycarbonylated.A more convenient synthesis of (8; R = Et, X =CO,Et, Y = Me) involves bromination and rearrange-ment of the propionylmalonic ester (7; R = Et, X =CO,Et, Y = Me). The availability of both propionylchloride and diethyl malonate in several isotopicallylabelled forms with both 14C and 13C makes this approachvery useful for labelling studies.The acyclic bromo-esters were best cyclised by use ofdilute alkali rather than thermally, a result which hasrecently been reported inde~endently.~ The a-ethoxy-carbonyltetronic acid (9; X = CO,Et, Y = Me) wasconverted by dilute alkali, acidification, and warminginto y-methyltetronic acid (9; X = H, Y = Me), pre-sumably via the carboxylic acid (9; X = CO,H, Y =Me).Acetylmalonic ester l7 (7; R = Et, X = CO,Et,Y = H) was converted into the bromo-ester (8; R = Et,X = CO,Et, Y = H); cyclisation then gave the a-ethoxycarbonyltetronic acid (9; X = C03Et, Y = H)and de-ethoxycarbonylation afforded tetronic acid (9;X = Y = H).Phenylacetylmalonic ester (7; R = Et,X = CO,Et, Y = Ph) l8 was similarly converted into y-l6 E. Benary, Ber., 1911, 44, 1759.l7 A. Ogata, Y . Nosaki, and K. Takagi, J . Phavm. SOC. Japapz,lB W. Borsche and U. Wannagat, Chem. Ber., 1952, 85, 193.1939, 59, 105 (Chem. Abs., 1939, 33, 4230’).phenyltetronic acid (9; X = H, Y = Ph). The ethoxy-carbonylpropionylmalonicester (7; R = Et, X = CO,Et,Y = EtO,C*CH,) was also prepared l9 but attempts tocyclise the derived bromo-compound to the a-ethoxycarbonyltetronic acid (9; X = CO,Et, Y = MeO,C*CH,)failed, presumably owing to the ease or elimination ofHBr, since a considerable quantity of maleic acid wasisolated on the usual work-up.The second major synthetic approach to the tetronicacid nucleus, viz.the cyclisation of acetoacetates, was dueto Lacey,15 who cyclised the ester (10; R = Et, X = Me)to the tetronic acid (11; X = Me). The latter could beconverted into the a-unsubstituted tetronic acid (RS)-( 1)via bromination [to the a-bromotetronic acid (12;W = Br, X = Me)], and catalytic reduction of thebromo-group.ll The bromination and reduction stepswere high-yielding, but cyclisation by use of sodium asreported gave only a 20-30% yield of the tetronic acid(11; X = Me). However, use of potassium t-butoxidein t-butyl alcohol improved the yield to 95%.0 0~ ( I o ) = racemic formj [ (11) = racemic form]HO W1 (12a) [ ( 1 2 ) = racemic form0R e0Although Haynes had reported a-acylation of y-phenyltetronic acid (13; R = Ph) to (14; R = Ph) byuse of acetyl chloride-tin(1v) chloride, the method didnot give significant quantities of a-acylated materialwhen applied to y-methyltetronic acid (15) [i.e.(M)-(l)].Since the O-acylation of thallium enolates with acylchlorides had been reported for phenols,21 we investigatedthe application of this reaction to the tetronic acids.The thallium enolate (16) was prepared from the tetronicacid (15) in essentially quantitative yield with thallium(1)l9 U. Eisner, J. A. Elvidge, and R. P. Linstead, J . Chenz. Soc.,1950, 2223.2o L. J. Haynes and J. Jamieson, J . Chem. Soc., 1968, 4132.21 E. C. Taylor, G. H. Hawks, tert., and A. McKillop, J. Amev.Chenz. SOC., 1968, 90, 24221976 1487ethoxide in benzene.Although an attempted C-acyl-ation failed with acyl fluorides, the acid chloridesPhCOCl , PrCOCl , and MeO,CCH,*CH,COCl convertedthe thallium salt into the corresponding O-acylated, , a 0 0 - Me **(15) (16)0(18) (17)tetronic acids (17; R = Ph, Pr, or iMeO,C-CH,) in yieldsof 94, 89, and 99%, respectively.We next attempted the conversion of the O-acylatedtetronic acids into C-acylated derivatives via Fries-typearrangements. Though the yields were disappointing [a29% yield of or-butyryl-y-methyltetronic acid (18; R =Pr) by TiC1, in PhNO, a t 50 "C], we were able to extendthe process to a synthesis of (RS)-carolic acid (2) via y-bromobutyronitrile 23 and y-bromobutyric acid.24 They-bromo-acid was converted into the acid chloride in theusual manner and used to form the O-acylated tetronicacid, which was then taken directly for rearrangement.The intermediate, presumably the bromobutyryltetronicacid (18; R = BrCH,*CH,*CH2) , was best converteddirectly into (RS)-carolic acid (2) by dilute alkali.Theoverall yield was 22% from the enolate (16).In an attempt to prepare terrestric acid, a homologueof carolic acid , 4-bromohexanoyl chloride was preparedas follows. The Doebner modification of the Knoeven-age1 reaction 25 on butyraldehyde gave hex-2-enoic acid,which was converted into y-caprolactone z6 and 4-bromo-hexanoic acid 27 as previously reported. The acid wasconverted into the chloride and thence into the O-acylderivative (17; R = MeCH,*CHBr.CH,*CH,) as de-scribed above in 92% yield.Attempts to rearrange andand cyclise this material as for the carolic acid series werenot successful.The a-acylation reaction was considerably improvedby omitting the thalliation step. Thus treatment of thecc-unsubstituted tetronic acid (15) directly with PrCOCland TiCl, in PhNO, at 50 "C gave a 72% yield of the CC-butyryltetronic acid (18; R = Pr). Likewise a similaracylation with BrCH,*CH,*CH,*COCl followed by thedilute alkali treatment described above gave a 53% yieldof (RS)-carolic acid (2). Attempts to extend the pro-cedure to the carolinic ester (19; R = Me) were notsuccessful, possibly because MeO,C*CH,*CH,*COCl cyc-40, 537.m J. Cason, Org. Synth., COIL Vol. 111, 1955, p.169.23 C. G. Derick and R. W. Hess, J . Amer. Chew. SOC., 1918,24 C. S. Marvel and E. I<. Birkheimer, J . Anzer. Che;+t. SOC.,1929, 51, 260.lised intramolecularly under even these mild conditions.Attempts to extend the synthesis to terrestric acid (20)from the 4-bromohexanoyl chloride described abovewere not successful, possibly owing to the greaterdifficulty in achieving the intramolecular cyclisation tothe final product, which would presumably require theSN2 displacement of a secondary rather than primaryhalide by a tetronate anion.In order to test whether the stereochemical integrityof position 5 in the tetronic acid nucleus was maintained,the tetronic acid (1) obtained from natural carolic acid(2) as described above l1 was used to resynthesise carolicacid (2).The yield was comparable and the product,of 97% optical purity, was identical in m.p. and i.r. andn.m.r. spectra with the natural carolic acid.Since our biosynthetic studies suggested that carolinicacid (19 ; R = H) and terrestric acid (20) were metabolictermini rather than intermediates, the syntheses of (19;R = H) via the ester (19; R = Me) and of (20) via (21)were not pursued, but our efforts were directed insteadto the synthon ( 5 ) .In addition to the biogenetic and synthetic advantagesoutlined above, the synthon (5) offered the added possi-bility of establishing the absolute configuration ofcarlosic acid (3), a point of considerable importance inelucidating the detailed biogenesis of carolic acid (2).0Me.Me.H H(19) (20)121) ( 2 2 )The acetoacetate (10; X = Me) of ethyl (RS)-lactatecould be prepared and cyclised by sodium as reported byLacey,15 but the yield of the cyclisation step was low.The yield of the cyclisation product, a-acetyl-y-methyl-tetronic acid (11; X = Me) was improved to 95% byuse of potassium t-butoxide in t-butyl alcohol at reflux.(RS)-Malic acid was readily esterified and convertedby treatment with keten dimer and triethylamine intothe triester (10; R = Me, X = MeO,C*CH,); however,application of the above cyclisation conditions resultedin almost total elimination of the equivalent of aceto-acetic acid, to give dimethyl fumarate. In thet-butoxide system the lowest temperature which could be35 S. E.Boxer and R. P. Linstead, J . Chem. SOC., 1931, 740.36 R. P. Linstead, J . Chem. SOC., 1932, 115.37 J. I?. Lane and H. W. Heine, J . Amer. Chem. SOC., 1951,73, 13481488 J.C.S. Perkin Iused was ca. 0 "C, giving a 39% yield of the tetronic acid(11 ; X = MeO,C*CH,). Doubtless, use of other basesand lower temperature could improve this yield, butthis was not pursued since the starting materials wereinexpensive and all subsequent steps were simple andhigh-yielding. Conversion of the a-acetyltetronic acid(11 ; X = MeO,C*CH,) into the corresponding a-bromo-tetronic acid (12; W = Br, X = MeO,C*CH,) bybromination and thence into the a-unsubstituted tetronicacid (12; W = H, X = MeO,CCH,) by reduction wascarried out as in the y-methyl series discussed above.11,28As in the y-methyl series, modification of the brominationby use of less bromine allowed the isolation of the inter-mediate a-bromoacetyltetronic acid (12 ; W = CO*CH,-Br, X = MeO,C*CH,).Treatment of the a-acetyltetronic acid (11; X =MeO,CCH,) with dilute alkali saponified the aliphaticester group to give the carboxylic acid (11; X =HO,C*CH,).As in the ester series, bromination gavethe a-bromotetronic acid bearing a free carboxy-group(12 ; W = Br, X = HO,C*CH,) , and catalytic reductionthen gave the a-unsubstituted tetronic acid (12; W =H, X = HO,C*CH,), which is the RS form of the syn-thon (5).Attempts to convert the synthon (RS)-(5) directlyinto carlosic acid were not successful. Difficulty wasalso encountered when attempts were made to O-acylatethe thallium salt of the ester (12; W = H, X = Me0,-CCH,) with butyryl chloride.Spectroscopic evidenceindicated that an O-butyryl derivative had been formed,but the yield was low. However direct a-acylation of theester (12; W = €I, X = MeO,C*CH,) with butyryl chlorideand TiCl, in nitrobenzene followed by gentle saponificationgave (RS)-carlosic acid [(RS)-(3)] in 59% overall yield.Similarly, natural (S)-malic acid was converted intoits methyl ester and thence into the acyclic acetoacetate(10a; X = MeO,C*CH,, R = Me), which was cyclisedto the tetronic acid ( l l a ; X = MeO,C*CH,). Saponific-ation of the ester group gave the acid ( l l a ; X = H0,-C*CH,), froin which the a-acetyl group could be removedby bromination [to give the a-bromotetronic acid (12a;W = Br, X = HO,C*CH,)] and catalytic hydrogenation[to yield the cc-unsubstituted tetronic acid (12a; W = H,X = HC0,C-CH,), i.e.( 5 ) ] . The last two compoundswere identical with materials derived from carlosic acid(3) 28 in i.r. and n.m.r. properties and closely similar inm.p. and optical rotation, thus establishing the absoluteconfiguration of (3) to be S as shown.The stereochemical result was unexpected, as a simpledecarboxylation of (3) in the organism should have given(2) with the same, rather than opposite, stereochemistryat position 5. This biosynthetic problem is underinvestigation.The ester (12; W = H, X = Me0,C-CH,) was acylatedas in the methyl series to (12; W = COPr, X = Me0,-C-CH,) which was converted by gentle saponificationinto (RS)-carlosic acid (3).Repetition with the optic-ally active ester (12a) gave carlosic acid (3) of the appro-priate absolute configuration.The synthetic procedures described should be especi-ally suitable for the production of (3) with specificisotopic labels for biosynthetic studies. Extension ofthese studies to specifically labelled viridicatic acid (12a;W = CO*[CH,],CH,, X = HO,C*CH,) via the methylester (12a; W = CO*[CHJ,-CH,, X = MeO,C*CH,)should be particularly easy and should allow us to testwhether viridicatic acid plays a critical role in the bio-synthesis of terrestric acid (20) similar to that of carlosicacid (3) in the biosynthesis of carolic acid (2).Conversion of (12; W = H, X = Me0,C*CH2), ob-tained by an alternative synthetic route, into methyl(RS)-carlate [ (RS)-(22)] has been reported,s under condi-tions essentially the same as ours except for the substitu-tion of 4-chlorobutyroyl chloride for the 4-bromo-material.Our combined studies constitute a formalsynthesis of carlic acid (22) in its correct absoluteconfiguration.The only synthetic limitation to our direct a-acylationprocedure for tetronic acids involved the use of a labileacylating agent, vix. p-methoxycarbonylpropionyl chlor-ide which, when treated under the standard conditionsdescribed with y-methyltetronic acid (15), allowed thelatter to be recovered unchanged. Substitution of asuitably masked carboxy-function in the acylatingreagent should allow synthesis of carolinic acid (19;R = H), though we have not pursued this further owingto our limited biosynthetic interest in this material.EXPERIMENTAL1.r.spectra were recorded with a Beckman IR-5A or aPerkin-Elmer Infracord spectrophotometer, and n.m.r.spectra with a Varian A-60A or XL-100 spectrophotometer(Me,Si as internal standard). 1.r. and n.m.r. data for com-pounds marked with an asterisk are available as Supple-mentary Publication No. SUP 21747 (4 pp.).* M.p.s weremeasured with a Kofler hot-stage apparatus. Micro-analyses were performed by Schwarzkopf MicroanalyticalLaboratories, Woodside, New York. Analytical t.1.c. wasperformed with Eastman Chromagram sheets coated withfluorescent silica gel, and preparative t .l.c.with glass plates(20 cm x 20 cm x 2 mm) coated with fluorescent silica gel(Brinkmann GFZs4). For column chromatography Woelmsilica gel was used (activity I1 unless otherwise specified).Optical rotations were measured with a Bendix-NPL auto-matic polarimeter. Organic solutions were dried prior toevaporation by addition of sodium sulphate (5-10% wlw)unless otherwise specified. Solvents were evaporated off at30-40 "C and 20-30 mmHg.Ethyl 2,5-Dihydro-4-hydroxy-5-methyl-2-oxofuran-3-carbox-ylute (9; X = CO,Et, Y = Me).-Theester ( 7 ; R = Et, X =CO,Et, Y = Me) was prepared from propionyl chloride anddiethyl ethoxymagnesiomalonate by the procedure used forthe ester (7; R = Et, X = CO,Et, Y = H) (below). Bro-mine (5 ml) was added over 45 min to a stirred solution of theester (21.6 g) in carbon disulphide (25 ml). After a further10 min the mixture was evaporated and the residue pouredinto N-sodium hydroxide (300 ml).After 12 h at room* For detailsof Supplementary Publicationssee Notice to AuthorsNo. 7, J.C.S. Perkin I, 1975, Index issue.** P. W. Clutterbuck, H. Raistrick, and F. Reuter, Bioclzem. J.,1936, 29, 8711976 1489temperature the mixture was poured onto ice (100 g) andconcentrated hydrochloric acid (25 ml) . Extraction withchloroform and evaporation afforded a residue which wastaken up in ether; the solution was slowly concentrateduntil crystallisation commenced, then set aside to give thetetronic acid (9; X = CO,Et, Y = Me) * (14.1 g, 76%)(from ethyl acetate), m.p. 91-91.5' (1it.,l6 89-90'>.4-Hydroxy-5-methyZfurun-2(5H)-one (9; X = H, Y = Me).-The ester (9; X = CO,Et, Y = Me) (14 g) was refluxedin 2r;-sodium hydroxide (100 ml) for 45 min; the mixturewas cooled, acidified, and extracted with ether to give thetetronic acid (9; X = H, Y = Me) * (8.3 g, 1 0 0 ~ o ) , m.p.119-120" (from ethyl acetate) (1it.,l6 117-119').4-Hydroxy-5-phenyZfuran-2(5H)-one (9; X = H, Y =Ph).-The ester in the corresponding phenyl series (7;R = Et, X = CO,Et, Y = Ph) l8,* (27.8 g) was brominatedin chloroform as for the Y = Me series and cyclised asabove t o give the a-ethoxycarbonyltetronic acid (9; X =C02Et, Y = Ph) * (22.5 g, 91yo), m.p.140-141" (lit.,2s140"). De-ethoxycarbonylation as in the methyl seriesgave the y-phenyltetronic acid (9; X = H, Y = Ph) *(86y0), m.p.127" (lit.,29 127.5-128.5').4-Hydroxyfurun-2( 5H) -one (9 ; X = Y = H) .-Alcohol-free diethyl ethoxymagnesiomalonate was treated with 1equiv. of acetyl chloride in ether. The mixture was refluxedfor 2 h, acidified with sulphuric acid, and extracted withether, and the extract was distilled to give diethyl acetyl-malonate * (7; R = Et, X = CO,Et, Y = H) l7 in 66%yield. This ester (20.2 g) was brominated in carbon di-sulphide as for the Y = Me series. Evaporation of thesolvent left a residue which was heated at 130 "C (50 mmHg)for 1.5 h, cooled, and triturated with ether to give the a-ethoxycarbonyltetronic acid * (9; X = CO,Et, Y = H)(10.2 g) (twice recrystallised from EtOAc), m.p.122-123'(lit. ,30 124425"). De-ethoxycarbonylation as in theY = Me series, with 2~-sodium hydroxide, gave the tetronicacid (9; X = Y = H) (4.8 g ; 80% from acetonitrile), m.p.140-141" (lit.,30 m.p. 141').Improved Procedure for 3-AcetyZ-4-hydroxy-5-methyZfi.4~an-2(5H)-one ( 1 1 ; X = Me).-The ester (10; R = Et, X =Me) 1 5 9 * (8 g) was stirred under reflux with t-butyl alcohol(10 nil) and potassium t-butoxide (4.5 g) for 2 h. Thecooled solution was acidified ( ~ N - H C ~ ) , water was added (todissolve KCl), and the acidified solution was extracted withether and chloroform. The extracts were evaporated andthe residue treated with ether and chilled to -78 'C to givethe product (11; X = Me) * (5.9 g , 95%), m.p. 51-53"(from ether-hexane) (lit.,16 52-54").Thulliwn Salt ( 16) of PHydroxy-5-methylfuran-2(5H)-one(15).-The tetronic acid (15) (114 mg) was stirred withbenzene (2 ml), ethanol (0.5 ml), and thallium(1) ethoxide(0.07 nil, 0.25 g), for 15 min. The salt (16) was filtered offand washed with benzene and hexane; yield 296-317 mg(93-100%); m.p. 175-177", vmx. (Nujol) 1 670 cm-l, andwas used for the experiments below without further charac-terisation.4-Be~zzoyZoxy-5-methyZfuran-2(5H)-one (17; R = Ph).-The thallium salt (16) (317 mg, 1 mmol), benzoyl chloride(140 mg), and benzene (20 ml) were refluxed for 2 h ; themixture was then cooled and the thallium chloride filteredoff. The benzene solution was washed with saturatedsodium hydrogen carbonate and water, dried, and evapor-ated to give the product (17; R = Ph) (204 mg, 94y0), m.p.115--116", vmax.(Nujol) 1760, 1 740 (GO), and 1625 cm-1(C=C); 8=(CDC13) 1.61 (3 H, d, J 7 Hz, CH,*CH*O), 5.14(1 H, m, CH,*CH*O) , 6.27 (1 H, d, J 1.5 Hz, :CH), and 7.55-8.27 (5 H, m, Ph) (Found: C, 65.75; H, 4.55. Cl2H1,O4requires C, 66.05; H, 4.2%), Mi 218. Haynes20 hasclaimed to have prepared this compound, but gives m.p.39-43".4-ButyryZoxyZ-5-methylfuran-2( 5H) -one ( 17 ; R = Pr) .-The thallium salt (16) (296 mg), ether (5 ml), and butyrylchloride (0.1 ml) were stirred for 30 min. Thallium chloridewas filtered off, and the ether solution was washed withsaturated sodium hydrogen carbonate and water, dried, andevaporated to give the $yoduct (17; R = Pr) as a paleyellow oil (153 mg, 89%); vmk (film) 2 940 (C-H), 1 790,1 760 (GO), and 1 625 cm-l (C=C); SH(CDC~,) 1.02 (3 H, t,(2 H, m, CH,), 2.59 (2 H, t, J 6 Hzj CH,.CH,*CO), 4.97(1 H, m, CH-0), and 6.00 (1 H, d, J 1.5 Hz, vinylic) (Found:C, 58.5; H, 6.5.C,H,,O,, requires C, 58.7; H, 6.55%).2,5-Dihydro-2-methy2-5-oxo-3-furyZ P-Methoxycarbonylpro-pionate ( 17 ; R = MeO,C*CH,*CH,) .-P-Methoxycarbonyl-propionyl chloride (1.32 g) was treated as butyryl chlorideabove to give the diester (17; R = MeO,C*CH,*CH,) (1.9 g,99yo), m.p. 40-41'; vmX. (Nujol) 1 790, 1 755, 1 725 (C=O),and 1625 cm-1 (C=C); S=(CDC13) 1.40 (3 H, d, J 7 Hz,(1 H, q, J 7 Hz, CH,CH*O), and 6.15 (1 €3, d, J 1.5 Hz,vinylic) (Found: C, 52.45; H, 5.2. CloHl,O, requires C,52.65; H, 6.3%).2,5-Dihydro-2-methyl-5-oxo-3-furyZ 4-Bromohexanoate (1 7 ;R = EtCHBrCH,*CH,) .-4-Bromohexanoic acid * (23.9 g)and thionyl chloride (15 ml) were heated on a steam-bathfor 1 h.The excess of reagent was removed, and the residuedistilled to give the acid chloride (20.2 g, 73%), b.p. 111-114" at 34 mmHg, vmX. 1 790 cm-l (GO), which was takendirectly for the subsequent step without further characteris-ation. The acid chloride (1.26 g) was treated with thethallium salt (16) as for butyryl chloride to give the ester(17; R = EtCHBr*CH,*CH,) (yellow oil; 1.6 g, 92%),vmx. (film) 1 780, 1755 (GO), and 1 625 cm-1 (C=C); SH(CDCl,) 1.13 (3 H, t, J 7 Hz, CH3*CH2), 1.72-3.0 (6 H, m,alkyl chain), 4.05 (1 H, m, CH,*CHBr*CH,), 4.92 (1 H, q,J 7 Hz, CH,*CH*O), and 6.04 (1 H, d, J 1.5 Hz, vinylic).3-ButyryZ-4-hydroxy-5-methylfuran-2( 5H) -one ( 18 ; R =Pr) .-(A) via Fries rearrangement of the ester ( 17 ; R = Pr).The ester (17; R = Pr) (153 mg), nitrobenzene (2 ml),and titanium tetrachloride (0.3 ml) were stirred and heateda t 50 "C for 35 min.The mixture was then cooled to 0 "Cand acidified (~M-HC~). The organic layer was separatedand the aqueous phase extracted with ether and chloroform.The combined extracts were washed with water, andextracted with saturated sodium hydrogen carbonate ; thissolution was in turn extracted with ether (25 ml), acidified(conc. HCl), and extracted with chloroform. Evaporationof the chloroform solution gave a residue (85 mg), whichwas chromatographed on preparative silica gel plates (twice)in chloroform-acetic acid ( 9 : 1) to give the product (18;R = Pr) (45 mg, 29%) as pale yellow crystals, m.p. 58-60°.Thetetronic acid (15) (1 14 mg) , butyryl chloride ( 1 10 mg), nitro-benzene (5 ml), and titanium tetrachloride (0.3 ml) werestirred and heated a t 50 "C for 2.5 h.The cooled solutionwas poured over ice and concentrated hydrochloric acid andthe aqueous phase was extracted with chloroform. Theacidic material was extracted with sodium hydrogen29 R. Anschutz and R. Bocker, Annalen, 1909, 368, 63.J 6 Hz, CH,*CH,), 1.47 (3 H, d, J 7 Hzj CH,*CH*O), 1.77CHJ, 2.72 (4 H, m, CH,*CH,), 3-70 (3 H, S , CH,O), 4-96(B) via Direct acylation of y-methyltetronic acid ( 15).E. Benary, B e y . , 1907, 40, 1079J.C.S.Perkin Icarbonate, and this solution was washed with chloroform,acidified, and extracted with chloroform. This last extractwas dried and evaporated to give the product (18; R = Pr)(130 mg, 72y0), identical with the material from method (A).3,4-Dihydro-8-methylfikro[3,4-b]oxepin-5,6(2H,8H)-dione[(RS)-Carolic Acid (2)].-(A). The thallium salt (16) was sus-pended in ether (50 ml), y-bromobutyryl chloride (1.6 g)was added, and the solution was stirred a t room temperaturefor 30 min. The thallium chloride was removed by filtration,and the ethereal solution was washed with water (2 x 20ml) and saturated sodium hydrogen carbonate (2 x 20 ml),dried, and evaporated to give a liquid ester (2 g). The esterwas dissolved in carbon disulphide (15 ml) and added drop-wise to a refluxing solution of carbon disulphide (40 ml)and titanium tetrachloride (2 ml).The solution turnedorange, and a viscous red oil separated. Refluxing wascontinued for 1 h. The solution was then cooled in iceand N-hydrochloric acid (40 ml) was added. The carbondisulphide layer was separated, and the aqueous phase wasextracted with chloroform (3 x 50 ml). The organicextracts were combined and extracted with 2~-sodiumhydroxide (2 x 50 ml). The basic extracts were washedwith ether, acidified with concentrated hydrochloric acid,and extracted with chloroform (1 x 50 ml; 3 x 25 ml).The chloroform layer was extracted with saturated sodiumhydrogen carbonate (5 ml), washed with water (10 ml),dried, and evaporated to give crude carolic acid (510 mg).Recrystallisation from ethanol gave (RS)-(2) (32 mg, 22y0),m.p.114--116" (lit.,31 113"), identical with the naturalproduct in i.r. and n.m.r. spectra.(B). The tetronic acid (RS)-(1) (114 mg, 1 mmol), y-bromobutyryl chloride * (187 mg), nitrobenzene (5 ml), andtitanium tetrachloride (0.3 ml) were stirred and heated a t50 "C for 2.5 h. After cooling, the solution was poured overice and concentrated hydrochloric acid. The aqueous phasewas extracted with chloroform. The chloroform wasevaporated off and the residue stirred with 2~-sodiumhydroxide (10 ml) a t room temperature for 30 min. Afteracidification with 3~-hydrochloric acid, the solution wasextracted with chloroform, which was back-extracted withsaturated sodium hydrogen carbonate (2 ml), dried, andevaporated to leave crude carolic acid, (RS)-(3) (135 mg).Recrystallization from ethanol gave (RS)-(2) (97 mg, 53Y0),m.p.1 1&116", identical in i.r. and n.m.r. spectra with thenatural material.Repetition of the above procedure with the R-isomer (1)[derived degradatively l1 from natural carolic acid (2)]gave carolic acid (2) (117 mg) identical in i.r. and n.m.r.spectra with the natural material. For optical activitydetermination this material was chromatographed on silicagel plates (RF 0.53 in acetone-hexane, 3 : 2) ; the product(80 mg) had m.p. 124-128' [129-130" when thrice re-crystallised from ethanol (yield 15 mg)], [aIDz1 +81.90 (c1.26) (lit.,32 [alDZo +84.0" (c not given)).Dimethy2 Malate.-A solution of (RS)-malic acid (27 g)in methanol (100 ml) was treated with gaseous hydrogenchloride until homogeneous and then stirred for 18 h.Removal of solvent and distillation gave (RS)-dimethylmalate * (2.38 g, 74%), b.p. 110-116" at 4.5 mmHg (lit.,33116" a t 11 mmHg).Repetition with (S)-malic acid gave (S)-dimethyl malate31 R.Sudo, A. Kaneda, and N. Itoh, J. Org. Claem., 1967, 32,32 P. W. Clutterbuck, W. N. Haworth, H. Raistrick, G. Smith,1844.and M. Stacey, Biochem. J., 1934, 28, 94.(79y0), b.p. 105-108" a t 2.5 mmHg (lit.,34 90-92' at 2mmHg), [b(IDz1 -9.2' (c 1.31 in MeOH) {lit.,34 [aID21 -8.9"(c 6.37 in MeOH)}.Acetoacetate (10; R = Me, X = MeO,C.CH,) of (RS)-Dimethyl Malate.-(RS)-Dimethyl malate (16.2 g), benzene(50 ml), and triethylamine (0.1 ml) were stirred and heatedto reflux.Keten dimer (50 ml) was added dropwise, andstirring and refluxing were continued for 1 h. The benzenewas removed under reduced pressure, and the residue passedthrough alumina (90 g) (ether as eluant) to yield the triester(10; R = Me, X = MeO,C*CH,) (19.7g, 80%). (Attempteddistillation of this, even in high vacuum, yielded dimethylfumarate as the only identified product.) Subsequentchromatography over alumina (90 g; ether as eluant) gavethe product (10; R = Me, X = MeO,CCH,) (19.2 g, 78%);v,,,. (film) 3 485 (C-H) and 1 725 cm-l (GO); 8=(CDCl,)2.25 (3 H, s, MeCO), 2.90 (2 H, d, J 6 Hz, COCH,CH), 3.50(2 H, s, COCH,*CO), 3.67 (3 H, s, CH,O), 3.72 (3 H, s,CH,O), and 5.47 (1 H, t, J 6 Hz, CH,CH*O).Because of its lability this material was not purifiedfurther but taken directly for the subsequent step.Repeti-tion with (S)-dimethyl malate gave (10a; R = Me, X =MeO,CCH,), [0(lD2l 2 1.6O (c 1.45).&!etlzyE 4-Acety~-2,5-dihydvo-3-hydro~~y-5-oxof.uran-2-ace-tate (11; X = MeO,C*CH,).-To the ester (10; R = Me,X = MeO,C.CH,) (5.17 g) was added t-butyl alcohol (60ml), by distillation directly from calcium hydride into thereaction vessel. The flask was cooled in ice and stirredunder nitrogen. Potassium t-butoxide (25 g) was addedrapidly in 2 g portions during 10-15 min. Stirring a t 0 "Cwas continued for 1 h. The mixture was allowed to warmto room temperature and stirred for an additional 30 min,then warmed on a steam-bath for 20 min, cooled again to0 OC, and acidified with 4~-hydrochloric acid (60 ml).Stirring was continued for 30 min, and the precipitate wasfiltered off and washed with ether (150 ml).The filtratewas shaken and the ether layer removed. The aqueousphase was then extracted with ether (2 x 150 ml). Theethereal extracts were then combined and dried. Solventswere removed a t 25-30" under reduced pressure. Thewater aspirator was replaced with an oil pump to removethe last traces of solvent. This gave a viscous orangeresidue which was dissolved in ether (40 ml) and chilled toyield yellow crystals. More ether (50 ml) was added, andthe solid was crushed, filtered off, powdered, and washedwith ether to give the ester (11; X = MeO,C.CH,) (16.6 g,39%), m.p.54-45".mmHg gave material of m.p. 58-58.5"; vmaU. (Nujol) 1 750,1725 (GO), 1 665, and 1 615 cm-l (C=C); 8H(CDC1,) 2.38(3 H, s, CH3CO), 2.72-2.80 (2 H, m, COCH,*CH), 3.60(3 H, s, CH,O), and 4.57-4.76 (1 H, m, CH,.CH*O) (in thisand some of the following spectra, multiplicities of theCH,.CH*O and CH,.CH-O signals are increased owing torestricted rotation) (Found: C, 50.1; H, 4.75. C,H,,O,requires C, 50.05; H, 4.65%). Repetition with (10a; R =Me, X = MeO,C*CH,) gave ( l l a ; X = MeO,CCH,), m.p.Methyl 4-Bronzoacetyl-2,5-dihyd~o-3-?~ydroxy-5-oxofu~an-2-acetate (12; X = MeO,C*CH,, W = CO*CH,Br).-To asolution of the ester (1 1 ; X = Me0,C-CH,) (2.14 g) in aceticacid (5 ml) at 40 "C was added bromine in acetic acid ( 1 ~ ;33 H.Havekoss, 0. Bayer, and H. Wolz, Ger. P. 738,92211943(Chem. Abs., 1946, 89, 30g7).34 H. Arakawa, Natuvwiss., 1963, 50, 441 (Chem. Abs., 1963,59, 7419e).Sublimation a t 50 "C and 2 x86-87O, -54.4" (C 1.21)1976 149110 ml) over 1 h. The acetic acid was evaporated off (25 "C;2.5 mmHg) to leave a viscous oil which crystallised whentriturated with ether. Filtration gave the product (12;X = MeO,CCH,, W = COCH,Br) C1.7 g from ethylacetate-hexane (twice recrystallized)], m.p. 95.5-97.5" ;vmX. (Nujol) 1740 (GO), 1665, and 1 610 cm-l (C=C);iSH[(CD,),CO] 3.02 (2 H, m, CO*CH,.CH*O), 3.67 (3 H, s,CH,O), 4.57 (2 H, s, COCH,Br), 4.90 (1 H, s, enol), and 5.24(1 H, M, CH,*CH*O).Methyl 4-BYOWO-2, 5-dilzydvo-3-hydroxy-5-oxofu~an-2-ace-tate (12; X = MeO,C*CH,, W = Br).-To a solution of theester (1 1 ; X = Me0,C-CH,) (2.14 g) in acetic acid ( 5 ml) at40 "C was added bromine in acetic acid ( 1 ~ ; 10 ml) over1 h.Water (10 ml) was added, followed by a second portionof the bromine solution (IM; 10 ml) at room temperature.Solvents were removed at 25-30 "C and 2.5 mmHg togive a white solid, which was triturated with ether (20 ml),broken up, and filtered off. The precipitate was furtherpowdered in the funnel and washed with ether (50 ml)and hexane (20 ml). This gave compound (12; X =MeO,C*CH,, W = Br) (1.14 g). Concentration of thewashings gave an additional 0.63 g. This was recrystallisedfrom ethyl acetate-hexane; yield 1.35 g (54%), m.p. 114-116". Sublimation at 96 "C and 1.5 x mmHg gavematerial of m.p.115-117"; vmax (Nujol) 2 590 (0-H),1730, 1720 (C=O), and 1655 cm-l (CX); ~H[(CD~),SO]2.44-3.24 (2 H, m, CO*CH,.CH.O), 3.60 (3 H, s, CH,O),5.10-5.32 (1 H, m, CH,-CH-O), and 7.42br (1 H, s, enol)(Found: C, 33.35; H, 2.75; Br, 31.6. C,H,BrO, requiresC, 33.45; H, 2.8; Br, 31.85%). Repetition with (lla;S = Me0,C-CH,) gave (12a; X = MeO,C*CH,, W = Br),m.p. 119-121".Methyl 2,5-Dihyd~o-3-I~yd~on.y-5-oxofu~an-2-acetate (12 ;X = MeO,CCH,, W = H).-Compound (12; X = Me0,-CCH,, W = Br) (100 mg) dissolved in 50% aqueous aceticacid (8 ml) containing 10% palladium-carbon (20 mg)was hydrogenated at room temperature and atmosphericpressure. The catalyst was removed and the filtrate ex-tracted with ether (3 x 10 ml). The extracts were driedand evaporated t o a white solid.Recrystallization fromtoluene gave the product (12; X = MeO,CCH,, W = H)(45 mg, 26y0), m.p. 112-114" (lit.,35 116-118"); vmax.(Nujol) 2 690 (0-H), 1740, 1695 (GO), and 1640 cm-1CO*CH,CH.O), 3.72 (3 H, s, CH30), 4.99 (1 H, s, vinylic),and 5.10 (1 H, m, CH,*CH-O). Repetition with (12a;X = MeO,CCH,, W = Br) gave (12a; X = MeO,CCH,,W = H), m.p. 116-117".4-A cety2-2,5-dilzydro-3-ltydioxy-5-oxo furan-2-acetic Acid(11; X = HO,CCH,).-The ester (11; X = MeO,CCH,)(462 mg) was dissolved in 6~-sodium hydroxide (1 ml) andstirred a t room temperature for 36 h. The resulting yellowsolution was poured over ice (2 g) and concentrated sulphuricacid (0.5 ml) . This solution was then continuously extractedwith ether for 18 h.The extract was dried and evaporated,and the residue recrystallised from ethyl acetate-hexane(2 x ) to give the acid (11; X = HO,C*CH,) (260 mg),m.p. 174-176". Sublimation at 140 "C and 3.5 xmmHg raised the n1.p. to 177-178"; vmx. (Nujol) 3 355(O-H), 1755, 1695 (GO), 1670, and 1615 cm-l (C=C);iS=[(CD,),CO] 2.49 (3 H, s, CH,*CO), 2.90-3.17 (2 H, m,CO*CH,CH*O), 5.05 (1 H, t, J 5 Hz, CH,-CH*O), 7.92 (1 H,s, CO,H), and 10.3 (1 H, s, enol) (Found: C, 47.8; H, 4.25.35 R. Nicoletti and L. Bsiocchi, Ann. Chim. (ItuZy), 1962, 52,716 (Chem. A h . , 1963, 59, 5014~).(C=C) ; 8~[CDC13-5~~ (CD3)2SO] 2.46-3.31 (2 H, m,C,H,O, requires C, 48.0; H, 4.05%). Repetition with (1 la;X = Me02C*CH2) gave (lla; X = HO,C*CH,), m.p.182-184", -10.3" (C 1.93).4-Brorno-2,5-di~zydro-3-hydroxy-5-oxofuran-2-a~et~c Acid(12; X = HO,C*CH,, W = Br).-The acid (11; X =H0,C-CH,) (65 mg) was dissolved in 50% acetic acid (1 ml).Bromine (106 mg) in 50% acetic acid (1 ml) was added.The solution was swirled and set aside for 5 min, thenevaporated to dryness at 25-30 "C and 2 mmHg. Theresidue was sublimed at 120 "C and 5 x mmHg to givethe product (12; X = H02C.CH,, W = Br) (42 mg, 53%),m.p. 198-199"; vmax. (Nujol) 2 680 (O-H), 1720, 1700(GO), and 1 655 cm-l (C=C) ; iS,[(CD,),SO] 2.52-3.29 (2 H,m, CH,*CH*O), 5.18-5.37 (1 H, m, CH,*CH.O), 9.60br (2 H,s, C0,H plus enol OH) (Found: C, 31.05; H, 2.4; Br,33.7. Calc. forC,H,BrO,: C, 30.4; H, 2.1; Br, 33.75%).Repetition with (lla; X = H02C*CH,) gave (12a; X =H0,C-CH2, W = Br), m.p.190-192", [aID2l -107.4" (c2.15) (lit.,28 m.p. 194"; [a], -117" (c not given)).2,5-Dihydro-3-hydroxy-5-oxofu~an-2-acetic Acid (1 2 ; X =HO,C.CH,, W = H) [(RS)-(5)].-Compound (12; X =HO,C*CH,, W = Br) (200 mg) was dissolved in 2~-sodiumhydroxide (5 ml) and shaken in hydrogen over 5% palla-dium-carbon (25 mg). After reduction was complete, thecatalyst was removed and the filtrate acidified with dilutesulphuric acid and extracted with ether (5 x 5 ml). Theextract was dried and evaporated to give material (130 mg)which on recrystallisation from ethyl acetate-hexaneafforded the acid (RS)-(5) (116 nig, 88%)) m.p. 185-187";vma. (Nujol) 2 665 (O-H), 1 725,l 695 (GO), and 1 620cm-1(C=C) ; G*[(CD,),CO] 2.38-3.30 (2 H, m, CO.CH,CH*O),4.98 (1 H, s, vinylic), 5.18-5.37 (1 H, m, CH,.CH*O), and10.28br (2 H, s, C0,H plus enol OH) (Found: C, 45.6; H,4.05. Calc. for C6H605: C, 45.6; H, 4.07;). Repetitionwith (12a; X = HO,CCH,, W = Br) galre (12a; X =H0,C-CH,, W = H) [i.e. (5)], m.p. 180-181", [aID21 -41.6"(c 1.86) {lit.,,, m.p. 182"; [a], -52' (c not given)).4-ButyryZ-2,5-dihydro- 3-hydroxy-5-0x0 fuian- 2-acetic Acid[(RS)-Carlosic Acid, (RS)-(3)].-To the ester (12; X =MeO,C*CH,, W = H) (180 mg) and butyryl chloride (110mg) in nitrobenzene (5 ml) , was added titanium tetrachloride(0.3 ml). The solution was stirred and heated at 50 "Cfor 2.5 h, then cooled and poured over ice and hydrochloricacid. The mixture was extracted with chloroform, whichin turn was extracted with saturated sodium hydrogencarbonate. Acidification of the aqueous extract followedby extraction with chloroform gave the crude ester (193mg), which was dissolved in chloroform and passed througha silica gel column (1 x 10 cm) (chloroform as eluant) toafford a pale yellow oil (151 mg). This was dissolved in5~-sodium hydroxide (15 ml) ; the solution was swirled for30 min, cooled in ice, and acidified. Extraction of theaqueous solution with ether gave (R.5)-carlosic acid [(RS)-(3)] (135 mg, 59y0), m.p. 169-172". Recrystallisation fromethyl acetate (2 X ) raised the m.p. t o 178-179". I.r.,n.m.r., and t.1.c. properties were identical with those ofnatural carlosic acid. Repetition with (12a; X = Me0,-CCH,, W = H) gave (S)-carlosic acid (3), m.p. 174-176" 17(lit.,,, m.p. 181", [aID21 - 137" (c 1.21) ; lit.,,& [a], -160" (cThis investigation was supported by a U.S. Public HealthService grant from the National Institute of Allergic andInfectious Diseases, The National Institutes of Health.[5/984 Received, 23rd May, 197610.2 1) }

ISSN:1472-7781    

DOI:10.1039/P19760001485

出版商:RSC    

年代:1976

数据来源: RSC