Syntheses based on 1,2-secopenicillins. Part 4. A new tricyclic beta;-lactam derivative

摘要

1977 189Syntheses Based on 1.2-Secopenicillins. Part 4.l A New Tricyclic p-Lactam DerivativeBy Michael J. Pearson, Beecham Pharmaceuticals, Research Division, Brockham Park, Betchworth, SurreyHeating (3R,4R)-I- azido-(t-butoxycarbonyl)methyl-4-(prop-2-ynylthio)-3-(triphenylmethylamino)azetidin-2-one (4) in refluxing toluene resulted in smooth intramolecular cycloaddition of the azido-group to the acetylenicf u n ct i o n to afford ( 5 aR, 6 R, 9 4 ) - t - b u t y I 6.7 - d i h y d r o - 7 - 0x0 - 6 - t r i p hen y I met h y I a m i n o - 4H, 5 a H- aze t o 2,l-61- v-triazolo 3,4-e 1.3,5 thiadiazepine-9-carboxylate (5).RH3 7AJSEVERAL fused tricyclic p-lactam derivatives have beendescribed recently.2-12 In most cases the compoundswere obtained by modification of the thiazolidine ordihydrothiazine ring system already present in penicillinsand cephalosphorins.Alkyl azides are known to reactwith activated acetylenes to afford v-triazoles,13 and sucha process has now been utilised in an intramolecularcycloaddition to provide a tricyclic p-lactam of novelstructure.The prop-2-ynylazetidinone (1) l4 was condensed withan excess of t-butyl glyoxylate in refluxing benzene togive the a-hydroxy-ester (2), which with thionyl chlorideafforded the a-chloro-derivative (3), both products beingmixtures of isomers. Treatment of (3) with tetramethyl-guanidinium azide in chloroform then gave the azido-acetylene (4). Although the product was homogeneousby t.l.c., its n.m.r. spectrum indicated a mixture ofdiastereoisomers.Signals at 6 5.08 and 5.34 are assignedto CHN,. Trituration with ether provided one isomer(4a) as a white crystalline solid. Evaporation of thePart 3, J. H. C. Nayler, N. F. Osborne, M. J. Pearson, andE. H. W. Bohme and J. E. Dolfini, J.C.S. Chem. Comm.,R. Heymamp;s. G. Amiard, and G. Nomind, Bull. SOC. chim.A. K. Bose, J. C . Kapur, S. D. Sharma, and M. S. Manhas,J. C . Sheehan, H. C. Dalzell, J. M. Greenwood, D. R.E. R. Farkas, E. T. Gunda, and J. Cs. JBszberdnyi, Tetra-R. Southgate, J.C.S. Perkin I, 1976, 1615.1972, 941.France, 1974, 563.Tetrahedron Letters, 1973, 2319.Ponzi, J . Org. Chem., 1974, 39, 277.hedron Letters, 1973, 5127.mother liquors gave an amorphous solid, consisting of(4a) (30) and the other diastereoisomer (4b) (70).When (4a) was refluxed in toluene for 3 h, simultaneousdisappearance of the characteristic azide and acetyleneC-H i.r.bands was accompanied by formation of a singlenew product (t.1.c. ; n.m.r.). Crystallisation from ethylacetate-light petroleum gave an SOY0 yield of a crystal-line solid assigned structure (5a) on the basis of molecularmodels (the orientation at C-9 has not been determined).It is known that the direction of this type of dipolaraddition is governed by both electronic and stericfactors.13 In our case it was surmised that the latterconsideration would be of major importance because ofthe inherent inflexibility of the p-lactam and triazolerings. The alternative mode of cycloaddition to givethe triazole (6) was discounted, since the trans-doublebond could not be accommodated in the rigid eight-membered ring.Heating the mixture of isomers (4a and b) in tolueneR. J.Stoodley and N. S. Watson, J.C.S. Perkin I , 1974,8 J. A. Edwards, A. Guzman, R. Johnson, P. J. Beeby, and9 D. 0. Spry, J.C.S. Chem. Comm., 1973, 671.l o D. 0. Spry, J . Org. Chem., 1975, 40, 2411.l1 S. R. Lammert and S. Kukolja, J . Amer. Chem. SOL, 1975,l2 R. D. Carroll and L. L. Reed, Tetrahedron Letters, 1975,13 G. Lrsquo;Abbe, Chew. Rev., 1969, 69, 345.l4 M. J. Harris, I. McMillan, J . H. C . Nayler, N. F. Osborne,1632.J. H. Fried, Tetrahedron Letters, 1974, 2031.97, 5582.3435.M. J. Pearson, and R. Southgate, J.C.S. Perkin I , 1976, 1612190 J.C.S. Perkin Ialso afforded a cycloaddition product.N.m.r. spectro-scopy indicated a mixture of triazoles (5a and b), consis-tent with the isomeric ratio in the starting azide (4).H H0(1) R = H(2) R = CH(OH)X02But(3) R = CHCI-CO~BU'(4) R = CHN~~CO~BU'H HThis acid-catalysed isomerisation was also evident inthe preparation of the acylamino-derivative (8). Detri-tylation of either (5a) or (5b) with toluene-fi-sulphonic15)I( 7 ) R'=H, R2= But( 8 ) R1=Ph0.CH2-C0, R 2 = d(9)R1=PhO-CH2-C0, R2=HH HP h 3C'(10) R =Ph3C( 1 4 ) R = HPh,CH+(11) R =CHN~*CO~BU'(13) R =CH(OH).CO~BU'(12)R =HHowever, chromatography (Merck silica gel H) gave an acid in methylene chloride-methanol gave the same freeamorphous product which consisted entirely of the base (7).This was acylated with phenoxyacetyl chloridetriazole (5b). Indeed similar treatment of the crystalline to provide the amide (8). The t-butyl group was re-triazole (5a) caused isomerisation to the amorphous (5b). moved by treatment with trifluoroacetic acid, to giv1977 191the free acid (9), which showed no antimicrobialactivity .The reason for this isomerisation is not clear, sinceDreiding models indicate that there are no particularlyunfavourable steric interactions in either isomer.However the phenomenon is not purely electronic, sincethe triazole (lo), prepared by the reaction of dimethylacetylenedicarboxylate with the azide (1 1) preparedfrom the azetidinone (12) l5 via (13) was obtained as amixture of isomers, the ratio of which was unchanged bychromatography on silica gel.Furthermore the removalof the triphenylmethyl group by toluene-9-sulphonicacid afforded the free base (14), as a mixture of isomers inthe same ratio. Thus it seems that steric factors areimportant in the tricyclic series, and although theisomerisation mechanism has not been studied a tent-ative suggestion is outlined in the Scheme.EXPERIMENTALGeneral experimental procedures were as outlined inPart 1.15 (3R,4R)- l-Hydroxy-(t-butoxycarbonyl)~thyamp;4-prop-2-ynylthio-3-triphenylmethylaminoazet~d~n-2-one (2) .-t-Butyl glyoxylate monohydrate (7.4 g, 10 equiv.) wasrefluxed in benzene (70 ml) for 0.5 h under a Dean-Starkhead. The lactam (1) l4 (1.99 g) was then added and thesolution was refluxed for 3amp; h.The mixture was washedwith water (5 x 20 ml), dried, and evaporated. Chromato-graphy afforded the amorphous a-hydroxy-ester (2) as amixture of isomers (1.72 g), v,,,. 3 500, 3 300, 1 772, and1 738 cm-1; 6 1.47 (9 H, s), 2.22 (1 H, t, J 2.5 Hz), 2.83-3.3(3 H, m, 1 H exch.), 4.49 ( 1 H, m, becoming two d, 4.47 and4.5, J 5 Hz, on D20 exch.), 4.73 and 4.77 (1 H, two d, J 5 Hz),5.27 (1 H, m, becoming two s, 5.2 and 5.35 on D20 exch.),and 7.1-7.7 (15 H, m).( 3R, 4R) - 2- A zido- (t-butoxycarbony 1) methyl-4-prop- 2-ynyl-thio-3-triphenylmethylaminoazetidin-2-one (4) .-The alcohol(2) (2.99 g) was dissolved in dry tetrahydrofuran (100 nil) at-15 "C and 2,6-lutidine (1.97 ml) was added, followeddropwise by thionyl chloride (1.23 ml) in tetrahydrofuran(20 ml) over 15 min.The mixture was filtered and thefiltrate evaporated ; the residue was dissolved in toluene andthe solution evaporated to afford the chloride (3) as anamorphous solid (3.01 g). This was dissolved in drychloroform (40 ml) and tetramethylguanidinium azide(950 ma) added. After 15 min the solution was washedwith dilute hydrochloric acid and brine, dried, and evapor-ated. Chromatography gave the azide (4) as a mixture ofisomers (2.31 g) . Trituration with ether provided oneisomer as a crystalline solid (4a) (900 mg), m.p. 154-155 "C(from ethyl acetate-light petroleum), vmaX. (Nujol) 3 370,3 270, 2 140, 1780, and 1762 cm-1; 6 1.49 (9 H, s), 2.21( l H , t , J 2.5Hz), 2.95(2H,t,J 2.5Hz), 2.97 (1 H , d , J 9Hz,exch.), 4.63 (1 H, m, collapsing to d, J 5 Hz, on D20 exch.),4.8 (1 H, d, J 5 Hz), 5.08 ( 1 H, s), and 7.1-7.8 (15 H, m)(Found: C, 67.3; H, 5.7; N, 12.7; S, 5.6.C,,H,,N,O,Srequires C, 67.2; H, 5.6; N, 12.7; S, 5.8).N.m.r. showed that the mother liquors from the crystal-lisation contained a further quantity (30) of the crystal-line isomer (4a), the remaining material being the otherdiastereoisomer (4b). No further enrichment in (4b) couldbe obtained since t.1.c. separation was not possible. Then.m.r. spectrum of (4b) was deduced from that of the mix-ture: 6 1.49 (9 H, s), 2.22 (1 H, t, J 2.5 Hz), 3.0 (1 H, m,exch.), 3.11 (2 H, t, J 2.5 Hz), 4.68 (1 H, m, collapses to d,J 5 Hz, on D20 exch.), 4.95 (1 H, d, J 5 Hz), 5.34 (1 H, s),and 7.1-7.8 (15 H, m).Cyclisation of the Crystalline Azide (4a) .-The azide (4a)(346 mg) was refluxed in toluene (10 ml) for 34 h.Thesolvent was evaporated off and the residue crystallised fromethyl acetate-light petroleum to give (5aR. 6R, 9Q-t-butyl6,7-dihydro- 7-oxo- 6-triph enylmethylamino-4H, 5aH-azeto-2, l-b-v-triazolo3,4-e 1,3,5thiadiazePine-9-carboxylate (5a)as white crystals (271 mg), m.p. 183 "C, aD23 + 148" (c 1.13in CHCl,), v,,,. (Nujol) 3 300, 1 787, and 1 761 cm-l; 6 1.48(9 H, s), 3.42 (1 H, d, J 10 Hz, exch.), 3.48 and 4.15 (2 H,ABq, J 15 Hz), 4.14 (1 H, m, collapses to d, J 5 Hz, on D,Oexch.), 4.58 (1 H, d, J 5 Hz), 6.14 (1 H, s ) , and 7.1-7.8(15 H, m) (Found: C , 67.2; H, 5.8; N, 12.6; S, 5.9.C,1H31N,0,S requires C, 67.2; H, 5.6; N, 12.7; S, 5.8).Cyclisation of the Mainly Non-crystalline Azide (4b) .-Theazide (4) (159 mg) consisting of (4b) and (4a) in the ratio7 : 31 was refluxed in toluene (10 ml) for 24 h.The solventwas evaporated off and chromatography gave starting mater-ial (19 mg) and pure triazole (5b) as an amorphous solid(126 mg), aIn23 +55.6" (c 1.28 in CHC1,); vmx. 1780 and1740 cm-1; 6 1.40 (9 H, s ) , 2.98 (1 H, d, J 10 Hz, exch.),3.95 (2 H, s), 4.80 (1 H, dd, J 5 and 10 Hz, collapsing to d,J 5 Hz, on D20 exch.), 5.3 (1 H, d, J 5 Hz), 7.03 (1 H, s),and 7.1-7.8 (15 H, m) (Found: M', 553.2156. C31H31-N503S requires M, 553.2148).Isomerisation of the Triazole (5a) .-(a) The crystallinetriazole (5a) (40 mg) was passed through a silica column(5 g; Merck silica gel H) (elution with chloroform). Then.m.r.spectrum of the recovered material indicated ca. 50isomerisation to (5b).(b) Treatment of the crystalline triazole (5a) (40 mg) inchloroform (2 ml) containing silica gel H (200 mg) for 2 hcaused complete conversion into (5b), as judged by n.m.r.spectroscopy.Detritylation of the Triazoles (5a and b) .-(a) Crystallinecompound (5a). The crystalline triazole (5a) (215 mg) wasdissolved in methylene chloride (10 ml) and cooled to - 15"C. Toluene-p-sulphonic acid (81 mg) in the minimumvolume of methanol was added over 1-2 min, and thesolution kept a t - 5 "C for 16 h. The solvent was evapor-ated off and the residue dissolved in ethyl acetate. Thesolution was washed with dilute aqueous sodium hydrogencarbonate and brine, dried, and evaporated.Chromato-graphy gave thefree base (7) (90 mg), vmx. 3 400, 3 350, 1 780,and1745cm-l; 61.5(9H,s), 1.84(2H,s,exch.),4.11(2H,s ) , 4.85 (1 H, d, J 5 Hz), 5.68 (1 H, d, J 5 Hz), 7.12 (1 H, s)and 7.59 (1 H, s) (Found: M+, 311.1075. C,,H1,N50,Srequires M , 3 1 1.1052).(b) The amorphous triazole (5b). Detritylation of theamorphous triazole (5b) as in (a) gave the same free base (7)Acylation of the Free Base (7).-The free base (7) (58 mg)was dissolved in methylene chloride (5 ml) at -10 "C.Triethylamine (20 mg) was added, followed by phenoxy-acetyl chloride (33 mg) in methylene chloride (0.5 ml) . Thesolution was washed with dilute aqueous sodium hydrogencarbonate followed by brine, dried, and evaporated.Chromatography gave the acylamino-derivative (8) as anamorphous solid (63 mg), vmx.3 380, 1 785, 1 740, and 1 685cm-l; 6 1.47 (9 H, s), 4.04 (2 H, s), 4.55 (2 H, s), 5.71 (1 H, d,(95 mg) -l5 E. G. Brain, A. J . Eglington, J. H. C. Nayler, M. J. Pearson,and R. Southgate, J.C.S. Perkin I, 1976, 447192 J.C.S. Perkin IJ 5 Hz), 5.97 (1 H, dd, J 5 and 10 Hz), and 6.82-7.73 (7 H,m) (Found: C, 53.8; H, 5.0; N, 15.3; S, 7.6. C,,H,,-N@,S requires C, 53.9; H, 5.2; N, 15.7; S, 7.2).Preparation of the Free Acid (9) .-The acylated derivative(8) (118 mg) was dissolved in trifluoroacetic acid (3 ml).After 30 min the solvent was evaporated off, the residuetreated with toluene, and the mixture evaporated; thisprocedure was repeated.Trituration with ether gave thefree acid (9) as a white amorphous solid (90 mg) , v,,,. (Nu jol)3 260br, 1 780, 1 745sh, and 1 680 cni-1; 6 CDCl,-(CD,),SO4.07 (2 H, s), 4.53 (2 H, s), 5.75 (1 H, dd, J 5 and 9 Hz,collapsing to a d, J 5 Hz, on D,O exch.) , 5.98 (1 H, d, J 5 Hz),6.75-7.6 (7 H, m, 1 H exch.), and 8.18 (1 H, d, J 9 Hz,exch .) .(3R, 4R) - 1 -Hydroxy- (t-butoxycarbonyl) wxethyZI-4-methyl-thio- 3-triphenylmethylaminoazetidin-Zone ( 1 3) .-Treatmentof the azetidinone (12) l5 (5.6 g) with t-butyl glyoxylate(17.7 g ) in refluxing benzene (120 ml) as described for theazetidinone (1) afforded the alcohol (13) as an amorphoussolid (5.68 g), v,,, 3 460, 3 250, 1 770, and 1 735 cm-1, amixture of isomers in the ratio ca. 5 : 3 the two CH-OHsignals appeared a t 6 5.05 and 5.24 (after D20 exch.);6 1.48 (9 H, s), 1.69 and 1.90 (3 H, two s), 3.0 (1 H, d, J 10Hz, exch.), 4.03br (1 H, s, exch.), 4.22-4.78 (2 H, m),4.96-5.38 (1 H, m, collapses to two s, 5.05 and 5.24, onD20 exch.), and 7.08-7.73 (15 H, m).(3R, 4R) - 1 - Azido- (t-butoxycarbonyl) methyl-4-methylthio-3-triphenylmethylaminoazetidin-2-one ( 11) .-The alcohol ( 13)(5.6 g ) was converted into the azide (1 1) by the procedurefor the preparation of (4).The product was an amorphoussolid (4 g), vmaX 3 280, 2 120, 1770, and 1 745 cm-l; 6 1.4(9 H, s), 1.7 and 1.87 (3 H, two s), 2.93 (1 H, d, J 10 Hz,exch.), 4.27-4.67 (2 H, m), 5.1 and 5.23 (1 H, twos, methineprotons of major and minor isomers, respectively) , and7.13-7.87 (15 H, m).The material was a mixture ofisomers in a ratio consistent with that in the starting alcohol(13). The two isomers could not be crystallised and wereinseparable by t.1.c.Dimethyl 1-2-MethyZthio-4-ox0-3-tri~henyZmethyZam~no-azetidin- l-yl- (t-butoxycarbony2)methyZ- 1,2,3-triazole-4,5-di-carboxylate (lo).-The azide (11) (529 mg) was refluxed intoluene (5 ml) containing dimethyl acetylenedicarboxylate(284 mg) for 1 h. The solvent was evaporated off and theresidue chromatographed on silica to give the amorphoustriazole (10) (490 mg), vmax. 1 775 and 1 750 crn-l; 8 1.43(9 H, s), 1.72 and 1.78 (3 H, two s), 3.0br (1 H, s, exch.),3.97(6H,s),4.33-5.07(2H,m), 6.83and6.97(1HJtwos),and 7.2-7.92 (15 H, m). The isomer ratio, as judged by therelative intensities of the methine singlets at 6 6.83 and 6.97,was identical with that of the starting azide (1 1).Dimethyl 1-3-Amino-2-methylthio-4-oxoazetidin-l-yl-(t-butoxycarbonyZ)methyEJ- 1,2,3-triazoZe-4,5-dicarboxylate ( 14) .-Detritylation of the triazole (10) (355 mg) with toluene-p-sulphonic acid (105 mg) as described for (5) afforded the freebase (14) as a gummy foam (176 mg), v,,, 3 450, 3 330,1 780, 1 750, and 1 740sh cm-l; 6 1.45 (9 H, s), 1.75 (2 H, s,exch.), 2.07 and 2.12 (3 H, two s), 3.98 (6 H, s), 4.65 (1 H,d, J 5 Hz), 5.18 and 5.48 (1 H, two d, J 5 Hz), and 6.88 and7.03 (1 H, two s, methine proton of each isomer). Theproduct, homogeneous by t.l.c., was again a mixture ofisomers, the ratio being identical with that of the startingazide (11).I thank my colleagues a t Beecham Pharmaceuticals foradvice and encouragement.6/1382 Received, 15th July, 1976