CEPHALOSPORINS
September 196s
933
The Chemistry of Cephalosporins. XII. Configuration of the Carboxyl Group in A2-Cephalosporins E. VAN HEYNINGEN AND LIXDAKAYAHERN Ldly Kpaeurch LuDorutories, Elz Lzlly und Company, IndzanapolLs, I n d m n u 46,206
Recewed J l a y 8, 1968 Although A3-cephaloaporins are potent antibiotics, Az-cephalosporins possess little or no microbiological activity. The work herein described establishes that the 4-carboxyl in Az-cephalosporins has the same absolute configurntion as the 3-carboxyl in penicillins. Consequently, the lack of antimicrobial activity is ascribed to stability of the p-lact,am ring in A2-cephalosporins rather than a disadvantageous configuration of the carboxyl group.
Abrahani :tnd Kewtonl showed that cephalosporin C (I),the cephalosporin produced naturally by Cephalosporium acrewioniuiiz, contained a dihydrothiazine ring in which the double bond was in the A3 position. Since that time several groups of investigators have prepared
valine residue as proved by X-ray crystallographic analysis6 and by degradation.' The niolecular configuration of a penicillin is designated as shown in I1 where the protons on the @-lactamring and the carboxyl all lie below the plane of the penani ring system which is concave with respect to the plane of the paper.
I1 cephalosporiii derivatives in which the double bond has been in the A 2 position.2 Whereas many of the :malogous A3-ccphalosporanic acids possessed good aritiinicrobial activity, the isomeric A2-cephalosporins were alniost completely inactive. Alt'hough seemingly trivial changes in molecular struct'ure often do cause dramatic changes in biological activit'y, such results are only infrequently predictable. The inactivity of A2-cephalosporins, because of their very close similarity to A3-cephalosporins, was unanticipated and surprising. Orice known, however, this inact'ivity requires explanation. Cooper3 and Collins and Richmond4 postulated that @-lactani antibiotics inhibit cell wall synthesis because they combine irreversibly with and thus inactivate an enzyme crucial to t'he synthesis. The covalent bond is formed by @-lact'aiiiacylation of the active site. Cocker and coworkers have observed that A*-cephalosporins have a @-lactam ring that is much niore stable toward basic hydrolysis than the @-lactam ring in A3-cephalosporins. llorinj has correlated @lactani stability with both antibiotic activity and @-lactam absorption in the infrared: in @-lactamantibiotics the niore labile the @-lactamring the lower the wavelength of @-lactam absorption and in general the greater the biological activity. A*-Cephalosporin esters have @-lactam absorptions at high wavelength conipnred to A3-cephalosporin analogs. To conclude that A2-cephalosporin inactivity is attributable to an unreactive @-lactani is therefore eminently reasonable. There does exist, however, another possible factor that must also be considered. Penicillins contain a D( 1 ) E. P. A h a h a m and G. G. F. Newton, Biochem. J . , 79, 377 (1961). (2) ( a ) R . R. Cliauvette and E. H. Flynn, J . .Wed. Chem., 9, 741 (1966);
011 A . B. Taylor. J . Chem. Soc.. 7020 (1965); (c) R . A. Archer a n d B. 5. liitcliell. . I . Oro. Ciirm.. 31, 3409 (1966); (d) J. D. Cocker, S. Eardley, G. I. Gregory, RI. E. Hall, and .I. G. Long, J . Chem. S o c . , C , 1142 (1966). ( 3 ) P. I). Coog'r.
Bacterial. Res., 20, 28 (1956).
( 4 ) . I . F. Collin* a n < ]AI. H . Richmond, Y a t u r e , 196, 142 (1962). (51 R . H. Alorin ( a h iiiioteri Ijy E.: Van Heyningen), . l d r n ~ ~n.r u y J < C Y . , 4 ,
48 (1967).
X-Ray analyses of a penicillin6 and a cephalosporin,s as well as the conversion of a penicillin into a cephalosporin through a process in which the @-lactam ring remained i n t a ~ t proved ,~ that the @-lactam rings in both antibiotics are stereocheiiiically identical. Cocker arid coworkers2d deduced from ninr data that the carboxyl group in A2-cephalosporinswas in an axial position, but flexibility of the dihydrothiazine ring permits an axial configuration for either an CY- or @-carboxylso nnir analysis does not provide conclurive assignment of carboxyl configuration. Consequently, A2-cephalosporins niay have an a-carboxyl as in 111, opposite to that in penicillin, or a @-carboxyl as in IV, with the same configuration as in penicillin 11.
I11
IV
If one were to assuiiie that proper carboxyl configuration were crucial for antibiotic act'ivity, then, alniost certainly, the @ configuration as in natural penicillins would provide the active form. The inactivity of A*cephalosporins niay possibly be due, then, to the configuration of the carboxyl if it is in the CY configuration. If, however, A2-cephalosporins have a @-carboxyl,then their inactivit,y is probably not due to "wrong" carboxyl configuration because molecular models show that a A2-cephalosporin with @-carboxylis niore nearly super-
.*.
(6) D. Crowfoot, C. W. Bunn. B. I\*. Rogers-Low, and Turner-.Jones in "The Chemistry of Penicillin," H. T. Clarke, d. R . Johnson, and R . Robinson. Ed., Princeton University Press, Princeton, N . J., 1919, p 310. (7) E. Kaczka and K. Folkers, ref 6, p 243. (8) D. C. Hodakin and E. N. Maslen, Biocircm. .I.. 79, 393 (1961). (4) R . B. l l o r i n , R. G. .Jackson, R. .2. Xliieller, IC. R . Lai-agnino. \\'. 1:. Sicanlon. and 5 . 1,. .\iidre\vh, .I. A m . Chem. 5 o r . . 86, 18!Ii (1Uti:jj.
CH
Pd-C
'H 6OOH
Sh,h
K = C,H,OCH, KCONH H
3-l VI1
C,H,OCH,CONH
- - P -
v
product', X1:t ant1 Xlb, ret:iiiied the b:tiiie coiifiguratioii :it C'-4, but \\-ere isomeric at C-3. LiItancy nickel desulfurization of the i w i w iiiixture. S I a itnd b, yiclclecl a crude product which, when treated I\ ith diazoinethaiie, gave ;L niethyl ebter (VII). Tlc -hon.ed OTIC' very predoiiiiiiant spot accoiiipanied with faint spots of tracc coiitamiriutlit~,+ubstantiating the coiiclusion that XTa and XIb differed only ut (1-3 which hecoiiie5 :t i) nimetrical isopropyl group in the tiehulfurized product. The methyl ehter T'II was purified \)y coluiiiri chroinutography. The iiiethyl plieiiox~iiiet1i~ldesttiioperiicilliii:~te~ \'I :tiid TIL posses.: identical iiiiii', ir, arid iiv spectra. Their X-rny powder diagrams coincide exactly. ,iltliougli T'I nielted at 110" a r i d TI1 at 112', their iiiisture iiieltirig point not tlcpresml. S:tiiiplei i'1 iiiuy h a w hid t rac ~itaniiriants. 111 order t o tw certain tli:it our :wignnient \vas correct, subjected \'I1 to acid hydrolysis and isolntcd the valirie so obtained. L41though its rotatioil was low colripiired to pure D-v:ilin(>,the rotation n'ai strougly negative lilic, ~)-wIiiie. T h e 11) drol i i i : ~ ? I i u s t ~ c:iused some raccniizatioti. 'l'lc show.c~lonly oiic >pot for ~ ~ l i n e . ('oiiwliwrit 1) , t hc :ii.;igrinient abovc iiiust corrc~h. \TP caii onl) coiwlude, therefoi*c>,that it is riot tlir l-carhosyl coiiligurwtion 1)uL fi-l:tct:mi \td)ility that explains ~?-cephsloiporinbiological iiiact ivii y. \\-(I
Experimental Section'? Methyl 6-Phenoxymethyldesthiopenicilloate (VI).--The tlc.iilfiiiization method described for b e n ~ y lpenicillin was applied to plienouymei hylpenirillin Phenoxymethylpenicilliri pot demum salt 13 g, 7 74 minoleb) was diesolved in H201250 ml) arid approximately 18 g of Etaney Xi (stored under 1 1 2 0 ) was added The mixture was heated to reflux in an oil bath (165') for 15 i n i n and then qriichly cwoled. The reaction mixture was cam>m n .j i l o i e i la rotarv e ~ a p w d t o r1 1 121 Ill i \ d j ~ ~ r d ~ i t < I < ) I k.11.7kadntl I\ 1 olkrrs trf 6. 11 2 3 3 .
It-.-
ilidn
50
935
C'EPHALOYPOHIXS
September 1968
TABLEI NMR SPECTRA^ -c
tk-2:
CQH,OCHzCONH
0 h- H, CsHs-H
OCHrH
6-H
VI
409-433 m
269 s
295 m
VI1
407-462 m
269 s
294 m
KO.
H
5. H
237 t 211 q 237 t 212 q
COOCH,
2-CHa
3-H
2-H
(5 3 ) (3 0) (5 5) (3.0)
120-160 ha
255 d ( 7 . 5 )
120-1 60 hd
255 d (7.5)
61 d 59 d 62 d 60d
(6 (6 (6 (6
OCHD
3) 3) 5) 5)
223 a 233 e
CgH50CH2CONH#s]
--
CO OR R (OCH:
NO
NH
CnHs-H
OCH?
7-H
6-H
2-H
3-H
3-CHa
... 4.57 d (9 .5) 406-450 m 271 5 s 349 q ( 5 0 ) 297.5 d ( 5 . 0 ) 200 q (18) . .. 477 d (9 0 ) 410-453 m 274 s 339 q (4 0 ) 319.0 d ( 4 . 0 ) 357 m 451 d (9 5 ) 405-450 m 272 s XIaf 339 q (4 0 ) 322.5 d ( 4 . 0 ) 295 d (12) 137-160 m h h, XIbg h 405-451 m 275 s 338 d (4 2) 314.5 d (4.2) a DCC1, wab solvent. The instrument was a HA-60. b The desthiooeiiicillin is numbered as shifts in ips; coupling constants in parentheses. Heptuplet. e A3-C'ephalosporin. f CDC13 D20. h The multiplets do not permit assignment. IXe Xf
4-H
or H)
127 s ... 228 s 117d (1.0) 28'2 m ... 64 d (7.0) 272 d ( 6 . 0 ) . . . 76 d (7.0) 261 d (2.0) ... Chemical if it were a uenicillin. 0.02 ml of'DAISO-ds. CDClJ-
+
fully filtered through a Buchner funnel coated with talc and the catalyst was washed four times with 20-ml portions of H20. The catalyst was then suspended in 500 ml of 0.05 M NaOH solution and stirred for 0.5 hr and again filtered from the solution. The aqueous filt'rates were layered with EtOAc and acidified to p H 2 with 1 S HCI, and the organic layers were separated and dried (Pja,S04). Evaporation of the combined EtOAc solutions in vacuo left a colorless oil that weighed 2.18 g, an 88yc yield of crude desthiopenicillin. The product did not crystallize readily but tlc (silica; Et20-AcOH-H2OJ 18:3: 1) showed one major and one minor spot. The oil, dissolved in EtOAc (50 ml), was chilled in an ice bath and CHsNz in CH2C1z was added portionwise until the yellow color persisted. After the solution had stirred in the cold for 15 min, glacial AcOH was added, dropwise, until the yellow color of excess CH2N2 disappeared. The oily residue obtained after evaporation of the solvents in vaczco crystallized on standing. Tlc (silica; EtOAc) separated the material into one major spot and several very minor spots. The product' was therefore chromatographed over 70 g of silica. The eluting solvent progressed from C& to 30Cr Et0,4c-C& The residue after evaporation of solvents, which showed only one spot in tlc, was recrystallized from benzene-petroleum ether (bp 60-68" ), 250 ~ : 61.06; H, mg, mp 108". Anal. Calcd for C I ; H ~ ~ Y ? OC, 6.63: K,8.38. Found: C , 60.79; H, 6.75; N, 8.24. The product was recrystallized three times from benzenepetroleum ether and the melting point was raised to 109-110'. The chemical shift values for the protons in the nmr spectrum of VI are in Table I. 7-Phenoxyacetamido-3-methyl-3-cephem-4-carboxylic Acid (VIII).-A solution of 21.4 g (0.1 mole) of i-aminodesacetoxycephalosporanic acid" in a mixture of 800 ml of H20 and 600 ml of acetone containing 28 g (0.332 mole) of NaHC03 was cooled in an ice-alcohol bath and treated by dropwise addition with 17 g (0.1 mole) of phenoxyacetyl chloride in 200 ml of acetone. After 3 hr in the cold, the solution was evaporated in 2 " ~ oto remove acetone, layered with EtOAc, and acidified to p H 2.7 with concentrated HCI. The EtOAc layer was washed (HzO), dried (hIgS04), and evaporated in uacuo. The crystalline residue was recrystallized from AkzCO-Et20-C&,, yield 25 g (747,), mp 187-188" dec. Another sample made similarly had the same melting point. Anal. Calcd for CI~HI&O;S: C, 55.17; H, 4.63; N, 8.04. Found: C, 55.16; H, 4.77: X, 8.01.
After 1 hr in the cold the reaction mixture was allowed to warm to room temperature. Excess CHJTZ wab decomposed by careful addition of glacial AcOH. The residue that remained after vacuum evaporation of solvents was dissolved in C&, washed (5% NaHC03 solution, H,O), and after drying (MgSOa), recovered by evaporation. The product crystallized in an 8.85 g yield (717,), m p 137-139". The signals for an nmr spectrum are reported in Table I. Anal. Calcd for C1,Hl8N2OsS: C, 56.35; HI 5.01; N, 7.73. Found: C, 56.39; H I 5.09; N, 7.68.
7-Phenoxyacetamido-3-methyl-2-cephem-4-carboxy~ic Acid (X).'4-Methyl 7-phenoxyacetamido-3-methyl-3-cephem-4-carboxylate ( 2 g, 5.53 mmoles) in 50 ml of H 2 0and 50 ml of pyridine was chilled in an ice bath and 5.53 ml of 1 S S a O H was added. After 5 hr in the cold the solution was diluted (H20), layered with EtOAc, and while still cold acidified to p H 3 with ('oncentrated HCl. The ethyl acetate layer over u-ater was adju3ted to pH 8.0 and the water layer was separated, again acidified to pH 2.5, and extracted with ethyl acetate. The organic layer after being washed (H20) and dried (RlgS04)was evaporated zn vacuo to give the Al-acid which crystallized from chloroform-hexane, 1.45 g ( ~ E I ~mp ~ )167-169". , Its chromatogram on silica (EtzOAcOH-H~O, 15:3:1) showed one spot. The nmr chemical shift values for X are in Table I. Anal. Calcd for C1&N20sS: C , 55.17; H , 4.63; N, 8.04. Found: C, 55.28; H, 4.81; X, 8.08.
7-Phenoxyacetamido-3-methylcepham-4-carboxylic Acid (XI).
-A solution of 7 g (0.02 mole) of 7-phenoxyacetamido-3-methyl2-cephem-4-carboxylic acid in 700 ml of EtOH containing 14 g of 5% Pd-C was heated and shaken at 60" for 7 hr under 3 atm of Ht. After filtration to remove the catalybt, the product was isolated by evaporation of the solvent. Tlc (silica; EtzOAcOH-H20, 15:3:1) showed one maior spot and two more mobile, minor spots. Of the two minor spots, the slower had the same Rr value as starting material. The material corresponding to that in the forerunning minor spot crystallized from a warm CHsCN solution of the product mixture, mp 187-189a, 50 mg (XIa). An nmr spectrum showed this material to be one of the expected tetrahydrothiazine products, isomeric to the main product a t C-3. (See Table I.) Anal. Calcd for ClsHlsK20,S: C, 54.85; H, 5.18; N, 8.00. Found: C, 54.66; HI 5.21; N, 7.96. The nmr spectrum of the residue after MeCN had been evaporated (4 g, 58%) showed the presence of a small amount of starting material XI based on vinyl proton and &proton and Methyl 7-Phenoxyacetamido-3-methyl-3-cephem-4-carboxyl- 3-methyl proton integrals. The residue also contained isomer ate (IX).-7-Phenoxyacetamido-3-meth~.l-3-cephem-4-carboxylic XIa above as shown by its 3-methyl and 4 and 6-proton signals, acid (12 g, 0.035 mole) suspended in 300 ml of EtOAc and cooled in an ice bath was treated with CH,N, in CH2C1, by dropwise (14) IVe are indebted to Dr. R . B. Morin and Dr. B. G . Jackson for the directions tor this experiment. addit.ion until the acid disbolved and the solution became yellow.
