Journal of Medicinal Chemistry, 1971, Vol. 14, No. 7 641
NOTES
EtzO to give 2.3 g (77%): mp 185"; [ a l a 1 ~ -0.55" (e 0.92, comparable to those of the cofactor forms of vitamin DMF). Anal. (CaoHasClsNiOs) C, H, N. Be (pyridoxal or pyridoxamine phosphates) we synPoly(O-Me-Tyr-Glu-Ala-G1y)Gly.l -14C-OEt(l).-To a soln of thesized the chloro ketone V (Scheme I) and have also 1 mg (7.6 mmoles) of glycine-l-14C-OEt.HCl (specific activity attempted to obtain the lower homologs of this compd. 3.15 mCi/mmole) and 1.33 g (15.1 mmoles) of Et3N in 5 in1 of The latter were considered as potential irreversible DMSO was slowly added a soln of 2.3 g (2.9 mmoles) of the polymerizing unit 4 in 15 ml of DMSO. The transfer vessels were inhibitors, primarily for enzymes that have nonphoswashed with 9.5 ml of DMSO which was added to the reaction phorylated forms of vitamin Be as substrates. mixt giving a final concn of 100 mmoles/l. The reaction mixt As the starting materials for the syntheses of the was shaken for 6 days and then centrifuged to yield the product a-chloro ketones, we used the homologous carboxylic which was washed with three 36-m1 portions of HzO and three 35-mi portions of Et,O and dried to give the blocked polymer. acids (I, VI, XIII),4 which were protected with an This material was diasolved in 50 ml of 90% FICCOZHand stirred a4,3-0-isopropylidene group. The higher homologs for 1hr, and then concd under reduced pressure to yield the crude (I and VI) could be readily converted to the correpolypeptide 1. This material was washed with EtzO, suspended sponding acid chlorides by carefully reacting the acids in 20 ml of H 2 0 , and dissolved by the addition of 1 N NaOH to with SOClz in C ~ H Bbut , we could not obtain an acid p H 7.8. The soln was dialyzed against distilled H20 for 12 hr, acidified to pH 2.5 with HC1, and dialyzed for 3 days. The chloride from the lowest homolog XI11 by this propptd polypeptide was collected by centrifugation and then lyocedure. Variations of reaction conditions with different philized to yield 0.3 g (247,): radioassay, 35.2 nCi/mg indicates chlorinating reagents did not give the desired result.6 47y0 incorporation of the radioactive label. Anal. ( C ~ O H Z ~ N ~ O ~ * Addition of the acid chlorides I1 and VI1 to a large HzO) C, H, N. Mol Wt Determination.--Calibrated columns of Sephadex excess of CHzNz in Et20 resulted in the formation of the G-100 (2.5 X 38.5 cm) and of Corning CPG 10-240 glass granules diazo ketones I11 and VIII. The higher homolog (2.0 X 28 cm) were employed for the mol wt detn. Using 0.1 I1 also gave an appreciable amt of the a-chloro ketone M NaC1-0.05 iM KHzPO4 corrected to pH 8.0 as eluent, 4.2 mg of IV, which was isolated; the mixt of I11 and IV, howpoly(O-?*le-Tyr-Glu-Ala-Gly)Gly-1-14C-OE t was passed through ever, gave the desired a-chloro ketone V on treatment each of these columns. The polypeptide was eluted from each column in a vol equiv to that corresponding to a mol w t of at with HC1, as indicated by a positive Baker's test6 least 1 x 106. (a test for active halogen, applied as a tlc spray reagent), Immunachemical Results.-Two rabbits were treated a t weekly and by ir and nmr spectra. The lower homolog VIII, intervals with 500 pg of p~ly(O-?rle-Tyr-Glu-Ala-Gly)Gly-l-~~Con being similarly treated with HCl, gave a mixt of OEt 1. The first 2 weeks they were injected intradermally using complete Freunds adjuvant as suspension medium and the products, with a negative Baker's test. Treatment 3rd week they were injected sc. The injection on the 4th week of the diazo ketone VI11 with ethereal HC1 resulted in was done iv using buffered saline. Bleedings were conducted on the formation of the blocked a-chloro ketone IX, as the following week and the serum from each animal was found indicated by a positive Baker's tests. Mild hydrolysis to give a precipitin reaction with the antigenic polypeptide 1. of the protecting isopropylidene group gave exactly The preimmunized sera under the same conditions gave a negative precipitin reaction. The quantitative determination of the total the same mixt of products as was obtained directly precipitate was obtained by the addition of dils of poly(0-Mefrom the diazo ketone VI11 by treatment with HCl. Tyr-Gl~-Ala-Gly)Gly-I-~~C-OEt (1) to 2-ml samples of the The 2 products were separated by preparative tlc, pooled rabbit sera. The samples were inrubated a t 37" for 1 hr and the main product was shown to be the bicyclic hemiand then stood at 4" for 48 hr. The ppts were collected by centrifugation, washed twice with buffered saline, and collected. acetal X of the a-chloro ketone, and was further charThe total amount of protein precipitated was estimated by analyacterized as the diacetyl derivative XII. The minor sis for N (Kjeldahl). From these results the precipitin curve component of the reaction mixt was probably formed shown in Figure 1 was obtained.
,
Acknowledgments.-This work was supported by a grant from the National Science Foundation.
Analogs of Vitamin B6w i t h Reactive Groups' W. KORYTNYK* AXD B . LACHMANN Department of Experimental Therapeutics, Roswell Park Memorial Institute, Buffalo, New York Received iyovember 13, 1970
Pyridoxol analogs that have alkylating groups in appropriate positions may react irreversibly with cofactor sites of appropriate apoenzymes, or, more generally, with any of the receptor sites that are available for the various biologically active forms of vitamin BeS2j3I n order to obtain an analog with dimensions (1) Pyridoxine Chemistry. 25. (a) Preceding paper in this series: R. Korytnyk, H . Ahrens, and N. Angelino, Tetrahedron, 28, 5415 (1970); (b) a brief report of this study has appeared: 157th National Meeting of the American Chemical Society, Minneapolis, Minn., April 1969, M E D I 11. (2) For a review of the synthesis and biological activity of vitamin BB analogs see W. Korytnyk and M. Ikawa, Methods Enzymol., 18A, 524 (1970). (3) Two vitamin Be analogs that have bifunctional alkylating groups in the 5 position have been described earlier and found t o have weak antitumor
by dehydration of the major compd X, and it was indicated by nmr to be XI (see Experimental Section). The small amt of the sample precluded the further work necessary for making an unequivocal structural assignment. The a-chloro ketone V was tested as an inhibitor of several enzymes. It did not inhibit apotryptophanM ; but when the concn was raised to ase7v8a t 5 X 4X M , 99% of the enzyme activity was abolished. After being filtered through a Sephadex G-25 column,
effects: N~,N~-bis(Z-chloroethyl)-O'-methyliaopyridoxamine (methoxypyridoxyl N mustard) and N~,N~-bis(Z-ohloroethyl)-4-deoxyisopyridoxamine (4-deoxypyridoxyl N mustard) [C. C. Stock, 9. Buokley, K. Suguira, a n d C. P. Rhoads, Cancer Res. 11, 432 (1951)l. (4) (a) W. Korytnyk, E. J. Kris, and R. P. Singh, J. Ore. Chem., 29, 574 (1964); (b) W.Korytnyk, J. Med. Chem., 8 , 112 (1965); (0) W. Korytnyk, B. Paul, A. Bloch, and C. A. Nichol, ibid., 10, 345 (1967). (5) Attempted preparation of the acid chloride XIV from the acid XI11 according t o the method of I. Tomita, H. G. Brooks, and D. E. Metsler [ J . Heterocycl. Chem., 3, 178 (1966)l, gave the hydrochloride of the starting material XIII. Reaction with oxalyl chloride w a s also negative. Compare the unsuccessful efforts a t obtaining acid chlorides from pyridine carboxylic acids by E . Wenkert, F. Haglid, and S. L. Mueller, J. Oro. Chem., 34, 247 (1969). (6) B. R. Baker, D. V. Santi, J. K. Coward, H. S. Shapiro, and J. H. Jordaan, J. Heterocycl. Chem., 3, 434 (1966). (7) (a) Crystalline apotryptophanase was prepared according t o the method of 1%A. '. Newton, Y. Morino, and E . E . Snell, J . Biol. Chem., 240, 1211 (1965); (b) compd XV was synthesized by C. I w a t a a n d D. E. Metsler, J. Heterocycl. Chem., 4 , 319 (1969). (8) The assay method was t h a t of W. A. Newton and E. E. Snell, P r o c . Nat. Acad. Sei. U.S.,48, 1431 (1962).
612 Journal of Medicinal Chemistry, 1971, Vol. 14, N o . 7'
the enzyme had, in addition t o the protein uv-absorption peak at 280 mp, a small peak a t 325 mp for the cofactor analog, indicating covalent binding to the enSCHEME I P C HI ,
p C HI 2
o&CH2CHzCOzH
I c1-
xn1
H XJY
of)--CHzCH2C0Cl
VHO CH3
CH,
?+
Ho~CH2CHzC02H CH,
xv
zyme. I n this connection it is of interest that a similar compd (but without a reactive group), p-(2-methyl-3hydroxy-4-formylpyridin-5-yl)propionicacid (XV), was found to be a weak inhibitor of this enzyme.' This enzyme apparently requires an analog with a strongly ionizable group in the 5 position, such as the phosphate group, to be an effective inhibitor. Such a group is absent in V, and even the &carboxylic acid group in XV apparently cannot effectively simulate the pliosphate group of the cofactor. Pyridoxamine-oxaloacetic transaminaseg was not inhibited by the analog V a t 4 X low2A I , nor was pyriCHZOH doxal phosphokinaseT0inhibited a t 2 X M . Pyridoxine phosphate oxidase from rabbit liver was also Ho~CHzCH2COCHzCl found to be uneffected at 31. The compd inhibSaccharomyces carlsberyited half-maximal growth of CH3 N+ ensis at 5 X AI; under similar conditions, the IH c1organism was inhibited by 4-deoxypyridoxol at, 5 X v 10-6 A i . 1 1
IV
Y \ C HI ,
) i O ' C HI2
Experimental Section
I c1-
VI
H
VI1
VI11
"
/
I c1-
H
X:z'
IX OH
OH
X
OAc
x I1
Where analyses are indicated only by symbols of the elements, analytical results obtained for these elements were within +0.4y0 of the theoretical values. Tlc (silica gel) was used routinely as described earlier.', I r spectra were determined with a PerkiiiElmer 457 spectrophotometer, arid iimr spectra with a Variari A60A instrument, as S-13% solns in CDCL, DMSO, or D2O; positions of peaks are expressed in cps from ThIS, or from sodium 3-(trimethylsilyl)-l-propanesulforiic acid, as an iriteriial std. Peaks were assigned 011 the basis of previous work."' a4,3-0-Isopropylidene-a5-pyridoxylacetylChloride Hydrochloride (II).-Compd I (1.37 g, 3.5 mmoles)*~was suspended . SOClr (1.2 g, 10 mmoles) in CsH, ( 3 ml) to the stirred 5uspensioii, and stirring w:ts room temp and for another 30 miri at 30". After cooling, the mixt was filtered, and the ppt was washed with 1790 cm-'. The dry CaH6. The yield was 1.5 g (91%); product was used for the react,ion about to be described, without further purification. 1-( a4,3-0-Isopropylidene-a6-pyridoxyl)-3-diazo-2-propanone f 111) and 3-Chloro-l-( ~~~,3.O.i~0pr0pylidene-a~-pyridoxyl)-2-propanone (IV).-A suspension of I1 (1.5 g, 4.9 mmoles) in Et20 (10 ml) was slowly added, for approx 20 min, to a stirred and cooled (ice-salt mixt) soln of C H ~ N (36 Z mmoles, EtOH free) ill Et20 (160 ml). After standing for 12 hr, the solti was filtered and the filtrate was evapd to 15 ml arid kept at 2' overnight. The deposited crystals (830 mg) consisted of I11 and IV, which
XI
(9) The enzyme was prepared by J. E . Ayling and E. E. Snell, Biochemistry, 3, 1616 (1968), and assayed by the method of H . Wada and E. E. Snell, J. B i d . Chem., 991, 127 (1962). (10) From rat liver, as described by D. B. McCormick, Vi. E. Gregory, and E. E. Snell, ibid.. 2S6, 2076 (1961). (11) Dr. A. Bloch. personal communication. (12) H. Ahrens and 1%'. Korytnyk, Methods Enzymol.. 18A, 489 (1970). (13) W. Korytnyk and H. Ahrens, i b i d . , ISA, 475 (1970).
NOTES were sepd on a silica gel column (1 x 45 cm) and eluted with EtOAc; 100 mg of the mixt gave 20 mg of the a-chloro ketone IV (mp 129'), which was eluted first (indicated by positive Baker's test) and 75 mg of the cy-diazo ketone 111 (mp 69'). ~ (diazo), ~ ' 1655 Data for diazo ketone 111 were: ir A ~ 2125 cm-1 (C=O); nmr (CIIcb) 142 (%CHI), 228 (4-CHz), 155-165 (m) (,5-CHzCH2), 312 (COCHNZ), 92 (CHI, isopropylidene), 248 mp ( e 13,200), 277 (sh, 8600). Anal. 468 (Ce-H); uv A":;,.", (CiiHi7N303) C, H, N. Data for a-chloro ketone IV were: ir: A":' 1740 cm-1; nmr (CDCL) 142 (2-CH3), 289 (4-CHz), 470 (Ce-H), 92 (CH3, isopropylidene), 168-177 (m) (.5-(CHz)z), 242 (5-COCHzCl). Anal. (C14HlsNC103)C, H, C1, N. 3-Chloro- 1-id-pyridoxyl)-2-propanoneHydrochloride (V 1.To the crude CHzNz reaction product (111 and IV, 200 mg), dissolved in EtzO (10 ml), 1 g of concd aq HCl was added within 20 min, and the mixt was stirred at room temp. After standing for 3 hr, the solvent was evapd in vacuo, and the oily residue was taken up in a small amt of MeOH and shaken with Ihrco. After filtration and evapn of the soln, a small amt of MeCN was added till turbidity developed and let crystallize. The yield was 135 mg (GO%), mp 149". The compd gave a positive Baker's test? nmr (DMSO-de) 157 (2-CH3) 289 (4-CHz), 180 (5-(CH2)2), 488 (Ce-H), 273 (COCHZCI). Anal. (CllHl~ClzNO~) C, HI C1, N. 2,2,8-Trimethyl-4H-3-dioxine [4,5-c]pyridine-5-acetyl Chloride Hydrochloride (VII).-To a stirred suspension of VI (500 mg, 2.1 mmoles) in CH&N ( 5 ml), SOCl, (600 mg, 5 mmoles) was added dropwise in ca. 5 min. After stirring for 15 min a t room temp, the mixt was heated to 50' and was kept at this temp for 30 min. The cooled soln was filtered, and the filtrate was evapd to dryness. The residue crystd after being refluxed with dry Me2C0. The yield was 350 mg (57y0): mp 210-212' dec; 1805 cm-1 (C=O). ir 2,2,8-Trimethy1-4H-3-dioxino [4,5-c]pyridine-5-(3-diazo-2propanone) (VIII).-The acid chloride VI1 (380 mg, 1.3 mmoles) was suspended in EtzO ( 5 ml), and the suspension was added drop by drop to a stirred CH2N2soln (8-10 mmoles, alcohol free) cooled to - 15' with an ice-salt mixt. The soln was filtered to remove a small amt of tarry material. Tlc (EtOAc) of the filtrate showed only 1spot. After keeping for 45 min at room temp, the reaction mixt was evapd to dryness, and the product was crystd from EtlO-petr ether, yielding 275 mg (81%) of pale yellow crystals: 2110 cm-l (Nl), 1630 cm-l (C=O); nmr (CDCb) mp 70"; ir ::A: 144 (2-CH3) 92 (CH3, isopropylidene) 288 (4-CH2), 208 (5-CHz), 314 (COCHN2),474 (Ce-H). Anal. (C13H15N303)C, H, N. 2,2,8-Trimethyl-4H-3-dioxino [4,5-c]pyridine-5-(3-chloro-2propanone) Hydrochloride (IX).-The diazo ketone VI11 was prepd, as just described, from 380 mg (1.3 mmoles) of the acid chloride VII. The filtered ethereal s o h of VI11 was evapd to a small v01, and the latter was slowly added to a slight excess of ethereal HC1 soln (dry), with stirring. The reaction mixt was stirred for another 15 min, and was kept at 2" overnight. Filtration and washing with a small amt of dry MezCO yielded 240 mg (60y0)of IX, mp 205' (from MeZCO). I t gave a positive Baker's 1723 cm-'; nmr (DMSO-de) 154 (2-C&), 93 test? ir ::A: (CHa, isopropylidene), 285 (COCHZCI), 250 (~-CHZCO),298 (4-CH2), 493 (C6-H). Ana!. (ClaH1&1~N03)C, H I N. 3-(Chloromethyl)-7-methyI-1,4-dihydropyrano [4,3-c]pyridine3,8-diol ( X ) and the By-product (XI).-Compd I X (61 mg, 0.2 mmole) was dissolved in 0.2 N HCl(7 ml), and the soln was stirred a t room temp for 25 hr. Tlc indicated the formation of a new compd (Rr 0.42 in 80:20 CHCL-MeOH), giving a pos Gibbs test. The solvent was evapd to dryness, and the residue was taken up in MeOH and again spotted on tlc. I n addition to the previous spot, another spot (Rr 0.72) was obtained, which was also Gibbs pos. The 2 products were separated by prep tlc. The compd with the lower Rr value (0.42) was extd from the tlc scrapings with MeOH. The MeOH soln was evapd, and the residue was treated with MezCO, giving 25 mg (54%) of product (X), mp 189 (from biezCO-hIeOH). Baker's test on the compd was negative and the compd was not retarded by boric acid strip on tlc plate,12 indicating that the 4-CHzOH group is not free. Its ir spectrum shows no CO absorption; nmr (DMSO-&,) 141 (7-CH3), 466 (Cb-H), 285 (Cl-Hz), 221 (C4-HZ)t 170 (3-CHzC1) (doublet, J = 3 cps). Anal. ( C ~ O H ~ Z C ~C,NH, O ~N, ) C1. The by-product of high Rr value (0.72) was isolated from the plate, but the small amt of material obtained (15 mg) was not adequate to establish the structure unequivocally as 3-(chloro-
Journal of Medicinal Chemistry, 1971, Vol. 14, No. Y
643
methyl)-7-methyl-lH-pyrano[4,3-c] pyridin-8-01 (XI) :
ir 1640 cm-1 (C=C); nmr (DMSO-de) 141 (CHI), CHZ groups (singlets) a t 257 and 312, 1 H peaks a t 367 and 462 cps. The by-product is formed directly from compd X by treatment with 0.2 IV HC1. Tlc of the product indicated a mixt of X I and X after 1 day at room temp. I t was impossible, however, to achieve complete conversion of X to XI. Likewise a mixt of X and X I was obtd when the diazo ketone VI11 was treated with 3894 HCl. 3-(Chloromet hyl)-7-methyl-l,4-dihydropyrano[4,3-c]pyridine3,S-diol Diacetate (XU).-Compd X (35 mg, 0.15 mmole) was dissolved in a 4: 1 mixt of pyridine and Ac20, and the resulting mixt was kept a t room temp for 3 days. It was evapd in vacuo, treated with an NaHC03 soln, and extd with EtzO. After drying (MgS04), the EtOAc was removed in vacuo, and the residual oil was dissolved in Et20-petroleum ether. The yield of cryst material was 25 mg (53%): mp 112-113' (from EtzO-petr ether); ir 1740, 1760 cm-1 (C=O); nmr (CDCla) 494 (5-H), 142 (7-CH3), 139 (8-OCOC&), 118 (3-OCOC&), 118 (1-Hz) (s), 254 (3-CHzC1) (d, J 11 CPS), 239 (d, J = 11 CPS), 211 (4-Hz) (d, J = 17 cps) 181 (d, J = 17 cps). Anal. (ClrHleNCIOa) C, HI N.
Acknowledgments.-This study was supported in part by research grants (CA-08793 and CA-11047) from the National Cancer Institute, U. S. Public Health Service. We would like to thank Dr. A. Bloch under whose supervision the microbiological testing of compound V was performed. One of the authors (W. K.) is greatly indebted to Dr. Esmond E. Snell for an opportunity to carry out experiments with enzymes in his laboratory under his guidance. Tests on pyridoxine phosphate oxidase and pyridoxal phosphokinase were carried out by Mr. Norman Angelino at this laboratory. Microsomal 3-Hydroxylation of 1,4-Benzodiazepines WOLFGANG S A D ~ WILLIAM E,~ GARLAND, AND NEALCASTAGNOLI, JR.* Department of Pharmaceutical Chemistry, School of Pharmacy, University of California, Sun Francisco, California 941.98 Received December 8, 1370
Liver microsomal oxidations of a variety of substrates are known to be mediated by a mixed function oxygenase system which utilizes molecular oxygen and requires NADPH as a reducing e q ~ i v a l e n t . ~I n the case of tertiary amines 1 it has been proposed that microsomal oxygenation leads to the formation of a carbinolamine 2, which, because of its inherent instability, decomposes spontaneously to the observed products, the secondary amine 3 and the aldehyde 4.4 Evidence consistent with this pathway was recently reported by McR/lahon5 who studied the incorporation of l*Oenriched O2 into benzaldehyde formed from the microsomal oxidative dealkylation of 1-benzyl-4-phenyl-4 carbethoxypiperidine. I n order to minimize exchange (1) Presented in part a t the 160th National Meeting of the American Chemical Society, Chicago, Ill., Sept 1970. Financial support from U. S. Public Health Service Grant G M 16496 is gratefully acknowledged. (2) NATO Postdoctoral Research Fellow, 1969-1970. (3) A recent symposium on this subject has been published: "Microsomes and Drug Oxidations," J. R. Gillette, Ed., Academic Press, New York, N. Y.,1969. (4) J. R. Gillette, Aduan. Pharmacol., 4, 219 (1966). ( 5 ) R. E. McMahon, H. W. Culp, and J. C. Ocoolowts, J . Amer. Cham. Soc., 91, 3389 (1969).
