PETERB. RUSSELL, GERTRUDE B. ELIONAND GEORGEH. HITCHINGS
474
[CONTRIBUTION FROM
Vol. 71
THEWELLCOME RESEAFXH LABORATORIES]
A New Condensed Pyrimidine System: Some p-Oxazino [2,3-d]pyrimidines’ PETERB. RUSSELL, GERTRUDE B. ELIONAND GEORGE H. HITCHINGS The synthesis of a new isomer of dihydroxanthopterin (11) by the ring closure of 5-chloroacetamido-2,4diamino-6-hydroxypyrimidine(I) with sodium bicarbonate solution has been reported.2 OH H I
OH
Xfl\JX
I
\&O
I1
Similar compounds (VI, VI1 and VIII) were formed from the 2-amino-, 2-methylamino- and 2dimethylaminopyrimidines corresponding to IV, and two 4,7-dimethyl-p-oxazino[2,3-d]pyrimidines, IX and X, were prepared from the appropriate 5-a-bromopropionamido derivatives. The reaction failed to occur with the 2-methyl analog of IV or with 5-(a-~hlorophenylacetamido)-4methyl-2,fbdihydroxypyrimidine. CHI H N,
s-.J
X=NHz X = CH3PU” VI11 X = ICH,)& IX X = OH X X=NHz VI
\‘I1
R = H R =H
‘OH R =H During the investigation of methods for closure R = CH, of the dihydropyrazine ring other alkaline reX R=CHI agents were studied, and barium hydroxide was found t o give a product whose ultraviolet absorpIn order to prove the structure of these subtion spectrum was different from that of I, I1 or stances, compound V was chosen for more intenthe a-dihydroxanthopterin of O’Dell and co-work- sive study. The absorption spectrum of V (Fig. 1) e r ~ .Although ~ this new compound was not iso- was found to have two bands, one of them in the lated in pure form, the possibility that ring closure region usually associated with the pyrimidine ring, involving the 6-hydroxyl group had taken place the second in the higher wave lengths. Hence, was suggested. By analogy with the formation of the spectrum was consistent with the formulation 3-hydroxy- [benz-$-oxazine] from o-chloroaceta- of the compound as a condensed pyrimidine sysmidophenol by treatment with potassium hydrox- tem. Hydrolysis of V with concentrated hydroide solution, i t appeared probable that the product chloric acid at 100’ resulted in recovery of the obtained by the action of barium hydroxide on pyrimidine (111) showing that there had been no I was a member of the hitherto unknown group of structural changes in the original pyrimidine nusubstances, the $-oxazino [2,3-d]pyrimidines. cleus during treatment with barium hydroxide. I n order t o eliminate the possibility of pteridine It also ruIed out the possibility that one of the formation, 5-chloroacetamido-6-hydroxypyrimi-nitrogens of the pyrimidine ring may have been dines were prepared which did not carry amino alkylated either inter- or intramolecularly by the groups at position 4 and these were heated with chloroacetyl group with a subsequent hydrolysis aqueous barium hydroxide. After failure to iso- of the amide linkage or linkages. Hydantoin late any of the desired compounds from 5-chloro- formation would give a product with the observed acetamidouracil or from 5-chloroacetamido-2- empirical composition; however, the pyrimidineamino-6-hydroxypyrimiidine, an experiment with 5- N-acetic acids do not hydantoinize. Moreover, chloroacetamido- 2,6- dihydroxy -4- methylpyrimi- such a pyrimidine-N-acetic acid derivative would dine, (IV) produced a colorIess, crystalline com- not be split to the parent pyrimidine 111under the which was believed to be 2,6- rather mild conditions employed. pound, C7H703N3, dihydroxy-4-methyl -p-oxazino[2,3-d]pyrirnidine Additional evidence in favor of the p-oxazino(17). pyrimidine structure was obtained when V was CH1 heated with concentrated ammonium hydroxide at 120” in a closed system. The product obtained gave an analysis corresponding to the formula C7Hx02N4and had an absorption spectrum resembling that of a dihydropteridine (Fig. 2 ) , cf. 111 ref. ( 2 ) . I t is therefore considered to be 2,6-diliydroxy-4-methyl-7,8-dihydropteridine(XI). CH; 1 ; :
---
(1) Presented before the Chicago, Ill., meeting of the American Chemical Society, April, 1948. 12) Hitchings and Elion, THISJOURNAL, 71, 467 (1949). (3) O’Dell, Vandenbelt, Bloom and Pfiffuer. ihid.. 69, 250 (1947) c l \ Asrhan Brr ZOs 1624 (18811
FI XI
XI1
There remained one further possibility to be ( 5 ) \%heeler and 1,iddle Tars TOURNAL.
80, 1164 (1908)
Feb., 1949
of compounds are so widely different as to leave little doubt that this oxazolo structure can be eliminated from consideration here. In view of the limitations of this method for the synthesis of p-oxazinopyrimidines, i t was decided to investigate further the mechanism of the ring closure. The oxazines are formed as a result of reaction A, but the low yields indicate that other reactions, of which B and C are the most probable, are competing with it.
14,000
h,-
2000
--
considered before the structure of V could be regarded as unequivocally established, namely, that a hydroxymethyloxazolo [2,3-d]pyrimidine of type XI1 had been formed. The mode of formation,6 the structural requirements for ring closure, and the physical properties7of the two groups
Y
I
I 6-
4"
2600 3000 3400 Wave length in A. Fig. 1.-Ultraviolet absorption spectra of 2,6-dihydroxy-L-methyl-p-oxazin0[2,3d]pyrimidine: -, at PH 1.0; -, a t pH 11.0.
\\
NHCOCHzOH
+
9
8
8oool 6000
'
AO/CH2
B
\
A O H
1
v " 2
f HOOCCHzCl
A O H
The extent to which reaction A proceeds would be expected to depend on the ease of ionization of the 6-hydroxyl group, and the acid strength of this group would be governed by resonance of the type shown CHI
HNANHCOR * x=,Np-oI /
7
' Y I !
475
SOME P-OXAZINO [2,3-d]PYRIMIDINES
,$'
HN \ "COR -X-!,=O
Thus a greater excess of barium hydroxide is necessary to obtain the maximum yield of (VI), X = NH, than is necessary for (V), X = 0 (Fig. 3). Where X is a group incapable of taking part in resonance of this type (e. g., methyl), the formation of the oxazine would be expected only a t com-
-
.,. I
2600 3000 3400 Wave length in d. Fig. 2.-Ultraviolet absorption spectra of 2,6-dihydroxy-4-methyl-7,8-dihydropteridine:- - - at PH l.0; -, a t pH 11.0. 1
(6) Johnson, Am. Chcm. J . , 34, 203 (1905). (7) Falco and Hitchings, forthcoming publication.
5t 1 2 3 4 5 Equivalents of barium hydroxide. Fig. 3.-Relationship between barium hydroxide concentration and yield of: -0-, 2-amino-6-hydroxy-4methyl-@-oxazino[2,3-d]pyrimidine; - - -El- -, 2.6-dihydroxy-4-methyl-@-oxazino[2,3-d]pyrirnidine.
-
I'ETER13. RUSSELL,GERTRUDE B. ELIONAND GEORGE H. HITCHINGS
476
Vol. 7 1
hours. After precipitation of the barium as sulfate, the solution was filtered and acidified with acetic acid. T h r oxazinopyrimidine crystallized slowly (0.08 g.) . \Vlieii recrystallized from water it fornied stout colorless needles, m . 1). :138-340" (dec.). , .1m/. Calcd. for C I H ~ O ~ I TC, ~ : 49.2; H, 4.6; 3, 21.4. l ~ o u n d :C, 49.3; 1-1, 4.6; hT,21.4. 5-Benzeneazo-2 -amino-6-hydroxy-4-methylpyrimidine. --Guanidine hydrochloride (19.2 g.) was added to a solution of 44 g. of ethyl benzeneazoacetoacetateg in 250 ml. of absolute alcohol. The mixture was stirred while a sodium Experimental ethoxide solution (9.2 g. of sodium in 200 ml. of absolute 5 -Chloroacetamido-2,6-dihydroxy-4-methylpyrimidine alcohol) was added over three and one-half hours. After (IV) .-An intimate mixture of 0.5 g. of 5-amino-2,6-di- the mixture had stood two days a t room temperature, the hydroxy-4-methylpyrmidine8 and 2 g. of chloroacetic acid orange solid was filtered off and dissolved in 1liter of water. was heated in a boiling mater-bath under vacuum for two The solution was acidified with acetic acid and the prehours. The cooled mass was extracted with ether, washed cipitate collected. After washing with 1 liter of 5% sowith sodium bicarbonate solution and recrystallized from dium bicarbonate solution and then with water, it w i i s 50% aqueous ethanol. It formed colorless needles, m. p. dried a t 100" (30 g.). On recrystallization from pyri257-258" (dec.), (0.75 g., Y8$$l,). dine, it formed orange-red prisms, m. p. 274" (dec.). .4nal. Calcd. for C?IT8O3N,C!: C, 38.6; H 9 3 . 5 . .Inul. Calcd. for CIiH11ONa: C, 57.6; H, 4.8. Found: Found: C, 38.9; H, 3.7. C, 57.9; H, 4.8. 2,6-Dihydroxy-4-methyl-p-oxazino[ 2,3-d]pyrimidine 2,5-Diamino-6-hydroxy-4-methylpyrimidine .-The (V) .-~-Chloroacetamido-2,6-dihydroxy-4-methylpyrimidine (0.5 g.) was heated with 30 ml. of a solution con- above azo compo4nd (3.5 9.) was suspended in 100 ml. 0: 50% aqueous ethanol and reduced catalytically a t GO taining 1.12 g. of barium hydroxide hydrate (3 equivs. of using a platinum catalyst and two to three atmospheres of Ba(OH)&H20). After about one-half hour a crystalline barium salt separated ; heating was continued for hydrogen. Reduction was complete after three hours. another hour. The warm solution was treated with an The solution was filtered and evaporated to tiyness on a steam-bath under diminished pressure. The residue wa excess of sodium sulfate, the barium sulfate filtered off and the solution made faintly acidic with acetic acid. On washed with acetone t o remove aniline, and the almos standing, the oxazinopyrirnidine separated as colorless colorless residue recrystallized from water (1.0 g.) . The p. was 280". This compound had previously been prisms (0.10 g., 30v/o). Oii crystallization from water, in. prepared by Jaegerioin poor yield from the corresponding i t formed elongated prisms, m. p., 348-319' (dec.). 5-bromo compound and ammonia and reported t o h a w Anal. Calcd. for C7H70JN3: C, 46.4; I€, 3.9: X, 23.2. a In. p. of 2S50e Found: C, 46.5; H,3.7; N,23.5. Anal. C;tlcd. for CbHsON4: C, 42.9; H, 5.7. Pounti: Hydrolysis of V with Hydrochloric Acid.-One hundred C, 43.0: H , 5.8. mg. of V was dissolved in 5 ml. of concentrated hydro5-Chloroacetamido-2-amino-6-hydroxy-4-methylpychloric acid and the solution was heated on the steamabove diaminopyrimidine (0.5 g.) wab bath for twenty hours. 'The hydrochloric acid and water rimidhe.-The were removed by evaporation to dryness several times in an heated with 2 g. of chloroacetic acid in a boiling water-bath open dish. The residue was dissolved in 2 ml. of water under vacuum. After two hours, the mixture was cooled and saturated sodium acetate solution mas added until the and the excess chloroacetic acid dissolved in ether. Thc pH was about 7. Slightly yellow plates separated (80 mg.) residue w.is recrystallized from 30% aqueous ethanol and with a m. p. of 257' (dec.), Mixed with an autheiitic formed colorless needle
