COMMUNICATIONS TO THE EDITOR
Nov. 20, 1954
methyl cytosine (HMC) nucleoside, in the face of good yields of the deoxyribonucleosides of the other three bases, was reported by Cohen,2 while the seemingly unrelated work of Jesaitis and Goebe13 indicated the presence of a significant quantity of glucose in T 4 DNA. With the notion that the presence of glucose in the DNA molecule somehow inhibited the action of the combined enzymes, we have attempted to separate the products of such enzyniatic hydrolysates by ion-exchange and to identify them. Such experiments have led us to the conclusion t h a t glucose is associated mole for mole with HMC in T4 DNA and, further, that the glucose is probably linked as an 0-glycoside to the 5-hydroxymethyl group of HMC, rather than substituted for deoxyribose in the DNA sugar-phosphate chain. We have extended the observation of Jesaitis and Goebel of the presence of glucose in T 4 to T2 and T 6 DNA’s (but not T1 DNA) and have noted that the ratio of glucose to DNA phosphorus is about 0.17 in these phages. The course of hydrolysis of 40 mg. of highly purified T4* DNA by pancreatic DXase was followed by electrometric titration. A maximum of 11% of the phosphoryl linkages was hydrolyzed by the enzyme, this value being about half of that obtained by a similar digestion of DNA from other source^.^ Half of the digest was hydrolyzed with whole Crotalus adamanteous venom (containing both diesterase and 5’-monoesterase) and the other half was hydrolyzed with a diesterase preparation essentially free of monoesterase activity.6 Both hydrolyses were again followed by titration and reached an end-point when an additional 70y0 of the phosphoryl linkages of the DNase-produced polynucleotides was broken. With calf thymus DNA such procedures yield nucleosides and 5’-nucleotides, respectively, in quantitative yield.6 With T 4 DNA4, however, many polynucleotides remain. The ion-exchange analyses7 of the two digests are summarized in Table I. It is clear that the diesterase does not cleave most of the HMC nucleotide linkages. However, a variety of nucleo-
5893
sides, nucleotides and polynucleotides containing HMC was obtained. Within each such compound or mixture of compounds, HMC was determined by spectrophotometric observations and glucose by the anthrone test.s Wherever HMC was present, glucose was also present in equimolar quantity. I n addition, there was a t least one HMC residue per polynucleotide. Glucose was identified as such after liberation from T 4 DNA by boiling for 7 hours with a sulfonic acid cation exchanger (Dowex-50-H). The identification was made by anion-exchange chromatography in a borate ~ y s t e mand , ~ by the hexokinasefirefly luminescence system. lo These analyses agree with each other and with the colorimetric anthrone method8 in indicating about 90% release of the glucose as such. Indirect evidence on the mode of linkage of glucose to HMC in T 4 was given by conventional acid hydrolysis (1 N HCI, 1 hour, 100°), which yielded 88% of the glucose as free sugar whereas 87% of the HMC remained in mononucleotide, nucleoside diphosphate and polynucleotide form. From the latter observation, glucose cannot be the sugar involved in internucleotidic linkage of HMC. Furthermore, its appearance without simultaneous HMC destruction is more consistent with the properties of 0-glycosides than with those of pyrimidine-N-glycosides, which are markedly acidstable. The sugar is therefore tentatively allocated to the 5-hydroxymethyl group. NOTEADDED IN PROOF.-Recently a report of similar findings with T2 bacteriophage DNA has appeared by Sinsheimer, Science, 120, 551 (1954), dealing with the identification of glucose associated with BMC mononucleotide. Our results confirm this and extend the observation t o the polynucleotide-bound BMC. ( 8 ) D. L. Norris, Science, 107, 254 (1948). JOURNAL, 74,2090 (1952). (9) J. X.Khym and L. P. Zill, THIS (10) B. L. Strehler and J. R. Totter, Arch. Biockcm. and Bioghys., 40, 28 (1952).
OAKRIDGENATIONAL LABORATORY OAKRIDGE, TENN. ELLIOTVOLKIN RECEIVED OCTOBER 4, 1954
CI4 AND C1* INTRAMOLECULAR ISOTOPE TABLE I EFFECTS I N THE DECARBOXYLATION O F LIQUID OF THE BASES OF T4 DNA B Y ION-EXMALONIC ACID AT 140.5’ PERCENTRECOVERY CHANGE ANALYSISFOLLOWING THE COMBANED ACTIONOF Sir: PANCREATIC DNASE A N D (a) WHOLE SNAKE VENOM AND The theoretical value for the ratio of CI4 to CI3 (b) SNAKE VENOMDIESTERASE isotope effects in reactions involving rupture of Base
(HMC)~ Guanine Thymine Ad enin e
(a) Whole venom Per cent recovered as Mononucleotide Polynucleotide
hTucleoside
4 79 72 78
12 0 0 0
84 21 27 21
(b) Venom diesterase
(HMC)~ .. 17 82 Guanine 8 74 18 Thymine 9 64 26 Adenine 8 72 19 a Nucleoside formation with diesterase is ascribed to resi5-hydroxymethylcytosine. dual 5’-monoesterase action. (4) R.M . Herriott and J. L. Barlow, J. Gen. Physiol., 36, 17 (1952). (5) M. Kunitz, ibid., 39, 363 (1949-1950). (6) R.0.Hurst and G. C. Butler, ibid., 193, 91 (1951). (7) E.Volkin and W. E. Cohn,i b i d . , 206,767 (1953).
C-C, C-N or C - 0 bonds is near 2.’ Reported experimental values are : decarboxylation of mesitoic acid2 -2.8 f 0.3; dehydration of oxalic acid3 -2.0 f 0.2; enzymatic hydrolysis of urea4 -3.2 f 0.4; non-enzymatic hydrolysis of urea5 -1.8 f 0.2; decarboxylation of malonic acid6 -3.5 f 0.4. Only for malonic acid are other results available for the C14 isotope effect; Roe and Hellman’ gave a figure of 6 f 2% for the intramolecular (1) J. Bigeleisen, J . Phys. Chem., 66, 823 (1952). (2) W. H. Stevens, J. M. Pepper and M. Lounsbury, J . Chcm. Phys., 40, 192 (1952). ( 3 ) A. Fry and M. Calvin, ibid., 66, 901 (1952). ( 4 ) J. A. Schmitt, A. L. Myerson and F. Daniels, ibid., 66, 917 (1952). JOURNAL, 76, 3564 (1953). (5) J. A. Schmitt and F. Daniels, THIS (6) P.E.Yankwich and E. C. Stivers, J . Chcm. Pkys., 21, 61 (1953). (7) A. Roe and M. Hellman, ibid., 19,660 (1951).
5894
COMMUNICATIONS TO THE EDITOR
Vol. 76
THE SYNTHESIS OF SPHINGOSINE effect, a value which stands in approximate theoretical agreement with the 2.92 f 0.07% result for Sir: the related CI3 effect according to a recent study.8 The structure of sphingosine (I) was conclusively Since the isotope effect ratio 2 is a clear prediction established by Carter and his co-workers.' Eviof the absolute rate theory, we have attempted to dence for the trans configuration of the double bond establish more firmly whether the malonic acid case has been presented recently by other investigadoes deviate by redetermining the CI3and C14intra- t o r ~ . ~We , ~ wish to announce the synthesis of molecular isotope effects, employing only mass trans - DL - erythro- 1,3-dihydroxy-2-amino-4-octadecspectrometric measurements. The results given ene (I). below are preliminary. Ethyl a-(trans-2-hexadecenoyl)-acetoacetate (II) Labeled malonic acid was synthesized v i a the (m.p. 33-34') was obtained in a 75% yield from cyanide-cyanoacetate route employed previously.6 the acid chloride and ethyl acetoacetate. Anal. The crude malonic acid was purified on a silicic Calcd. for C22H38O4: C, 72.1; H, 10.4. Found: acid column, using the n-butanol-chloroform sol- C, 72.2; H, 10.6. The Japp-Klingemann revent systemg; after preliminary concentration, action4 of (11) with benzenediazonium chloride the product was freed of solvents by sublimation afforded a 60-70% yield of the a-phenylazo ester in vacuo a t 95-100'. This material was diluted (111), m.p. 39-41'. Anal. Calcd. for C~H40N203: to the desired isotopic composition (about 1% C, 72.8; H, 9.3; N, 6.5. Found: C, 72.9; H, CI4, total carbon basis) with Eastman Kodak Co. 9.5; x, 6.2. b-hite Label acid which had been subjected to the CHI(CH~)Y~-CH=CH-CH-CH-CH~~H same manipulations. The mixture was recrystalI lized from acetone-benzene, resublimed, then OH AH2 (1) stored a t room temperature in a vacuum desiccator. CHI(CHZ)I~~!H=CH-CO-C-COOCZH~ The batch decomposition, combustion and deII X-XH-CtjHL (111) gradation techniques employed have been described in detail elsewhere8; all samples of carbon Reductive acetylation with zinc in acetic acid dioxide submitted for mass analysis were equili- gave a quantitative yield of the corresponding abrated first with standard water to eliminate the acetamido ester, m p . 62-63'. Anal. Calcd. for effect of variation in 0's content. C22H39N04: C, 69.3; H , 10.2; N, 3.6. Found: C, 69.8; H, 10.6; X, 4.0. TABLE I Selective reduction of the @-oxo group was effected with sodium borohydride under mild MOLEFRACTIONS FROM IONCURRENTRATIOS X lo6 Acetic acid Methyl carbon of conditions. The two diastereoisomeric carbinols COi evolved combustion acetic acid could be separated by crystallization. One of them 10420 10684 10661 (m.p. 64-65') was obtained in a pure state. Anal. e13 10416 10675 10657 Calcd. for C Z ~ H ~ ~ N C,O68.9; ~ : H, 10.7; S , 3.6. 10424 10681 Found: C,68.9; H, 10.8; hT,3.6. Saponification of the pure isomer with diluted 16313 8543 hydrochloric acid yielded ethyl 2-amino-3-hydroxyC14 16290 8581 4-octadecenoate hydrochloride, m.p. 110-1 12". 16299 8645 Anal. Calcd. for C20H40N03C1:C, 63.4; H, 10.7; In Table I are given the C13 and CI4 mole frac- N , 3.7; C1, 9.4. Found: C, 63.5; H , 10.6; N, tions obtained from three runs in which yields were 4.0; C1, 9.6. Direct treatment of the hydrochloride with an nearly quantitative; the results for carboxyl excess of lithium aluminum hydride gave the carbon of acetic acid were erratic, however. From (I \ . the data shown we calculate for the C14 effect desired lJ3-dihydroxy-2-aniino-4-octadecenc 3.45 f 0.46%, for the CI3 effect 2.85 f 0.09%; The erythro configuration of this base was proven the ratio of these is 1.91 f 0.17. Since the two by its hydrogenation to the corresponding dihydroheavy isotopes are present in approximately the sphingosine. The triacetyl derivative of (I) melted sharply same concentration in the carboxyl carbon atoms, the results obtained for C13 depend upon those for a t 90-91 O after one crystallization from methanol. CI4; accordingly, agreement of the present CI3 Anal. Calcd. for C24H43r\j06:C, 67.7; H, 10.1; ' , 8.3. Found: C, 6'7.9; H, 10.3; S , 3.6. figure with that obtained previously by the same h The presence of an allylic system involving the techniques applied to unlabeled starting material* increases our confidence in the CI4 result. We carbon atoms 3 , 4 and 5 was demonstrated by believe that these results indicate strongly that the hydrogenolysis of the secondary acetoxy group (1). the triacetyl derivative was shaken with iiialonic acid case can no longer be considered deviant from the predictions of the absolute rate platinum oxide and hydrogen, approxiitiately 50-607& of the theoretical amount o f acetic acid was theory. produced. SOYES LABORATORY OF CHEMISTRY PETER E. YANKWICH The infrared spectrum of the triacetyl coi~ipound UXIVERSITY O F ILLINOIS ALBERTL. PROMISWW URBANA, ILLINOIS ROBERT F. NYSTROhf showed the characteristic trans-peak2 near 1 0 . 3 ~ . (1) 13. E . Carter, E'. G . Glick, W. P. Sotris and G. E. Phillips, RECEIVED SEPTEMBER10, 1954 (8) i
P E I'ankaich and A. L Promislow, 'Ims J O U R V ~ I . 76, 4648
1554)
( 9 ) C 3 l f a f v e l and K D
an&, h d , 7 3 . 2642 (I55U)
J . Bioi. Chem., 170, 285 (1547,. ( 2 ) K. Mislow.THrs J O U R N A L , 74, 5155 ( l Y s 5 2 1 i Y ) G. Fodor, Ndture, 171, 651 (1853). , A j R . Jepp and F, Klinpcmann, Bzi.., 20, ?Y4'?, 2'234, 33YY I I Y Y 7 1 .
