3168
C'OMXZUNICATIONS TO TIIE
EDITOR
L'ol. 77
tion about 20 Ing. of a solid was isolated from 4000 clines, a molecular weight of 282 was calculated liters of urine that supported half-maximum growth from the ultraviolet absorption. The infrared o f C. f(tcsc-icdntn at a concentration o f 0.03 milli- spectrum measured on a KRr pellet showed strong ~iiic~ograiiis per 1111. T h e 1 i ; i n i c I.)iopterin is sug- bands at X50-~;3~4X, 3 3 10, 2975-29.30, 1695-1680, gested for this substance. The material precipi- 1538, 1490, 1418, 1370, 1295, 121.5, 1173, 1130, tated as pale yellow spheres from hot water and 1063 and 823 cm. -- I . decomposed without melting at 250-2SOO. Tt was Riopteriti titrated with periodate consumed 1.7 osidation equivalents per molecular weight optically active, [ c Y ] ~ ~- D3 0 (0.1 S HCl; C, 0.4). and l.,i I t was insoluble in the conirnori organic solvents, of Z37 at PH 3 and 8.5, respectively. Formaldeslightly soluble in water, soluble in dilute acid and hyde, formic acid and ammonia were not detected alkali. Analysis calculated for C S H ~ ~:S C ~, O ~in the oxidation mixtures. Upon adjusting the -I.i.(i; H, 4.64; ?;, 29.5; mol. wt., 237. Found on oxidation mixtures to pH 5 , yellow precipitates two separate samples, C, 41.8,4 i . G ; H, 5.00, 4.S1; came down, and the ultraviolet absorption spectra of these indicated that the material from the pH N,29.:3; S, negative. The ultraviolet absorption spectrum of biopterin 2 osidatjon was 2-amino-4-hydroxy-6-formylpteriw a s typical of 3-aniino-4-hydroxy-alkylpteridinesdine and the material from the pH 8.5 oxidation Thus with maxima a t 2.54 mp and 305 mfi in 0.1 -Y was 2-amin0-4-hydroxy-6-carboxypteridine. NaOH. Assuming biopterin has the same molec- biopterin appears to consume about two inore periodate oxidation equivalents in alkali h y the ular extinction as 2-atnino-4-hydroxy-alkylpteripteridine aldehyde being oxidized to the correTABLE I sponding acid. From these data atid studies on THE EFFECT O F PGA O N THE GROWTHRESPOXSEOF models it is concluded the structure of biopterin is Crilhidinf~isrirzilntnTO CRINE OR TO CERTAIN 6-SUBSTITTTED 2-ainino-4-hydrox?-~(1 ,?-dihydroxypropyl)-pteridPTERIDISES ine. PO.< Additions t o medium-- -The microbiological response to biopterin is 10 100 1000 now obtained in the absence of added PG-k3 Additions per tube (2.a ml.) Sone m; my m-r -_ final strength mediiim" ~. .. . Optical Density The involvement of PGA4in the metabolism of C. 0.02 0 03 0 11 0 88 .fasciculnzta was demonstrated indirectly, however, .\one 0.001 ml. urine o n2 o 0-1 by adding an inhibitory amount of 2,4-diamino-50.01 mi. urine 00,j 0.98 p-chloro-phenoxy-A-ethylpyrimidine (I) to the cul10.7 0 1 nil. uritie ture medium. Under these conditions growth was obtained by adding sufficient PGA to counteract I only if biopterin was present (Table 2). The data of Tables I and I1 illustrate that C. fusciculata required both PGA and a pteridine for growth, implying that these structurally-related substances OH have independent metabolic functions for this R = CH2011" 1 0 in; 0 09 0.25 0.65 0.87 organism. IOOin-; n.07 0.58 0.81 0.88 We are indebted to Mrs. L. H . Smith for some of Test conditions were those described by Xathan and the microbiological assays and to Dr. S. H. Hutner, Cowperthwaite2 except that the adenosine was replaced by ' . for his 10Oy adenine, guanine and uracil per tube and the assay Haskins Laboratories, New York, N. 1 was carried out in slanted test tubes rather than flasks. suggestions concerning the culturing of the test Microbiological findings with other pteridines tested with organism.
-
~
. I
10 m y PGA in the medium were as follows: When R = H , OH or COOH, no activity a t l0y/tube, when R = CHO or CHs, 0.3 to 0.1 as much activity as when R = CHlOH (see table). 2-Amino-4-hydroxy-5formyl-A-methyltetrahydropteridine, the methylpteridine moiety of leucovorin, was inactive a t 10yltube.
TABLE I1 TOXICIT'i PYRIMIDIXE
Rir~pterin
. . . . 10 m:
I00 In., ...
..
100 m y 100m?
OF
2,4-DIAMIS@-~~-~-CHl,oROPHESOXY-6-ETHYL-
GROWTH OF Crifhidici REVERSAL BY PGA"
FOR
PGA
10 m7 100 m: 10 in-, 100mr
0 03 .4:3 .46
0.02 .02
.n:3
.. .. .49
.09 .49 .iA
.oo
0.01
0.03
.. ..
..
.. .. .04 .38
.. ..
.. ..
.03 a Tert conditions :is i n footnote t o T:il,le 1 ; 2.,5 ml. final volume .4j
-.
( 2 ) H. A. Nathan M p d . , 86, I I7 ( I 9 X )
xiid
.I
NUTRITION A K D PHYSIOLOGY SECTION E. L. PATTERSOX H. P . BROQUIST RESEARCH DIVISION ALBERTA M. ALBRECHT M. H . VON SALTZA LEDERLELABORATORIES E. L. R. STOKSTAD PEARL RIVER,S E T V Y O R K RECEIVED MARCH24, 1955
COMPANY ~ a s c i c u . l n t n : AMERICAS CYASAMID
2,4-Diamino-j-p-chloropheno~y-6-ethylpyrimidine (I) 0 l Y 3.7 10 7 Optical Density
... .... ....
(3) Within t h e past year the test organism apparently "lost" its nutritional requirement for preformed folic acid. Factors responsible for this change are unknown. (4) ADnExDGM.-After this Communication was submitted for publication, t h e Editors informed us t h a t Forrest and Mitchell have isolated and characterized what appears t o be t h e same pteridine from wild type Drosophila.
Ci,n.iiprtIiwiiitc, P r o r . .Snr. E.rpi!.
Bioi.
BOND ORDER-LENGTH
RELATIONSHIPS
Sir: The importance of establishing reliable bond order-length relationships for as many atom-pairs as possible is generally recognized. There are several ways of defining bond order (w). Examples are the original method suggested by Pauling, Brockway and Beach, and the more recent method of molecular orbitals. Bond order length curves have been usually constructed by plotting
COMMUNICATIONS TO THE EDITOR
Juiic 5 , 19.55
bond order against bond length, and attempts have been made to give these curves a theoretical or semi-theoretical basis. Empirical equations connecting these two internuclear properties have been suggested. I t is assumed now, as in the past, that single, double and triple bonds have their conventional orders, and second order refinements are properly neglected. For the hydrogen molecule, re = 0.74 B.,n =oI, and for the hydrogen molecule ion, re = 1.06 A., ?a = 0.5. Also, (l/re2) = 1.826 and 0.890, which figures suggest that bond order and l/'Ye2 have some connection. This. possibility may be tested rigorously for carbon-carbon bonds. The carboncarbo? bond length in ethane is 1.543 A., in ethylene 1.353 A., and in acetylene 1.207 A. Also, using the rotational Raman spectrum, Stoicheff has found that the C-C distance in benzene is 1.397 A., and by the molecular orbital method; Coulson2 has evaluated the C-C bond order in this molecule as 1.67. If now bond order is plotted against l / r e 2 an excellent straight line is obtained, having the equation l/re2 = 0.2868
+ 0.1334%
Similar relationships probably cxist for carbonoxygen and carbon-sulfur bonds. THEUNIVERSITY MANCHESTER, ENGLAND RECEIVED MAY9, 1955
Ye,
71
1 1.G7 2 3
STRUCTURE OF PRISTIMERIN AND CELASTROLI
Pristimerin is an antibiotic isolated from Pris-
timera indica and P.grahami (Celastraceae),2 which besides being active against the common grampositive organisms, is characterized by its activity against the Virzdans group of Streptococci. Comparison of the melting point, ultraviolet and infrared spectra of pristimerin with celastrol monomethyl ether3showed that the two compounds were identical.* We wish to present structure I for pristimerin and I' for celastrol as a working hypothesis.
A One isolated nuclear double bond, probably tri- or tetrasubstituted
(1)
1
A.
I/l.e2
1.543 1.397 1.353 1.207
0,1202 0.5121 0 5462 0.6S68
(I.&.
A.
1.543 1,401 1.344 1.207
For the carbon-nitrogen pair,j bond length information is on a lower plane of accuracy. A careful assessment of the data available pn the C-N single bond has given the value 1.475 A., and for the triple bond spectroscopic methods have led to 1.156 A. Again, theoretical bond orders calculated by the molecular orbital method, and experimental bond lengths are available for the molecules, melamine and pyridine. For melamine, the bond order of e5ch C-N bond is 1.658, and the bond length 1.346 A., and for pyridine the values are 1.334, and 1.37 A. Table I1 gives the data, the equation to the line being; l / r e * = 0.316
+ 0.144%
(2)
TABLE I1 11
YC,
I
A.
1.475 1.37 1.346 1,156
1 ,53
1.66 3
I/YC=
0.460 0,533 0.552 0.748
(l.r)c,
A.
1,475 1.37 1.35 1,156
For nitrogen-nitrogen bonds, the single bond distance is known from electron diffraction measurements on hydrazine to be 1.48 A. The double bond distance in azomethane is 1.24 A., %nd the triple bond distance in nitrogen is 1.094 A. -4 plot of bond order against 1,'re2 gives a i almost perfect straight line having the equation ; ~
... ..
~
~
~~
.~
1 ! r F 2 = 0 Bfi3 -1- 0,1F).?n
(3 )
11) B. P. Stoichei?, Corn . I . Phys., 32, 633 ( 1 9 5 4 ) ; G ITerzl,eri: .ind H. P. Stoicheff, N a f r i r e , 176, 7 ! ) (1933). ( 2 1 C . .\ Coulsou, I'roc. R o y . SIC., 1 6 9 8 , 413 (l!13!1). (,':) I <
