COMMUNICATIONS TO THE EDITOR
July 5, 1958
5 . 3 0 ~(bridge) as well as the bands commonly associated with the triphenylmethyl phosphonium ion. Acidification of I11 with dilute hydrochloric acid produced immediate decolorization and decaborane was recovered in 5oY0 yield. Air oxidation and hydrolysis of I11 was not apparent after three months a t ambient temperature. I t appears that this material is the triphenylmethylphosphonium salt of decaborane and is the first such material to be isolated as a pure stable (C6Hj)aP'CHz-
+ BioHir
---f
(C6Ha)3PtCHsBloH13-
substance. The reactions of other phosphine methylenes with decaborane are under investigation.
3481
when chromatographed with butanol-acetic acidwater,8 phenol-water,8 or pyridine-ethyl acetatewater.g I n addition, an ultraviolet absorbing compound was formed which exhibited the same chromatographic and spectral properties as guanosine diphosphate. Longer hydrolysis led to the formation of a second ultraviolet absorbing conipouiid which exhibited the same properties as guanosine monophosphate. I n view of the well-established role of the uridine sugar nucleotides as glycosyl donors in the biosynthesis of many complex saccharides, i t is an interesting variation to find fucose occurring in a guanosine nucleotide. The only other guanosine sugar nucleotide known a t present is guanosine diphosphate mannose.'O- l1
ROHM& HAASCOMPANY REDSTONE ARSENALRESEARCH DIVISION ALABAMA M. FREDERICK HAWTHORNE (8) S. M. Partridge, Biochem. J . , 42, 238 (1948). HUNTSVILLE, (9) M. A. Jermyn and F. A. Isherwood, ibid., 44, 402 (1949). RECEIVED MAY23, 1958 (IO) E. Cabih and L. P.Leloir, J . B i d . Chern., 206, 779 (1954).
Acta, 17, 283 (1965). (12) U. S. Public Health Service Postdoctoral Fellow. (11) J. L. Strominger, Biochim. e l Biophrs.
ISOLATION OF GUANOSINE DIPHOSPHATE FUCOSE FROM AEROBACTER AEROGENES
NATIONAL INSTITUTE OF ARTHRITIS& METABOLIC DISEASES NATIONAL INSTITUTES OF HEALTH
Sir: UNITEDSTATES PUBLICHEALTHSERVICE V. GINSBURG 14, MARYLAND H . N. KIRK MAN^^ We wish to report the isolation of a new sugar BETHESDA RECEIVED APRIL 5, 1958 nucleotide, guanosine diphosphate fucose, from a strain of Aerobacter aerogenesl that produces a polysaccharide containing L-fucose.2 ORGANOBORON COMPOUNDS. X. MIXED The nucleotides from 150 g. wet weight of bacTRIALKYLBORANES DISTILLABLE WITHOUT DISPROPORTIONATIONIJ teria were extracted with boiling 70% ethanol, precipitated with mercuric ion, and chromato- Sir: graphed on a Dowex 1-C1- c ~ l u m n . ~ Elution with It was suggested recently' that mixed trialkyl0.01 N HC1 and increasing concentrations of NaCl boranes characterized by the presence of a t-butyl yielded an ultraviolet absorbing peak which was group may manifest unusual stability to dispropordetermined to be 75% uridine diphosphate glucose tionation. One such substance, diisobutyl-t-butyland uridine diphosphate galactose by enzymatic borane, was described previously and its stability a n a l y ~ i s . ~This peak contained a total of 10 pM. to disproportionation was attributed to steric interof nucleotides calculated as uridine from spectral ference with the disproportionation mechanism. data. A minor component of this fraction amountWe wish to describe now the first distillable ing to 10% of its optical density a t 260 mp could be trialkylborane containing three dissimilar alkyl isolated b y paper electrophoresis or paper chro- groups, namely, t-butyl-isobutyl-n-amylborane. matography. The ultraviolet absorption spectrum This substance, b.p. 43.544.0" a t 0.5 mm., n2'D of this component was typical of a guanosine de- 1.4296, d Z 50.7506, was fractionally distilled twice rivative. I t s mobility during electrophoresis in in vacuo without decomposition, rearrangement or sodium formate buffer, p H 3.5, was less than guano- disproportionation. Anal. Calcd. for CI3H29B: sine diphosphate, but greater than guanosine B, 5.52. Found: B, 5.56. MRD: c a l ~ d . 67.30; ,~ monophosphate while it had a higher Rt than obsd., 67.48. Oxidation with alkaline hydrogen guanosine monophosphate when chromatographed peroxide' produced equimolar quantities of t-butyl, with ethanol-neutral ammonium acetate s o l ~ t i o n . ~isobutyl and n-amyl alcohols in high yield. The isolated component was analyzed colort-Butyl-isobutyl-n-amylborane was prepared in imetricallyo and found to contain approximately two ways: (a) in 50% yield by the alkylation of 0.8 pM. of 6-deoxyhexose per pM. guanosine. The n-amyldifluoroborane with t-butylmagnesium chloabsorption peak a t 400 mp given by the guanosine ride in anhydrous ether; (b) in 3Oy0 yield by the derivative in this test was identical in shape with reaction of t-butyl-di-n-amylborane with isobutylthat given by authentic fucose and also disap- magnesium bromide. The physical constants and peared a t the same rate upon dilution with water.I the infrared spectra of the two samples were pracHydrolysis of the guanosine derivative in 0.01 tically identical. I n connection with method (a), N HCI at 100" for 10 minutes liberated a com- i t is noteworthy that one t-butyl group derived pound having an Rr identical with that of fucose from the Grignard reagent rearranges to isobutyl (1) Strain AsSi (ATCC 12057). during the alkylation reaction. Concerning (2) J. F. Wilkinson, W. F. Dudman and G. 0. Aspinail, Biochem. J., 59, 446 (1955). (3) E . Cabib, L. F. Leloir and C. E. Cardini, J . B i d . Chem., 803, 1055 (1953). (4) H. M. Kalckar, E. P. Anderson, and K.J. Isselbacher, Biockim. Biophys. Acfa,20, 262 (1956). (5) A. C. Paladini and L. P.Leloir, Biochem. J . , 51,426 (1952). (6) 2. Dische and L. B. Shettles, J . B i d . Chem., 176, 595 (1918). (7) 2. Dische and L. B. Shettles, ibid., 192, 579 (1951).
(1) Previous paper, G. F. Hennion, P. A. McCusker and A. J. Rutkowski, THISJOURNAL, 80, 617 (1958). (2) Contribution from the Radiation Project operated by the University of Notre Dame and supported in part under Atomic Energy Commission Contract AT-(1 1)-38. (3) The B-C bond refraction was taken a s 1.93. (4) S . L. Clark and J. R . Jones, Abstracts, 133rd Meeting, American Chemical Society, San Francisco, April, 1958, p. 34-L.
3482
COMXUNICATIONS TO
THE
EDITOR
VOl. 80
method (b), it should be noted that t-butyl-di-n1.4333, amylborane (b.p. 47.5” at 0.14 mm., n 2 5 ~ dZ5 0.7585) is also a new mixed trialkylborane stable to disproportionation when distilled in vacuo and was prepared by the alkylation of t butyldichloroborane with n-amylmagnesium bromide in ether. Anal. Calcd. for C I ~ H ~ ~B, B: 5.15. Found: B, 5.26; M R D calcd.,371.93; obsd., 72.06. A number of other mixed trialkylboranes stable to distillation have been prepared in This Laboratory. Details regarding their properties and methods of preparation will be reported at a later date.
mones2 to demonstrate that their respective biological activities do not reside in relatively large proteins but in smaller peptides. We wish to report a simple micromethod of electrodialysis utilizing membranes implanted in starch gels which has allowed us to study this question with regard to pituitary thyrotropin. Based on the passage or non-passage of thyrotropic activity through cellophane membranes whose permeabilities were calibrated with proteins of known molecular weight, we have tentatively assigned a molecular weight in the range of 2(i,OO030,000 to the active principle. The method makes use of the starch gel electroDEPARTMENT OF CHEMISTRY G. F. HEYNION phoresis of Smithies4and is carried out in a trough, UNIVERSITYOF XOTREDAME P. X MCCUSKER 2.3 by 30 cm., 1 cm. deep. The membrane is SOTRE DAME, ISDIANA J. V . MARRA placed in the gel in either of two ways. One way RECEIVED JUSE 2, 1958 is to insert the membrane across half the width of the gel, parallel to and about 1 cm. away from the A MICROMETHOD OF ELECTRODIALYSIS AND ITS filter paper4 containing the sample. The effect of the membrane can thus be seen in direct comparison APPLICATION T O THYROTROPIC HORMONE with the electrophoretic pattern of the sample. Sir: If the membrane, however, is impermeable to the Electrodialysis has been used successfully with protein, the protein tends to travel around it. corticotropin’ and the posterior pituitary hor- This can be prevented by placing the nicmbrane in a semicircle held in place by masking tape and + S a m p l e / JMernbrane + Sample1 l M e m b r o n e - then pouring the hot starch solution in the appara- ._ tus. The latter method was used when bioassay 5 1 UNSTRETCHED MEMBRANE experiments were carried out. 1 q -3 __ Visking 20,‘32 dialysis tubing, which has been LYSOZYME (14,OOOi k o Z y M E t OVALEUM;N used by Craig and co-workers in studying dialysis of proteins by diffusion,j was found to be most suitable for the studies with thyrotropin (30 U.S.P. units per mg.”). With this membrane it was found that neither the biological activity nor any stainI __ able material would pass through a t pH 5.0 or a t pH 9.5. If stretching by hydrostatic pressure is f _ _ _ ~ - carried out,5 the membrane allows both the acCHYMOTRYPSINOGEN CHYMOTRYPSINOGEN f24.000) tivity and the staining components to pass a t pH 5.0. These results afford strong evidence that the I ‘b. I j\: thyrotropic activity does not reside in a small peptl - 3 __I __. PROLACTIN tide bound solely by electrostatic forces to a larger protein and thus separable by electrodialysis. The results together with those on known proteins, IL-upon which the tentative molecular weight is THYROTROPIN PREP THYROTROPIN PREP based, are shown in Fig. 1. The resolving power of the starch gel electrophoresis and the selectivity of the membranes6 are clearly illustrated in the case of the prolactin preparation, which separates into three components. Xone of these will pass the unstretched membrane but one component passes through the stretched membrane a t both PH 9.5 and 5.0. With respect to thyrotropin it should be rememFig. 1.-Diagram of starch gel electrodialysis with a series of proteins of increasing molecular weight (mol. wts. from bered that other methods may still demonstrate ref. 5) in acetate buffer, pH 5.0, r/2 = 0.012: lysozyme that the activity resides in a small molecule. .-I ovalbumin was run a t PH 3 . 5 ; duration of runs 2-3 hr.; full report of this work together with other at260-300 volts; 13 ma. Proteins were commercial prepara- tempts to dissociate the thyrotropic activity from
r-
i
,
’
2%.
-
+
tions except prolactin.^ Those proteins not passing the membrane were detected as heavily stained areas just behind the membrane. With pH 9..5 glycine buffer, analogous results were obtained with trypsin, prolactin, in-iimuciiict and ovalbumin.
4. (1)
G. P. Hess, J . I. Harris, I?. H. Carpenter and C . H. Li, THIS 73, 5918 (1951).
JOURNAL,
( 2 ) C . H . Haselbach and .4 K . P i g u e t , ,Vel>~,Ciiint. A c l n , 36, 2131 (1952)): R . Acher, J . Chauvet and G O l i v r y , l?iochinz. Biop1iy.s Arlil,
22,401 (1966). (3) J , G . Pierce, L K , U’ynston and h l . 12. Carsten, Riociiiiii. Iliophys. A t l n , 28, 134 (1958). (4) 0. Smithies, B i o c h ~ mJ. , , 61, 629 (193J). ( 3 ) L. C . Craig, T. P. King and A . StrachPr, T H I s J O U R N A I . , 7 9 , 3729 (1957).
