3826
COMMUNICATIONS TO THE EDITOR
We have made what seems to us a more reasonable approximation: that ec is the same in all solvents. By calculating the activity coefficients of the various species,2 we are able to account for the Benesi-Hildebrand measurements with ec = 14,000. With this extinction coefficient we calculate from their data concentration equilibrium constants K’ (at about 20’) of 2.3 in pure benzene, 1.9 at infinite dilution in carbon tetrachloride, and 1.4a t infinite dilution in n-heptane. We have measured the 297 mp absorption of the complex in carbon tetrachloride solution a t various temperatures ranging from 0’ to 44.5’, and for various concentrations of benzene; and using e, = 14,000, have calculated concentration equilibrium constants. By plotting log K’against 1/T, the heats of formation were obtained from the slope and are given in Table I.
Vol. 72
We wish to acknowledge the helpful suggestions of Professors J. B. Ramsey and W. G. McMillan, Jr., and Messrs. J. T. Denison and S. C. Furman. DEPARTMENT OF CHEMISTRY UNIVERSITY OF CALIFORNIA THOMAS M. CROMWELL Los ANGELES,CALIFORNIA ROBERTL. SCOTT RECEIVED JUNE 30, 1950
Sir:
THE NUCLEIC ACID OF AVIDIN
Avidin, the biotin-binding protein of egg white, has been c h a r a c t e r i ~ e d as ~ * a~ ~water-insoluble ~ protein, isoelectric a t approximately pH 10, containing 10-13% N, and also phosphorus, carbohydrate and ash. A contaminant, iso-electric near @H 4.3, was regularly detected in the Tiselius apparatus and could be separated successfully only by this technique.2 By an entirely different method of isolation TABLEI . Z (bentonite adsorption, elution with M K2HP04, mole [Ill fraction K’ mole/liter AH,, dialysis and refractionation of the water-insoluble benzene at 25.0° calories at 25.0° protein with ammonium sulfate) we have regu1.00 17.35 X -1452 * 80 2.17 larly obtained water-insoluble avidin (avidin NA) 0.620 5.04 X -1416 * 14 1.87 of an activity similar to the highest previously 0.0434 141 X -1349 34 1.84 r e p ~ r t e d . ~ J ,The ~ electrophoretic behavior of 0.0217 264 X -1317 * 50 1.91 such preparations corresponded to that of Woolley The standard deviations for each concentration and Longsworth’s preparations. The ultraviolet absorption spectrum of avidin (using a t least four temperatures) are given in Table I. We estimate the probable error, allow- NA showed a steep maximum near 260 mp. The ing for uncertainties in the measurement of tem- presence of desoxypentose nucleic acid was thereperature and optical density, to be about 50 calo- upon demonstrated by a variety of methods. ries. The regular-solution calculation referred to Most of the 1%of phosphorus occurring in avidin above predicts a difference of approximately 90 NA could be split off by hot, but not cold, trior precipitated calories between the heat of formation in benzene chloroacetic4 or perchloric and in carbon tetrachloride, which is in the right with acid after treatment with alkali.e The direction and in reasonable agreement with the Dische reaction’ accounted for over 50y0 of the measured values. It should be pointed out that phosphorus in the various nucleic acid extracts. with increasing dilution of the benzene, the calcu- Paper chromatography combined with U.V. lation of AH becomes less sensitive to the choice spectral analysis clearly indicated the presence of of e,, and below xa = 0.05 is essentially independ- thymine, adenine, cytosine and probably guanine, as well as a small amount of uracil, and pentose. ent of it. Electrophoretic fractionations of avidin NA All the solutions of benzene and iodine in car(free, or as the biotin complex), performed in 0.2 bon tetrachloride were prepared and thermostatted at 25.0‘. By using such a procedure i t M KzHPOa, has yielded the main protein moiety was easy to make the appropriate corrections for showing the typical ultraviolet spectrum of a prothe variation of the volume concentration with tein containing tryptophane and tyrosine, with a temperature. Spectrophotometer measurements maximum a t 281 mp. I n contrast, a material were made in a Beckman Quartz Spectrophotome- containing almost all the phosphorus, and absorbter, equipped with the attachment for thermo- ing much more strongly in the ultraviolet, with a statting the solutions described by McCullough maximum only a t 258 mp was found to make up the minor peak of the “contaminant”* (Table I). and Barsh.8 We have investigated the shape of the absorp- Both components were non-dialyzable and, in tion curve a t 1.0 and 23.8’ and found no measur- contrast to the unfractionated material, largely water-soluble in both experiments. able temperature broadening. The biotin-binding activity of avidin is inherent We are continuing to study these interactions in the protein and independent of the nucleic acid between iodine and aromatic hydrocarbons and (1) Eakin, Sndl and Williams, J. Bid. Chem., 140, 535 (1941). will discuss these and further experiments and cal(2) Woolley and Longsworth, i b X , 143, 285 (1942). culations in a inore detailed report a t a later date. (3) Penningon, Snell and Eakin, THISJOURNAL, S4, 469 (1942). f
(2) J. H.Hlldebrand and R. L. Scott, ”Solubility of Non-electrolytea,” third edition, 1960. (a) f . b. M&u!leudh ~ t l dM.r ~BUMII, , Tnrr JOUnXAI,, TI, 8028 08SdP>,
(4) Schneider, J. BioL Ckcm., M i , 293 (1946); 104, 747 (1946). (6) Ogur and Rorea, Arch. Biochsm., S I , 168 (1949). (6) Schmidt nnd Tnnahaeuaer, J . B i d . Chcm., 161, 88 (1946). (7) HonJnnd, J q Bxp. M i d . , 73, 139 (19400).
.lug., 1950
COMMUNICATIONS TO
THE
EDITOR
3827
Sarmentogenin (0.15%; m. p. 270-274'; mixed with authentic sample, no m. p. depression) and a new aglycone, sarverogenin (0.10'%), recently deActiv- Soly. CornMoN/P -Ultraviolet absorptionity in scribed by Reichstein and his associates,z were obponent bilityo ratio E Epb ENb water tained from seeds of S. courmontii (sample number Main - 2.0 100 Max.281 130 91 Sol. 50R344). Strophanthidol (0.15% was isolated Max. 290 from seeds of S. mortehanii (50R1241) and from Min. 250 Minor -15.4 2.gd Max.258 8,100 1360 4 Sol. seeds of S. arnoldianus (9R7837) in very small Min. 232 yields. No sarmentogenin was isolated from the Unfrac14 Max.260 10,000 310 100 Insol. seeds of either S. thollonii (9R9113 and 9R9114) or tionated material Min. 243 S. preussii (50R581 and 50R582). The samples, Mobilities X lo6 cm.* volt-' a t 0" and PH 8.9 in 0.2 ilf however, were small and further work on larger K 2 H P O ~ .Similar mobilities were observed for avidin PiA and its biotin complex a t approximately 0.5 and 1% samples is in progress. protein concentration, respectively. The mobilities vary Isolation studies with seeds of S. congoensis with the solvent; in veronal-chloride buffers a t 0.1 ionic (50R580) yielded a major Legal-positive steroid strength, mobilities similar to those reported by Woolley (1.02%) and two minor Legal-positive steroids and Longsworth2 have been observed. * E , calculated according to Vischer and Chargaff (J.Bid. Chem., 176,703, (0.03% and 0.01%). The major steroid has been 715 (1949)), E N correspondingly on the basis of the S characterized and it appears to be sarverogenin ; content of the preparations. In terms of yo of the un- m. p. 219-222'C. (micro-block) ; Anal. Found : fractionated material, 1 mg. of which inactivated 7.4 y bio(c Upon fractionation of the biotin complex of avidin c , 65.S4, 65.91; H, 7.48, 7.27; [crIz3D f49.3' tin. XA the minor component obtained had N/P ratio of = 1.0; methanol); A, 2160 A. (ethanol), E% 2.1, which indicates that it is largely nucleic acid. 433. The molecular weight by the ebullioscopic method in methanol was 437 * 14. An infrared moiety. This is evident not only from the electro- spectra comparison of our steroid with sarverophoretic fractionations, but also from the finding genin showed the two to be almost identical, exhithat a readily water-soluble biotin-binding fracbands a t 5.78, 5.88, 6.12 and 2.97 p as soltion can be isolated from egg white free from nu- biting cleic acid, yet almost as active as avidin NA. ids in Nujol mull. ,4 mixture of the two comThis albumin (avidin A) is probably identical pounds gave no melting point depression and each gave the same sequence of color changes when a with the protein component of avidin NA. The question whether avidin is naturally asso- few crystals were contacted with sulfuric acid. ciated with a nucleic acid, or whether the com- Attempts to prepare the acetyl derivative failed plex is formed only during the isolation, regardless to give a crystalline product. However, a crysbenzoate was prepared: m. p. 187-191' of the method employed, is under investigation. talline Anal. Found: C, 70.61; H, Nucleoproteins appear not to have been previously (micro-block); [ c x ] ~ ~+30.7' D 1" (c = 1.301; acetone). 6.69; recognized and isolated from egg white. Their These properties and * the sequence of colors propossible embryological significance, particularly in duced with this derivative in sulfuric acid are in association with avidin, invites speculation. agreement with sarverogenin dibenzoate.2 WESTERNREGIONAL RESEARCHLABORATORY In addition to seeds, other parts of plants H. L. FRAENKEL-CONRAT ALBANY,CALIFORNIA (branchlets with leaves, stem bark, root bark, CHEMISTRY BUREAUOF AGR.AND INDUSTRIAL \v. H. WARD stem wood, root wood, inner shells of fruits and AGRICULTURAL RESEARCH ADMINISTRATION UNITED STATES DEPARTMENT S . S. SNELL awns of seeds) were examined for steroid content. OF AGRICULTURE E. D. DUCAY From numerous samples of S. sarmentosus, S. barJ U N E 26, 1950 RECEIVED feri, gratus, hispidus and S. preussii minute quantities of Legal-positive reacting fractions were obtained in only a few cases. Apparently the glycoside resides almost exclusively in the seeds, a t least in the species which were investiSTROPHANTHUS AGLYCONES gated. A detailed report of our work will be preSir : sented a t a later date. In a search for sources of sarmentogenin and The samples investigated in this work were colother 11-oxygenated steroidal aglycones, we have lected and identified by B. A. Krukoff. The boexamined a number of different species of Strotanical information and the places of deposit of phanthus. Because of the current interest in the Strophanthus species, we should like to report our the herbarium specimens have been reported in (2) von Euw, Katz, Schmutz and Reichstein, "Festschrift Prof findings a t this time. We employed a modificaPaul Casparis" S. 178 ( Z h c h , 1949); Buzas, von Euw and Reichtion of the procedure of Katzl for the isolation of stein, Hclu. Chim Acta, 88, 465 (1950) Since submission of this the steroidal aglycones. With seeds of S. sarmen- manuscript, von Euw and Reichstein reported the isolation of Savertoms, S. hispidus, S. kombe and S. eminii our data oaide, sarmentocymarin and sarmentogenin From S. courmonlii; Help. agree in general with those published previously chim. ucta, 88, 1006 (1950) (3) We are indebted to Dr. A Katz, presently n t The Exppuimenby other investigators. tnl Biology and Medical Institute, National Institute of Health, TABLE I ELECTROPHORETIC FRACTIONATION OF AVIDINNA
s.
(1)
A. Kntr, Hdr. Chfm. Arlo, 81, 998-1004
(1948).
s.
Betheads, Muylmd, for carrying out these comp.ri8orn.
