COMMUNICATIONS TO THE EDITOR
A%ug.20, 195s
4433
Compound I (3.00 g., hexahydrate) was allowed resolves into two peaks on the descending side to react for 30 minutes with 400 ml. of 0.295 N while the B protein shows only a pronounced hydrochloric acid a t 100 & 0.5", and the reaction skewness of the boundary. From this they conwas then quenched in ice-water containing sodium cluded that @lactoglobulin A is primarily responbicarbonate. The organic material was isolated sible for the aggregation between pH 3.7 to 5.2,5,8 by continuous extraction with ether, and the un- while P-lactoglobulin B aggregates to a considerably reacted pinacol (1.2 g.) was isolated from the ether lesser extent. Klostergaard and Pasternakg reported electrosolution as the hexahydrate, after addition of a minimum quantity of water. Periodate' cleavage phoretic patterns identical with those of Ogston of the recovered pinacol afforded acetone-carbonyl- and Tombs,' and also some ultracentrifugal data, C14 whose 2,4-dinitrophenylhydrazone (m.p. 125- with the opposite conclusion that only p-lacto126') had a radioactivity assay of 2.240 0.004 globulin B associates. In the course of studies on the molecular behavior mc./mole. The pinacolone remaining in the foregoing ether solution was converted to the 2,4- of /3-lactoglobulin between @H 1.5 and 5.5E9L0 dinitrophenylhydrazone, m.p. 126-127' (depres- we have examined for evidence of aggregation sion when mixed with acetone-2,4-dinitrophenyl- eight samples of @-lactoglobulinA and twelve of hydrazone!), radioactivity assay, 4.388 f 0.012 P-lactoglobulin B prepared in our laboratory from the milk of individual cows as well as samples of the mc. mole. Compound I1 (hexahydrate, m.p. 45-46.5 ") was two proteins kindly given to us by Dr. R. Aschaffenburg. The results obtained showed that all prepared from redistilled biacetyl (b.p. 83-84') and methyl-C'4-magnesium iodide. A repetition samples of @-lactoglobulin A aggregate strongly of the experiment described above (by which I a t PH 4.65 and 2' while none of the samples of was subjected to rearrangement) led to the reisola- @-lactoglobulin B do. A correlation of the ultracentrifugal and electrotion of 2.1 g. of 11-hexahydrate. Cleavage of reisolated I1 by periodate produced acetone-methyl- phoretic patterns is given in Fig. 1. The ultraC14 whose 2,4-dinitrophenylhydrazone(m.p. 125- centrifugal patterns of the @-lactoglobulinprepared 126") had a radioactivity assay of 2.062 It 0.001 mc./mole. The pinacolone 2,4-dinitrophenylhyu s l o b = 1.9, 2.5 ~ 4 . 8 drazone isolated from the same reaction mixture had a m.p. of 126-127", and a radioactivity assay of 4.122 f 0.017 mc.i'mole. From the kinetic data of Duncan and Lynn4 the conditions for both of the N O R M A L foregoing experiments correspond to values of a / ( n - x ) in the first-order rate equation of 1.36, or a value o f f (fraction reacted) of 0.264. From these data, therefore, it can be calculated6 that the isotope effects ( k * / k ) are, within experimental error of f170,for the rearrangement of I, equal to or less than 0.97, for the rearrangement of 11,unity.6
*
(4) J. F. Duncan and K. R . Lynn. J . Chem. SOC.,3.513 (1956), Table 2. ( 5 ) W. Stevens and R. Attree, Can. J. Reseaych, B27, 807 (1949); J. Ying-Peh Tong and P. E. Yankwich, J . Pkys. Chem., 6 1 , 540 (1957). (6) This paper is based upon work performed at Oak Ridge Sational Laboratory which is operated by Union Carbide Corporation for the Atomic Energy Commission.
I
+
SED.; 2 O C . E L E C T R O P H . ; I 'C. OAKRIDGENATIONAL LABORATORY VERNONF. RAAEN RIDGE,TENNESSEE CLAIRJ. COLLIXS Fig. 1.-Tracings of ultracentrifugal and electrophoretic RECEIVED JULY 7, 1958 patterns (descending) of various @-lactoglobulins in $H 4.65 acetate buffer, r/2 = 0.1. Both sedimentation and electrophoretic migration proceed from left to right: sediTHE ASSOCIATION BEHAVIOR OF 8-LACTOGLOBU- mentation, 59,780 r.p.m.; "?\Tormal" and p-A, 1.4% proLINS A AND B tein, 160 min.; p-B, 7 % protein. 352 min.; electrophoresis, Sir: 1.6% protein, 8,000 sec. at 9.7 volts/cm 0.4K
The discovery of Aschaffenburg and Drewry that P-lactoglobulin consists of two genetically different proteins' has led to a reexamination of its Ogston and electrophoretic heterogeneity.2' Tombs7 found that a t pH 4.63 @-lactoglobulinA' 3,4,536
(1) R . Aschaffenburg and J Drewry, Noturc, 176, 218 (1955): 180, 376 (1957). (2) C. H.Li, THISJOURNAL, 68, 2746 (1946). (3) L. G . Longsworth and C. F. Jacobsen, J . Phys. Colloid Chem., 63, I26 (1949). (4) B . D. Polis, H. W. Schmukler, J. H Custer and T . L. McMeekin, THISJOIJRNAL,7% 4965 (1960) (5) A. G . Ogston and J. M. A. Tilley, Biochem. J . , 69, 644 (1955). (6) S. N. Timasheff, unpublished experiments. (7) A. G.Ogston and M. P. Tombs,Biochem. J . , 66, 399 (1957).
from pooled milk (designated as "normal") and of P-lactoglobulin A exhibits two peaks with ~ ~ values a t 2% protein of 2.8 and 3.3 S , corresponding to monomer and aggregate, respectively, while p lactoglobulin B gives a single peak with ~ 2 0 ?f , ~ 2.7 S for 2% protein. Increasing the protein concentration up to 7y0 resulted in no evidence of aggregation. The electrophoretic patterns are (8) R. Townend and S. N. Timasheff, .47ck. Biochem. B i o p h y s . , 63,482 (1956). (9) H.Klostergaard and R . A. Pasternak, THISJOURNAL, 79, 5671 (1957). (10) R . Townend and S. N. Timasheff, i b i d . , 79, 3613 (1957).
0
.
~
4432
COMMUNICATIONS TO THE EDITUR
essentially identical with those reported prev i o ~ s l y . ~The ~ ~ two peaks observed with plactoglobulin iz can be identified a s monomer (slow) and aggrcgatc (rapid). In the "normal" protein an intcrniediate peak due to @-lactoglobulin J3 is also present. The ultracentrifugal and c.lectrophoretic data on @-lactoglobulin A as a lutiction of concentration were analyzed in terms d the Gilbert theory, yielding equilibrium constants for the aggregation in good agreement with those obtained from light scattering.12,13 Prom our ultracentrifugal data wc conclude, therefore, that the association of @-lactoglobulinin the PH range of 3.7 to 5.2 is due primarily to plactoglobulin A, while pure p-lactoglobulin B does not aggregate, This is in direct contradiction of the conclusion of Klostergaard and Pasternakg :ind in agreement with that of Ogston and Tombs7 One should remark, however, that the latter rexchcd their conclusion from electrophoretic data done, which could be open to question in the :lbsence of supporting measurements.
Vol. 80
I t has been found that diacetyldelcosine, when oxidized with potassium permanganate in acetone gave a lactam, diacctyloxodelcosiric., t 1 i . p 1 03-~fL5°, [c~]*~D +%4.6 :k 1 .-I" ( f , 0.71 i i i chloroforni!. infrared band a t l(i44 cin---1(1;ici:~iiicar1;ciiiyl). A n d Calcd. for C28H41010-2:: C, ilO.OcI; I-I, 7,4!). Found: C, 60.95; H, 7.42. T h e formation oi ; I lactam shows the presence of a methylene group next to the nitrogen. Hydrolysis of diacetylosodelcosine with aq U ~ O U S rnc thanolic potassi ~1111 hydroxide produced oxodelcosine, n1.p. 243-246 ', [ o I z 3 D $44.3 0.6' (c, 7.81. in chloroform), infrared absorption, 1649 and 1622 C1iI.--' (split lactam band). -4naZ. Calcd. for C2JlsiOe?:: C, 61.65; H, 7.98. Found: C , 61.75; 13, i.RS. Oxodelcosine was oxidized by sodium bichromate in acetic acid to a diketonic product, didehydrooxodelcosine, m.p. 211-212°, [ C Y ] ~+l35.1 ~D f O.!Jo (c, 1.11in chloroform), infrared bands a t 3446 cm. (OH), 1757 cm.-' (cyclopentanone], 1 i 2 O cni.--' (lactam c:d!rm)-I) ; (cyclohexanonc), 1653 cm. ultraviolet, A,, 2% mp, log e 2.07. .4nal. Calcd. for Ct4H330sN: C, (32.19; 13, 7.18. Found: C , ( 1 1 ) C;. A. Gilbert, Disc. Faraday Sou., No. 20, 68 (1955). (12) K . Townend and S. N. Timasheff, t o be published. 62.35; H, 7.12. (13) The concentration dependence of Klostergaard and PasterThe oxidation product is a diketone as espected, nak'sa electrophoretic data on &lactoglobulin A is also strong evidence but one of the keto groups is in a five-membered f o r associatinn.1 ring while the other is in a six-membered ring. E.4STERX REGIONAL RESEARCH LABORATORY -kcording to structure I (R = CHR,R' = K" = EASTERS UTILIZATIOS RESEARCH AND DEVELOPMENT DIVISIOX SERGE Ii, TIMASHEFFH), the oxidation product of oxodelcosine should U. s. DEPARTMENT O F AGRICULTURE ROBERTTOWNEND have contained two cyclopentanone rings. Hence, PIIIL.4DEI.PIIIA 18,PENNSYLVANIA the substituents cannot be as shown in that strucRECEIVED J U N E 13, 1958 ture and, to account for the experimental result, i t is necessary to alter the suggested structure of delcosine to I (R = R " = H, R' = CI13). If this T H E INTERRELATION OF DELCOSINE AND structure be the correct one, then delcosine must be DELSOLINE very closely related to delsoline which has been Sir: assigned structure I (R = H; R' = R" = C H S ) . ~ The previous investigation of delcosine (isolated Indeed, delsoline should be an 0-niethyldelcosine. from Delphinium consolida L.) had led Anet, In order to ascertain this point, delcosine v a s Clayton and Marion' to assume that the alkaloid methylated by means of sodium hydride and possessed the same carbon-nitrogen nucleus as methyl iodide in dioxane.6 The product obtained lycoctonine2 and to suggest for it the tentative after chromatography on alumina melted a t 215structure I (R = CH3, R' = R" = H). In this 216" undepressed on admixture with delsoline, forniula although R" = H had been established and its optical activity, [ c Y ] ~ ~+53.4 D f 0.5" (c, 2.04 in chloroform) was the same as that of delsoline [cx]*~D +53.6 -C 0.5" (c, 2.11 in chloroform). -4naZ. Calcd. for C&4lOiN: C, 64.22; H , 8.84. Found: C, 64.43; H, 8.79. I t s infrared spectrum was superimposable on that of delsoline, and its Debye-Scherrer powder diagram was identical with that of delsoline. The methylation was therefore selective and affected only the secondary hydroxyl OH located on the five-membered ring. The conversion of delcosine to delsoline thus supports the new OCH3 structure I (R = R" = H, R' = CH3) suggested for delcosine. I the evidence for R' = H was meager and rested only on the interpretation of the formation of a carbinolamine ether.' The presence of a &-tertiary vicinal glycol has been demonstrated chemically.a An attempt now has been made to find further evidence for R' = H. (1) R. Anet, D. a'. Clayton and L. Marion, Can. J. Ckcm., 36, 397 (1957). (2) M. Przybylska and L. Marion, ibid., Sl, 185 (1956). (3) R. Anet and L. Marion. ibid., 36, 766 (1958).
*
PUBLISHED AS S.R.C. No. 4874 THEDIVISIONO F P U R E CHEMISTRY NATIONAL RESEARCH COUNCIL O F CAN-4D.4 OTTAWA, CANADA RECEIVED JULY 18, 1958
v. SKARIC' LBO MARION
( 4 ) W. I. Taylor, W. E. Walles and L. Marion, ibid., S2, 780 (1954).
( 5 ) F. Sparatore, R. Greenhalgh and L. Marion, Tetrahedron, in press. (6) M. Carmack, J. P. Ferris, J. Harvey Jr., P. L. Magat, E. W. Martin and D. W. Mayo, THIS JOURNAL, 80,497 (1958), have used this methylation procedure t o establish the relationship of deltaline, delpheline and lycoctonine. (7)National Research Council of Canada Postdoctorate Fellow.
