2072
PAUL H. MAURER,MICHAELHEIDELBERGER AND DANH. MOORE
aspect of the problem which has been studied in proteins such as Ea, @-lactoglobulin and horse serum pseud~globulin.~~ Only 70% of the amino groups of Ea are affected by PhNC028and HCHO, whereas acetic anhydride blocks 85% of the amino groups.13 Brand, et al., noted a difference in the rates of deamination of proteins by HN02.29 Differences in reactivity of the -NH2 groups toward formalin have been observed with diphtheria toxin30and tobacco mosaic virus.31 The third factor influencing the removal of aiilino groups may be the PH. Kornblum and Iffland have shown that aliphatic primary -NH2 (27) R . Porter, Biochem. et Bioghys. Acta., 2, 105 (1948). (28) A. Kleczkowski, Brit.J . Expil. Path., Zl, 1 (1940). (29) E. Brand, L. J. Saidel, U‘. H. Goldwater, B. Kassel a n d F. J. R y a n , THIS JOURNAL. 67, 1511: (1945). (30) A. M. Pappenheimer, J r . , J . Biol. C h i n . , 136,201 (1937). (31) G. L. Miller a n d W. M . Stanley, ibid., 141, 905 (1041).
VOl. ’ 7 3
groupsa2do not react below PH 3.0, indicating a narrow pH zone in which deamination is achieved. This probably explains the smaller extent of deamination of acid DnEa a t pH 3.0 than a t pH 3.5 (Table I). Since more groupings are detectable in the denatured than in the native state,33 it was hoped to obtain a highly deaminated protein by reacting free HNO, with acid DnEa a t pH 3.0-3.5. However, deamination merely equalled that of Ea under the same conditions, but was greater than the deamination of Ea in the milder buffer mixture. Even 8A which had been deaminated 44% by the buffer mixture lost only an additional 23% of its -NH2 groups after exposure to free HN02. (32) N. Kornblum a n d D . C. Iffland, THISJOURNAL, 71, 2137 (1949). (33) M. 1,. Anson, Adv. in Prolein Ckem., 2, 361 (1946).
SEWYORK,X. Y.
RECEIVED JULY 11, 1950
[CONTRIBUTION FROX T H E DEPARTMENrS OF BIOCHEMISTRY AND h’fEDICINE, AND THE ELECTROPHORESIS COLLEGE OF PHYSICIANS AND SURGEONS, COLUMBIA UNIVERSITY]
LABORATORY,
The Deamination of Crystalline Egg AlbUfnin.lb2 11. Physical and Chemical Properties of the Soluble and Denatured Derivatives BY PAUL H. MAURER, MICHAEL HEIDELBERGER AND DANH. MOORE The partially deaminated (27 to 36%) and soluble derivative of egg albumin, fraction B, has physical and chemical properties like those of native Ea and may therefore be termed a n undenatured deaminated Ea. The insoluble derivative, fraction A, formed during deamination (33-56y0)of Ea has many properties characteristic of the denatured state, but its configuration appears less extended than that of acid denatured Ea (DnEa) as indicated by nitroprusside tests, sedimentation and diffusion constants, and the reduced tendency to aggregate in salt solutions. This less completely unfolded protein is relatively stable to further treatment with acid a t PH 1.5 whereas the B fraction is readily transformed into an acid DnEa derivative different from the one obtained by deaminating acid DnEa.
In the preceding paper3the preparation of various and other proteins7, and has been ascribed to deaminated and denatured derivaties of crystalline changes in the asymmetry of the “backbone” of egg albumin (Ea) was described. Since the solu- the peptide chain6p8but may also be due to a n bilityr3 optical activity, viscosity, diffusion con- unfolding of the molecule, as appears to be the stant, sedimentation constant, electrophoretic mo- cause of the parallel increase in ultraviolet absorpbility, ion binding and sensitivity to salt are a11 tiong and in the number of detectable -SH groups. lo useful for the characterization of the protein The [a]Dof the soluble, or B, fractions resembled molecule and detection of changes, these proper- that of native Ea. After acid treatment of B ties were studied. a t pH 1.5 [.ID was like that of acid denatured Ea 1. Optical Activity.-The specific rotation of (DnEa) and of the A, or insoluble, deaminated the protein solution was determined a t room fractions. No detectable change was observed temperature, 22-25’, a t about 1% concentration, after further acid treatment of A (Table I) nor after dialysis against phosphate buffer a t pH 7.5, was there significant change in the [ . I D of DnEa 0.05 ionic strength. The factor G.4S4 was used to upon deamination (DnEa Deam -S-S-,for the oxiconvert mg. N to nig. protein. An increase in dized form; -SH, for the reduced form (Table I ) ) . negative rotation o€ Ea upon denaturation (Table 2. Viscosity.-Eight to ten ml. of a 1% pro1) is i n accord with previous reports on this5i6*6a tein solution were dialyzed a t 0 to 5’ for 4 to 5 days in the presence of toluene against two changes ( I ) Submitted by Paul H. hIaurer in partial fulfillment of t h e requirements for t h e degree of Doctor of Philosophy i n t h e Faculty of daily of 0.02 M phosphate buffer at PH 7.5, ionic P u r e Science, Columbia University. strength 0.05, a concentration suflicient to suppress (2) Presented before t h e 40th Annual Meeting of the American Sothe electroviscous effect.” Viscosity determinaciety of Biological Chemists, Detroit, Michigan, April 18-22, 1949, and before t h e 41st Annual Meeting of t h e American Society of Biologitions were made with 5 ml. of dialyzed solution in an cal Chemists, Atlantic City, N. J . , April 17-21, 1950. Ostwald-Fenske type viscometer a t 25 * 0.5’. (3) P. H. Maurer a n d M . Heidelberger, THISJ O U R N A L 73, 2070 (1951). (4) M. Heidelberger a n d F. E. Kendall, J . E z g t l . M e d . , 62, 697 (1935). (Sj 1%.F. Holden and M Freedinan, A u t . J . Ex,hlZ. B i d . onii M c d . Sei,, 7, 13 (1930). ( 6 ) 13. A. Barker, .I. A d . i ‘ h ~ i i i 103, I , I .I GCU.Pkysioi., 27, 1 0 1 (l’3U1.
(7) A. W. Aten, Jr., C. J. Dippel, IC. J. Keunig a n d J. Van Drenen, J. Colloid Sci., 3, 65 (1948); K. Linderstr$m-Lang, Cold Spring Harbor Symposia on Quant. Bwl., 14, 117 (1949). ( 8 ) W.Pauli a n d R. Weiss, Biochem. 2..233, 381 (1931). (9) J. I,. Crammer a n d A. Seuberger, Biochem. J . , 37, 302 (1042). 110) ,\I. L. Anson, Adu. iic Protein Chefn., 2, 361 (1046). ( I I ) h. Pc>lson. Zi dinid Z.,88, 31 :1!139)
May, 1951
THEDEAMINATION OF CRYSTALLINE EGGALBUMIN
TABLE I PROPERTIES OF Ea, DEAMINATED Ea [alD
Preparation Ea 3A 3B 3BW 3B(B) 4A 4B
5A 5B 6A 6B 8A
8B 10A 5A Dn 6 A I)n 8A Dn 6 B Dn 8 B Dn DnEa 105
$2
-2 8 O -51 - 40 53 37 51 36 -51 36 - 52 36 52 -37 54 57 57 55 63 62 53
-60
DI
x
10’ 7.8b 3.0 8.0
3.9
AND
DENATURED Ea
Particle Sto f/jo weight 3.55b 1 . 1 45,000 4.2 1 . 9 135,OOoh 1 . 1 38,000 3.1
3.4 7.8 3.2 7.5 4.0 7.3
3.0 3.0 4.3 3.3 4.4 3.0 3.1 3.2
1.8 75,000 1 . 0 33,000 1 . 7 125,oooh 1 . 1 41,000 1 . 8 135,OOoh 1 . 1 39,000 1 . 7 76,000 1 . 1 43,000
2.7 2.5 3.5 2.6 3.1 1.5
3.9 4.0 6.2 7.8 8.9 9.2
2.1 2.2 1.5 1.7 1.4 2.3
8.8‘
140,000 175,000 175,Oooh Z90,OOoh 280,000 600,OOoh
%el. at 2 mg. N/ml. 1.044 1.136 1.056 1,134 1.051 1.136 1.053 1.181 1.058 1.134 1.064 1.134d 1.049 1.224 1.186 1.143 1.163 1.149 1.149 1.227”
2073
. 1.00
$ 0.95 .r(
3 0.90
.E 0.85 4 0.80 Y
D
0.75
1.2 1.6 2.0 2.4 Mg. N per ml. Fig. 1.-Comparison of fluidities of Ea, DnEa and deaminated fractions-key to chart: A 3B, A 4B, 0 3A, 4A, 0 5A, lOA, 0 Ea, DnEa 101, $ DnEa 102, 0DnEa 105; line D represents average curve for DnEa 101 and 102.
0
0.4
0.8
e
such as that shown by acid DnEa and other DnEa preparations on “ageing,)’14may be taken to indicate disaggregation. Although the A fractions resembled DnEa in many respects, their 1.315 relative fluidity had not changed after a year. The relative fluidities of the B fractions resembled - 50 1.428 Deam (SS) those of native Ea (Fig. l), whereas those of the - 55 2.0 10 1 . 8 490,OOoh 1.198’ ADn preparations differed little from the fluidity - 59 2 . 0 13 1 . 7 630,OOoh 1.398f of the original A fractions. On the other hand, when Fraction B was treated with acid its viscosity 1.5 - 59 13 2 . 0 850,000h 1.575 Deam (SS) increased to about that of Aand DnEa, indicating 8A F N A - 50 6.3 2.2 2 . 0 280,OOoh 1.145“ the denaturation coniirmed by other tests. As for - 55 2 . 4 5.0 2 . 1 205,O0Oh 1.16od the deaminated denatured Ea, DnEa Deam, -S-Sg a y %Ea) - 65 1.650 2 70 form, its viscosity was greater than that of the Other values of [ a ]reported ~ are: -37” (refs. 6, 8); -SH form, which, in turn, was greater than that -27” t o -32” (ref. 6a); -31” (ref. 12). "ram Sved- of the DnEa used. The viscosity of DnEa Deam, berg and Pedersen, ref. 27. The value closest to that of -SH form was not reduced appreciably by a second Ea is given. All other S ~ values O are averages of two or treatment with thioglycolic acid. The difference three calculations. Same value for the -SH and -S-Sforms. e Value after three months of ageing. f Twice in viscosity between the -S-S- and -SH forms of treated with thioglycolate. rprepared by Dr. C. F. C. DnEa Deam is not exhibited by the -S-S- and -SH MacPherson, ref. 14. "slue given t o nearest 5,000 forms of the A and B fractions, perhaps because particle weight. of differences in the nature of the -S-S- linkages. Two concentrations of protein were used, the As will be shown in connection with the nitroweaker being obtained by twofold dilution with prusside tests, the A fractions are not “unfolded” buffer. The N content of one solution was deter- to the same extent as is acid DnEa. Therefore, mined by micro-Kjeldahl analysis.I2 The time of fewer -SH groupings are available for intermolecuoutflow for buffer alone was about 260 sec. No lar -S-S- formation. However, upon oxidation, the corrections were made for small differences in less exposed -SH groups in Ea and fraction A density between the protein solutions and the could undergo intramolecular -S-S- bond formation, buffer as the over-all accuracy was limited to and this would not change the viscosity.18,19 3. Electrophoresis.-Recrystallized Ea is elec1 to 201, by the method of analysis. Concentration in mg. N per ml. was plotted against relative trophoretically homogeneous except between PH fluidity. This yields a linear relation up to a t 5.0 and 1O.OZo where two as yet inseparable components appear. The ratio of these components least 2 mg. N/m1.13114 The viscosity of a protein solution is believed varies with age, preparation, PH and medium.21 to reflect the particle shape and extent of hydration The isoelectric point of Ea is between 4.6 and 4.ELZ2 of the protein.16 A marked increase in viscosity, The electrophoretic properties of DnEa have also Le., a decrease in relative fluidity, would be in- been investigated. l4lZ2 dicative of the unfolding of the protein into a (1936); W. T. Astbury, S. Dickinson and K. Bailey, Biochem. J . , 19, more asymmetric form characteristic of the de- 235 (1935): H. B. Bull, Cold Spring Harbor S y m p . Quotil. Biol., 6, 140 natured protein. I6,l7 A decrease in viscosity, (1938). *
(I
(12) E. A. Kabat and M. M. Mayer, “Experimental Immunochemistry,” C. C. Thomas, Springfield, Ill., 1948. 62, 1405 (1940). (13) H. P. Treffers, THISJOURNAL, (14) C. F. C. MacPherson, M. Heidelberger and D. H. Moore, ibid., 61, 578 (1945). (15) Reviewed by M. A. Lauffer, Chem. Rcv., 81, 561 (1942). (16) H. Neurath, J. P. Greenstein, F. W. Putnam and J. 0. Erickson, ibid., 34, 157 (1944). (17) A. E. Mirsky and L. Pauling, Pmc. Nall. Acad. Sti.,11, 439
(18) M. L. Anson, J. Gcn. Physiol., 10, 355 (1941). (19) E. Fredericq and V. Devreau, Bull. SOC. chim. Beig., 88, 389 (1949). (20) L. G. Longswortb, R. K. Cannan and D . A. MacInnes, THIS 62, 2580 (1940); A. C. Chibnall, Proc. Roy. SOC.London, JOURNAL, B181. 136 (1942). (21) C. F. C. MacPherson, D. H. Moore and L. G. Longsworth. J . Bid. Chcm., 166, 881 (1944). (22) H. A. Abramson, L. S.Moyer and hf. H. Gurin, “Electrophoresis of Proteins,” Reinhold Pub]. Co..New York, N. Y . . 1942.
2074
I'atiL
H. MAURER,MICHAEL HEIDELBERGER AND DANH.
MOORE
VOl. 73
4. Diffusion.-One per cent. protein solutions were dialyzed as for viscosity. Measurements of diffusion were carried out with the usual modificaPrepara2% detion amination Buffer pH p X 10s tions in the Tiselius apparatus a t l * 0.05'. At 0.05 ionic strength The scanning method of LongsworthZ5was used. 4-4 42 Citrate-HCl 2.9 +6.71 The diffusion constants tabulated are averages 4A 42 Phosphate 6.2 -7.77 from the two arms of the U-tube and were cal4A 42 Phosphate 7. -g. culated by the inflection points-maximum ordinatez6 4B 31 Citrate-HC1 2.9 +6.28 method. The patterns obtained indicated afair de'1n 31 Phosphate 6.2 -7'38 gree of homogeneity. However, diffusion constants 413 31 Phosphate 7.5 -g.49 are insensitive as an indicator of polydispersity, which is more readily detected by measurements of At 0.10 ionic strength" 3 , o5 $6.25 sedimentation properties.27 I3a Glycine 3.62 +3.89 From Table I it will be seen that the D20 of Ea Acetate-NaC1 4,64 -o,20 and the B fractions are about the same, ;.e., 7.8 Acetate 3,65 -3,53 X The A fractions, which are denatured Acetate tj,12 -1,46 by definition (see Paper I) and by other criteria, Phosphate 6,80 -5,y2 have diffusion constants of about 3.2. The ADn Phosphate 7, -6, samples show a slight decrease in Dzo which apPhosphate proaches that of acid DnEa. However, B, on con* Data taken from L. G . Longsworth, ant^. N. Y.A c u d . version to DBn exhibits greatly altered D2o values in Sci.,41,267 (1941). the range of those of DnEa and ADn. A decrease Mobilities were determined in a modified com- in diffusion constant is ordinarily ascribed to an pact Tiselius electrophoresis apparatus of 2 ml. increase in the asymmetry of the molecule and/or capacity.23 The buffers used were of 0.05 ionic to a s s o ~ i a t i o nrather l~~~~ than to hydration.28 strength (Table 11), the low concentration being 5. Sedimentation.-Sedimentation runs were used for comparison with the salt sensitive acid made in an air-driven ultracentrifugez9operating DnEa investigated under the same condition^.'^ at 48,000 r.p.m. All runs were made in 0.02 &I The protein concentration was approximately 0.5%. phosphate buffer a t pH 7.5 with a lY0 protein solution. Sedimentation patterns of the prof 8 teins were recorded by the Longsworth scanning 4-6 * methodz6 as adapted to the ~ltracentrifuge.~~ 2 . sedimentation constants were calculated from +4 g the equation of Svedberg and Pedersen.2i It @ +x z .3 1: is evident from Table I that the soluble de0 45 -4. 0 E aminated B fractions have sedimentation con5m stants like that of native Ea, but slightly lower, -2 " h: + possibly owing to the low salt concentration -4 40 2 The patterns (Fig. 3 ) show a sharp 3. sedimenting boundary indicative of a monodis-6 3 perse The insoluble deaminated A .- 8 2 fractions which show other physical properties 35 5 similar to those of DnEa exhibit differences - 10 was not from DnEa in sedimentation. The SZO 1 2 3 4 5 6 7 8 9 to 121J as with the least aggregated acid DnEa preparations, but instead was 3 to 4, like that of Pa. Fig. 2.--Mobilities of Ea (Curve 1) and fractions 4A and 4B native Ea. There was also less spreading of the (curves 2), and precipitation of A and B mixture (curve 3)-key sedimenting boundary than with DnEa (Fig. 3 ) . 6 . Particle Weights and Shapes.-The parto chart: 0 Ea, 0 4A, V 4B. ticle weights of the preparations were calculated Conductivities of protein solution and buffer from the sedimentation and diffusion constants at were determined after 4 days' dialysis. At p H 20' according to the equationz7 2.9 where Ea is homogeneous both A and B were JI = XTSzo/Dm(l - Vp) also homogeneous, and were much less heteroFrom S z o and D20 it is possible to calculate the geneous a t the other two $H values than is crystalline Ea. Two definite peaks were never ob- frictional ratio f/fo by the formula tained. No determinations were made between f , , f o = ] 0 - 8 (!I$?) "" D 'SzoV pH 3.5 and 3.5 as fraction A is insoluble in this range. The isoelectric points of both the A and B J. T. EdsalI, "Proteins, Amino Acids and Peptides," Reinhold P u b l . M ~ B I L I T I E S OF
TAELE I1 DEAMISATED Ea FRACTIONS AND Ea
#
