CUNNINGHAM, JR., TIETZE, GREEN AND NEURATH 67 ELECTROPHORETIC STUDIES ON ENZYMATICALLY MODIFIED OVALBUMIN AND CASEIN BY GERTRUDE E. PERLMANN The Rockefeller Institute for Medical Research, New York, U.S.A. Received 13th May, 1952 Electrophoretic analysis which has been used for a characterization of enzymatically modified proteins, has proved to be a sensitive tool for following enzymatic reactions involving charged groups. It also presents a qualitative picture of the type of groups concerned. The modifications to be discussed include reactions in which carboxylic and phosphate groups are removed from the proteins. The present paper describes recent work in which electrophoresis has been used to follow the enzymatic modifications of the two proteins, ovalbumin and casein.68 OVALBUMIN A N D CASEIN Unlike most studies of this type, e.g.proteolysis with the aid of pepsin and papain,l-S only those enzymes have been used that attack specific linkages in the protein and thus leave the substrate relatively intact. Although the change in molecular weight accompanying the modification of ovalbumin, say, appears to be within the limits of error of the measurement,6 other properties of the protein, such as the electrophoretic mobility undergo appreciable alteration. This probably results from the fact that one, or more, of the relatively few groups that are split off in the presence of the enzyme contribute to the net charge of the intact protein. The reactions to be considered are (1) the ovalbumin-plakalbumin transforma- tion, (2) the dephosphorylation of ovalbumin and plakalbumin, and (3) the de- phosphorylation of casein.Since much of the work has been with ovalbumin as the substrate, its electro- phoretic behaviour will be illustrated with the aid of the patterns of fig. 1. Here a represents egg white, which is a mixture of several proteins.7 From this mixture Egg White a- -d Ovalbumin b. a- r---------;td FIG. 1 .-Electrophoretic patterns of egg white and ovalbumin in sodium phosphate buffer of pH 6.8 and 0.1 r/2. Electrophoresis was carried out at a 1 % protein con- centration for 14,400 sec at 6-4 V/cm. ovalbumin is crystallized as needles by ammonium sulphate precipitation,g but 6 shows that the crystalline material is electrophoretically complex, consisting of two components, A1 and A2.9 It will be noted that these two components are also present in the egg white in much the same proportions as in the crystalline protein.In some preparations a slower moving component A3 is present in the electrophoretic patterns. THE OVALBUMIN-PLAKALBUMIN REACTION.-AS shown by Linderstrpm-Lang and Ottesen 10911 ovalbumin can be transformed, by the action of a proteolytic enzyme from B. subtilis, into a new protein, plakalbumin, which crystallizes as plates. The reaction involves the liberation of alanine, valine, glycine, aspartic acid and glutamic acid.12 On electrophoretic analysis plakalbumin gives patterns similar in appearance to those of ovalbumin except that at an ionic strength of 0-1 in the pH range from 3 to 7 the mobilities of the two components PI and P2 are 0.6 x 10-5 cm2 sec-1 V-1 less negative than the corresponding constituents, A1 and A2, of the parent substance13 This is shown for theG E R T R U D E E .PEKLMANN 69 major components, A1 and PI, of each protein by the pH-mobility curves of fig. 2. Since the mobility change of 0-6 x 10-5 is the same as that observed when two equivalents of acid are added to a mole of ovalbumin at I72 = O.l,14 this change has been interpreted 13 as being in agreement with the chemical finding that an aspartic and a glutamic acid residue are lost in the A -+ P transformation.* DEPHOSPHORYLATION OF OVALBUMIN.-GrOUpS containing phosphorus are known to be present in both ovalbumin and casein. Since these groups may contribute to the net charge over a considerable range of pH, their selective removal might also modify the mobility of the protein without seriously affeciing other 6.0 4.0 2.0 - 2.0 - 4.0 - 6.0 Flakalbumin 0 Ovalbumin 4.0 60 8.0 FIG.2.-Mobilities of plakalbumin and ovalbumin as function of pH. properties. Moreover, if ovalbumin is assumed to have a molecular weight of 44,000, the phosphorus content of this protein does not correspond to an integral number of atoms of this element per mole. This fact, together with the electro- phoretic complexity of ovalbumin, led Linderstrdm-Lang and Ottesen to suggest that A1 contains two atoms of phosphorus per mole and A2 one.11 Thus an oval- bumin preparation such as that of fig. l b with 85 % A1 and 15 % A2, should con- tain 2 x 0.85 + 1 x 0.15 = 1-85 atoms of phosphorus per mole, a value in good * With the aid of a different enzyme preparation than that used in the work described here, Ottesen and Villee (Compt.rend. trav. lab. Carlsberg, 1951, 27, no. 18) observed a plakalbumin formation in which a di- and tetrapeptide are released ; one being alanyl- alanine, the other containing valine, alanine, aspartic acid and glycine in the ratio of 1 : 3 : 1 : 1. No extensive electrophoretic studies have been carried out yet on these pre- parations and therefore no comparison can be made. Further information about the ovalbumin-plakalbumin process may be found in an article by Ottesen and Wollenberger (Nature, in press) and a more detailed discussion is presented by Linderstrsm-Lang (Lane MedicuZ Lecture (Stanford University Press, 1952), in press).70 O V A L B U M I N AND C A S E I N agreement with the chemically dcternliiied one of 1-82.With all preparations in- vestigated over a period of several years such agreement was found.15. To test further the hypothesis of Linderstr$m-Lang and Ottesen, phosphatases were added to salt-free ovalbumin solutions of pH 5.4 and the reaction was followed Time Flectrophoretic Atoms phosphorus In hours composltiori per mole protein Compurea Obsewed 85% A] 14% A2 18, 1.8* t M = A3 47% A, 49% 143 135 4% A3 y 34 I FIG. 3.-Dephosphorylation of ovalbumin at pH 5.4 with prostate phosphatase as function of time (taken from J. Gen. Physiol., 1952, 35, 711). Each reaction mixture contained 4-6 % ovalbumin and 0.01 % enzyme. by electrophoretic and chemical analyses.Two representative series of experiments are illustrated with the aid of fig. 3 and 4. As can be seen from fig. 3, if prostate Time Electrophoretic Atoms phosphorus in hours composition per male protein Computed Observed 85% A1 14% A2 184 1.82 i2 2 O truce A3 Az 50% A2 40% A3 0.50 0.58 30% * A, FIG. 4.-Dephosphorylation of ovalbumin at pH 5-4 with intestinal phosphatase as function of time. Each reaction mixture contained 4.6 % ovalbumin and 0.006 % enzyme. * may represent a proteolytic degradation product. phosphatase is added to ovalbumin the proportion of the slower moving compon- ent A2 increases with time at the expense of Al, until A1 is completely transformed to a protein with the electrophoretic properties of A2 and with one phosphorus per mole.This monophosphorus ovalbumin, A2, crystallizes from ammonium sulphate as needles.GbKTRUDE E . PERLMANN 71 In the experiments shown in fig. 4, a phosphatase from calf intestine was added to ovalbumin. However, dephos- phorylation continues and a new component appears which moves more slowly than A2 and is designated as A3. That A3 does not contain phosphorus is revealed by the chemical analysis of the mixture and of A3 whose patterns are shown in lines 3 and 4 of fig. 4. After 24 h, ovalbumin has been converted to a phosphorus- free protein A3 which may still be crystallized as needles and whose mobility is that of the slowest component present as a trace in the original ovalbumin preparation. Although chemical differences other than the phosphorus content, e.g.small variations in the amino acid composition may exist between Al, A2 and A3, it is very probable that the mobility increments are due to the loss of charged phosphate groups and that the electrophoretic complexity of ovalbumin is adequately explained by the number of phosphorus atoms present in each component. The different mobilities of A1, A2 and A3 in a sodium phosphate buffer of pH 6.8 and 0.1 ionic strength, i.e. uAl = - 5.9 x 10-5, uA2 = - 5.2 x 10-5 and U A ~ = - 4.5 x 10-5cm2sec-1 V-1 are further illustrated with the patterns of fig. 5. These were obtained with a mixture of equal quantities of A1, A2 and A3 in a sodium phosphate buffer of pH 6.8 and 0.1 r/2. Fig. 5a is the pattern at a total protein concentration of 1.6 % and has the apparent composition of 45 % Al, Here A1 is again rapidly transformed into A2. rotol protein Klect~opho~etlc Atom4 phosphwu3 concentrution composition per mole protein per cent Computed Observed 45% Al 1.22 1.0 32% A ? 23% A3 Q ~ 1.6 FIG. 5.-Electrophoretic patterns of an artificial mixture of equal amounts of A1, .A2 and A3 in sodium phosphate buffer of pH 6-8 and 0.1 r/2.Electrophoresis was carried out for 12,600 sec at 6-1 V/cm. 32 % A2 and 23 % A3. If the phosphorus content of this mixture, column 3 and 4 of fig. 5, is now computed from this apparent composition, the value of 1-22 atoms per mole deviates considerably from the chemically determined value of 1.0. If, however, this same mixture is analyzed at a total protein concentration of 043 %, the pattern of fig. 5b is obtained and the apparent composition of 33 % Al, 34 % A2 and 33 % A3 approaches closely the true value.These results illustrate rather strikingly the large differences to be expected between the apparent and the true composition at high protein concentrations, especially if the mobility differences are small as in this instance.ls-l* Consequently, in all of the electrophoretic analyses reported here the total protein concentration has been kept as low as possible while still permitting the desired accuracy in the determination of the pattern areas. In these analyses phosphate buffers have been used because they afford a convenient means for stopping the action of phosphatases. In fig. 6, the electrophoretic mobilities of A1, the major component of oval- bumin, A2 and A3 are plotted as a function of pH.From the relative position of these curves it can be seen that the isoelectric pH values of the three ovalbumins are different and that the curves diverge until a constant mobility difference of Au = 0.6 x 10-5 is found in the pH range of 7 to 9. Comparison of this mobility decrement at pH 7.0 with the titration data again indicates that the net charge is altered by a value of - 2 at each step in the dephosphorylation. This result indicates that the two phosphate groups of A1 are esterified through the phosphate hydroxyl whose pK is > 12, i.e. each phosphate group is present as a monoester72 OVALBUMIN A N D CASEIN with two hydroxyls whose hydrogen ions are dissociated above pH 3. The con- vergence of the curves of fig. 6 at pH 5.0 suggests that the pK for the first of these hydroxyls is somewhat above its value in the free acid or simple organic esters, i.e.No //O \O- \O- protein-P-OH ~- H+ protein-P-0- PH 5 PH 7 However, electrophoretic analysis does not reveal the nature of the amino acid residue in the protein to which a phosphate is linked. Preliminary experiments have indicated that the two groups in A1 are attached to different amino acids.ls9 19 6.0 4.0 2.0 - 2.0 - 4.0 - 6.0 FIG. 6.-Mobilities of the ovalbumin component A1 and the dephosphorylated ovalbumins A2 and A3 as function of pH. This conclusion is foreshadowed by the results in fig. 3 and 4 since the prostate enzyme attacks only one phosphate group in Al, whereas the intestinal enzyme removes both. DEPHOSPHORYLATION OF PLAKALBum.-Since no phosphorus is lost in the ovalbumin --f plakalbumin transformation 1 1 9 20 it may be anticipated that plakal- bumin could be dephosphorylated in much the same manner as ovalbumin and with similar changes in mobility. Both the prostate and the irdestinal phosphatase dephosphorylate plakalbumin to monophosphorus P2 and phosphorus-free P3.As shown in fig. 7 by the super- imposed tracings of the patterns of these proteins, the relationship of PI, P2 and P3 is similar to that of the corresponding ovalbumins, AI, A2 and A3. The arrows This has proved to be the case.GERTRUDE E . PERLMAN 73 in fig. 7 are the positions taken from fig. 5 that boundaries due to Al, A2 and A3 have under identical experimental conditions. It is therefore not surprising that the mobility decrements of the plakalbumins over the entire pH range, as shown in table 1, are the same as those previously described for the ovalbumins,l5 but that the actual mobility values are less negative.This is further demonstrated by the isoelectric pH values given in table 2. For comparison those of the cor- responding ovalbumins are also listed. The shift of the isoelectric pH of the six proteins with the ionic strength is the same, but a constant difference of 0.08 pH d ascending 4 I 3 FIG. 7.-superimposed tracings of the patterns of plakalbumin and the dephosphorylated plakalbumins P2 and P3 in sodium phosphate buffer of pH 6.8 and 0.1 r/2. Electro- phoresis at 1.2 % protein concentration for 12,600 sec at 6.4 V/cm. The arrows indicate the boundary positions of the corresponding ovalbumins.units was found in the A ~ J A ~ and P1-+P2 transformation and of 0.1 in the A2-fA3 and p2-tP3 reaction. In the neighbourhood of the isoelectric pH the mobility change accompanying the removal of each phosphate group corresponds to a loss of about one negative charge, suggesting that under these conditions only one of PLAKALBUMINS P2 AND P3 IN BUFFER SOLUTIONS OF IONIC STRENGTH 0.1 TABLE 1 .-MOBILITIES OF THE PLAKALBUMIN COMPONENT AND THE DEPHOSPHORYLATED buffer 0.1 N HClS-0.5 N glycine 0-02 N NaAc+ 0-1 N HAc +0-08 N NaCl 0.1 N NaActO.1 N HAc 0.1 N NaAcS-0.09 N HAc 0-1 N NaAcf0.07 N HAc 0-1 N NaActO-06 N HAc 0.1 N NaAc t0.05 N HAc 0.1 N NaAc+ 0.01 N HAc 0.02 N NaCacf0-004 N HCac+O.OS N NaCl 0.02 N NaV+O.02 N HV +0*08 N NaCl 0.1 N NaVS-0.02 N HV 0.1 N NaVf0-005 N HV P H 3.0 3.9 4.64 4-72 4.8 4.84 4.95 5.64 6.78 7.8 8.6 9.2 U X 105 p1 f 6.54 -- 4-50 - 5-62 - 5-84 - 5.93 p2 6.53 - 3.93 p3 6.90 3.40 0.8 0.4 - 0.2 - 2.0 - 3-50 - 4.35 - 4-60 Ac = acetate ; Cac = cacodylate ; V = diethylbarbiturate.* taken from J. Gen. Physiol., 1952, 35, 711. 1 taken from J. Amer. Chem. Soc., 1949,71, 1146. C74 OVALBUMIN A N D CASEIN TABLE 2.-&MPARISON OF THE ISOELECTRIC pH VALUES OF OVALBUMINS ANI) PLAKALBUMINS APH ____ - -~ ionic strength PH A* A2 * A3* Ai-+Az A2jA3 0.1 4-58 4-65 4-74 0.07 0.09 0.05 4-63 4.70 4.80 0.07 0.1 0.02 4.68 4.75 4-85 0.07 0.1 0.01 4-7 1 4.80 4.90 0-09 0.1 P$ p2 p3 Pl-fPZ p2-'p3 0.1 4-72 4.80 4.88 0.08 0.08 0.05 4.77 4.86 4.95 0.09 0.09 0-02 4.82 4.9 1 5.0 0.09 0.09 0.0 1 4-86 4.96 5-06 0.I 0 . 1 0.1 0.05 0.02 0.01 Ai+b A2-fP2 A3+P3 0-14 0.15 0.14 0.14 0.16 0.15 0.14 0.16 0.15 0.15 0.16 0.16 * taken from .I. Gen. Physiol., 1952, 35, 711. $ taken from J . Arner. Chern. SOC., 1949, 71, 1146. the two hydroxyls of a phosphate has lost its proton, i.e. the phosphate group is present as -P-0-. The constant shift of 0.14 to 0.16 of the isoelectric pH \OH values in the Al-tP1, Az-tP2 and A3-tP3 processes, as pointed out previously, is due to the removal of the two carboxyl groups lost in the plakalbumin trans- formation. DEPHOSPHORYLATION OF CASEIN.-That casein is a mixture of several distinct proteins was first demonstrated by the solubility studies of Linders tr#m-Lang and Kodama.21.22 Mellander later showed that casein has three electrophoretic components,23 a-, B- and y-casein, which in 1944 were separated by Warner.24 The electrophoretic patterns of these purified fractions are reproduced in fig.8. Fig. 8a is that of unfractionated casein consisting of 75 % of the a-component and 25 % /%casein. The proteins of fig. 8b and c are those of the isolated fractions, the phosphorus content of the a- and /%casein being 1.0 and 0.6 % respectively. In the sodium phosphate buffer of pH 6.8 and 0.1 r/2, u, = - 7.6 x 10-5 and uB = - 3.4 x 10-5 cm2 sec-1 V-1. Although casein has thus far been considered to be resistant toward the action of purified phosphatases from mammalian tjssues,25-27 it has now been found that a-casein, the fraction with the higher phosphorus content and the more negative mobility, is readily dephosphorylated in the pH range from 5.6 to 6-6 by prostate phosphatase.28 This enzyme, however, has no effect on /%casein.During the dephosphorylation of a-casein the solubility of the protein decreases. Simultaneously, as shown in fig. 9, the electrophoretic behaviour of the protein changes and several new components with lower mobilities appear. In the experi- ments represented by fig. 9, 0-5 % a-casein in acetate buffer of pH 5.6 and T/2 0.1 was exposed to the action of the prostate enzyme and the enzymatic process followed by electrophoretic and chemical analyses. Fig. 9a is the tracing of the pattern of a-casein, 9b, c, d and e those after the protein has been in contact with the enzyme for 1, 3, 6 and 12 h, respectively.The electrophoretic composition and the loGERTRUDE E. PERLMANN 75 chemical analyses are listed in table 3. Here the components of the partially de- phosphorylated mixtures are identified on the basis of their mobilities, the values of which in a 0.1 ionic strength phosphate buffer are given in parentheses. As seen Unfractionated Casein a: (3 0 a a. ~ d a- -d a- casein a @-casein (3 C. a- -d FIG. 8.-EIectrophoretic patterns of unfractionated casein, a-casein and 8-casein in sodium phosphate buffer of pH 6.8 and 0.1 r/2. Electrophoresis was carried out at 0.5 % protein for 14,400 sec at a potential gradient of 4.95 Vlcm. a' -casein 12 ~OUPS e. FIG. 9.-Tracings of electrophoretic patterns of a-casein before and after treatment with prostate phosphatase.Electrophoresis was carried out in sodium phosphate buffer of pH 6.8 and 0.1 r J 2 for 10,800 sec at 4-75 V/cm. from fig. 9b and line 2 of table 3, after 1 h 6-7 % of the phosphorus is liberated. The main component has the same mobility as that of the a-casein, but two com- ponents, tentatively designated as u3- and a4- are present. After 3 h contact with the enzyme the boundary with the mobility of u-casein has disappeared. The major component of this mixture has a mobility of u = - 7.1 x 10-5 cm2sec-1 V-1.76 OVALBUMIN AND CASEIN TABLE 3.-DEPHOSPHORYLATION OF X-CASEIN WITH PROSTATE PHOSPHATASE AS FUNCTION OF TIME Each reaction mixture contains 0.5 o/, casein and 0.005 ;( enzyme in acetate buffer of pH 5.6 and 0.1 Ti2 time of apparent relative composition of components ( %) % % inorganic non-protein exposure to the enzyme a(-7.6) ai(-7.1) d2(-6.6) a3(-5.6) 4 - 3 .6 ) Lw5(-2.9) phosphorus nitrogen released formed (hours) - - - - 0 100.0 - - - 1 82.4 - - 3 - 74-2 - 6 Electrophoresis carried out in sodium phosphate buffer of pH 6-8 and 0.1 T/2. 8.5 9.1 - 6.7 2.2 17.2 8.6 - 14.9 3-6 44.0 6*8* 30.9* 13.7 4.6 25.3 6.7 12 - 41.1 4.5 38.0 6.1 10-3 28.2 12.6 - * not clearly resolved. As dephosphorylation continues, the apparent concentrations of the slower moving components increase. Due to the uncertainty of the molecular weight of casein, the number of phos- phate groups per mole is unknown, although the phosphorus content indicates that this number is much greater than in ovalbumin. However, if the removal of each phosphate group is accompanied by a definite mobility change, as with ovalbumin, the mobilities listed in table 3 may then possibly correspond to the loss of one phosphate group in the cc--ta~ transformation, one for the CCI+CC~, two for the ~~2-+tc3, four for the a3-fa4 and one for cc4-fcc5.The complexity of the partially dephosphorylated a-casein is hardly surprising in view of the high phos- phorus content of the starting material. Moreover, most of this phosphorus is linked to serine. In the last column of table 3 it is further shown that the dephosphorylation of the protein is accompanied by the formation of trichloroacetic acid soluble nitrogen. No explanation of this phenomenon can be given at the present time. Although the action of the prostate enzyme on ovalbumin was not accompanied by the production of non-protein nitrogen, a proteolysis of casein cannot be excluded, particularly since some trichloroacetic acid soluble nitrogen appears even after liberation of phosphorus has ceased.The mechanism of the dephosphorylation of a-casein is still obscure, but a few general conclusions can be drawn from this work. It is clear that cc-casein contains a certain number of phosphate groups with ionizable hydroxyls which contribute to the net charge and thus also to the electrophoretic mobility of the protein. Such phosphate groups, as shown by the investigation of the ovalbumin, are readily attacked by phosphatases. Finally, it should be stressed that electrophoresis can be of considerable im- portance in the study of enzymatic reactions which induce minute changes in a protein molecule.Moreover, the combination of this method with chemical analysis will facilitate studies directed toward gaining insight into protein structure. These conditions are not fulfilled with P-casein. 1 Svedberg and Eriksson-Quensel, J . Amer. Chem. Sac., 1934, 56,409. 2 Annetts, Biachern. J., 1936, 30, 1807. 3 Tiselius and Eriksson-Quensel, Biochem. J,, 1939, 33, 1752. 4 Petermann and Pappenheimer, J. Physic. Chem., 1941, 45, 1. 5 Petermann, J. Physic. Chem., 1942, 46, 183. 6 Giintelberg and Linderstrfjm-Lang, Cornpt. rend. trav. lob. Carlsberg, 1949, 27, no. 1. 7 Longsworth, Cannan and MacInnes, J. Arner. Chem. Sac., 1940, 62, 2580. 8 S~rensen and H+yrup, Compt. rend trav. lab. Carlsberg, 1915-17, 12. 9 Longsworth, J. Amer. Chem. Sac., 1939, 61, 529. 10 Linderstr$m-Lang and Ottesen, Nature, 1947, 159, 807. 11 Linderstr$m-Lang and Ottesen, Compt. rend. trav. lab. Carlsberg, 1949, 26, 403. 12 Eeg-Larsen, Linderstrgjm-Lang and Ottesen, Arch. Biochern., 1948, 19, 340.G E R T R U D E E. PERLMANN 77 13 Perlmann, J . Amer. Chem. SOC., 1949, 71, 1146. 14 Cannan, Kibrick and Palmer, Ann. N. Y. A c d . Sci., 1941, 41, 243. 15 Perlmann, J. Gen. Physiol., 1952, 35, 71 1. 16 Svensson, Arkiw. Kemi. Min. Geol. A, 1943, 17, no. 14 ; 1946, 22, no. 10. 17 Perlmann and Kaufman, J. Amer. Chern. SOC., 1945, 67, 638. 18 Longsworth, J. Physic. Chem., 1947, 51, 171. 19 Perlmann, Abstr. 119th Meeting, Amer. Chem. SOC., April, 1951. 20 Perlmann, Nature, 1949, 164, 961. 21 LinderstrQm-Lang and Kodama, Compt. rend. trav. lab. Carlsberg, 1925, 16, no. 1. 22 Linderstr#m-Lang, Compt. rend. trav. lab. Carlsberg, 1929, 17, no. 9. 23 Mellander, Biochem. Z., 1939, 300, 240. 24 Warner, J. Amer. Chem. Sac., 1944, 66, 1725. 25 Rimington and Kay, Biochern. J., 1926, 20, 777. 26 Schmidt and Thannhauser, J. Biol. Chem., 1943, 149, 369. 27 Anagnostopoulos, Pacht, Bourland and Grabar, Bull. SOC. Chim. Bioi., 1951, 28 Perlmann, J. Amer. Chem. Soc., 1952, 72. 33, 699.