IMMUNOCHEMICAL RELATION OF BENCE-JONES PROTEINS TO NORMAL SERUM PROTEINS*

Transcription

1 IMMUNOCHEMICAL RELATION OF BENCE-JONES PROTEINS TO NORMAL SERUM PROTEINS* BY II. F. DEUTSCH, C. H. KRATOCHVIL,? AND ARNOLD E. REIFS (From the Departnient of Physiological Chemistry, cniversity of Wisconsin, Madison, Wisconsin) (Received for publication, January 17, 1955) The Bence-Jones proteins have been reported to contain two immunological entities. However, in view of their varying physicochemical properties, it might be anticipated that they would show great er complexities in this respect. Although they are generally considered to be discrete substances and unrelated antigenically to the normal human serum proteins, the previous experimentation employed permits a reasonable doubt as to the validity of the conclusions reached. Thus this point appeared worthy of reinvestigation. Collier and Jackson (1) have indicated that Bence-Jones proteins probably appear exclusively in the urine of all patients with multiple myeloma. The disease is usually concomitant with a hyperglobulinemia of one or more serum globulins, and the possibility presents itself that the Bence- Jones proteins may be fragments of these molecules. This was investigated by quantitative immunochemical studies of Bence-Jones proteins, r-globulin fractions, and pooled normal human serum prot,eins. The results show that antisera to the two Bence-Jones proteins gave marked cross-reactions with normal serum proteins. This indicates a structural similarity and strongly suggests that the Bence-Jones proteins may indeed be fragments of certain normal serum protein molecules. EXPERIMENTAL Electrophoretic analyses were carried out in diethyl barbiturate buffer of.1 ionic strength, ph 8.6, at potential gradients near 5.5 volts per cm. The sedimentation analyses were performed in the Spinco apparatus at gravitational fields near 2.5 X lo6 and employed a potassium phosphate buffer,.2 ionic strength, ph 7.4. The normal yz- and rl-globulin fractions of human serum were analogous to those previously employed (2, 3). In addition to these, two Bence- Jones proteins, which will be designated BJ-LP and BJ-ZE, the rz-globu- Downloaded from by guest on October 18, 218 * This work was supported in part by a grant from the United States Public Health Service. A preliminary report was presented at the Sixth International Congress of Cancer, Sao Paulo, Brazil, July, t Captain, United States Air Force (Medical Corps). $ Present address, The Lovelace Foundation, Albuquerque, Xew Mexico. 13

2 34 BENCE-JONES PROTEIN lin of a multiple myeloma serum designated rz-kl, and normal human serum were employed as precipitinogens. The properties of BJ-LP have been described in the previous report (4). The BJ-ZE was prepared by a combination of salt and ethanol fractionation techniques. It was found to possess less than.3 per cent methionine when subjected to microbiological assay. This means that the BJ-ZE protein has less than.5 mole V2 -KL BJ-ZE E 1 FIG. 1. Electrophoretic (left) and ultracent.rifuge (right) diagrams of proteins r2-kl and BJ-ZE. Time of electrophoresis,?n-kl 29 minutes, BJ-ZE 2 minutes; time of ultracentrifugation,?z-kl 83 minutes, BJ-ZE 156 minutes. ~ BJ-ZE Downloaded from by guest on October 18, 218 of this amino acid per 35, gm. and may be considered to lack it. Results of electrophoretic and ultracentrifugal experiments on antigens BJ- ZE and rz-kl appear in Fig. 1. BJ-ZE shows two components in the ultracentrifuge. The slower of these sediments with a velocity which suggests a molecular weight near 35,. The faster component may be an aggregate of the first if the ultracentrifugal properties of this protein are analogous to those of BJ-LP in 1 We wish to thank B. E. Kline for carrying out this analysis. The limits of the assay employed would not allow for the detection of less than.3 per cent methio nine.

3 H. F. DEUTSCH, C. H. KRATOCHVIL, AND A. E. REIF 15 this respect (4). Its low solubility below ph 6 prevented experimentation in the range in which BJ-LP has been found to exist as a monomer. Both the Bence-Jones proteins and the y-globulins were sufficiently pure, and their physical properties were such that they could be stated to be free of contamination with each other. This obviates the consideration that the immunological cross-reactions experienced could be due to contamination. R.abbit and chicken antisera were prepared against BJ-LP, BJ-ZE, and +yz-kl by the adjuvant technique of Freund and McDermott (5). In some cases the y-globulins of the rabbit antisera were separated by the method of Nchol and Deutsch (6). Quantitative precipitin reactions were carried out by methods previously described (7). 6 z LP ANTIGEN E 4 k a 5 2 a 2 P /.+. ANTIGEN N ADDED FIG. 2. The reactions of rabbit anti-bj-lp serum with various precipitinogens Results The data will be considered with respect to the reactions of various precipitinogens with antiserum to BJ-LP, BJ-ZE, and r2-kl. Reactions of BJ-LP Antibody-The results of the reaction of BJ-LP with its rabbit antiserum are presented in Fig. 2. This antigen stimulated good antibody production in rabbits to give a precipitin system which obeys the Heidelberger and Kendall (8) relationship. At the point of maximal precipitation the composition of the specific precipitate closely approximates AbzAn when the rabbit antibody (Ab) is considered to have a molecular weight of 16, (9) and the BJ-LP antigen (An) to be 31, (4). This antiserum also reacted strongly to normal yz- and rl-globulins, normal human serum, and BJ-ZE. rz-kl, however, failed t.o give any significant precipitate with a rabbit BJ-LP antibody preparation when studied over antigen additions from 5 to 353 y of?z-kl antigen iy. The latter amount gave only 15.4 y of N precipitate, the amount at 118 y of antigen S being 5.6 y. The amount of specific precipitate given by normal human serum appears to be in proportion to its content of r-globulins. Downloaded from by guest on October 18, 218

4 16 BENCE-JONES PROTEIN The precipitin reaction of rabbit anti-bj-lp serum with y-globulins is not due to the presence of Bence-Jones protein. Thus the addition of 9 y of yz-globulin N to 1 ml. of BJ-LP rabbit antiserum yields an amount of specific precipitate equal to 15 y of the homologous antigen. The y2- globulin employed would have had to contain near 17 per cent BJ-LP, an amount far in excess of that which could be readily detected both by electrophoretic and ultracentrifugal analysis. The BJ-LP antiserum does not contain antibody directed against a y- globulin impurity of the antigen. This is shown by the results of what will be termed inhibition reactions. These consist in determining the amount of specific precipitate given by BJ-LP in the presence of a second Results of Block& I Reactions of Rabbit BJ-LP Antibody Reaction by Various Proteins - B J-LP added Blocking protein ~-.- I Amount Ppt. Ppt. blocked.- Y N yz-globulin wkl y1 -Globulin Human serum < BJ-ZE - -I- TABLE TN protein. The homologous antigen. is employed at a level that will give maximal specific precipitation or be in the region of slight antigen excess. The simultaneous addition of a second antigen that combines with antibody produced against the homologous antigen will result in a marked decrease in specific precipitation because of conditions that simulate an excess of antigen. If a specific antibody to the second protein is present, the precipitate formed will be the sum of the individual reactions. Data of this type of experiment are presented in Table I and clearly show that all of the proteins tested block the homologous reaction. The y2-kl antigen gave no specific precipitation at any antigen level, but exerted a strong inhibiting effect. This indicates that the two proteins in question possess immunochemical similarities, even though they are vastly different in their ultracentrifugal and electrophoretic properties. The reactions of chicken anti-bj-lp serum under isotonic conditions and in 8 per cent NaCl are presented in Fig. 3. Goodman et al. (1) have I None Downloaded from by guest on October 18, 218

5 H. F. DEUTSCH, C. H. RRATOCHVIL, AND A. E. REIF.I7 reported that far greater specific precipitation occurs for chicken antigenantibody systems at higher salt levels. The marked differences previously reported do not obtain for the BJ-LP system, and the broad zone of maximal precipitation at the higher salt concentration is indicative of a decreased sensitivity. For this reason no further studies were carried out in 8 per cent NaCl. Chicken anti-bj-lp serum reacts weakly with normal human serum, and, in contrast to rabbit anti-bj-lp serum, yields a small amount of precipitation with r2-kl. However, the data for blocking reactions with normal human serum and yn-kl, as presented in Fig. 3, reveal that these pro- 56 = 4 t" CQ a 5,,, 24 ii? 1" 8 J+-. ANTIGEN N ADDED FIG. 3 FIG. 4 FIG. 3. The reactions of chicken anti-bj-lp serum with BJ-LP at 8 per cent NaCI (A) and at.15 M NaCl-borate buffer (). The blocking reactions with ~s-kl () and normal human serum (m) were carried out at a constant level of 23 y of BJ-LP nitrogen. FIG. 4. The reaction of BJ-ZE with its rabbit anti-r-globulin. teins very strongly block the homologous reaction, normal human serum being much more effective than -y2-kl. Antisera to BJ-ZE-BJ-ZE did not prove to be an effective antigen. A weak rabbit antiserum was obtained after prolonged immunization. The r-globulin fraction was separated and reconstituted to twice its serum concentration to give a workable precipitin system. The results of the immunochemical reaction of this antibody preparation are presented in Fig. 4. The antibody also reacted with all of the other antigens tested. From the data of Table II it is apparent that the specific precipitation experienced with heterologous antigens is cross-reactive in nature. It is interesting that BJ-LP blocks the reaction of BJ-ZE with its antibody, whereas the reverse was not true (see Table I). Once again the amount of specific precipitation given by heterologous Downloaded from by guest on October 18, 218

6 18 BENCE-JONES PROTEIN antigens was too great to be attributed to contamination with Bence-Jones protein. Furthermore, the results of the blocking reactions also failed to support this possibility. Antisera to rz-kg-both rabbit and chicken antibodies to this protein were utilized. The specific precipitin results for chicken antiserum (Fig. TABLE Results of Blocking Reactions of Rabbit BJ-ZE Antibody Reaction by Various Proteins.- - BJ-ZE added Blocking protein Amount ~ rz-globulin w-i= I -rt-globulin I Human serum I I BJ-LP I II T.- -~-- Ppt Ppt. blocked 7N Downloaded from by guest on October 18, 218 FIG. 5. The reactions of ~Q-KL with its chicken antiserum in 8 per cent NaCl (A) and under isotonic conditions (). l, the results of the precipitin reactions for the r-globulin preparation of this serum. 5) reveal that the system is very sensitive to an increase in salt concentration, but the antigen excess region does not exhibit the usual decreased precipitation. Since the titer of the antiserum under isotonic conditions was somewhat low, the yz-globulin portion of the antiserum was separated and reconstituted to give a precipitin potency under isotonic conditions of the level shown in Fig. 5. This preparation was employed in further studies.

7 H. F. DEUTSCH, C. H. KRATOCHVIL, AND A. E. REIF 19 The two Bence-Jones proteins blocked the chicken antibody only weakly (Table III). It is interesting that BJ-ZE reacted strongly with the chicken rz-kl antibody at the level employed, but only a portion of this was blocking antibody. The proteins of normal human serum blocked the reaction completely. This was also found for rabbit anti-rz-kl globulin and will be reported more fully later. The rabbit antibody to r2-kl was only TABLE Results of Blocking Reactions of Chicken yz-kl Antibody Reaction by Human Serum and Bence-Jones Proteins III -/vkl added Blocking protein Amount Ppt. Ppt. blocked Human serum I BJ-LP B J-ZE TABLE Results of Blocking Reactions of Rabbit yz-kl Antibody Reaction by Bence-Jones Proteins 7,.KL added Blocking protein 18.2 BJ-LP 18.2 I BJ-ZE 18.2 Amount IV Ppt Ppt. 15 blocked (+6)* Downloaded from by guest on October 18, 218 * Limit of assay method. slightly blocked by BJ-LP, as shown by the data of Table IV. Thus the two Bence-Jones proteins block the homologous reaction of a yt-globulin with its chicken antibody only slightly and show little or no interference with the analogous rabbit precipitin system. DISCUSSION The two Bence-Jones proteins utilized in the present studies have uniquely different immunological properties. One of them, BJ-ZE, would appear to be similar to the proteins usually studied which are low or lacking in methionine (11-14). Immunologically it was similar to the Bence-

8 11 BENCE-JONES PROTEIN Jones proteins studied by Collier and Jackson (1) and by Moore, Kabat, and Gutman (15) in that it was a relatively poor antigen. In the present studies it is apparent from the results of blocking experiments that a portion of its antibody combines with serum globulins. Thus in such systems certain of the serum proteins behave as incomplete antigens by showing an ability to combine with antibody, even though they may fail to form specific precipitates. BJ-LP is a much better antigen than BJ-ZE, and it shows considerable cross-reaction with the serum globulins. A serum protein such as rz-kl, which fails to form a specific precipitate, does, however, combine with a portion of the antibody, as illustrated by the results of the blocking experiments. Such results indicate that the Bence-Jones proteins possess structural similarities to the serum globulins. It should be stressed that the difference in immunological response of rabbits and chickens to serum yz-globulins (3) is likewise reflected when Bence-Jones proteins are used as antigens. A likely interpretation of the results presented in the present studies is that the Bence-Jones proteins are actually pieces (port.ions) of certain globulin molecules, as previously suggested by Rundles et al. (16). It is possible that in the accelerated synthesis of certain globulins in the condition of multiple myeloma a step in the normal synthetic mechanism may be limiting. Such limitation could lead to the formation of a fragment of the usual globulin molecule. The Bence-Jones proteins possess the electrophoretic mobility of the elevated serum globulins, and, when a Bence- Jones-positive myeloma patient exhibits a serum hyperglobulinemia characterized by two elevated globulin components, the urinary Bence-Jones protein likewise shows two components.2 The specific fault involved in the formation of Bence-Jones proteins may be a failure to incorporate methionine adequately under certain conditions of accelerated globulin synthesis. Numerous studies have demonstrated the low level or absence of this amino acid in Bence-Jones protein (ll-14), whereas the serum globulins of similar electrophoretic mobility always contain methionine (17, 18). Other workers have also shown that the total liver tumor proteins are characterized by a marked decrease in methionine compared to the normal tissue (19). The fact that many different types of Bence-Jones proteins exist (4, 2-22) indicates that a possible metabolic failure or limitation with respect to methionine incorporation may involve a number of specific blood serum proteins. At times a number of proteins may be affected in such manner in a given individual. Downloaded from by guest on October 18, 218 The authors wish to acknowledge the technical assistance of Mr. George S. Steinmetz during a portion of this investigation. 2 Kratochvil, C. H., and Deutsch, H. F., unpublished data.

9 H. F. DEUTSCH, C. H. KRATOCHVIL, AND A. E. REIF 111 SUMMARY Highly purified Bence-Jones proteins have immunological similarities to those of normal serum globulins. This indicates structural similarities and suggests that they may be portions of normal serum protein molecules. BIBLIOGRAPHY 1. Collier, F. C., and Jackson, P., New England J. Med., 248, 49 (1953). 2. Cohn, M., Wetter, L. R., and Deutsch, H. F., J. Immunol., 64, 381 (195). 3. Deutsch, H. F., Arch. Biochem. and Biophys., 49,356 (1954). 4. Deutsch, H. F., J. Biol. Chem., 216, 97 (1955). 5. Freund, J., and McDermott, K., Proc. Sot. Exp. Biol. and Med., 49, 548 (1942). 6. Nichol, J. C., and Deutsch, H. F., J. Am. Chem. Sot., 7, 8 (1948). 7. Cohn, M., Wetter, L. R., and Deutsch, H. F.,.I. Immunol., 61, 283 (1949). 8. Heidelberger, M., and Kendall, F. E., J. Exp. Med., 62, 467 (1935). 9. Kabat, E. A., and Pedersen, K. O., Science, 87, 372 (1938). 1. Goodman, M., Wolfe, II. R., and Norton, S., J. Immunol., 66, 225 (1951). 11. Dent, C. E., and Rose, G. A., Biochem. J., 44, 61 (1949). 12. Roberts, E., Ramasarma, G. B., and Lewis, H. B., Proc. Sot. Exp. Biol. and Med., 74, 237 (195). 13. Schreier, K., and Pltickthun, II., KZin. H ochschr., 3, 677 (1952). 14. Agren, G. A., Acta them. &and., 6, 1232 (1952). 15. Moore, D. II., Kabat, E. A., and Gutman, A. B., J. Clin. Invest., 22, 67 (1943). 16. Rundles, R. W., Cooper, G. R., and Willett, R. W., J. CZin. Invest., 3,1125 (1951). 17. Brand, E. R., Kassell, B., and Saidel, L. J., J. CZin. Invest., 23,437 (1944). 18. Grisolia, F. T., and Cohen, P. P., Cancer Res., 13, 851 (1953). 19. Schweigert, B. S., Guthneck, B. T., Price, J. M., Miller, J. A., and Miller, E. C., Proc. Sot. Exp. Biol. and Med., 72,495 (1949). 2. Kabat, E. A., and Mayer, M. M., Experimental immunochemistry, Springfield, 476 (1948). 21. Putnam, F. W., and Stelos, P., J. BioZ. Chem., 23,347 (1953). 22. Putnam, F. W., and Miyake, A., Science, 12,848 (1954). Downloaded from by guest on October 18, 218

10 IMMUNOCHEMICAL RELATION SERUM PROTEINS NORMAL OF BENCE-JONES PROTEINS TO H. F. Deutsch, C. H. Kratochvil and Arnold E. Reif J. Biol. Chem. 1955, 216: Access the most updated version of this article at Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's alerts This article cites references, of which can be accessed free at tml#ref-list-1 Downloaded from by guest on October 18, 218