/ . Biochem., 79, 91-105 (1976)
Proteins and Immunoglobulins1 Fumio KISHIDA,2 Takachika AZUMA, and Kozo HAMAGUCHI' Department of Biology, Faculty of Science, Osaka University, Toyonaka, Osaka 560 Received for publication, July 23, 1975
The formation of the interchain disulnde bonds in partially reduced Bence Jones proteins and immunoglobulins was studied in the presence of glutathione. It was found that only oxidized glutathione (GSSG) was effective for the formation of the interchain disulnde bonds in type X Bence Jones proteins and IgG. In type * Bence Jones proteins, on the other hand, no formation of the inter L-L disulnde bond was observed in the presence of GSSG at above pH 6. The kinetic pattern of disulfide bond formation of Bence Jones proteins was well interpreted by assuming that two monomers of a type X protein dimer are discriminated (monomers 1 and 2) and that only an intermediate in which the SH group on monomer 1 is blocked with GSSG can form a disulfide-bonded dimer and the intermediate in which the SH group on monomer 2 is blocked with GSSG can not. Comparison of the kinetic data for the formation of the interchain disulfide bonds of IgG with those for Bence Jones proteins suggested that H chain-GSSG mixed disulfide is a principal intermediate for the formation of the inter H-L disulfide bond.
covalent assembly of H and L chains of a myeloma protein have been studied in our laboratory (1—4). Studies on the formation of the interchain disulfide bonds should also be very important for understanding the formation of the three-dimensional structure of immunoglobulins. Kinetic studies on the biosynthesis of immunoglobulins show that covalent intermediates such as Hi, HL, and HjL are formed during the assembly of H and L chains, whereas covalent Lt is not formed at all within the cell (5, 6). On the other hand, disulfide-bonded dimers of L chains are frequently found in human Bence Jones proteins, especially in type X proteins ( 7 ) . These facts suggest that the characteristics of the inter
It is well-known that immunoglobulin G consists of two heavy (H) and two light (L) chains which interact by both noncovalent interactions and covalent disulfide bonds. The kinetic and equilibrium properties of the non1
This work was supported in part by a grant from the Ministry of Education, Science and Culture, of Japan. 1 Present address: Research Department, Pesticides Division, Sumitomo Chemical Co., Ltd., Takarazuka, Hyogo 665. ' To whom correspondence should be sent. Abbreviations: DTNB, 5,5'-dithiobis(2-nitrobenzoic acid); DTT, dithiothreitol; EDTA, ethylenediaminetetraacetic acid; GSH, reduced glutathione; GSSG, oxidized glutathione; SDS, sodium dodecyl sulfate. Vol. 79, No. 1, 1976
91
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
Formation of Interchain Disulfide Bonds in Bence Jones
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
92
MATERIALS AND METHODS Proteins — Purification and determination of the antigenic type of Bence Jones proteins were done according to the procedure described previously (17, 18). Six type c proteins, Ta, Ham, Ya, Iwa, Tew, and Mat, and seven type X proteins, As, Fu, Ori, Tod, Sh, Ni, and Nag, were used. SDS gel electrophoresis showed that Ta, Ham, As, Fu, Ni, Sh, and Nag are dimers with a disulfide bond but contain a small amount of monomer with no disulfide bond. The other specimens contain both dimers with a disulfide bond and monomers.
It was confirmed that Ham, Sh, Nag, Ni, Tod, Ta, and Fu remain dimers even after reduction of the interchain disulfide bond at the protein concentrations used in the present experiments. The preparation of human myeloma protein Jo (IgGl, K) was described in a previous paper (2). Normal human immunoglobulin G (Fraction II) was obtained from Sigma Chemical Co. Reagents—Sodium dodecyl sulfate (SDS), 5, 5' - dithiobis(2 - nitrobenzoic acid) (DTNB), iodoacetamide, and ethylenediaminetetraacetic acid (EDTA) were purchased from Wako Pure Chemicals Co. Dithiothreitol (DTT) and urea were from Nakarai Pure Chemicals Co. and GSSG and GSH were from Sigma Chemical Co. Urea was recrystallized from 80% ethanol. All other reagents were of reagent grade and were used without further purification. Reduction of Interchain Disulfide Bonds— About 1-5% protein in 0.2 M Tris-HCl buffer containing 2 mM EDTA (pH 8.0) was allowed to react with 20 mM DTT for 40 min at room temperature. Under these conditions, no intrachain disulfide bonds of Bence Jones proteins and immunoglobulins were reduced. We refer to this procedure as partial reduction. The partially reduced protein was separated from residual reagents by passage through a column of Sephadex G-25 equilibrated with 0.01 M Tris-HCl buffer containing 1 mM EDTA at pH 8.0 or with 0.1 M KC1 containing 1 mM EDTA. Interchange Reaction of the SH Groups of Bence Jones Proteins and Immunoglobulins in the Presence of Glutathione—A solution of a partially reduced protein was mixed with the same volume of a solution containing a desired amount of glutathione. Unless otherwise specified, a glutathione was dissolved in 0.2 M Tris-HCl buffer containing 2 mM EDTA at pH 8.0. In order to vary the pH or the ionic strength of the reaction mixture, glutathione was dissolved in an appropriate buffer and the ionic strength was adjusted with KC1. The buffers used were as follows glycine-HCl (pH 2-4), acetate (pH 4-6), Tris-HCl (pH 7-9), and glycine-KOH (above pH 9). The reaction was continued for a given time at 25°. The progress of the reaction was followed by SDS / . Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
L-L disulfide bond are different from those of the inter H-L or inter H-H disulfide bond. The relative lability of interchain disulfide bonds in immunoglobulins has been studied employing various reducing reagents, and several intermediates were found under reducing conditions (8, 9). However, studies on the formation of disulfide bonds from the protein in which interchain disulfide bonds are reduced seem necessary to interpret the results of the covalent assembly of immunoglobulins in vivo. Saxena and Wetlaufer (10) employed a system containing reduced and oxidized glutathione (GSH and GSSG) for the regeneration of disulfide bonds from reduced lysozyme [EC 3. 2.1.17]. Recently, Petersen and Dorrington { 77) applied this system to the reformation of the interchain disulfide bonds in immunoglobulin G. Similar experiments were also performed by Sears et al. (72). However, the mechanism of the formation of the interchain disulfide bonds in immunoglobulins is less well understood. Recent X-ray crystallographic studies (13—16) show that the three-dimensional structure of a type X Bence Jones protein is very similar to that of the Fab' fragment of immunoglobulin. Therefore, comparative studies on the formation of inter L-L and inter L-H disulfide bonds may be interesting in view of the similarity of the conformation. In the present paper, we report the kinetics of disulfide bond formation in Bence Jones proteins and immunoglobulin G in the presence of glutathione and offer a possible mechanism for these reactions.
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
In order to study the effect of urea on the inter L-L disulfide bond formation, a partially reduced Bence Jones protein was incubated in urea solution at a given concentration containing 1 mM EDTA for 8.5 hr at room temperature. Then a urea solution of the same concentration containing GSSG was added and the reaction mixture was left to stand for 3 hr at 25°. The extent of dimer formation was analyzed by SDS gel electrophoresis. Oxidation in the Presence of Cu2+—In this experiment, EDTA was omitted from the system. Degassed and deionized water was used throughout the experiment. Reoxidation of partially reduced protein was allowed to proceed in the presence of 10 ^M CuI+ (CuSO4) at room temperature in a polyethylene beaker which was covered with moistened gauze. An aliquot of the reaction mixture was taken at a given time and the SH content was determined by titration with DTNB. No effect of Cu t+ ions on the results of SH group titration was observed since the titration was carried Vol. 79, No. 1, 1976
out in the presence of a large excess of EDTA (1 mM) over CuI+ ions, as described below. Titration of Free SH Groups with DTNB —To 3 ml of about 0.1-0.3% protein either in 0.01 M Tris-HCl buffer at pH 8.0 or in 0.01 M acetate buffer at pH 6.5 (both containing 1 mM EDTA) was added 0.5 ml of freshly prepared DTNB solution in 1 M Tris-HCl buffer at pH 8.0. The mixture was left to stand for 30 min at room temperature and its absorbance was measured against a reagent blank using a Hitachi automatic recording spectrophotometer, model 323. A value for the molar extinction coefficient of reduced DTNB of 13,600 was used to convert the absorbance at 412 nm to SH content (79). SDS Polyaerylamide Gel Electrophoresis— Electrophoresis in the presence of 0.1% SDS was carried out at pH 7.5 in 7% polyacrylamide gel using a glass tube (0.6x8 cm) according to the method described by Weber and Osborn (20), but 2-mercaptoethanol was omitted. Before electrophoresis, samples were incubated with 1% SDS containing 50% glycerol for 24 hr at room temperature (for Bence Jones protein) or for 1 min at 100° (for immunoglobulins). After the run, the gels were stained overnight with 0.25% Coomassie brilliant blue in 50% methanol-9.2% acetic acid and destained with 5% methanol-7.5% acetic acid. Densitometric scanning of the gels was carried out at 600 nm using a Fuji Riken densitometer, type FD-4. pH Measurements—A Hitachi-Horiba pH meter, model F-5, was used. RESULTS Effect of Glutathione Concentration—Figure 1A shows the effect of GSSG concentration on the formation of the inter L-L disulfide bond of Nag protein (type X) at pH 8.0 as determined by SDS gel electrophoresis. In this figure, the percent of the dimer with a disulfide bond obtained after 180 min is plotted against the concentration of GSSG. The protein concentration was about 4x 10~5 M. In the absence of GSSG, no inter L-L disulfide bond was formed within 180 min when EDTA was present in the reaction medium. The extent
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
gel electrophoresis and by titration of the free SH groups with DTNB. For analysis by SDS gel electrophoresis, the reaction was stopped by the addition of a sufficient amount of iodoacetamide with respect to the free SH groups. The reaction mixture was then dialyzed against water at 5° and analyzed. For the titration of SH groups, the reaction was terminated by lowering the pH to 6.5 with 1M acetic acid followed by gel nitration on a Sephadex G-25 or G-50 column equilibrated with 0.01 M acetate buffer containing 1 mM EDTA at pH 6.5, at which the formation of interchain disulfide bond is minimized, as will be described in the "RESULTS." An aliquot of the eluted protein solution was withdrawn and the number of free SH groups was immediately determined with DTNB. Another aliquot of the solution was then incubated at 25° after increasing the pH to 8.0 using 1 M Tris-HCl buffer (final concentration of Tris, 0.1 M). After a given reaction time, the solution was acidified to pH 6.5 with 1M acetic acid and passed through a Sephadex G-25 column equilibrated with 0.01 M acetate buffer containing 1 mM EDTA at pH 6.5, then the number of free SH groups was determined by titration.
93
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
94 100
100
0.2
0.4
0.6
0.8
1.0
o
0.2
CONCENTRATION OF GSSG (mM)
0.4
0.6
0.8
1.0
X =GSSG/(GSSG+GSH)
Fig. 1. Reformation of the inter L-L disulfide bond from partially reduced Nag protein (type X) as a function of GSSG concentration (A) and mole fraction of GSSG (x) in the mixture of GSSG and GSH (total concentration, 1 mM) (B). All the solutions contained 1.5 mM EDTA. The yield of the dimer at 180 min was determined by SDS polyacrylamide gel electrophoresis. Protein concentration, 0.1%, pH 8.0, 25°.
of disulfide bond formation increased up to 0.2 mM GSSG and became constant above 0.4 mM. The maximum yield of the dimer was about 50% of the total protein. Similar experiments were done with various molar ratios of GSSG and GSH. Figure IB shows a plot of the extent of inter L-L disulfide bond formation at 180 min against the mole fraction of GSSG (x=GSSG/(GSSG+ GSH)). The total concentration of glutathione (GSSG+GSH) in the reaction mixture was fixed at 1 mM. At values of x below 0.1, no appreciable formation of inter L-L disulfide bond was observed. At values of x above 0.1, the
yield of the dimer with a disulfide bond increased and became constant above *=0.5. The maximal yield of the dimer, however, was about 50%, which was essentially the same as that in the presence of GSSG alone (Fig. 1A). This indicates that only GSSG is effective in the formation of inter L-L disulfide bond, and further experiments were done in the presence of 1 mM GSSG. Table I summarizes the percent of the dimer with a disulfide bond formed after 180 min for six type K and seven type X Bence Jones proteins. It is remarkable that all the type X proteins except Tod protein can form
TABLE I. The yields of the dimers with an interchain disulfide bond determined at 180 min from partially reduced Bence Jones proteins in the presence of 1 mM GSSG and 1.5 mM EDTA. The yields were determined by SDS polyacrylamide gel electrophoresis and are expressed as per cent of the total protein. pH 8.0, 25°. Protein
Type
Yield of dimer with S-S bond (%)
Protein
Ta Ham
c
3
As
K
3
Fu
Ya
£
0
Sh
Iwa Tew
C
Nag
c
0 0
Mat (D)» Mat (M)»
c
0
Ori Ni
c
0
Tod
Type
Yield of dimer with S-S bond (%)
X X X X X X X
96 83 64 54 39 32 4
* The properties of the dimer (D) and monomer (M) of Mat protein are described in Ref. 18. J. Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
0
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
95
O-SH
SSG
+ GSSG
+ GSSG O-SSG
+ GSH
(1)
• S j j j + GSH ( 2 ) O-s
+ GSH
(3)
Reactions 1 and 2 correspond to the formation of the mixed disulfide and reaction 3 represents the formation of the interchain disulfide bond. Since all the reactions involve the disappearance of SH groups, the results of the titration of free SH groups at a given time after removal of GSSG and GSH by gel filtration offer a measure of the progress of the thiol-disulfide interchange reactions. Moreover, if GSSG is separated from the reaction mixture and the reaction mixture is then incubated, reactions 1 and 2 will not occur and only reaction 3 may proceed. Figure 2 shows the percent of the remaining free SH groups of partially reduced Nag and Fu proteins (type Z) titrated immediately after the removal of GSSG (curve 1) and after the separation of GSSG followed by incubation of the solution for about 15 hr (curve 2). The time course of the formation of the dimer with a disulfide bond as determined by SDS gel electrophoresis is also shown in this figure {curve 3). It is apparent that dimer formation occurs with an accompanying decrease in the titratable SH groups. However, the extent of disappearance of the SH groups did not completely parallel the extent of formation of the inter L-L disulfide bond. This is more marked in the case of Nag protein. Both the rate and the final yield of dimer formation for Nag protein were different from those Vol. 79, No. 1, 1976
90 120 150 180 TIME (min) Fig. 2. Disappearance of SH content and reformation of the inter L-L disulfide bond from partially reduced Fu protein (type X) (A) and Nag protein (type X) (B) in the presence of 1 mM GSSG and 1.5 mM EDTA. Protein concentration, 0.1-0.3%, pH 8.0, 25°. Curve 1, remaining SH content determined immediately after the removal of GSSG; curve 2 remaining SH content in solutions incubated for 15 hr after the removal of GSSG; curve 3, yield of the dimer as determined by SDS polyacrylamide gel electrophoresis. 30
60
for Fu protein. The disappearance of SH groups occurred even after separation of GSSG from the reaction medium. This shows that reaction 3 proceeds in the course of the formation of the inter L-L disulfide bond of these proteins. The time course of reaction 3 for Fu and Nag proteins was measured (Fig. 3A). Before these measurements, the reduced proteins were reacted with GSSG for 15 min, at which time 55% of the total SH groups for Nag protein and 52% for Fu protein could be titrated with DTNB. The SH contents titrated after incubation for 15 hr were 32% for Nag protein
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
the inter L-L disulfide bond to a considerable extent, while no appreciable disulfide bond formation was observed for all the type K proteins. In addition, the yield of the dimer with a disulfide bond differed from specimen to specimen with the type X proteins. The elementary reactions between the SH groups of a partially reduced Bence Jones protein and GSSG may be represented by Eqs. 1-3.
96
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI 100
30 60 90 TIME (min)
Fig. 3. A: Time course of the dimer formation of Fu protein (O) and Nag protein ( • ) , corresponding to reaction (3) in the text. After the partially reduced proteins had been reacted with GSSG for 15 min, GSSG was removed from the reaction mixture by gel nitration, and then the progress of the reaction was followed with time by measuring the SH content. The solutions contained 1.5 mM EDTA. pH 8.0, 25°. B: First-order plots of the data shown in A.
and 28% for Fu protein. It is clear that the rate of reaction 3 of Fu protein is more rapid than that of Nag protein. The data given in Fig. 3A were analyzed in terms of first-order kinetics (Fig. 3B). The rate constants obtained from the slopes of the plots were 3.1 X 10"s min" 1 for Fu protein and 1.2 xlO" 2 min"1 for Nag protein. 100
•4100
30
60 90 120 TIME (min)
150
180
Fig. 4. Disappearance of the SH content of partially reduced Ta protein (type r) in the presence of 1.5 mM GSSG and 1.5 mM EDTA at pH 8.0 and 25°. Protein concentration, 0.1-0.3%. A , in the absence of GSSG; O, remaining SH content determined immediately after the removal of GSSG ; 3 , remaining SH content determined for solutions incubated for 15 hr after the removal of GSSG; • , yield of the dimer.
Figure 4 shows the time course of the reaction of partially reduced Ta protein with GSSG at pH 8.0. Although the titratable SH groups disappeared with time in a way similar to that for Fu or Nag protein (Fig. 2), essentially no inter L-L disulfide bond formation occurred within 180 min. Furthermore, no decrease in the SH content was observed when the solution was incubated for 15 hr at pH 8.0 after excluding GSSG from the reaction medium. These results are quite different from those for type X proteins shown in Fig. 2. The disappearance of the SH content with time shown in Fig. 4 followed first-order kinetics and the apparent rate constant was found to be 1.5X10"1 min"1. Effect of pH—The effect of pH on the formation of the inter L-L disulfide bond of Nag (/!) and Ta (*) proteins was studied between pH 2 and 10 in the presence and absence of GSSG. As shown in Fig. 5A, the pH profile for Nag protein in the presence of of GSSG shows a minimum near pH 6. On raising the pH from 6 to 8, a sharp increase in the formation of the disulfide-bonded dimer was observed, corresponding to deprotonation of the SH group of Nag protein. A further increase in the pH, however, decreased dimer formation. In the absence of GSSG, no inter L-L disulfide bond was formed in the pH re/ . Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
90 120 TIME (min)
97
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
A
50 40 -30 "ao
> \
.O
•
v
1 •
Z 2°
\
S10
X
Effect of Ionic Strength—As shown in Fig. 7, the extent of the disulfide bond formation of Nag protein in the presence of GSSG
/
Q
i °
1
•
Cr ' " '
8
9
10
10
Fig. 5. Effect of pH on the reformation of the inter L-L disulfide bond from partially reduced Nag protein (A) and Ta protein (B). The yields of the dimers at 180 min were determined by SDS polyacrylamide gel electrophoresis. Ionic strength, 0.075, 25°. O, in the absence of GSSG and presence of 1.5 mM EDTA; • , in the presence of 1 mM GSSG and 1.5 mM EDTA. Vol. 79, No. 1, 1976
served for Nag protein, the formation of inter L-L disulfide bond for Ta protein occurred even in the absence of GSSG. The pH profiles in the acidic pH region for Nag and Ta proteins were very similar. Effect of Urea—The effect of urea on the inter L-L disulfide bond formation of Nag and Ta proteins is shown in Fig. 6. Addition of urea caused a decrease in the disulfide bond formation of Nag protein in the presence of GSSG. With increasing urea concentration, the extent of disulfide bond formation decreased gradually and became constant above 2 M urea. On the other hand, in the absence of GSSG, no formation of the dimer was detected between 0 and 2.5 M urea. No disulfide-bonded dimer was formed for Ta protein up to 2.5 M urea in the presence or absence of GSSG. Since Bence Jones proteins begin to denature from about 2 M urea at pH 8 (21), the intrachain disulfide bonds may participate in the thiol-disulfide interchange reaction leading to the formation of higher molecular weight species than the dimer above 2 M urea. Therefore, no experiments were carried out above 3 M urea.
1 2 CONCENTRATION OF UREA ( M )
3
Fig. 6. Reformation of the inter L-L disulfide bond of Nag protein (squares) and Ta protein (circles) at various urea concentrations, pH 8.0 and 25°. The yields of the dimers were determined at 180 min by SDS polyacrylamide gel electrophoresis. Closed symbols, in the presence of 1 mM GSSG and 1.5 mM EDTA; open symbols, in the absence of GSSG and presence of 1.5 mM EDTA.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
gion between 6 and 10. On lowering the pH from 5.5 to 3.5, an increase in the amount of the dimer was again observed. More than 25% of the protein formed disulfide-bonded dimer after 180 min near pH 4. The acceleration of dimer formation in the acidic pH region was seen even in the absence of GSSG. This is contrast to dimer formation in the alkaline pH region. The plots of dimer formation in the presence and absence of GSSG below pH 5.5 fall on the same profile. In the case of Ta protein, no inter L-L disulfide bond was formed in the pH region between 6 and 10 in the presence or absence of GSSG. This is markedly different from the case for Nag protein. Lowering of the pH from 6 to 2 increased the extent of dimer formation, with a maximum at pH 3. As ob-
98
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
Effect of Cu2+ Ions—The effect of Cu*+ ions on the inter L-L disulfide bond formation of Nag and Ta proteins was examined (Fig. 8). In this case, EDTA was omitted from the medium. Although the rates of oxidation were different, both Nag and Ta proteins formed the inter L-L disulfide bond in the presence of 10 fiM Cu!+ ion. Interchain Disulfide Bond Formation in Immunoglobulin G—The formation of interchain disulfide bonds between polypeptide chains in partially reduced normal immunoglobulin G was studied in the presence of 100 80 60
0.2 0.3 0.4 IONIC STRENGTH
40
0.5
20
Fig. 7. Effect of ionic strength on the reformation of the inter L-L disulfide bond from partially reduced Nag protein (circles) and Ta protein (squares). pH 8.0, 25°. Ionic strength was adjusted with KC1. The yields of the dimers at 180 min were determined by SDS polyacrylamide gel electrophoresis. Closed symbols, in the presence of 1 mM GSSG and 1.5 mM EDTA; open symbols, in the absence of GSSG and presence of 1.5 mM EDTA.
0 20
40
60 180
100
10
20
30 40 TIME (min)
50
60
Fig. 8. Effect of Cu'+ ions on the reformation of the inter L-L disulfide bond from partially reduced Nag protein (triangles) and Ta protein (circles). pH 8.0, room temperature. The yields of the dimers were determined by titration of free SH groups. Closed symbols, in the presence of 10 fiM CuSO4; open symbols, in the absence of CuSO4 and presence or absence of 1.5 mM EDTA.
20
40 TIME (min)
60 180
Fig. 9. Kinetics of the reoxidation of partially reduced normal immunoglobulin G in the absence of glutathione (A), in the presence of GSSG and GSH (GSSG/GSH = l/10, total concentration, 1 mM) (B), and in the presence of 1 mM GSSG (C). All the solutions contained 1.5 mM EDTA. pH 8.0, 25°. • , H; A, L; D, H,; O, HL; O, H ,L; • , H,L,. The amount of each species was determined by SDS polyacrylamide gel electrophoresis. / . Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
was increased by an increase in the ionic strength of the medium. The yield of the disulfide-bonded dimer varied from 44 to 72% with variation of the ionic strength from 0.05 to 0.58. On the other hand, formation of the dimer was not observed for Nag protein in the absence of GSSG and for Ta protein in the presence or absence of GSSG at any ionic strength.
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
Figure 10 shows the kinetics of formation of the interchain disulfide bonds from partially reduced myeloma protein (Jo) (IgGl, *). As has been observed in the reaction of normal immunoglobulin G, no appreciable interchain disulfide bonds were formed in the absence of GSSG (not shown). However, in the presence of GSSG, the free H and L chains disappeared within 10 min and the extent of formation of H2Li was 93% at 60 min. It is interesting to note that the L chain (type «) of myeloma protein (Jo) can easily form a di100
10
20
30
40
50
60
180
TIME (min)
Fig. 10. Kinetics of reoxidation of partially reduced myeloma protein (Jo) in the presence of 1 mM GSSG and 1.5 mM EDTA. pH 8.0, 25°. Symbols used are the same as in Fig. 9. Vol. 79, No. 1, 1976
sulfide bond with its own H chain, in spite of the fact that no disulfide bond is formed between type K L chains. The kinetic pattern of disulfide bond formation of Jo protein was very similar to that for normal immunoglobulin G, although the rate of formation of HJLI for the former was more rapid than that for the latter. DISCUSSION It is well-known that the oxidation of sulfhydryl groups does not proceed easily in the absence of catalysts (22). We have studied the formation of disulfide bonds between polypeptide chains from partially reduced Bence Jones proteins and immunoglobulin G in the presence of glutathione. For several proteins, including immunoglobulins, it is known that the rate of formation of intrachain or interchain disulfide bonds is accelerated by the addition of GSSG and GSH through the thioldisulfide interchange reaction (23). In order to exclude effects due to contaminating metal ions, all the present experiments were carried out in the presence of EDTA. As can be seen in Figs. 2 and 9A, essentially no interchain disulfide bonds of Bence Jones proteins and immunoglobulin G are formed in the presence of EDTA. This finding is incompatible with the result obtained by Petersen and Dorrington (11) that EDTA accelerates the formation of interchain disulfide bonds of a myeloma protein. Saxena and Wetlaufer (10) reported that the regeneration of the intrachain disulfide bonds of reduced lysozyme was markedly accelerated in the presence of both GSH and GSSG at a molar ratio of 10/1. In contrast to their results, the formation of the interchain disulfide bonds of partially reduced Bence Jones proteins and immunoglobulin G proceeded in the presence of GSSG alone. As shown in Figs. IB, 9B, and 9C, the interchain disulfide bonds of a type 1 Bence Jones protein (Nag) and normal immunoglobulin G form more rapidly in the presence of GSSG alone than in the presence of both GSSG and GSH. This may arise from the fact that the reformation of the interchain disulfide bonds in Bence
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
glutathione and 1.5 mM EDTA. While the four-chain structure is maintained in the partially reduced IgG molecule, no disulfide bond formation was observed in 60 min in the absence of GSSG (Fig. 9A). In the presence of glutathione, the formation of the interchain disulfide bonds was accelerated, as shown in Fig. 9, B and C. However, it is apparent from these figures that the formation of disulfide bonds proceeds more rapidly in the presence of GSSG alone (1 mM) than in the presence of both GSSG and GSH (GSSG+GSH = 1 mM and GSSG/GSH = l/10). In the presence of GSSG, the free H and L chains disappeared in 20 min and the extent of formation of H*Lj was 81% at 60 min. It can be seen from Fig. 9, B and C that HL and HjL are the principal intermediates leading to the formation of HiLi, and the formation of HL precedes that of HiL.
99
100
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
fe
l)-SH fa(GSSG) (lj-SSG £)-SSG
(2)-SSG
(C)
(E)
Scheme 1
In this scheme, we assume that the dimer of a Bence Jones protein consists of two distinct monomers which are designated as monomer 1 and monomer 2, and that only intermediate B, in which the SH group on monomer 1 is blocked with GSSG, can form a dimer with an inter-monomer disulfide bond (D) through thiol-disulfide interchange reaction, while intermediate C, in which the SH group on monomer 2 is blocked with GSSG, can not. E represents a product in which both SH groups are blocked with GSSG. ku klt kt, and kt are the rate constants. On the basis of Scheme 1, curve 1 in Fig. 2 represents the remaining SH content of species A, B, and C. Curve 2 in Fig. 2 was obtained for solution incubated for a prolonged
time after excluding glutathione, in which reactions A—»B, A—»C, and C—>E had not occurred and all of species B was converted to D. Furthermore, A can not be converted to D, since EDTA is present in this system. Therefore, curve 2 represents the SH content of species A and C. On the basis of the molecular weight of the monomer, the SH contents represented by curves 1 and 2 may be expressed by ((A)+(B)/2+(C)/2) and ((A)+ (C)/2), respectively, since A contains two SH groups and B and C contain one SH group per dimer. The difference between curve 1 and curve 2 may thus represent the yield of species B. Curve 3 in Fig. 2 represents the yield of D determined experimentally by SDS gel electrophoresis. The total protein concentration ((A)o) maybe expressed by Eq. 4, (4)
where the concentration of each species is expressed as monomer units of the protein. The concentrations of A and C cannot be determined independently. Beyond 60 min, however, no SH groups could be titrated in the system without GSSG (curve 2 in Fig 2). This, means that the interchange reaction mixture does not contain species A and C, since noloss of SH content is expected from these, species when GSSG has been excluded. A t />60 min, therefore, Eq. 4 reduces to Eq. 5and the yield of E can be estimated.
The time dependence of the yield of each species thus estimated is illustrated in Fig. 11. If the following scheme is employed toanalyze the interchange reaction between a. partially reduced type A protein and GSSG,
SH (A)
Scheme 2 / . Biochem^
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
Jones proteins and immunoglobulins is not required for the correction of mismatched disulfide bonds, since each polypeptide chain interacts noncovalently and the pair of SH groups are located at fixed positions in the threedimensional structure of the proteins. In addition to this, the correct intrachain disulfide bonds of lysozyme, once formed, are not attacked by GSH, because the disulfide bonds are buried in the interior of the protein molecule. On the other hand, interchain disulfide bonds of Bence Jones proteins and immunoglobulins are easily reduced by GSH, since they are located on the surface of the protein molecule. Therefore, only GSSG may be effective for the formation of the interchain disulfide bonds. Formation of Inter L-L Disulfide Bond— We analyzed the kinetics of formation of the dimer with a disulfide bond from partially reduced type X protein, Fu or Nag, (A), according to the following scheme.
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
101
(6),
(A)=(A)oe-
B—>D is a consecutive reaction, the rate constants {ku kt, and ks) and the maximum value of (B) ((B)max) are related as follows. (B) n
(A)o 30
60
90
120
150
180
TIME (min)
Fig. 11. Formation of each species in Scheme 1 in the text from partially reduced Fu protein (A) and Nag protein (B) in the presence of 1 DIM GSSG. O, (A) + (C)/2; O, (B); # , ( 0 ) ; D, (E). Solid lines {A to E) represent theoretical curves; see the text for details.
it can be similarly deduced that curves (A+C), B, D, and E in Fig. 11 represent the yields of A, I, D, and E, respectively. In this scheme, we do not discriminate the two monomers in the dimer and assume that intermediate I, in which either of the two SH groups is blocked with GSSG, can form both the dimer with a disulfide bond (D) and a product in which both SH groups are blocked with GSSG (E). The reaction Scheme 2 predicts that the formation of D must proceed in parallel with the formation of D, depending on the rate constants for the reactions I—>D and I—»E. However, the results obtained (Fig. 11) show that the yield of E is constant after 60 min, when the formation of D is still proceeding. As shown in Fig. 1A, an increase in GSSG concentration Vol. 79, No. 1, 1976
(9).
Ui+*J\ A
In Eq. 9, the values of ks and (B)max are known experimentally (Figs. 3 and 11). The ratio of the rate constants, kjkt, is given by Eq. 10, because the reaction is of a competitive type, as shown in Scheme 1, (10) *» At t—>oo, the values of (B) and (C) should become zero and Eq. 10 reduces to Eq. 11, (D) (E)
(11)
The ratio (D)/(E) can be estimated from the results of SDS gel electrophoresis. Using Eqs. 9 and 11, the values of ki and kt can be determined. From Eqs. 4 and 10, the following equations are obtained. (12)
(D)=(A)o-{(AH(B)-
1
(*•
Proteins and Immunoglobulins1 Fumio KISHIDA,2 Takachika AZUMA, and Kozo HAMAGUCHI' Department of Biology, Faculty of Science, Osaka University, Toyonaka, Osaka 560 Received for publication, July 23, 1975
The formation of the interchain disulnde bonds in partially reduced Bence Jones proteins and immunoglobulins was studied in the presence of glutathione. It was found that only oxidized glutathione (GSSG) was effective for the formation of the interchain disulnde bonds in type X Bence Jones proteins and IgG. In type * Bence Jones proteins, on the other hand, no formation of the inter L-L disulnde bond was observed in the presence of GSSG at above pH 6. The kinetic pattern of disulfide bond formation of Bence Jones proteins was well interpreted by assuming that two monomers of a type X protein dimer are discriminated (monomers 1 and 2) and that only an intermediate in which the SH group on monomer 1 is blocked with GSSG can form a disulfide-bonded dimer and the intermediate in which the SH group on monomer 2 is blocked with GSSG can not. Comparison of the kinetic data for the formation of the interchain disulfide bonds of IgG with those for Bence Jones proteins suggested that H chain-GSSG mixed disulfide is a principal intermediate for the formation of the inter H-L disulfide bond.
covalent assembly of H and L chains of a myeloma protein have been studied in our laboratory (1—4). Studies on the formation of the interchain disulfide bonds should also be very important for understanding the formation of the three-dimensional structure of immunoglobulins. Kinetic studies on the biosynthesis of immunoglobulins show that covalent intermediates such as Hi, HL, and HjL are formed during the assembly of H and L chains, whereas covalent Lt is not formed at all within the cell (5, 6). On the other hand, disulfide-bonded dimers of L chains are frequently found in human Bence Jones proteins, especially in type X proteins ( 7 ) . These facts suggest that the characteristics of the inter
It is well-known that immunoglobulin G consists of two heavy (H) and two light (L) chains which interact by both noncovalent interactions and covalent disulfide bonds. The kinetic and equilibrium properties of the non1
This work was supported in part by a grant from the Ministry of Education, Science and Culture, of Japan. 1 Present address: Research Department, Pesticides Division, Sumitomo Chemical Co., Ltd., Takarazuka, Hyogo 665. ' To whom correspondence should be sent. Abbreviations: DTNB, 5,5'-dithiobis(2-nitrobenzoic acid); DTT, dithiothreitol; EDTA, ethylenediaminetetraacetic acid; GSH, reduced glutathione; GSSG, oxidized glutathione; SDS, sodium dodecyl sulfate. Vol. 79, No. 1, 1976
91
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
Formation of Interchain Disulfide Bonds in Bence Jones
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
92
MATERIALS AND METHODS Proteins — Purification and determination of the antigenic type of Bence Jones proteins were done according to the procedure described previously (17, 18). Six type c proteins, Ta, Ham, Ya, Iwa, Tew, and Mat, and seven type X proteins, As, Fu, Ori, Tod, Sh, Ni, and Nag, were used. SDS gel electrophoresis showed that Ta, Ham, As, Fu, Ni, Sh, and Nag are dimers with a disulfide bond but contain a small amount of monomer with no disulfide bond. The other specimens contain both dimers with a disulfide bond and monomers.
It was confirmed that Ham, Sh, Nag, Ni, Tod, Ta, and Fu remain dimers even after reduction of the interchain disulfide bond at the protein concentrations used in the present experiments. The preparation of human myeloma protein Jo (IgGl, K) was described in a previous paper (2). Normal human immunoglobulin G (Fraction II) was obtained from Sigma Chemical Co. Reagents—Sodium dodecyl sulfate (SDS), 5, 5' - dithiobis(2 - nitrobenzoic acid) (DTNB), iodoacetamide, and ethylenediaminetetraacetic acid (EDTA) were purchased from Wako Pure Chemicals Co. Dithiothreitol (DTT) and urea were from Nakarai Pure Chemicals Co. and GSSG and GSH were from Sigma Chemical Co. Urea was recrystallized from 80% ethanol. All other reagents were of reagent grade and were used without further purification. Reduction of Interchain Disulfide Bonds— About 1-5% protein in 0.2 M Tris-HCl buffer containing 2 mM EDTA (pH 8.0) was allowed to react with 20 mM DTT for 40 min at room temperature. Under these conditions, no intrachain disulfide bonds of Bence Jones proteins and immunoglobulins were reduced. We refer to this procedure as partial reduction. The partially reduced protein was separated from residual reagents by passage through a column of Sephadex G-25 equilibrated with 0.01 M Tris-HCl buffer containing 1 mM EDTA at pH 8.0 or with 0.1 M KC1 containing 1 mM EDTA. Interchange Reaction of the SH Groups of Bence Jones Proteins and Immunoglobulins in the Presence of Glutathione—A solution of a partially reduced protein was mixed with the same volume of a solution containing a desired amount of glutathione. Unless otherwise specified, a glutathione was dissolved in 0.2 M Tris-HCl buffer containing 2 mM EDTA at pH 8.0. In order to vary the pH or the ionic strength of the reaction mixture, glutathione was dissolved in an appropriate buffer and the ionic strength was adjusted with KC1. The buffers used were as follows glycine-HCl (pH 2-4), acetate (pH 4-6), Tris-HCl (pH 7-9), and glycine-KOH (above pH 9). The reaction was continued for a given time at 25°. The progress of the reaction was followed by SDS / . Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
L-L disulfide bond are different from those of the inter H-L or inter H-H disulfide bond. The relative lability of interchain disulfide bonds in immunoglobulins has been studied employing various reducing reagents, and several intermediates were found under reducing conditions (8, 9). However, studies on the formation of disulfide bonds from the protein in which interchain disulfide bonds are reduced seem necessary to interpret the results of the covalent assembly of immunoglobulins in vivo. Saxena and Wetlaufer (10) employed a system containing reduced and oxidized glutathione (GSH and GSSG) for the regeneration of disulfide bonds from reduced lysozyme [EC 3. 2.1.17]. Recently, Petersen and Dorrington { 77) applied this system to the reformation of the interchain disulfide bonds in immunoglobulin G. Similar experiments were also performed by Sears et al. (72). However, the mechanism of the formation of the interchain disulfide bonds in immunoglobulins is less well understood. Recent X-ray crystallographic studies (13—16) show that the three-dimensional structure of a type X Bence Jones protein is very similar to that of the Fab' fragment of immunoglobulin. Therefore, comparative studies on the formation of inter L-L and inter L-H disulfide bonds may be interesting in view of the similarity of the conformation. In the present paper, we report the kinetics of disulfide bond formation in Bence Jones proteins and immunoglobulin G in the presence of glutathione and offer a possible mechanism for these reactions.
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
In order to study the effect of urea on the inter L-L disulfide bond formation, a partially reduced Bence Jones protein was incubated in urea solution at a given concentration containing 1 mM EDTA for 8.5 hr at room temperature. Then a urea solution of the same concentration containing GSSG was added and the reaction mixture was left to stand for 3 hr at 25°. The extent of dimer formation was analyzed by SDS gel electrophoresis. Oxidation in the Presence of Cu2+—In this experiment, EDTA was omitted from the system. Degassed and deionized water was used throughout the experiment. Reoxidation of partially reduced protein was allowed to proceed in the presence of 10 ^M CuI+ (CuSO4) at room temperature in a polyethylene beaker which was covered with moistened gauze. An aliquot of the reaction mixture was taken at a given time and the SH content was determined by titration with DTNB. No effect of Cu t+ ions on the results of SH group titration was observed since the titration was carried Vol. 79, No. 1, 1976
out in the presence of a large excess of EDTA (1 mM) over CuI+ ions, as described below. Titration of Free SH Groups with DTNB —To 3 ml of about 0.1-0.3% protein either in 0.01 M Tris-HCl buffer at pH 8.0 or in 0.01 M acetate buffer at pH 6.5 (both containing 1 mM EDTA) was added 0.5 ml of freshly prepared DTNB solution in 1 M Tris-HCl buffer at pH 8.0. The mixture was left to stand for 30 min at room temperature and its absorbance was measured against a reagent blank using a Hitachi automatic recording spectrophotometer, model 323. A value for the molar extinction coefficient of reduced DTNB of 13,600 was used to convert the absorbance at 412 nm to SH content (79). SDS Polyaerylamide Gel Electrophoresis— Electrophoresis in the presence of 0.1% SDS was carried out at pH 7.5 in 7% polyacrylamide gel using a glass tube (0.6x8 cm) according to the method described by Weber and Osborn (20), but 2-mercaptoethanol was omitted. Before electrophoresis, samples were incubated with 1% SDS containing 50% glycerol for 24 hr at room temperature (for Bence Jones protein) or for 1 min at 100° (for immunoglobulins). After the run, the gels were stained overnight with 0.25% Coomassie brilliant blue in 50% methanol-9.2% acetic acid and destained with 5% methanol-7.5% acetic acid. Densitometric scanning of the gels was carried out at 600 nm using a Fuji Riken densitometer, type FD-4. pH Measurements—A Hitachi-Horiba pH meter, model F-5, was used. RESULTS Effect of Glutathione Concentration—Figure 1A shows the effect of GSSG concentration on the formation of the inter L-L disulfide bond of Nag protein (type X) at pH 8.0 as determined by SDS gel electrophoresis. In this figure, the percent of the dimer with a disulfide bond obtained after 180 min is plotted against the concentration of GSSG. The protein concentration was about 4x 10~5 M. In the absence of GSSG, no inter L-L disulfide bond was formed within 180 min when EDTA was present in the reaction medium. The extent
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
gel electrophoresis and by titration of the free SH groups with DTNB. For analysis by SDS gel electrophoresis, the reaction was stopped by the addition of a sufficient amount of iodoacetamide with respect to the free SH groups. The reaction mixture was then dialyzed against water at 5° and analyzed. For the titration of SH groups, the reaction was terminated by lowering the pH to 6.5 with 1M acetic acid followed by gel nitration on a Sephadex G-25 or G-50 column equilibrated with 0.01 M acetate buffer containing 1 mM EDTA at pH 6.5, at which the formation of interchain disulfide bond is minimized, as will be described in the "RESULTS." An aliquot of the eluted protein solution was withdrawn and the number of free SH groups was immediately determined with DTNB. Another aliquot of the solution was then incubated at 25° after increasing the pH to 8.0 using 1 M Tris-HCl buffer (final concentration of Tris, 0.1 M). After a given reaction time, the solution was acidified to pH 6.5 with 1M acetic acid and passed through a Sephadex G-25 column equilibrated with 0.01 M acetate buffer containing 1 mM EDTA at pH 6.5, then the number of free SH groups was determined by titration.
93
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
94 100
100
0.2
0.4
0.6
0.8
1.0
o
0.2
CONCENTRATION OF GSSG (mM)
0.4
0.6
0.8
1.0
X =GSSG/(GSSG+GSH)
Fig. 1. Reformation of the inter L-L disulfide bond from partially reduced Nag protein (type X) as a function of GSSG concentration (A) and mole fraction of GSSG (x) in the mixture of GSSG and GSH (total concentration, 1 mM) (B). All the solutions contained 1.5 mM EDTA. The yield of the dimer at 180 min was determined by SDS polyacrylamide gel electrophoresis. Protein concentration, 0.1%, pH 8.0, 25°.
of disulfide bond formation increased up to 0.2 mM GSSG and became constant above 0.4 mM. The maximum yield of the dimer was about 50% of the total protein. Similar experiments were done with various molar ratios of GSSG and GSH. Figure IB shows a plot of the extent of inter L-L disulfide bond formation at 180 min against the mole fraction of GSSG (x=GSSG/(GSSG+ GSH)). The total concentration of glutathione (GSSG+GSH) in the reaction mixture was fixed at 1 mM. At values of x below 0.1, no appreciable formation of inter L-L disulfide bond was observed. At values of x above 0.1, the
yield of the dimer with a disulfide bond increased and became constant above *=0.5. The maximal yield of the dimer, however, was about 50%, which was essentially the same as that in the presence of GSSG alone (Fig. 1A). This indicates that only GSSG is effective in the formation of inter L-L disulfide bond, and further experiments were done in the presence of 1 mM GSSG. Table I summarizes the percent of the dimer with a disulfide bond formed after 180 min for six type K and seven type X Bence Jones proteins. It is remarkable that all the type X proteins except Tod protein can form
TABLE I. The yields of the dimers with an interchain disulfide bond determined at 180 min from partially reduced Bence Jones proteins in the presence of 1 mM GSSG and 1.5 mM EDTA. The yields were determined by SDS polyacrylamide gel electrophoresis and are expressed as per cent of the total protein. pH 8.0, 25°. Protein
Type
Yield of dimer with S-S bond (%)
Protein
Ta Ham
c
3
As
K
3
Fu
Ya
£
0
Sh
Iwa Tew
C
Nag
c
0 0
Mat (D)» Mat (M)»
c
0
Ori Ni
c
0
Tod
Type
Yield of dimer with S-S bond (%)
X X X X X X X
96 83 64 54 39 32 4
* The properties of the dimer (D) and monomer (M) of Mat protein are described in Ref. 18. J. Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
0
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
95
O-SH
SSG
+ GSSG
+ GSSG O-SSG
+ GSH
(1)
• S j j j + GSH ( 2 ) O-s
+ GSH
(3)
Reactions 1 and 2 correspond to the formation of the mixed disulfide and reaction 3 represents the formation of the interchain disulfide bond. Since all the reactions involve the disappearance of SH groups, the results of the titration of free SH groups at a given time after removal of GSSG and GSH by gel filtration offer a measure of the progress of the thiol-disulfide interchange reactions. Moreover, if GSSG is separated from the reaction mixture and the reaction mixture is then incubated, reactions 1 and 2 will not occur and only reaction 3 may proceed. Figure 2 shows the percent of the remaining free SH groups of partially reduced Nag and Fu proteins (type Z) titrated immediately after the removal of GSSG (curve 1) and after the separation of GSSG followed by incubation of the solution for about 15 hr (curve 2). The time course of the formation of the dimer with a disulfide bond as determined by SDS gel electrophoresis is also shown in this figure {curve 3). It is apparent that dimer formation occurs with an accompanying decrease in the titratable SH groups. However, the extent of disappearance of the SH groups did not completely parallel the extent of formation of the inter L-L disulfide bond. This is more marked in the case of Nag protein. Both the rate and the final yield of dimer formation for Nag protein were different from those Vol. 79, No. 1, 1976
90 120 150 180 TIME (min) Fig. 2. Disappearance of SH content and reformation of the inter L-L disulfide bond from partially reduced Fu protein (type X) (A) and Nag protein (type X) (B) in the presence of 1 mM GSSG and 1.5 mM EDTA. Protein concentration, 0.1-0.3%, pH 8.0, 25°. Curve 1, remaining SH content determined immediately after the removal of GSSG; curve 2 remaining SH content in solutions incubated for 15 hr after the removal of GSSG; curve 3, yield of the dimer as determined by SDS polyacrylamide gel electrophoresis. 30
60
for Fu protein. The disappearance of SH groups occurred even after separation of GSSG from the reaction medium. This shows that reaction 3 proceeds in the course of the formation of the inter L-L disulfide bond of these proteins. The time course of reaction 3 for Fu and Nag proteins was measured (Fig. 3A). Before these measurements, the reduced proteins were reacted with GSSG for 15 min, at which time 55% of the total SH groups for Nag protein and 52% for Fu protein could be titrated with DTNB. The SH contents titrated after incubation for 15 hr were 32% for Nag protein
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
the inter L-L disulfide bond to a considerable extent, while no appreciable disulfide bond formation was observed for all the type K proteins. In addition, the yield of the dimer with a disulfide bond differed from specimen to specimen with the type X proteins. The elementary reactions between the SH groups of a partially reduced Bence Jones protein and GSSG may be represented by Eqs. 1-3.
96
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI 100
30 60 90 TIME (min)
Fig. 3. A: Time course of the dimer formation of Fu protein (O) and Nag protein ( • ) , corresponding to reaction (3) in the text. After the partially reduced proteins had been reacted with GSSG for 15 min, GSSG was removed from the reaction mixture by gel nitration, and then the progress of the reaction was followed with time by measuring the SH content. The solutions contained 1.5 mM EDTA. pH 8.0, 25°. B: First-order plots of the data shown in A.
and 28% for Fu protein. It is clear that the rate of reaction 3 of Fu protein is more rapid than that of Nag protein. The data given in Fig. 3A were analyzed in terms of first-order kinetics (Fig. 3B). The rate constants obtained from the slopes of the plots were 3.1 X 10"s min" 1 for Fu protein and 1.2 xlO" 2 min"1 for Nag protein. 100
•4100
30
60 90 120 TIME (min)
150
180
Fig. 4. Disappearance of the SH content of partially reduced Ta protein (type r) in the presence of 1.5 mM GSSG and 1.5 mM EDTA at pH 8.0 and 25°. Protein concentration, 0.1-0.3%. A , in the absence of GSSG; O, remaining SH content determined immediately after the removal of GSSG ; 3 , remaining SH content determined for solutions incubated for 15 hr after the removal of GSSG; • , yield of the dimer.
Figure 4 shows the time course of the reaction of partially reduced Ta protein with GSSG at pH 8.0. Although the titratable SH groups disappeared with time in a way similar to that for Fu or Nag protein (Fig. 2), essentially no inter L-L disulfide bond formation occurred within 180 min. Furthermore, no decrease in the SH content was observed when the solution was incubated for 15 hr at pH 8.0 after excluding GSSG from the reaction medium. These results are quite different from those for type X proteins shown in Fig. 2. The disappearance of the SH content with time shown in Fig. 4 followed first-order kinetics and the apparent rate constant was found to be 1.5X10"1 min"1. Effect of pH—The effect of pH on the formation of the inter L-L disulfide bond of Nag (/!) and Ta (*) proteins was studied between pH 2 and 10 in the presence and absence of GSSG. As shown in Fig. 5A, the pH profile for Nag protein in the presence of of GSSG shows a minimum near pH 6. On raising the pH from 6 to 8, a sharp increase in the formation of the disulfide-bonded dimer was observed, corresponding to deprotonation of the SH group of Nag protein. A further increase in the pH, however, decreased dimer formation. In the absence of GSSG, no inter L-L disulfide bond was formed in the pH re/ . Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
90 120 TIME (min)
97
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
A
50 40 -30 "ao
> \
.O
•
v
1 •
Z 2°
\
S10
X
Effect of Ionic Strength—As shown in Fig. 7, the extent of the disulfide bond formation of Nag protein in the presence of GSSG
/
Q
i °
1
•
Cr ' " '
8
9
10
10
Fig. 5. Effect of pH on the reformation of the inter L-L disulfide bond from partially reduced Nag protein (A) and Ta protein (B). The yields of the dimers at 180 min were determined by SDS polyacrylamide gel electrophoresis. Ionic strength, 0.075, 25°. O, in the absence of GSSG and presence of 1.5 mM EDTA; • , in the presence of 1 mM GSSG and 1.5 mM EDTA. Vol. 79, No. 1, 1976
served for Nag protein, the formation of inter L-L disulfide bond for Ta protein occurred even in the absence of GSSG. The pH profiles in the acidic pH region for Nag and Ta proteins were very similar. Effect of Urea—The effect of urea on the inter L-L disulfide bond formation of Nag and Ta proteins is shown in Fig. 6. Addition of urea caused a decrease in the disulfide bond formation of Nag protein in the presence of GSSG. With increasing urea concentration, the extent of disulfide bond formation decreased gradually and became constant above 2 M urea. On the other hand, in the absence of GSSG, no formation of the dimer was detected between 0 and 2.5 M urea. No disulfide-bonded dimer was formed for Ta protein up to 2.5 M urea in the presence or absence of GSSG. Since Bence Jones proteins begin to denature from about 2 M urea at pH 8 (21), the intrachain disulfide bonds may participate in the thiol-disulfide interchange reaction leading to the formation of higher molecular weight species than the dimer above 2 M urea. Therefore, no experiments were carried out above 3 M urea.
1 2 CONCENTRATION OF UREA ( M )
3
Fig. 6. Reformation of the inter L-L disulfide bond of Nag protein (squares) and Ta protein (circles) at various urea concentrations, pH 8.0 and 25°. The yields of the dimers were determined at 180 min by SDS polyacrylamide gel electrophoresis. Closed symbols, in the presence of 1 mM GSSG and 1.5 mM EDTA; open symbols, in the absence of GSSG and presence of 1.5 mM EDTA.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
gion between 6 and 10. On lowering the pH from 5.5 to 3.5, an increase in the amount of the dimer was again observed. More than 25% of the protein formed disulfide-bonded dimer after 180 min near pH 4. The acceleration of dimer formation in the acidic pH region was seen even in the absence of GSSG. This is contrast to dimer formation in the alkaline pH region. The plots of dimer formation in the presence and absence of GSSG below pH 5.5 fall on the same profile. In the case of Ta protein, no inter L-L disulfide bond was formed in the pH region between 6 and 10 in the presence or absence of GSSG. This is markedly different from the case for Nag protein. Lowering of the pH from 6 to 2 increased the extent of dimer formation, with a maximum at pH 3. As ob-
98
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
Effect of Cu2+ Ions—The effect of Cu*+ ions on the inter L-L disulfide bond formation of Nag and Ta proteins was examined (Fig. 8). In this case, EDTA was omitted from the medium. Although the rates of oxidation were different, both Nag and Ta proteins formed the inter L-L disulfide bond in the presence of 10 fiM Cu!+ ion. Interchain Disulfide Bond Formation in Immunoglobulin G—The formation of interchain disulfide bonds between polypeptide chains in partially reduced normal immunoglobulin G was studied in the presence of 100 80 60
0.2 0.3 0.4 IONIC STRENGTH
40
0.5
20
Fig. 7. Effect of ionic strength on the reformation of the inter L-L disulfide bond from partially reduced Nag protein (circles) and Ta protein (squares). pH 8.0, 25°. Ionic strength was adjusted with KC1. The yields of the dimers at 180 min were determined by SDS polyacrylamide gel electrophoresis. Closed symbols, in the presence of 1 mM GSSG and 1.5 mM EDTA; open symbols, in the absence of GSSG and presence of 1.5 mM EDTA.
0 20
40
60 180
100
10
20
30 40 TIME (min)
50
60
Fig. 8. Effect of Cu'+ ions on the reformation of the inter L-L disulfide bond from partially reduced Nag protein (triangles) and Ta protein (circles). pH 8.0, room temperature. The yields of the dimers were determined by titration of free SH groups. Closed symbols, in the presence of 10 fiM CuSO4; open symbols, in the absence of CuSO4 and presence or absence of 1.5 mM EDTA.
20
40 TIME (min)
60 180
Fig. 9. Kinetics of the reoxidation of partially reduced normal immunoglobulin G in the absence of glutathione (A), in the presence of GSSG and GSH (GSSG/GSH = l/10, total concentration, 1 mM) (B), and in the presence of 1 mM GSSG (C). All the solutions contained 1.5 mM EDTA. pH 8.0, 25°. • , H; A, L; D, H,; O, HL; O, H ,L; • , H,L,. The amount of each species was determined by SDS polyacrylamide gel electrophoresis. / . Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
was increased by an increase in the ionic strength of the medium. The yield of the disulfide-bonded dimer varied from 44 to 72% with variation of the ionic strength from 0.05 to 0.58. On the other hand, formation of the dimer was not observed for Nag protein in the absence of GSSG and for Ta protein in the presence or absence of GSSG at any ionic strength.
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
Figure 10 shows the kinetics of formation of the interchain disulfide bonds from partially reduced myeloma protein (Jo) (IgGl, *). As has been observed in the reaction of normal immunoglobulin G, no appreciable interchain disulfide bonds were formed in the absence of GSSG (not shown). However, in the presence of GSSG, the free H and L chains disappeared within 10 min and the extent of formation of H2Li was 93% at 60 min. It is interesting to note that the L chain (type «) of myeloma protein (Jo) can easily form a di100
10
20
30
40
50
60
180
TIME (min)
Fig. 10. Kinetics of reoxidation of partially reduced myeloma protein (Jo) in the presence of 1 mM GSSG and 1.5 mM EDTA. pH 8.0, 25°. Symbols used are the same as in Fig. 9. Vol. 79, No. 1, 1976
sulfide bond with its own H chain, in spite of the fact that no disulfide bond is formed between type K L chains. The kinetic pattern of disulfide bond formation of Jo protein was very similar to that for normal immunoglobulin G, although the rate of formation of HJLI for the former was more rapid than that for the latter. DISCUSSION It is well-known that the oxidation of sulfhydryl groups does not proceed easily in the absence of catalysts (22). We have studied the formation of disulfide bonds between polypeptide chains from partially reduced Bence Jones proteins and immunoglobulin G in the presence of glutathione. For several proteins, including immunoglobulins, it is known that the rate of formation of intrachain or interchain disulfide bonds is accelerated by the addition of GSSG and GSH through the thioldisulfide interchange reaction (23). In order to exclude effects due to contaminating metal ions, all the present experiments were carried out in the presence of EDTA. As can be seen in Figs. 2 and 9A, essentially no interchain disulfide bonds of Bence Jones proteins and immunoglobulin G are formed in the presence of EDTA. This finding is incompatible with the result obtained by Petersen and Dorrington (11) that EDTA accelerates the formation of interchain disulfide bonds of a myeloma protein. Saxena and Wetlaufer (10) reported that the regeneration of the intrachain disulfide bonds of reduced lysozyme was markedly accelerated in the presence of both GSH and GSSG at a molar ratio of 10/1. In contrast to their results, the formation of the interchain disulfide bonds of partially reduced Bence Jones proteins and immunoglobulin G proceeded in the presence of GSSG alone. As shown in Figs. IB, 9B, and 9C, the interchain disulfide bonds of a type 1 Bence Jones protein (Nag) and normal immunoglobulin G form more rapidly in the presence of GSSG alone than in the presence of both GSSG and GSH. This may arise from the fact that the reformation of the interchain disulfide bonds in Bence
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
glutathione and 1.5 mM EDTA. While the four-chain structure is maintained in the partially reduced IgG molecule, no disulfide bond formation was observed in 60 min in the absence of GSSG (Fig. 9A). In the presence of glutathione, the formation of the interchain disulfide bonds was accelerated, as shown in Fig. 9, B and C. However, it is apparent from these figures that the formation of disulfide bonds proceeds more rapidly in the presence of GSSG alone (1 mM) than in the presence of both GSSG and GSH (GSSG+GSH = 1 mM and GSSG/GSH = l/10). In the presence of GSSG, the free H and L chains disappeared in 20 min and the extent of formation of H*Lj was 81% at 60 min. It can be seen from Fig. 9, B and C that HL and HjL are the principal intermediates leading to the formation of HiLi, and the formation of HL precedes that of HiL.
99
100
F. KISHIDA, T. AZUMA, and K. HAMAGUCHI
fe
l)-SH fa(GSSG) (lj-SSG £)-SSG
(2)-SSG
(C)
(E)
Scheme 1
In this scheme, we assume that the dimer of a Bence Jones protein consists of two distinct monomers which are designated as monomer 1 and monomer 2, and that only intermediate B, in which the SH group on monomer 1 is blocked with GSSG, can form a dimer with an inter-monomer disulfide bond (D) through thiol-disulfide interchange reaction, while intermediate C, in which the SH group on monomer 2 is blocked with GSSG, can not. E represents a product in which both SH groups are blocked with GSSG. ku klt kt, and kt are the rate constants. On the basis of Scheme 1, curve 1 in Fig. 2 represents the remaining SH content of species A, B, and C. Curve 2 in Fig. 2 was obtained for solution incubated for a prolonged
time after excluding glutathione, in which reactions A—»B, A—»C, and C—>E had not occurred and all of species B was converted to D. Furthermore, A can not be converted to D, since EDTA is present in this system. Therefore, curve 2 represents the SH content of species A and C. On the basis of the molecular weight of the monomer, the SH contents represented by curves 1 and 2 may be expressed by ((A)+(B)/2+(C)/2) and ((A)+ (C)/2), respectively, since A contains two SH groups and B and C contain one SH group per dimer. The difference between curve 1 and curve 2 may thus represent the yield of species B. Curve 3 in Fig. 2 represents the yield of D determined experimentally by SDS gel electrophoresis. The total protein concentration ((A)o) maybe expressed by Eq. 4, (4)
where the concentration of each species is expressed as monomer units of the protein. The concentrations of A and C cannot be determined independently. Beyond 60 min, however, no SH groups could be titrated in the system without GSSG (curve 2 in Fig 2). This, means that the interchange reaction mixture does not contain species A and C, since noloss of SH content is expected from these, species when GSSG has been excluded. A t />60 min, therefore, Eq. 4 reduces to Eq. 5and the yield of E can be estimated.
The time dependence of the yield of each species thus estimated is illustrated in Fig. 11. If the following scheme is employed toanalyze the interchange reaction between a. partially reduced type A protein and GSSG,
SH (A)
Scheme 2 / . Biochem^
Downloaded from https://academic.oup.com/jb/article-abstract/79/1/91/772269 by University of Western Sydney user on 10 January 2019
Jones proteins and immunoglobulins is not required for the correction of mismatched disulfide bonds, since each polypeptide chain interacts noncovalently and the pair of SH groups are located at fixed positions in the threedimensional structure of the proteins. In addition to this, the correct intrachain disulfide bonds of lysozyme, once formed, are not attacked by GSH, because the disulfide bonds are buried in the interior of the protein molecule. On the other hand, interchain disulfide bonds of Bence Jones proteins and immunoglobulins are easily reduced by GSH, since they are located on the surface of the protein molecule. Therefore, only GSSG may be effective for the formation of the interchain disulfide bonds. Formation of Inter L-L Disulfide Bond— We analyzed the kinetics of formation of the dimer with a disulfide bond from partially reduced type X protein, Fu or Nag, (A), according to the following scheme.
FORMATION OF S-S BONDS IN BENCE JONES PROTEINS AND IgG
101
(6),
(A)=(A)oe-
B—>D is a consecutive reaction, the rate constants {ku kt, and ks) and the maximum value of (B) ((B)max) are related as follows. (B) n
(A)o 30
60
90
120
150
180
TIME (min)
Fig. 11. Formation of each species in Scheme 1 in the text from partially reduced Fu protein (A) and Nag protein (B) in the presence of 1 DIM GSSG. O, (A) + (C)/2; O, (B); # , ( 0 ) ; D, (E). Solid lines {A to E) represent theoretical curves; see the text for details.
it can be similarly deduced that curves (A+C), B, D, and E in Fig. 11 represent the yields of A, I, D, and E, respectively. In this scheme, we do not discriminate the two monomers in the dimer and assume that intermediate I, in which either of the two SH groups is blocked with GSSG, can form both the dimer with a disulfide bond (D) and a product in which both SH groups are blocked with GSSG (E). The reaction Scheme 2 predicts that the formation of D must proceed in parallel with the formation of D, depending on the rate constants for the reactions I—>D and I—»E. However, the results obtained (Fig. 11) show that the yield of E is constant after 60 min, when the formation of D is still proceeding. As shown in Fig. 1A, an increase in GSSG concentration Vol. 79, No. 1, 1976
(9).
Ui+*J\ A
In Eq. 9, the values of ks and (B)max are known experimentally (Figs. 3 and 11). The ratio of the rate constants, kjkt, is given by Eq. 10, because the reaction is of a competitive type, as shown in Scheme 1, (10) *» At t—>oo, the values of (B) and (C) should become zero and Eq. 10 reduces to Eq. 11, (D) (E)
(11)
The ratio (D)/(E) can be estimated from the results of SDS gel electrophoresis. Using Eqs. 9 and 11, the values of ki and kt can be determined. From Eqs. 4 and 10, the following equations are obtained. (12)
(D)=(A)o-{(AH(B)-
1
(*•
