THE JOURNAL OF BIOLOGICAL CHEMISWY Vol. 247, No. 18, Issue of September 25, pp. 570B-5708, 3872

Printed in U.S.A.

Effect of Chemical Modification of Lysine Residues on the

Conformation of Human Immunoglobulin G*

(Received for publication, MayI12, 1972)

YASUSHI WAKAGAWA, SYLVIA CAPETILLO, AND BRUNO JIRGENSONS

From The University of Texas, M. Biochemistry, Houston, Texas 77026

SUMMARY

D. Anderson Hospital and Tumor Institute at Houston, Department of

Human IgG (immunoglobulin G) was chemically modified by citraconic anhydride and the conformational transitions were tested by measuring CD (circular dichroism), hydrody- namic properties, and immunoprecipitation reactions. It was found that mild conversion of the lysine e-amino groups (at pH 7.7 and room temperature) into the negatively charged citraconyl groups caused expansion and loss of the fl struc- ture of IgG. The positive CD band at 202 nm became nega- tive, while the weak bands in the near-ultraviolet zone changed only slightly; the intrinsic viscosity increased, the sedimentation coefficient decreased, and the immunore- sponse against goat antiserum was lost. The citraconylated IgG was reconverted to native IgG by dialysis against 0.05 M acetate of pH 4.0 for 72 hours. The reversal to the native conformation was achieved also by raising the ionic strength to about 1.1. Carbamylation of 88% of the e-amino groups of IgG did not change the CD spectrum, hydrodynamic properties, or immunoprecipitation reaction. The results indicate that the conformational change on citraconylation was caused by electrostatic repulsion of the introduced car- boxylate ions into the macromolecules. Moreover, accord- ing to these results the e-amino groups of lysine residues are not essential for the immunoprecipitation reaction, whereas the native rigid conformation of the macromolecules is decisive in this reaction.

The importance of reactive groups in the side chains of im- munoglobulins in maintaining the native conformation and im- munochemical reactivity has been and is a broad subject of study. In this report we are concerned with the e-amino groups of lysine residues. These groups have been modified previously by other authors (l-6). None of these papers, however, report a detailed study of the conformational transitions that presumably could accompany these chemical modifications, even if the modifica- tions are accomplished avoiding extreme pIT values, common de- naturing agents, and high temperatures. The purpose of the

* This work was support,ed by Grant, CA-01785 from the Na- tional Cancer Institute, United St,ates Public Health Service, and Grant G-051 from The Robert A. Welch Foundation, Houston, Tex.

present paper is to report such conformational transitions as tested by CD’ and hydrodynamic methods after modifying amino groups with citraconic anhydride and potassium cyanate.

EXPERIMENTAL PROCEDURE

Materials-Normal human IgG was obtained from Pentex, Kankakee, Ill. (Fraction II, lot No. 33). Myeloma IgG was the same sample prepared by Ross and Jirgensons (7). Citraconic anhydride, l-fluoro-2,4-dinitrobenzene, and potassium cyanate were purchased from Eastman-Kodak. The latter compound was recrystallized from water-ethanol mixture as described by Stark (8). Goat antiserum to human IgG was obtained from Hyland, Los Angeles, Calif. Other chemicals used were analyti- cal grade.

Preparation of Citraconyl-IgG-In a typical experiment, ap- proximately 3 mg per ml of IgG solution in 0.1 M sodium phos- phate (pH 7.7) was prepared. The protein was modified by citraconic anhydride at room temperature with continuous stir- ring. A 1000-fold molar excess of citraconic anhydride was dis- solved in absolute ethanol and was added to the IgG solution in small portions. The pH was maintained to 7.7 with 5 N NaOH by using a Radiometer pa-stat equipped with a SBR 2c titri-

graph. The glass electrode was a Radiometer type G222B. The alcohol concentration after addition of the reagent was 2%. After 20 min of reaction the protein solution was dialyzed against 2 liters of sodium phosphate buffer, p1-I 7.7, and p = 0.1, at 4” for 24 hours with one change.

To remove the modifying citraconyl groups, citraconyl-IgG was dialyzed against 2 liters of 0.05 M sodium acetate buffer, pH 4.0, at 4’ for 72 hours with one change. After 72 hours of dialysis against sodium acetate buffer, the protein solution was transferred to 2 liters of sodium phosphate buffer, pH 7.7, and F = 0.1, for 24 hours of dialysis with one change. The purpose of this buffer change was to eliminate the acetate which absorbs strongly the short wave ultraviolet light.

Carbamylation of IgG-IgG was dissolved in 0.1 M sodium phosphate buffer, pH 7.7, to make a protein concentration ap- proximately 5 mg per ml, and recrystallized potassium cyanate was added in small portion to make final concentration of 0.5 M. The pH of the solution was monitored in a Radiometer pH-stat with 1 M acetic acid and maintained to pH 7.9. The reaction was carried out at room temperature with continuous

1 The abbreviations used are: CD, circular dichroism; IgG, immlmoglobulin (:; ORD, optical rotatory dispersion.

5703

by guest on May 14, 2020

http://ww

w.jbc.org/

Dow

nloaded from

5704

stirring. The reaction mixture was dialyzed against 2 liters of sodium phosphate buffer, pH 7.7, and p = 0.1, at 4’ for 24 hours with one change.

Analyses of Proteins-The myeloma IgG concentration was determined by a Beckman DU spectrophotometer at 280 nm with an extinction coefficient value of 14.0 and the molecular weight of 1.78 x 10” (7). Amino acid analyses were performed following the method of Spackman et al. (9). Proteins were hydrolyzed in evacuated tubes in 6 N HCI at 110” for 22 hours. Amino acid analysis was performed by a Spinco amino acid analyzer with an accelerating system and a high sensitivity card. To determine the number of e-amino groups modified by citra- conic anhydride, the dinitrophenylation method was used as described by Lenard and Singer (10). Homocitrulline was determined after carbamylation by the method described by Stark et al. (11).

Determination of S Values-The sedimentation coefficients were determined in a Beckman Spinco model E analytical ultra- centrifuge at 20” and 52,640 or 59,780 rpm with an An-D rotor and a single sector cell. The buffer solution used was sodium phosphate, pH 7.7, p = 0.1. The protein solution was dialyzed against the same buffer for 24 hours before centrifugation. The S values were reduced to standard conditions as described by Schachman (12).

Viscosities-The viscosities were determined in a Cannon- Ubbelohde dilution viscometer at 28 f 0.1”. The intrinsic viscosity, [VI, was obtained from several determinations in the range of protein concentration between 0.3 to 0.9%, extrapo- lated at infinite dilution. Ionic strength was raised by adding anhydrous sodium sulfate.

Circular Dichroism Spectra-A 1 )urrum-Jasco dichrograph model CD-SP was used. Measurements were made with sensi- tivity scale setting 2 x lop5 or 5 X lop5 dichroic optical density difference (AE) per cm on the recorder chart. All recordings were made at 23-25”. Mean residue ellipticity [0] was computed by using a mean residue weight of 109 for all proteins. An av- erage AE was measured from triplicate recordings. The cell having a l.O-cm path length was used for the measurements in near-ultraviolet region (250 to 340 nm); a O.l- or 0.05.cm cell was used in far-ultraviolet region (190 to 250 nm).

Immunod$usion-An immunodiffusion plate was prepared on a polystyrene dish (dimension: 2.5 X 7.6 cm) with lo/o agar in diethylbarbital-sodium acetate buffer, pH 8.2, and p = 0.1. Results of the precipitation reactioll were examined after 24

hours. Ionic strength on the agar plate was adjusted by adding required amounts of sodium sulfate in the buffer solution.

Cellulose Acetate Electrophoresis-Cellulose acetate electro- phoresis was carried out in barbital buffer (pH 8.6, p = O.Oi5) by using a Beckman Microzone electrophoresis cell, model R-101. The running conditions were constant voltage of 250 volts at room temperature for 30 min. Ponceau S was used for staining the proteins.

Citraconyl-IgG

Normal IgG-Preliminary experiments were carried out to find the optimum conditions to modify e-amino groups of normal IgG reversibly. Some typical results of citraconylation of IgG are shown in Table I. When a protein concentration of 0.72% and approximately a 2000.fold molar excess of citraconic an- hydride were used, the CD spectrum of this preparation had the negative trough shifted from 216 to 208 nm. Precipitates were observed during the regeneration step and only 217, of the total protein could be recovered in solution. When the concentration of citraconic anhydride was reduced to a half of that in the previous experiment, the 216-nm negative band was shifted to 212 nm. This preparation also produced pre- cipitates during the regeneration step but the protein recovery was improved to 49u/,. Finally when the reagent concentration was reduced to about 1000.fold molar excess and the protein concentration was decreased to 0.317,, 84.4 lysine residues (925%) were modified. For t,he modified protein, si,,, decreased to 4.89 S and [?I] increased to 0.227 dl per g; however, the negative trough at 216 nm did not shift, and the protein was completely recovered upon regeneratioll. Cellulose acetate electrophoresis of the citraconyl-IgG showed a single zone which migrated to- ward the anode 2 times faster than the unmodified protein. After dialysis against a pII 4.0 buffer, 10 lysine residues remained covered, but the values of s:~,~, [s], and the CD spectra were the same as that of the starting protein. It was concluded that the protein concentration of 0.3% and an approximately lOOO- fold molar excess of citraconic anhydride at pH 7.7 for 20-min reaction are the most suitable conditions for reversible citra- conylation of IgG.

The changes in the Cl> spectra on citraconylation in the non- specific (normal) human IgG were similar to those observed in the individual myeloma IgG (Table I, Fig. 1 and Fig. 3). Be-

TABLE I Mottijication oJ lysine residues o,f normal IgG by citracon,ic anhydride

Normal IgG was dissolved in 0.1 M sodium phosphate buffer, pH 7.7, and the protein was modified by citraconic anhydride at room temperature. The pH was maintained to 7.7 with 5 N NaOH by using a Radiometer pH-stat. After 20 min of reaction the protein solution was dialyzed against sodium phosphate buffer, pH 7.7, and /* = 0.1. To remove the modifying citraconyl groups, citraconyl- IgG was dialyzed against 0.05 M sodium acetate buffer, pH 4.0.

I& Molar concen- ratio of tration reagent

% 0.72

0.72

0.31

x 1930

x 1038

x 1034

No. of lysine

residues modified

90.7

80.7

84.4

4.55

4.67

4.89

After modification

[VI x PI

0.218

nm 208 200 212

- 4400 -5100 -3900

0.227 216 - 3200

CD

%

21.1

49.1

100.0

No. of lysine

residues modified

17.6 6.35

10.9 6.74

10.7 6.74

After regeneration

dl/g wn

216 202

0.122 216 / 202

0.108 1 216 202

- 1400 +3400 - 3200 +3200 - 2200 +3300

by guest on May 14, 2020

http://ww

w.jbc.org/

Dow

nloaded from

3

2

I

0 I) b - x -I

m

-2

-3

-4 I I

- -20

- -30

- -40

0.9

0.8

0.7

0.2

0. I

190 210 230 250 250 270 290 310

WAVELENGTH (nm)

FIG. 1 (left). CD spectra of myeloma IgG at various ionic strengths. For the near ultraviolet (250 to 310 nm), the light path length was 1.0 cm with the protein concentration of 0.09 to 0.16%. For the far ultraviolet (190 to 250 nm), the optical path was 0.1 or 0.05 cm and the protein concentration was 0.015 to 0.020%. The protein was dissolved in sodium phosphate (pH 7.7, p = 0.1) (C--C); the ionic strength was increased to 0.4 with sodium

TAI+LE II

lilodi$cation of lysine residues of myeloma IgG

by citraconic anhydride

Approximately 3 mg per ml of myeloma IgG was dissolved in 0.1 M sodium phosphate buffer, pH 7.7, and a lOOO-fold molar ex- cess of eitraconic anhydride was used for citraconylation. After modification, dialysis and regeneration were followed in the same manner as described in Table I.

No. of bll Precipita- My&ma IgG modified tion in

lysine 4o.w

i I

immuno- residues ,, = 0.1 ,, = 0.4 p = 1.1 diffusion

dl/g

Native. . . 5” (87.9)b 6.54 Modified.. . 88.4 5.56 Regenerat.ed. 9.6 6.68

a Free lysine residue obtained after dinitrophenylation. * Control IgG was hydrolyzed in 6 N HCl at 110” for 22 hours. c The result of immunodiffusion was positive in an agar plate

of p = 1.1.

cause of the multiplicity of molecular species in the normal non- specific IgG, the individual myeloma IgG is preferred as a ho- mogeneous object. Further studies on the myeloma IgG are described in the following section.

Myeloma I&+--Analyses of the native myeloma IgG are shown in Table II. The sir,,, was 6.54 S and the intrinsic viscosity was 0.070 dl per g. It contained 88 lysine residues, and 5 resi- dues were not susceptible to dinitrophenylation in the native state. The effect of ionic strength on the intrinsic viscosity was

, / 230 250 270 290 310 330

WAVELENGTH (nml

sulfate (O-O); the ionic strength was adjusted to 1.1 (A-A).

FIG. 2 (right). Ultraviolet absorption spectra of native and modified myeloma IgGs. Myeloma IgG solutions were dialyzed against 2 liters of sodium phosphate buffer (pH 7.7, p = 0.1) at 4” for 24 hours. The protein concentrations were 0.28 mg per ml. Control IgG (0-O); citraconyl IgG (O-O); carbamylated IgG (A--A).

-“Ia , , , ,I, , , , , , J-40

190 210 230 250 250 270 290 310

WAVELENGTH (ml

FIG. 3. CD spectra of citraconyl IgG at various ionic strengths. The conditions for measurements are the same as described in Fig. 1. Regenerated IgG in sodium phosphate buffer, pH 7.7 and p = 0.1, is shown (U-0).

small. The CD spectra are shown in Fig. 1. Slightly increased molar ellipticity values were observed at very high ionic strength in some parts of the CD spectrum (Fig. 1). However, these changes are small in comparison to those of citraconyl-IgG (Fig. 3).

To modify myeloma IgG, the optimal conditions described above were applied. After modification, ultraviolet absorptions of modified and control myeloma IgG’s were compared. No change of absorbance at 278 nm was observed as shown in Fig. 2. This indicates that no acylation took place on tyrosyl groups. Further, no change at 278 nm was observed by adding hydroxyl-

by guest on May 14, 2020

http://ww

w.jbc.org/

Dow

nloaded from

5706

amine as described by Riordan and Vallee (13). The absorb- ance minimum at 250 nm shifted to 260 nm upon modification, as observed previously with pepsinogen (14). Dansylation followed by thin layer chromatography (15) indicated no free amino terminals present in the modified IgG. The value of a&, decreased to 5.56 S and [q] increased to 0.173 dl per g upon citraconylation. The citraconyl-IgG did not react with goat antiserum. The modified IgG gave a single fast moving elec- trophoretic band on cellulose acetate. Drastic changes in [r] were observed by increasing ionic strength. When the ionic strength was increased to 0.4, [q] became the same as that of the native protein, 0.064 dl per g; and it slightly increased when ionic strength was raised to 1.1. The CD spectra of citraconyl myeloma IgG are shown in Fig. 3. In the near-ultraviolet re- gion, the CD bands were slightly enhanced. In the far-ultra- violet region the prominent positive peak at 202 nm became negative. The effect of ionic strength observed in CD spectra corresponds to the changes in hydrodynamic properties. With increasing ionic strength the CD bands decreased in the near- ultraviolet region and the ellipticity at 202 nm increased. At ionic strength of 1.1, the CD spectra in the far-ultraviolet re- gion became practically identical with those of the native IgG.

Citraconyl myeloma IgG was regenerated by dialyzing against 0.05 M sodium acetate buffer, pH 4.0, for 72 hours, and it was found that 10 lysines remained modified under these regenera- tion conditions. Electrophoresis showed a slightly diffused single band and the mobility was almost the same as that of the control. s$,,~ and [r] values of regenerated myeloma IgG were 6.68 S and 0.070 dl per g, respectively, which were the same as the values of native IgG. Viscosities of the regenerated protein were not affected by changes in the buffer ionic strength, in agree- ment with the results for the native IgG. Immunological pre- cipitation was restored upon regeneration.

Carbamyl-IgG

Carbamylation of e-amino groups of lysines required longer time than citraconylation. With progressing reaction time, more lysine residues were modified, as shown in Table III. The reaction rate was similar to that for carbamylation of ribonu- clease observed by Stark et al. (11). According to these authors,

TABLE III Carbamylation of lysine residztes of myeloma IgG

Myeloma IgG solution (5 mg of protein per ml in 0.1 M sodium phosphate buffer, pH 7.7) was carbamylated in 0.5 M potassium cyanate at room temperature. The pH of the solution was main- tained to 7.9 in a Radiometer pa-stat. The modified protein solu- tions were dialyzed in sodium phosphate buffer, pH 7.7, and p = 0.1, for 24 hours. Acid hydrolysis for amino acid analysis was carried out in evacuated tubes containing 6 N HCl for 22 hours at 110”.

I

hrs w&T

24 45.9 (35.5) 42.3 (52.5) 6.72 48 34.2 (21.4) 53.3 (66.1) 6.64 72 25.0 (10.0) 62.7 (77.7) 6.49

a The numbers of lysine and homocitrulline residues were deter- mined by amino acid analyses. The numbers in parentheses are corrected for 24% decomposition of homocitrulline into lysine during acid hydrolysis.

24% of the homocitrulline decomposed to lysine during acid hydrolysis in 6 N HCI at 110” for 22 hours. Taking this de- composition rate into consideration, 52.5 lysine residues could be modified for 24 hours carbamylation, 66.1 lysine residues for 48 hours, and 77.7 lysine residues for 72 hours, respectively. At the 72-hour reaction period, 88.2% of lysines could be modi- fied at most. These corrected rates are shown in Table III. The values of s&, and [T] were close to the values of the native myeloma IgG. Ultraviolet absorption spectrum of carbamyl- ated myeloma IgG was identical with the native one as shown in Fig. 1. CD spectrum of IgG carbamylated for 72 hours also resembled the spectrum of native IgG, as shown in Fig. 4. Ionic strength variation had no effect on the CD spectra and viscosity. Immunochemical reactivity tested with goat antiserum showed positive precipitation reactions that were independent of the number of modified lysine groups.

DISCUSSION

In this paper the studies deal with conformation of IgG modi- fied by citraconic anhydride and potassium cyanate. The former modification introduces acidic groups into e-amino groups re- versibly, and the latter compound converts the e-amino group to an electrostatically neutral one. The changes are shown schematically as follows:

cl-

O=C-NH, H3C-C - c=o CH=C-C=O

NH;

/I I

b I I

O=C CH3 NH

I HC -C=O NH

I

coo-

Carbamylated IgG

pH-4

coo-

Cttraconylated IgG

According to our experimental findings, conversion of the e-amino groups of lysine residues into negatively charged groups on citraconylation resulted in expansion and partial disorgani- zation of the immunoglobulin molecules. The ability to form precipitate with goat antiserum also was lost on such conversion. However, the precipitability and the conformational features of the native globulin were restored in two ways: either by in-

-4 t 11 z 8 ‘I/ I

190 210 230 250 250 270 290 310

WAVELENGTH hml

FIG. 4. CD spectra of carbamylated myeloma IgG at various ionic strengths. The conditions are identical with those of Fig. 1.

by guest on May 14, 2020

http://ww

w.jbc.org/

Dow

nloaded from

5707

creasing the ionic strength or by cleaving off the citraconyl res- ever, impossible to state that all 0 structures were disrupted idues at pH 4.0. Moreover, we observed that carbamylation and to what extent the unspecified other rigid conformations of the t-amino groups of the lysine residues did not affect the were involved in the conformational transition. CD spectra, viscosity, and immunological precipitability. We did not examine in detail the 230. to 250.mu spectral zone, Hence we concluded that (a) the native conformation rather but others (22, 37) have paid much attention to a positive band than the presence of free c-amino groups is essential for the im- centered near 232 nm. In comparison to the previously de- rnunoprecipit’ation reaction, and (b) that the conformational scribed bands (observed at 200 and 216 nm), this-band is very transition incited by citraconylation is caused by electrostatic weak as indicated by a hump in the CD curve in our Fig. 3. repulsion. According to Cathou et al. (22) and Dorrington and Srnith (37),

A detailed specification of the reversible conformational transi- this band should be attributed to tyrosine chromophorcs in asym- tion on citraconylation is presently impossible, because of the metrical protein environtncnt. limit,ed information that is available on conformation of native The changes of the CD in the near-ultraviolet spectral zone immunoglobulins. In spite of the many efforts, the three-di- also are difficult to explain in detail, chiefly because of overlap mensional structure (conformation) of the immunoglobulins of the CD bands caused by the vicinal effects on tyrosinc, tryp- is only partially elucidated. httempts to obtain good immuno- tophan, and phenylalanine side chains and the inherent nsym- globulin crystals suitable for x-ray structural analysis have been metry of the disulfide bonds. The bands near 284 and 291 nm disappointing; and the low resolution results on imperfect crys- can be attributed to tryptophan (37, 38), but the effects of the tals have not revealed much about conformation (16-18). The other aromatic side chains cannot be fully excluded in this zone. limited information presently available on the secondary struc- The CD bands in the near-ultraviolet zone, depending on the ture of these proteins is from ORD (7, 18-21) and CD data on protein asymmetrical environment, can be either positive or solutions (22-24). ORD and CD are quite useful methods in negative. Thus any of the weak bands we observed in the 250. detecting changes in conformation, especially if supplemented to 280-nm spectral zone (Fig. 3) may be caused by cancellation by other methods. Expansion of the serum y-globulins of bands of opposite sign as it has been concluded for ot,her in- in strongly acidic and alkaline solutions has been detected as stances by other authors (39,40). Increased freedom of rotation early as 1954 (25-29). However, more interesting are the con- of the aromatic chromophores upon disorganization of a rigid formational transitions which occur in nearly neutral solutions, conformation should result in decrease in magnitude of the CD especially if one wants to consider immunochemical reactivity. effects, as it often is observed in denaturation. The fact that For this reason, attempts have been made to study the confor- the bands became more positive on citraconylation (Fig. 3) may mational transitions in the absence of strong acids, alkalies, be due to the unequal cancellation of the positive and negative concentrated urea, or guanidine salts. One of the ways is to CD in the same zone, i.e. that the disorganization presumably modify the side chains under mild conditions. This approach resulted in weak cancellation of the positive effects and strong allows also to test the importance of the various side chains in cancellation of the “hidden” negative bands in the 260. to 290.nm the immunochemical reactions. Expansion of the macromole- spectral region. cules of human y-globulin on mild succinylation has been de- A detailed analysis of the near-ultraviolet zone would bc ex- scribed by Habeeb (2); also he observed that this conformational tremely difficult at present because of the complexity of struc- transition was accompanied by a moderate loss of reactivity tures and large number of the chromophoric groups in the var- toward antibody. More recently, similar observations were ious parts of these macromolecules. Work is in progress on reported on maleylation of IgG (6). In none of these reports, the CD effects of chemically modified Fragment Fab and Fc, however, were the conformational changes tested by chiroptical as well as on light chains of definite antigenic type (Rence-Jones methods. proteins). It is hoped that conformational analysis of these

While the increase in intrinsic viscosity and decrease of sedi- mentation coefficient are clear indications of the expansion of IgG, as we observed on citraconylation, the changes in the sec- ondary structure cannot be interpreted with the same degree of confidence. The positive CD band centered near 200 nm and the negative band at 216 to 218 nm are indicative of the p structure (30-32) and replacement of the positive band with a negative one indicates loss of t,he fl form and disorganization of the polypeptide backbone. However, the amount of the pleated sheet 0 conformation camlot be very high in these globu- lins; neither is there any indication of a high LY helix content, as it, has been concluded by several groups of investigators (7, 22, 23, 29, 33, 34). Although some parts of the IgG macromolecule are flexible (35), the major portions are rigid. Thus it is likely that the major secondary structure in the immunoglobulins is neither the /3 form, cy helix, nor the flexible random coil but some unspecified aperiodic conformation, as it now is disclosed by x-ray structural analysis in many proteins (e.g. in chymo- trypsin by Matthews et al. (36)). From the changes we ob- served in the far-ultraviolet CD spectra (Fig. 3) we can conclude that some of the secondary structure was lost on citraconylation, and that the disorganization probably was due to the hydrogen bond disruption of the pleated sheet conformation. It is, how-

macromolecules will yield more meaningful results about the CD bands in the near-ultraviolet zone than the present data on the whole IgG.

Previously lysine residues were converted to amidine which causes no electrostatic changes (1) and it was found that, amidi- nation of e-amino groups did not change physical and immuno- logical properties. In the present work, carbamyl groups were introduced onto the +amino groups and the resulting protein also showed no change in physical, hydrodynamic, and immuno- logical properties (Table III and Fig. 4). Singer and Thorpe (41) have shown that tyrosinc is involved in the active sit,e of immunoglobulin. hlt,hough the tyrosyl residues of tjhe citra- conyl-IgG were not modified, no precipitation react>ion against goat antiserum was observed. Thus neither the intact tyrosine residues nor lysine side chnills secure the immunoprecipitation if the specific native conformation is not maintained. The find- ing that native IgG, citraconyl-IgG at high ionic st,rengt,h, and carbamyl-IgG in a wide range of salt concentration yielded prac- tically identical CD spectra in the far- and near-ultraviolet spec- tral zones is noteworthy. It shows that the chemical modifica- tion of a significant .numbcr of lysine side chains has no effect on the conformation of t~hese macromolecules.

by guest on May 14, 2020

http://ww

w.jbc.org/

Dow

nloaded from

5708

1. 2.

3.

4.

5.

6.

7.

8. 9.

10. 11.

12. 13.

14.

15.

16.

17.

18. 19.

20.

REFERENCES 21. STEINER, L. A., AND LOWEY, S. (1966) J. Biol. Chem.

WOFSY, L., AND SINGER, S. J. (1963) Biochemistry 2, 104-116 241, 231-240

HAREEB, A. F. S. A. (1967) Arch. Biochem. Biophys. 121, 652- 22. CATHOU, R., KULCZYCKI, A., JR., AND HABER, E. (1968) Bio-

664 chemistry 7, 3958

FREEDMAN, M. H., GROSSBERG, A. L., AND PRESSMAN, D. 23. DOI, E., AND JIRGENSONS, B. (1970) Biochemistry 9, 1066

(1968) Biochemistry 7, 1941 24. DORRINGTON, K. J., AND TANFORD, C. (1970) Advan. Zmmunol.

BEZKOROVAINY, A., ZSCHOCKE, R., AND GROHLICH, D. (1970) 12, 333

J. Zmmunol. 104, 648 25. JIRGENSONS, B. (1954) Arch. Rio&em. Biophys. 48, 154-166

HABEER, A. F. S. A., SCHHOHENLOHER, R. E., AND BENNETT, 26. YANG, J. T., AND FOSTER, J. F. (1955) J. Awier. Chem. Sot. 77,

J. C. (1971) Fed. Proc. 30, 593 2374

BEZKOROVAINY, A., AND GROHLICH, D. (1972) Zmmunochem- 27. EDELHOCH, H., LIPPOLDT, R. E., AND STEINER, R. F. (1962)

istry 9, 137 J. Amer. Chem. SOC. 84, 2133

Ross, D. L., AND JIRGENSONS, B. (1968) J. Biol. Chem. 243, 28. CALLAGHAN, P., AND MARTIN, N. H. (1963) Biochem. J. 87,

2829-2836 225

STARK, G. R. (1967) Methods E?azymol. 11, 590 29. GOULD, H. J., GILL, T. J., III, AND DOTY, P. (1964) J. Biol.

SPACKMAN, D. H., STEIN, W. H., AND MOORE, S. (1958) And. Chem. 239, 2842-2851

Chem. 30, 1190 30. SARKAR, P. K., AND DOTY, P. (1966) Proc. Xat. Acad. sci.

LENARD, J., AND SINGER, S. J. (1966) Nature 210, 536 U. S. A. 66, 981

STARK, G. R., STEIN, W. H., AND MOORE, S. (1960) J. Biol. 31. IIZUKA, E., AND YANG, J. T. (1966) Proc. Nat. Acud.

Chem. 236, 3177-3181 Sci. U. S. A. 66, 1175

SCHACHMAN, H. K. (1957) Methods Enzymol. 4, 32 32. WOODY, R. W. (1969) Biopolymers 8, 669

RIORDAN, J. F., AND VALXX:, B. L. (1964) Biochemistry 3, 33. TROITSKII, G. V. (1965) Bioj%ka 10, 895

1768-1774 34. IKEDA, K., HAMAGUCHI, K., AND MIGITA, S. (1968) J. Biochem.

NAKAGA~A, Y., AND PERLMANN, G. E. (1972) Arch. Biochem. (Tokyo) 63, 654

Biophys. 149, 476 35. NOELKEN, M. E., N~;LSON, C. A., BUCKLEY, C. E., III, AND

WOODS, K. R., AND WANG, K-T. (1967) Biochim. Biophys. TANFORD, C. (1965) J. Biol. Chem. 240, 218-224

Acta 133, 369370 36. MATTHEWS, B. W., SIGLER, P. B., HENDERSON, R., AND BLOW,

SOLOMON, A., MCLAUGHLIN, C. L., WEI, C. H., AND EINSTEIN, D. M. (1967) Nature 214, 652

J. R. (1970) J. Biol. Chem. 246, 5289-5291 37. DORRINGTON, K. J., AND SMITH, B. R. (1972) Biochim. Biophys.

Acta 263, 70 DAVIES, D. R., SARMA, R., LABAW, L. W., SILVERTON, E.,

&GAL, D., AND TERRY, W. D. (1971) Alan. N. Y. Acad. Sci. 38. STRICKLAND, E. H., HOR~ITZ, J., AND BILLUPS, C. (1968)

Biochemistry 8, 3205 190, 122 39. STRICKLAND, E. H., KAY, E., AND SHANNON, L. M. (1970) J.

JIRGENSONS, B. (1958) Arch. Biochem. Biophys. 74, 57-69 Biol. Chem. 246, 1233 WINKLER, M., AND DOTY, P. (1961) Biochim. Biophys. Acta 40. HORWITZ, J., STRICKLAND, E. H., AND BILLUPS, C. (1970) J.

64, 448454 Amer. Chem. Sot. 92, 2119 IMAHORI, K., AND MOMOI, H. (1962) Arch. Biochem. Biophys. 41. SINGER, S. J., AND THORPE, N. 0. (1968) Proc. Kat. Ad. Sk.

97, 236-242 U. S. A. 60, 1371

by guest on May 14, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Yasushi Nakagawa, Sylvia Capetillo and Bruno JirgensonsHuman Immunoglobulin G

Effect of Chemical Modification of Lysine Residues on the Conformation of

1972, 247:5703-5708.J. Biol. Chem. 

  http://www.jbc.org/content/247/18/5703Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/247/18/5703.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on May 14, 2020

http://ww

w.jbc.org/

Dow

nloaded from