THE J0”Rm.L OF BioLoGicAL CHEMrsTnY “a,. 252, No. 3, Issue of February 1IJ, pp 883-889, 1977

Printed in U.S.A

Primary Structure of the “Hinge” Region of Human IgG3 PROBABLE QUADRUPLICATION OF A SAMIN ACID RESIDUE BASIC UNIT*

(Received for publication, July 28, 1976)

TERJE E. MICHAELSEN,~ BLAS FRANGIONE,§ AND EDWARD C. FRANKLIN§

From the §Irvington House Institute, Departments of Medicine and Pathology, New York University Medical Center, New York, New York 10016

The middle part of the heavy chain of IgG3 (hinge region) which covalently links the two y3 chains to each other, is about 4 times larger than the same region in the three other human IgG subclasses. This is probably due to a quadrupli- cation of a 45-nucleotide DNA segment resulting in a y3 hinge region which is 62 amino acid residues long and con- sists of an NH,-terminal 17-residue segment followed by a S-residue segment which is identically and consecutively repeated three times. The NH,-terminal 17-residue segment shows 70% homology with the repetitive 15.residue segment and appears to be the result of a small insertion and several point mutations of the same 45-nucleotide DNA stretch. Since this unit of repetition shows 60 to 70% homology with the hinge of the other IgG subclasses, it may represent the primitive IgG hinge.

It has recently become evident that human IgG3 has struc- tural and biological features not shared by the other three subclasses of human IgG. Functionally, IgG3 does not bind to staphylococcal protein A as do the other subclasses (11, has a more rapid turnover (21, and is the most potent subclass in activating the Cl component of complement (3). Structurally,

the y3 chain has a larger molecular weight than other human y chains (4-6) and IgG3 molecules have an unusually large radius of gyration in aqueous solutions (6, 7). Both features are a result of an unusually large middle part (hinge region) present in the y3 chain (6, 7). Furthermore, this region of y3 has a strikingly well ordered secondary and tertiary structure revealed by very simple circular dichroism spectra of an iso- lated hinge fragment, ctFh’ (8). The IgG3 subclass is also

* This work was supported by United States Public Health Service Research Grants AM 01431 and AM 02594 and United States Public Health Service Training Grant AM 05064 and Grant C98 99-12 from the Norwegian Council for the Sciences and Humanities.

$: United States Public Health Service International Research Fel- low. Present address, Institute of rmmunology and Rheumatology, F. Qvamsgt. 1, Oslo 1, Norway.

I The abbreviations and fragment nomenclature used are: ABA, a-aminobutyric acid, PTH, phenylthiohydantoin; dansyl, 5-dimeth- ylaminonaphthalene-1-sulfonyl; Me, allylamine, N,N-dimethyl-N- allylamine. ctFh, a 55-residue hinge fragment obtained by a-chymo- trypsin digestion; pa ctFh, a 29-residue hinge fragment obtained by papain digestion followed by a-chymotrypsin digestion; pFh, a 64. residue hinge fragment obtained by pepsin digestion; tF(ab),, a divalent Fab-type fragment obtained by trypsin digestion; ctFab, a monovalent Fab-type fragment obtained by u-chymotrypsin diges-

characterized by a great genetic polymorphism (9). Previous studies of y3 heavy chain disease proteins have shown that these unusual features are the result of a series of duplications involving the cysteine residues of the hinge (10, 11). The relationship of the extended hinge region to the biological features of IgG3 is still an open question. The aim of the present report is to determine the complete primary structure of the hinge region of normal sized y3 chains and to define precisely the number of cysteine residues, the order of the repeated sequences, and the relationship of the hinge region to the Cyl and Cy2 domains.

MATERIALS AND METHODS

Myeloma Proteins (IgG3) -1gG3 proteins were isolated from the sera of three patients (HER, FRO, and JON) with multiple myelo- matosis as previously described (6) using DEAE-cellulose (What- man DE52) or quaternary aminoethyl (QAEl-Sephadex. For some experiments, a fourth protein, a cryoglobulin (KUP) was used which was isolated by repeated precipitation at low temperature (12). The allotypes of the four IgG3 preparations were determined by agglu- tination-inhibition techniques (9) to be HER Gm(g), JON Gm(g), FRO Gm(b?, and KUP Gm(b”).

Gel Filtration- Columns packed with different kinds of Sephadex were used and operated under neutral or dissociating conditions as previously described (6). Standard curves showing the relationship between K,, (13) and molecular weights were constructed for each of the columns employing standard proteins (6).

Isoelectric Focusing- Isoelectric focusing was performed using a LKB (Bromma, Sweden) 8101 (110 ml) column, focusing in 1% Am- pholine (LKB) at pH 5 to 7. Focusing was done at 500 V for 3 days.

Isolation of Chymotryptic Hinge Fragment (ctFh) -The ctFh frag- ment was isolated from three of the IgG3s essentially as described before (14). Briefly, the IgG3 proteins were digested with a-chymo- trypsin (Sigma, crystallized three times) (l/50, w/w, enzyme/sub- strate ratio) in 0.1 M phosphate buffer, pH 7, for 24 h at 37”. The digest was dialyzed and subjected to gel filtration on a column of Sephadex G-100 in 5 M guanidine HCl, 1 M acetic acid (Fig. la). The main optical density peak contained monovalent ctFab 04, = 44,000) and was well separated from the ctFh-containing peak which was detected by an anti-Fh specific antiserum (14). The ctFh fragments were further purified (after dialysis against distilled water) by isoe- lectric focusing (Fig. lb). All analyses and further digestions were done on the isoelectrically focused, purified ctFh (with p1 5.85) after removal of the Amoholine and sucrose bv eel filtration on Senhadex G-25 in 10% formic acid.

” I

Isolation of Peptic Hinge Fragment (pFhi - IgG3 KUP and HER were digested with pepsin (l/100 w/w) for 20 h in 5% formic acid at 37”. The pFh was purified by gel filtration on Sephadex G-75 in 10%

tion; Cyl-Fh, a fragment containing the Cyl domain and the hinge region of IgG3; Fch, a papain fragment which includes the Fc frag- ment and part of the hinge.

883

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

y3 Hinge

formic acid and was further purified by ion exchange chromatogra- phy on Whatman DE52 and eluted with 0.01 M phosphate buffer, pH 7.6.

Isolation of Fd’ Fragments from IgG3 - Fd’ fragments from IgG3 were isolated as previously described (6, 7). Briefly, the F(ab’), peptic fragment was partially reduced and alkylated with [“Cl- iodoacetic acid before separating the Fd’ fragment from light chains by gel filtration on Sephadex G-100 in 5 M guanidine HCl, 1 M acetic acid (6, 7).

Isolation of Fch fragments from IgG3-Fch fragments were pre- pared as described previously by digesting IgG3 (20 mglml in 0.1 M phosphate, pH 7.0) with papain (crystallized three times, Sigma) (l/ 100, w/w) for 45 min without reducing agents (15). Fch which con- tains the Fc region and part of the hinge was isolated by gel tiltra- tion on Sephadex G-150, followed by ion exchange chromatography on CM-Sephadex (15). The final preparations were often slightly contaminated with Fc fragments which are the end product of Fch if the papain digestion is continued (15).

Citraconylation - To prevent trypsin digestion at lysine residues, the lysine residues were modified by introducing citraconylic groups (16). Citraconylation was performed with the protein dissolved (5 to 20 mg/ml) in 6 M guanidine HCl, 0.005 M borate, pH 8.5, by adding citraconylic anhydride (Pierce) in 30 molar excess over the lysine content. The total volume of citraconylic anhydride was added over a period of about 15 min under constant stirring. The pH was kept at 8.0 to 8.5 by adding 2 M NaOH. After standing at 4” for another hour, the samples were extensively dialyzed against 0.001 M NH,HCO,, pH 8.5, before lyophilization and storing at -30” until use.

Reduction and Alkylation - Partial and total reduction and alkyl- ation was performed as previously described (6).

Amino Acid Analysis - Amino acid analyses of peptides and pro- teins were performed after hydrolysis in 6 M HCl for 16 to 78 h (6, 17) either with a Beckman 120 or a Durrum D-500 automatic amino acid analyzer.

NH,-terminal Amino Acid and Amino Acid Sequence Analysis- NH,-terminal amino acid analysis was performed by the dansyl chloride method and identification was done on polyamide thin layer plates (15 x 15 cm) with standards plus samples on one side and only samples on the other (18, 19). The plates were run in two dimensions in three different solvent mixtures (19). Automated amino acid sequence analysis was done on a Beckman 89OC Sequencer using the Beckman Me, allylamine peptide program (10/29/74). Identification of amino acids was generally done by three and a minimum of two methods. Gas-liquid chromatography was performed with a 7620A Hewlett-Packard gas chromatograph on columns packed with Beck- man coated support No. 56796 (201. Thin layer chromatography of PTH-derivatives was done on polyamide plates (5 x 5 cm), in two dimensions (21). Standards were placed on one side and samples on the other. The amides of aspartic acid or glutamic acid were clearly discriminated by thin layer chromatography but for unknown rea- sons, PTH-Ser and PTH-Thr were difficult to detect. The remainder of the aliquot was hydrolyzed under vacuum with hydriodic acid (HI), and the products identified by amino acid analysis (22). All PTH-derivatives were converted to the native amino acids except for PTH-Ser which was converted to alanine and PTH-Thr which was converted to a-aminobutyric acid (ABA) and PTH-Trp which was destroyed. Cysteine was detected by counting radioactivity of PTH- [“Clcarboxymethylcysteine and thin layer chromatography. The runs were terminated when clear-cut identification of a residue was no longer possible.

High Voltage Electrophoresis - High voltage electrophoresis was done on Whatman No. IMM or 3MM paper at pH 2.1, 3.5, or 6.5 by employing Savant (Hicksville, N.Y.) liquid-cooled electrophoresis tanks (23). Peptides were located by autoradiography if they con- tained 14C-labeled cysteine or by staining a guide strip of the paper with 1% ninhydrin in acetone containing 0.1% CdSO,. Regions of the paper containing peptides were cut out for elution or further purifi- cation by electrophoresis at a different pH. Peptides were eluted from the paper with 6 M HCl if they were to be used only for amino acid analysis, or with 1% acetic acid if used for other purposes.

Diagonal Electrophoresis - Cysteine-containing peptides were spe- cifically isolated by diagonal electrophoresis taking advantage of a change in their charge after performic acid oxidation (24).

Hinge Specific Antiserum-A y3 hinge specific antiserum (anti- Fh) was raised in a rabbit by immunizing with Ffab’), fragments of IgG3 (HER) and absorbing with insolubilized Fab fragments of IgG3 (HER) (14).

RESULTS

Analysis of Chymotryptic Hinge Fragment, ctFh, from IgG3 Proteins of Different Allotypes - The ctFh fragment was iso- lated from the proteins HER, FRO, and JON as described under “Materials and Methods” by gel filtration in 5 M guani- dine, 1 M acetic acid (Fig. la) followed by isoelectric focusing (Fig. lb). The molecular weight of ctFh obtained by gel filtra- tion (Fig. la) was 10,000 to 12,000. The amino acid composition was calculated (Table I) assuming 1 residue of leucine per chain (M, = 6,000). The amino acid compositions of ctFh fragments were identical for all IgG3 proteins regardless of allotype. The NH,-terminal amino acid sequence of completely

1 reduced and W-alkylated ctFh was shown to be: Thr-Cys-Pro-

10 Arg-Cys-Pro-Glu-Pro-Lys-Ser-Cys-Asp-Thr-Pro-Pro-Pro-Cys. In addition, on the basis of radioactivity, [Wlcarboxymethyl- cysteine was identified at position 20 and 26.

Isolation and Amino Acid Sequence of 15-Residue Peptide Which Is Repeated Three Times in ctFh Fragment - Since the complete sequence of ctFh was not obtained by automated sequencing analysis, the ctFh (from HER) was citraconylated and further digested with trypsin (l/50, w/w for 2 h at pH 8.5). The digest was then totally reduced and W-alkylated and gel- filtered on Sephadex G-15 in 5% formic acid (Fig. 2). The

40

I- n E 3.0 c

8 D 2o Cl

IgG J.Her, chymotrypsin digestion 24 h

~~~

3.4 7

1.0 6

0.6 5

0.6 4

0.4 3

0.2 2

1 i0 i0 i0 Fraction M.

FIG. 1. a, gel filtration on Sephadex G-100 in 5 M guanidine HCI, 1 M acetic acid of a-chymotryptic digest (l/50, w/w for 24 h at pH 7.0) of IgG3 (HER). Column dimensions: 3.2 x 95 cm. Sample: 10 ml. Sample concentration: 50 mglml. Elution rate: 9 ml/h. Volume per fraction: 4.8 ml. ctFh (indicated by arrow) was eluted as a dimer corresponding to a M, = 10,000 to 12,000. b, isoelectric focusing of material from the Fh peak (a) at pH 5 to 7 in 1% Ampholine (LKB) on an LKB 8101 isoelectric focusing column after dialysis against distilled water. Each fraction was assayed for the presence of “hinge”-containing fragments by radial immunodiffusion (28) with anti-Fh antiserum (O-O). The pH was also measured in each fraction (0-O). Absorbance = -.

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

y3 Hinge 885

amino acid composition of material from each peak (I, II, and III) is shown in Table I. The amino acid sequence of material

1 in Peak I was shown to be Cys-Pro-Glu-Pro-Lys-Ser-Cys-Asp-

10 Thr-Pro-Pro-Pro-Cys-Pro-Arg. The repetitive yield for proline was 95% and that of Cys(Cm) based on counts per min was 82%. Thus, in order to account for the amino acid composition of ctFh (Table I), this 15-residue segment must be present three times in ctFh. Amino acid composition of material in Peaks II and III (Table I) indicated that they could be placed at the COOH-terminal and NH,-terminal ends of ctFh, respec- tively (see below and Fig. 61, and that they contain 1 carboxy- methylcysteine residue each. Consequently, based on the amino acid composition of ctFh (Table I) and the radioactivity (Fig. 2), ctFh fragment contains 11 S-S bonds and must contain a triplication of the peptide in Peak I.

Amino Acid Sequence of ctFh Fragment-In order to estab- lish the complete primary structure of the ctFh fragment, intact IgG3 (HER) was digested with papain in the absence of

TABLE I

Amino acid composition of different fragments and peptides of y3 hinge

The numbers are given as moles of residues per mol of peptide.

Fig. 2 Fig. .?c Fig. 4 __

Amino ctFh acids - PFh

Peak Peak Peak I IIb IIIb

Asp 3.0 4.1 1.1 2.0 1.0 Thr 3.9 6.6 1.1 0.9 1.7 1.0 1.0 Ser 2.9 3.0 1.0 1.4 0.9 Glu 4.1 4.7 1.1 1.1 2.0 1.0 1.0

Pro 21.7 22.8 5.6 2.0 0.9 9.1 2.0 4.2 2.0 0.9 Gly 1.0

Ala 1.0 1.2 1.0 1.0 1.0 ‘I2 cys 10.6” 11.4” 2.5” 0.5” 0.6” 4.0’ 0.8c 1.7’ 0.9’ 0.8’ Val 0.3 Leu 1.1 2.0 1.1 0.9 1.0 His 1.3 0.3 LYS 2.9 4.1 1.2 1.1 I.1 Al%! 3.8 3.3 1.1 1.2 1.1 1.0 1.0

U Average of three different IgG3s (HER, FRO, and JON). b Analysis after further purification by paper electrophoresis, pH

6.5. ’ Determined as cysteic acid after performic acid oxidation (26). ” Determined as carboxymethylcysteine, thus giving too low val-

ues.

10 a 30 40 50 Fractla” Number

FIG. 2. Gel filtration on Sephadex G-15 in 5% formic acid of a trypsin digest of citraconylated ctFh. The sample was totally reduced and alkylated with [‘Qiodoacetic acid before being applied to the column. Column dimensions: 2.4 x 45 cm. Sample concentration: 3 mglml. Fraction volume: 1.8 ml. Thirty microliters from each frac- tion were taken for liquid scintillation counting.

cysteine. It has been previously shown (15) that papain under these conditions cleaves the hinge near the NH,-terminal end as well as other places, and yields Fch fragments which con- tain the Fc region and part of the hinge (for purification, see “Materials and Methods”). The Fch preparation was further digested with a-chymotrypsin (l/50, w/w for 4 h at pH 7.0) and liltered on a Sephadex G-150 column under neutral conditions (Fig. 3~). Radial immunodiffusion analysis of the gel filtration fractions revealed a broad anti-Fh reactive peak which was divided into two fractions (1 and 2). After desalting on a Sephadex G-25 column in 10% formic acid, materials in the two pools were further separated by isoelectric focusing (pH 5 to 7) and both showed three peaks reacting with anti-Fh antiserum (Fig. 3, b and c). Based on their elution position on Sephadex G-150 and their amino acid analysis, it was shown that the material with pI 5.8 is identical with ctFh (Table I). The fragment with ~14.7 has a lower molecular weight, which together with the amino acid composition (Table I, pa ctFh) indicate that this represents the COOH-terminal half of ctFh. This was proven by sequence analysis which yielded the se-

1 10 quence: Asp-Thr-Pro-Pro-Pro-Cys-Pro-Arg-Cys-Pro-Glu-Pro- Lys-Ser-Cys-Asp (Fig. 6, see arrow pa). [W]Carboxy- methylcysteine was also detected at positions 21 and 24. Thus, this sequence analysis gives direct proof for the existence of a repeated sequence in the y3 hinge by revealing glutamic acid in position 11 in the pa ctFh fragment, instead of alanine (231 yl numbering) which should be in this position if there were no repetitions (see Fig. 6).

Arrangement of S-S Bonds in ctFh Fragment-To study the arrangement of the S-S bonds in the hinge, diagonal electrophoresis was performed on a tryptic digest of ctFh (Fig. 41. All the cysteines appeared by this analysis to form inter- chain bonds based on the independent electrophoretic mobility and amino acid composition of the peptides (Table I, peptides 1, 2, 3, and 4). Since trypsin does not split between the 2 cysteines in peptide 2 which has the sequence: Ser-Cys-Asp- Thr-Pro-Pro-Pro-Cys-Pro-Arg-, these cysteines could form an intrachain S-S bond. To test this, material in Band 2 ob- tained after the first dimension (Fig. 4) was eluted and incu- bated with papain in order to split the Cys-Asp bond as noted when intact IgG3 molecules were digested to get pa ctFh. Then diagonal electrophoresis was run of the incubation mixture. Under a variety of incubation conditions (variations in incuba- tion time, substrate enzyme ratio, and pH), no such cleavage was noted since either no digestion or random cleavage at many bonds occurred. Thus, while the first and third pair of cysteine residues of peptide 2 (see Fig. 6) could be involved in intrachain S-S bonds, this possibility is unlikely since the corresponding cysteines in the second repetition are not, as shown by the appearance of pa ctFh fragments by treatment with papain without using reducing agents.

Characterization of COOH and NH, Termini of Hinge- In order to evaluate the possibility that there could be additional repeated sequences, COOH-terminal to the ctFh, two IgG3 myeloma proteins of different Gm types (HER and KUP) were digested with trypsin (11100 for 2 h at pH 7.0) followed by gel filtration on Sephadex G-200 under neutral conditions. This yielded two components: one with a molecular weight of about 100,000 reacted with an antiserum to the Fab fragment and represented tF(ab),; the other, reactive with anti-Fc had a molecular weight of about 50,000 and represented a disulfide- linked dimer of the COOH-terminal part of the heavy chain. After complete reduction and alkylation, the NH,-terminal

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

~3 Hinge

Fc + Fch chymotrypsin 1:50, 4h 0 a Sephadex G 150

1 L,!.+Y-Ty-. \ ’ , , , 1 *

300 400 500 600 700 Elution volume ml

1 2 isoejectric focusing pH 5-7

r2

FIG. 3. a, gel filtration on Sephadex G-150 under neutral conditions of an a- chymotrypsin digest (l/50, w/w, 4 h at pH 7.0) of an Fch preparation contami- nated with Fc fragment. Column di- mensions: 3.2 x 93 cm. Quantitation (radial immunodiffusion) (25) with an anti-Fh antiserum showed that the Fh peak (0-O) was eluted at a position corresponding to M, = 37,000. 5, iso- electric focusing at pH 5 to 7 with 1% Ampholine of Pool 1 from a. Each frac- tion was assayed for the amount of Fh- related fragments (O-O). pH in each fraction (0-O). Absorbance = -. c, isoelectric focusing at pH 5 to 7 with 1% Ampholine of Pool 2 from a. (0-O) Quantitation with anti-Fh. The quantitative results in a, b, and c are arbitrarily given as the square of the radius (r) of the precipitation circle with the antiserum. pH in each fraction (0-O). Absorbance = -.

3b lb Fraction number

1 sequence of tFc was found to be: Cys-Pro-Ala-Pro-Glu-Leu-

10 20 Leu-Gly-Gly-Pro-Ser-Val-Phe-Leu-Phe-Pro-Pro-Lys-Pro-Lys- Asp-Thr-Leu-Met-Ile-Ser. This indicates that trypsin cleaved the molecule immediately prior to the last cysteine of the hinge, and that following this the sequence became identical with the Cy2 of y3 (27). The finding rules out the possibility of further duplications COOH-terminal to the ctFh fragment.

Since, based on the previous results from heavy chain dis- ease proteins (10, ll), ctFh apparently does not include the whole hinge at the NH,-terminal side, it was necessary to obtain an overlap peptide stretching from the C-y1 to the hinge. To do so, Fd’ fragments of +y3 (HER and KUP) were prepared (see “Materials and Methods”) and treated with ci- traconic anhydride to mask lysine residues followed by tryp- sinization and gel filtration on Sephadex G-50 fine in 10% formic acid (Fig. 5). The main radioactive Peak IV contains the l&residue hinge repetitive unit and also another cysteine- containing peptide which was further purified by paper elec- trophoresis at pH 6.5, followed by electrophoresis at pH 2.1. The amino acid composition of this peptide was: Asp,,,, Thrs,T, G~u~.~, Pro,.,, Gly,.,, CY%;, VaLx, Len,.,, His,.,, LYS,.,, h-g,.” and the amino acid sequence of the first residues was deter-

1 10 mined to be: Val-Glu-Leu-Lys-Thr-Pro-Leu-Gly-Asp-Thr-. This peptide begins NH2-terminal to ctFh (see Fig. 6) since a 64-residue pFh fragment obtained by pepsin digestion of intact IgG3 proteins HER and KUP (Table I, pFh) had an NH,-

1 terminal sequence: Lys-Thr-Pro-Leu-Gly-Asp-Thr-Thr-His- 10

Thr-Cys-Pro-Arg-Cys, thus placing its start at the 4th position

of the above NH,-terminal hinge peptide. Furthermore, a tryptic peptide with the amino acid composition Val, Asp, Lys, Arg was recovered from Cyl-Fh (the purification of this frag- ment which contains the Cyl domain and the hinge will described elsewhere). This peptide could correspond to the Val- Asp-Lys-Arg sequence in yl (29) which immediately precedes residue Val (215) and is probably located in the Cyl domain NH,-terminal to the Val-Glu-Leu sequence (Fig. 6). These results taken together speak against the possibility of addi- tional repetitions.

Determination of Number of Disulfide Bonds in Hinge-To confirm the results of the sequence studies, the number of S-S bridges in the hinge was estimated by subjecting the whole IgG3 molecule to partial reduction and W alkylation and separating the heavy and the light chain on Sephadex G-75 in 10% formic acid. Reduction was done at dithiothreitol concentrations ranging from 5 to 50 mM and alkylation was performed with a 20% molar excess of [“Cliodoacetic acid. The total content of radioactivity in the heavy chain peak was compared with that of the light chain peak. Assuming that 1 cysteine was alkylated in the light chain, the number of alkylated cysteines in the heavy chain was calculated as shown in Table II. At 10 mM dithiothreitol and higher, 12 cysteines in the heavy chain (11 from the hinge and 1 from H-L bond) were alkylated while at 5 InM dithiothreitol, one of the hinge S-S bonds was not reduced.

DISCUSSION

Previous studies have clearly demonstrated that some of the unique structural features of the y3 heavy chain are due to the presence of an extended hinge which is unusually rich in cysteine and proline residues as a result of a series of duplica- tions (6-8, 10, 11). However, most of the work on the primary

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

y3 Hinge 887

FIG. 4. Diagonal electrophoresis at pH 3.5 of a tryptic digest of ctFh. After electrophoretic separation, one paper was stained with 1% ninhydrin with CdSO,. Another paper was stained with 0.05% ninhydrin to localize the peptides which were subsequently cut out from the paper and eluted with 6 M HCl, hydrolyzed, and subjected to amino acid analysis (Table I).

structure of the y3 hinge has been done with defective heavy chain disease proteins and has not permitted a precise esti- mate of the number of cysteine residues, nor a clear-cut deflni- tion of the beginning and the end of the hinge in relation to the Cyl and Cy2 domains (10, 11).

The present study elucidates the structure of the hinge region of intact y3 heavy chains based on an analysis of four IgG3 myeloma proteins of varying Gm types, and resolves the discrepancies in earlier reports (6, 10, 17). The results provide clear-cut evidence that the hinge of intact y3 chains contains 11 cysteine residues and that the unusual structure appears to be the consequence of quadruplication of a l&residue segment as previously postulated (11). This proof of repetitive se- quences was obtained from ctFh, a chymotryptic fragment which appears to include all the cysteine residues of the hinge, by digestion with trypsin at arginine residues. In addition, by using a variety of proteolytic enzymes, it was possible to show that there are no additional unexpected sequences at the COOH- and probably not at the NH,-terminal ends of the hinge and that in all probability the arrangement of the cysteine residues is in the form of a series of parallel inter- chain disulfide bonds.

Fig. 6 compares the linear sequence of the hinge of the -y3 chain to that of the yl chain of protein Eu (~1) (29) and provides a possible interpretation of the findings. Starting after the valine which corresponds to residue 215 (yl number- ing) there is an insertion of 47 residues, marked by the solid

FIG. 5. Gel tiltration on Sephadex G-50 iine in 10% formic acid of a trypsin digest (l/50, w/w for 2 h at pH 8.5) of citraconylated, totally reduced, and ‘*C-alkylated HER Fd’ fragment. Column dimensions: 1 x 96 cm. Sample volume: 1 ml. Sample concentration: 20 mg/ml. Fraction volume: 2.0 ml. Thirty microliters of each fraction were taken for liquid scintillation counting. Materials in Peak IV were further purified by paper electrophoresis.

bars, which ends after the Cys-Pro at the end of Line 3. Following this, and starting with the glutamic acid corre- sponding to residue 216 (yl numbering), there is a striking homology between the remainder of the hinge and the hinge of yl. Within this 62-residue region, one can identify a &residue repetitive sequence whose beginning can be placed at any point in the first Pro-Arg-Cys-Pro sequence.

Several hypotheses can be put forward to explain the struc- ture of the hinge, all postulating tandem duplications of a basic DNA sequence or perhaps unequal crossing over be- tween two hinge genes.

Based on the observed homologies, the hinge could have evolved by a tandem quadruplication of a 45nucleotide region of DNA. The alternative which requires the smallest number of mutations, would code for the 15 residues corresponding to the region from Glu 216 to Pro 230 (yl numbering). As seen in Fig. 7, this would give identical sequences for the three COOH- terminal units and requires six substitutions and an insertion of 2 residues in the NH,-terminal segment. This interpretation is favored since it is consistent with, and perhaps explains, the recently described mutant IgGl human myeloma proteins which have deletions of the region from residue 216 to 230 (30, 30, a mutant mouse myeloma protein MOPC 21, IF2 (32) which has a deletion of the Cyl domain and begins normal sequence at a valine corresponding to Glu 216, and the finding in the majority of the y and (Y heavy chain disease proteins of deletions of most of the variable regions and the whole Cyl domain with resumption of normal sequence commonly at Glu 216 (33). Interestingly, one y3 heavy chain disease protein (ZUC) has only retained the last 3 cysteines of the hinge beginning with the Glu-Pro-Lys sequence corresponding to residue 216, the rest of the hinge region being deleted together with the Cyl domain (23). Several other points could also serve as the beginning of this basic subunit of DNA although there is no compelling evidence in favor of any of them.

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

888 73 Hinge

w Y3

& Val-Glu-Leu-Lys-Thr-Pro-Leu-Gly-Asp-Thr-Thr-His-Thr =’ +-Pro-Ar&$Pro-

Asp-Thr-Pro-Pro-Pro

H

-&$-Pro-Arg&$bo-

Y3 Glu-Pro-Lys-Ser FIG. 6. Comparison between the

“hinge” sequence of y3 and yl chains. The sites of digestion of the y3 chain by the various enzymes are indicated as: pe = pepsin; ct = a-chymotrypsin; tr = trypsin digestion after citraconylation; pa = papain. Numbers correspond to those of yl Eu-residues 215 to 254 (29). I = a 47-residue insertion in y3; m = sequence difference between yl and y3; 0 = cysteine residues are numbered consecutively. L = HL, H = HH bond.

Y3

Yl

w ct

Ala-Pro-Glu-LeuLLeu-Gly-Gly-Pro-Ser-Val-Ph~-Le~-Ph~-Pro-Pro-

231 Ala-Pro-Glu-Leu-Leu-Gly-Gly-Pro-Ser-Val-Phe-Leu-Phe-Pro -Pro-

Y3

Yl

Lys-Pro-Lys-Asp-Thr-Leu-Met-Ile-Ser

254 Lys-Pro-Lys-Asp-Thr-Leu-Met-Ile-Ser

TABLE II

Radioactivity in light and heavy chain peaks ofpartially reduced and “C-alkylated IgG3 (HER)

The reduction was done with varying concentrations of dithiothre- itol; alkylation was done with 20% molar excess of [‘Y&doacetic acid. Each result is given as the average of two exDeriments.

Heavy chain peak Light chain peak Concentration of

dithiothreitol No. of Total No.

lkt3 Total No. No. of S-

S bonds

M

0.005 0.01 0.05

cpm

296,000 333,000 321.000

CPm

11.0 26,900 1.0 12.1 27,400 1.0 12.0 26.700 1.0

These structural features of the hinge of ~3, together with the observed homology of 60 to 70% between the hinge regions of other subclasses of IgG and an almost equal degree of homology with the hinge of the cx chains might indicate that the 45nucleotide DNA segment coding for Glu 216 to Pro 230 represents an ancient hinge “gene” which has mutated during the evolution of classes and subclasses. In the case of IgG3, as shown in this report, this region has quadruplicated, and in the human IgAl subclass, a duplication has occurred (34). The explanation of the ZUC protein might then be a deletion of three of the DNA segments.

By several criteria, including diagonal electrophoresis and selective proteolytic digestion, no evidence for any intrachain disulfide bridges was obtained and therefore, it seems likely that all cysteine residues are involved in symmetrical inter-

heavy chain bonds. Nevertheless, as previously reported, low concentrations of dithiothreitol (5 mM) apparently left one S-S bond unreduced (17). Due to the repetition in the hinge, it will be difficult to clarify which S-S bond is more resistant than the rest.

The previously observed homology of greater than 90% in the corresponding domains as compared to 60 to 70% homology in the hinge regions of the IgG subclasses suggests that forces leading to the evolution of human IgG subclasses may have affected the hinge preferentially. This relatively high muta- tion rate in the hinge could either indicate that the hinge can tolerate mutations without affecting its function, or might indicate the existence of forces favoring the structural and perhaps also the functional diversification of the hinge. A specific functional requirement might explain the conserva- tion of the quadruplicated hinge unit of the y3 chain which otherwise would be expected to have been deleted during evolution.

Although the biological function of the hinge remains to be elucidated, its unique structure, apparently independent evo- lution, and conservation in most of the classes and subclasses of immunoglobulin suggest that it may serve some important functions.

Among several different functions for the hinge, one can suggest a possible role in modulating and coordinating the action of the binding (Fab) and effector (Fc) parts of the molecule, perhaps by transmitting a signal from the binding part to the effector part of the molecule when the antibody binds to antigen (35) or in providing a flexible or specific angle between the two binding sites of the molecule (36). Lastly, in

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

y3 Hinge 889

y3 Hinge

Glu 17 15 15 15 PI-0

. i: :::

. :: . . ::i H I4 I4 I I I I I I II I

i = substitution .

I = insertion

FIG. 7. Possible model for the evolution of the hinge through a quadruplication of a 45nucleotide DNA segment coding for 15 amino acid residues. The proposed model (see Fig. 6) is based on the results of human and mouse IgG variants (see “Discussion”). Substitutions compared to the “prototype” hinge unit are marked by dotted lines and insertions by solid lines. Cysteine residues are indicated by solid vertical lines below the units. -

the absence of evidence localizing any specific function to the 16. hinge, it seems possible that it can modify and influence the effector functions of the adjacent domains (3’7).

$

19. 20.

Acknowledgments -The skillful technical assistance of Mrs. E. Rosenwasser and Mr. R. Kutny is gratefully acknowl- 21. edged, and we also thank Ms. M. Chavis and Ms. B. Cooper- smith for secretarial assistance. We are grateful to Dr. K. 22. Sletten for his contribution in the initial phases of the study.

1.

2.

3.

4. 5.

6.

7.

8.

9.

10.

11.

12.

13. 14.

15.

REFERENCES

Kronwall, G., and Williams, R. C. (1969) J. Zmmunol. 103, 828- 833

Sniegelberg, H. L., Fishkin, B. D., and Grey, H. M. (1968) J. Clin. Znuest. 47, 2323-2330

Ishizaka. T.. Ishizaka. K.. Salmon. S., and Fudenbera. H. H. (1967) 2. Zmmunol. 49, 82-91

-.

Saluk. P. H.. and Clem, L. N. (1971)J. Zmmunol. 107.298-301 Virella, G., and Parkhouse, R. M. E. (1972) Immunology 23,857-

860 Michaelsen, T. E., and Natvig, J. B. (1974) J. Biol. Chem. 249,

2778-2785 Michaelsen, T. E., and Natvig, J. B. (1972) FEBS Lett. 28, 121-

124 Johnson, P. M., Michaelsen, T. E., and Scopes, P. M. (1975)

Stand. J. Zmmunol. 4, 113-119 Natvig, J. B., Kunkel, H. G., and Joslin, F. G. (1969) J. Zmmu-

nol. 102, 611-617 Adlersberg, J. B., Franklin, E. C., and Frangione, B. (1975)

Proc. Natl. Acad. Sci. U. S. A. 72, 723-727 Frangione, B. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 1552-

1555 Dammacco, F., Franklin, E. C., and Frangione, B. (1972) J.

Zmmunol. 109, 565-569 Andrews, P. (1964) Biochem. J. 91, 222-233 Michaelsen, T. E., Natvig, J. B., and Sletten, K. (1974) Stand.

J. Zmmunol. 3, 491-498 Michaelsen, T. E., and Natvig, J. B. (1973) Stand. J. Zmmunol.

2, 299-312

23. 24. 25.

26. 27. 28. 29.

30.

31.

32.

33.

34.

35.

36.

37.

Gibbons, I., and Perham, R. N. (1970)Biochem. J. 116,843-849 Michaelsen, T. E. (1973) Stand. J. Zmmunol. 2, 523-529 Gray, W. R. (1967) Methods Enzymol. 11, 469-474 Hartley, B. S. (1970) Biochem. J. 119, 805-822 Pisano, J. J., and Bronzert, T. J. (1969) J. Biol. Chem. 244,5597-

5607 Summer, M. R., Smythers, G. W., and Oroszlan, S. (1973) Anal.

Biochem. 53, 624-628 Smithies, O., Gibson, D., Fanning, E. M., Goodflesh, R. M.,

Gilman, J. G., and Ballantyne, D. L. (1971) Biochemistry 10, 4912-4921

Frangione, B., and Milstein, T. (1968) J. Mol. Biol. 33,893-906 Brown, J. R., and Hartley, B. S. (1966) Biochem. J. 101,214-228 Mancini, G., Carbonara, A. O., and Heremans, J. F. (1965)

Immunochemistry 2, 235-254 Hirs, C. W. (1967) Methods Enzymol. 11, 197-206 Frangione, B., and Milstein, C. (1969) Nature 224, 597-599 Ouchterlony, 0. (1958) Prog. Allergy 5, l-78 Edelman, G. M., Cunningham, B. A., Gall, W. E., Gottlieb, P.

B., Rutishauser, H, and Waxdal, M. J. (1969) Proc. Natl. Acad. Sci. U. S. A. 63, 78-85

Fett, J. W., Deutsch, H. F., and Smithies, 0. (1973) Zmmuno- chemistry 10, 115-118

Rivat, C., Schiff, C., Rivat, L., Ropartz, C., and Fougereau, M. (1976) Eur. J. Zmmunol. 6, 545-551

Milstein, C., Adetugbo, K., Cowan, N. J., and Secher, D. S. (1974) Second International Congress of Immunology, Vol. 1, pp. 157-168

Franklin, E. C., and Frangione, B. (1975) Contemporary Topics in Molecular Immunology, Vol. 4, pp. 89-126, Plenum Press, New York

Frangione, B., and Wolfenstein-Todel, C. (1972) Proc. N&l. Acad. Sci. U. S. A. 69, 3673-3676

Givol, D., Pecht, I., Hochman, J., Schlessinger, J., and Stein- berg, I. Z. (1974) Second International Congress of Zmmunol- ogy, Vol. 1, pp. 39-48

Cathou, R. E., Holowka, D. A., and Chan, L. M. (1974) Second International Congress of Immunology, Vol. 1, pp. 63-73

Michaelsen, T. E., Wisloff, F., and Natvig, J. B. (1974) Stand. J. Zmmunol. 4, 71-78

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from

T E Michaelsen, B Frangione and E C Franklinof a 15-amino acid residue basic unit.

Primary structure of the "hinge" region of human IgG3. Probable quadruplication

1977, 252:883-889.J. Biol. Chem. 

  http://www.jbc.org/content/252/3/883Access 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/252/3/883.full.html#ref-list-1

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

by guest on August 20, 2018

http://ww

w.jbc.org/

Dow

nloaded from