of September 8, 2018.This information is current as

Interaction StrengthCH3−Affecting the Noncovalent CH3

CH3 Domain Enable Fab-Arm Exchange by Species-Specific Determinants in the IgG

Janine Schuurman and Paul W. H. I. ParrenWinkel,Rob C. Aalberse, Albert J. R. Heck, Jan G. J. van de

van den Bremer, Joost Neijssen, Tom Vink, Ignace Lasters,Rose, Tamara H. den Bleker, Stefan Loverix, Ewald T. J. Aran F. Labrijn, Theo Rispens, Joyce Meesters, Rebecca J.

http://www.jimmunol.org/content/187/6/3238doi: 10.4049/jimmunol.1003336August 2011;

2011; 187:3238-3246; Prepublished online 12J Immunol 

MaterialSupplementary

6.DC1http://www.jimmunol.org/content/suppl/2011/08/12/jimmunol.100333

Referenceshttp://www.jimmunol.org/content/187/6/3238.full#ref-list-1

, 10 of which you can access for free at: cites 37 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2011 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Species-Specific Determinants in the IgG CH3 Domain EnableFab-Arm Exchange by Affecting the Noncovalent CH3–CH3Interaction Strength

Aran F. Labrijn,* Theo Rispens,† Joyce Meesters,* Rebecca J. Rose,‡ Tamara H. den Bleker,†

Stefan Loverix,x,1 Ewald T. J. van den Bremer,* Joost Neijssen,* Tom Vink,*

Ignace Lasters,x,2 Rob C. Aalberse,† Albert J. R. Heck,‡ Jan G. J. van de Winkel,*,{

Janine Schuurman,* and Paul W. H. I. Parren*

A distinctive feature of human IgG4 is its ability to recombine half molecules (H chain and attached L chain) through a dynamic

process termed Fab-arm exchange, which results in bispecific Abs. It is becoming evident that the process of Fab-arm exchange is

conserved in several mammalian species, and thereby represents a mechanism that impacts humoral immunity more generally than

previously thought. In humans, Fab-arm exchange has been attributed to the IgG4 core-hinge sequence (226-CPSCP-230) in

combination with unknown determinants in the third constant H chain domain (CH3). In this study, we investigated the role of the

CH3 domain in the mechanism of Fab-arm exchange, and thus identified amino acid position 409 as the critical CH3 determinant

in human IgG, with R409 resulting in exchange and K409 resulting in stable IgG. Interestingly, studies with IgG from various

species showed that Fab-arm exchange could not be assigned to a common CH3 domain amino acid motif. Accordingly, in rhesus

monkeys (Macaca mulatta), aa 405 was identified as the CH3 determinant responsible (in combination with 226-CPACP-230).

Using native mass spectrometry, we demonstrated that the ability to exchange Fab-arms correlated with the CH3–CH3 dissoci-

ation constant. Species-specific adaptations in the CH3 domain thus enable Fab-arm exchange by affecting the inter-CH3 domain

interaction strength. The redistribution of Ag-binding domains between molecules may constitute a general immunological and

evolutionary advantage. The current insights impact our view of humoral immunity and should furthermore be considered in the

design and evaluation of Ab-based studies and therapeutics. The Journal of Immunology, 2011, 187: 3238–3246.

Immunoglobulin molecules are typically composed of two Hchains and two L chains that are covalently connected viadisulfide bonds: one disulfide bond connecting each L chain to

a H chain and a number of disulfide bonds linking the two H chains(1). The cysteines involved in the inter-H chain disulfide bondformation are located in the hinge region and differ in numberbetween IgG subclasses with two disulfide bonds present in bothhuman IgG1 and IgG4. Unlike IgG1, however, the disulfide bondsin the IgG4 hinge can easily adopt an intra-H chain conforma-tion, facilitated by the 226-CPSCP-230 core-hinge sequence ofIgG4 (compared with 226-CPPCP-230 in IgG1) (2). As a result,

a fraction of IgG4 molecules exists as noncovalently associatedhalf molecules, which becomes apparent when analyzed by non-reducing SDS-PAGE (2–4). A single S228P mutation enhances thedisulfide linkage in the core-hinge of IgG4 and reduces the amountof half molecules (2–4).Another characteristic of the human IgG4 molecule is its ability

to exchange half molecules with other IgG4 molecules. This pro-cess, termed Fab-arm exchange, occurs naturally in vivo and canbe mimicked in vitro by the addition of mild reducing agents suchas reduced glutathione (GSH) (5, 6). As a consequence, blood-derived IgG4 is characterized by an inability to cross-link identicalAgs (homologous cross-linking), which has been referred to asfunctional monovalency and, as a result, is nonprecipitating (7).Furthermore, through Fab-arm exchange, natural IgG4 Abs ac-quire the ability to cross-link two different Ags (heterologouscross-linking), making them bispecific (5). More recent studiessuggest that IgG subclasses from other mammalian species sharethis ability (6, 8–10), making it a mechanism that could impacthumoral immunity more generally.Human IgG4 is classically considered nonactivating because of

its limited ability to bind C1q and FcRs, resulting in a poor abilityto induce complement and cell activation (11, 12). As such, IgG4is commonly chosen for immunotherapeutic applications where re-cruitment of the immune system’s effector functions is undesired(13, 14). On infusion in patients, however, unmodified (bivalent)IgG4 molecules engage in Fab-arm exchange with the patient’sendogenous IgG4 pool (15). Thus, in time, bivalent IgG4 thera-peutics convert into functionally monovalent IgG4 therapeutics,which could potentially affect both their initial binding strengthand cross-linking behavior. This kind of unpredictability is an

*Genmab, 3584 CM Utrecht, The Netherlands; †Department of Immunopathology,Sanquin Research–Landsteiner Laboratory, 1066 CX Amsterdam, The Netherlands;‡Biomolecular Mass Spectrometry and Proteomics Group, Department of Chemistry,Utrecht University, 3584 CH Utrecht, The Netherlands; xAlgonomics, 9052 Ghent,Belgium; and {Immunotherapy Laboratory, Department of Immunology, UniversityMedical Center, 3508 GA Utrecht, The Netherlands

1Current address: Mabmole, Ternat, Belgium.

2Current address: Complix, Ghent, Belgium.

Received for publication October 7, 2010. Accepted for publication July 9, 2011.

Address correspondence and reprint requests to Dr. Paul W. H. I. Parren, Genmab,Yalelaan 60, 3584 CM Utrecht, The Netherlands. E-mail address: p.parren@genmab.com

The online version of this article contains supplemental material.

Abbreviations used in this article: CH2, second constant H chain domain; CH3, thirdconstant H chain domain; CHO, Chinese hamster ovary; EGFR, epidermal growthfactor receptor; GSH, reduced glutathione; HEK, human embryo kidney; KD, disso-ciation constant; PBS-TB, PBS/0.05% Tween 20/1% BSA; PDB, Protein Data Bank;PK, pharmacokinetic; rhCh, rhesus monkey of Chinese origin; rhIn, rhesus monkeyof Indian origin.

Copyright� 2011 by TheAmericanAssociation of Immunologists, Inc. 0022-1767/11/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1003336

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

undesired trait for a therapeutic Ab, and eliminating the molecularfeatures enabling Fab-arm exchange should be considered whendesigning therapeutic IgG4 Abs.The ability to exchange half molecules has been attributed to the

IgG4 core-hinge sequence (226-CPSCP-230) in conjunction withdeterminants in the third constant H chain domain (CH3) (6). Inaccordance, mutating the core-hinge sequence (S228P) inhibitedFab-arm exchange to undetectable levels in vivo (8, 9, 15),whereas the molecular determinants in the CH3 enabling Fab-armexchange remain to be elucidated. In this article, we identify R409as the crucial CH3 residue governing Fab-arm exchange in humanIgG4. We furthermore confirm that the ability to engage in Fab-arm exchange is conserved in a number of other species, but thatthis does not correspond to a particular CH3 consensus sequence.We demonstrate that in IgG4 from rhesus monkeys (Macacamulatta) of Chinese origin, the 226-CPACP-230 core-hinge se-quence and the CH3 residue L405 are responsible. Moreover, weestablish that the ability to exchange Fab-arms is largely deter-mined by the strength of the noncovalent CH3–CH3 interaction.

Materials and MethodsCells

FreeStyle 293F (human embryo kidney [HEK]-293F) and FreeStyle Chi-nese hamster ovary (CHO)-S cells were cultured in FreeStyle 293 ex-pression medium and FreeStyle CHO expression medium, respectively(Invitrogen, Carlsbad, CA).

Commercial Abs

Purified human, pig, horse, bovine, sheep, rabbit, rat, goat, and guinea pigIgG were obtained from Jackson ImmunoResearch (West Grove, PA).Purified cynomolgus monkey and rhesus monkey IgG were purchased fromSera Laboratories International (Haywards Heath, U.K.).

Cloning and production of Abs and His-second constantH chain domain-CH3 constructs

Expression vectors for IgG1-epidermal growth factor receptor (EGFR),IgG4-EGFR, IgG4-S228P-EGFR, IgG1-CD20, and IgG4-CD20 have beendescribed previously (6, 15). For the expression of His-second constantH chain domain (CH2)-CH3(rhChG4), His-CH2-CH3(G1), and His-CH2-CH3(G4), constructs were designed that contained N-terminal poly-histidine (6xHis) tags followed by aa 230–447 (EU numbering conventionsare used throughout the article) of rhesus IgG4, human IgG1, and humanIgG4, respectively. All His-CH2-CH3 constructs were codon optimized,synthesized de novo by Geneart (Regensburg, Germany), and cloned intopEE12.4 (Lonza, Slough, U.K.). A Quikchange site-directed mutagenesiskit (Stratagene, La Jolla, CA) was used to introduce the Q355R, R409K,E419Q, L445P, K370E, R409M, K370Q, R409L, S364D, R409A, L351K,Y349D, F405A, F405L, R409W, E357T, R409T, and T411V mutations inIgG4 using pTomG42F8, pTomG47D8, and pEE12.4His-CH2-CH3-G4 astemplates. In the same way, the R355Q, K409R, Q419E, P445L, T350I,K370T, F405L, and T350I/K370T/F405L mutations were introduced inIgG1 using pConG1f2F8, pConG1f7D8, pG1f2F8CPSC, pG1f7D8CPSC,and pEE12.4His-CH2-CH3-G1 as templates. Rhesus IgG4 constructs werebased on the sequences described by Scinicariello et al. (16) and synthe-sized de novo by Geneart AG.

All Abs were produced under serum-free conditions (FreeStyle medium)by cotransfecting relevant H and L chain expression vectors in HEK-293Fcells, using 293fectin (Invitrogen), or CHO-S cells, using FreeStyle MAXReagent (Invitrogen), both according to the manufacturer’s instructions. AllHis-CH2-CH3 constructs were produced by transfecting relevant expres-sion vectors in HEK-293F cells as described earlier.

Abs were purified by protein A affinity chromatography (rProtein A FF;GE Healthcare, Uppsala, Sweden), dialyzed overnight to PBS, and filtered-sterilized over 0.2-mM dead-end filters. Concentration of purified IgGs wasdetermined by nephelometry and absorbance at 280 nm. Purified proteinswere analyzed by SDS-PAGE, mass spectrometry, and glycoanalysis.Batches of IgG were tested by size-exclusion chromatography and shownto be at least 94% monomeric. Endotoxin levels of batches used in vivowere ,0.1 endotoxin units/mg IgG. The His-CH2-CH3 proteins werepurified by immobilized metal ion (Ni2+) affinity chromatography(Macherey-Nagel, Duren, Germany), desalted using PD-10 columns (GEHealthcare) equilibrated with PBS, and filtered-sterilized over 0.2-mM

dead-end filters. The concentration of the purified proteins was determinedby absorbance at 280 nm. The quality of the purified proteins was analyzedby SDS-PAGE.

GSH-mediated Fab-arm exchange in vitro

As described previously (6), combinations of Abs were mixed and in-cubated with GSH (Sigma, St. Louis, MO) at a final concentration of 4 mg/ml per Ab. The final concentration of GSH was 0.5 mM. The mixtureswere incubated at 37˚C for 24 h, and samples were drawn in PBS-TB(PBS/0.05% Tween 20/1% BSA), in which (bi)specific IgG concen-trations were measured. For inhibition experiments, Ab mixtures weresupplemented with 160 mg (200-fold excess) of purified animal IgG eitherduring or after the 24-h incubation at 37˚C. All purified animal IgG werefirst dialyzed against PBS to remove additives and preservatives.

Fab-arm exchange in vivo

Female SCID mice (6- to 8-wk-old) were obtained from Charles RiverLaboratories (Maastricht, The Netherlands) and housed in a barrier unit ofthe Central Laboratory Animal Facility (Utrecht, The Netherlands). Themice were kept in filter-top cages with water and food provided ad libitum.All experiments were approved by the Utrecht University animal ethicscommittee. Mixtures of Abs (600 mg in total per mouse) were administeredto mice (n = 4) and blood samples were drawn from the saphenous veinat 3, 24, 48, and 72 h after administration. Blood was collected in hepa-rin-containing vials, which were kept on ice, and centrifuged (5 min at10,000 3 g) to separate the plasma from cells. Plasma was transferred to anew vial and stored at 220˚C for determination of bispecific Ab levels.

Binding assay for the detection of CD20/EGFR bispecific Abs

The presence of CD20/EGFR bispecific Abs was determined using asandwich ELISA as described previously (6). In short, ELISA plates(Greiner bio-one, Frickenhausen, Germany) were coated overnight with 2mg/ml recombinant EGFR (extracellular domain) in PBS at 4˚C. The plateswere washed and incubated with serial diluted plasma samples or purifiedAb mixtures (in PBS-TB) for 90 min at 20˚C under shaking conditions(300 rpm). Next, the plates were washed and incubated with 2 mg/mlmouse anti-idiotype mAb 2F2 SAB1.1 (directed against human mAb-CD20; Genmab) diluted in PBS-TB for 75 min at 20˚C. Bound bispe-cific Abs were detected with HRP-labeled goat anti-mouse IgG (JacksonImmunoResearch) and ABTS substrate (Roche Diagnostics). The colordevelopment reaction was stopped by addition of an equal volume ofoxalic acid (Riedel de Haen, Buchs, Germany), and absorbance wasmeasured at 405 nm. Bispecific Abs in plasma samples were quantified bynonlinear regression curve fitting (GraphPad Software, San Diego, CA)using an in vitro exchanged Ab mixture as reference (with the assumptionthat the maximal expected concentration of bispecific IgG4 was 50% oftotal IgG4 concentration).

Total human IgG ELISA

The total amount of human Abs in the mice was determined by sandwichELISA. In short, ELISA plates were coated overnight with 2 mg/ml mouseanti-human IgG (MH16-1; Sanquin) in PBS at 4˚C. The plates were sub-sequently washed and blocked with PBS-C (PBS/2% normal chicken se-rum; Life Technologies) for 1 h at 37˚C. Next, the plates were washed andincubated with diluted plasma samples in PBS/0.05% Tween 20/2% nor-mal chicken serum for 120 min at 20˚C under shaking conditions (300rpm). Bound Abs were detected by HRP-labeled goat anti-human IgG(Jackson ImmunoResearch) and ABTS substrate (Roche Diagnostics). Thecolor development reaction was stopped by addition of an equal volume ofoxalic acid, and absorbance was measured at 405 nm. IgG was quantifiedby nonlinear regression curve fitting (GraphPad Software) using the in-jection mixtures as reference.

Determination of dissociation constant by native massspectrometry

Apparent dissociation constant (KD) values were determined by nano-electrospray ionization mass spectrometry, as described elsewhere (17).In brief, purified proteins (at 50 mM stock concentrations) were bufferexchanged into 100 mM NH4Ac (pH 6.9), serial diluted to concentrationsranging from 25 nM to 20 mM (monomer equivalent), and analyzed bynano-electrospray ionization mass spectrometry. Effect of time betweenanalysis and sample dilution was investigated and did not significantlyinfluence the results. Signals corresponding to the monomeric (MS) anddimeric (DS) states were integrated, and the relative proportion of eachstate at each concentration of total protein in monomer equivalents ([M]0)was determined using the following equations: [M]eq = MS/(MS + DS)×[M]0

The Journal of Immunology 3239

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

(concentration monomer at equilibrium), and [D]eq = ([M]0 2 [M]eq)/2(concentration dimer at equilibrium). The apparent KD was subsequentlyextracted by plotting the [D]eq against [M]eq

2 values of each concentrationand determining the gradient by least squares linear regression (MicrosoftOffice Excel 2007 software; Microsoft, Redmond, WA).

Modeling in silico

Crystal structures were studied using the Brugel modeling package (18).To propose mutations that would lead to a desired stabilization (ordestabilization) of IgG4, we used a quantitative structure-based scoringmethodology (19). In brief, each position in the CH3–CH3 dimer interfacewas subjected to in silico mutagenesis to all natural amino acids, exceptcysteine and proline. Subsequent to in silico mutagenesis, exploration ofthe conformational space was obtained by interdependent optimization ofthe side chains of all residues located in a sphere of 12 A of the mutatedresidue, using the FASTER algorithm (20), whereby the side chain underinvestigation was kept fixed at a macrorotameric state. Subsequently, oneach macrorotameric state thus obtained, 300 steps of steepest descentminimization was carried out, and finally a scoring function for the sidechain under investigation was evaluated, as described previously (19).Finally, per position in the CH3–CH3 dimer interface, the highest scoresfor each mutation were compared, and visual inspection of the resultingconformation was carried out in selected cases.

Statistical analysis

Data analysis was performed using GraphPad Prism for Windows, version4.03 (GraphPad). Data sets were compared by using one-tailed pairedStudent t tests. Statistical significance was accepted when p , 0.05.

ResultsIdentification of the CH3 residue mediating Fab-arm exchangein humans

IgG4 Fab-arm exchange was previously shown to depend on theCPSCP sequence in the core-hinge combined with the IgG4 CH3

domain (6). To identify the specific amino acid residues enabling

Fab-arm exchange, we aligned the IgG4 CH3 sequence with that

of IgG1 (which does not support Fab-arm exchange) and identified

sequence differences (Fig. 1A). The whole CH3 domain of IgG1 or

individual IgG1-specific CH3 residues (excluding those resulting

from allelic variation) were introduced into IgG4 by mutagenesis.

Vice versa, the CH3 domain of IgG4 or individual IgG4-specific

CH3 residues were introduced into IgG1 (in combination with the

P228S mutation to mimic the IgG4 core-hinge). The mutations

were introduced into fully human mAbs 2F8 (21), directed against

EGFR, and 7D8 (22), directed against the CD20 Ag. The resulting

proteins (listed in Fig. 1A), were compared with wild-type IgG4 in

their ability to support Fab-arm exchange (Fig. 1B, 1C). Swapping

CH3 domains between IgG1 and IgG4 activated Fab-arm ex-

change for IgG1-P228S and abrogated the activity for IgG4. In-

troduction of individual subclass-specific point mutations showed

that residues at positions 355, 419, and 445 in the CH3 domains of

both IgG1 and IgG4 had no influence on Fab-arm exchange ac-

tivity. In contrast, mutations in the 409 position abrogated or en-

abled Fab-arm exchange activity in IgG4 and IgG1-P228S,

respectively. Thus, residue R409 in the CH3 domain is crucial for

Fab-arm exchange in vitro.To investigate whether the effects observed in vitro also apply to

in vivo conditions, we injected mixtures of IgG4-CD20 with IgG1-

EGFR, IgG4-EGFR, IgG4-S228P-EGFR, IgG4-R409K-EGFR, or

IgG1-P228S-K409R-EGFR (in a 5:1 molar ratio) into mice. Blood

samples were drawn at 3, 24, 48, and 72 h after injection, and

bispecific Abs were quantified using in vitro exchanged 5:1 mix-

tures (IgG4-CD20/IgG4-EGFR) as reference standard (Fig. 1D).

FIGURE 1. Residue R409 in the CH3 enables Fab-arm exchange in humans. A, Selected sequences of the human IgG1 and IgG4 subclasses, including

IgG1-specific allotypes located in the CH3, and a list of human IgG4 and IgG1 mutants in which either the entire CH3 was swapped or individual subclass-

specific CH3 residues were exchanged. B, Equimolar mixtures of identical IgG4-CD20 and IgG4-EGFR mutants or (C) IgG1-CD20 and IgG1-EGFR

mutants incubated in the presence of 0.5 mM GSH at 37˚C. The generation of bispecific Abs was followed over time by ELISA and expressed as percentage

of control (wild-type IgG4 at 24 h). Data represent mean 6 SEM of at least three separate experiments. D, Groups (n = 4) of SCID mice injected with Ab

mixtures of IgG4-CD20 and IgG4-EGFR mutants or IgG1-EGFR mutants in a 5:1 ratio (600 mg in total). The generation of bispecific Abs and the total

amount of Abs was followed over time and quantified by ELISA. Samples were measured at least twice in separate experiments. Exchange was expressed as

percentage of total Ab concentration. Data represent mean 6 SEM values of four mice from one example experiment. The detection limit of the assays is

indicated (dotted line) and represents serum levels of 750 ng/ml.

3240 ROLE OF THE CH3 DOMAIN IN THE MECHANISM OF FAB-ARM EXCHANGE

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Bispecific Abs appeared in the blood of mice injected with IgG4-EGFR or IgG1-P228S-K409R-EGFR, whereas samples containingIgG1-EGFR, IgG4-S228P-EGFR, or IgG4-R409K-EGFR testednegative. These results confirm the importance of residues S228and R409 for Fab-arm exchange.

In silico model for IgG4 CH3–CH3 interaction

In human IgG1, the noncovalent interaction between the CH3domains involves 16 residues located on 4 antiparallel b-strandsthat make intermolecular contacts (23). Alanine scanning muta-genesis showed that stabilization of the IgG1 CH3–CH3 in-teraction was largely mediated by five of these residues, includingK409 (24). To get a better understanding of the role of K409 in theIgG1 CH3–CH3 interaction, we studied the 1.65-A resolutioncrystal structure of IgG1 (Protein Data Bank [PDB] code 1L6X)(25) in more detail. In this structure (Fig. 2A), the volume of theK409 side chain is accommodated by a cavity that is bordered byL368’, K370’, D399’, F405’, and Y407’ in the opposite CH3domain. K409 forms hydrogen (H-) bonds with D399’ and a bur-ied water molecule, which prevents an electrostatic clash betweenK409 and K370’ in the opposite CH3 domain. The buried watermolecule also forms H-bonds with S364 and T411.Because the 3.15-A resolution of the only available IgG4

structure (PDB code 1ADQ) (26) is too low for direct comparison,the K409R substitution was modeled in the 1L6X structure byoptimizing side-chain conformations of the arginine residue andits surrounding residues, using the FASTER algorithm (20). Theresulting model (Fig. 2B), confirmed the R409 and K370’ side-chain conformations observed in structure 1ADQ (26) and sug-gested that the R409 guanidinium group maintains a favorableelectrostatic interaction with the D399’ carboxylate side chain, asis also the case for the K409 ammonium group in IgG1. Contraryto the latter, however, the R409 guanidinium group in IgG4 takesup the space of the water molecule observed in high-resolutionIgG1 structures, thereby negating the H-bonds with S364, T411,and K370’, and causing an electrostatic clash between R409 andK370’.

Fab-arm exchange is conserved in other species but cannot beexplained by the 409 position

Homologs of the human IgG4 subclass have been identified inother species, including several nonhuman primates often used tomodel human disease (16, 27, 28). Furthermore, previous studiessuggest that Fab-arm exchange occurs in rhesus monkeys (Macacamulatta) (6), cynomolgus monkeys (Macaca fascicularis) (8, 9),

and rabbits (Oryctolagus cuniculus) (10). In addition, non-precipitating Abs have been reported in guinea pig (Caviaporcellus), rat (Rattus norvegicus), horse (Equus caballus), sheep(Ovis aries), cow (Bos taurus), donkey (Equus asinus), mouse(Mus musculus), goat (Capra hircus), and pig (Sus scrofa) (29,30). To investigate whether other species contain IgG subclassesthat engage in Fab-arm exchange, purified IgG preparations fromdifferent species were assessed for their ability to interfere within vitro exchange of an equimolar mixture of human IgG4-EGFRand IgG4-CD20 (Fig. 3). We found that purified monkey, pig,horse, rabbit, and rat IgG was able to significantly inhibit theformation of bispecific Abs to varying degrees, suggesting theability to undergo Fab-arm exchange. Of all tested species, cyn-omolgus and rhesus monkey IgG preparations were most efficient,inhibiting the formation of bispecific Abs by .50% when addedin 200-fold excess. Mouse IgG was not included in these experi-ments because it interfered with detection of bispecific Abs, forwhich an anti-mouse conjugate was used as secondary Ab. Ina different experimental setup, however, we did confirm thatmouse IgG3 engages in Fab-arm exchange with human IgG4 (datanot shown), as previously described by Lewis et al. (8).In an attempt to correlate the ability to engage in Fab-arm ex-

change to specific CH3 residues, available sequences of IgG sub-classes from multiple species were aligned (Supplemental Table I).Examination of the 96 CH3 sequences, however, revealed noobvious consensus sequence relating to Fab-arm exchange withinthe 16 aa implicated in the CH3–CH3 interface. R409 was ob-served only in the CH3 of human IgG4 (and one human IgG3allotype) and sheep IgG1 and IgG2. The combination of S228 inthe core-hinge and R409 in the CH3 was specific for human IgG4.Thus, the ability to engage in Fab-arm exchange appears to beconserved in several species, but presumably regulated by differ-ent molecular determinants.

Identification of the molecular determinants mediatingFab-arm exchange in rhesus monkeys (Macaca mulatta)

To directly investigate Fab-arm exchange in rhesus monkeys, theV regions of both 2F8 and 7D8 were cloned into rhesus IgG4backbones corresponding to polymorphisms found in rhesusmonkeys of Chinese (rhCh) and Indian (rhIn) origin (16). Becausethe available sequences (Fig. 4) were incomplete, we supple-mented the H chain with the appropriate C-terminal amino acidresidues derived from the rhesus macaque genome sequence (31).Furthermore, because the rhesus IgG4 H chain unexpectedly didnot appear to contain a cysteine for L chain pairing, variants

FIGURE 2. In silico model for the IgG4 CH3–CH3 interface. A, The IgG1 CH3–CH3 interface centered around residue K409 as resolved in the 1.65-A

resolution crystal structure (PDB code 1L6X) (25) and (B) an in silico model of the IgG4 CH3–CH3 interface centered around residue R409, based on the

same structure. Only polar H atoms (white) of relevant side chains and relevant H-bonds (black dashed lines) are depicted. The H2O molecule depicted is

typically present in crystal structures at resolutions,2.8 A [e.g., PDB codes 1L63 (1.65 A) (25), 1H3T (2.4 A) (37), 1H3U (2.4 A) (37), 1H3X (2.44 A) (37),

and 1HZH (2.7 A) (38)], but not observed in crystal structures of poorer resolution [e.g., 1E4K (3.2 A) (39), 1FC1 (3.5 A) (23), 1FC2 (2.8 A) (23), 1H3V (3.1

A) [37], 1H3W (2.85 A) (37), 1H3Y (4.1 A) (37)]. Structural images were generated with PyMOL software (DeLano Scientific, South San Francisco, CA).

The Journal of Immunology 3241

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

containing a cysteine (at the position equivalent to human IgG4)were constructed to address potential H–L chain mispairing (Fig.4). The resulting proteins, rhChIgG4, rhChIgG4-S131C, andrhInIgG4-S131C, were expressed and analyzed by nonreducingSDS-PAGE (Fig. 5A). The rhChIgG4 molecules existed as a mix-ture with major species of ∼150, ∼90, and ∼44 kDa as a result ofheterogeneous disulfide linkage between H and L chains (Fig. 5A).The S131C mutation promoted the disulfides linking the H chainsto the L chains, as demonstrated by one major protein species of∼150 kDa. As expected (15), the rhInIgG4-S131C Abs, containingthe CPPCP core-hinge sequence, contained stabilized inter-Hchain disulfides as demonstrated by the low abundance of halfmolecules compared with the rhChIgG4-S131C Abs (containinga CPACP core-hinge sequence). Furthermore, the rhesus IgG4molecules were compared with human IgG4 in their ability tosupport Fab-arm exchange (Fig. 5B). Both IgG4 molecules basedon rhCh enabled Fab-arm exchange, whereas IgG4 based on rhIndid not. Furthermore, rhChIgG4 did engage in Fab-arm exchange,despite the presence of protein species containing an atypicaldisulfide linkage between the two half molecules including aninter-L chain disulfide bond next to two inter-H chain bonds.

To identify the specific CH3 aa residues enabling Fab-arm ex-change, we aligned the amino acid residues involved in CH3–CH3interface of human and rhesus IgG (Fig. 5C). The three rhChIgG4-specific amino acids in the CH3 domain were introduced intohuman IgG1-P228S-EGFR and IgG1-P228S-CD20, as individualmutations or in combination, and assessed for their ability tosupport Fab-arm exchange (Fig. 5D). Whereas the T350I andK370T substitutions had no effect, the F405L substitution alone orthe combination of all three mutations did enable Fab-arm ex-change in an IgG1-P228S background.Together, these results show that the CH3 of rhesus IgG4 en-

ables Fab-arm exchange (in the context of A228 in the core-hinge),and that residue L405 mediates this ability (despite the pres-ence of K409).

Probing and quantifying the CH3–CH3 interaction

Because the mechanism of Fab-arm exchange involves the dis-sociation and reassociation of CH3 domains, we hypothesized thatthe CH3–CH3 interaction strength would correlate with the abilityto exchange half molecules. To investigate this correlation, wedesigned his-tagged constructs, lacking the hinge region to preventcovalent inter-H chain disulfide bonds, based on the Fc domains ofrhChIgG4, human IgG1, and human IgG4 (termed His-CH2-CH3[rhChG4], His-CH2-CH3[G1], and His-CH2-CH3[G4], respec-tively). Subsequently, variants of these constructs containing themutations in the CH3–CH3 interface described earlier were gen-erated by site-directed mutagenesis. Furthermore, additional mu-tations were designed based on the K409R model (Fig. 2B,Supplemental Fig. 1) and introduced into the CH3 domains ofIgG4-EGFR, IgG4-CD20, and the His-CH2-CH3(G4) construct.The IgG4 variants were mixed pairwise and compared with wild-type IgG4 in their ability to support Fab-arm exchange (Fig. 6A,6B). The resulting His-CH2-CH3 proteins listed in Table I wereused to measure the strength of the noncovalent CH3–CH3interactions by determining the apparent KD by native massspectrometry as described elsewhere (17) and ranked according tointeraction strength (Table I).CH3 mutations affected interdomain KDs and Fab-arm ex-

change differentially and could be divided into four groups.First, the R409A and S364D mutations only minimally affectedCH3–CH3 binding affinity and did not impact Fab-arm ex-change. Second, the R409K, R409L, R409M, K370E, and K370Q

FIGURE 4. The amino acid se-

quences of the constant domains of

the (A) H chain and (B) k L chain, as

described by Scinicariello et al. (16),

were used to construct rhesus mon-

key IgG4 expression vectors. Allo-

typic differences were incorporated

in the constructs based on highest

residue frequency (red highlighted

residues). The IgG4 sequences were

supplemented (blue highlighted amino

acids), where amino acid residues

were missing (C terminus based on

M. mulatta whole-genome sequence

NW_001199565.1) or to asses the

atypical absence of the cysteine in-

volved in H–L pairing (C131 in-

volved in disulfide linkage between

H and L chains in human IgG4, as

opposed to S131 reported for rhesus

IgG4). EU numbering conventions

are used for annotation.

FIGURE 3. IgG from different species engage in Fab-arm exchange.

Equimolar mixtures of IgG4-CD20 and IgG4-EGFR were incubated for

24 h (37˚C, 0.5 mM GSH) in the presence of a 200-fold excess (black bars)

of purified IgG from different species. As control, a 200-fold excess of

purified IgG from different species was added after, instead of during, the

exchange reaction (white bars). The generation of bispecific Abs was

followed over time by ELISA and expressed as percentage exchange by

human IgG4 only. Data represent mean 6 SEM of at least three sep-

arate experiments. Statistical significance was assessed by paired t test

(*p , 0.05, **p , 0.005, ***p , 0.0005).

3242 ROLE OF THE CH3 DOMAIN IN THE MECHANISM OF FAB-ARM EXCHANGE

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

mutations decreased the apparent KD of the CH3–CH3 interaction(3.2- to 24-fold in the His-CH2-CH3[G4] constructs) and blockedFab-arm exchange to undetectable levels. The third group ofmutations, consisting of L351K, E357T, Y394D, F405A, F405L,R409T, and R409W, all increased the apparent KD (16.7–833.3-

fold) and influenced both maximal exchange levels and exchangekinetics. Thus, the maximal observed exchange was alreadyreached after 6 h, followed by a reduction in the levels of bispe-cific reactivity. The fourth group consists of IgG1-based mole-cules, where all mutations, except T350I, increased the apparent

FIGURE 6. The ability to engage in Fab-arm

exchange is correlated to the CH3–CH3 interaction

strength. A and B, Equimolar mixtures of identical

IgG4-CD20 and IgG4-EGFR mutants incubated in

the presence of 0.5 mM GSH at 37˚C. The genera-

tion of bispecific Abs was followed over time by

ELISA and expressed as percentage of control

(IgG4 at 24 h). Data represent mean 6 SEM of at

least three separate experiments. Relation between

apparent KD (Table I) and bispecific Ab generation

after 6 h of (C) IgG4-derived constructs (Fig. 5A,

5B) or (D) IgG1-CPSC–derived constructs (Fig. 4B,

4D). Vertical dotted line indicates the total Ab

concentration used in the exchange reaction (5.3 31028 M).

FIGURE 5. Rhesus monkey IgG4 can mediate Fab-arm exchange. A, Rhesus monkey IgG4 formats of mAbs directed to EGFR and CD20 (rhChIgG4,

rhChIgG4-S131C, and rhInIgG4-S131C) analyzed on nonreducing SDS-polyacrylamide gel. The molecular sizes of intact Abs (H2L2), H chain dimers (H2),

half molecules (HL), L chain dimers (L2), and H and L chains are indicated. Possible conformations of disulfide linkages (red lines) between H chains

(black lines) and L chains (gray lines) are depicted as cartoons. One representative experiment out of two is shown. B, Equimolar mixtures of identical

rhesus monkey-derived IgG4-CD20 and IgG4-EGFR molecules, or (D) equimolar mixtures of IgG1-P228S-CD20 and IgG1-P228S-EGFR mutants, in-

cubated in the presence of 0.5 mM GSH at 37˚C. The generation of bispecific Abs was followed over time by ELISA and expressed as percentage of control

(human IgG4 at 24 h). Data represent mean6 SEM of at least three separate experiments. C, Sequence alignment of relevant amino acid residues of human

and rhesus monkey IgG subclasses.

The Journal of Immunology 3243

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

KD, in which increases of .15-fold (K409R, F405L, and ITL)enabled Fab-arm exchange. Together, these data suggest that ef-ficient Fab-arm exchange can occur only if the CH3–CH3 in-teraction is sufficiently weak. Increasingly weaker interactions,however, also make the generation of bispecifics less efficient(Figs. 1C, 5D, 6C, 6D).

DiscussionThe ability of human IgG4 to engage in Fab-arm exchange wasshown to depend on the IgG4 core-hinge sequence, 226-CPSCP-230, in combination with determinants in the CH3 domain (6). Inthis study, we identified residue R409 as the crucial CH3 de-terminant enabling Fab-arm exchange in human IgG4 (comparedwith K409 in the other human IgG subclasses). This can be ra-tionalized because R409 is the only IgG4-specific residue directlyinvolved in the noncovalent CH3–CH3 interaction. It is, however,striking that what appears to be a rather conservative K to Rmutation has such a profound impact, although large effects of Kto R substitutions on protein–protein interactions are not withoutprecedence (32, 33). To explain the dramatic effect of this rela-tively conservative substitution, we used in silico modeling. Themodel suggests that because of the larger size of the R409 sidechain, a buried water molecule is displaced that in IgG1 is in-volved in the stabilization of the CH3–CH3 interaction by formingfour short H-bonds with S364, K409, T411, and K370’, and pre-venting electrostatic clashes between K409 and K370’ (Fig. 3).Previous studies suggested that IgG from rhesus monkeys

(Macaca mulatta) (6), cynomolgus monkeys (Macaca fas-cicularis) (8, 9), and IgG3 from mice (Mus musculus) (8) couldengage in Fab-arm exchange with human IgG4. In accordance, wenow demonstrate that a variety of other species share this ability.The observed variation in efficiency (Fig. 2) could be explained bydifferences in the following factors: 1) interspecies compatibility,2) relative concentrations of permissive subclasses, and 3) stabi-lity of the hinge disulfide bonds. A more detailed analysis withrecombinant rhesus monkey IgG4 constructs identified the 226-

CPACP-230 core-hinge sequence in combination with residueL405 in the CH3 domain as the molecular determinants enablingFab-arm exchange in this species. Thus, these data, together withsequence analysis of available CH3 sequences of different species,suggest that ability to engage in Fab-arm exchange is regulated byspecies-specific molecular determinants.The redistribution of Ag-binding domains by Fab-arm exchange

results in bispecific IgG4 molecules that, in the polyclonal setting,are functionally monovalent. We have previously proposed thatthe resulting inability of IgG4 to cross-link homologous Ags con-tributes to the anti-inflammatory nature attributed to human IgG4by interfering with immune complex formation (6). In this way, thefunctional monovalency of IgG4, typically induced on chronic Agstimulation, may act as a natural buffer and dampen inflammatoryreactions induced by other Ab subclasses. Because Fab-arm ex-change is also observed in other species and because it is mediatedby species-specific molecular determinants, one might speculatethat this anti-inflammatory mechanism represents an evolutionaryadvantage. The fact that polymorphisms that block Fab-arm ex-change exist in both humans (K/R409) and rhesus monkeys (P/A228), however, also suggests that there has been selective pres-sure benefiting the preservation of bivalent IgG4 molecules andraises the possibility that such polymorphisms might also exist inother species, including those used to model human diseases.Although evaluation of available CH3 sequences did not reveal

specific common residues enabling Fab-arm exchange, all aminoacids implicated are located at the CH3–CH3 interface. In ac-cordance, we established a relation between the noncovalentCH3–CH3 interaction strength and the ability to exchange Fab-arms, consistent with a model in which strong interactions, such asfound between IgG1 CH3 domains, block Fab-arm exchange bypreventing the dissociation of half molecules. Weaker interactions,such as found between IgG4 CH3 domains, then allow the effi-cient exchange of Fab-arms. This interpretation is supported bya mechanistic study that showed dissociation of the CH3 domainsto be the rate-determining step of IgG4 Fab-arm exchange (34).Increasingly weaker interactions, however, make the generation ofbispecifics less efficient by precluding effective (re)association.The ∼10-fold difference between the apparent KDs of IgG4 andIgG4-R409K/IgG1 CH3 domains reported in this study (and thusthe difference between allowing and inhibiting Fab-arm exchange)may be an underestimation, because the same mass spectrometrictechnique using IgG4Dhinge constructs reported similar IgG4apparent KDs, but ∼100-fold stronger apparent KDs for IgG4-R409K (17), which is consistent with the previously reportedIgG1 CH3 interaction strength of ,10210 M as determined bysize-exclusion chromatography (35).Consistent with a previously reported alanine scan of the IgG1

CH3 interface (24), our results show that mutating residues K409and F405 considerably affect the KD of the CH3–CH3 interaction,ranging from 3.0 3 1029 to 3.4 3 1025 M (Table I). Substitutionof K370 with alanine had a significant destabilizing effect on CH3stability in IgG1 (24), but conversely, the same mutation increasedthe CH3–CH3 interaction in IgG4 (17), consistent with our pro-posed model (Fig. 2B) where a K370A substitution would relievethe electrostatic strain with R409’ in the partner CH3 domain.Likewise, the K370E and K370Q substitutions described in thisarticle stabilize the CH3–CH3 interaction in IgG4 and are thoughtto act in a similar way (see Supplemental Fig. 1 for a detaileddiscussion of all mutations).The ability of therapeutic IgG4 Abs to recombine with endog-

enous IgG4 in vivo may affect their pharmacokinetic (PK) andpharmacodynamic properties. Although PKmodeling suggests thatunder normal conditions this will unlikely lead to biological effects,

Table I. CH3–CH3 apparent KD of His-CH2-CH3 constructs based onrhesus IgG4 and human IgG1 and IgG4

His-CH2-CH3Fold Differencea

Construct KD (M) R2 G1 G4

G1-T350I 1.0 3 1029 0.812 3G4-K370E 2.0 3 1029 0.8058 24G4-R409M 3.0 3 1029 0.7968 16G1 3.0 3 1029 0.992 1G1-K370T 5.0 3 1029 0.88 0.6G4-R409K 8.0 3 1029 0.9669 6G4-K370Q 1.1 3 1028 0.994 4.3G4-R409L 1.5 3 1028 0.9513 3.2G4-S364D 4.7 3 1028 0.9882 1.02G4 4.8 3 1028 0.9755 1rhChG4 1.1 3 1027 0.987 NA NAG1-K409R 1.1 3 1027 0.974 0.03G4-R409A 1.6 3 1027 0.9677 0.3G4-R409T 5.8 3 1027 0.896 0.08G4-L351K 7.4 3 1027 0.9171 0.06G1-F405L 8.5 3 1027 0.967 0.0035G1-ITL 1.2 3 1026 0.973 0.0025G4-Y349D 9.9 3 1026 0.9747 0.0048G4-F405A 1.9 3 1025 0.9976 0.0025G4-F405L 2.5 3 1025 0.941 0.0019G4-R409W 3.4 3 1025 0.9993 0.0014G4-E357T 4.1 3 1025 0.9773 0.0012

aCompared with wild-type CH2-CH3(G1) or CH2-CH3(G4) molecules.

3244 ROLE OF THE CH3 DOMAIN IN THE MECHANISM OF FAB-ARM EXCHANGE

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

it cannot be excluded (15). Thus, Fab-arm exchange introducesunpredictability when using wild-type IgG4-based therapeutics,which is undesired for human immunotherapy. Therefore, muta-tions that completely prevent Fab-arm exchange in vivo should beconsidered when designing therapeutic IgG4 Abs. We and othershave previously shown that a S228P mutation in the IgG4 core-hinge represents a tool for preventing Fab-arm exchange to lessthan detection limits (8, 9, 15). The mutations in the CH3 domainidentified in this article, R409K, R409L, R409M, K370E, andK370Q, may be used for additional stabilization or as an alter-native approach to hinge stabilization. The R409K mutation is ofspecial interest for stabilizing therapeutic IgG4 Abs because thismutation represents a naturally occurring allotype of IgG4 (36)and is thus least likely to increase immunogenicity in a therapeuticsetting. The existence of the IgG4-K409 allotype also suggeststhat some individuals may possess IgG4 species that do not en-gage in Fab-arm exchange, remain (monospecific) bivalent, andwill not recombine with IgG4-based therapeutics, making generalPK/pharmacodynamic predictions even more difficult.In summary, we identified residue R409 as the CH3 determinant

responsible for IgG4 Fab-arm exchange in humans (together withthe previously identified S228 in the core-hinge). In rhesus monkey,the ability to engage in Fab-arm exchange was mediated by res-idues A228 and L405 in the IgG4 core-hinge and CH3, re-spectively. We could not detect obvious consensus CH3 sequencespredicting Fab-arm exchange in IgG from different species. Wecould, however, demonstrate a relation between the CH3–CH3interaction strength and the ability to engage in Fab-arm ex-change. The redistribution of Ag-binding domains between mol-ecules may constitute a general immunological and evolutionaryadvantage, impacting our view of humoral immunity, the in-terpretation of Ab-based studies, and the design of Ab-basedtherapeutics.

AcknowledgmentsWe thank M. Roza, S. Achoyan, L. Wiegman, J. van den Brakel, and

A. Ortiz Buijsse for expert technical assistance, W. Bleeker and G. Perdok

for advice, and members of the SPARREN team for helpful discussions.

DisclosuresA.F.L., J.M., J.N., E.T.J.v.d.B., T.V., J.G.J.v.d.W., J.S., and P.W.H.I.P.

have stock and/or warrants in Genmab. T.S., R.C.A., S.L., I.L., R.J.R.,

and A.J.R.H. received consulting fees and/or were awarded Genmab-

sponsored research funding.

References1. Saphire, E. O., R. L. Stanfield, M. D. Crispin, P. W. Parren, P. M. Rudd,

R. A. Dwek, D. R. Burton, and I. A. Wilson. 2002. Contrasting IgG structuresreveal extreme asymmetry and flexibility. J. Mol. Biol. 319: 9–18.

2. Bloom, J. W., M. S. Madanat, D. Marriott, T. Wong, and S. Y. Chan. 1997.Intrachain disulfide bond in the core hinge region of human IgG4. Protein Sci. 6:407–415.

3. Angal, S., D. J. King, M. W. Bodmer, A. Turner, A. D. Lawson, G. Roberts,B. Pedley, and J. R. Adair. 1993. A single amino acid substitution abolishes theheterogeneity of chimeric mouse/human (IgG4) antibody. Mol. Immunol. 30:105–108.

4. Schuurman, J., G. J. Perdok, A. D. Gorter, and R. C. Aalberse. 2001. The inter-heavy chain disulfide bonds of IgG4 are in equilibrium with intra-chain disulfidebonds. Mol. Immunol. 38: 1–8.

5. Schuurman, J., R. Van Ree, G. J. Perdok, H. R. Van Doorn, K. Y. Tan, andR. C. Aalberse. 1999. Normal human immunoglobulin G4 is bispecific: it hastwo different antigen-combining sites. Immunology 97: 693–698.

6. van der Neut Kolfschoten, M., J. Schuurman, M. Losen, W. K. Bleeker,P. Martınez-Martınez, E. Vermeulen, T. H. den Bleker, L. Wiegman, T. Vink,L. A. Aarden, et al. 2007. Anti-inflammatory activity of human IgG4 antibodiesby dynamic Fab arm exchange. Science 317: 1554–1557.

7. van der Zee, J. S., P. van Swieten, and R. C. Aalberse. 1986. Serologic aspects ofIgG4 antibodies. II. IgG4 antibodies form small, nonprecipitating immunecomplexes due to functional monovalency. J. Immunol. 137: 3566–3571.

8. Lewis, K. B., B. Meengs, K. Bondensgaard, L. Chin, S. D. Hughes, B. Kjaer,S. Lund, and L. Wang. 2009. Comparison of the ability of wild type and sta-bilized human IgG(4) to undergo Fab arm exchange with endogenous IgG(4)in vitro and in vivo. Mol. Immunol. 46: 3488–3494.

9. Stubenrauch, K. G., U. Wessels, J. T. Regula, H. Kettenberger, J. Schleypen, andU. Kohnert. 2010. Impact of molecular processing in the hinge region of ther-apeutic IgG4 antibodies on disposition profiles in cynomolgus monkeys. DrugMetab. Dispos. 38: 84–91.

10. Wang, W., R. Xu, and J. Li. 2010. Production of native bispecific antibodies inrabbits. PLoS ONE 5: e10879.

11. Jefferis, R. 1990. Molecular structure of human IgG subclasses. In The HumanIgG Subclasses. F. Shakib, ed. Pergamon Press, Oxford, U.K. p. 15–31.

12. van der Zee, J. S., P. van Swieten, and R. C. Aalberse. 1986. Inhibition ofcomplement activation by IgG4 antibodies. Clin. Exp. Immunol. 64: 415–422.

13. Labrijn, A. F., R. C. Aalberse, and J. Schuurman. 2008. When binding is enough:nonactivating antibody formats. Curr. Opin. Immunol. 20: 479–485.

14. Salfeld, J. G. 2007. Isotype selection in antibody engineering. Nat. Biotechnol.25: 1369–1372.

15. Labrijn, A. F., A. O. Buijsse, E. T. van den Bremer, A. Y. Verwilligen,W. K. Bleeker, S. J. Thorpe, J. Killestein, C. H. Polman, R. C. Aalberse,J. Schuurman, et al. 2009. Therapeutic IgG4 antibodies engage in Fab-arm ex-change with endogenous human IgG4 in vivo. Nat. Biotechnol. 27: 767–771.

16. Scinicariello, F., C. N. Engleman, L. Jayashankar, H. M. McClure, andR. Attanasio. 2004. Rhesus macaque antibody molecules: sequences and het-erogeneity of alpha and gamma constant regions. Immunology 111: 66–74.

17. Rose, R. J., A. F. Labrijn, E. T. van den Bremer, S. Loverix, I. Lasters, P. H. vanBerkel, J. G. van de Winkel, J. Schuurman, P. W. Parren, and A. J. Heck. 2011.Quantitative analysis of the interaction strength and dynamics of human IgG4half-molecules by native mass spectrometry. Structure. In press.

18. Delhaise, P., M. Bardiaux, and S. J. Wodak. 1984. Interactive computer ani-mation of macromolecules. J. Mol. Graph. 2: 103–106.

19. Desmet, J., G. Meersseman, N. Boutonnet, J. Pletinckx, K. De Clercq,M. Debulpaep, T. Braeckman, and I. Lasters. 2005. Anchor profiles of HLA-specific peptides: analysis by a novel affinity scoring method and experimentalvalidation. Proteins 58: 53–69.

20. Desmet, J., J. Spriet, and I. Lasters. 2002. Fast and accurate side-chain topologyand energy refinement (FASTER) as a new method for protein structure opti-mization. Proteins 48: 31–43.

21. Bleeker, W. K., J. J. Lammerts van Bueren, H. H. van Ojik, A. F. Gerritsen,M. Pluyter, M. Houtkamp, E. Halk, J. Goldstein, J. Schuurman, M. A. van Dijk,et al. 2004. Dual mode of action of a human anti-epidermal growth factor re-ceptor monoclonal antibody for cancer therapy. J. Immunol. 173: 4699–4707.

22. Teeling, J. L., R. R. French, M. S. Cragg, J. van den Brakel, M. Pluyter,H. Huang, C. Chan, P. W. Parren, C. E. Hack, M. Dechant, et al. 2004. Char-acterization of new human CD20 monoclonal antibodies with potent cytolyticactivity against non-Hodgkin lymphomas. Blood 104: 1793–1800.

23. Deisenhofer, J. 1981. Crystallographic refinement and atomic models of a humanFc fragment and its complex with fragment B of protein A from Staphylococcusaureus at 2.9- and 2.8-A resolution. Biochemistry 20: 2361–2370.

24. Dall’Acqua, W., A. L. Simon, M. G. Mulkerrin, and P. Carter. 1998. Contributionof domain interface residues to the stability of antibody CH3 domain homo-dimers. Biochemistry 37: 9266–9273.

25. Idusogie, E. E., L. G. Presta, H. Gazzano-Santoro, K. Totpal, P. Y. Wong,M. Ultsch, Y. G. Meng, and M. G. Mulkerrin. 2000. Mapping of the C1q bindingsite on rituxan, a chimeric antibody with a human IgG1 Fc. J. Immunol. 164:4178–4184.

26. Corper, A. L., M. K. Sohi, V. R. Bonagura, M. Steinitz, R. Jefferis, A. Feinstein,D. Beale, M. J. Taussig, and B. J. Sutton. 1997. Structure of human IgM rheu-matoid factor Fab bound to its autoantigen IgG Fc reveals a novel topology ofantibody-antigen interaction. Nat. Struct. Biol. 4: 374–381.

27. Asada, Y., Y. Kawamoto, T. Shotake, and K. Terao. 2002. Molecular evolution ofIgG subclass among nonhuman primates: implication of differences in antigenicdeterminants among Apes. Primates 43: 343–349.

28. Shearer, M. H., R. D. Dark, J. Chodosh, and R. C. Kennedy. 1999. Comparisonand characterization of immunoglobulin G subclasses among primate species.Clin. Diagn. Lab. Immunol. 6: 953–958.

29. Franek, F. 1987. The subclass heterogeneity of pig IgG. Vet. Immunol. Immu-nopathol. 17: 79–89.

30. Margni, R. A., and R. A. Binaghi. 1988. Nonprecipitating asymmetric anti-bodies. Annu. Rev. Immunol. 6: 535–554.

31. Rhesus Macaque Genome Sequencing and Analysis Consortium; Gibbs, R. A.,J. Rogers, M. G. Katze, R. Bumgarner, G. M. Weinstock, E. R. Mardis,K. A. Remington, R. L. Strausberg, J. C. Venter, et al. 2007. Evolutionary andbiomedical insights from the rhesus macaque genome. Science 316: 222–234.

32. Chacko, S., E. Silverton, L. Kam-Morgan, S. Smith-Gill, G. Cohen, andD. Davies. 1995. Structure of an antibody-lysozyme complex unexpected effectof conservative mutation. J. Mol. Biol. 245: 261–274.

33. Teillet, F., M. Lacroix, S. Thiel, D. Weilguny, T. Agger, G. J. Arlaud, andN. M. Thielens. 2007. Identification of the site of human mannan-binding lectininvolved in the interaction with its partner serine proteases: the essential role ofLys55. J. Immunol. 178: 5710–5716.

34. Rispens, T., P. Ooijevaar-de Heer, O. Bende, and R. C. Aalberse. 2011. Mechanismof Immunoglobulin G4 Fab-arm Exchange. J. Am. Chem. Soc. 133: 10302–10311.

35. Ellerson, J. R., D. Yasmeen, R. H. Painter, and K. J. Dorrington. 1976. Structureand function of immunoglobulin domains. III. Isolation and characterization ofa fragment corresponding to the Cgamma2 homology region of human immu-noglobin G1. J. Immunol. 116: 510–517.

The Journal of Immunology 3245

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

36. Brusco, A., S. Saviozzi, F. Cinque, M. DeMarchi, C. Boccazzi, G. de Lange,A. M. van Leeuwen, and A. O. Carbonara. 1998. Molecular characterizationof immunoglobulin G4 gene isoallotypes. Eur. J. Immunogenet. 25: 349–355.

37. Krapp, S., Y. Mimura, R. Jefferis, R. Huber, and P. Sondermann. 2003. Structuralanalysis of human IgG-Fc glycoforms reveals a correlation between glycosyla-tion and structural integrity. J. Mol. Biol. 325: 979–989.

38. Saphire, E. O., P. W. Parren, R. Pantophlet, M. B. Zwick, G. M. Morris,P. M. Rudd, R. A. Dwek, R. L. Stanfield, D. R. Burton, and I. A. Wilson. 2001.Crystal structure of a neutralizing human IGG against HIV-1: a template forvaccine design. Science 293: 1155–1159.

39. Sondermann, P., R. Huber, V. Oosthuizen, and U. Jacob. 2000. The 3.2-A crystalstructure of the human IgG1 Fc fragment-Fc gammaRIII complex. Nature 406:267–273.

3246 ROLE OF THE CH3 DOMAIN IN THE MECHANISM OF FAB-ARM EXCHANGE

by guest on September 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from