We report the structure of the Fc fragment of rabbit IgG at 1.95 Å (1 Å=0.1 nm) resolution. Rabbit IgG was the molecule for which Porter established the four-chain, Υ-shaped structure of the antibody molecule, and crystals of the Fc (‘Fragment crystallisable’) were first reported almost 50 years ago in this journal [Porter, R. R. (1959) Biochem. J. 73, 119–126]. This high-resolution analysis, apparently of the same crystal form, reveals several features of IgG-Fc structure that have not previously been described. More of the lower hinge region is visible in this structure than in others, demonstrating not only the acute bend in the IgG molecule that this region can mediate, as seen in receptor complexes, but also that this region has a tendency to adopt a bent structure even in the absence of receptor. As observed in other IgG-Fc structures, the Cγ2 domains display greater mobility/disorder within the crystals than the Cγ3 domains; unexpectedly the structure reveals partial cleavage of both Cγ2 intra-domain disulphide bonds, whereas an alternative conformation for one of the cysteine residues in the intact bridge within the more ordered Cγ3 domains is observed. The N-linked oligosaccharide chains at Asn297 are well-defined and reveal two alternative conformations for the galactose units on each of the α(1–6)-linked branches. The presence of this galactose unit is important for stabilizing the structure of the entire branched carbohydrate chain, and its absence correlates with the severity of autoimmune conditions such as rheumatoid arthritis in both human clinical studies and in a rabbit model of the disease. Rabbit IgG, through this high-resolution structure of its Fc region, thus continues to offer new insights into antibody structure.

INTRODUCTION

The rabbit (Oryctolagus cuniculus) immunoglobulin gamma, IgG, was one of the first molecules of the immune system to be investigated, and the pioneering studies of Porter [1] produced the very first information about the structure of the antibody molecule. Proteolytic cleavage of rabbit IgG by papain into three fragments, one of which crystallized and was termed Fc (‘Fragment crystallisable’), together with two identical fragments, termed Fab, (‘Fragments antigen-binding’), led to the first proposal of the four-chain, Υ-shaped structure for IgG [2].

Even before the hetero-tetrameric model of two identical heavy and two identical light chains was developed, the first immunoglobulin allotypes were described for the rabbit IGHG (immunoglobulin gamma heavy chain) gene [3]. This observation led Oudin [3] to hypothesize that IgG production was regulated by two independently segregating loci (named a and b). Several years later these two loci were identified respectively in the variable region of the heavy chain and the constant region of the κ-light chain genes, confirming the presence of two distinct protein chains in the molecule.

Moreover, amino acid sequence studies on rabbit γ chains were pivotal in clarifying the nature of immunoglobulin domains [4] and the arrangement of disulphide bonds present in the molecule [5]. Further enzymatic digestion and functional characterization studies of rabbit IgG-Fc provided the first attempts to map the binding sites for complement and cytophilic activity [68]. Later, the same molecule was used as a model for the determination of the composition and topology of the carbohydrate chains present on the Fc fragment [9,10], which play a role in stabilizing the structure of the Fc and are required for binding to cell-surface receptors.

Despite the key role played by this molecule in the history of immunology, no crystal structure is currently available for the Fc fragment of rabbit IgG. We describe here the 1.95 Å (1 Å=0.1 nm) resolution structure, one of the highest resolution Fc structures yet reported, from a crystal form that appears identical to the crystals pictured almost 50 years ago by Porter [1].

EXPERIMENTAL

Protein purification

Purified polyclonal rabbit IgG (Sigma–Aldrich) was digested with papain according to the manufacturer's instructions (ImmunoPure Fab preparation kit; Pierce Corporation). The Fab portion was removed from the resulting mixture using a Protein A column. Fractions containing the Fc fragment and undigested IgG were then loaded on to a Superdex-200 gel-filtration column (Amersham Pharmacia Biotech) equilibrated with 0.5 M Tris/HCl (pH 7.2), 0.25 M NaCl and 0.1% (w/v) sodium azide. The purity of the eluted sample was assessed by non-reducing SDS/PAGE; a single band indicated the presence of intact dimer (results not shown). Fractions containing the purified Fc were pooled and concentrated to 2 mg/ml at 4 °C by centrifugation through an Amicon 10-kDa-molecular-mass-cut-off membrane (Millipore) and stored at the same temperature until further use.

Crystallization of the Fc fragment

Crystallization conditions for the IgG-Fc fragment have previously been reported [8,9], but initial crystallization trials using these conditions were unsuccessful. However, crystals were obtained under several conditions from a commercial sparse matrix screen (Crystal Screen I; Hampton Research) and promising hits were further optimized using the hanging-drop vapour-diffusion method. Conditions that yielded crystals suitable for data collection were as follows: 1 ml of 0.1 M sodium acetate at pH 4.7 and 2 M sodium formate (reservoir) and 1 μl of protein solution with an equal volume of reservoir (drop). The drops were stored at 18 °C and plate-like crystals measuring approx. 0.1 mm ×0.1 mm typically appeared after 2 days. Their morphology was identical with those pictured [1,6] and characterized [11,12] previously.

Data collection and processing

Crystals were fished using nylon loops, soaked in a cryoprotectant solution [0.1 M sodium acetate, pH 4.7, 2 M sodium formate and 30% (v/v) glycerol] and flash-cooled in liquid nitrogen. Data were collected at beamline ID29 at the ESRF (European Synchrotron Radiation Facility; Grenoble, France). The data were processed with MOSFLM [13], SCALA [14] and the CCP4 program suite [15]. The space group and cell parameters were found to be very similar to those observed previously [12]. Calculation of the Matthews coefficient (2.53; 51.1% solvent) indicated the presence of one Fc molecule per asymmetric unit. Data collection and statistics are presented in Table 1. The structure was solved by molecular replacement with MOLREP [15,16] using protein atoms from the human IgG-Fc structure as a search model (PDB code 1H3T [17]). Refinement was performed initially with REFMAC [18] and later with PHENIX [19]. TLS (translation/libration/screw) refinement was implemented using 10 TLS groups per chain, as suggested by the TLSMD server [20]. Cycles of refinement were alternated with rounds of manual model building with COOT [21]. Protein atoms were modelled first, followed by carbohydrate, ligands and water molecules. The refined structure was analysed with COOT, PROCHECK [22] and MolProbity [23]. Secondary structure was assigned with the KSDSSP algorithm implemented in Chimera [24]. Omit maps were prepared in CNS (Cranked Nilsson–Strutinsky) [25]. Figures were prepared with PyMOL (http://pymol.org/). Refinement statistics are presented in Table 2.

Table 1
Data collection and processing statistics
Parameter  
Beamline ID29 
Detector ADSC Q315r 
Detector distance (mm) 230 
λ (Å) 0.976 
Resolution limit (Å) 1.95 
Space group P21 (4) 
Cell dimensions (Å; °) a=58.66, b=71.21, c=69.03; 
  α=γ=90.00, β=104.78 
Unique reflections 40194 
Outer shell (Å) 2.00–1.95 
Completeness (%): overall (outer shell) 99.9 (99.9) 
Multiplicity: overall (outer shell) 7.3 (7.2) 
I/σ(I): overall (outer shell) 17.2 (3.6) 
Rmerge: overall (outer shell) 0.107 (0.449) 
Parameter  
Beamline ID29 
Detector ADSC Q315r 
Detector distance (mm) 230 
λ (Å) 0.976 
Resolution limit (Å) 1.95 
Space group P21 (4) 
Cell dimensions (Å; °) a=58.66, b=71.21, c=69.03; 
  α=γ=90.00, β=104.78 
Unique reflections 40194 
Outer shell (Å) 2.00–1.95 
Completeness (%): overall (outer shell) 99.9 (99.9) 
Multiplicity: overall (outer shell) 7.3 (7.2) 
I/σ(I): overall (outer shell) 17.2 (3.6) 
Rmerge: overall (outer shell) 0.107 (0.449) 
Table 2
Refinement statistics
Parameter Value 
Resolution range (Å) 56.7–1.95 
Number of:  
 Protein atoms 3479 
 Carbohydrate atoms 186 
 Water molecules 398 
 Formate ions 
 Azide ions 
 Glycerol molecules 
Average B factors (Å2 
 Protein 24.6 
 Carbohydrate 42.5 
 Water molecules 35.1 
 Formate ions 39.9 
 Azide ions 38.3 
 Glycerol molecules 48.5 
 Protein Cγ2 (chain A) 28.5 
 Protein Cγ2 (chain B) 27.0 
 Protein Cγ3 (chain A) 21.8 
 Protein Cγ3 (chain B) 20.5 
Rcryst (%) 16.9 
Rfree (%) (5% of reflections) 20.4 
rmsd bond lengths (Å) 0.004 
rmsd bond angles (°) 0.928 
Ramachandran plot (% of total residues) Favoured: 99.5, allowed: 0.5, disallowed: 0 
Parameter Value 
Resolution range (Å) 56.7–1.95 
Number of:  
 Protein atoms 3479 
 Carbohydrate atoms 186 
 Water molecules 398 
 Formate ions 
 Azide ions 
 Glycerol molecules 
Average B factors (Å2 
 Protein 24.6 
 Carbohydrate 42.5 
 Water molecules 35.1 
 Formate ions 39.9 
 Azide ions 38.3 
 Glycerol molecules 48.5 
 Protein Cγ2 (chain A) 28.5 
 Protein Cγ2 (chain B) 27.0 
 Protein Cγ3 (chain A) 21.8 
 Protein Cγ3 (chain B) 20.5 
Rcryst (%) 16.9 
Rfree (%) (5% of reflections) 20.4 
rmsd bond lengths (Å) 0.004 
rmsd bond angles (°) 0.928 
Ramachandran plot (% of total residues) Favoured: 99.5, allowed: 0.5, disallowed: 0 

Atomic co-ordinates and structure factors have been deposited in the PDB database with the accession code 2VUO.

RESULTS AND DISCUSSION

Overall structure

The 1.95Å resolution structure of rabbit IgG-Fc displays the typical horseshoe-like arrangement of the Cγ2 and Cγ3 domains in the two polypeptide chains (Figure 1). Superimposition with the highest resolution human IgG-Fc structure (PDB code 1L6X; 73% sequence identity with rabbit IgG-Fc), revealed rmsd (root mean square deviation) values of 0.81 and 0.76 Å for the two chains (performed with LSQMAN software [26]). Superimpositions of the individual domains (rmsd 0.37 and 0.47 Å for Cγ3; 0.80 and 0.85 Å for Cγ2) indicated that most of the differences between the rabbit and human structure occur in the Cγ2 domains.

Overall structure of rabbit IgG-Fc

Figure 1
Overall structure of rabbit IgG-Fc

In this stereo representation of the structure, the two polypeptide chains are shown in green and blue, with their oligosaccharide chains (indicated as C and D) in yellow. The position of the N-termini on both chains is indicated by the letter N. The N-terminal lower hinge region in each chain points directly towards the reader.

Figure 1
Overall structure of rabbit IgG-Fc

In this stereo representation of the structure, the two polypeptide chains are shown in green and blue, with their oligosaccharide chains (indicated as C and D) in yellow. The position of the N-termini on both chains is indicated by the letter N. The N-terminal lower hinge region in each chain points directly towards the reader.

In the rabbit IgG-Fc structure, interpretable density was present for residues Pro230–Ser444 in chain A, and residues Pro231–Ser444 in chain B; the three C-terminal residues were disordered. At the N-termini, although no density was observed in either chain for Cys229, the residue involved in the inter-chain disulphide bond, much more of the ‘lower hinge’ region is visible here than in other Fc structures (usually defined from Pro238), except those complexed with receptors (FcγR) that bind to this part of the Fc molecule. However, Pro230 and Pro231 (in chains A and B respectively) are only 7.8 Å apart (distance between closest atoms), consistent with an intact inter-chain disulphide bridge between Cys229 residues. In Figure 1, this lower hinge region of the polypeptide chain points towards the viewer and demonstrates the flexibility and acute bend in the IgG molecule that it can mediate. It is striking that the conformation of this region in rabbit Fc is very similar to that observed in Fc-receptor complexes (PDB codes 1E4K, 1T89, 1T83) [27,28] and in intact Igs (1HZH, 1IGT) [29,30]. In the receptor complexes, a bend in this lower hinge region is demanded by the presence of the receptor. In the rabbit Fc structure, a single hydrogen bond between Glu233 (chain B) and Arg443 in the Cγ3 domain of a symmetry-related molecule may assist in stabilizing this part of the structure, but the only other crystal-packing contacts with any lower hinge region residues on either chain are a total of five van der Waals interactions (Glu233 on both chains and Leu235 on chain A; cut-off <3.5 Å). Thus, although the presence of a symmetry-related molecule in this crystal form may restrict the conformational freedom of the lower hinge compared with other Fc crystal structures, it appears that this part of the hinge has a natural tendency to adopt this ‘bent’ conformation even in the absence of bound receptor.

The Cγ2 domains are more mobile (or disordered) than the Cγ3 dimer, as reflected by their higher overall B factors (average values 27.8 and 21.2 Å2 respectively; Table 2). This feature has been seen in almost all other Fc structures, from the first reported human IgG-Fc [31] to the recent high-resolution structure 1L6X (average values 23.8 and 18.7 Å2 respectively). No significant difference in the B values was observed between the two chains of rabbit IgG-Fc, however, implying that the values for Cγ2 and Cγ3 reflect an intrinsic difference between the two domains, and are not the result of different crystal-packing environments.

The conformation of the main chain for residues 283–296, adjacent to residue Asn297, to which carbohydrate is covalently attached, has been found to adopt one of two different conformations in other IgG-Fc structures; these have been termed the ‘classical’ and the ‘complex’ conformation [32]. In rabbit IgG-Fc, the structure conforms to the ‘complex’ conformation observed in the previously published high-resolution human IgG-Fc structures [17,32].

Intra-domain disulphide bonds

Each domain contains a conserved intra-chain disulphide bond, bridging the two β-sheets of the immunoglobulin fold. In both Cγ2 domains, the electron density indicates the presence of both an intact covalent bond and a broken state for the Cys261 to Cys321 bridge (Figure 2A). The differences in conformation between the two cysteine residues derives principally from a rotation around the χ1 (N-Cα-Cβ-Cγ) torsion angle, and are almost certainly the result of radiation damage during data collection. In the Cγ3 domains, there is no evidence of intra-domain disulphide bond cleavage, but Cys425 in chain A (Figure 2B) displays two alternative conformations, indicating that some degree of flexibility is present even in the core of this Ig domain. The partial cleavage of both bonds within the Cγ2 domains is clearly consistent with the higher B values observed for this domain, as discussed above.

Conformations of the intra-domain disulphide bridges

Figure 2
Conformations of the intra-domain disulphide bridges

Stereo images of the region surrounding the intra-chain disulphide bridges in Cγ2 (A) and Cγ3 (B) of chain A is shown. The electron density was calculated from a composite simulated annealing omit map, contoured at 1 σ, generated omitting 4% of the structure per cycle. Carbon atoms are represented in green, oxygen in red, nitrogen in blue and sulphur atoms in orange. The highly conserved tryptophan residues (277 and 381) adjacent to the disulphide bonds can be seen towards the back of the image. In Cγ2 (A), the disulphide bond is partially broken, whereas in Cγ3 (B), alternative conformations for the intact bridge are observed.

Figure 2
Conformations of the intra-domain disulphide bridges

Stereo images of the region surrounding the intra-chain disulphide bridges in Cγ2 (A) and Cγ3 (B) of chain A is shown. The electron density was calculated from a composite simulated annealing omit map, contoured at 1 σ, generated omitting 4% of the structure per cycle. Carbon atoms are represented in green, oxygen in red, nitrogen in blue and sulphur atoms in orange. The highly conserved tryptophan residues (277 and 381) adjacent to the disulphide bonds can be seen towards the back of the image. In Cγ2 (A), the disulphide bond is partially broken, whereas in Cγ3 (B), alternative conformations for the intact bridge are observed.

Allelic variation

Allotypes of the IGHG gene, which codes for the only IgG heavy chain subclass in the rabbit, have been described in the constant region of the γ heavy chain. They correlate with single amino-acid substitutions at positions 228 and 309 in the hinge and Cγ2 domain respectively [33]. Each locus presents two serologically defined alleles: d11 and d12 at position 228 and e14 and e15 at position 309. The two allelic variants at position 309, which is defined in the crystal structure, correspond to an alanine/threonine substitution. Given the nature of the rabbit IgG sample, no information was available regarding the sequence or the allelic proportions present in the crystallized protein. As the allelic variant with alanine is the more widespread [34], this residue (corresponding to the e15 allotype) was refined at position 309. No positive density in the Fo–Fc map (even at a contour level of ∼2.0 σ) was observed, suggesting that the allele with alanine was the one exclusively present in the crystal. The presence of an alanine residue at position 309 was later confirmed by MS analysis (results not shown). An allotypic variant (nG4m; leucine/valine) is also found at position 309 in human IgG4 [35]. This amino acid is located at the Cγ2/Cγ3 interface, in the binding site for the neonatal Fc receptor [36]. Although the two variants characterized in human Fc both code for hydrophobic residues, the e14 allotype in rabbit Fc codes for a polar residue, which may affect the binding of the neonatal receptor, although there are no reported functional studies that test this hypothesis. Incidentally, we observed positive density on both chains at Ala396, indicative of a threonine or valine residue, although no variability has been noted previously at this position.

Glycosylation

Rabbit IgG-Fc possesses a conserved glycosylation site at Asn297 in each Cγ2 domain, at which a complex biantennary-type oligosaccharide is covalently attached. The carbohydrate covers a hydrophobic patch on the Cγ2 domain surface and extends into the cavity between the two Cγ2 domains (Figure 3). Its presence stabilizes and maintains the structure of the protein, as proposed initially from studies of the human IgG-Fc fragment [31] and later confirmed by microcalorimetry experiments [37,38] and structural studies of sequentially deglycosylated Fc glycoforms [17]. Moreover, the conserved oligosaccharide plays a crucial role in IgG effector functions: it is required for optimal Fcγ receptor binding, and mediates interactions with serum lectins, such as MBP (mannose-binding protein), which leads to complement activation [39,40].

The carbohydrate structure of rabbit IgG-Fc

Figure 3
The carbohydrate structure of rabbit IgG-Fc

(A) Electron density for the N-linked oligosaccharides at Asn297 in each polypeptide chain, calculated from a composite simulated annealing omit map (contour level: 1 σ), is shown together with a representation of the two Cγ2 domains. (B) Schematic representation of the complex biantennary oligosaccharide chain present in rabbit IgG-Fc. The residues modelled in the electron density are highlighted in light grey (GlcNAc-8 is only modelled on chain C).

Figure 3
The carbohydrate structure of rabbit IgG-Fc

(A) Electron density for the N-linked oligosaccharides at Asn297 in each polypeptide chain, calculated from a composite simulated annealing omit map (contour level: 1 σ), is shown together with a representation of the two Cγ2 domains. (B) Schematic representation of the complex biantennary oligosaccharide chain present in rabbit IgG-Fc. The residues modelled in the electron density are highlighted in light grey (GlcNAc-8 is only modelled on chain C).

Although the core of the oligosaccharide is conserved, heterogeneity is observed in the fucosylation (found in ∼30–40% of rabbit IgG-Fc molecules), galactosylation (∼50–60%) and sialylation of the chain (∼25%); an additional bisecting N-acetylglucosamine residue may also be present (∼25–30% of rabbit IgG-Fc molecules) [41,42]. On both chains, carbohydrate units up to and including the galactose residue at the end of each of the α(1–6) branches (Gal-6) were built into good electron density (Figure 3), whereas on the α(1–3) branches, only Man-7 and GlcNAc-8 (the latter only on chain C, Figure 3) were built. The lack of interpretable electron density for fucose, terminal sialic acid or a bisecting N-acetylglucosamine is consistent with the low fractional compositions reported. The α(1–3) branches extend into the area between the two Cγ2 domains and lie close to each other, but none of the contacts between the two chains described previously in the much lower resolution analysis [9] were seen. Although some unliganded human Fc structures show inter-chain carbohydrate contacts, others do not [43]; whether or not such contacts occur appears to depend upon domain quaternary structure, which in turn may be determined by crystal-packing contacts. The α(1–6) branches lie close to the first two strands (A and B) of the Cγ2 domain, their core regions interacting with residues Phe241, Phe243 and Glu265.

The extent of galactosylation in IgG molecules is highly variable, with a reported 40–50% of the molecules entirely lacking galactose in healthy rabbits [41,42]. We observed density at the end of each α(1–6) branch, clearly indicating the presence of a galactose residue (Gal-6). When refined with an occupancy of 100% (as for all the other carbohydrate units), higher B factors and some negative difference electron density indicated a slightly lower occupancy and/or the presence of alternative conformations (refinement with occupancy values set between 40 and 60% revealed positive difference electron density). It was clear, however, that the principal conformation adopted by Gal-6 is different in the two chains (Figures 1 and 3). In chain A, the O-6 atom of Gal-6 points away from the Cγ2 domain, whereas in chain B the O-6 atom points towards the Cγ2 surface. These two conformations are related by an approx. 180° rotation of the plane of the hexose ring, but in both, Gal-6 lies in the same cavity, forming (different) hydrogen bonds to the same residues Pro244, Lys246 (although density for CE and CZ of Lys246 on chain A is ill-defined), Glu258 and Thr260. The galactose conformation in chain A is identical to that seen in the 1.65 Å human IgG-Fc (PDB code 1L6X), but no asymmetry was observed in that structure as the two chains are related by a crystallographic dyad. There is thus some variability in the nature of the interaction between this carbohydrate unit and the protein surface.

Fc galactosylation and functional implications

The galactose content of human IgG-Fc has been shown to correlate inversely with disease progression in RA (rheumatoid arthritis) [44,45] and other auto-immune diseases [46]. Various mechanisms have been proposed: the absence of galactose may expose hydrophobic surfaces that promote Fc-mediated IgG aggregation, or generate new Fc epitopes that are recognized by RF (rheumatoid factor) autoantibodies [47]. The galactose unit is also known to be a key determinant for stabilizing the carbohydrate chain's interaction with the protein domain. Its absence leads to greater mobility of the chains, as shown by NMR [48] and crystallographic [49] studies, and this enhanced accessibility permits interaction with lectins such as MBP, leading to complement activation [40]. Similar changes in the galactose content of rabbit IgG have been reported after hyperimmunization, and enhanced avidity of rabbit IgG RF autoantibodies for IgG-Fc was shown to be due to decreased galactose content [50]. In a rabbit model of RA, a long-term immunization study that monitored RF titre, avidity and IgG-Fc galactose content during disease progression showed the same inverse correlation between RF avidity for IgG-Fc and galactose content [51]. The crystal structure of rabbit IgG-Fc has revealed a conformational variability in the mode of interaction of this key carbohydrate residue and the Cγ2 domain, underlining the tenuous nature of this contact and its role in tipping the balance between a mobile carbohydrate chain and an ordered structure in contact with the Cγ2 domain surface.

Rabbit IgG provided the very first insights into antibody structure and, through the high-resolution crystal structure of its Fc region reported here, it continues to reveal new structural details and enhance our understanding of the antibody molecule.

We are grateful to Sophia Karagiannis for assistance with the Fc fragment preparation and to Roberto Steiner, Stella Fabiane, Balvinder Dhaliwal and Silviana Comsa for helpful discussions. We are also grateful to Karen Homer (Dental Institute, King's College London) for the MS analysis. We thank the staff at ID29 at ESRF (Grenoble, France) for assistance with data collection.

Abbreviations

     
  • Fc

    fragment crystallizable

  •  
  • IGHG

    IgG heavy chain

  •  
  • rmsd

    root mean square deviation

  •  
  • MBP

    mannose-binding protein

  •  
  • RA

    rheumatoid arthritis

  •  
  • RF

    rheumatoid factor

  •  
  • TLS

    translation/libration/screw

FUNDING

E. G. is funded by a King's College London Strategic Development Studentship.

References

References
1
Porter
R. R.
The hydrolysis of rabbit γ-globulin and antibodies with crystalline papain
Biochem. J.
1959
, vol. 
73
 (pg. 
119
-
126
)
2
Fleischman
J. B.
Porter
R. R.
Press
E. M.
The arrangement of the peptide chains in γ-globulin
Biochem. J.
1963
, vol. 
88
 (pg. 
220
-
228
)
3
Oudin
J.
The allotype of certain blood protein antigens
C. R. Hebd. Seances Acad. Sci.
1956
, vol. 
242
 (pg. 
2606
-
2608
)
4
Hill
R. L.
Delaney
R.
Lebovitz
H. E.
Fellows
R. E.
Jr
Studies on the amino acid sequence of heavy chains from rabbit immunoglobulin G
Proc. R. Soc. London B Biol. Sci.
1966
, vol. 
166
 (pg. 
159
-
175
)
5
O'Donnell
I. J.
Frangione
B.
Porter
R. R.
The disulphide bonds of the heavy chain of rabbit immunoglobulin G
Biochem. J.
1970
, vol. 
116
 (pg. 
261
-
268
)
6
Utsumi
S.
Stepwise cleavage of rabbit immunoglobulin G by papain and isolation of four types of biologically active Fc fragments
Biochem. J.
1969
, vol. 
112
 (pg. 
343
-
355
)
7
Colomb
M.
Porter
R. R.
Characterization of a plasmin-digest fragment of rabbit immunoglobulin gamma that binds antigen and complement
Biochem. J.
1975
, vol. 
145
 (pg. 
177
-
183
)
8
Ovary
Z.
Saluk
P. H.
Quijada
L.
Lamm
M. E.
Biologic activities of rabbit immunoglobulin G in relation to domains of the Fc region
J. Immunol.
1976
, vol. 
116
 (pg. 
1265
-
1271
)
9
Sutton
B. J.
Phillips
D. C.
The three-dimensional structure of the carbohydrate within the Fc fragment of immunoglobulin G
Biochem. Soc. Trans.
1983
, vol. 
11
 (pg. 
130
-
132
)
10
Rademacher
T. W.
Homans
S. W.
Fernandes
D. L.
Dwek
R. A.
Mizuochi
T.
Taniguchi
T.
Kobata
A.
Structural and conformational analysis of immunoglobulin-derived N-linked oligosaccharides
Biochem. Soc. Trans.
1983
, vol. 
11
 (pg. 
132
-
134
)
11
Poljak
R. J.
Dintzis
H. M.
Papain fragmentation of rabbit gamma-g-globulin: A crystallographic study
J. Mol. Biol.
1966
, vol. 
17
 (pg. 
546
-
547
)
12
Aschaffenburg
R.
Lewis
M.
Phillips
D. C.
Press
E. M.
Smith
S. G.
Sutton
B. J.
Mountford
C. W.
Crystallographic studies of immunoglobulins: crystallization of the Fc fragment of rabbit IgG with and without cleavage of the inter-chain disulphide bridge
J. Mol. Biol.
1979
, vol. 
135
 (pg. 
1033
-
1036
)
13
Leslie
A. G. W.
Recent changes to the MOSFLM package for processing film and image plate data
Joint CCP4+ESF-EAMCB Newsletter on Protein Crystallography, no. 26
1992
Warrington
Daresbury Laboratories
14
Evans
P. R.
Wilson
K. S.
Davies
G.
Ashton
A. W.
Bailey
S.
Scaling of MAD data
Proceedings of CCP4 Study Weekend on Recent Advances in Phasing
1997
Warrington
Daresbury Laboratory
(pg. 
97
-
102
)
15
Collaborative Computational Project
Number 4
The CCP4 suite: Programs for protein crystallography
Acta Crystallogr. D Biol. Crystallogr.
1994
, vol. 
50
 (pg. 
760
-
763
)
16
Vagin
A. A.
Teplyakov
A.
MOLREP: An automated program for molecular replacement
J. Appl. Crystallogr.
1997
, vol. 
30
 (pg. 
1022
-
1025
)
17
Krapp
S.
Mimura
Y.
Jefferis
R.
Huber
R.
Sondermann
P.
Structural analysis of human IgG-Fc glycoforms reveals a correlation between glycosylation and structural integrity
J. Mol. Biol.
2003
, vol. 
325
 (pg. 
979
-
989
)
18
Murshudov
G. N.
Vagin
A. A.
Dodson
E. J.
Refinement of macromolecular structures by the maximum-likelihood method
Acta Crystallogr. D Biol. Crystallogr.
1997
, vol. 
53
 (pg. 
240
-
255
)
19
Adams
P. D.
Grosse-Kunstleve
R. W.
Hung
L. W.
Ioerger
T. R.
McCoy
A. J.
Moriarty
N. W.
Read
R. J.
Sacchettini
J. C.
Sauter
N. K.
Terwilliger
T. C.
PHENIX: Building new software for automated crystallographic structure determination
Acta Crystallogr. D Biol. Crystallogr.
2002
, vol. 
58
 (pg. 
1948
-
1954
)
20
Painter
J.
Merritt
E. A.
Optimal description of a protein structure in terms of multiple groups undergoing TLS motion
Acta Crystallogr. D Biol. Crystallogr.
2006
, vol. 
62
 (pg. 
439
-
450
)
21
Emsley
P.
Cowtan
K.
Coot: Model-building tools for molecular graphics
Acta Crystallogr. D Biol. Crystallogr.
2004
, vol. 
60
 (pg. 
2126
-
2132
)
22
Laskowsky
R. A.
MacArthur
M. W.
Moss
D. S.
Thornton
J. M.
PROCHECK: A program to check the stereochemical quality of protein structures
J. Appl. Crystallogr.
1993
, vol. 
26
 (pg. 
283
-
291
)
23
Davis
I. W.
Leaver-Fay
A.
Chen
V. B.
Block
J. N.
Kapral
G. J.
Wang
X.
Murray
L. W.
Arendall
W. B.
3rd
Snoeyink
J.
Richardson
J. S.
Richardson
D. C.
MolProbity: all-atom contacts and structure validation for proteins and nucleic acids
Nucleic Acids Res.
2007
, vol. 
35
 (pg. 
W375
-
W383
)
24
Pettersen
E. F.
Goddard
T. D.
Huang
C. C.
Couch
G. S.
Greenblatt
D. M.
Meng
E. C.
Ferrin
T. E.
UCSF chimera – a visualization system for exploratory research and analysis
J. Comput. Chem.
2004
, vol. 
25
 (pg. 
1605
-
1612
)
25
Brunger
A. T.
Adams
P. D.
Clore
G. M.
DeLano
W. L.
Gros
P.
Grosse-Kunstleve
R. W.
Jiang
J. S.
Kuszewski
J.
Nilges
M.
Pannu
N. S.
, et al. 
Crystallography & NMR system: A new software suite for macromolecular structure determination
Acta Crystallogr. D Biol. Crystallogr.
1998
, vol. 
54
 (pg. 
905
-
921
)
26
Kleywegt
G. J.
Use of non-crystallographic symmetry in protein structure refinement
Acta Crystallogr. D Biol. Crystallogr.
1996
, vol. 
52
 (pg. 
842
-
857
)
27
Sondermann
P.
Huber
R.
Oosthuizen
V.
Jacob
U.
The 3.2-Å crystal structure of the human IgG1 Fc fragment-FcγRIII complex
Nature
2000
, vol. 
406
 (pg. 
267
-
273
)
28
Radaev
S.
Motyka
S.
Fridman
W. H.
Sautes-Fridman
C.
Sun
P. D.
The structure of a human type III Fcγ receptor in complex with Fc
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
16469
-
16477
)
29
Saphire
E. O.
Parren
P. W.
Pantophlet
R.
Zwick
M. B.
Morris
G. M.
Rudd
P. M.
Dwek
R. A.
Stanfield
R. L.
Burton
D. R.
Wilson
I. A.
Crystal structure of a neutralizing human IGG against HIV-1: a template for vaccine design
Science
2001
, vol. 
293
 (pg. 
1155
-
1159
)
30
Harris
L. J.
Larson
S. B.
Hasel
K. W.
Day
J.
Greenwood
A.
McPherson
A.
The three-dimensional structure of an intact monoclonal antibody for canine lymphoma
Nature
1992
, vol. 
360
 (pg. 
369
-
372
)
31
Deisenhofer
J.
Crystallographic refinement and atomic models of a human Fc fragment and its complex with fragment B of protein A from Staphylococcus aureus at 2.9- and 2.8-Å resolution
Biochemistry
1981
, vol. 
20
 (pg. 
2361
-
2370
)
32
Matsumiya
S.
Yamaguchi
Y.
Saito
J.
Nagano
M.
Sasakawa
H.
Otaki
S.
Satoh
M.
Shitara
K.
Kato
K.
Structural comparison of fucosylated and nonfucosylated Fc fragments of human immunoglobulin G1
J. Mol. Biol.
2007
, vol. 
368
 (pg. 
767
-
779
)
33
Esteves
P. J.
Alves
P. C.
Ferrand
N.
van der Loo
W.
Hotspot variation at the CH2-CH3 interface of leporid IgG antibodies (Oryctolagus, Sylvilagus and Lepus)
Eur. J. Immunogenet.
2002
, vol. 
29
 (pg. 
529
-
535
)
34
Esteves
P. J.
Alves
P. C.
Ferrand
N.
van der Loo
W.
Restriction fragment alleles of the rabbit IGHG genes with reference to the rabbit IGHGCH2 or e locus polymorphism
Anim. Genet.
2002
, vol. 
33
 (pg. 
309
-
311
)
35
Brusco
A.
Saviozzi
S.
Cinque
F.
DeMarchi
M.
Boccazzi
C.
de Lange
G.
van Leeuwen
A. M.
Carbonara
A. O.
Molecular characterization of immunoglobulin G4 gene isoallotypes
Eur. J. Immunogenet.
1998
, vol. 
25
 (pg. 
349
-
355
)
36
West
A. P.
Jr
and Bjorkman
P. J.
Crystal structure and immunoglobulin G binding properties of the human major histocompatibility complex-related Fc receptor
Biochemistry
2000
, vol. 
39
 (pg. 
9698
-
9708
)
37
Ghirlando
R.
Lund
J.
Goodall
M.
Jefferis
R.
Glycosylation of human IgG-Fc: Influences on structure revealed by differential scanning micro-calorimetry
Immunol. Lett.
1999
, vol. 
68
 (pg. 
47
-
52
)
38
Mimura
Y.
Church
S.
Ghirlando
R.
Ashton
P. R.
Dong
S.
Goodall
M.
Lund
J.
Jefferis
R.
The influence of glycosylation on the thermal stability and effector function expression of human IgG1-Fc: Properties of a series of truncated glycoforms
Mol. Immunol.
2000
, vol. 
37
 (pg. 
697
-
706
)
39
Mimura
Y.
Sondermann
P.
Ghirlando
R.
Lund
J.
Young
S. P.
Goodall
M.
Jefferis
R.
Role of oligosaccharide residues of IgG1-Fc in Fcγ RIIb binding
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
45539
-
45547
)
40
Malhotra
R.
Wormald
M. R.
Rudd
P. M.
Fischer
P. B.
Dwek
R. A.
Sim
R. B.
Glycosylation changes of IgG associated with rheumatoid arthritis can activate complement via the mannose-binding protein
Nat. Med.
1995
, vol. 
1
 (pg. 
237
-
243
)
41
Hamako
J.
Matsui
T.
Ozeki
Y.
Mizuochi
T.
Titani
K.
Comparative studies of asparagine-linked sugar chains of immunoglobulin G from eleven mammalian species
Comp. Biochem. Physiol. B
1993
, vol. 
106
 (pg. 
949
-
954
)
42
Raju
T. S.
Briggs
J. B.
Borge
S. M.
Jones
A. J.
Species-specific variation in glycosylation of IgG: evidence for the species-specific sialylation and branch-specific galactosylation and importance for engineering recombinant glycoprotein therapeutics
Glycobiology
2000
, vol. 
10
 (pg. 
477
-
486
)
43
Oganesyan
V.
Damschroder
M. M.
Leach
W.
Wu
H.
Dall'Acqua
W. F.
Structural characterization of a mutated, ADCC-enhanced human Fc fragment
Mol. Immunol.
2008
, vol. 
45
 (pg. 
1872
-
1882
)
44
Parekh
R. B.
Dwek
R. A.
Sutton
B. J.
Fernandes
D. L.
Leung
A.
Stanworth
D.
Rademacher
T. W.
Mizuochi
T.
Taniguchi
T.
Matsuta
K.
Association of rheumatoid arthritis and primary osteoarthritis with changes in the glycosylation pattern of total serum IgG
Nature
1985
, vol. 
316
 (pg. 
452
-
457
)
45
Axford
J. S.
Glycosylation and rheumatic disease
Biochim. Biophys. Acta
1999
, vol. 
1455
 (pg. 
219
-
229
)
46
Cremata
J. A.
Sorell
L.
Montesino
R.
Garcia
R.
Mata
M.
Cabrera
G.
Galvan
J. A.
Garcia
G.
Valdes
R.
Garrote
J. A.
Hypogalactosylation of serum IgG in patients with coeliac disease
Clin. Exp. Immunol.
2003
, vol. 
133
 (pg. 
422
-
429
)
47
Newkirk
M. M.
Fournier
M. J.
Shiroky
J.
Rheumatoid factor avidity in patients with rheumatoid arthritis: identification of pathogenic RFs which correlate with disease parameters and with the gal(0) glycoform of IgG
J. Clin. Immunol.
1995
, vol. 
15
 (pg. 
250
-
257
)
48
Wormald
M. R.
Rudd
P. M.
Harvey
D. J.
Chang
S. C.
Scragg
I. G.
Dwek
R. A.
Variations in oligosaccharide-protein interactions in immunoglobulin G determine the site-specific glycosylation profiles and modulate the dynamic motion of the Fc oligosaccharides
Biochemistry
1997
, vol. 
36
 (pg. 
1370
-
1380
)
49
Corper
A. L.
Sohi
M. K.
Bonagura
V. R.
Steinitz
M.
Jefferis
R.
Feinstein
A.
Beale
D.
Taussig
M. J.
Sutton
B. J.
Structure of human IgM rheumatoid factor Fab bound to its autoantigen IgG Fc reveals a novel topology of antibody-antigen interaction
Nat. Struct. Biol.
1997
, vol. 
4
 (pg. 
374
-
381
)
50
Murray
J. S.
Brown
J. C.
Evidence that the Fc region of autologous rabbit IgG isolated before and after hyperimmunization is structurally different: Recognition by rheumatoid factor and monoclonal antibodies
Clin. Exp. Immunol.
1990
, vol. 
81
 (pg. 
286
-
292
)
51
Ciric
D.
Milosevic-Jovcic
N.
Ilic
V.
Petrovic
S.
A longitudinal study of the relationship between galactosylation degree of IgG and rheumatoid factor titer and avidity during long-term immunization of rabbits with BSA
Autoimmunity
2005
, vol. 
38
 (pg. 
409
-
416
)

Author notes

The structural co-ordinates reported will appear in the Protein Data Bank under accession code 2VUO.