Conserved Fc gamma R- glycan discriminates between fucosylated and afucosylated IgG in humans and mice

(1)

Contents lists available atScienceDirect

Molecular Immunology

journal homepage:www.elsevier.com/locate/molimm

Conserved Fc γR- glycan discriminates between fucosylated and afucosylated IgG in humans and mice

^☆

Gillian Dekkers

^a

, Arthur E.H. Bentlage

^a

, Rosina Plomp

^b

, Remco Visser

^a

,

Carolien A.M. Koeleman

^b

, Anna Beentjes

^a

, Juk Yee Mok

^c

, Wim J.E. van Esch

^c

, Manfred Wuhrer

^b

, Theo Rispens

^d

, Gestur Vidarsson

^a,⁎

aDepartment Experimental Immunohematology, Sanquin Research and Landsteiner Laboratory, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands

bCenter for Proteomics and Metabolomics, Leiden University Medical Center, Leiden, The Netherlands

cSanquin Reagents, Department R&D, Amsterdam, The Netherlands

dDepartment Immunopathology, Sanquin Research and Landsteiner Laboratory, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands

A R T I C L E I N F O

Keywords:

Fc gamma receptors N-Glycosylation Immunoglobulin Fucosylation Murine Human

A B S T R A C T

The binding strength between IgG and FcγR is influenced by the composition of the N-linked glycan at position N297 in the Fc-domain of IgG. Particularly, afucosylation increases the binding affinity of human IgG1 to human FcγRIIIa up to ∼20 fold, and additional galactosylation of the afucosylated IgG increases the affinity up to ∼40 fold. The increase in affinity for afucosylated IgG has previously been shown to depend on direct carbohydrate- carbohydrate interactions between the IgG-Fc glycan with an N-linked glycan at position 162 unique to hFcγRIIIa and hFcγRIIIb. Here we report that the N162 glycosylation site is also found in the orthologous mouse FcγR, mFcγRIV. The N162-glycan in mFcγRIV was also responsible for enhancing the binding to mouse IgG with reduced fucose similar to hFcγRIIIa. However, unlike hFcγRIIIa, mFcγRIV did not bind more avidly to IgG with increased galactose and reduced fucose. Overall, these results suggest the N162-glycan in the human FcγRIII family and its orthologous mouse FcγRIV to be functionally conserved.

1. Introduction

Immunoglobulin G (IgG) is the main antibody class found in mammalian sera like humans and mice. IgG signals to the immune cells byfirst binding of its variable domains to foreign pathogens and then by binding of the constant domain (Fc) and subsequent crosslinking of Fc- gamma receptors (FcγR) on myeloid and NK cells. The capacity of IgG to perform this task is dependent on the affinity of each IgG subclass to the different subtypes of FcγR. Furthermore, this affinity is also influenced by the composition of the conserved glycan at position Asn297 in the Fc of IgG (Dekkers et al., 2017b; Jefferis, 2009; Shields et al., 2002;

Shinkawa et al., 2003; Subedi and Barb, 2015; Yasuma et al., 2016).

This bi-antennary glycan is composed of a core structure containing N- Acetylglucosamine (GlcNAc) and mannose groups and can be variably

extended with fucose, galactose, sialic acid and an additional bisecting GlcNAc (bisection). Particularly the absence of core fucose (referred to as afucosylation from here on) on human IgG1 increases the binding aﬃnity to human FcγRIIIa and FcγRIIIb (collectively referred to here- after as hFcγRIII) by up to ∼20 fold (Dekkers et al., 2017b; Shields et al., 2002). Addition of galactose to afucosylated IgG increases the aﬃnity even further for hFcγRIIIa by another ∼2 fold, resulting in up to

∼40x better aﬃnity of this glycoform for FcγRIIIa (Dekkers et al., 2017b; Houde et al., 2010). Intriguingly, galactose addition has no effect on binding of fucosylated IgG to hFcγR, and also limited eﬀect on binding of afucosylated IgG to hFcγRIIIb (Dekkers et al., 2017b).

Of all human FcγR, FcγRIIIa and FcγRIIIb bear a glycan at position N162 in the binding site for Fc. For those FcγR lacking the N162 glycan, the IgG Fc glycan does not seem to make strong direct contact in the

https://doi.org/10.1016/j.molimm.2017.12.006

Received 1 August 2017; Received in revised form 29 November 2017; Accepted 6 December 2017

☆The authors declare that they have no conﬂicts of interest.

⁎Corresponding author at: Department Experimental Immunohematology, Sanquin Research and Landsteiner Laboratory, Academic Medical Centre, University of Amsterdam, Plesmanlaan 125, 1066CX Amsterdam, The Netherlands.

E-mail address:G.Dekkers@sanquin.nl(G. Vidarsson).

Abbreviations: Asn, asparagine; Bisection, bisecting GlcNAc; Fc, fragment crystallizable; FcγR, Fc gamma receptor; GlcNAc, N-acetylglucosamine; HEK, human embryonic kidney;

hFcγR, human FcγR; hIgG, human IgG; IgG, immunoglobulin G; mFcγR, —mouse FcγR; mIgG, mouse IgG; N162, asparagine at position 162; N162A, asparagine at position 162 substituted by alanine; N297A, asparagine at position 297 substituted by alanine; PBS, phosphate buﬀered saline; Rmax, maximum response; SPR, surface plasmon resonance; V158, valine at position 158; WT, wild type

Available online 19 December 2017

T

(2)

interaction, and only plays a minor contribution to affinity (Sondermann et al., 2000). However, a clear interaction has been observed between the IgG Fc-glycan and the N162-glycan of hFcγRIII (Ferrara et al., 2006). Interestingly, artificial removal of this hFcγRIII- N162-glycan increases affinity to fucosylated human IgG (dominant glycoform found in both human and mice), suggesting this glycan in human hFcγRIIIa to sterically interfere with binding. However, the hFcγRIII-N162-glycan also discriminates between fucosylated and afucosylated human IgG (Ferrara et al., 2011). When IgG without the Fc- core-fucose residue is compared to IgG with Fc-core-fucose closer glycan-glycan interactions are observed as well as enhanced glycan- protein interactions, which explains the higher affinity (Ferrara et al., 2011). If these interactions are also explaining the additional effects observed upon IgG-Fc galactosylation, is unknown.

For humans, the eﬀect of the glycan composition of IgG1, in particularly fucosylation, on binding to the diﬀerent FcγRs has already been thoroughly explored (Dekkers et al., 2017b; Subedi and Barb, 2016;

Thomann et al., 2015). For mice this has been less extensively studied.

The mouse repertoire of FcγRs consists of four FcγRs: mFcγRI, mFcγRIIb, mFcγRIII and mFcγRIV (Nimmerjahn et al., 2005) and their function and expression pattern differs slightly from humans, as re- viewed by Bruhns and Jönsson (2015). The mouse orthologue of hFcγRIIIa is the mouse mFcγRIV (Nimmerjahn et al., 2005). Interest- ingly, it has been reported that afucosylated mouse IgG2a and IgG2b have an at least 10 fold increased affinity for mFcγRIV compared to normally fucosylated mouse IgG. Unlike the human counterpart, a slight increase in binding affinity to mFcγRIIb and mFcγRIII was observed for the afucosylated antibodies (Nimmerjahn and Ravetch, 2005). Galactosylation had no or minimal effect on binding of mouse IgG to mFcγR (Nimmerjahn et al., 2007). However, sialylated mIgG showed an almost 10 fold decrease in affinity for mFcγRIII and mIgG2a for mFcγRIV (Kaneko et al., 2006). The molecular mechanism by which glycosylation and sialylation of mouse IgG affect binding to mouse mFcγRIV is unknown.

Here, we investigated the eﬀect of Asn297-glycosylation of human and mouse IgG on binding to FcγRs, with particular focus on the contribution of the N162-FcγR glycan.

2. Material and methods

2.1. Human and mouse IgG constructs

All restriction enzymes, ligases and accompanying buﬀers were from Thermo Fisher Scientiﬁc (Waltham, MA USA). Anti-human rhesus D (anti-D clone 19A10) IgG1 heavy and kappa light chain were both cloned into pEE14.4 expression vector (Lonza) as described previously (Dekkers et al., 2017a,b). Mouse anti-KEL1 (PUMA1 clone) expression vectors for mouse IgG1 and IgG2a heavy chain and kappa light chain were cloned as described previously (Dekkers et al., 2017a; Howie et al., 2016).

2.2. Human and mouse fusion Fc-FcγR constructs

In order to make the fusion Fc-FcγR constructs the extracellular domains of the human FcγRIIIa V158 (NCBI reference sequence:

NP_001121065.1) and mouse FcγRIV (NP_653142.2) were reverse translated and codon optimized. Additionally, for both constructs a variant with N162 mutation was created. All four constructs were ordered from Geneart (Life Technologies, Paisley, UK) (FcγRIIIa variants) or Integrated DNA Technologies (Coralville, IA, USA) (FcγRIV variants) including 5′ HindIII restriction site, Kozak and 3′ EcoRI restriction site.

The IgG2-Fc domain, composed of a human IgA1a hinge (PVPSTPPTPSPSTPPTPSPSCCH), human IgG2 Fc CH2 and CH3 domains including a mutation deleting the Fc-glycan (N297A), C-terminal biotinylation tag (BirATag) (GLNDIFEAQKIEW), 5′ EcoRI and 3′ EcoRV restriction sites was reverse translated, codon optimized at Geneart, and

ordered from Integrated DNA Technologies (Dekkers et al., 2017b).

FcγR and IgG-Fc constructs were ligated at EcoRI restriction site and cloned into pcDNA3.1 expression vector (Invitrogen, Carlsbad, Cali- fornia, USA) usingﬂanking HindIII and EcoRV restriction sites.

2.3. IgG and Fc-FcγR production and isolation

All IgGs and Fc-Fusion FcγRs were produced by transient transfection of HEK-freestyle cells (Thermo Fisher Scientific), as previously described byVink et al. (2014)andDekkers et al. (2016). After 5 days the IgG-containing or Fc-Fusion-containing cell supernatant from these cells was harvested by spinning twice at maximum speed (> 4000 g) and subsequentfiltration with 0.45 nm puradisc syringe filter (What- mann, GE Healthcare, Little Chalfront, UK).

IgG and Fc-Fusion FcγR were isolated from cell supernatant with affinity chromatography columns HiTrap Protein A HP (GE Healthcare) for human IgG1, and all Fc-Fusion FcγR constructs or HiTrap Protein G HP (GE Healthcare) for mouse IgG1 and IgG2a on ÄKTA prime (GE Healthcare) according to standard procedures. Purified fractions were concentrated to concentration > 0.5 mg/mL using Protein Concentrators, 9K MWCO (Pierce, Thermo Fisher Scientific) and subsequently dialyzed against phosphate buffered saline (PBS) overnight using Slide-A-Lyzer™ Dialysis Cassettes, 10K MWCO (Thermo Fisher Scientific). Concentration of purified protein was determined using Nanodrop 2000c UV/VIS spectrophotometer (Thermo Fisher Scientific).

2.4. IgG glyco-engineering

Glyco-engineering of human IgG1 was optimized as described by Dekkers et al. (2016). In short: To decrease either fucosylation or galactosylation 0.4 mM (for hIgG1) or 0.2 mM (for mIgG) 2-deoxy-2- ﬂuoro-^L-fucose (2FF) (Carbosynth, Berkshire, UK) or 1 mM 2-deoxy-2- ﬂuoro-^D-galactose (2FG) (Carbosynth) respectively was added to the cell suspension. To increase galactose, 1% pEE6.4 + B4GALT1 en- codingβ-1,4-galactosyltransferase 1 (B4GALT1) enzyme was co-trans- fected with 99% IgG1-κ HC + LC vector and 5 mM^D-galactose (Sigma Aldrich, Saint Louis, MO, USA) was added to the cell suspension 1 h before transfection.

2.5. Mass spectrometric analysis of IgG glycan composition

IgG Fc glycan compositions of produced human IgG1 and mouse IgG1 and IgG2a were determined by mass spectrometry as described previously byDekkers et al. (2016)andde Haan et al. (2017). In short, trypsin-digested glycopeptide samples were analyzed by nanoLC-ESI- QTOF-MS. The separation was performed on an RSLCnano Ultimate 3000 system (Thermofisher, Breda, the Netherlands) with a gradient pump, loading pump and an autosampler. The resulting co-elution of the different glycoforms of the IgG1 Fc glycosylation site warrants fair comparison by ensuring identical ionization conditions for the various glycopeptide species. The LC was coupled to the MS detector via a CaptiveSpray source with a NanoBooster (Bruker Daltonics, Bremen, Germany). The Maxis Impact quadrupole-TOF-MS (micrOTOF-Q, Bruker Daltonics) was used as detector. MSConvert (Proteowizard 3.0) (Chambers et al., 2012) was used to convert the datafiles to mzXML format, and an in-house alignment tool (Plomp et al., 2015) was used to align the retention times of the datafiles.

A list of previously observed IgG glycans (Dekkers et al., 2016) was used for relative quantiﬁcation, after manual examination of the cur- rent data did not reveal additional glycan structures. Peptide moieties of the glycopeptides diﬀered based on the type of IgG: hIgG1 (EE- QYNSTYR) (Dekkers et al., 2016), mIgG1 (EEQFNSTFR), and mIgG2a (EDYNSTLR) (de Haan et al., 2017). The highest intensity of selected peaks (within an m/z window of ± 0.04 and within a time window of ± 12 s surrounding the retention time) was extracted using the in- house developed 3D Max Xtractor software tool. If above a

(3)

signal:background ratio of 3, the background-subtracted area of the ﬁrst 3 isotopic peaks of each glycopeptide in both 2+ and 3+ charge state were summed, and this summed value was then divided by the total summed value of all glycopeptides per IgG subclass to arrive at a percentage for each glycopeptide. From these percentages we calculated several derived traits (fucosylation, bisection, galactosylation and sialylation) as described in (Dekkers et al., 2016).

2.6. Fc-FcγR biotinylation

After puriﬁcation, the protein was site-speciﬁcally biotinylated on the BirA tag using BirA enzyme as described by (Rodenko et al., 2006).

For biotinylation of 1μM FcγR protein 0.00657 μM BirA ligase was used. After biotinylation overnight at 25 °C the FcγR sample was buﬀer- exchanged and subsequently concentrated in PBS pH 7.4 using Amicon Ultra centrifugal ﬁlter units (MWCO 30 kDa) (Merck, Millipore, Darmstadt, Germany).

2.7. Acquired receptors

The following FcγR were acquired from Sino Biological Inc. (Beijing, China); mouse CD64/FcγRI, His & AVI Tag, Biotinylated (50086-M27H- B-50), mouse CD32/FcγRIIb, His & AVI Tag, Biotinylated (50030- M27H-B-50), mouse CD16/FcγRIII, His tag (50326-M08H-50), mouse CD16-2/FcγRIV (50036-M27H-B-50), human CD16a/FcγRIIIa 158Val, His & AVI Tag, Biotinylated (10389-H27H1-B-50).

2.8. Surface plasmon resonance

Surface plasmon resonance (SPR) measurements were carried out on a IBIS MX96 (IBIS technologies, Enschede, the Netherlands) as described by (de Lau et al., 2011; Dekkers et al., 2017a). Biotinylated human FcγRIIIa V158, mouse FcγRI, FcγRIIb and FcγRIV were pur- chased from SinoBiologicals (Bejing, China), biotinylated mouse FcγRIII was not available therefore His-tag conjugated mouse FcγRIII (Sino- Biologicals) was used. Either human or mouse receptors along with human fusion Fc-FcγRIIIa WT and N162A or mouse fusion Fc-FcγRIV WT and N162A respectively, were simultaneously spotted using a Continuous Flow Microspotter (Wasatch Microﬂuidics, Salt Lake City, UT) onto a single SensEye G-streptavidin sensor (Senss, Enschede, Netherlands) allowing for binding aﬃnity measurements of each antibody to all FcγR of one species simultaneously on the IBIS MX96 (IBIS technologies) as described by (de Lau et al., 2011). Spotting was always in duplicate and in three-fold dilutions, ranging from 100 nM to 3 nM for mouse FcγRIIb and 30 nM to 1 nM for mouse FcγRIIb, mouse FcγRIV human fusion Fc-FcγRIIIa WT, human fusion Fc-FcγRIIIa N162A, mouse

fusion Fc-FcγRIV WT and mouse fusion Fc-FcγRIV N162A in PBS 0.075% Tween-80 (Amresco, Solon, OH, USA), pH 7.4. For the His- tagged mouse FcγRIII, biotinylated anti-His IgG1 (GenScript, Piscat- away, NJ, USA) was spotted in duplicate and three-fold dilution onto the sensor, ranging from 1 nM to 30 nM, and 100 nM mouse FcγRIII (equally diluted in PBS 0.075% Tween-80, pH 7.4) was loaded onto the sensor before each antibody injection.

The IgGs were then injected over the chip at dilution series in PBS supplemented with 0.075% Tween-80 ranging from 0.7 nM to 1500 nM for all mouse IgG1 and IgG2a glycoforms (2-fold dilutions, 12x), and ranging from 5.8 nM to 506 nM for human IgG1 glycoforms (1.5-fold dilutions, 12x). Regeneration after each sample was carried out with acid buffer (10 mM Gly-HCL pH 2.3). The system was calibrated for slight changes in refractive index due to environmental factors by buffer composition or temperature referencing the signals on spots to signals outside spots. Calculation of the apparent dissociation constant (KD)‘KDapparent’ was performed by fitting a 1:1 Langmuir binding model to the RU360values at each antibody concentration. For consistent re- porting, thesefits were carried out at each receptor density, and the final reported estimated KDs were calculated by interpolating to Rmax= 500 (Dekkers et al., 2017a; Schasfoort et al., 2016). In the case of mouse FcγRIII, anti-His association and dissociation curves were subtracted before calculation of IgG-binding affinity using SPRINT 1.9.4.4 software (IBIS technologies). Analysis and calculation of all binding data was carried out with Scrubber software version 2 (Biologic Software, Campbell, Australia).

2.9. Statistical analysis

All statistical analyses were performed using Graphpad Prism software version 7.02 (San Diego, CA, USA). Error bars represent one standard deviation and signiﬁcance was determined by a p-value <

0.05.

3. Results

3.1. Aﬃnity of human IgG glycovariants to human FcγRIIIa

To probe the effect of the N162-glycan found in human FcγRIII, we first generated recombinant hFcγRIIIa variants with or without the N- linked glycan (WT or N162A) as Fc-Fusion proteins, with the Fc part lacking binding affinity for FcγR (Fc: N297A), including a site-specific c-terminal biotin-label (Supplementary Fig. 1).

As reported previously, removing the N162-glycan by the N162A mutation resulted in an approximately 4-fold increased aﬃnity of un- modiﬁed fucosylated IgG1 to FcγRIIIa, with average KDapparent

from Fig. 1. Binding of human IgG1 glycoforms to FcγRIIIa^V158and FcγRIIIa^V158-N162A.

a) Apparent binding affinity (KDapparent) of unmodified human IgG1 to wild type Fc-FcγRIIIa^V158 and Fc-FcγRIIIa^V158-N162A and b) fold change in binding affinities of hIgG1 glycoforms relative to binding of unmodified IgG1. As measured by SPR, with both FcγR being coupled to streptavidin sensors through a c-terminal site-specific biotin tag, expressed as FcγR-IgA1-hinge-IgG2Fc-N297A fusion protein. Data represents means and S.D. of at least 3 individual measurements, * or *** denote a statistical significance of p < 0.05 or p < 0.001 respectively, as tested by unpaired t-test. U = unmodified IgG.

(4)

1.0 × 10⁻⁷ to 2.6 × 10⁻⁸nM respectively (Fig. 1a) (Ferrara et al., 2011), as analyzed by surface plasmon resonance (SPR). This mod- ification was known to hamper the discrimination of FcγRIIIa between fucosylated and afucosylated human IgG1. However, to test whether the FcγRIII-N162 glycan also affects the discrimination between galactosylated IgG. We generated IgG variants, engineering the bi-antennary glycan at the level of fucose and galactose as described previously (Dekkers et al., 2017b, 2016), resulting in 6 different glycoforms as summarized in Supplementary Table 1.

The relative affinity for Fc-FcγRIIIa WT was 5–10-fold higher after afucosylation, with a 2–3-fold variation in binding being observed for the galactose variants. However, for FcγRIIIa N162A, affinities were virtually independent of levels of fucosylation and galactosylation (Fig. 1b and Supplementary Fig. 2). This suggest the affinity-modula- tion by the hIgG glycan was solely dependent on the FcγRIII-glycan at position 162.

3.2. The N162-glycan site in hFcγRIII is conserved in mouse FcγRIV

We then investigated if a similar mechanism is operational in mice, and therefore analyzed all human and mouse FcγR for their presence of the N162-glycan found in both human FcγRIIIa and FcγRIIIb. This predicted glycosylation site was only found in the mouse orthologue receptor of hFcγRIIIa– mFcγRIV – but not in any other mouse FcγR (Fig. 2).

3.3. Aﬃnity of mouse IgG glycovariants to mouse FcγR

Next we tested if the combination of afucosylation and galactosylation also aﬀects binding to mouse FcγR in a similar way as in humans (Dekkers et al., 2017b) and, in particular, if the conserved FcγRIV-N162 glycan was also responsible for discriminating between fucosylated and afucosylated IgG (as seen inFig. 1). We therefore recreated those IgG- glycovariants as mouse IgG, using the same methods as before (Dekkers et al., 2017b, 2016). This resulted in 6 mIgG1 and 6 mIgG2a variants with high or decreased fucosylation and either decreased or increased galactosylation (Supplementary Tables 2 and 3).

For mouse IgG1 we detected only binding to mFcγRIIb and mFcγRIII. The aﬃnity of mIgG1 to these receptors was similar as reported earlier (Dekkers et al., 2017a; Nimmerjahn et al., 2005) (Fig. 3a). IgG produced with low amount of fucose, with high, inter- mediate or low levels of galactose of mIgG1 did not strongly inﬂuence its binding to neither receptor (Fig. 3b, Supplementary Fig. 3a and b).

mIgG2a bound all mouse FcγR and aﬃnity was also similar to what we and others reported earlier (Dekkers et al., 2017a; Nimmerjahn et al., 2005) (Fig. 3c). Similar variation in glycosylation of mIgG2a aﬀected

the aﬃnity to all mFcγR (Fig. 3d, Supplementary Fig. 3c–f), although the changes for FcγRI, FcγRIIb and FcγRIII were only minor (≤∼2x).

For mFcγRI the affinity seemed slightly decreased for afucosylated IgG, while it was slightly increased for mFcγRIIb and FcγRIII, also reported previously byNimmerjahn and Ravetch (2005). On the other hand, the increased affinity was especially pronounced for binding of afucosylated mIgG2a to mFcγRIV, but unlike the binding of human IgG1 to human FcγRIIIa (Fig. 1b), we observed no additional effect of galactosylation for binding of mIgG2a to mFcγRIV (Dekkers et al., 2017b) (Fig. 3d, Supplementary Fig. 3f).

To investigate if the N162-glycan in mFcγRIV has similar function as in hFcγRIIIa, we generated recombinant FcγRIV variants with- or without the N-linked glycan (WT or N162A) as Fc-Fusion proteins, comparable to the hFcγRIIIa Fc-Fusion constructs (Supplementary Fig. 4). Similar to hFcγRIIIa (Fig. 1a), an increase in binding (25x) was observed for fucosylated mIgG2a to the FcγRIV-N162A compared to WT FcγRIV (Fig. 3e and Supplementary Fig. 3g and h), suggesting that the FcγRIV-162-glycan also sterically interferes with recognition of fucosylated IgG.

We observed no increased binding aﬃnity to FcγRIV-without the N162-glycan after afucosylation of mIgG2a (Fig. 3f and Supplementary Fig. 3g and h). Collectively this strongly suggests the conserved N162- glycan found in the human FcγRIII family and its orthologue in mice to be the primary sensor discriminating between fucosylated and afucosylated IgG.

4. Discussion

The presence of the IgG-Fc glycan at position 297 is required for binding to both human and mouse FcγR in general (Shields et al., 2001;

Tao and Morrison, 1989). However, a more complex interaction takes place between human IgG and hFcγRIII (Ferrara et al., 2011), and as we show here, also for mouse IgG and mFcγRIV. This is because these two orthologous receptors have an glycan at position N162 that for humans has been shown to interact with the IgG-Fc glycan. Here we confirmed that the glycan composition of the Fc-N297 glycan is important for the affinity of human IgG to hFcγRIIIa and that this is conserved for the affinity of mouse IgG to mFcγRIV. The main effect seen was that low fucosylation increases the affinity for these receptors, which was fully attributable to the presence of the FcγR-N162 glycan. For humans we registered an additional effect of the IgG1-Fc galactosylation on affinity for hFcγRIIIa: highly galactosylated and afucosylated IgG1 binds sig- nificantly better than low galactosylated, afucosylated IgG. However, this additional effect of galactose was not found for the mouse IgG2a – mFcγRIV interaction.

Mouse IgG glycosylation is known to be a little more complex than that of humans. The bi-antennary glycan has the same core structure as that of human IgG, but can also be extended withα1,3-galactose (α-gal) and sialic acid in the form of N-glycolylneuraminic acid (Maresch and Altmann, 2016; Mimura et al., 2016). For this research we solely fo- cused on engineering the residues which are known to affect human FcγRIII binding (fucose and galactose). It would be interesting to investigate the effect of the mouse specific glycan additions, but this was not possible due to the lack of available glyco-engineering tools.

By deleting the N162 FcγR-glycan sites in both human and mouse FcγRIIIa and FcγRIV, respectively, we detected increased binding to fucosylated IgG (Shibata-Koyama et al., 2009). Even though afucosylation of IgG did not further increase binding aﬃnity to receptors lacking the N162-glycan, it did increase the aﬃnity to WT FcγRIIIa or FcγRIV beyond the apparent steric impediment asserted by the N162- FcγR glycan to fucosylated IgG. This strongly suggest that the presence of this N162-glycan allows these FcγR to discriminate between fucosylated and afucosylated IgG. The N162 FcγR-glycan does not seem to be able to further discriminate for the addition of galactose in neither the human IgG1/FcγRIIIa nor the mouse IgG2a/FcγRIV interaction.

Ferrara et al. showed that the discriminatory capability of the Fig. 2. Alignment human- and mouse FcγR sequences.

Aligned partial protein sequence of human and mouse FcγR, starting at amino acid position 136, encompassing amino acid position 162 of FcγRIIIa and partial interface with IgG, predicted N-glycosylation sites are indicated in bold, the consensus sequence for the conserved N162 glycosylation site in human FcγRIII/CD16 and mouse FcγRIV are indicated as bold and underlined residues.

(5)

FcγRIIIa through the N162-glycan is due to a direct carbohydrate-carbohydrate interaction, and allows for closer contact between receptor and ligand when the core-fucose is absent from the IgG N297 Fc glycan (Ferrara et al., 2011). Although we see similar range of affinity changes with afucosylated IgG, the apparent affinities we measure are slightly different than abovementioned study. This is likely due to the differ- ential experimental setup and/or possibly due to different states of bisection and galactosylation of IgG of which the latter is also known to affect binding to FcγRIII (Dashivets et al., 2015; Houde et al., 2010;

Subedi and Barb, 2016; Thomann et al., 2015), especially when afucosylated (Dekkers et al., 2017b; Houde et al., 2010). However, the level of IgG-galactosylation and other possible present glycan traits was not determined by Ferrara et al. (2011). In this study, we did not analyze the composition of the FcγR-N162 glycans which may also arguably aﬀect binding to IgG. Recently, it was shown that FcγRs

produced in diﬀerent cell lines vary in their glycan composition and this indeed aﬀected binding of therapeutic antibodies (Hayes et al., 2017).

Variation of FcγRIIIa glycosylation has been found to occur in vivo on different immune cells and may thus affect IgG affinity (Edberg and Kimberly, 1997; Huizinga et al., 1990), but the details of in vivo FcγR glycovariation are unfortunately unknown.

In this study, we made use of two different FcγRIV constructs, both with a C-terminal biotin tag. These differed, with the commercial product being a monomeric construct produced in an undisclosed human cell line and our in-house FcγRIV being a hIgG2-Fc-FcγR-fusion, unable to bind FcγR through its Fc portion because it lacks the N297-linked glycan (Dekkers et al., 2017b) and separating the Fc and FcγR portions spatially through a relatively long IgA1-hinge, produced in HEK Free- style cells. We have generated such Fc-Fusion constructs previously, without remarkable differences in affinities for IgG subclasses Fig. 3. Binding of mouse IgG glycoforms to mouse FcγR.

a) Apparent binding affinities (KDapparent) of unmodified mouse IgG1 to mouse FcγRs and b) fold change in binding affinities of mIgG1 glycoforms relative to binding of unmodified IgG1.

c) Apparent binding affinities (KDapparent) of unmodified mouse IgG2a to mouse FcγR and d) fold change in binding affinities of mIgG2a glycoforms relative to binding of unmodified IgG2a. e) Binding of unmodified mouse IgG2a to wild type Fc-mFcγRIV and Fc-mFcγRIV-N162A and f) fold change in binding affinities of mIgG2a glycoforms relative to binding of unmodified IgG2a. As measured by SPR, with all FcγR being coupled to streptavidin sensors through a c-terminal site-specific biotin tag except FcγRIII, which was captured using biotin coupled anti-HIS (see methods), and Fc-FcγRIVs expressed as FcγR-IgA1-hinge-IgG2Fc-N297A fusion protein. Data represents means and S.D. of at least 2 individual measurements, *, **,

*** and **** denote a statistical significance of p < 0.05, p ≤ 0.01, p ≤ 0.001 and p ≤ 0.0001, respectively, as tested by tested by unpaired t-test (a, c, e, f) or one-way ANOVA against unmodified IgG1, using Dunnett’s multiple comparisons test (b, d). U = unmodified IgG.

(6)

compared to those reported by others (Bruggeman et al., 2017; Dekkers et al., 2017a,b). Here, however, we noticed a 3x increased affinity of afucosylated IgG2a for the Fc-FcγRIV WT construct compared to the monomeric produced FcγRIV, although the qualitative discrimination between the binding to fucosylated and afucosylated IgG was observed for both constructs. The elevated affinity of our Fc-FcγRIV might either be due to the altered configuration but also due to differences in glycan composition between the two FcγRIV constructs (Edberg and Kimberly, 1997; Zeck et al., 2011).

As hFcγRIIIa and mFcγRIV have additional glycans besides the N162-glycan it is possible that additional glycans also influence the affinity, although their effect appears minor and do not seem to affect the discrimination of afucosylated IgG1 (Shibata-Koyama et al., 2009).

The importance of the FcγR glycosylation clearly needs to be further addressed by future studies.

In summary, here we describe that the N162 found in hFcγRIIIa, which is known to provide this receptor with its discriminatory capacity for fucosylated and afucosylated IgG, is functionally conserved in mFcγRIV. Although we and others have recently described that galactosylation further increases the aﬃnity of afucosylated human IgG to FcγRIII (Dekkers et al., 2017b; Houde et al., 2010), this is not the case for mFcγRIV. The conserved discrimination between fucosylated and afucosylated IgG in humans and mice through a single glycan in the orthologous receptors FcγRIIIa and FcγRIV in mice and humans, respectively, is interesting in the light of recent data showing that human immune responses can be mounted with afucosylated antibodies (Ackerman et al., 2013; Kapur et al., 2015, 2014a; Sonneveld et al., 2017, 2016; Wang et al., 2017). This suggest a mechanism to be in place that can trigger exceptionally strong eﬀector functions in immune responses, now described not only for allo-immune responses against red blood cells and platelets after transfusion (Kapur et al., 2015; Sonneveld et al., 2017) and in pregnancy (Kapur et al., 2014a,b; Sonneveld et al., 2016; Wuhrer et al., 2009), but also HIV (Ackerman et al., 2013) and Dengue (Wang et al., 2017). Understanding of those mechanisms will allow us to tap into these to generate more potent immune responses, or prevent them when inappropriately triggered.

5. Conclusions

The conserved N162 glycan in human FcγRIIIa and mouse FcγRIV is responsible for the discrimination between fucosylated and afucosylated IgG Fc glycoforms. For both species afucosylated binds with much higher aﬃnity to these homologous receptors. In humans, additional galactosylation of afucosylated IgG Fc further directs strength of aﬃ- nity, this is not observed for mouse afucosylated IgG Fc.

Disclosure of potential conﬂicts of interest

The authors declare no potential conﬂict of interest.

Author contributions

G.D., T.R., and G.V. designed the study, G.D., A.E.H.B., R.P., R.V., C.A.M, A.B., J.Y.M. performed experiments, G.D., A.E.H.B., R.P., C.A.M., T.R. and G.V. analyzed the data, J.Y.M. and W.J.E.E. provided vital reagents, M.W., T.R., and G.V. supervised the study.

Funding

Gestur Vidarsson: Sanquin Product and Process Development, # 12- 001.

Acknowledgement

The authors would like to thank David Falck for helpful discussions.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, athttps://doi.org/10.1016/j.molimm.2017.12.006.

References

Ackerman, M.E., Crispin, M., Yu, X., Baruah, K., Boesch, A.W., Harvey, D.J., Dugast, A.-S., Heizen, E.L., Ercan, A., Choi, I., Streeck, H., Nigrovic, P.A., Bailey-Kellogg, C., Scanlan, C., Alter, G., 2013. Natural variation in Fc glycosylation of HIV-speciﬁc antibodies impacts antiviral activity. J. Clin. Invest. 123, 2183–2192.http://dx.doi.

org/10.1172/JCI65708.

Bruggeman, C.W., Dekkers, G., Bentlage, A.E.H., Treffers, L.W., Nagelkerke, S.Q., Lissenberg-Thunnissen, S., Koeleman, C.A.M., Wuhrer, M., van den Berg, T.K., Rispens, T., Vidarsson, G., Kuijpers, T.W., 2017. Enhanced effector functions due to antibody defucosylation depend on the effector cell Fcγ receptor profile. J. Immunol.

199, 204–211.http://dx.doi.org/10.4049/jimmunol.1700116.

Bruhns, P., Jönsson, F., 2015. Mouse and human FcR eﬀector functions. Immunol. Rev.

268, 25–51.http://dx.doi.org/10.1111/imr.12350.

Chambers, M.C., Maclean, B., Burke, R., Amodei, D., Ruderman, D.L., Neumann, S., Gatto, L., Fischer, B., Pratt, B., Egertson, J., Hoﬀ, K., Kessner, D., Tasman, N., Shulman, N., Frewen, B., Baker, T.a., Brusniak, M.-Y., Paulse, C., Creasy, D., Flashner, L., Kani, K., Moulding, C., Seymour, S.L., Nuwaysir, L.M., Lefebvre, B., Kuhlmann, F., Roark, J., Rainer, P., Detlev, S., Hemenway, T., Huhmer, A., Langridge, J., Connolly, B., Chadick, T., Holly, K., Eckels, J., Deutsch, E.W., Moritz, R.L., Katz, J.E., Agus, D.B., MacCoss, M., Tabb, D.L., Mallick, P., 2012. A cross-platform toolkit for mass spectrometry and proteomics. Nat. Biotechnol. 30, 918–920.http://dx.doi.org/10.1038/

nbt.2377.

Dashivets, T., Thomann, M., Rueger, P., Knaupp, A., Buchner, J., Schlothauer, T., 2015.

Multi-angle eﬀector function analysis of human monoclonal IgG glycovariants. PLoS One 10, e0143520.http://dx.doi.org/10.1371/journal.pone.0143520.

de Haan, N., Reiding, K.R., Krištić, J., Hipgrave Ederveen, A.L., Lauc, G., Wuhrer, M., 2017. The N-glycosylation of mouse immunoglobulin G (IgG)-fragment crystallizable diﬀers between IgG subclasses and strains. Front. Immunol. 8, 1538–1547.http://dx.

doi.org/10.3389/ﬁmmu.2017.00608.

de Lau, W., Barker, N., Low, T.Y., Koo, B.-K., Li, V.S.W., Teunissen, H., Kujala, P., Haegebarth, A., Peters, P.J., van de Wetering, M., Stange, D.E., van Es, J.E., Guardavaccaro, D., Schasfoort, R.B.M., Mohri, Y., Nishimori, K., Mohammed, S., Heck, A.J.R., Clevers, H., 2011. Lgr5 homologues associate with Wnt receptors and mediate R-spondin signalling. Nature 476, 293–297.http://dx.doi.org/10.1038/

nature10337.

Dekkers, G., Plomp, R., Koeleman, C.A.M., Visser, R., von Horsten, H.H., Sandig, V., Rispens, T., Wuhrer, M., Vidarsson, G., 2016. Multi-level glyco-engineering techni- ques to generate IgG with deﬁned Fc-glycans. Sci. Rep. 6, 36964.http://dx.doi.org/

10.1038/srep36964.

Dekkers, G., Bentlage, A.E.H., Stegmann, T.C., Howie, H.L., Lissenberg-Thunnissen, S., Zimring, J., Rispens, T., Vidarsson, G., 2017a. Aﬃnity of human IgG subclasses to mouse Fc gamma receptors. MAbs 9, 767–773.http://dx.doi.org/10.1080/

19420862.2017.1323159.

Dekkers, G., Treﬀers, L., Plomp, R., Bentlage, A.E.H., de Boer, M., Koeleman, C.A.M., Lissenberg-Thunnissen, S.N., Visser, R., Brouwer, M., Mok, J.Y., Matlung, H., van den Berg, T.K., van Esch, W.J.E., Kuijpers, T.W., Wouters, D., Rispens, T., Wuhrer, M., Vidarsson, G., 2017b. Decoding the human immunoglobulin G-glycan repertoire re- veals a spectrum of Fc-receptor- and complement-mediated-eﬀector activities. Front.

Immunol. 8, 877.http://dx.doi.org/10.3389/ﬁmmu.2017.00877.

Edberg, J.C., Kimberly, R.P., 1997. Cell type-speciﬁc glycoforms of Fc gamma RIIIa (CD16): diﬀerential ligand binding. J. Immunol. 159, 3849–3857.

Ferrara, C., Stuart, F., Sondermann, P., Brünker, P., Umaña, P., 2006. The carbohydrate at FcgammaRIIIa Asn-162. An element required for high aﬃnity binding to non-fucosylated IgG glycoforms. J. Biol. Chem. 281, 5032–5036.http://dx.doi.org/10.1074/

jbc.M510171200.

Ferrara, C., Grau, S., Jäger, C., Sondermann, P., Brünker, P., Waldhauer, I., Hennig, M., Ruf, A., Rufer, A.C., Stihle, M., Umaña, P., Benz, J., 2011. Unique carbohydrate- carbohydrate interactions are required for high aﬃnity binding between FcgammaRIII and antibodies lacking core fucose. Proc. Natl. Acad. Sci. U. S. A. 108, 12669–12674.http://dx.doi.org/10.1073/pnas.1108455108.

Hayes, J.M., Frostell, A., Karlsson, R., Müller, S., Millan-Martin, S., Pauers, M., Reuss, F., Cosgrave, E., Anneren, C., Davey, G.P., Rudd, P.M., 2017. Identiﬁcation of Fc gamma receptor glycoforms that produce diﬀerential binding kinetics for rituximab. Mol.

Cell. Proteom.http://dx.doi.org/10.1074/mcp.M117.066944.mcp.M117.066944.

Houde, D., Peng, Y., Berkowitz, S.a., Engen, J.R., 2010. Post-translational modifications differentially affect IgG1 conformation and receptor binding. Mol. Cell. Proteom. 9, 1716–1728.http://dx.doi.org/10.1074/mcp.M900540-MCP200.

Howie, H.L., Delaney, M., Wang, X., Er, L.S., Vidarsson, G., Stegmann, T.C., Kapp, L., Lebedev, J.N., Wu, Y., AuBuchon, J.P., Zimring, J.C., 2016. Serological blind spots for variants of human IgG3 and IgG4 by a commonly used anti-immunoglobulin reagent.

Transfusion 56, 2953–2962.http://dx.doi.org/10.1111/trf.13812.

Huizinga, T.W., Kleijer, M., Tetteroo, P.A., Roos, D., von dem Borne, A.E., 1990. Biallelic neutrophil Na-antigen system is associated with a polymorphism on the phospho- inositol-linked Fc gamma receptor III (CD16). Blood 75, 213–217.

Jeﬀeris, R., 2009. Recombinant antibody therapeutics: the impact of glycosylation on mechanisms of action. Trends Pharmacol. Sci. 30, 356–362.http://dx.doi.org/10.

1016/j.tips.2009.04.007.

Kaneko, Y., Nimmerjahn, F., Ravetch, J.V., 2006. Anti-inﬂammatory activity of

(7)

immunoglobulin G resulting from Fc sialylation. Science 313, 670–673.http://dx.

doi.org/10.1126/science.1129594.

Kapur, R., Della Valle, L., Sonneveld, M., Hipgrave Ederveen, A., Visser, R., Ligthart, P., de Haas, M., Wuhrer, M., van der Schoot, C.E., Vidarsson, G., 2014a. Low anti-RhD IgG-Fc-fucosylation in pregnancy: a new variable predicting severity in haemolytic disease of the fetus and newborn. Br. J. Haematol. 166, 936–945.http://dx.doi.org/

10.1111/bjh.12965.

Kapur, R., Kustiawan, I., Vestrheim, A., Koeleman, C.A.M., Visser, R., Einarsdottir, H.K., Porcelijn, L., Jackson, D., Kumpel, B., Deelder, A.M., Blank, D., Skogen, B., Killie, M.K., Michaelsen, T.E., de Haas, M., Rispens, T., van der Schoot, C.E., Wuhrer, M., Vidarsson, G., 2014b. A prominent lack of IgG1-Fc fucosylation of

platelet alloantibodies in pregnancy. Blood 123, 471–480.http://dx.doi.org/10.

1182/blood-2013-09-527978.

Kapur, R., Della Valle, L., Verhagen, O.J.H.M., Hipgrave Ederveen, A., Ligthart, P., de Haas, M., Kumpel, B., Wuhrer, M., van der Schoot, C.E., Vidarsson, G., 2015.

Prophylactic anti-D preparations display variable decreases in Fc-fucosylation of anti- D. Transfusion 55, 553–562.http://dx.doi.org/10.1111/trf.12880.

Maresch, D., Altmann, F., 2016. Isotype-speciﬁc glycosylation analysis of mouse IgG by LC-MS. Proteomics 16, 1321–1330.http://dx.doi.org/10.1002/pmic.201500367.

Mimura, Y., Kelly, R.M., Unwin, L., Albrecht, S., Jeﬀeris, R., Goodall, M., Mizukami, Y., Mimura-Kimura, Y., Matsumoto, T., Ueoka, H., Rudd, P.M., 2016. Enhanced sialylation of a human chimeric IgG1 variant produced in human and rodent cell lines. J.

Immunol. Methods 428, 30–36.http://dx.doi.org/10.1016/j.jim.2015.11.009.

Nimmerjahn, F., Ravetch, J.V., 2005. Divergent immunoglobulin g subclass activity through selective Fc receptor binding. Science 310, 1510–1512.http://dx.doi.org/

10.1126/science.1118948.

Nimmerjahn, F., Bruhns, P., Horiuchi, K., Ravetch, J.V., 2005. FcyRIV: a novel FcR with distinct IgG subclass speciﬁcity. Immunity 23, 41–51.http://dx.doi.org/10.1016/j.

immuni.2005.05.010.

Nimmerjahn, F., Anthony, R.M., Ravetch, J.V., 2007. Agalactosylated IgG antibodies depend on cellular Fc receptors for in vivo activity. Proc. Natl. Acad. Sci. U. S. A. 104, 8433–8437.http://dx.doi.org/10.1073/pnas.0702936104.

Plomp, R., Dekkers, G., Rombouts, Y., Visser, R., Koeleman, C.A.M., Kammeijer, G.S.M., Jansen, B.C., Rispens, T., Hensbergen, P.J., Vidarsson, G., Wuhrer, M., 2015. Hinge- region O-glycosylation of human immunoglobulin G3 (IgG3). Mol. Cell. Proteom. 14, 1373–1384.http://dx.doi.org/10.1074/mcp.M114.047381.

Rodenko, B., Toebes, M., Hadrup, S.R., van Esch, W.J.E., Molenaar, A.M., Schumacher, T.N.M., Ovaa, H., 2006. Generation of peptide-MHC class I complexes through UV- mediated ligand exchange. Nat. Protoc. 1, 1120–1132.http://dx.doi.org/10.1038/

nprot.2006.121.

Schasfoort, R.B.M., Andree, K.C., van der Velde, N., van der Kooi, A., Stojanović, I., Terstappen, L.W.M.M., 2016. Interpolation method for accurate aﬃnity ranking of arrayed ligand-analyte interactions. Anal. Biochem. 500, 21–23.http://dx.doi.org/

10.1016/j.ab.2016.01.023.

Shibata-Koyama, M., Iida, S., Okazaki, A., Mori, K., Kitajima-Miyama, K., Saitou, S., Kakita, S., Kanda, Y., Shitara, K., Kato, K., Satoh, M., 2009. The N-linked oligosaccharide at Fc gamma RIIIa Asn-45: an inhibitory element for high Fc gamma RIIIa binding aﬃnity to IgG glycoforms lacking core fucosylation. Glycobiology 19, 126–134.http://dx.doi.org/10.1093/glycob/cwn110.

Shields, R.L., Namenuk, A.K., Hong, K., Meng, Y.G., Rae, J., Briggs, J., Xie, D., Lai, J., Stadlen, A., Li, B., Fox, J.a., Presta, L.G., 2001. High resolution mapping of the binding site on human IgG1 for Fc gamma RI, Fc gamma RII, Fc gamma RIII, and FcRn and design of IgG1 variants with improved binding to the Fc gamma R. J. Biol.

Chem. 276, 6591–6604.http://dx.doi.org/10.1074/jbc.M009483200.

Shields, R.L., Lai, J., Keck, R., O’Connell, L.Y., Hong, K., Meng, Y.G., Weikert, S.H.a., Presta, L.G., 2002. Lack of fucose on human IgG1 N-linked oligosaccharide improves

binding to human Fcgamma RIII and antibody-dependent cellular toxicity. J. Biol.

Chem. 277, 26733–26740.http://dx.doi.org/10.1074/jbc.M202069200.

Shinkawa, T., Nakamura, K., Yamane, N., Shoji-Hosaka, E., Kanda, Y., Sakurada, M., Uchida, K., Anazawa, H., Satoh, M., Yamasaki, M., Hanai, N., Shitara, K., 2003. The absence of fucose but not the presence of galactose or bisecting N-acetylglucosamine of human IgG1 complex-type oligosaccharides shows the critical role of enhancing antibody-dependent cellular cytotoxicity. J. Biol. Chem. 278, 3466–3473.http://dx.

doi.org/10.1074/jbc.M210665200.

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

dx.doi.org/10.1038/35018508.

Sonneveld, M.E., Natunen, S., Sainio, S., Koeleman, C.A.M., Holst, S., Dekkers, G., Koelewijn, J., Partanen, J., van der Schoot, C.E., Wuhrer, M., Vidarsson, G., 2016.

Glycosylation pattern of anti-platelet IgG is stable during pregnancy and predicts clinical outcome in alloimmune thrombocytopenia. Br. J. Haematol. 174, 310–320.

http://dx.doi.org/10.1111/bjh.14053.

Sonneveld, M.E., Koelewijn, J., de Haas, M., Admiraal, J., Plomp, R., Koeleman, C.A.M., Hipgrave Ederveen, A.L., Ligthart, P., Wuhrer, M., van der Schoot, C.E., Vidarsson, G., 2017. Antigen speciﬁcity determines anti-red blood cell IgG-Fc alloantibody glycosylation and thereby severity of haemolytic disease of the fetus and newborn.

Br. J. Haematol. 176, 651–660.http://dx.doi.org/10.1111/bjh.14438.

Subedi, G.P., Barb, A.W., 2015. The structural role of antibody N-glycosylation in receptor interactions. Structure 23, 1573–1583.http://dx.doi.org/10.1016/j.str.2015.

06.015.

Subedi, G.P., Barb, A.W., 2016. The immunoglobulin G1 N-glycan composition aﬀects binding to each low aﬃnity Fc γ receptor. MAbs 8, 1512–1524.http://dx.doi.org/10.

1080/19420862.2016.1218586.

Tao, M.H., Morrison, S.L., 1989. Studies of aglycosylated chimeric mouse-human IgG: role of carbohydrate in the structure and eﬀector functions mediated by the human IgG constant region. J. Immunol. 143, 2595–2601.

Thomann, M., Schlothauer, T., Dashivets, T., Malik, S., Avenal, C., Bulau, P., Rüger, P., Reusch, D., 2015. In vitro glycoengineering of IgG1 and its eﬀect on Fc receptor binding and ADCC activity. PLoS One 10, e0134949.http://dx.doi.org/10.1371/

journal.pone.0134949.

Vink, T., Oudshoorn-Dickmann, M., Roza, M., Reitsma, J.J., de Jong, R.N., 2014. A simple, robust and highly eﬃcient transient expression system for producing antibodies. Methods 65, 5–10.http://dx.doi.org/10.1016/j.ymeth.2013.07.018.

Wang, T.T., Sewatanon, J., Memoli, M.J., Wrammert, J., Bournazos, S., Bhaumik, S.K., Pinsky, B.A., Chokephaibulkit, K., Onlamoon, N., Pattanapanyasat, K., Taubenberger, J.K., Ahmed, R., Ravetch, J.V., 2017. IgG antibodies to dengue enhanced for FcγRIIIA binding determine disease severity. Science 355, 395–398.http://dx.doi.org/10.

1126/science.aai8128.

Wuhrer, M., Porcelijn, L., Kapur, R., Koeleman, C.A.M., Deelder, A., de Haas, M., Vidarsson, G., 2009. Regulated glycosylation patterns of IgG during alloimmune responses against human platelet antigens. J. Proteome Res. 8, 450–456.http://dx.doi.

org/10.1021/pr800651j.

Yasuma, R., Cicatiello, V., Mizutani, T., Tudisco, L., Kim, Y., Tarallo, V., Bogdanovich, S., Hirano, Y., Kerur, N., Li, S., Yasuma, T., Fowler, B.J., Wright, C.B., Apicella, I., Greco, A., Brunetti, A., Ambati, B.K., Helmers, S.B., Lundberg, I.E., Viklicky, O., Leusen, J.H., Verbeek, J.S., Gelfand, B.D., Bastos-Carvalho, A., De Falco, S., Ambati, J., 2016.

Intravenous immune globulin suppresses angiogenesis in mice and humans. Signal Transduct. Target. Ther. 1, 15002.http://dx.doi.org/10.1038/sigtrans.2015.2.

Zeck, A., Pohlentz, G., Schlothauer, T., Peter-Katalinić, J., Regula, J.T., 2011. Cell type- speciﬁc and site directed N-glycosylation pattern of Fc RIIIa. J. Proteome Res. 10, 3031–3039.http://dx.doi.org/10.1021/pr1012653.