Haplotypes of FcγRIIa and FcγRIIIb Polymorphic Variants Influence IgG-Mediated Responses in Neutrophils

This information is current as Joris van der Heijden, Sietse Nagelkerke, Xiwen Zhao, Judy of September 28, 2021. Geissler, Theo Rispens, Timo K. van den Berg and Taco W. Kuijpers J Immunol published online 19 February 2014 http://www.jimmunol.org/content/early/2014/02/18/jimmun

ol.1203570 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2014/02/19/jimmunol.120357 Material 0.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 28, 2021

*average

Subscription Information about subscribing to The Journal of is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2014 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published February 19, 2014, doi:10.4049/jimmunol.1203570 The Journal of Immunology

Haplotypes of FcgRIIa and FcgRIIIb Polymorphic Variants Influence IgG-Mediated Responses in Neutrophils

Joris van der Heijden,* Sietse Nagelkerke,* Xiwen Zhao,* Judy Geissler,* Theo Rispens,† Timo K. van den Berg,* and Taco W. Kuijpers*,‡

Human blood neutrophils normally express two FcgRs (FcgRIIa and FcgRIIIb) that, upon multivalent binding of IgG in immune complexes or on opsonized targets, mediate responses such as phagocytosis, Ab-dependent cellular cytotoxicity, and respiratory burst. Allelic variants have been described for both FcgRIIa (131H/R) and FcgRIIIb (NA1/NA2/SH), with different binding affinity for IgG subclasses. Because neither of these variants acts alone, we have set out to systematically analyze in a large cohort of healthy FCGR2/3-genotyped volunteers how the different haplotypes of neutrophil FcgRs functionally interact. Maximal IgG-

induced H2O2 production by neutrophils from individuals with different (homozygous) haplotypes was observed in the following

order: 131HH-NA2NA2 > 131RR-NA1NA1 > 131HH-NA1NA1 > 131RR-NA2NA2. Although FcgRIIa 131H is known to bind Downloaded from IgG1 and IgG2 more avidly, no such differences in affinity are known for FcgRIIIb variants. Nonetheless, a remarkable impact of the FcgRIIIb variants on IgG-mediated neutrophil activity was thus demonstrated, which was not explained by differences in FcgR surface expression. The FcgR expression profile was changed by overnight G-CSF/IFN-g activation of the neutrophils and eliminated any haplotypic impact on the respiratory burst. To our knowledge, our results are the first to provide an integrated functional analysis of neutrophil FcgR haplotypes and suggest that particularly the early phase of IgG-mediated neutrophil

reactivity is influenced by FCGR2/3 genotypic variation. The Journal of Immunology, 2014, 192: 000–000. http://www.jimmunol.org/

he FcgRs are receptors for IgG and they are expressed on Functionally relevant polymorphic variants have been described virtually all immune cells. These receptors mediate var- for all low-affinity receptors (1–6), and copy number var- T ious cellular responses to IgG, including phagocytosis, iation (CNV) is found in the encoding FcgRIIc, FcgRIIIa, Ab-dependent cellular cytotoxicity, and respiratory burst. In hu- and FcgRIIIb, but not FcgRIIa and FcgRIIb (7). mans, there are three types of FcgR. Based on their affinity for Neutrophils constitute the first and major line of human defense monomeric IgG, they can be divided into the high-affinity FcgRI against bacterial and fungal pathogens. These cells constitutively and the low-affinity FcgRII and FcgRIII. FcgRII exists in three express two main FcgRs, that is, FcgRIIa and FcgRIIIb, whereas by guest on September 28, 2021 isoforms: FcgRIIa, FcgRIIb, and FcgRIIc. FcgRIII exists as FcgRI is only expressed on activated cells. FcgRIIb is not expressed a transmembrane FcgRIIIa and as a GPI-linked FcgRIIIb. Sig- on circulating neutrophils. In contrast, FcgRIIc can be detected naling by FcgR is mediated by ITAM or ITIM motifs. ITAMs are in 15–20% of healthy individuals, because they carry an FCGR2C found either in adaptor such as FcRg, which associates open reading frame instead of the more common pseudogene with with FcgRI and FcgRIIIa, or within the cytoplasmic tail of the astopcodoninexon3(3,8,9). FcgR itself, in the cases of FcgRIIa and FcgRIIc. The only in- Apart from this basic make-up of IgG receptors of human hibitory FcgRisFcgRIIb, as it has an ITIM in its cytoplasmic tail. neutrophils, functionally important single nucleotide polymor- The genes encoding for the low-affinity FcgR are found in one phisms (SNPs) have been described for the major FcgRs on cluster located on 1q23 and are highly polymorphic. neutrophils. For instance, an SNP in FCGR2A results in either a histidine or an arginine at position 131 (131H/R), located in the ligand-binding domain of FcgRIIa (1). FcgRIIa-H131 has a higher *Department of Blood Cell Research, Sanquin Research and Landsteiner Laboratory, Academic Medical Center, University of Amsterdam, 1105 AZ Amsterdam, The binding affinity for IgG1 and more in particular for IgG2, when Netherlands; †Department of Immunopathology, Sanquin Research and Landsteiner compared with FcgRIIa-R131, whereas binding of IgG3 and IgG4 Laboratory, Academic Medical Center, University of Amsterdam, 1105 AZ Amster- are similar (10). Functionally, phagocytes from individuals ho- dam, The Netherlands; and ‡Department of Pediatric Hematology, Immunology and Infectious Disease, Emma Children’s Hospital, Academic Medical Center, University mozygous for H131 (FcgRIIa-131HH) have an increased capacity of Amsterdam, 1105 AZ Amsterdam, The Netherlands to respond with phagocytosis, degranulation, or cytokine release Received for publication January 14, 2013. Accepted for publication January 16, in response to IgG-opsonized bacteria or erythrocytes, when 2014. compared with cells from FcgRIIa-131RR individuals (11, 12). This work was supported by Sanquin and Landsteiner Foundation for Blood Trans- Moreover, the neutrophil-specific FcgRIIIb has three polymorphic fusion Research Grant LSBR2009-0916. variants, NA1, NA2, and SH, also known as HNA1a, -1b, and -1c, Address correspondence and reprint requests to Prof. Taco W. Kuijpers, Department respectively. FcgRIIIb-NA1 and -NA2 nucleotide sequences differ of Pediatric Hematology, Immunology and Infectious Disease, Emma Children’s Hospital, Academic Medical Center, Room H7-230, Meibergdreef 9, 1105 AZ at five positions, resulting in four amino acid differences. As a con- Amsterdam, The Netherlands. E-mail address: [email protected] sequence, the NA2 variant has two additional N-linked glycosylation The online version of this article contains supplemental material. sites compared with NA1 (5). The SH variant differs from NA2 Abbreviations used in this article: CNV, ; MLPA, multiplex at only one additional position, resulting in an A→Daminoacid ligation–dependent probe amplification; ROS, reactive oxygen species; SNP, single change that may affect the tertiary structure of the (6). Although nucleotide polymorphism. binding affinities for IgG1 and IgG3 are similar for all three variants Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00 (10), functionally, neutrophils from FcgRIIIb-NA1NA1 individuals

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1203570 2 NEUTROPHIL FcgR HAPLOTYPES IN THE RESPONSE TO IgG bind and phagocytize IgG-opsonized bacteria and IgG-sensitized FCGR2C gene and a nonspecific FCGR2B/C probe to detect the open erythrocytes more efficiently than do those from NA1NA2 and reading frame in exon 3. NA2NA2 individuals (12, 13). Apart from the differences in bind- Isolation and activation of neutrophils ing and function, both FCGR2A and FCGR3B polymorphisms have been associated with numerous autoimmune and infectious dis- Granulocytes were isolated from whole blood by Percoll density gradient eases (14). centrifugation. As determined by flow cytometry, neutrophil purity was always .95% and FcgRI expression was always absent, indicating that the Because combinations of FcgR isoforms rather than a single IgG cells had not been preactivated in vivo (for instance by an intercurrent receptor type are expressed on neutrophils, we studied the response infection or cold). For activation, neutrophils were cultured for 16 h at of neutrophils to IgG in a large cohort of healthy FCGR2/3-geno- a concentration of 5 3 106 cells/ml in complete medium, supplemented typed volunteers expressing different FcgRIIa/FcgRIIIb haplotypes. with 10 ng/ml clinical grade G-CSF (Neupogen; Amgen, Breda, The Netherlands) and 50 ng/ml recombinant human IFN-g (PeproTech, Rocky As demonstrated for the respiratory burst in neutrophils, the cellular Hill, NJ). activation by IgG was not determined by a single SNP but instead by the combination of SNPs in a given haplotype. NADPH oxidase activity assay NADPH oxidase activity was compared between neutrophils from donors Materials and Methods with different FcgRIIa/FcgRIIIb combinations by measuring reactive ox- Subjects ygen species (ROS) as H2O2 production for 60 min in an Amplex Red assay (Invitrogen, Carlsbad, CA), essentially as described before (17). Healthy adult white volunteers (n = 171) volunteered to donate whole Cells were compared unstimulated or stimulated with either 100 ng/ml blood. These donors were not known to have hematological disorders of PMA or 17.5 mg/ml IgG. Neutrophils stimulated with IgG were also Downloaded from any kind in the past or at present. The study was approved by the Medical tested in the presence of blocking Abs against FcgRI, FcgRII, and FcgRIII, Ethics Committee of the Academic Medical Center and Sanquin in Amsterdam, which were added at 10 mg/ml final concentration each. Where indicated and it was performed in accordance with the Declaration of Helsinki. the same mAbs were tested as F(ab9)2 fragments to avoid any potential influence of the Fc domain (referred to as the so-called Kurlander effect) Abs and reagents (18) and to confirm the genuinely blocking capacity of the mAbs used. The following mAbs were used in flow cytometry: anti-FcgRI clone 10.1 F(ab9)2 fragments were used at an optimal dose of 7 mg/ml final concen- (mouse IgG1; BD Pharmingen, San Diego, CA), anti-FcgRII clone AT10 tration. (mouse IgG1; AbD Serotec, Oxford, U.K.), anti-FcgRIIb/c clone 2B6 For each experiment, the maximum slope of H2O2 production was de- http://www.jimmunol.org/ (mouse IgG1; a gift of Macrogenics, Rockville, MD) (15), and anti- termined by calculating the maximum value of the derivative of the H2O2 FcgRIII clone 3G8 (mouse IgG1; BD Pharmingen). Blocking Abs used in concentration over time. Measurements were performed every 30 s. The NADPH oxidase assays were anti-FcgRI (mouse IgG1, low-endotoxin maximum slope was based on five sequential time points. azide-free, clone 10.1; BioLegend, San Diego, CA), anti-FcgRII clone AT10 (mouse IgG1, azide-free; AbD Serotec), and anti-FcgRIII clone 3G8 Statistical analysis (mouse IgG1, low-endotoxin azide-free; BioLegend). F(ab9) fragments 2 Flow cytometry and NADPH oxidase activity data were analyzed with a of AT10 were prepared from the intact Ab mentioned above by pepsin Student t test; p , 0.05 was considered statistically significant. Data are 9 a g cleavage at pH 3.5 for 20 h; F(ab )2 fragments of -Fc RIII were from expressed as means 6 SEM. Ancell (anti-human CD16 F(ab9)2; Ancell, Bayport, MN) The purity of the

F(ab9)2 fragments was confirmed by SDS-PAGE gel electrophoresis and by guest on September 28, 2021 Coomassie staining. Results Annexin V (BD Pharmingen) was used to determine neutrophil apoptosis status. Haplotypes of FcgRIIa and FcgRIIIb polymorphisms HEPES+ buffer contained 132 mM NaCl, 6 mM KCl, 1 mM CaCl2, determine neutrophil response to IgG 1 mM MgSO4, 1.2 mM potassium phosphate, 20 mM HEPES, 5.5 mM We used clinical-grade IgG preparations for use in patients to glucose, and 0.5% (w/v) human serum albumin (pH 7.4). Complete medium used for overnight culturing was IMDM (PAA Lab- activate human neutrophils in vitro isolated from healthy volun- oratories, Pasching, Austria) supplemented with 10% (v/v) FCS (Bodinco, teers. The protein fraction of these IgG preparations consisted of Alkmaar, The Netherlands), penicillin (100 U/ml), streptomycin (100 mg/ml; .90% IgG (54.0% IgG1, 39.6% IgG2, 2.7% IgG3, and 3.8% PAA Laboratories), and 2 mM L-glutamine (PAA Laboratories). IgG4), with minor contaminations of IgA (2.4% of total protein High-performance size exclusion chromatography content) and traces of other plasma proteins. The IgG preparations contained mainly monomeric IgG (68%), but also some dimers Ig products were analyzed for IgG monomer, IgG dimer, and IgG polymer (28.4%) and a small amount of polymers (3%), as was determined content using high-performance size exclusion chromatography. Samples weredilutedto5mg/mlinPBSandimmediately thereafter applied to by high-performance size exclusion chromatography (Supplemental a Superdex 200 HR 10/300 column (Amersham Biosciences, Uppsala, Fig. 1). For reasons of simplicity, we have used in our studies Sweden) connected to an HPLC system (A¨ KTAexplorer, Amersham Bio- preparations for s.c. use only, which gave a slightly higher neutrophil sciences) at room temperature with a flow rate of 0.5 ml/min and PBS as response when compared with the i.v. Ig preparations of the same running buffer. A computer program (Unicorn version 5.11) was used to integrate the areas under the curve of the peaks shown in the chromato- plasma source and with similar Ig composition (data not shown). grams. For estimation of protein size, the column was calibrated with a Using our MLPA for FCGR2 and FCGR3 genes, we have high molecular weight calibration (GE Healthcare). genotyped CNV and SNPs within the genes coding for the dif- ferent low-affinity FcgRIIa and FcgRIIIb in 171 healthy adult Genotyping white individuals. Subsequently, we divided these individuals Genomic DNA was isolated from whole blood with the Gentra Puregene kit into groups, representing different combinations, or haplotypes, of (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. polymorphic variants of FcgRIIa (131H/R) and FcgRIIIb (NA1/ CNV and SNPs in the low-affinity FCGR genes FCGR2A, FCGR2B, NA2), the two FcgRs expressed on resting neutrophils. To prevent FCGR2C, FCGR3A, and FCGR3B were determined with an FCGR-spe- cific multiplex ligation–dependent probe amplification (MLPA) assay. The creating too many groups and groups with insufficient numbers of MLPA assay was performed essentially as described previously (7, 16). individuals, we selected the 117 individuals that had two copies The FCGR MLPA includes gene-specific probes to determine the CNV of of any gene at the FCGR2/3 locus (in particular for FCGR3B the genes. It also includes probes to detect the following SNPs: FCGR2A encoding neutrophil-specific FcgRIIIb) and were also genotyped 27Q/W (rs9427397 plus rs9427398), FCGR2A 131H/R (rs1801274), FCGR2B 232I/T (rs1050501), FCGR2B/C-386C/G (rs3219018) and -120A/T (rs34701572), as FCGR2C-Stop (hence, without FcgRIIc expression on their FCGR3A 158V/F (rs396991), and FCGR3B (HNA1a/HNA1b/HNA1c). The neutrophils) (9). The distribution of the resulting nine haplotypes assay also contains a probe specific for the stop codon in exon 3 of the in this selected group of healthy individuals is shown in Table I. The Journal of Immunology 3

The compound heterozygous FcgRIIa-131HR/FcgRIIIb-NA1NA2 haplotype was most prevalent. Within the selected 31 homozygous FcgRIIa-131RR individuals, only 2 were FcgRIIIb-NA1NA1 whereas 15 were NA2NA2 (Table I). The genotype frequencies of both FcgRIIa and FcgRIIIb in this selection of individuals are similar to what has been found in other studies concerning white individuals (19). Upon stimulation of neutrophils with IgG we found clear dif- ferences in NADPH oxidase activity between the various hap- lotypes (Fig. 1). The amount of H2O2 produced by neutrophils from individuals that were genotyped as homozygous FcgRIIIb- NA2NA2 was found to depend on the functional polymorphism in FcgRIIa; that is, the neutrophils of those that were typed as FIGURE 1. H2O2 production by neutrophils expressing different FcgR FcgRIIa-131HH produced significantly more H2O2 compared with haplotypes in response to IgG. Data are depicted as a ratio of the maximum those that were FcgRIIa-131RR, with FcgRIIa-131HR individuals slopes of H2O2 production in response to IgG or PMA. H2O2 production showing an intermediate response. in FcgRIIIb-NA2NA2 individuals was significantly different between In contrast, in FcgRIIIb-NA1NA1 individuals this order of FcgRIIa-HH individuals and those that are FcgRIIa-HR (p = 0.0425) and , , NADPH oxidase activity from high to low activity was completely FcgRIIa-RR (p = 0.0078). *p 0.05, **p 0.01. reversed, being FcgRIIa-131RR . -131HR . -131HH. Although Downloaded from proper statistical analysis could not be performed due to a lack of mAb was a result of the Kurlander effect (18), in which one mAb sufficient healthy FcgRIIa-RR individuals to be tested in our co- can both bind its targeted receptor with its epitope-specific Fab hort, a clear trend was observed. Individuals typed heterozygous domains and another Fc receptor by the Fc tail. In this case this for FcgRIIIb (NA1NA2) all had equal responses when compared resulted in partial blocking the IgG-binding epitope of FcgRIIa by among each other, regardless of their FcgRIIa genotype. the Fc tails of CD16 mAbs upon binding to the surface-expressed

To exclude the possibility that the differences in H2O2 pro- FcgRIIIb molecules on neutrophils. The F(ab9)2 fragments of CD16 http://www.jimmunol.org/ duction are the result of differences in FcgR expression levels, mAb 3G8 have previously been described to block IgG binding to neutrophils were analyzed by flow cytometry (Fig. 2). We did not FcgRIII (22). Moreover, the F(ab9)2 fragments of the CD16 mAb detect any differences in expression of FcgRIIa between the dif- still bound to the same extent to surface-expressed FcgRIIIb on ferent haplotypes. FcgRIIIb expression was higher in FcgRIIIb- human neutrophils as indicated by flow cytometry (data not shown). NA2NA2 individuals than in those homozygous for NA1, and The results obtained with blocking Abs were similar for the dif- heterozygotes had an intermediate expression. Because the mAb ferent haplotypes tested (Supplemental Fig. 2). we used to detect expression levels does not differ in binding be- The above-mentioned experiments indicate the relevance of the tween FcgRIIIb-NA1 and FcgRIIIb-NA2 (13), we know that this genetic haplotypes on the IgG-binding capacity of the neutrophil, is truly a difference in surface expression levels. However, within which seems to depend totally on the function of FcgRIIa for the by guest on September 28, 2021 the groups of individuals with the same FcgRIIIb genotype, there subsequent signaling capacity toward NADPH oxidase activity. were no differences in expression between carriers of the different FcgRIIa variants. Although some differences in FcgR expression Preactivation of neutrophils results in a loss of differences levels were observed, these findings cannot explain the differences in between haplotypes IgG-mediated NADPH oxidase activity as observed in Fig. 1. As To investigate the impact of preactivation on FcgR expression and expected, FcgRI was absent on these circulating neutrophils. The IgG-mediated neutrophil activation, we also cultured the neu- absence of FcgRI (CD64) expression confirms that the freshly iso- trophils from these genotyped individuals overnight with a com- lated neutrophils were not preactivated in vivo (20, 21). bination of G-CSF and IFN-g, which had previously been shown To determine the individual contribution of FcgRIIa and FcgRIIIb to most effectively induce FcgRI (20, 23, 24). The number of in the response to IgG, we performed similar experiments in the apoptotic neutrophils was ,20% under these conditions (data not presence of blocking mAbs or F(ab9)2 fragments for both FcgRs on shown), whereas FcgRI expression was maximally upregulated, circulating neutrophils from healthy volunteers (Fig. 3). Blocking as determined by flow cytometry. Preactivation also changed the FcgRII by either intact or F(ab9)2 CD32 mAb completely inhibited FcgRIIa and FcgRIIIb expression patterns (Fig. 4), as reported H2O2 production, showing that FcgRIIa is absolutely required for before (20, 21, 25). Total FcgRII (CD32) expression was increased ROS production in this setting. Blocking FcgRIII by intact CD16 only slightly, whereas FcgRIIIb (CD16) expression was reduced by mAb inhibited this response much less efficiently, and blocking ∼50%, probably due to shedding (25). Although expression of the FcgRIII with F(ab9)2 of CD16 mAb did not inhibit the response at only inhibitory IgG receptor FcgRIIb was undetectable on resting all, suggesting that the inhibition of the response by intact CD16 cells, low levels of expression were reproducibly detected after preactivation (Fig. 4). When expression levels were compared be- tween individuals having different haplotypes, FcgR expression following preactivation was not identical among these individuals. Table I. Distribution of FcgRIIa/FcgRIIIb haplotypes in healthy white individuals FcgRI expression was found to be significantly higher in FcgRIIa- 131RR individuals than in those that were 131HH (Fig. 4). Even NA1NA1 NA1NA2 NA2NA2 though the total FcgRII (CD32) expression was very similar among all donors, the expression of FcgRIIb (CD32b) was highest in in- HH 8 17 9 dividuals typed FcgRIIIb-NA1NA1 and almost absent in those typed HR 12 23 17 RR 2 14 15 FcgRIIIb-NA2NA2. In contrast, even after partial shedding by the overnight preactivation, the FcgRIIIb expression was most pro- Only individuals with two copies of each gene are represented in this table (n = 117). H, FcgRIIa-H131; NA1, FcgRIIIb-HNA1a; NA2, FcgRIIIb-HNA1b; nounced in FcgRIIIb-NA2NA2 compared with NA1NA1 and R, FcgRIIa-R131. NA1NA2 individuals, similar to the differences in expression levels 4 NEUTROPHIL FcgR HAPLOTYPES IN THE RESPONSE TO IgG

FIGURE 2. FcgRII and FcgRIIIb expression levels on neutrophils. Data are shown as mean fluorescence intensity (MFI) of fluorescently labeled mAbs against FcgRII (AT10, pan-CD32) or FcgRIIIb (3G8). Binding of FcgRIIIb mAbs was significantly lower in individuals that are FcgRIIIb-NA1NA1 when compared with those that are FcgRIIIb-NA1NA2 (p = 0.0115) or FcgRIII-NA2NA2 (p , 0.0001). Binding to FcgRIIIb-NA2NA2 was also significantly higher when compared with FcgRIIIb-NA1NA2 (p = 0.0114). Binding of FcgRIIIb mAbs was not different between individuals when compared with the polymorphism in FcgRIIa. Binding of mAbs against FcgRII was similar between all donors. 131HH, 131HR, and 131RR and NA1NA1, NA1NA2, and NA2NA2 indicate the polymorphic variants of FcgRIIa (131H/R) and FcgRIIIb (NA1/NA2), respectively. All values were corrected by subtracting the background values measured with isotype mAb controls. Data are expressed as means 6 SEM of 10–12 experiments per polymorphic variant. *p , 0.05,

***p , 0.001. Downloaded from observed in freshly isolated neutrophils from these individuals. tivation. Although FcgRIIa-131H binds more avidly to IgG1 and Without preactivation with cytokines, the number of apoptotic neu- IgG2, we found a significant impact of FcgRIIIb variants on the trophils after overnight culture was .70–80%, thus precluding respiratory burst activity, with maximal IgG-induced H2O2 pro- functional studies (data not shown). duction by neutrophils from individuals with (homozygous) haplo-

To investigate whether the altered FcgR expression pattern on type 131HH-NA2NA2 . 131RR-NA1NA1 . 131HH-NA1NA1 . http://www.jimmunol.org/ preactivated neutrophils influenced the response to IgG, these 131RR-NA2NA2. These haplotypic differences in respiratory burst cells were again tested for NADPH oxidase activity. Compared to activity disappeared upon preactivation of the neutrophils. freshly isolated neutrophils, H2O2 production was increased 2-fold Many polymorphic variants of low-affinity FcgRhavebeen upon stimulation with IgG but not with PMA (Fig. 5 and data not described to be functionally different from one another. Changes shown). Remarkably, the observed differences between haplotypes in binding affinity (10), receptor expression (3), promoter activity in freshly isolated neutrophils were not found anymore with the (26, 27), and localization in the plasma membrane (28) are all preactivated cells. Blocking mAbs against FcgRIIa, either used among the consequences of such polymorphisms. By far the most alone or combined with mAbs against FcgRI and/or FcgRIIIb, studies reporting on these issues analyze one receptor at a time, by guest on September 28, 2021 resulted in ∼50% reduction of H2O2 production. although most immune cells express multiple FcgRs with different In contrast to the situation in freshly isolated neutrophils, the affinities for IgG and different downstream signaling pathways, response to IgG could not be significantly inhibited by intact some activating and some inhibitory (29). FcgRIIIb mAbs. The combination of FcgRII mAbs with FcgRI When we compared the respiratory burst after stimulation with and/or FcgRIIIb mAbs was not more effective than blocking IgG, a response that both FcgRIIa and FcgRIIIb have been re- FcgRII alone, irrespective of the haplotype (Supplemental Fig. 3), ported to induce (30), we found that the neutrophil response is suggesting that, similarly to what we observed in freshly isolated ultimately determined by the combination of polymorphisms. For neutrophils, FcgRIIa was the dominant activating FcgR in the pre- instance, FcgRIIIb-NA2NA2 individuals that are also FcgRIIa- activated neutrophils as well. 131HH had an almost 2-fold higher respiratory burst than did those that were FcgRIIa-131RR, whereas in FcgRIIIb-NA1NA1 Discussion donor cells these differences were reversed. Minor differences in Our studies indicate an unpredicted yet relevant interaction be- FcgR expression were detected, but these could not explain the tween common FcgR haplotypes in IgG-mediated neutrophil ac- differences we observed in the respiratory burst.

FIGURE 3. The effect of blocking anti-FcgR Abs on the IgG-induced H2O2 production by neutrophils. Data are depicted as a percentage of the ratio of the maximum slopes of H2O2 production in response to IgG or PMA without blocking Abs. H2O2 production was measured in the presence of blocking mAbs or F(ab9)2 fragments against FcgRII (CD32) or FcgRIII (CD16). The significance of differences in H2O2 production between neutrophils with and without blocking mAbs or F(ab9)2 is indicated in the figure. Blocking mAbs or F(ab9)2 against FcgRII (CD32) are indicated by anti-CD32, those against FcgRIII (CD16) by anti-CD16. Data are expressed as means 6 SEM, n = 8 for each blocking mAb or F(ab9)2. The Journal of Immunology 5

FIGURE 4. Changes in neutrophil FcgR expression after overnight preactivation with G-CSF and IFN-g.At the left side of each of the four panels, the data are shown as mean fluorescence intensity (MFI) of fluo- rescently labeled mAbs against FcgRI (CD64), FcgRII (pan-CD32), FcgRIIb (CD32b), or FcgRIIIb (CD16). Day 0, freshly isolated, nonactivated neutrophils (open bars); day 1, neutrophils cultured overnight in the pres- ence of G-CSF and IFN-g (black bars). At the right side of each panel the MFI of fluorescently labeled mAbs against the different FcgR is indicated for all genotypes separately on the neutrophils cultured overnight in the presence of G-CSF and IFN-g (dark gray bars). 131HH, 131HR, and 131RR and NA1NA1, NA1NA2, and NA2NA2 indicate the polymorphic variants of FcgRIIa (131H/R) and FcgRIIIb (NA1/NA2), respectively. Sig- nificant differences in FcgR expression are indicated in the figure. All values were corrected by subtracting the background values measured with isotype mAb controls. Data are expressed as means 6 SEM of 32 independent Downloaded from experiments with freshly isolated (day 0) neutrophils, and eight to nine independent experiments per poly- morphic variant with overnight cultured neutrophils (day 1, total of 26). http://www.jimmunol.org/

Blocking FcgRII, either by intact mAbs or F(ab9)2 fragments, FcgRIIIb is much higher than FcgRIIa, it will bind many IgG almost completely inhibited the response. In contrast, blocking molecules that are consequently not available for binding by FcgRIIIb by intact mAbs only partially inhibited the response and FcgRIIa. Because FcgRIIIb is not able to signal directly, this may F(ab9)2 fragments did not inhibit at all. This suggests that the negatively influence the level of the response. On the other hand, blocking effect observed with intact FcgRIIIb mAbs is due to a so- FcgRIIIb could increase the binding of neutrophils to immune called Kurlander effect, that is, the Fc tails of bound FcgRIIIb complexes, which may increase the response. In general, these mAbs also bind and consequently block FcgRIIa. The fact that the mechanisms could be in balance, explaining why no difference in response was not increased in experiments with intact mAbs response is observed after blocking. Nevertheless, there is a pos- by guest on September 28, 2021 compared with F(ab9)2 fragments shows that possible induction of sibility that this balance is slightly different between the hap- the response by Kurlander effect by crosslinking receptors on the lotypes. Alternatively, pre-existing interactions between FcgRIIa surface was also not the case in this experimental setting. The and FcgRIIIb and other receptors in plasma membrane micro- blocking experiments suggest a predominant role of FcgRIIa in domains may contribute to the observed effects. Differences in the the production of ROS by neutrophils in response to IgG. Al- functional capacity of the FcgR complexes containing the NA1 though the experiments with blocking FcgRIIIb F(ab9)2 fragments and NA2 FcgRIIIb haplotypes to transduce outside–in signals may at first glance may suggest that FcgRIIIb is not involved in the explain the magnitude of the downstream response. For example, neutrophil response of H2O2 production, the influence of the ge- neutrophil FcgR and the integrin CR3 (CD11b/CD18) are known netic polymorphisms in this receptor clearly shows that FcgRIIIb to coexist in such signaling complexes and influence each other in modulates this response. Therefore, when taken together, the cellular responses (31–34). These interactions have been proposed blocking experiments do not fully explain the differences in IgG to be due to lectin-saccharide–like binding between the receptors responses between haplotypes. Possibly, FcgRIIIb plays a dual (34). Because two of the five nucleotides in FcgRIIIb-NA2 that role in this response. On the one hand, because expression of differ from FcgRIIIb-NA1 result in additional N-glycosylation sites (5), it might well be that glycosylation differences affect the interaction with CR3. It will clearly be of interest to study the composition and signaling capacity of the various FcgRIIa and FcgRIIIb haplotype complexes. Large changes in receptor expression were induced by pre- activating the neutrophils with a combination of IFN-g and G-CSF. Induced expression of FcgRI and ∼50% reduction in FcgRIIIb surface expression profoundly altered the FcgR expression profile and IgG-induced response. Although changes in FcgR expression were variable among haplotypes, no significant differences be- tween the haplotypes in H2O2 production were detected anymore, irrespective of the FcgR expression levels. Overall, the response of preactivated neutrophils to IgG was about twice as high compared with nonactivated neutrophils. When FIGURE 5. H2O2 production by preactivated neutrophils. Neutrophils were preactivated by culturing overnight in the presence of G-CSF and used as intact murine Abs, the results showed that the anti-CD32

IFN-g. Data are depicted as a ratio of the maximum slopes of H2O2 pro- mAb was still the most potently blocking mAb. In our experimental duction in response to IgG or PMA. Data are expressed as mean 6 SEM of setting we found that blocking FcgRI by intact anti-CD64 mAb two to four independent experiments per haplotype (26 in total). alone had no effect on the response at all. Reasoning from the 6 NEUTROPHIL FcgR HAPLOTYPES IN THE RESPONSE TO IgG genetic haplotypes, it is mainly the reduction in FcgRIIIb in re- ciation studies. The MLPA assay that we have developed and lationship to the relatively unchanged FcgRIIa expression that has described before (7, 16), and that we have also employed in the a large influence on how these receptors interact with and respond present study, provides an important and powerful tool to perform to IgG. The role of FcgRI is unclear in this scenario, and it is such integrated analyses. difficult to address experimentally by blocking Abs, as there are no anti-CD64 Abs that completely block IgG binding with their Acknowledgments Fab alone. The 10.1 clone against FcgRI did block IgG-coated We thank Prof. Dirk Roos for helpful discussions and critical reading of erythrocyte binding to FcgRI-transfected cell lines using intact the manuscript. mAb, but not when the F(ab9)2 fragments of this mAb were ap- plied, demonstrating a Kurlander effect to be involved in the blocking effects of intact anti-CD64 mAb as well (data not shown). This may Disclosures explain that blocking of FcgRI IgG binding by mAb 10.1 is partially The authors have no financial conflicts of interest. dependent on IgG concentration (35). It might well be that the high concentration of IgG that is present in our assays was such that References binding of mAb 10.1 did not significantly reduce the IgG-mediated 1. Warmerdam, P. A., J. G. van de Winkel, E. J. Gosselin, and P. J. Capel. 1990. respiratory burst in preactivated neutrophils, explaining why the IgG Molecular basis for a polymorphism of human Fcg receptor II (CD32). J. Exp. response in preactivated neutrophils could not be completely blocked Med. 172: 19–25. 2. Li, X., J. Wu, R. H. Carter, J. C. Edberg, K. Su, G. S. Cooper, and R. P. Kimberly. even in the presence of mAbs against all three FcgRs. 2003. A novel polymorphism in the Fcg receptor IIB (CD32B) transmembrane With respect to the genotypic impact and the surface expression region alters receptor signaling. Arthritis Rheum. 48: 3242–3252. Downloaded from of the various FcgRs on neutrophils, we observed some remark- 3. Metes, D., L. K. Ernst, W. H. Chambers, A. Sulica, R. B. Herberman, and P. A. Morel. 1998. Expression of functional CD32 molecules on human NK cells able correlations in the induction of FcgRI and FcgRIIb. FcgRI is determined by an allelic polymorphism of the FcgRIIC gene. Blood 91: 2369– induction was linked to the FcgRIIa-131H/R phenotype, whereas 2380. 4. Wu, J., J. C. Edberg, P. B. Redecha, V. Bansal, P. M. Guyre, K. Coleman, FcgRIIb induction was linked to the FcgRIIIb-NA1/NA2 pheno- J. E. Salmon, and R. P. Kimberly. 1997. A novel polymorphism of FcgRIIIa type. Thus, the FcgRIIa-131RR FcgRIIIb-NA2NA2 haplotype (CD16) alters receptor function and predisposes to autoimmune disease. J. Clin. Invest. 100: 1059–1070. seemed to result in increased FcgRI and lower FcgRIIb induction http://www.jimmunol.org/ 5. Ory, P. A., M. R. Clark, E. E. Kwoh, S. B. Clarkson, and I. M. Goldstein. 1989. upon activation, hence changing the weaker IgG-induced H2O2 Sequences of complementary DNAs that encode the NA1 and NA2 forms of Fc production of this haplotype under resting conditions to a similar receptor III on human neutrophils. J. Clin. Invest. 84: 1688–1691. level of respiratory burst activity compared with the other hap- 6. Bux, J., E. L. Stein, P. Bierling, P. Fromont, M. Clay, D. Stroncek, and S. Santoso. 1997. Characterization of a new alloantigen (SH) on the human lotypes upon activation of neutrophils. Although very reproduc- neutrophil Fcg receptor IIIb. Blood 89: 1027–1034. ible, with heterozygous genotypes showing an intermediary effect 7. Breunis, W. B., E. van Mirre, J. Geissler, N. Laddach, G. Wolbink, E. van der Schoot, M. de Haas, M. de Boer, D. Roos, and T. W. Kuijpers. 2009. Copy compared with both homozygous genotypes, an explanation for number variation at the FCGR locus includes FCGR3A, FCGR2C and FCGR3B such linkage at the genomic level in FcgR protein expression but not FCGR2A and FCGR2B. Hum. Mutat. 30: E640–E650. regulation remains unclear to date. In the cases of FcgRIIb and 8. Breunis, W. B., E. van Mirre, M. Bruin, J. Geissler, M. de Boer, M. Peters, D. Roos, M. de Haas, H. R. Koene, and T. W. Kuijpers. 2008. Copy number by guest on September 28, 2021 FcgRIIIb, some intergenic polymorphisms in or near the promoter variation of the activating FCGR2C gene predisposes to idiopathic thrombocy- sites of these adjacent, head-to-head arranged genes may be linked topenic purpura. Blood 111: 1029–1038. to the NA2 allele. In contrast, such regulatory mechanisms at 9. van der Heijden, J., W. B. Breunis, J. Geissler, M. de Boer, T. K. van den Berg, and T. W. Kuijpers. 2012. Phenotypic variation in IgG receptors by nonclassical the genomic level seems less likely to be involved in case of the FCGR2C alleles. J. Immunol. 188: 1318–1324. FcgRIIa–FcgRI interaction in FcgRI upregulation because of the 10. Bruhns, P., B. Iannascoli, P. England, D. A. Mancardi, N. Fernandez, S. Jorieux, . and M. Dae¨ron. 2009. Specificity and affinity of human Fcgamma receptors and distance of 11 megabases between both loci. their polymorphic variants for human IgG subclasses. Blood 113: 3716–3725. Numerous reports have been published about the association 11. Nicu, E. A., U. Van der Velden, V. Everts, A. J. Van Winkelhoff, D. Roos, and of FcgR polymorphic variants with disease susceptibility and se- B. G. Loos. 2007. Hyper-reactive PMNs in FcgRIIa 131 H/H genotype peri- odontitis patients. J. Clin. Periodontol. 34: 938–945. verity (reviewed by Bournazos et al. in Ref. 36). As is the case 12. Bredius, R. G., C. A. Fijen, M. De Haas, E. J. Kuijper, R. S. Weening, J. G. Van with studies on receptor function, nearly all of these studies have de Winkel, and T. A. Out. 1994. Role of neutrophil FcgRIIa (CD32) and focused on variations in a single receptor type. As our study FcgRIIIb (CD16) polymorphic forms in phagocytosis of human IgG1- and IgG3- opsonized bacteria and erythrocytes. Immunology 83: 624–630. indicates, specific FcgR haplotypes can lead to a response that is 13. Salmon, J. E., J. C. Edberg, and R. P. Kimberly. 1990. Fcg receptor III on human clearly different from the expectations raised by some of the ex- neutrophils. Allelic variants have functionally distinct capacities. J. Clin. Invest. 85: 1287–1295. cellent studies investigating IgG binding by single receptor var- 14. van Sorge, N. M., W. L. van der Pol, and J. G. van de Winkel. 2003. FcgR iants (10). polymorphisms: Implications for function, disease susceptibility and immuno- Our studies have been performed on healthy white volunteers therapy. Tissue Antigens 61: 189–202. 15. Veri, M. C., S. Gorlatov, H. Li, S. Burke, S. Johnson, J. Stavenhagen, K. E. Stein, only. It is known that genotype and haplotype frequencies of E. Bonvini, and S. Koenig. 2007. Monoclonal antibodies capable of discrimi- FcgRIIa and FcgRIIIb variants can be different in other pop- nating the human inhibitory Fcg-receptor IIB (CD32B) from the activating Fcg- ulations (19). However, we do not expect that the differences in receptor IIA (CD32A): biochemical, biological and functional characterization. Immunology 121: 392–404. response we have observed between haplotypes will be different 16. Schouten, J. P., C. J. McElgunn, R. Waaijer, D. Zwijnenburg, F. Diepvens, and in populations in which a different genetic distribution of these G. Pals. 2002. Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res. 30: e57. haplotypes is found, especially because our blocking experiments 17. van Bruggen, R., A. Drewniak, M. Jansen, M. van Houdt, D. Roos, H. Chapel, suggest that the IgG-induced respiratory burst in our assay is A. J. Verhoeven, and T. W. Kuijpers. 2009. Complement receptor 3, not Dectin- dependent on FcgRs only. 1, is the major receptor on human neutrophils for b-glucan-bearing particles. Mol. Immunol. 47: 575–581. In conclusion, our systematic integrated functional analysis of 18. Kurlander, R. J. 1983. Blockade of Fc receptor-mediated binding to U-937 cells FcgRIIa and FcgRIIIb genetic variants in an extensive cohort of by murine monoclonal antibodies directed against a variety of surface antigens. healthy subjects demonstrates, to our knowledge for the first time, J. Immunol. 131: 140–147. 19. Lehrnbecher, T., C. B. Foster, S. Zhu, S. F. Leitman, L. R. Goldin, K. Huppi, and a novel level of functional interaction between haplotypes of al- S. J. Chanock. 1999. Variant genotypes of the low-affinity Fcg receptors in two lelic FcgRIIa and FcgRIIIb variants on human neutrophils. These control populations and a review of low-affinity Fcg receptor polymorphisms in control and disease populations. Blood 94: 4220–4232. findings provide a clear incentive for a more complete and inte- 20. Repp, R., T. Valerius, A. Sendler, M. Gramatzki, H. Iro, J. R. Kalden, and grated analysis of FcgR haplotypes in clinical and genetic asso- E. Platzer. 1991. Neutrophils express the high affinity receptor for IgG (FcgRI, The Journal of Immunology 7

CD64) after in vivo application of recombinant human granulocyte colony- 28. Floto, R. A., M. R. Clatworthy, K. R. Heilbronn, D. R. Rosner, P. A. MacAry, stimulating factor. Blood 78: 885–889. A. Rankin, P. J. Lehner, W. H. Ouwehand, J. M. Allen, N. A. Watkins, and 21. Perussia, B., E. T. Dayton, R. Lazarus, V. Fanning, and G. Trinchieri. 1983. K. G. Smith. 2005. Loss of function of a lupus-associated FcgRIIb polymor- Immune interferon induces the receptor for monomeric IgG1 on human mono- phism through exclusion from lipid rafts. Nat. Med. 11: 1056–1058. cytic and myeloid cells. J. Exp. Med. 158: 1092–1113. 29. Nimmerjahn, F., and J. V. Ravetch. 2008. Fcg receptors as regulators of immune 22. Fossati, G., R. J. Moots, R. C. Bucknall, and S. W. Edwards. 2002. Differential responses. Nat. Rev. Immunol. 8: 34–47. role of neutrophil Fcg receptor IIIB (CD16) in phagocytosis, bacterial killing, 30. Crockett-Torabi, E., and J. C. Fantone. 1990. Soluble and insoluble immune and responses to immune complexes. Arthritis Rheum. 46: 1351–1361. complexes activate human neutrophil NADPH oxidase by distinct Fcg receptor- 23. Cohen, D. M., S. C. Bhalla, E. J. Anaissie, J. P. Hester, C. A. Savary, and specific mechanisms. J. Immunol. 145: 3026–3032. J. H. Rex. 1997. Effects of in vitro and in vivo cytokine treatment, leucapheresis 31. Huang, Z. Y., S. Hunter, P. Chien, M. K. Kim, T. H. Han-Kim, Z. K. Indik, and and irradiation on the function of human neutrophils: implications for white A. D. Schreiber. 2011. Interaction of two phagocytic host defense systems: Fcg blood cell transfusion therapy. Clin. Lab. Haematol. 19: 39–47. receptors and complement receptor 3. J. Biol. Chem. 286: 160–168. 24. Gericke, G. H., S. G. Ericson, L. Pan, L. E. Mills, P. M. Guyre, and P. Ely. 1995. 32. Marzocchi-Machado, C. M., C. M. Alves, A. E. Azzolini, A. C. Polizello, Mature polymorphonuclear leukocytes express high-affinity receptors for IgG I. F. Carvalho, and Y. M. Lucisano-Valim. 2002. Fcgamma and complement (FcgRI) after stimulation with granulocyte colony-stimulating factor (G-CSF). J. receptors: expression, role and co-operation in mediating the oxidative burst and Leukoc. Biol. 57: 455–461. degranulation of neutrophils of Brazilian systemic lupus erythematosus patients. 25. Homburg, C. H., M. de Haas, A. E. von dem Borne, A. J. Verhoeven, Lupus 11: 240–248. C. P. Reutelingsperger, and D. Roos. 1995. Human neutrophils lose their surface 33. Voice, J. K., and P. J. Lachmann. 1997. Neutrophil Fcg and complement receptors FcgRIII and acquire annexin V binding sites during apoptosis in vitro. Blood 85: involved in binding soluble IgG immune complexes and in specific granule release 532–540. induced by soluble IgG immune complexes. Eur. J. Immunol. 27: 2514–2523. 26. Su, K., J. Wu, J. C. Edberg, X. Li, P. Ferguson, G. S. Cooper, C. D. Langefeld, and 34. Zhou, M., R. F. Todd, III, J. G. van de Winkel, and H. R. Petty. 1993. Cocapping R. P. Kimberly. 2004. A promoter haplotype of the immunoreceptor tyrosine-based of the leukoadhesin molecules complement receptor type 3 and lymphocyte inhibitory motif-bearing FcgRIIb alters receptor expression and associates function-associated antigen-1 with Fcg receptor III on human neutrophils. with autoimmunity. I. Regulatory FCGR2B polymorphisms and their asso- Possible role of lectin-like interactions. J. Immunol. 150: 3030–3041. ciation with systemic lupus erythematosus. J. Immunol. 172: 7186–7191. 35. Dougherty, G. J., Y. Selvendran, S. Murdoch, D. G. Palmer, and N. Hogg. 1987. 27. Su, K., X. Li, J. C. Edberg, J. Wu, P. Ferguson, and R. P. Kimberly. 2004. A The human mononuclear phagocyte high-affinity Fc receptor, FcRI, defined by Downloaded from promoter haplotype of the immunoreceptor tyrosine-based inhibitory motif- a monoclonal antibody, 10.1. Eur. J. Immunol. 17: 1453–1459. bearing FcgRIIb alters receptor expression and associates with autoimmunity. 36. Bournazos, S., J. M. Woof, S. P. Hart, and I. Dransfield. 2009. Functional and II. Differential binding of GATA4 and Yin-Yang1 transcription factors and clinical consequences of Fc receptor polymorphic and copy number variants. correlated receptor expression and function. J. Immunol. 172: 7192–7199. Clin. Exp. Immunol. 157: 244–254. http://www.jimmunol.org/ by guest on September 28, 2021