Human IgE does not bind to human FcRn

The neonatal Fc receptor (FcRn) is known to mediate placental transfer of IgG from mother to unborn. IgE is widely known for triggering immune responses to environmental antigens. Recent evidence suggests FcRn-mediated transplacental passage of IgE during pregnancy. However, direct interaction of FcRn and IgE was not investigated. Here, we compared binding of human IgE and IgG variants to recombinant soluble human FcRn with β2-microglobulin (sFcRn) in surface plasmon resonance (SPR) at pH 7.4 and pH 6.0. No interaction was found between human IgE and human sFcRn. These results imply that FcRn can only transport IgE indirectly, and thereby possibly transfer allergenic sensitivity from mother to fetus.

www.nature.com/scientificreports/ IgG is one of the most abundant serum proteins and the most abundant immunoglobulin found in human serum 10 . Mutational analysis have led to the identification of IgG1-Fc variants influencing IgG binding to FcRn. IgG1-MST-HN has been found to exhibit increased binding to FcRn 11 , whereas IgG1-IHH cannot bind FcRn 12 . IgG can mediate a non-cellular response by engaging the complement system as well as it can bind Fc gamma receptors (FcγRs) and cross-link them on effector cells in the form of IgG immune complexes (ICs), triggering cellular effector functions 10 . In contrast to IgG, IgE is the immunoglobulin with the lowest abundance in serum. Next to its role in the defense of parasites 13 , IgE is known for its involvement in immune reactions against environmental antigens, causing type I hypersensitivity 14,15 . There are two main IgE Fc receptors, the high affinity FcεRI and the low affinity FcεRII (CD23). Cross-linking upon binding of antigens to FcεRI-bound IgE or binding of IgE ICs on the surface of e.g. basophils and mast cells has been reported to initiate cellular immune responses 14 . CD23 on the other hand exists in both in a trimeric membrane-bound and soluble mono-and trimeric forms, regulating IgE synthesis and homeostasis 16 . Membrane-bound CD23 is expressed on B cells and intestinal epithelial cells, where it also controls IgE synthesis and mediates transfer of IgE-ICs to the intestinal lumen, respectively 14,17 .
Whereas it is clear that the active transport of IgG across the placenta to the unborn is FcRn-mediated 8,9,18 , transport of other isotypes such as IgA and IgM is generally not considered relevant, most likely passive, as only a small fraction of what is found in maternal sera can be found in cord blood 19 . However, FcRn involvement has been reported for the transfer of tolerance to food allergens from mother to offspring in mice 20 as well as for the transfer of IgE in anti-IgE IgG/IgE ICs in mice 21 and humans 22 . A recent study suggested FcRn-dependent placental transport of IgE from mother to offspring in mice 23 .
In this study we investigated the binding of human IgE and IgG variants to human sFcRn using SPR aiming to complement the already published cellular data from a physicochemical point of view.

Materials and methods
Generation of anti-biotin IgG1-Fc variants, anti-biotin IgE and human sFcRn. Linear DNA strands encoding for mutated IGHG1*03 and IGHE*02 Fc-regions were ordered from Integrated DNA Technologies and cloned into a pcDNA3.1 expression vector containing anti-biotin heavy chain variable regions obtained from 24,25 , as described previously 26,27 . Linear DNA strands encoding for the soluble FcRn α-chain with a C-terminal BirA deca-Histidine (His) tag and ß2-microglobulin were ordered accordingly and cloned separately into pcDNA3.1 expression vectors, as described elsewhere 28 . In brief, expression vectors and DNA inserts were digested with EcoRI and NheI FastDigest restriction enzymes (Thermo Scientific). The expression vector backbone was isolated by gel purification using a 1% UltraPure agarose (Thermo Scientific) gel with 1:10.000 SYBR Safe (Invitrogen). DNA was extracted from the gel using the NucleoSpin Gel and PCR clean-up kit (Macherey-Nagel) according to the manufacturer's protocol. The DNA fragments were isolated using the same kit but without prior gel purification.
The DNA fragments were ligated into the pcDNA3.1 backbone overnight at 16 °C using T4 DNA Ligase (New England Biolabs) in 1 × T4 DNA Ligation buffer (New England Biolabs) with a molar ratio of insert to vector of 3:1. 5 µL of the ligation reaction was transformed into 50 µL DH5α competent cells (Thermo Scientific) by heat shock.
The cells were plated on LB-agar plates containing 50 µg/mL ampicillin (Thermo Scientific) and incubated overnight at 37 °C. Colonies were picked and grown in 2 mL LB medium containing 50 µg/mL ampicillin (Thermo Scientific) overnight at 37 °C, shaking at 180 RPM. DNA was isolated from the bacterial culture using the NucleoSpin Plasmid EasyPure kit (Macherey-Nagel) according to the manufacturer's protocol and sequenced.
Sequence-confirmed DNA was used for transformation of DH5α competent cells as described above and colonies were picked and grown in a 5 mL preculture. This was subsequently used to inoculate 200 mL LB medium containing 50 µg/mL ampicillin (Thermo Scientific), after which the culture was grown overnight at 37 °C, shaking at 180 RPM. DNA was isolated using the NucleoBond Xtra Maxi kit (Macherey-Nagel) according to the manufacturer's protocol and again sequenced.

Production of anti-biotin IgG1-Fc variants, anti-biotin IgE and
BirA-His-tagged human sFcRn. Antibodies were produced as described previously 26 . In brief, 31.35 µg of the heavy chain vector, 37.65 µg light chain vector (pcDNA3.1 anti-biotin VLCL 24,25 ) and 31 µg pSVLT/p21/p27 mix 29 were added to 6.66 mL opti-MEM (Thermo Scientific) per 100 mL of transfection cell culture. 300 µL Polyethylenimine (PEI) MAX (linear, MW 4.000, Polysciences) was added, the mixture was immediately vortexed and incubated for 20 min at room temperature. 100 mL of HEK293F cells (Thermo Scientific) at 1*10 6 cells/mL in fresh FreeStyle 293 Expression Medium (Thermo Fisher Scientific) were transfected with the mixture and incubated at 37 °C at 8% CO2 and shaking. After 4 h, 100 units/mL penicillin and 100 µg/mL streptomycin (Thermo Fisher Scientific) were added to the culture. Human sFcRn was produced as described previously 28 , using equimolar amounts of both the soluble FcRn α-chain and ß2-microglobulin expression vectors. The culture supernatants were harvested 6 days after transfection by spinning down the cells twice for 5 min at 3.100 × g and filtering through a 0.45 µm syringe filter (Whatman).

Purification of recombinant anti-biotin IgGs and human sFcRn.
Anti-biotin IgGs were purified from culture medium with AKTA prime (GE Healthcare) by affinity chromatography using either a 5 mL HiTrap HP protein A (IgG1-WT and -MST-HN) or protein G (IgG1-IHH) column (GE Healthcare), as described previously 30 , or a HisTrap HP column (sFcRn) (GE Healthcare). Fractions containing the antibodies or sFcRn were combined and concentrated using a 10 K MWCO Pierce Protein Concentrator PES (Thermo Scientific). Antibodies were fractionated by HPLC-SEC using an AKTA UPC-900, P-920 and Frac-950 (GE Healthcare)

SDS-PAGE.
Each antibody was tested in SDS-PAGE under reducing and non-reducing conditions. Samples were incubated for 5 min at 70 °C or 95 °C in the presence of 20 mM Iodoacetamide (to prevent reduction during denaturation) 31 or 0.25% (w/v) ß-Mercaptoethanol in NuPAGE™ LDS Sample Buffer (Thermo Scientific) for non-reduced and reduced conditions, respectively. Samples were loaded on a NuPAGE™ 4-12% Bis-Tris Gel (Invitrogen) and run with MOPS SDS running buffer (Thermo Scientific) for 10 min at 100 V followed by 45 min at 120 V. The gel was stained overnight in Blue-Silver solution (10% phosphoric acid, 10% ammonium sulfate, 0.12% CoomassieBlue G-250 (Sigma-Aldrich) and 20% methanol (Thermo Scientific) and thoroughly destained in distilled water.

Anti-biotin IgG variants and IgE are functional and show expected molecular assembly. Anti-
biotin IgG1 variants were tested in HPLC-SEC in order to confirm their integrity. All molecules showed expected sizes and were monomeric (Fig. 1A). Furthermore, integrity and size of all IgG molecules assayed were confirmed in SDS-PAGE (Fig. 1B, left). For the IgE supernatant we observed some additional bands on the gel (Fig. 1B, right), presumably due to other protein species in the culture supernatant and the fact that a non-native gel was used. We therefore further confirmed the integrity of the IgE in the culture supernatant in an IgE-specific Western Blot after separation on a native gel next to a contact allergy serum sample (Fig. 1C) of a previously described patient cohort 32 . We then employed our SPR platform to test antigen binding and further validate the molecular identity of the IgE in the culture supernatant. As shown in the schematic overview in Fig. 1D, BSAbiotin was spotted on the sensor and purified anti-biotin IgG or IgE culture supernatant was injected, followed by a subsequent injection of anti-hIgE or PBS (left panel) or-in independent experiments-soluble FcεRI or PBS (right panel). Both anti-biotin IgG and IgE bound BSA-biotin at pH 7.4 and pH 6.0. Only the IgE was recognized by anti-IgE antibody and soluble FcεRI, confirming its molecular identity (Fig. 1E).
Anti-biotin IgE does not bind to human sFcRn. Next, we tested binding of the antibodies to human sFcRn in our SPR system after antigen capture, as described previously 34 . In order to have a valid set of controls, we included two previously described IgG1-Fc variants, anti-biotin IgG1-MST-HN 11 and IgG1-IHH 12 , which have enhanced 11 and no binding 12 to FcRn. We captured the antibodies on the sensor in the same manner as in Fig. 1E 11,12,35 . Importantly, no binding of human sFcRn was observed to IgE (Fig. 2).

Discussion
Human FcRn mediates placental transcytosis of IgG from mother to unborn 3,7-9 . Yet, it remains not fully understood how and to what extent isotypes other than IgG, e.g. IgE, cross the placental barrier 22,23,36 . In this study, we compared binding of human sFcRn to different Fc variants of IgG1 and to IgE at both neutral and acidic pH. Unlike IgG, which shows a clear pH-dependent binding to FcRn, no binding was seen to human IgE. The concept of FcRn-dependent transfer of tolerance from mother to offspring is well established in mice: After sensitization of mothers, antigen-specific IgG has been described to confer tolerance to e.g. ovalbumin (OVA) or food allergens to offspring. Transfer of such protective IgG and IgG-ICs has been reported to occur via placental passage and via the mother milk, respectively, both in a FcRn-dependent manner 20,[37][38][39] .
The presence of IgE in cord blood has been reported in humans, albeit at low concentrations compared to paired maternal blood 36,[40][41][42] . Possibly, some placental transcytosis of IgE molecules takes place 22,23,36,40 . Evidence has been reported that this happens in a FcRn-dependent manner in humans 22 and in mice 21,23 . Two recent studies implied a role of maternal IgE in priming allergic responses in offspring, providing evidence for such indirect Figure 2. Titration of human sFcRn in SPR reveals no binding to IgE at pH 7.4 or 6.0. Anti-biotin IgG variants and IgE were captured on a sensor with different concentrations of BSA-biotin spotted, leading to comparable levels of antigen-bound IgG and IgE (not shown due to blank subtraction). Human sFcRn was injected in a twofold serial dilution covering a concentration range from 15.63 to 1000 nM and K D values were calculated fitting a 1:1 Langmuir binding model after blank subtraction. A representative of three independent experiments is shown. www.nature.com/scientificreports/ FcRn-dependent placental transport in mice 21,23 and suggested a direct FcRn-mediated transport of IgE 43 . It is unclear to which extent the high amounts of antigen-specific IgE administered to the parental animals during pregnancy in these studies 21,23 -more than 100-fold higher than naturally occurring total IgE levels in mice 44 -can be extrapolated to more physiological concentrations, as recently noted elsewhere 45 . If direct FcRn-mediated transport of IgE does not occur, are there other ways FcRn might play a role in IgE transport across the placenta? One possibility is FcRn-mediated placental passage of complexes of IgE with IgG, forming anti-IgE IgG/IgE complexes 21,22 . The presence of such complexes in serum is controversial 46 . Next to anti-IgE IgG/IgE ICs, FcRn-mediated placental passage of antigen-specific IgE could also occur in the form small ICs formed by IgG and IgE binding the same antigen. Interestingly, a study by Weil et al. provides indirect evidence for this scenario in the context of parasitic infection. Although they found remarkable levels of parasitic IgE in cord blood, often exclusively antigen-specific but generally ~ 50-100 times less than found in maternal blood, the data was interpreted as an antigen-specific IgE response originating from the fetus. However, such IgE levels could also be interpreted as evidence for ICs formed by IgG and IgE bound to the same antigen 36 , but do not favor the hypothesis of IgE being commonly transported across the placental barrier.
Invariably, the relative total IgE concentrations found in cord blood compared to matched parental samples are very low 36,40 , comparable to or even lower than relative levels of IgA or IgM, respectively 19 , not suggestive of an active, directly FcRn-mediated placental transport as described for IgG 3,[7][8][9]19 .
Our results complement previously published data showing no binding of hIgE to hFcRn when overexpressed on MDCK cells by FACS 22 with binding data to human sFcRn in SPR at both pH 7.4 and pH 6.0. We conclude that it is very unlikely that human FcRn directly mediates placental transcytosis of human IgE, but rather through ICs containing IgG, possibly as anti-IgE IgG/IgE ICs, as suggested elsewhere 21,47 .

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. www.nature.com/scientificreports/