Characterization of the interactions of rabbit neonatal Fc receptor (FcRn) with rabbit and human IgG isotypes

Abstract

Despite the increasing importance of rabbit as an animal model in pharmacological studies like investigating placental transfer of therapeutic IgGs, little is known about the molecular interaction of the rabbit neonatal Fc receptor (FcRn) with rabbit and human IgG molecules. We analyzed the interactions of the rabbit and human FcRn with rabbit and human IgG isotypes using surface plasmon resonance assay. Similar to FcRn of other species, rabbit FcRn functions in pH-dependent manner, as it binds IgGs at pH 6.0, but no binding occurs at pH 7.4. We also showed that rabbit FcRn binds rabbit IgG and human IgG1 with nearly identical affinity, whereas it has stronger interactions with the other human IgG isotypes. The similar affinity of rabbit IgG and human IgG1 for rabbit FcRn was confirmed by in vitro FcRn-mediated recycling assay. These data verify that rabbit is an appropriate animal model for analyzing the pharmacokinetics of human therapeutic monoclonal antibodies.

Fig 1. Domain-by-domain alignment of the amino acid sequences for human and rabbit variants of FcRn α-chain, β2m and IgG isotypes.

Structural and functional features are highlighted in the amino acid sequences of hFcRn and rbFcRn α-chains (Panel A), rabbit and human β2m (Panel B) and CH2, as well as CH3 domains of human and rabbit IgG isotypes (Panel C). Consensus residues are assigned based on the number of occurrences of the character in the column, emphasizing the degree of conservation. The higher the conservation in a column the darker the background of the character [32]. Potential N-linked glycosylation sites are indicated by filled triangle. Residues at the interface between FcRn and Fc region of IgG based on a crystallography analysis of a rat FcRn-heterodimeric Fc complex [10] are labelled with asterisks. Conserved His at position 166 in FcRn α-chain sequences (Panel A) is considered to bind to albumin [33] is indicated by a + sign. Numbering is based on the hFcRn α-chain sequence.

Fig 2. Expression and SEC analysis of the soluble form of rbFcRn purified from Sf9 cell supernatant.

The expression of rbFcRn was detected by 15% SDS-PAGE followed by Coomassie staining. The rbFcRn α-chain, as well as the rbβ2m was observed around 28 and 12 kDa, respectively (Panel A). By using anti-His tag monoclonal Ab in Western blot experiments only the rbFcRn α-chain containing 6xHis-tag can be detected (Panel B). As a control, soluble hFcRn was used in both experiments. The purity of rbFcRn (Panel C) and rbIgG (Panel D), as well as hFcRn (Panel E) and hIgG1 (Panel F) samples was verified by size-exclusion chromatography.

Fig 3. pH-dependent binding of rbIgG and hIgG1 by rbFcRn.

Soluble rbFcRn (Panel A and B), as well as rbIgG (Panel C) and hIgG1 (Panel D) were immobilized on a GLC chip at densities of 1200, 1900 and 2100 RU, respectively. rbIgG (Panel A) and hIgG1 (Panel B) at 133 nM, while soluble rbFcRn (Panel C and D) at 600 nM concentration in PBS-T buffer were injected at pH 6.0, as well as at pH 7.4 and interactions were monitored at 25°C. The curves were corrected by subtracting the non-specific binding responses obtained from control channel (n = 2–3, one representative figure is shown in each cases).

Fig 4. Sensorgrams for the interaction of soluble rbFcRn with rbIgG and hIgG1.

Varying rbIgG and hIgG1 (8.34 to 133 nM) or soluble rbFcRn (37.5 to 600 nM) concentrations were injected over immobilized soluble rbFcRn or rbIgG and hIgG1, respectively, on the chip surface. In both experimental set-ups, the interactions were monitored at pH 6.0 and at 25°C and the curves were corrected by subtracting the non-specific binding responses obtained from control channel. Heterogeneous ligand model (Panel A and B) or Langmuir 1:1 binding model (Panel C and D) was fitted to the sensorgrams by grouped analysis using BIAevaluation software (n = 2–3, one representative figure is shown in each cases).

Fig 5. FcRn-mediated recycling studies of rbIgG and hIgG1 in rabbit peritoneal macrophages.

Cells were pulsed with Alexa 488-conjugated rbIgG and hIgG1 for 20 minutes and chased for 30 minutes. Confocal images were collected at chase starting point (rbIgG on Panel A and hIgG1 on Panel C) and after 30 minutes (rbIgG on Panel B and hIgG1 on Panel D) and were processed using Fiji software [30]. Cells were drawn around, and the median level of pixel intensity were determined, and for comparison, relative fluorescent intensity of the cells were calculated in all cases (Panel E). Values shown are the means ± SEM (**, p< 0.01; ***, p<0.001).