Generation of high affinity ICAM-1-specific nanobodies and evaluation of their suitability for allergy treatment

Abstract

The nasal cavity is an important site of allergen entry. Hence, it represents an organ where trans-epithelial allergen penetration and subsequent IgE-mediated allergic inflammation can potentially be inhibited. Intercellular adhesion molecule 1 (ICAM-1) is highly expressed on the surface of respiratory epithelial cells in allergic patients. It was identified as a promising target to immobilize antibody conjugates bispecific for ICAM-1 and allergens and thereby block allergen entry. We have previously characterized a nanobody specific for the major birch pollen allergen Bet v 1 and here we report the generation and characterization of ICAM-1-specific nanobodies. Nanobodies were obtained from a camel immunized with ICAM-1 and a high affinity binder was selected after phage display (Nb44). Nb44 was expressed as recombinant protein containing HA- and His-tags in Escherichia coli (E.coli) and purified via affinity chromatography. SDS-PAGE and Western blot revealed a single band at approximately 20 kDa. Nb44 bound to recombinant ICAM-1 in ELISA, and to ICAM-1 expressed on the human bronchial epithelial cell line 16HBE14o- as determined by flow cytometry. Experiments conducted at 4°C and at 37°C, to mimic physiological conditions, yielded similar percentages (97.2 ± 1.2% and 96.7 ± 1.5% out of total live cells). To confirm and visualize binding, we performed immunofluorescence microscopy. While Texas Red Dextran was rapidly internalized Nb44 remained localized on the cell surface. Additionally, we determined the strength of Nb44 and ICAM-1 interaction using surface plasmon resonance (SPR). Nb44 bound ICAM-1 with high affinity (10^-10 M) and had slow off-rates (10^-4 s^-1). In conclusion, our results showed that the selected ICAM-1-specific nanobody bound ICAM-1 with high affinity and was not internalized. Thus, it could be further used to engineer heterodimers with allergen-specific nanobodies in order to develop topical treatments of pollen allergy.

Keywords: ICAM-1; VHH; allergy; high affinity; nanobody.

Figure 1

Generation of ICAM-1-specific nanobody Nb44. (A) shows the timeline of immunization. Intervals of injections with ICAM-1 as well as of obtaining pre-immune and immune sera are indicated (in days). The time point of isolating peripheral blood mononuclear cells (PBMCs) for generating a VHH-cDNA library is marked. (B) Reactivity of IgG antibodies specific for ICAM-1 in pre-immune serum (taken on day 0) and in immune serum (taken on day 58 after four injections with ICAM-1) was determined by ELISA. OD values (y-axis) correspond to the amount of bound IgG antibodies and are shown as mean of duplicates with a variation of less than 5%. Displayed data are representative of 2 independent experiments. (C) Amino acid sequence of Nb44. Framework regions (FR) 1-4 and complementarity determining regions (CDR, red boxes) 1-3 are indicated. Hallmark VHH amino acid substitutions (V37F, G44E, L45R, W47G) are marked in blue, cysteine residues forming the intradomain disulfide bridge are labeled in orange, the linker sequence is marked in gray and tag sequences (HA-tag, His-tag) are highlighted in yellow.

Figure 2

Expression and purification of nanobodies. (A) Coomassie Brilliant Blue-stained SDS-PAGE of ICAM-1-specific nanobody (Nb44) under reducing and non-reducing conditions and (B) nitrocellulose-blotted Nb44 under reducing and non-reducing conditions, detected with HRP-coupled mouse anti-HA antibody. Lane 1: protein molecular mass marker protein molecular mass marker (M), lane 2: purified Nb44, 1µg. Molecular masses are indicated on the left margin in kDa. (C) Reactivity of Nb44 to plate-bound ICAM-1 and control protein BSA. OD values (y-axis) correspond to the amount of bound nanobodies and are shown as means of triplicates ± SD. Data shown are representatives of 4 independent experiments.

Figure 3

SPR-based study of the interaction between Nb44 and ICAM-1. (A) Separated and Coomassie Brilliant Blue-stained reaction mixtures of ICAM-1 deglycosylated by PNGase F. Lane 1: protein molecular mass marker (M), lane 2: ICAM-1, 1 µg, lane 3: ICAM-1 in reaction mixture without (-) PNGase F, lane 4: with (+) PNGase F, lane 5: protein molecular mass marker. Molecular masses are displayed on both margins in kDa. (B) Multi-cycle kinetics using ICAM-1 as ligand and Nb44 as analyte. Different colored lines correspond to different concentrations of Nb44. (C) Single-cycle kinetics with Nb44 as ligand (captured by anti-His-tag antibodies) and ICAM-1 as analyte. (B, C) Recorded curves (colored) were superimposed with calculated curves (black) according to a heterogeneous ligand model. Signal intensities (RU, y-axes) are shown versus time (in seconds, x-axes). Association and dissociation rate constants (k_a, k_d), dissociation constants (K_D) and T values for each parameter are indicated. Displayed graphs are representatives of 4 (B) or 3 (C) independent experiments.

Figure 4

Flow cytometric analysis of mouse anti-ICAM-1 antibody (A) and Nb44 (B) binding to ICAM-1 expressed on 16HBE14o- cells. (A) Cells were incubated with a purchased mouse anti-ICAM-1 antibody or a mouse isotype control at 4°C (upper panels) or 37°C (lower panels), and subsequently stained with Alexa Fluor 467-conjugated anti-mouse antibody. Plots show forward scatter (FSC-A) (x-axes) versus anti-mouse Alexa Fluor 467 (y-axes). (B) Cells were incubated with Nb44 (left column), an isotype control (Bet v 1-specific Nb, middle column) or buffer alone (right column) at 4°C (upper panels) or 37°C (lower panels). Plots show forward scatter (FSC-A) (x-axes) versus PE-labeled anti-His tag antibody (y-axes). 16HBE14o- cells (A, B) were previously selected for alive cells by negative staining with eFluor^® 780 viability dye. Experiments shown in (A, B) were performed in triplicates and representative plots of three independent experiments are shown.

Figure 5

Nb44 binds to ICAM-1 on cell surfaces and is not internalized up to 24 hours. 16HBE14o- cells were simultaneously incubated with Texas Red-labeled Dextran (lysine-fixable) to label the endolysosomal system and Nb44 (left columns) or an isotype control (Bet v 1-specific Nb, right columns) for 30 minutes (A–C) 4 hours (D) or 24 hours (E). After fixation with formaldehyde, cells were not permeabilized (A) or permeabilized with saponin (B-E). Nanobodies bound to the plasma membrane were detected with Alexa Fluor 488-labeled anti-His-tag antibody (green). (C) For control purposes, cells were not only incubated with Nb44 and detection antibody (column 1), isotype control and detection antibody (column 2), but also with detection antibody alone (column 3) and PBS (column 4). (A-C) Confocal microscopic images showing yz-axes and xy-axes are representative of 3 independent experiments. (D, E) Representative wide field fluorescence microscopy images indicating Nb44 immobilization to the cell surface for 4 hours and 24 hours without internalization. Experiments were performed in duplicates. (A-E) Texas Red Dextran was present in all settings and cell nuclei were visualized with DRAQ5 (blue); scale bar is 10 µm.