Molecular Insights of Nickel Binding to Therapeutic Antibodies as a Possible New Antibody Superantigen

Abstract

The binding of nickel by immune proteins can manifest as Type IV contact dermatitis (Ni-specific T cells mediated) and less frequently as Type I hypersensitivity with both mechanisms remaining unknown to date. Since there are reports of patients co-manifesting the two hypersensitivities, a common mechanism may underlie both the TCR and IgE nickel binding. Focusing on Trastuzumab and Pertuzumab IgE variants as serendipitous investigation models, we found Ni-NTA interactions independent of Her2 binding to be due to glutamine stretches. These stretches are both Ni-inducible and in fixed pockets at the antibody complementarity-determining regions (CDRs) and framework regions (FWRs) of both the antibody heavy and light chains with influence from the heavy chain constant region. Comparisons with TCRs structures revealed similar interactions, demonstrating the possible underlying mechanism in selecting for Ni-binding IgEs and TCRs respectively. With the elucidation of the interaction, future therapeutic antibodies could also be sagaciously engineered to utilize such nickel binding for biotechnological purposes.

Keywords: IgE; IgG1; allergy; antibody; glutamine; nickel (II); type I hypersensitivity.

Figure 1

Dissociation equilibrium constants (KD) of Pertuzumab and Trastuzumab IgE (A) and IgG1 (B) of different VH families with the Ni-NTA biosensor. The measurements were performed at concentrations from 200 nM to 3.125 nM of the antibodies. Values of KD (M), ka (1/Ms), and kd (1/s) were determined using the Octet RED96 system. The X-axis depicts the time (in seconds) while the Y-axis depicts the binding responses (nm). All the experiments were conducted in at least duplicates. “Poor response” reflects measurement responses that are less than 0.1 nm for at least three concentrations. Variants that could not be produced are left empty.

Figure 2

Structural analyses of glutamine from Pertuzumab and Trastuzumab VH3, VH5, and VH7 IgE variants in binding Ni-NTA. (A) Model of whole IgE (L-chain in white surface and H-chains in gray surface) reveals the formed pocket at the VHx–Vκ1 interface (dashed circle). The Ni-NTA bound region of the variants VH3 (red), VH5 (blue), and VH7 (green) shows the glutamine (Q) stretch constituted by the VH Q39 (underlined) and Vκ1 Q37-Q38 to form a stable cryptic pocket (capacity of which is shown in cyan transparent surface) in the Trastuzumab IgE variants or a transient non-cryptic pocket (yellow transparent surface) in the Pertuzumab VH7. The bound Ni-NTA is shown in magenta spheres and interacts at the pocket in both Pertuzumab and Trastuzumab. (B) Relative solvent accessible surface area (SASA) of the Pertuzumab and Trastuzumab VH Q39 residue in all the variants to demonstrate the deeply buried Q39 (SASA< 10%) of the Pertuzumab variants. (C) Distance d between residue 67 (R67 in VH3 or H67 in VH5) and H-chain CDR2 D50 of Pertuzumab VH3-Vκ1 IgE (red) and VH5-Vκ1 IgE (blue) during the simulation. Structural visualizations were generated using PyMOL 2.3.2 (46). An augmented reality (AR) presentation of the Ni-NTA binding to Trastuzumab VH5 IgE glutamine-forming pocket could be viewed using our ‘‘APD AR Holistic Review’’ app available on both Google and Apple app stores [for more details see ref (47, 48)].

Figure 3

The Ni-NTA binding region of the Pertuzumab VH5–Vκ1 IgE. The main residue for the interaction is H67 of VH5 FWR3, whereas for Trastuzumab VH5–Vκ1 IgE, it is at the Q-stretch containing VH5 Q39 at FWR2. (A) Distribution of the Ni-NTA docked conformers showing all the glutamine (GLN) and histidine (HIS) present in the V-region of the Pertuzumab and Trastuzumab VH5–Vκ1 IgE variants. The predominant binding residue locations are highlighted in dashed boxes. (B) KD of the Pertuzumab and Trastuzumab VH5–Vκ1 IgE variants including H67Q mutation.

Figure 4

Ni-NTA dissociation equilibrium constants (KD) of Pertuzumab and Trastuzumab VH3 (A) and VH5 (B) IgG1 variants paired with different Vκs. The experiments were performed at concentrations from 200 nM to 3.125 nM of the antibodies. Values of KD (M), ka (1/Ms), and kd (1/s) were determined using the Octet RED96 system. The X-axis depicts the time (in seconds) while the Y-axis depicts the binding responses (nm). All the experiments were conducted in at least duplicates. “Poor response” reflects measurement responses that are less than 0.1 nm for at least three concentrations. Variants that could not be produced are left empty.

Figure 5

Effects of the different VH5-Vκ pairings in the Pertuzumab and Trastuzumab IgG1 variants. (A) Distance-based difference maps [generated using CMView (49)] between Fab domains of Pertuzumab and Trastuzumab VH5-Vκ3 and VH5-Vκ6 IgG1 variants (superimposed against the respective VH5-Vκ1 IgG1 as the reference) indicates more disruption by the Vκ3 or Vκ6 for Pertuzumab VH5 IgG1 variants, leading to varying changes in the heavy and light chain pairing. (B) An alternative Ni-NTA binding region in Pertuzumab VH5-Vκ1 IgG1 formed a Q-stretch of Vκ1 Q3, Q89, and Q90. These Q-stretches were incomplete due to the Vκ mutation Q3V in the Pertuzumab VH5-Vκ3 and VH5-Vκ6 IgG1 variants.

Figure 6

Analysis of metal binding ability. (A) Binding responses of the Pertuzumab VH5 IgE variant to different metal ions (Ni²⁺, Cu²⁺, and Co²⁺) recharged on the NTA sensors. (B) Numbers of the metal interactions in the retrieved protein-metal complexes from Protein Data Bank. The smaller offset plot showed the metal interaction results of a subset of immune-related protein complexes found by using independent keywords: “immune”, “antibody”, “MHC”, and “TCR”. (C) Structural presentation of a VHH antibody complexed with Ni²⁺ (PDB: 4PPT), HLA-B*27:05 (PDB: 5IB4), and HLA-A68 (PDB: 6EI2) complexed with a peptide and Ni²⁺, where glutamines (magenta) are found to be involved in (e.g., in 4PPT) or closed to the Ni²⁺ binding region with histidine (green). The Ni²⁺ ions are represented by blue spheres.