Figure 1.
(A): Computer graphic representation of a part of the complex between a single staphylococcal protein A domain (domain B, closely related to the Z domain used in the present study) (yellow) and human IgG1 Fc (brown) (PDB file: 1FC2.pdb). The amino acid side chains corresponding to the seven positions addressed for substitutions in the Z domain are highlighted in cyan. Highlighted in purple, red and green, respectively, are three IgG1 Fc subregions in close contact with the B domain. (B): Alignment of amino acid sequences of Fc regions of human IgG1, mouse IgG1, mouse IgG2a and mouse IgG2b, respectively, corresponding to the three contact areas shown in (A), using the same colour code. (C): Amino acid sequence of the 58-residue Z domain, with the seven positions included in the engineering boxed. Indicated with red dots are the ten positions at which unique cysteine residues were introduced for site-specific labeling with a photoactivable maleimide benzophenone (MBP) group.
Figure 2.
Relative binding responses for Z domain variants to mouse IgG1.
Solutions of Z domain variants (2 µM concentration) obtained after single amino acid substitutions at the seven targeted positions (K4, F5, Q9, Q10, F13, H18 and K35) were injected over a sensor chip surface containing immobilized mIgG1 monoclonal antibody protein and the equilibrium responses (plateau values) were recorded and normalized to the response values obtained for the wild type Z domain (WT) A blank surface (activated/deactivated) was used as negative control and used for buffer effect subtraction. (A) A representative overlay sensorgram from injections of ZWT, ZF5R and ZF5I variants over mIgG1 monoclonal protein showing higher equilibrium response values for the two mutant variants than for the ZWT domain. The response obtained from buffer injection only is also indicated (BUFFER). (B) Results from the analysis of the different variants. The horizontal line in each panel corresponds to the normalized response obtained for the wild type Z domain ( = 100).
Figure 3.
Binding of antibodies to immobilized Z domain variants.
(A-D): Overlay of sensorgram traces obtained after injection of four different mouse IgG1 monoclonal antibodies (one per panel) over separate sensor chip surfaces containing similar amounts of the variants ZF5I, ZF5R or the wild type Z domain (ZWT), respectively. (E): Overlay of sensorgram traces obtained after injection of a TNF-alpha receptor:human IgG1 Fc fusion protein (etanercept, Enbrel®) over separate sensor chip surfaces containing similar amounts of the variants ZF5I, ZF5R or the wild type Z domain (ZWT), respectively.
Figure 4.
Site-specific conjugation of maleimide benzophenone to Z domain variants.
(A): Schematic showing how maleimide benzophenone is conjugated to a unique cysteine residue introduced into each of the Z variants by genetic engineering. (B): A representative result from analysis by LC-ESI mass spectrometry of the product obtained after conjugation (ZF5I-Q32C-MBP). (C): Schematic showing the 16 different Z-MBP probes (eight positions for the MBP group on both the ZWT and ZF5I variants) evaluated for photocoupling to Fc of mIgG1 using UV radiation.
Figure 5.
Photo-conjugation of Z domain probes to mouse IgG1.
(A): Analysis by SDS-PAGE of the mouse IgG1 mAb photocoupling efficiency of nine different ZWT- or ZF5I-based probes, differing in the position of the maleimide benzophenone (MBP) group. As indicated, variants containing the MBP group at positions 2, 3, 4, 9, 11, 15, 24 or 32 were investigated. (B): Analysis by SDS-PAGE of the photocoupling efficiency of the ZWT-Q32C-MBP and ZF5I-Q32C-MBP probes, both containing the MBP group at position 32, to 19 different mouse IgG1 mAbs (see Material and Methods for a list). The designations HC+P, HC, LC and P, refer to heavy chain+probe, heavy chain, light chain and probe, respectively. The M lanes refer to marker protein with molecular weights in kDa as indicated.
Figure 6.
Photo-conjugation efficiencies for the different Z domain probes to mouse IgG1.
The efficiencies of photoconjugation were analysed through scanning of the gel images in Figure 5 using ImageJ software (see Materials and methods section). (A) Photo-conjugation efficiencies for 16 probe variants, derived from either the ZWT-domain (white bars) or the ZF5I-domain, containing the F5I substitution (black bars). The position at which a cysteine was introduced to site-specifically position the MBP probe is indicated for each probe. (B) Mean values of photo-conjugation efficiencies (%) from the use of either the ZWT-Q32C-MBP (white bar; mean value = 45.4±4.1%) or ZF5I-Q32C-MBP (black bar; mean value = 64.4±5.3%) probes to a large panel of different mIgG1 antibodies (16 of the mAbs seen in Figure 5B).
Figure 7.
Analysis of labeling and comparison of the antigen binding potency of differently biotinylated antibody samples.
Two samples of a mouse anti-human interferon-gamma mIgG1 mAb were biotinylated using either a direct or a ZF5I-Q32C-MBP-BIO probe-mediated strategy and the resulting preparations compared with respect for their antigen binding potency. (A) Western blotting analysis of the samples and the probe, using a streptavidin horse radish peroxidase (HRP) conjugate. Lane 1: blank; lane 2: a sample of the anti-human interferon-gamma mIgG1 mAb biotinylated using a conventional amine reactive sulfo-NHS-ester-biotin reagent resulting in "global" biotinylation of both heavy (ca. 50 kDa) and light chains (ca. 25 kDa); lane 3: the ZF5I-Q32C-MBP-BIO probe alone, biotinylated using the same amine reactive sulfo-NHS-ester-biotin reagent; lane 4: a sample of the anti-human interferon-gamma mIgG1 mAb after photoconjugation to the ZF5I-Q32C-MBP-BIO probe (ca. 8 kDa) showing a selective biotinylation of only the heavy chains (ca. 50 + 8 kDa). A small amount of unreacted probe is also visible. (B) Biosensor analysis of the relative antigen binding potency of the biotinylated antibody preparations. Using a streptavidin coated biosensor chip, the biotinylated antibody proteins were selectively immobilized onto separate sensor chip surfaces, followed by injection over both surfaces of a common 7.5 nM solution of the antigen human interferon-gamma. This allowed for direct comparison of the effect on the antigen binding potency from the different biotinylation strategies. Sample (i): the anti-human interferon-gamma mIgG1 mAb biotinylated using a conventional amine reactive sulfo-NHS-ester-biotin reagent; sample (ii): the anti-human interferon-gamma mIgG1 mAb biotinylated via photoconjugation using the ZF5I-Q32C-MBP-BIO probe.