A stepwise mutagenesis approach using histidine and acidic amino acid to engineer highly pH-dependent protein switches

Abstract

Antibody-based drugs can be highly toxic, because they target normal tissue as well as tumor tissue. The pH value of the extracellular microenvironments around tumor tissues is lower than that of normal tissues. Therefore, antibodies that engage in pH-dependent binding at slightly acidic pH are crucial for improving the safety of antibody-based drugs. Thus, we implemented a stepwise mutagenesis approach to engineering pH-dependent antibodies capable of selective binding in the acidic microenvironment in this study. The first step involved single-residue histidine scanning mutagenesis of the antibody's complementarity-determining regions to prescreen for pH-dependent mutants and identify ionizable sensitive hot-spot residues that could be substituted by acidic amino acids to obtain pH-dependent antibodies. The second step involved single-acidic amino acid residue substitutions of the identified residues and the assessment of pH-dependent binding. We identified six ionizable sensitive hot-spot residues using single-histidine scanning mutagenesis. Nine pH-dependent antibodies were isolated using single-acidic amino acid residue mutagenesis at the six hot-spot residue positions. Relative to wild-type anti-CEA chimera antibody, the binding selectivity of the best performing mutant was improved by approximately 32-fold according to ELISA and by tenfold according to FACS assay. The mutant had a high affinity in the pH range of 5.5-6.0. This study supports the development of pH-dependent protein switches and increases our understanding of the role of ionizable residues in protein interfaces. The stepwise mutagenesis approach is rapid, general, and robust and is expected to produce pH-sensitive protein affinity reagents for various applications.

Keywords: Antibody engineering; Carcinoembryonic antigen; Hot-spot residues; Tumor microenvironments.

Fig. 1

Schematic diagram of mutagenesis strategy for the generation of pH-dependent mutants. a Construction of the anti-CEA chimera antibody. b Single-histidine scanning mutagenesis. c Expression of histidine mutants. d Measurement of the expression quantity of histidine mutants. e Determination of epitope residues using dual-pH capture ELISA. f Single-residue acidic amino acid mutagenesis at identified epitope residue. g Expression of acidic amino acid mutants. h Measurement of the expression quantity of acidic amino acid mutants. i Screening of pH-dependent mutants using dual-pH capture ELISA

Fig. 2

Screening of pH-dependent mutants. a, b Determination of the epitope residues for the CDR domain of the anti-CEA antibody using dual-pH capture ELISA. Error bars represent standard deviations between technical replicates. c Analysis of relevant anti-CEA antibody epitope residues. A ribbon diagram of the crystal structure of an anti-CEA T84.66 Fv fragment is presented. The heavy chain is in dark blue, and the light chain is in light blue. The IgG ionizable sensitive hot-spot residues (LC-F36, LC-R54, LC-T95, HC-P98, HC-G100, and HC-A107) are represented as red ribbons

Fig. 3

Screening of the pH-dependent mutants using dual-pH capture ELISA. Error bars represent standard deviations between technical replicates

Fig. 4

SDS-PAGE of purified antibody. The monomeric anti-CEA antibody heavy chain is indicated by 55KD, and the light chain is indicated by 25KD. Lane 1,12: size markers (kDa). Lane 2: wild-type antibodies. Lane 3–11: pH-dependent mutants

Fig. 5

Characterization of purified pH-dependent mutants binding to CEA. a The pH dependence of anti-CEA antibody mutants binding to CEA antigen was tested under acidic and physiological pH conditions, and CEA binding was analyzed through a titration ELISA. b EC50 values of anti-CEA mutants binding to human CEA by ELISA at pH6.0 and pH7.4 were calculated using the nonlinear fit (variable slope, four parameters) model built into GraphPad Prism software version 9.1.2. c The binding affinity of tested CEA antibody mutants from binding experiments to CEA antibody at a broad pH value using a pH range ELISA. Error bars represent standard deviations between technical replicates

Fig. 6

The pH selectivity of anti-CEA mutants was determined by FACS. a Binding activity of clone LF36D and LF36E to LS147T cells at pH 6.0 and pH 7.4. Error bars represent standard deviations between technical replicates. b EC50 values of anti-CEA mutants binding to human-CEA by FACS at pH6.0 and pH7.4 were calculated using the nonlinear fit (variable slope, four parameters) model built into GraphPad Prism software version 9.1.2. c The fold-change of EC50 Values was determined by pH Flow Cytometry at pH 6.0 and pH 7.4