Researchers at the University of Regensburg and the Centre for Protein Diagnostics (PRODI) at Ruhr University Bochum have developed an innovative method that combines computer-aided protein modelling with spectroscopic experiments. Their aim is to precisely validate antibody-antigen interactions at the atomic level, thereby enabling a more in-depth understanding of molecular binding mechanisms.

 

A centrepiece of this method is the targeted creation of so-called "dead mutants". These are variants of an antibody with a single, precisely placed mutation that completely cancels binding to its target antigen. By selectively disrupting specific bindings, the researchers can identify which atomic contacts are essential for antibody-antigen interaction.

 

As a case study, the researchers used the antibody solanezumab, which binds the amyloid-beta (Aβ) peptide, a major factor in Alzheimer's disease. They predicted that a single mutation (^G95AHC) in solanezumab would completely prevent Aβ binding. This prediction was extensively validated experimentally using immuno-infrared biosensors (iRS) and ELISA assays. The results confirmed that the "dead mutant" showed no binding to the antigen, while the native structure of the antibody remained intact. This experimental confirmation provides direct knowledge about the antibody-antigen interaction.

 

The iRS systems used are based on ATR-FTIR spectroscopy, are label-free and provide additional information about the secondary structure of the antigen. They can recognise Aβ misfolding years before the onset of clinical symptoms, which can enable early Alzheimer's diagnosis. One advantage of "dead mutants" in this context is their function as ideal negative controls: Since iRS systems detect all immobilised proteins, the mutants help to clearly distinguish specific binding from non-specific binding and thus significantly increase diagnostic reliability. Tests with cerebrospinal fluid (CSF) samples emphasised the specificity of this method.

 

In summary, this study demonstrates the power of combining computational modelling and experimental validation for the detailed understanding of antibody-antigen interactions at the atomic level. This interdisciplinary approach has great potential for diagnostic and possibly therapeutic applications and is in principle transferable to other antibodies and their antigens.

 

Further information can be found in our original publication:

Marvin Scherlo, Adrian Höveler, Marvin Mann, Grischa Gerwert, Jörn Güldenhaupt, Klaus Gerwert, Till Rudack, Carsten Kötting
Replacement of a single residue in an antibody abolishes cognate antigen binding, as predicted by theoretical methods
Computational and Structural Biotechnology Journal 2025 27:4363-4372 doi: 10.1016/j.csbj.2025.10.018 (external link, opens in a new window)