MD/BSc Thesis

Decoding the CHI3L1-RAGE Immunosuppressive Complex: Integrative Structural and Coevolutionary Insights into HER2+ Breast Cancer Resistance

In the complex landscape of HER2⁺ breast cancer, therapeutic resistance often outpaces intervention. My thesis zeroes in on a critical, yet poorly understood, molecular checkpoint: the interaction between the protein CHI3L1 and its receptor RAGE, which tumors hijack to suppress the immune system.

By integrating deep computational modeling with biophysical analysis, we developed the first high-confidence atomic model of this immunosuppressive complex. The model reveals an elegant mechanism of antagonism: CHI3L1 acts as an evolutionary repurposed immune checkpoint, locking the receptor in an altered oligomeric state. This work transforms a diffuse resistance mechanism into a tractable target, providing a molecular foundation for rational drug design.

An Integrative Approach

This project is the culmination of my MEDTEC training, bridging biology, medicine and bioengineering. To solve this puzzle, we designed and executed a multi-pronged strategy that combined:

  • Biophysical Characterization: Using ELISA, SEC, and mass photometry to quantify binding affinity, stoichiometry, and the role of glycosylation.
  • X-Ray Crystallography: Conducting experiments at two synchrotron facilities (SSRL and ESRF), successfully resolving the apo-structure of CHI3L1 and navigating the challenges of complex crystallization.
  • Computational & Co-evolutionary Modeling: Performing a systematic, deep exploration of AlphaFold-Multimer, optimizing predictions across hundreds of seeds and parameters to decode the co-evolutionary signals that define the complex's architecture.

This thesis demonstrates how an integrative, multidisciplinary and holistic approach can bring clarity to complex biological systems. Evolution wrote the code; our work was (and is) to read it.