Dissertation
Modeling and Targeting Tissue Dynamics in Fibrosis and Cancer
Drug development is hampered by the limited predictive value of conventional 2D culture systems, which fail to recapitulate the complexity of the disease microenvironment and contribute to high clinical attrition rates.
- Author
- C. Liao
- Date
- 17 April 2026
- Links
- Thesis in Leiden Repository
This thesis presents the development of clinically relevant 3D culture models and a high-throughput data acquisition and analysis platform to improve drug target discovery and validation in fibrosis and cancer.
A 3D fibroblast-based ECM remodeling assay was established to quantify fibrotic matrix remodeling and screen antifibrotic compounds. Using this model, we identified the non-canonical MEK–ERK pathway, rather than canonical TGFβ signaling, as a key driver of fibrotic ECM remodeling. Notably, high concentrations of TGFβ receptor inhibitors paradoxically enhanced MEK-mediated profibrotic remodeling, highlighting MEK–ERK signaling as a potential therapeutic target. The platform was further applied to colorectal cancer, where cancer-associated fibroblasts (CAFs) showed enhanced matrix remodeling compared to matched normal fibroblasts. Cathepsin H (CTSH) was identified as a critical mediator of CAF-driven remodeling and a potential biomarker for T1 colorectal cancer.
To investigate tumor immunotherapy, we developed ECM-embedded tumoroid arrays to evaluate bispecific antibodies (bsAbs) and T cell receptor (TCR)-engineered T cells. For CD3×HER2 bsAbs, both binding affinity and epitope location critically influenced T cell recruitment and cytotoxicity. High-affinity bsAbs targeting proximal HER2 epitopes were most effective, while reduced CD3 affinity was less easily compensated by increased antibody concentration. CD8+ and activated T cells exhibited superior cytotoxicity and motility compared to total or resting T cells. The platform also enabled validation of PRAME-specific TCR T cells in uveal melanoma and ROPN1-targeting TCR T cells in triple-negative breast cancer, demonstrating antigen-specific tumor elimination.
Finally, 3D mosaic tumoroid models were used to mimic tumor heterogeneity in HER2-positive breast cancer. BsAb-mediated interactions between T cells and HER2+ tumor cells induced substantial bystander killing of HER2− cells through IFNγ/TNFα-dependent paracrine signaling, even in the absence of direct T cell contact.
Together, this work establishes a versatile 3D high-throughput platform that captures key aspects of the disease microenvironment, enabling mechanistic insights and improving preclinical evaluation of antifibrotic therapies and cancer immunotherapies.