Research Overview

Melanoma is an aggressive tumor with dramatically reduced patient survival when disseminated to distant organs. Despite the historic success of modern cancer drugs, the cure of patients is hampered by the notorious tumor heterogeneity and plasticity of melanoma leading to therapy failure in most cases and limited diagnostic options to monitor therapy response/resistance over time.

Recent work and topics

Tumor cell state switching & Therapy resistance in melanoma
Accumulating evidence indicates that therapy and immune escape in cancer is not solely driven by genetic evolution, but also by the epigenetic adaptive plasticity of cell phenotypes. Current findings by us and others point to an understanding of melanoma as a highly plastic system, in which tumor cell subpopulations dynamically switch between multiple cellular identities (e.g. differentiated vs. de-differentiated, OXPHOS-driven vs. glycolytic, high vs. low ROS detoxifying, etc.) depending on the current therapeutic and microenvironmental context. As initiator and consortial member of the Clinical Research Unit CRU/KFO 337 (PhenoTImE), we are using established and patient-derived cell systems to study and track tumor cell states across different cancer entities in the context of signaling-targeted and immune checkpoint therapies. Technology-wise, our group applies for example cell state-directed fluorescent reporter constructs, multiplexed CODEX™ imaging, mass-spectrometry imaging, mass-spectrometric metabolomics, and, for cell manipulation, CRISPR/Cas9 editing and chemical compound libraries.
Dynamic therapy and immune escape. While the majority of cancer cells can be eliminated by first-line therapies and immune cell activation, a small portion of the tumor survives (minimal residual disease) and gives rise to clinical tumor relapse. The pre-resistant persister pool (DTP cells, dashed box) comprises a complex reservoir of different cell states from which repopulation of tumor heterogeneity dynamically emerges.
© Alexander Roesch

Regulation of tumor cell fate decisions
The phenotypic capacity to reversibly switch between different cell states provides a fundamental advantage for tumors to survive selective pressure. Our group has focused on the preclinical and clinical description of melanoma persister cell states over the past years. We have shown that melanomas irrespective of their genetic heterogeneity comprise multi-drug resistant, non-proliferating cells that repopulate the tumor and are identified by high expression of the histone H3K4 demethylase KDM5B (JARID1B). However, the exact role of KDM5B for regulating cellular identity and ways to exploit it for biomedical applications are still unclear; especially since the expression of KDM5B does not follow a static hierarchy, but is dynamically changing depending on the therapeutic con-text. Specifically, we hypothesize that KDM5B is a dynamic molecular switch that interconnects cell differentiation and cell cycle control. A deeper understanding of the switches between KDM5B levels and how this triggers transitions in cell states could lead to novel drugs with significant improvement in cancer treatment.

Dynamic cell fate decisions. Melanoma cells can dynamically shift between different differentiation, metabolic, and cell cycle states depending on the expression level of KDM5B/JARID1B.
© Alexander Roesch

Liquid biopsy and Tumor monitoring
Personalized oncology has not managed yet to overcome the limitation by inter- and intra-tumor heterogeneity in patients. This dilemma is even further complicated by the limited real-time information that translational research in humans but also in common mouse models can provide due to the high invasiveness of its mostly tissue biopsy-based design. In contrast, liquid biopsies, comprising the noninvasive analysis of circulating tumor-derived material in the blood, represent an innovative tool in precision oncology to overcome current limitations associated with tissue biopsies; however, mostly unspecific markers are used for the sole assessment of tumor load so far, while biology-specific read-outs are not yet implemented. As part of the EU Horizon 2020 Marie Curie Training Program MELGEN and an active member of the European Liquid Biopsy Society (ELBS), we have established a multi-analyte, multi-marker liquid biopsy platform for diagnostic and therapy monitoring application. The goal of our LB projects is to develop a multilateral liquid biopsy platform that allows tumor load monitoring and assessment of tumor biological processes, such as phenotype switching, in melanoma patients and mouse models.
ctDNA-based serial monitoring for early prediction of therapy outcome. Serial monitoring of mutant ctDNA copies under therapy predicts clinical progression earlier as compared to routine radiologic scans
© Alexander Roesch

Read more