Lassi Paavolainen

ROR1-PI3K/AKT signaling drives adaptive resistance to cell cycle blockade in TP53 mutated ovarian cancer

Abstract Drug resistance remains a major challenge to durable responses in ovarian cancer, the fifth leading cause of cancer-related death among women. In this study, we developed long-term resistant (lt-res, several months) pre-clinical models of two drugs inducing mitotic arrest in TP53 -mutated cells: adavosertib (ADA), an investigational WEE1 inhibitor targeting the DNA damage response and currently evaluated in clinical trials, and paclitaxel (PTX), a widely used chemotherapeutic agent in cancer care targeting microtubules. Through integrated multi-omics functional profiling, we identify a shared PI3K/AKT-regulated signaling node that governs drug adaptation across all lt-res models. This node modulates the activity of DNA-damage responses and genotoxic stress to toggle between two adaptive states: activated PI3K/AKT driving a proliferative “fast-bypass” program with sustained cell cycle progression and mitotic evasion, or reduced PI3K/AKT signaling initiating a “slow-repair” state characterized by DNA damage checkpoint engagement, replication slowdown, and increased drug efflux. Notably, upregulation of receptor tyrosine kinases, such as ROR1, was observed in both ADA and PTX lt-res models with activated PI3K/AKT signaling. Targeting ROR1 with zilovertamab-vedotin, a monoclonal antibody-drug conjugate, resulted in enhanced cytotoxicity, demonstrating a new approach against recurrent drug-resistant ovarian cancer.

Evaluating feature extraction in ovarian cancer cell line co-cultures using deep neural networks

Abstract Single-cell image analysis is crucial for studying drug effects on cellular morphology and phenotypic changes. Most studies focus on single cell types, overlooking the complexity of cellular interactions. Here, we establish an analysis pipeline to extract phenotypic features of cancer cells cultured with fibroblasts. Using high-content imaging, we analyze an oncology drug library across five cancer and fibroblast cell line co-culture combinations, generating 61,440 images and ∼170 million single-cell objects. Traditional phenotyping with CellProfiler achieves an average enrichment score of 62.6% for mechanisms of action, while pre-trained neural networks (EfficientNetB0 and MobileNetV2) reach 61.0% and 62.0%, respectively. Variability in enrichment scores may reflect the use of multiple drug concentrations since not all induce significant morphological changes, as well as the cellular and genetic context of the treatment. Our study highlights nuanced drug-induced phenotypic variations and underscores the morphological heterogeneity of ovarian cancer cell lines and their response to complex co-culture environments.