Engineered colorectal cancer tissue recapitulates key attributes of a patient-derived xenograft tumor line

The development of physiologically relevant in vitro colorectal cancer (CRC) models is vital for advancing understanding of tumor biology. Although CRC patient-derived xenografts (PDXs) recapitulate key patient tumor characteristics and demonstrate high concordance with clinical outcomes, the use of this in vivo model is costly and low-throughput. Here we report the establishment and in-depth characterization of an in vitro tissue-engineered CRC model using PDX cells. To form the 3D engineered CRC-PDX (3D-eCRC-PDX) tissues, CRC PDX tumors were expanded in vivo, dissociated, and the isolated cells encapsulated within PEG-fibrinogen hydrogels. Following PEG-fibrinogen encapsulation, cells remain viable and proliferate within 3D-eCRC-PDX tissues. Tumor cell subpopulations, including human cancer and mouse stromal cells, are maintained in long-term culture (29 days); cellular subpopulations increase ratiometrically over time. The 3D-eCRC-PDX tissues mimic the mechanical stiffness of originating tumors. Extracellular matrix protein production by cells in the 3D-eCRC-PDX tissues resulted in approximately 57% of proteins observed in the CRC-PDX tumors also being present in the 3D-eCRC-PDX tissues on day 22. Furthermore, we show congruence in enriched gene ontology molecular functions and Hallmark gene sets in 3D-eCRC-PDX tissues and CRC-PDX tumors compared to normal colon tissue, while prognostic Kaplan-Meier plots for overall and relapse free survival did not reveal significant differences between CRC-PDX tumors and 3D-eCRC-PDX tissues. Our results demonstrate high batch-to-batch consistency and strong correlation between our in vitro tissue-engineered PDX-CRC model and the originating in vivo PDX tumors, providing a foundation for future studies of disease progression and tumorigenic mechanisms.

Automated summary

The development of an in vitro tissue-engineered CRC model using PDX cells provides a cost-effective and high-throughput alternative to the in vivo model. This model demonstrates high batch-to-batch consistency and strong correlation with the originating PDX tumors, making it a valuable tool for studying disease progression and tumorigenic mechanisms.