3D maps of colorectal tumors uncover novel structures, molecular gradients and cellular interactions
A Ludwig Harvard team led by Peter Sorger and Sandro Santagata applied highly multiplexed tissue imaging and 3D reconstruction methods along with spatial statistics and machine learning to identify cell types and states underlying morphological features of known diagnostic and prognostic significance in colorectal cancer. Their large scale 2D and 3D maps, reported in a January issue of Cell, revealed key immune interactions at the invasive boundary of tumors, interconnections between histological structures and morphological and molecular gradients of the sort typically seen in developing tissues. Lymphoid structures that were previously thought to be isolated and pools of mucin, for example, were each revealed in 3D reconstructions to be highly interconnected. Peter, Sandro and their colleagues also showed that regions within tumors vary in their degree of malignancy and their invasive properties. Gradients in histological features and the expression of oncogenes and epigenetic regulators of gene expression reflect transitions between more and less invasive regions. On the immunological front, the Ludwig Harvard team found that it is mainly other immune cells—namely, myeloid cells—and not just cancer cells that suppress anti-tumor T cell responses by engaging the PD-1 immune checkpoint. The study is part of Ludwig’s Tumor Atlas project and the National Cancer Institute’s Human Tumor Atlas Network (HTAN). To learn more and access the dataset, visit the Harvard Tissue Atlas.