Advances in oncology increasingly show that cancer is not driven by genetic mutations alone. Just as a city depends on well-connected roads to function, cells rely on a highly organized 3D DNA structure to maintain healthy activity. When this genomic “infrastructure” deteriorates, even subtly, essential communication pathways break down, creating opportunities for disease to emerge.
At the 2025 American Society of Hematology (ASH) meeting, researchers from the Sylvester Comprehensive Cancer Center at the University of Miami presented groundbreaking findings that place genome architecture at the center of lymphoma development. Led by cancer researcher Dr. Martin Rivas, the team demonstrated how partial loss of two key architectural proteins, SMC3 and CTCF, can disrupt critical DNA loops that keep tumor suppressor genes active. Their study introduces the concept of architectural tumor suppression, marking a significant shift in how scientists understand the earliest steps of blood cancer.
Architecture as a form of tumor defense
In healthy cells, enhancer-promoter loops act like electrical wiring, allowing regulatory DNA regions to switch genes on at the right time. SMC3 and CTCF help maintain these loops. When the levels of these proteins drop, what the researchers refer to as haploinsufficiency, short-range loops begin to erode.
Rather than collapsing the entire genome, this selective breakdown silences key tumor suppressor genes such as Tet2, Kmt2d and Dusp4. As a result, developing B-cells get stuck in an immature state, creating conditions that can ultimately lead to malignancy.
AI exposes hidden patterns
To uncover these architectural disruptions, the team relied heavily on AI-powered computational tools. By integrating multi-layered datasets, including Hi-C chromatin maps, single-cell RNA sequencing and epigenetic profiling, they were able to visualize dynamic shifts in DNA structure that would be impossible to detect manually.
“AI allowed us to see patterns invisible to the human eye, how losing just one copy of a gene reshapes the 3D genome,” said Rivas. This convergence of computational biology and cancer genomics offers a deeper understanding of the molecular events that precede lymphoma.
New biomarker and new therapeutic avenues
The researchers also found that patients with diffuse large B-cell lymphoma (DLBCL) who exhibit lower SMC3 expression have significantly worse outcomes, suggesting genome architecture itself may serve as a prognostic biomarker. Beyond diagnostics, the findings could inspire an entirely new generation of therapies aimed at restoring or mimicking healthy DNA looping. Instead of simply correcting mutations, future treatments may focus on stabilizing the genomic scaffolding that keeps tumor suppressor genes active.
This research reframes cancer as a disease not only of genetic code but also of structural failure. By understanding how DNA architecture contributes to malignancy, the field moves closer to developing interventions that reinforce genomic stability. In other words, just as repairing a city’s infrastructure keeps neighborhoods connected and functioning, restoring DNA loops may be key to preventing cells from slipping into cancerous states.
