AI uncovers how DNA architecture failures trigger blood cancer - Oncology & Cancer - Genetics

Credit: Pixabay/CC0 Public Domain

Cancer isn't just about broken genes—it's about broken architecture. Imagine a city where roads suddenly vanish, cutting off neighborhoods from essential services. That's what happens inside cells when the 3D structure of DNA collapses.

A new study presented at the 2025 American Society of Hematology (ASH) meeting, by Martin Rivas, Ph.D., a cancer researcher at Sylvester Comprehensive Cancer Center, part of the University of Miami Miller School of Medicine, revealed that even subtle disruptions in genome architecture can predispose individuals to lymphoma. This finding offers a new perspective on understanding and eventually treating blood cancers.

The study, titled "SMC3 and CTCF Haploinsufficiency Drive Lymphoid Malignancy via 3D Genome Dysregulation and Disruption of Tumor Suppressor Enhancer-Promoter Loops," introduced a new idea: architectural tumor suppression.

Proteins like SMC3 and CTCF don't just organize DNA—they actively prevent cancer by maintaining loops that connect gene "switches" (enhancers) to the genes they control (promoters). Lose even half of these proteins, and the loops start disappearing, silencing critical tumor suppressor genes.

"We've long known that mutations drive cancer," said Rivas. "But this work shows that architecture—the way DNA folds—can be just as important. It's like losing the blueprint for a building while construction is under way."

Using AI-driven analytics to interpret massive datasets from Hi-C maps, single-cell RNA sequencing and epigenetic profiles, the team uncovered a striking pattern:

SMC3 or CTCF haploinsufficiency (partial loss) doesn't wreck the entire genome structure. Instead, it erodes short-range enhancer-promoter loops—the wiring that keeps tumor suppressor genes like Tet2, Kmt2d, and Dusp4 active.

Without these loops, B-cells hit a "decision bottleneck" and fail to mature into plasma cells, creating fertile ground for malignancy.

AI tools helped integrate these complex layers of data, revealing how architectural changes ripple through gene expression and cell fate.

"This is where computational biology shines," Rivas added. "AI allowed us to see patterns invisible to the human eye—how losing just one copy of a gene reshapes the entire 3D landscape."

The findings aren't just theoretical. Patients with diffuse large B-cell lymphoma (DLBCL) who have lower SMC3 expression fare worse. This suggests that genome architecture could become a biomarker for prognosis—and maybe a target for therapy. Instead of fixing mutations, future treatments might aim to restore proper looping or mimic its effects.

This research reframes cancer biology: it's not only about the genetic code but also about the scaffolding that holds it together. By understanding architectural tumor suppression, scientists can explore therapies that stabilize genome structure—an entirely new frontier in oncology.

"We're entering an era where cancer treatment could mean repairing architecture, not just fixing broken genes," said Rivas. "That's a paradigm shift."

In the end, think back to that city analogy: when the streets disappear, neighborhoods become isolated and life grinds to a halt. Inside cells, when DNA loops vanish, tumor suppressor genes lose their lifelines—and cancer finds a way in. Restoring those connections could be the key to keeping the city—and the cell—alive and thriving. 

Provided by University of Miami Leonard M. Miller School of Medicine 

by University of Miami Leonard M. Miller School of Medicine

edited by Sadie Harley, reviewed by Robert Egan 

Source: AI uncovers how DNA architecture failures trigger blood cancer