Salk Institute Study Reveals Dynamic DNA Motion's Pivotal Role in Cancer Development

Breakthrough Discovery: DNA's Restless Dance and Its Link to Cancer

The human genome, that vast repository of genetic information packed into every cell nucleus, is far from static. Recent research from the Salk Institute for Biological Studies has unveiled that DNA undergoes constant motion—folding, unfolding, and refolding in three-dimensional space. This dynamic process, driven by protein complexes like cohesin and its loader NIPBL (NIPBL, also known as nipped-B-like protein), plays a crucial role in regulating gene expression and maintaining cell identity. Disruptions in these movements could be a key driver in cancer development, where cells lose their specialized functions and proliferate uncontrollably. 73 75

Led by Jesse Dixon, MD, PhD, and first author Tessa Popay, PhD, the study published in Nature Genetics on February 16, 2026, demonstrates how these looping dynamics vary across the genome. Active genes associated with specific cell functions turn over loops rapidly, while silent regions maintain more stable structures. This finding challenges previous views of the genome as a fixed scaffold, highlighting its fluid nature essential for healthy cellular behavior. 116

Decoding the Basics of Genome Architecture

To grasp this discovery, consider the deoxyribonucleic acid (DNA) structure: two meters of it coiled into a nucleus smaller than a pinhead. Proteins organize it into loops via chromatin looping, where distant DNA segments come close in 3D space to interact. Cohesin, a ring-shaped complex, extrudes loops along DNA strands, pinned by CTCF (CCCTC-binding factor) proteins at specific sites. NIPBL loads cohesin onto DNA and fuels its movement, akin to a motor along a track. 73

Past studies using techniques like Hi-C (high-throughput chromosome conformation capture) mapped static snapshots, showing compartments and topologically associating domains (TADs). However, Dixon's team sought the kinetics—the speed of loop formation and dissolution—revealing it's not uniform. In human retinal pigment epithelial (RPE-1) cells, depleting NIPBL halted new loops, causing uneven unfolding: some regions relaxed in minutes, others persisted for hours. 74

Innovative Methods Uncover Hidden Dynamics

The researchers employed cutting-edge tools for real-time insights. Using the dTAG system, they inducibly degraded NIPBL or RAD21 (a cohesin subunit) in RPE-1 cells, tracking changes over 4-24 hours with Hi-C at 10-kb resolution. They synchronized cells for mitotic exit—a phase when loops reform—to pinpoint NIPBL's role. SLAM-seq (SLAM, selective labeling using alkylation for metabolic sequencing of RNA) measured nascent transcripts, linking structural changes to gene activity. 116

Extending to human induced pluripotent stem cells (iPSCs) differentiated into cardiomyocytes (heart cells) and neurons, they found cell-type-specific effects. Over 1,392 genes were dysregulated across lineages, 78% unique to each type, enriched in functions like cell migration and shape. ChIP-seq (chromatin immunoprecipitation sequencing) mapped protein bindings, revealing STAG1 (stromal antigen 1) stabilizes persistent loops. 75

This multi-omics approach (integrating Hi-C, RNA-seq, ChIP-seq) provided a dynamic view, clustering 16,860 loops by NIPBL dependency—from minimal to maximal—based on fold-changes and pile-up analyses.

Core Findings: Fast Loops for Active Genes, Stability for Silence

Key revelation: loop turnover rates correlate with activity. Dynamic (fast-unfolding) loops cluster at expressed genes coordinating cell functions, while stable loops mark repressed regions with repressive marks like H3K27me3 (histone H3 lysine 27 trimethylation). Upon NIPBL loss, 549 genes altered during mitotic exit, mostly downregulated, affecting morphology—cells elongated, hinting at identity shifts. 116

• Maximal-dependency loops lost rapidly, linked to enhancers.
• Mixed-dependency persisted, reliant on STAG1 and post-mitotic NIPBL.
• Sensitive genes showed closer super-enhancer proximity (within 500 kb of transcription start sites, TSS) and weaker TSS insulation, allowing cross-gene enhancer access.

In neurons, genes like BDNF (brain-derived neurotrophic factor) dropped; in heart cells, CASZ1 (castor zinc finger 1) for cardiac development. Continuous looping "remembers" identity by reinforcing enhancer-promoter contacts. 74

Photo by Ian on Unsplash

Dynamic DNA Motion: Guardian of Cell Identity

Cells differentiate into over 200 types despite identical DNA, via epigenetic regulation including 3D folding. Dynamic extrusion brings enhancers (gene boosters) to promoters, activating lineage genes. Popay noted, "The continuous folding and unfolding may help a cell ‘remember’ who it is supposed to be." Dixon added, "Genome folding machineries tightly control cell identity... mutations lead to syndromic conditions." 73

This preservation explains tissue-specific phenotypes in disorders. In cancer, hijacking these dynamics erodes identity—normal cells revert to proliferative states, mimicking embryonic growth.

Unraveling Cancer's Reliance on Faulty Folding

Cancer arises from mutations accumulating over time, with up to 66% from random DNA replication errors. Cohesin pathway alterations are frequent: STAG2 mutated in 15-40% bladder cancers, 19% endometrial. NIPBL overexpression correlates with poor prognosis, lymph node metastasis in some tumors. 79 82

The Salk study suggests dysfunctional extrusion promotes oncogene activation by altering enhancer access, akin to structural variants Dixon's lab previously linked to cancer genes. Globally, cancer caused 10 million deaths in 2022; US estimates for 2026 project 2 million new cases. Up to 5% Americans carry pathogenic variants raising risk, underscoring somatic changes like cohesin defects.Read the full Nature Genetics paper 116

For details on methods and data, see the Salk Institute press release. 73

From Developmental Disorders to Oncology Links

NIPBL mutations cause Cornelia de Lange syndrome (CdLS), affecting 1:10,000-30,000 births with growth delays, limb defects, intellectual disability—mirroring tissue-specific gene dysregulation. Cohesinopathies overlap autism spectrum disorders. Cancer exploits similar pathways: loss-of-function in tumor suppressors, gain in growth genes via misfolding. 80

Real-world case: bladder cancer patients with STAG2 mutations show distinct chromatin states, worse outcomes. This convergence positions folding dynamics as a therapeutic nexus.

Salk Institute: Hub for Genomic Innovation

Founded 1960 in La Jolla, California, Salk Institute pioneers plant biology to neuroscience, with Genomics Analysis group advancing Hi-C tech. Dixon, holding the Helen McLoraine Chair, builds on prior work mapping structural variants activating oncogenes. Collaborations with UC San Diego enhance translational impact. 86

The institute's Cancer Center integrates metabolism, epigenetics, fostering interdisciplinary training—ideal for aspiring researchers.

Photo by Google DeepMind on Unsplash

Future Directions: Targeting Dynamics Therapeutically

Prospects include drugs modulating NIPBL/cohesin activity, restoring folding in cancers. CRISPR screens could identify synthetic lethals, as cohesin mutants sensitize to WNT signaling inhibitors. Longitudinal studies tracking dynamics in patient tumors via single-cell Hi-C promise personalized medicine. 82

Challenges: tissue specificity demands precise interventions. Yet, as Dixon states, understanding "molecular machines" opens doors to preventing dysfunction in cancers and disorders.

Careers in Genome Dynamics Research

This study spotlights demand for experts in computational biology, stem cell differentiation, multi-omics. Universities like UC Riverside, UW Genome Sciences seek postdocs in integrative genomics. Roles span faculty in molecular biology to research associates analyzing Hi-C data, with salaries averaging $70,000-$120,000 for postdocs. 106

• Postdoctoral positions: Hi-C method development, iPSC models.
• Faculty tracks: Epigenetics, cancer genomics labs.
• Industry: Biotech firms like Illumina hiring for 3D genome sequencing.

Training in Python/R for bioinformatics, wet-lab skills in CRISPR/dTAG essential. Explore opportunities to contribute to next breakthroughs.