A UCSF study shows that the physical location of genes inside the nucleus can have a powerful influence on whether they are switched on or off, offering a new branch of investigation for studying disease
Open a genetics textbook and you will likely find an explanation for gene expression laid out along a flat line: a strand of DNA punctuated by genes, promoters, regulatory elements and chemical marks that help determine whether a gene is switched on or off.
But that flat line actually belongs to a complex 3D architecture within the spherical nucleus of a cell. Twisted into its iconic double helix and wrapped around histone proteins, DNA is packed in and anchored to different locations within the nucleus. Some genes wind up hugging the periphery, while others sit near specialized structures deeper inside.
New research from UC San Francisco shows that where a gene is located within this architecture can profoundly affect how it behaves.
In a study published July 22 in Nature, researchers mapped the changing positions of genes as stem cells in the developing human brain matured into neurons. Hundreds of genes important for neuronal development moved away from the inner lining of the nucleus, called the lamina, and relocated to structures within the nuclear interior called nuclear speckles. Genes that made this move could become more than eight times as active.
In total, 739 genes detached from the lamina, and 305 of them relocated to nuclear speckles as the cells matured into neurons. These included SATB2 and MEF2C, transcription factors with important roles in cortical development and autism spectrum disorder, along with neuronal genes such as SYT1, NRG3 and ANKS1B that have been implicated in neurodevelopmental disorders and schizophrenia.
“Our data establish a paradigm in which knowing the spatial location of a gene is necessary for understanding its regulation and function,” said Daniel Lim, MD, PhD, professor of neurological surgery and the senior author of the study. “This principle is fundamental enough to affect our understanding of gene expression in essentially every cell type, a wide range of diseases such as cancer, perhaps even age-related decline.”
A floor plan for the genome
Nearly every cell in the body contains the same DNA, but what makes a neuron different from a muscle cell or a skin cell is the combination of genes that are expressed. One of biology’s central problems is how cells make the genetic programs they need readily available while keeping thousands of others silent.
Lim likens pushing the unneeded sections of the genome into boxes in the attic. “Your cell doesn’t need most of the information,” he said. “And so that’s where you put things that you want durably off.”
Lim’s father was an architect and his brother is an engineer, and Lim suspects that growing up around people who thought about space and structure influenced the way he approaches biology.
“The space in an attic is very different from a kitchen or a living room,” he said. “And this concept of spatial compartmentalization applies to biology too.”
According to the new study, the attic appears to be the nuclear lamina. Large sections of the genome attach to it in regions called lamina-associated domains, or LADs. In the developing brain cells studied by Lim’s team, LADs encompassed about 30% to 40% of the genome and generally contained genes with very low levels of activity.
Meanwhile, genes located in nuclear speckle-associated domains (SPADs) were expressed at much higher levels. This area functions more like a kitchen with all the appliances you need to cook up a recipe, or a living room with stereo equipment that can play records. It contains the proteins that transcribe DNA to RNA.
Lim and his team tracked the genes’ movement from LADs to SPADs using a technique they developed called GO-CaRT (Genome Organization with CUT and RUN Technology), which uses antibodies to isolate DNA fragments associated with specific nuclear compartments so they can be analyzed using high-throughput sequencing methods.
Beyond biochemistry
The new study focused on bivalent genes, which carry both an activating histone mark and a repressive one. Scientists have long thought of these genes as being held in a “poised” state: mostly quiet, but prepared to turn on when their repressive marks are removed.
Strikingly, Lim and his colleagues discovered that developmental genes are held in their silent, poised state by their location at the lamina, not simply by a repressive chemical mark (in this case, H3K27me3). When they stripped the H3K27me3 mark away, genes anchored at the lamina stayed silent.
“A prevailing view for studying genetics and epigenetics is that everything is happening in solution, like in a test tube,” said Lim. “And while this view of biochemistry is important, we're now seeing that those chemical reactions that control gene expression are happening in different places inside the nucleus, and that this spatial location really matters.”
Implications for disease
The results of the new study extend well beyond brain development. Gene expression is disrupted in cancer, inflammation and aging, and the implications may be especially significant for brain disorders. Most cells periodically divide, dismantling and reconstructing much of their internal organization. Many of the neurons in an adult brain were formed before birth and must preserve the same genome—and its spatial arrangement—for a lifetime.
The nuclear lamina is known to erode over time, a feature of both neurodegenerative diseases and premature-aging syndromes. Lim suspects that the attic that once kept essential machinery organized may begin to crumble. Shelving may sag and the boxes that once safely stored genes away could become less secure, allowing genes to become inappropriately accessible.
“The genes that drive cancer are probably not in those boxes anymore,” Lim said.
That work is ongoing and was not part of the Nature study. But it illustrates why Lim believes spatial genome organization could be relevant across many areas of biology.
The technique his group developed is deliberately straightforward. Researchers already studying chromatin modifications can add GO-CaRT to their existing experiments without adopting an entirely new experimental system.
Lim hopes other laboratories will now apply the approach to their own studies in other tissues, stages of development and diseases.
Lim’s laboratory has begun studying age-related changes in mice and plans to investigate post-mortem human tissue, asking how nuclear architecture changes during normal aging and whether those changes are accelerated in neurodegenerative disease. Early evidence also implies that unpacking sections of the genome from LADs may be involved in some types of childhood cancers, including pediatric gliomas.
Curiosity that drives discoveries
For many years, Lim's lab studied chromatin regulators—proteins that modify the way DNA is packaged. Mutations in chromatin regulators are frequently associated with neurodevelopmental disorders, cancer, and autism spectrum disorder.
“We always got to this point where we could see the cellular effect of a mutation in a chromatin regulator, but we couldn’t figure out what that chromatin regulator was really doing to regulate gene expression because the traditional linear view of the genome didn’t explain the data,” Lim said. “So, I thought that we were missing something.”
That uncertainty led the team to ask what else might be regulating gene expression, ultimately revealing an unexpected role for spatial organization inside the nucleus.
UCSF has long recognized the value of fundamental discovery, supporting basic research that has changed our understanding of biology, transformed medicine and helped give rise to the modern biotechnology industry.
That legacy includes Elizabeth Blackburn’s discovery of telomerase, Ron Vale’s foundational work on molecular motors and Herb Boyer’s development of recombinant DNA technology. Each began as an effort to answer a fundamental question in biology, without an immediate clinical application in mind. Yet over time, those discoveries opened entirely new areas of medicine—from telomerase-targeted therapies and drugs that act on molecular motors to recombinant DNA medicines such as insulin, human growth hormone and vaccines.
This new paper fits into that tradition. “On its face, it's a study of how the genome is spatially organized inside the nucleus during human brain development,” said Lim. “But the principle we uncovered may fundamentally change our understanding of gene expression, and this may prove important to the development of therapeutic strategies for aging and disease.”
Open Access Publication: Sajad Hamid Ahanger, Evan R. Semenza, Chujing Zhang, Eugene Gil, Mitchel A. Cole, Serena Huei-An Lu, Li Wang, Arnold R. Kriegstein, Daniel A. Lim. Subnuclear genome compartmentalization controls bivalent chromatin activity. Nature. 2026 Jul 22;657(8131):539-548. doi: 10.1038/s41586-026-10832-w