The research, led by Lu-Yan Chan, Ming-Kin Wong, and Yong-Hong Yan as co-first authors in the laboratory of Zhongying Zhao at Hong Kong Baptist University, with collaborators at the National Institute of Biological Sciences in Beijing and the University of Science and Technology of China, focuses on CYL-1, the worm counterpart of cyclin L1. Cyclins are best known as the rhythmic partners of cyclin-dependent kinases, the enzymes that drive cells through the phases of the cell cycle. Yet a growing body of work has revealed that many cyclins moonlight in other processes, and cyclin L had previously been implicated in pre-mRNA splicing through its interaction with CDK-11 in cultured cell lines. What remained murky was whether this splicing connection mattered in a living, developing animal—and what else the protein might be doing there.

To find out, the team first established where and when CYL-1 operates. Fluorescently tagged versions of the protein revealed that CYL-1 is expressed ubiquitously, residing in cell nuclei throughout the entire course of worm development. That broad, persistent nuclear presence hinted that the protein is not a niche player confined to one tissue or stage, but a general component of the gene-expression machinery whose loss should ripple across the organism. The next step was to identify who CYL-1 actually works with inside those nuclei.

Using co-immunoprecipitation followed by mass spectrometry—a technique in which a bait protein is pulled down from worm extracts along with everything physically bound to it, and the captured partners are then identified by their peptide fingerprints—the researchers mapped CYL-1’s interaction network in vivo. The results confirmed that CYL-1 forms complexes with two worm CDK-11 isoforms, CDK-11.1 and CDK-11.2, and also identified a highly conserved protein called EELO-1, named for its association with early transcriptional elongation, as a core interactor. Together with the splicing-associated factor SAP30BP highlighted in the study’s title, these proteins define a complex that sits at the junction of transcription and RNA processing.

Genetic analysis then revealed how essential this machinery is. When the researchers interfered with cyl-1, cdk-11.1, cdk-11.2, or eelo-1 using RNA interference, embryos failed to develop normally, and the two CDK-11 paralogs were shown to act redundantly: knocking down either alone was survivable to a degree, but compromising both sharply reduced embryonic viability. This redundancy is a classic signature of duplicated genes that share a critical function, and it means the worm tolerates losing one copy of the kinase only because the other can carry the load. The four proteins, in other words, are not optional accessories but load-bearing pillars of mRNA biogenesis during embryogenesis.

The mechanistic heart of the study lies in what those pillars actually do to RNA. Immunostaining assays, which use antibodies to visualize the location of specific molecular marks on proteins, showed that loss of CYL-1 disrupts events at the very start of the transcription cycle, affecting transcriptional initiation and the elongation of RNA polymerase along the gene body. RNA sequencing of embryos depleted of each complex member added a genome-wide view: beyond changes in overall transcript levels, the researchers documented widespread defects in pre-mRNA splicing, with alternative splicing patterns shifting substantially and differently depending on which of the four genes—cyl-1, cdk-11.1, cdk-11.2, or eelo-1—had been silenced. That last detail matters. If the four proteins were simply interchangeable parts of one machine, depleting any of them would produce the same splicing fingerprint. Instead, each leaves a distinct signature, suggesting partially specialized roles within a shared complex.

Perhaps the most striking finding came from asking a developmental question rather than a biochemical one. Using established cell fate markers—reporter constructs that fluoresce only when a cell has committed to a particular identity—the team demonstrated that CDK-11.1, CDK-11.2, and EELO-1 are involved in regulating cell fate specification during development. In an embryo, every cell carries the same genome; what distinguishes a muscle cell from a neuron is which subset of genes is transcribed and how those transcripts are processed. A complex that influences both transcriptional output and splicing is therefore ideally positioned to shape that decision, and the new data indicate that it does. The result reframes splicing factors not merely as editors of genetic text but as participants in the regulatory logic that assigns identities to embryonic cells.

Importantly, the study also uncovers an asymmetry within the complex. While CYL-1 is required for transcriptional initiation, elongation, and splicing, its partners CDK-11.1, CDK-11.2, and EELO-1 carry the additional responsibility of governing cell fate determination—a role that CYL-1 itself did not show in these assays. This differential division of labor between a cyclin and its interactors adds a nuanced twist to cyclin biology. It suggests that the familiar cyclin-CDK pairing can be modular: the kinase and its associated factors may perform functions beyond, and independent of, the cyclin that helps recruit or activate them. For an organism as experimentally tractable as C. elegans, whose transparent embryos allow researchers to watch every cell division in real time, such a system offers a powerful platform for dissecting how those roles are partitioned at the molecular level.

The evolutionary implications extend well beyond worms. CDK-11 and cyclin L are conserved across metazoans, including mammals, where CDK11 has been implicated in cell cycle control, apoptosis, and RNA processing, and where splicing deregulation is increasingly recognized as a driver of developmental disorders and cancer. EELO-1’s conservation means the worm findings likely illuminate a broadly shared architecture of gene expression. If a CDK-11-containing complex couples transcriptional elongation and splicing to cell fate decisions in nematodes, the same coupling may operate in vertebrate embryos, stem cells, and disease states where differentiation goes awry. The study does not test these possibilities directly, but it establishes the in vivo framework that such future work in other organisms can build upon.

What makes the work resonate beyond its immediate field is the picture it paints of gene expression as a single, physically continuous process rather than a relay of disconnected stages. The RNA polymerase that reads a gene, the factors that splice its transcript, and the regulatory network that decides a cell’s destiny are, in this emerging view, held together in one macromolecular assembly. A defect anywhere in that assembly—an underactive kinase, a missing cyclin, an absent elongation factor—can propagate outward from a biochemical glitch to a developmental outcome: an embryo that fails to survive, or a cell that adopts the wrong identity. By grounding that logic in the concrete genetics of a living animal, the Hong Kong-led team has turned a splicing complex into a lens on one of biology’s deepest questions—how a genome becomes an organism—and provided a molecular handle that researchers studying development, RNA biology, and disease will now be eager to pull.

Subject of Research: The role of a CYL-1/CDK-11/EELO-1 complex in pre-mRNA splicing and cell fate specification during C. elegans development

Article Title: A CDK-11/SAP30BP/CYL-1 complex links pre-mRNA splicing to developmental cell fate decisions

Article References: Chan, L.-Y., Wong, M.-K., Yan, Y.-H., Ho, V. W. S., Ma, Y., Xie, D., Ye, P., Zeng, J., Dong, M.-Q., Guang, S., & Zhao, Z. (2026). A CDK-11/SAP30BP/CYL-1 complex links pre-mRNA splicing to developmental cell fate decisions. BMC Biology. https://doi.org/10.1186/s12915-026-02734-0

Image Credits: AI Generated

DOI: 10.1186/s12915-026-02734-0

Keywords: CYL-1, CDK-11, EELO-1, pre-mRNA splicing, cell fate determination, C. elegans, embryogenesis, cyclin-dependent kinase, transcriptional elongation, developmental genetics, mRNA biogenesis, BMC Biology

The problem with permanent control is well documented in the field. When engineers force a gene to remain continuously active or silenced, the cell may reach the desired phenotype, but it often loses plasticity, the adaptive flexibility that allows cells to respond to future environmental changes. Worse, permanent interventions can generate abnormal equilibrium states that do not exist in natural biology, raising safety concerns for any therapeutic application. Cho’s team set out to answer a deceptively simple question: is it possible to guide a cell to a stable desired state with a minimal, transient push, and then let the cell’s own regulatory machinery hold it there?

NUDGE, which stands for natural dynamics control of gene regulatory networks, approaches this question through computational logic modeling. The framework builds a model of how key genes inside a cell influence one another and uses that model to determine which stable state, or attractor, the cell will ultimately settle into. It then mathematically decomposes the logical functions describing these gene interactions to identify the minimal combination of control factors required to convert a cell from its current state to the desired one. In practical terms, if the goal is to differentiate a stem cell into a cardiomyocyte, NUDGE functions as a blueprint for minimal intervention, specifying exactly which molecular targets need to be briefly activated or inhibited and for how long.

One of the most striking features of the framework is its mathematical rigor. The exact version of NUDGE comes with a formal guarantee: it identifies all minimal control combinations that lead to the desired phenotype within the model. Because exhaustive search becomes intractable for large molecular networks, the team also developed an efficient approximation method that extends the approach to systems with hundreds or thousands of interacting components. This combination of exactness at small scale and scalability at large size positions NUDGE as a general-purpose tool rather than a bespoke solution for a single cell type.

The performance numbers reported in the study are substantial. When the researchers assessed permanent control across 63 published biological network models in which fixing a single node was sufficient to achieve the target phenotype, 55 of those models showed reduced plasticity or attractors absent from the uncontrolled network, confirming the limitations of conventional approaches. Separately, NUDGE’s approximation method achieved an average intervention error below 0.01 in more than 90 percent of 552 control problems spanning 69 large biological networks. Error in this context measured the fraction of simulated converged states that displayed the undesired phenotype. Among minimum-sized, error-free interventions pooled from three competing methods for each problem, NUDGE found an average of 83 percent, compared with 59 percent for IBMFA and 53 percent for LDOI, two established control-theory baselines.

Beyond identifying which factors to stimulate, NUDGE offers something rare among control frameworks: a window into the trajectory itself. The method can analyze which pathways a stimulated cell follows as it transitions toward the target state and how stably that state is maintained once reached. This means researchers can anticipate, before running a single experiment, whether unexpected or pathological states might emerge along the way. For therapeutic applications, where an off-target cell state could mean disease rather than repair, this predictive capability could prove as valuable as the control strategy itself.

To test whether the framework explains real biology rather than merely manipulating abstract models, the team applied NUDGE to three distinct processes. In cardiomyocyte differentiation, the framework reproduced the known pattern in which MESP1, a gene critical for heart development and regeneration, is briefly active early in differentiation but silent in mature cardiomyocytes. The model demonstrated how a single temporary intervention involving MESP1 and one additional regulatory target can guide differentiation without keeping MESP1 continuously active, mirroring what nature itself appears to do.

The second and third applications extended the framework into immunology. In modeling mast cell formation, the cells responsible for allergic responses, NUDGE identified GATA2 and GATA1 as key regulators of cell fate during differentiation. In the macrophage model, the team derived optimal strategies for a single temporary intervention, including one using the signaling molecule IL-4, to induce anti-inflammatory M2 states. Crucially, the intervention preserved the cell’s capacity to transition between inflammatory and anti-inflammatory phenotypes and among M2 subtypes, exactly the kind of retained plasticity that permanent gene locking destroys.

The significance of this work lies in its generality. NUDGE provides a framework that identifies where a single temporary intervention can guide a cell to a desired state, provided a stable state with the desired phenotype exists in the network’s uncontrolled dynamics. The anticipated applications are broad: steering stem cells into specific lineages such as cardiomyocytes, normalizing diseased or aged cells, regenerating damaged tissue, modulating immune responses in autoimmune disease, and designing cancer reversion therapies that return malignant cells to a near-normal state rather than killing them outright. Professor Cho emphasized that NUDGE is not limited to returning a specific cell to a healthy state but constitutes a general-purpose cell-control technology that guides a cell from its current state to a desired state through a single minimal stimulus, with potential as a foundational tool across stem cell biology, aging research, regenerative medicine, and oncology.

The study, conducted by co-first authors Ferio Brahmana and Corbin Hopper, with Woojeong Lee as co-author, all of KAIST’s Department of Bio and Brain Engineering, was published online on September 14 in the Proceedings of the National Academy of Sciences under the title Uncovering minimal control of cell fate by natural dynamics. The work was supported by National Research Foundation of Korea grants funded by the Korean government, the Ministry of Science and ICT, the Korea Dementia Research Project, and internal KAIST convergence research programs. If its predictions hold up in experimental and clinical settings, the idea that a brief nudge, rather than a permanent rewrite, is enough to redirect a cell’s destiny could reshape how regenerative medicine is practiced, moving the field from forced reprogramming toward something closer to a conversation with the cell’s own dynamics.

Subject of Research: Computational control of cell fate through minimal temporary intervention in gene regulatory networks

Article Title: KAIST finds a way to redirect cell fate with just a single stimulus

Article References: KAIST finds a way to redirect cell fate with just a single stimulus. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: cell fate control, gene regulatory networks, NUDGE, KAIST, stem cell differentiation, cardiomyocytes, macrophages, regenerative medicine, cancer reversion, systems biology, attractor control, PNAS