RNA edit swaps out OCT4 to make better stem cells

More than two decades after Shinya Yamanaka showed that four transcription factors could rewind adult cells back to an embryonic-like state, scientists have found a surprising new lever for controlling that transformation: a single chemical tweak on a noncoding RNA. In a study published in Nature Chemical Biology, a team led by Yanhong Shi at City of Hope and Chuan He at the University of Chicago reports that the enzyme PUS7, which installs pseudouridine modifications on RNA, acts as a molecular checkpoint for human somatic cell reprogramming. When PUS7 is missing, cells stall halfway to pluripotency; when it is supplied, reprogramming accelerates, and remarkably, PUS7 can even replace OCT4, the factor long considered indispensable for generating human induced pluripotent stem cells.

The discovery began with an unexpected observation. The researchers were reprogramming somatic cells from four individuals who carry pathogenic mutations in the PUS7 gene, mutations previously linked to intellectual disability and microcephaly. Using the canonical Yamanaka cocktail of OCT4, SOX2, KLF4 and MYC, known as OSKM, they found that cells from healthy donors readily produced colonies of NANOG-positive iPS cells, while the PUS7-mutant lymphoblastoid cell lines and fibroblasts generated dramatically fewer colonies, with sharply reduced expression of the pluripotency markers OCT4 and NANOG. Western blotting confirmed a complete loss of PUS7 protein in the mutant cells, pointing to the enzyme itself rather than any secondary effect as the culprit.

To trace where the reprogramming went wrong, the team performed RNA sequencing at five time points across the roughly three-week process. Reprogramming is known to proceed in two transcriptional waves: an early wave driven mainly by MYC and KLF4 that rewires metabolism and the cell cycle, and a later wave, driven by OCT4, SOX2 and KLF4, that activates the endogenous pluripotency network through events such as mesenchymal-to-epithelial transition and large-scale chromatin remodeling. In the PUS7-mutant cells, the first wave proceeded normally through day six, but the second wave never fired. Instead of activating pattern specification and morphogenesis programs, the mutant cells upregulated inflammatory and cytokine pathways and drifted into what the field calls a refractory state, a nonproductive intermediate that cannot complete the journey to pluripotency.

Crucially, rescue experiments showed that the enzymatic activity of PUS7 was required. Supplying wild-type PUS7 alongside OSKM restored colony formation and pluripotency marker expression in the mutant fibroblasts, whereas a catalytically inactive PUS7 mutant did not. That result redirected attention from PUS7’s known role in tRNA modification and translational control toward its RNA targets more broadly. Pseudouridine, the most abundant internal RNA modification in nature, is installed by a family of pseudouridine synthases and can alter RNA folding, stability and protein binding. The team therefore set out to map the pseudouridine landscape during reprogramming using BID-seq, a bisulfite-based technique that detects pseudouridine at single-nucleotide resolution.

Comparing isogenic wild-type and PUS7-knockout iPS cells, the researchers identified 84 pseudouridine sites dependent on PUS7, and after filtering for sites that also increase during reprogramming, one candidate rose to the top: uridine 250 of 7SK, a small nuclear RNA that associates with chromatin. The pseudouridylation level at this site rose from about 12 percent in fibroblasts to 29 percent in iPS cells, then collapsed to under 3 percent in PUS7-knockout cells, while the abundance of the 7SK transcript itself remained unchanged. The modification sits within a conserved PUS7 recognition motif in stem-loop 3 of the RNA, and targeted sequencing confirmed that it rises dynamically during reprogramming in healthy cells but not in PUS7-mutant ones.

Why would one pseudouridine on a noncoding RNA matter so much? The answer lies in 7SK’s role as a scaffold for chromatin regulators. Using RNA immunoprecipitation combined with targeted pseudouridine quantification, the team showed that subunits of the BAF chromatin-remodeling complex, including ARID1A, SMARCA4, SMARCD1 and SMARCC1, preferentially bind 7SK molecules carrying the pseudouridine at position 250, whereas binding of the canonical 7SK partners LARP7 and CDK9 was largely unaffected. An in vitro assay with purified ARID1A and synthetic 7SK RNAs, one unmodified and one site-specifically pseudouridylated, confirmed that the modification directly strengthens the interaction. In PUS7-knockout cells, the association between 7SK and BAF fell, and with it BAF occupancy on chromatin.

The downstream consequence was a genome-wide closing of chromatin. ATAC-seq revealed that accessibility loss in PUS7-knockout cells was concentrated at BAF-bound enhancers and super-enhancers rather than at all open regions, and CUT&Tag profiling showed reduced ARID1A and BAF155 binding together with diminished OCT4 occupancy at pluripotency-associated regulatory elements. Most strikingly, the team could bypass PUS7 entirely by fusing it to the RNA-binding protein dCas13b and guiding it to 7SK with a designed guide RNA. This programmable system restored pseudouridylation at position 250, re-recruited BAF, reopened chromatin and rescued iPS cell colony formation in PUS7-mutant fibroblasts, while a catalytically dead version of the fusion failed. The results invert the usual logic of gene regulation: rather than chromatin state dictating RNA output, an RNA modification feeds back to shape chromatin architecture.

Those mechanistic insights led to the boldest experiment. Because exogenous OCT4 has been associated with imprinting defects and restricted developmental potential, and because no factor had convincingly replaced it in human cells, the team swapped OCT4 out of the Yamanaka cocktail and substituted PUS7, creating a four-factor combination they call PSKM. From the same parental fibroblasts, PSKM generated typical iPS cell colonies with normal karyotypes, robust pluripotency marker expression and activation of endogenous OCT4, while SKM with an empty vector or catalytically inactive PUS7 produced almost nothing. Disrupting the BAF complex sharply reduced PSKM efficiency, confirming that the chromatin remodeler is a critical downstream effector. To the authors’ knowledge, PUS7 is the first RNA-modifying enzyme shown to functionally substitute for exogenous OCT4 in human reprogramming.

The quality of the resulting cells proved equally important. Compared with isogenic OSKM-derived iPS cells, PSKM-derived iPS cells more closely resembled human embryonic stem cells across multiple molecular benchmarks. At the DLK1-DIO3 imprinted locus on chromosome 14, a sensitive indicator of reprogramming fidelity, OSKM cells showed hypomethylation and aberrant overexpression of the long noncoding RNAs MEG3 and MEG8 along with nearby microRNA clusters, defects that were largely corrected in PSKM cells. PSKM cells also displayed lower levels of the DNA damage marker gamma-H2AX, comparable to embryonic stem cells. Functionally, PSKM-derived embryoid bodies grew larger and enriched developmental gene programs, and the cells differentiated more faithfully into neural progenitor cells and hematopoietic progenitor cells, with fewer off-target mesodermal signatures than their OSKM counterparts.

The most demanding test came from a three-dimensional model of postimplantation human development. When embedded in Matrigel, both embryonic stem cells and iPS cells self-organize into epiblast-like spheroids that undergo lumen formation and then a gastrulation-like program marked by the mesodermal factor Brachyury. PSKM-derived spheroids formed lumens similar in size to embryonic stem cell structures, polarized their F-ACTIN apically as mature epiblasts do, and matched embryonic stem cells in the timing and spatial pattern of Brachyury induction. OSKM-derived structures lagged on every measure, forming smaller lumens, diffuse actin polarity and prematurely delaminating Brachyury-positive cells, a pattern consistent with residual fibroblast memory. Together, these findings establish 7SK pseudouridylation as a previously hidden regulatory axis linking RNA chemistry to chromatin remodeling, and position PSKM as a next-generation platform for producing developmentally competent human iPS cells for disease modeling and regenerative medicine.

Subject of Research: PUS7-mediated pseudouridylation of 7SK RNA in human somatic cell reprogramming into induced pluripotent stem cells

Article Title: PUS7 pseudouridylation of 7SK orchestrates human somatic cell reprogramming into iPS cells

Article References: PUS7 pseudouridylation of 7SK orchestrates human somatic cell reprogramming into iPS cells. (n.d.). https://doi.org/10.1038/s41589-026-02339-2

Image Credits: AI Generated

DOI: 10.1038/s41589-026-02339-2

Keywords: PUS7, pseudouridine, 7SK RNA, iPS cells, reprogramming, BAF complex, chromatin accessibility, epitranscriptomics, OCT4, RNA modification, stem cells, regenerative medicine