Highly efficient XIST reactivation in female hPSC by transient dual inhibition of TP53 and DNA methylation during Cas9 mediated genome editing

Abstract

The irreversible erosion of X-chromosome inactivation (XCI) due to repression of the long non-coding RNA XIST presents a major challenge for disease modeling and raises safety concerns for the clinical application of female human pluripotent stem cells (hPSCs) due to the aberrant overexpression of X-linked genes. While Cas9-mediated non-homologous end joining (NHEJ) targeting the XIST promoter can induce DNA demethylation and restore XCI by reactivating XIST, its efficiency remains low. Here, we introduce a highly efficient strategy for XIST reactivation by combining TP53 inhibition with suppression of DNA methylation maintenance during Cas9-mediated NHEJ. This dual-inhibition approach increased the proportion of XIST-positive hPSCs from ~5 to ~43.7%, providing a robust method for stabilizing XCI in female hPSCs for diverse applications.

1. Introduction

Human pluripotent stem cells (hPSCs) offer valuable opportunities for both basic disease modeling and translational applications. Patient-specific induced pluripotent stem cells (iPSCs) enable genotype–phenotype correlation studies and facilitate autologous cell transplantation. However, under conventional culture conditions, the long non-coding RNA XIST is irreversibly silenced in female hPSCs, leading to the erosion of X-chromosome inactivation (XCI) [2, 4, 6, 15, 28]. Loss of XIST expression results in bi-allelic activation of X-linked genes and their subsequent overexpression [15, 18]. Given that the X chromosome harbors over 500 genes [27], XCI erosion can profoundly impact differentiation outcomes and limit the clinical utility of female hPSCs [15, 18]. This issue is particularly problematic in disease modeling using female patient-derived iPSCs, including those for X-linked disorders, where XCI erosion hampers the accurate recapitulation of disease phenotypes in vitro. Furthermore, overexpression of cancer related X-linked genes raises concerns about the safety and applicability of female hPSCs for clinical use [2]. Thus, XCI erosion represents a critical, yet often overlooked, source of variability that compromises experimental reproducibility and the therapeutic potential of female hPSCs.

Our previous work demonstrated that XIST repression in female hPSCs is caused by DNA methylation (DNAme) at XIST promoter regions, facilitated by de novo DNA methyltransferase activity [6]. Recently, we showed that DNA hypermethylation could be reversed via Cas9-mediated non-homologous end joining (NHEJ), leading to XIST reactivation through endogenous transcriptional mechanisms [18]. While DNA demethylation approach could reactivate XIST in female hPSCs, its efficiency was limited, with only ~5% of transfected cells showing reactivation. To enhance this, we employed homology-directed repair (HDR) in Cas9-mediated gene editing, introducing a small targeting vector containing a 1.5 kb region of the XIST promoter along with an antibiotic selection marker. Following selection, XIST reactivation efficiency increased to ~20% [18]. However, the inclusion of exogenous sequences in up-stream of XIST gene during HDR raises concerns about potential unintended effects on XIST expression during disease modeling and clinical applications.