CRISPR Prime Editing Models and Corrects Rare GDF11 Mutation Linked to Growth Disorder

Rare and ultra-rare genetic diseases affect millions globally, yet the path from genetic diagnosis to effective therapy remains enormously challenging. The Undiagnosed Diseases Network (UDN) has identified over 886 diagnoses through whole-genome and whole-exome sequencing, but therapeutic development for these highly individualized conditions lags far behind. This study addresses that gap by demonstrating a complete workflow — from disease modeling to mutation correction — using CRISPR prime editing for a single patient's de novo pathogenic variant in GDF11, a gene encoding a TGF-beta family member with roles in development, tissue homeostasis, and aging.

The specific variant studied is a C-to-G transversion at nucleotide position 6,529 of GDF11 (NM_005811.4), creating a premature stop codon at tyrosine 336 (Tyr336*). This mutation was identified in UDN participant 056, who presented with growth delay and multisystem abnormalities. Using prime editing, the team introduced this exact heterozygous mutation into HEK293T cells, generating clonal cell lines that faithfully model the patient's genotype. Western blot analysis confirmed that heterozygous (HET) cells exhibited significantly reduced GDF11 protein levels compared to wild-type controls, consistent with post-translational degradation likely mediated by endoplasmic reticulum- and Golgi-associated quality control pathways rather than nonsense-mediated mRNA decay alone.

Immunohistochemical analysis of Golgi morphology using the GOLPH2 marker revealed striking structural abnormalities in HET cells. Compared to wild-type cells, HET clones displayed a significantly increased number of compact, irregularly shaped Golgi structures per cell, consistent with Golgi fragmentation and stress. This finding is notable because GDF11 is a secreted protein that transits through the Golgi apparatus, and disruption of its processing may directly impair Golgi homeostasis. The Golgi phenotype provides a cellular mechanism linking the mutation to the patient's developmental and multisystem clinical features.

Transcriptomic profiling via RNA-seq (deposited at GEO: GSE312163) of wild-type versus HET HEK293T cells revealed broad dysregulation of gene networks. Downregulated pathways included metabolic and Golgi-linked biosynthetic genes, while upregulated pathways included cell-adhesion and extracellular matrix genes. These transcriptional shifts were interpreted as paralleling the participant's developmental, neural, and cardiovascular phenotypes, supporting a haploinsufficiency mechanism for the Tyr336* allele rather than a dominant-negative effect. Gene ontology enrichment analysis using ShinyGO and volcano plots generated with VolcaNoseR provided statistical support for these pathway-level changes.

For mutation correction, the team systematically tested multiple prime editing strategies varying the prime editor version (PE6 PEmaxΔRNaseH vs. PE7), pegRNA scaffold design (standard, tevopreq1, tmpknot), and guide RNA design tools (including Pridict AI-based design). PE7 combined with a Pridict-designed pegRNA emerged as the most effective ribonucleoprotein complex. Editing efficiency was further enhanced by incorporating a silent protospacer-adjacent motif (PAM)-disrupting mutation alongside the correction, which likely prevents Cas9 re-binding to the edited allele and reduces mismatch repair-mediated reversion. Editing outcomes were quantified using CRISPResso2 and EditR software. Together, these results establish a scalable, cost-effective pipeline for rare disease modeling and allele-specific correction that could be adapted across many UDN-identified variants.