AUTHORS
Bryan T Ciccarelli, MD, PhD, MA, Daniel Brown, PhD2, Nathan Welty, MD, PhD3, Nouraiz Ahmed, PhD4, Brandon William Simone, PhD5, Selina Luger, MD, FRCPC6, Saar Gill, MD, PhD7
AFFILIATIONS
1 University of Pennsylvania, Philadelphia, Pennsylvania, United States
2 University of Pennsylvania, Philadelphia, Pennsylvania, United States
3 University of Pennsylvania, Philadelphia, Pennsylvania, United States
4 Massachusetts Institute of Technology, Cambridge, Massachusetts, United States
5 University of Pennsylvania, Philadelphia, Pennsylvania, United States
6 University of Pennsylvania, Philadelphia, Pennsylvania, United States
7 University of Pennsylvania, Philadelphia, Pennsylvania, United States
ABSTRACT
Background: VEXAS syndrome is a severe, late-onset autoinflammatory disease caused by somatic UBA1 mutations in hematopoietic stem and progenitor cells (HSPCs), representing an extreme, single-gene model of CH. UBA1 encodes the primary E1 ubiquitin-activating enzyme in humans. Alternative translation initiation generates distinct nuclear (UBA1a) and cytoplasmic (UBA1b) isoforms. Most patients harbor alterations at UBA1 p.Met41 (M41), the translation initiation site for cytoplasmic UBA1b, resulting in decreased UBA1b, impaired ubiquitination, stress response signaling, and defective hematopoiesis. VEXAS affects ~1:4,000 men over age 50, causing profound morbidity and up to 50% 5-year mortality. Glucocorticoids and steroid-sparing agents show variable and often limited efficacy, and azacitidine can achieve durable hematologic and molecular responses in a subset of patients but is not curative. Allogeneic hematopoietic stem cell transplantation is the only potentially curative option but carries prohibitive treatment-related mortality in this older, comorbidity-laden population, and no approved disease-modifying therapy exists. Disease severity correlates with UBA1-mutant clonal burden, suggesting that even partial correction or depletion of mutant HSPCs may be therapeutic. We therefore developed complementary gene-editing approaches to model, correct, and selectively eliminate VEXAS clones.
Methods: Cell lines were cultured using standard practices. Male healthy donor HSPCs were cultured using standard practices with or without adjunctives to maintain quiescence and modulate proliferative state. Base editing (BE) and prime editing (PE) were delivered via transient, non-viral mRNA or protein to engineer
UBA1 mutations in cell lines and HSPCs. Editing was quantified by targeted amplicon sequencing. Engineered cells were evaluated for UBA1 isoform expression, ubiquitin homeostasis, and stress-response signaling, including the impact of complete UBA1 loss by Cas9-mediated knockout. Therapeutic approaches included correction of pathogenic alleles by BE and PE, BE-mediated suppression of UBA1a via start-codon disruption (Met1Thr), and allele-specific Cas9-mediated installation of nonsense mutations to selectively induce complete UBA1 loss in M41 mutation-bearing cells. Strategies were tested
in vitro and in an NSG xenotransplantation model.
Results: Optimized BE and PE editing enabled efficient installation of all 3 canonical UBA1 M41 mutations (M41Leu/Thr/Val) in cell lines and HSPCs, extending VEXAS models to encompass all major pathogenic variants in their endogenous context. High-efficiency correction of UBA1 M41 mutations was achieved in cell lines with BE and PE (up to ~80% relative correction); BE-mediated correction in HSPCs has reached ~25% to date, with PE optimization ongoing. Because the degree of clone reduction required for durable clinical benefit is unknown, we also investigated strategies to selectively eliminate VEXAS HSPCs. VEXAS phenotypes were proliferation-dependent: high-proliferation conditions attenuated molecular abnormalities via compensatory UBA1a upregulation, whereas low-proliferation conditions were required for faithful recapitulation. Complete UBA1 loss was lethal, whereas isolated UBA1a loss was well tolerated in wild-type HSPCs. These findings suggested a synthetic lethal dependency on residual UBA1a activity in UBA1b-deficient VEXAS HSPCs. Accordingly, dual editing to introduce UBA1a loss into UBA1 M41-mutant HSPCs resulted in rapid negative selection of the dual-edited population
in vitro. In xenotransplantation studies, isolated UBA1a disruption did not significantly impact weight, survival, engraftment, or persistence of edited cells. In contrast, cells harboring both UBA1a loss and UBA1 M41 mutations underwent near-complete eradication post-transplant, with variant allele frequency decreasing from ~90% to undetectable in 4/5 mice and 0.29% in the remaining animal (below most clinically reportable thresholds). To improve selectivity and engraftment, we developed allele-specific CRISPR approaches that selectively introduce UBA1-inactivating mutations in UBA1 M41-mutant cells, resulting in rapid negative selection
in vitro.
Conclusions: High-efficiency genome engineering enables faithful modeling and therapeutic targeting of VEXAS-associated CH. Direct correction of UBA1 mutations and mutation-selective synthetic lethality represent viable ex vivo gene therapy strategies for VEXAS syndrome.
DISCLOSURES
Bryan T Ciccarelli, MD, PhD, MA: No relevant conflicts of interest to declare.