The Mechanism of Cell Autonomous Inflammation in VEXAS Syndrome Is Mediated By Proteotoxic Stress

VEXAS (Vacuoles, E1 enzyme, X-linked, Autoinflammatory, Somatic) syndrome is a severe, multi-system disease caused by mutations in the UBA1 gene acquired in hematopoietic stem cells (HSCs) affecting 1/4000 men over the age of 50 (Beck DB, NEJM 2020; Beck DB, JAMA 2023). UBA1 is the E1 enzyme responsible for initiating the majority of ubiquitylation and has two isoforms: a nuclear isoform, UBA1a, and a cytoplasmic isoform, UBA1b. UBA1 mutations in VEXAS occur predominantly at the Met41 hotspot, lead to loss of the cytoplasmic isoform UBA1b, and are restricted to the myeloid lineage. We have previously shown that UBA1 mutations generate a bottleneck in transferring ubiquitin to cytoplasmic E2 enzymes and UBA1b isoform levels correlate with patient outcome (Collins J, EMBO 2024; Ferrada M, Blood 2022). However, how a reduction in cytoplasmic E2 charging and ubiquitylation causes inflammation is poorly understood. Here, we identify the cell autonomous mechanisms that initiate and drive inflammation in VEXAS.

First, we investigated the impact of defective cytoplasmic ubiquitylation on protein homeostasis by analyzing total proteomes of CD14+ monocytes isolated from healthy donors or VEXAS patients. Pathway analysis revealed upregulation of not only inflammation but also substrates associated with protein processing in the endoplasmic reticulum and the unfolded protein response (UPR). To validate these findings, we constructed a THP1, human monocyte, model of VEXAS utilizing a conditional Cas9-mediated knock-out of endogenous UBA1 in the background of constitutive lentivirally incorporated wild-type (UBA1^WT) or VEXAS disease associated UBA1 (UBA1^M41V). When induced, the UBA1^M41V line generated vacuoles, showed poly-ubiquitin defects, and produced an inflammatory response. We then performed unbiased proteomics and again observed upregulation of conserved pathways in ER protein processing, UPR, and inflammation. Immunoblot and transcriptomic analyses of both patient CD14+ monocytes and the cell model confirmed upregulated markers of UPR and inflammation including BiP, p-eIF2α, XBP-1s, ATF4, and p-STAT1.

To examine whether UPR activation is driven by loss of ubiquitin conjugation to a specific E2, we performed an unbiased molecular screen with the Chinese Hamster Ovary (CHO) line ts20. At restrictive temperatures endogenous UBA1 is lost. By complementing the system with UBA1^WT or UBA1^M41V we were able to screen all 28 E2 enzymes assayable in CHO cells. While most E2 enzymes showed a mild loss of E2~Ub thioester species when cytoplasmic UBA1b is lost due to UBA1^M41V, only two, UBE2J1 and UBE2G2, showed substantial loss of ubiquitin conjugation. These same E2s were found to be disproportionately undercharged in both our THP1 VEXAS model and patient cells when assayed by immunoblotting. This demonstrates the E2 defects to be conserved and specific to Met41 mutations across both model cell lines and patient monocytes.

UBE2J1, UBE2J2, and UBE2G2 are cognate members of the Endoplasmic Reticulum Associated Degradation (ERAD) pathway. Importantly, UBE2J1 and UBE2J2 are the only cytoplasmic restricted E2s in the entire proteome. Thus, we hypothesized that loss of cytoplasmic ubiquitylation might specifically inhibit ERAD. To examine this, we used an ERAD substrate reporter, CD3δ-GFP, and were able to show that in the THP-1 UBA1^M41V model, ERAD substrate processing is defective. Furthermore, selective chemical inhibition of a cognate E3 of ERAD, HRD1, as well as siRNA-mediated depletion of the ERAD E2s phenocopied ERAD defects as well UPR activation and inflammation. Finally, we posited that ERAD failure drives chronic UPR activation and inappropriate inflammation. Indeed, when we utilized UPR-specific small molecule inhibitors targeting ATF6, IRE1α, and PERK, in the settings of both the THP-1 UBA1^M41V model and CD14+ patient cells, we were able to dramatically reduce the inflammatory phenotype.

In summary, we identified that VEXAS mutations in the Met41 hotspot in UBA1 lead to an undercharging of the cytoplasmic-resident E2 enzymes of ERAD, subsequent ERAD failure, the accumulation of unfolded proteins, and chronic activation of the UPR leading to inflammation capable of being modulated by UPR-targeting small molecule drugs. These results identify the mechanisms initiating and propagating inflammation in VEXAS syndrome and highlight multiple avenues for therapeutic intervention.

Beck:GSK: Consultancy; Alexion: Consultancy; Sobi: Consultancy; Novartis: Consultancy.