UBA1 dysfunction in VEXAS and cancer

Abstract

UBA1, an X-linked gene, encodes one of the only two ubiquitin E1 enzymes, playing a pivotal role in initiating one of the most essential post-translational modifications. In late 2020, partial loss-of-function mutations in UBA1 within hematopoietic stem and progenitor cells were found to be responsible for VEXAS Syndrome, a previously unidentified hematoinflammatory disorder predominantly affecting older males. The condition is characterized by severe inflammation, cytopenias, and an association to hematologic malignancies. In this research perspective, we comprehensively review the molecular significance of UBA1 loss of function as well as advancements in VEXAS research over the past four years for each of the VEXAS manifestations - inflammation, cytopenias, clonality, and possible oncogenicity. Special attention is given to contrasting the M41 and non-M41 mutations, aiming to elucidate their differential effects and to identify targetable mechanisms responsible for each of the symptoms. Finally, we explore the therapeutic landscape for VEXAS Syndrome, discussing the efficacy and potential of clone-targeting drugs based on the pathobiology of VEXAS. This includes azacitidine, currently approved for myelodysplastic neoplasms (MDS), novel UBA1 inhibitors being developed for a broad spectrum of cancers, Protein Kinase R-like Endoplasmic Reticulum Kinase (PERK) inhibitors, and auranofin, a long-established drug for rheumatoid arthritis. This perspective bridges basic research to clinical symptoms and therapeutics.

Keywords: MDS; VEXAS; clonal cytopenia; inflammation; ubiquitin.

Figure 1. Conceptual representation of the differential effect of UBA1 mutations based on the degree of loss of function of ubiquitin E1 enzyme UBA1.

(left panel) UBA1, an E1 enzyme, activates ubiquitin and subsequently transfers the activated ubiquitin to up to approximately 30 E2 enzymes with various efficiency. The displayed heatmap illustrates the variability in ubiquitin transfer efficiency (dark green: low efficiency, brown: high efficiency) of UBA1 wild type (first column), UBA1 partial loss of function (second column) and UBA1 total loss of function (third column). Wild type UBA1 and partial loss of function mutations affect the ubiquitin transfer efficiency of a subset of E2 enzymes, whereas a total loss of function of UBA1 leads to a complete loss of loading of ubiquitin to E2 enzymes solely dependent on UBA1. (right panel) At the E2/E3-substrate transfer step, the effect of UBA1 loss of function is mediated by the decrease of available ubiquitin-loaded E2 enzymes. In the case of partial loss of function mutations, ubiquitylation of substrates can be variable due to the differential impairment of ubiquitin transfer to the E2 enzymes, which may result in imbalance of regulator proteins and altered cell fate.

Figure 2. Mechanism of cytoplasmic-specific loss of function mutations and comparison with the non-M41 mutations.

(A) UBA1 mRNA transcript contains three alternative start codons at position M1 (UBA1a), M41 (UBA1b), and M67 (UBA1c). The transcript starting from M1 contains the nuclear localization signal (NLS) and the translated protein is transferred to the nucleus. In physiological conditions UBA1 mRNA is also translated from position M41, lacking the NLS, and the cytoplasmic isoform UBA1b is produced (top panel). Mutations at position M41 greatly reduces the translation efficiency starting at M41 and more transcripts are translated from M67. This results in the translation product, which is the catalytically deficient cytoplasmic isoform UBA1c (bottom panel). The isoform lacks residues from M41 to A65, compared to UBA1b. (B) The effect of M41 mutations result in intact UBA1a in the nucleus and isoform swap in the cytoplasm of the catalytically active UBA1b to the more inactive UBA1c. This results in stable ubiquitylation in the nucleus and substrate accumulation in the cytoplasm (top panel). The effect of non-M41 mutations is equally present in UBA1a and UBA1b, respectively, and substrate accumulation should similarly be seen both in the nucleus and cytoplasm (bottom panel).

Figure 3. Cellular, tissue-level, immune-environmental, and systemic effect of UBA1 mutations.

UBA1 mutations lead to substrate accumulation, which result in activation of the unfolded protein response (UPR, top left panel). This affects the cell fate of different cell types carrying the mutations in a context dependent way. In addition, UPR results in inflammatory response, including cytokine production. A list of aberrations due to mutations which may lead to altered immune microenvironment is given in the top right panel. The aberrations impair the immuno-competence of the patient. The panel at the bottom left illustrates the alterations in cell type composition within the bone marrow resulting from the cell-type-specific effects of UBA1 mutations in hematopoietic stem cells (HSCs) and progenitor cells. Mutations in UBA1 lead to distinct outcomes depending on the lineage of the mutated cells. Specifically, mutated cells of the lymphoid and likely also erythroid lineages progressively decrease as the cells differentiate, whereas the myeloid cells carrying the mutations undergo clonal expansion. The pie charts on the right side of the panel provide an approximate quantification of the mutant to wild type ratio per lineage observed. In peripheral blood of VEXAS patients cytopenias are observed either as a consequence of differentiation aberrations in the bone marrow, inflammatory environment, or due to cytotoxic anti-inflammatory treatment (bottom right). In addition, aberrant activation of immune cells is observed, which aggravates both systemic and tissue inflammation.

Figure 4. Genotype-phenotype associations of pathogenic UBA1 mutations and possibilities of therapeutic targeting.

Four different clinical phenotypes of pathogenic UBA1 mutations are known: X-linked spinal muscular atrophy (XL-SMA) is caused by germline mutations which have hotspot in exon 15. VEXAS is caused by somatic mutations in the cells of the bone marrow. M41 is the most frequent genotype but recurrent non-M41 mutations are also reported. Blood cancers are associated with both M41 and non-M41 mutations. Lung cancer is also reported with UBA1 mutations in females, which include frameshift and nonsense mutations. Molecularly, the degree of the enzymatic dysfunction and alteration of localization are different by genotypes. Therapies targeting this level: Auranofin tries to ameliorate phenotype by improving enzymatic dysfunction, whereas UBA1 inhibitors targets to tip the balance of survival to apoptosis by preferentially in cells with severe dysfunction. The cellular phenotypes of the mutations not only depend on the nature of the mutations but also on the affected cell type. Neuronal cells are particularly sensitive to protein aggregates and the mutations may be toxic with only slight enzymatic dysfunction. Other phenotypic mechanisms may be binding defects to proteins important in neural development. Concerning VEXAS, myeloid cells seem to be more resistant to UPR-mediated apoptosis due to activation of the PERK arm of the UPR. Therapies targeting of this level: PERK inhibitors try to prevent the preferential escape from apoptosis of the myeloid cells. Azacitidine likely also restore the cell type composition, but the exact mechanism is not known. The diverse clinical phenotypes are likely associated with the variety of mutations and their nature of loss of function and affected cell types. Abbreviations: wt: wild type; mut: mutated; LOF: loss of function; UPR: unfolded protein response; i: inhibitors.