Autoimmunity against POLR3A and cytopenias in aplastic anemia and other bone marrow failure syndromes

Abstract

Abstract Idiopathic aplastic anemia (IAA) is mediated by cytotoxic T-cells directed toward unrecognized antigen(s) on hematopoietic stem and progenitor cells. B-cell directed therapies have not been effective in IAA suggesting autoantibodies do not drive the pathogenicity of the autoimmune destruction of progenitor cells. Nevertheless, antigens recognized by pathogenic T-cells are likely also recognized by antibodies and could be used to study targets of cellular autoimmunity in IAA or have diagnostic applications. When we applied high density protein and peptide arrays to screen sera of IAA patients for antibody reactivity, in addition to expected infectious antigens, (EBV, HPV, HCV) we found several interesting reactivities including DAF, GANP among others exclusively present in >10% of patients and absent in controls. Among self-antigens, anti-RNA polymerase III subunit A (POLR3A) showed the highest reactivity with up to four 12-mer peptides with 100% homology to POLR3A recognized by sera of selected patients and were not related to transfusion burden and history of autoimmune condition such as systemic sclerosis. Western blot provided confirmation of the specificity in index cases.1 The fact that POLR3A is a self-antigen recognized by autoantibodies in ~15% of patients with systemic sclerosis and other autoimmune conditions motivated us to examine this discovery further. POLR3A is involved in innate immunity against viral infections and loss of function genetic alterations have been described in certain forms of immunodeficiency characterized by impaired IFN-γ induction and increased susceptibility to viral infections. Herein we hypothesized this reactivity may be a marker of a maladaptive response to viral infection in the pathogenesis of IAA. We continued to explore the biological significance of our findings through patients who were seropositive for POLR3A and searched for signs of BMF or immune cytopenia. Of 4293 patients screened for POLR3A antibody during routine clinical workup, 486 tested positive. Among associated diagnoses, MCTD/overlap syndrome (371), scleroderma (257), and pulmonary fibrosis (114) were most common. In seropositive patients, anemia was documented in 44% vs 38% (p=0.01), thrombocytopenia in 9% vs 8%, and neutropenia in 5% in both groups. We expanded our initial array based screening of BMF patients by clinical grade ELISA. We screened a total of 204 patients (165 AA and AA/PNH, 39 PNH), and identified 66 seropositive patients: 36 patients with AA without PNH, 22 had AA/PNH and 8 with hemolytic PNH likely related to AA. In seropositive AA/PNH patients; compared to seronegative patients with the same spectrum of diseases, neutropenia (71% vs 49%, 45/66 vs 68/138, p=0.005) and pancytopenia 64% vs 42%, 42/66 vs 58/138, p=0.005) were more pronounced, while there was a trend towards anemia and thrombocytopenia (92% vs 84%, 62/66 vs 116/138, p=0.07). PNH clones were detected significantly more frequently in seropositive patients compared to seronegative patients (36.8% vs 23.8%, p=0.03). Screening of other BMF syndromes revealed 13 seropositive cases in: LGL (3/105), MDS/AML (2/15), CMML (1/4), PRCA (1/6), chronic cytopenia without additional diagnosis (6/12). Seropositive non-AA BMF patients were more likely to be thrombocytopenic and neutropenic than the seronegative counterparts (p=0.004). When we assessed expression of POLR3A and other related RNA polymerases in AA, we found that 17/30 of POLR subunits were overexpressed in AA/PNH patients compared to healthy controls (POLR3A expression increased 17% compared to control, p<0.001. Expression in 27/36 patients above mean of controls), when normalized for myeloid vs lymphoid compartments. In conclusion, while the pathogenic mechanism behind POLR3A seroreactivity in AA remains to be discovered, this serological marker links AA to other autoimmune diseases with cryptic pathogenesis. The physiological role of POLR3A in innate immunity against viral infections suggests that seropositivity in the context of BMF could be established during an acute fulminant or chronic viral infection without complete viral clearance. We plan to confirm any association of viral or other autoantigens with POLR3A seroreactivity in AA patients with temporal relation to disease onset or response to therapy to unravel more details about the phenotype of patients who present with humoral autoimmunity in AA.