Heterogeneous nuclear ribonucleoprotein (hnRNP) is composed of heterogeneous nuclear RNA and at least 30 associated proteins designated hnRNP-A1 (34 kd) through hnRNP-U (120 kd) according to their molecular weight and migration in two-dimensional gels.¹ The hnRNP proteins are abundant nuclear proteins involved in processing of mRNA precursors, regulation of alternative splicing and mRNA transport.¹ hnRNP-A2 is identical to the "RA-33" protein originally identified by studies of RA-33 autoantibodies, which were initially found exclusively in the sera of rheumatoid arthritis (RA) patients (especially in early disease prior to development of definitive clinical diagnostic criteria) and which occur independently of rheumatoid factor (RF).^1,2 RA-33 autoantibodies are reactive with hnRNP-A2, hnRNP-B1 and hnRNP-B2, which share >80% amino acid sequence identity and are probably the products of alternative splicing.¹

With the aim of investigating a potential pathogenic role of these Abs,  a recent report describes the Ab response to A2/B1 hnRNPs in different murine models of lupus.²  hnRNP autoantibodies occur primarily in RA, systemic lupus erythematosus (SLE) and mixed connective tissue disease (MCTD), and less frequently in other connective tissue diseases.^1,3 Although the roles of hnRNPA autoantibodies in the pathogenesis of RA, SLE and MCTD are unknown,¹ the presence of RA-33 autoantibodies is correlated with severe erosive arthritis in SLE patients.⁴ Because RA-33 autoantibodies can occur early in disease, they are clearly of value in the diagnosis of early RA, particularly RF-negative RA. EIA using partially purified native hnRNP antigens is more sensitive (50% versus 20% for RA) but less specific (85% versus 94% for RA) than immunoblotting for detection of hnRNP-A/B autoantibodies, especially in SLE and MCTD.¹ The specificity of immunoblotting is greater in hnRNP autoantibody-positive sera without U1-snRNP reactivity.¹ Approximately 50% of SLE patients with U1-snRNP autoantibodies also have hnRNP autoantibodies.⁵ hnRNP-A1 autoantibodies are found in ~25% of SLE and ~50% of RA;⁶ keratin antibodies, which are also present in ~50% of RA⁷ cross-react with an hnRNP-A1 epitope.^1,6 In SLE patients, hnRNP-A1 autoantibodies always occur with hnRNP-A2 autoantibodies.^1,5 Epitope mapping identified a major conformational epitope of hnRNP-A2 recognized by most RA sera, 60% of SLE sera and ~33% of MCTD sera.^8,9 Autoantibodies to hnRNP other than hnRNP-A/B in RA, SLE, MCTD and other connective tissue diseases are not yet identified.

A recent study demonstrated the cross-reactivity of anticentromere antibodies with hnRNP-A2/RA33 in the sera of  patients with longstanding rheumatoid arthritis (RA) and consecutive evolution of limited cutaneous systemic sclerosis.¹⁰  RF correlated with anti-RA33 antibodies in patients with Erosive arthritis (EA), but not with the presence of anti-RA33 alone. Thus, anti-RA33 antibodies may identify those patients with SLE who are at risk for EA and an association with RF suggests a common immune response or pathological mechanism in autoimmune erosive joint disease.¹¹