INTRODUCTION

Anti-Sm was one of the first autoantibodies to non-histone proteins described in systemic autoimmune rheumatic diseases.¹ Antibodies to the Sm antigen are highly diagnostic of systemic lupus erythematosus (SLE) and are present in 10 to 30% of unselected SLE populations.^2,3 The identification of Sm antigen as well-defined proteins bound to U-rich small nuclear RNAs (snRNAs) has been considered a significant advance in biology, and these autoantibodies have served as useful probes to help investigate the important components of the spliceosome [which is the multiprotein complex responsible for pre-mRNA splicing of heterogeneous nuclear RNA to mature messenger RNA (mRNA)].¹ In this chapter we discuss the major classes of anti-spliceosome antibodies [including anti-Sm, anti-U1 RNP (also known as anti-nRNP in older terminology), and anti-U2 RNP] and briefly review other minor autoantibodies to other classes of U RNPs, such as LSm (Like Sm) proteins.

SPLICEOSOMAL AUTOANTIGENS

Small nuclear ribonucleoproteins (snRNPs) are the major autoantigens in the spliceosome. They are classified by association with specific U-rich snRNAs, including the most abundant U1, U2, U4, U5, and U6 RNAs (Fig. 22G.1). These snRNPs associate with pre-mRNA in a sequential manner to assemble the spliceosome into a functional complex, which can catalyze the splicing reaction. Fig. 22G.1A illustrates the composition of the major snRNPs: U1, U2, U4/U6, and U5 snRNP sharing the Sm core proteins B or B′ (27/28kDa), D1/D2/D3 (14 kDa), E (12 kDa), F (11 kDa), and G (9 kDa), which are organized as seven-member ring structures (Sm ring). Analysis of individual snRNPs has disclosed that, in addition to the shared peptides, some proteins are specifically associated with certain snRNPs. U1 snRNPs contain three distinct proteins designated 70k (68/70 kDa), A (32 kDa), and C (22 kDa), shown in blue in Fig. 22G.1A. U2 snRNPs have two unique proteins: A’ (31 kDa) and B’’ (29 kDa, shown in green). U4 is always associated with U6 and is thus often designed U4/U6 snRNP. There are several other proteins associated with U5 or U4/U6 snRNPs³ but only the 200 kDa doublet of U5snRNPs is shown in Fig. 22G.1A because this can be used to differentiate sera with anti-U1 RNP alone versus anti-U1 RNP plus anti-Sm (Fig. 22G.1C).^5,6 U6 snRNP uses one of the LSm rings, which is structurally similar to the Sm ring, as its core complex (Fig. 22G.1D). Other snRNPs (such as U3, U8 and U13) are primarily localized to the nucleolus and are not part of the spliceosomal complex. U11 and U12 snRNPs occur in relatively lower abundance and are required for U2-snRNP-independent splicing. Other components of the spliceosome include SR proteins, which are serine-arginine–rich proteins involved in the mRNA splicing reaction.

Fig. 22G.1 Spliceosomal components. (A) Composition of U1, U2, U4, and U5 snRNPs illustrated as protein bands immunoprecipitated using anti-Sm or anti-U1 snRNP (nRNP) autoantibodies. The four snRNPs share the Sm core complex shown in orange. U1-snRNP-specific proteins are shown in blue, U2-snRNP-specific in green, and U5-snRNP–specific in black. Anti–Sm-positive sera immunoprecipitate all four snRNPs and associated proteins, whereas anti-nRNP–specific sera will only bring down U1 RNP. (B) snRNAs immunoprecipitated from a cell lysates using human anti-Sm or anti-nRNP, analyzed in a urea polyacrylamide gel, and detected by silver staining of RNA bands. (C) snRNP-associated proteins immunoprecipitated from lysates of cells metabolically labeled with [¹S]-methionine using human anti-Sm or anti-nRNP, fractionated in SDS-PAGE, and detected by autoradiography. (D) Seven-member Sm and LSm rings are illustrated showing structural similarity among these doughnut-shaped structures, with the center of the doughnut involved in binding of single-stranded RNA.

ANTI-SM AUTOANTIBODIES

Anti-Sm antibodies recognize the Sm ring core protein complex (Fig. 22G.1D) associated with the snRNA U1, U2, U4/U6, and U5 (Fig. 22G.1A). The Sm core proteins are the B/B′/N, D1, D2, D3, E, F, and G proteins described previously. Sm B and B′ are products of alternative splicing from a single gene, whereas Sm N shares 93% amino acid homology with B′ and is derived from a different gene with its expression restricted to certain cell types and stages of development. Earlier work of the crystal structures of two Sm protein complexes, D3/B and D1/D2, suggested that the seven Sm core proteins could form a closed ring and the snRNAs may be bound in the positively charged central hole. Because these core proteins are common to U1, U2, U4/U6, and U5 snRNPs, a wider range of snRNAs are immunoprecipitated by anti-Sm antibodies than are precipitated by anti-U1 RNP antibodies (Fig. 22G.1B).

Anti-Sm produces a nuclear speckled staining pattern on HEp-2 cell nuclei by conventional indirect immunofluorescence (IIF) (Fig. 22G.2). However, this staining pattern is practically indistinguishable from that of anti-U1 RNP by this technique. Double immunodiffusion (DID), Western blotting (WB), line immunoassay (LIA), analysis of precipitated RNAs (Fig. 22G.1B), enzyme-linked immunoassays (ELISA), or addressable laser bead immunoassays are preferred to definitively identify this autoantibody. Advances in cloning and epitope mapping of Sm-related peptides has led to the use of synthetic peptides in clinical studies. However, the clinical usefulness of these anti-peptide antibodies is unclear. Because most lupus sera bind to the glycine-arginine repeat of the Sm peptide regardless of the presence of anti-Sm antibodies under conventional methods such as DID or Western blot (reactivity with the D protein),^7,8 the significance of reactivity with the Sm peptide needs to be carefully evaluated. The techniques required to demonstrate antibody to Sm necessitate the use of standard reference sera that are available through the Centers for Disease Control in Atlanta, Georgia.⁹

Fig. 22G.2 Immunofluorescence of human anti-Sm using the CDC prototype serum on HEp-2 cells. Speckles are restricted primarily to nuclei and sparing nucleoli. snRNP complexes are concentrated in nuclear bodies known as Cajal bodies (arrows), which are often seen adjacent to nucleoli. M = mitotic cell. Anti-U1 RNP staining is practically identical.

Antibodies to both Sm and U1 RNP, which commonly coexist in the same serum,² are primarily detected in SLE patients. In SLE it is common to detect antibodies to U1 RNP alone or antibodies to both U1 RNP and Sm, but antibodies to Sm alone are rarely found.^2,6 Over the years, it has been repeatedly confirmed that this antibody is present almost exclusively in SLE and is considered a diagnostic marker for that disease to the extent it is included in the American College of Rheumatology (ACR) criteria for the classification of the SLE.¹⁰ Although anti-Sm is present in only 10 to 30% of patients with SLE,² it is a highly specific disease marker. It has only rarely been detected in normal sera or in patients with other systemic rheumatic diseases such as Sjögren’s syndrome (SjS), scleroderma, polymyositis, or drug-induced lupus. However, it is important to note that lupus-related autoantibodies (including anti-Sm) have been observed prior to the development of full clinical features or diagnosis.¹¹ Therefore, anti-Sm can be detected occasionally in a patient who does not have SLE but may develop SLE in the future. Some reports suggested that the titers of anti-Sm antibodies correlate with the disease activity.^3,12 Anti-Sm–positive SLE were associated with milder renal and central nervous involvement or late-onset renal disease¹³ in some reports, but this is controversial.³

The reactivity of anti-Sm antibodies has been refined and further based on the understanding that the B/B′ and D polypeptides are considered the major target autoantigens (although anti-U1 RNP antibodies often react with B/B′ in Western blot testing.^5,13,14 It is known that SmB/B′ share cross-reactive epitopes (PPPGMRPP) with U1-specific RNPs, which are more frequently targeted by antibodies present in patients with mixed connective tissue disease (MCTD). Thus, the SmD polypeptides are regarded as the more disease-specific Sm antigens. Recently, it was shown that the polypeptides D1 and D3 contain the modified amino acid symmetrical dimethylarginine within a major autoepitope of the SmD1 and SmD3 C-terminus.^15,16 A synthetic peptide of SmD3 containing symmetrical dimethylarginine at position 112 represents a promising tool for the detection of a highly specific subpopulation of anti-Sm antibodies.¹⁷

ANTI-U1 RNP (NUCLEAR RNP, NRNP) AUTOANTIBODIES

Autoantibodies to U1 RNP were first defined using DID as anti-Mo that recognize soluble nuclear ribonucleoprotein (nRNP).¹⁸ In the same year, Sharp and colleagues described a group of patients associated with high levels of antibodies to saline extractable nuclear antigen (ENA) detected by passive hemagglutination testing.¹ This group of patients was described as MCTD, which is characterized by overlapping symptoms and signs of SLE, scleroderma, and polymyositis.^19,20 The main clinical features of MCTD are a high prevalence of Raynaud’s phenomenon, edema of the fingers, arthritis/arthralgias, myositis, serositis, and a relative absence of renal disease. The characteristic serologic feature of MCTD is the presence of high titers of RNase-sensitive anti-ENA antibodies, which were later confirmed as anti-U1 RNP. This was distinguished from RNase-resistant anti-ENA antibodies, which is anti-Sm antibodies, or anti-dsDNA antibodies.^19–21

When anti-U1 RNP antibodies are present alone and in high titer, they are often associated with MCTD. Anti-U1 RNP antibodies are also detected, usually at a lower titer, in other systemic rheumatic diseases (including 30 to 40% of SLE) and in much lower frequency in rheumatoid arthritis, systemic sclerosis, Sjögren’s syndrome, polymyositis, discoid lupus, and (rarely) in drug-induced lupus.^3,22 Because of the clinical relevance of the titers observed in systemic rheumatic diseases, it became necessary not only to detect but quantify the level of anti-U1 RNP antibodies in patient’s sera.

It is important to appreciate that some SLE and MCTD sera bind most of the snRNP polypeptides whereas others bind to little, if any, 70k or C polypeptides. In one study, only 8% of SLE sera containing U1 RNP antibodies bound to the 70k antigen, but 76% of MCTD sera bound this protein.²³ The observed higher frequency of anti-70k antibodies in MCTD has been supported in several studies and has been reported to be as high as 95% in MCTD, whereas the range of reactivity in SLE is 20 to 50%. The frequency of anti-U1 RNP is highly dependent on the assay employed, as evidenced by the fact that when recombinant 70k protein was used in an ELISA, up to 85% of SLE patients showed elevated antibody levels.

It has been suggested that the presence of anti-70k is correlated with the presence of Raynaud’s phenomenon, esophageal dysmotility and myositis, and a negative indicator for the presence of renal disease.^23–25 Taken together, the data suggest that antibodies to the 70k protein may be more characteristic of patients with classic features of MCTD. These may be important observations for the clinician who attempts to identify patients at high risk of developing end-organ disease when they present with only a few features (e.g., Raynaud’s phenomenon or myositis) of other systemic rheumatic diseases.

Although antibodies to the 70k protein quantitatively vary during the disease course, there is little evidence that they correlate with disease activity or that they are involved in disease pathogenesis.^24,25 There has been interest in the possibility that anti-U1 RNP antibodies have a unique property of being able to “penetrate” living cells.²⁶ Although the pathogenic significance of such observations are not clear, this property may account for other occasional observations that anti-U1 RNP antibodies appear in nuclei of patients’ tissue specimens. However, whether this is in fact an in vivo phenomenon or an in vitro artifact is still controversial.

Antibodies to the A and C proteins are found in approximately 25% of unselected SLE cohorts and in 75% of SLE patients preselected for antibodies to U1 RNP.^24,27,28 Like antibodies to the 70k protein, the titers may vary during the course of the disease but there is no clear evidence that they predict disease activity or that they are involved in the pathogenesis.²⁴

U1 RNA has been described as acting as an endogenous adjuvant for the development of a pro-inflammatory state via activation of toll-like receptor 3 (TLR3) signaling.²⁹ More recent studies have suggested that U1 RNA may induce innate immunity through other cellular RNA sensors, including TLRs 7 and 8 and RNA-dependent protein kinase.^30–32 This may explain why RNA-protein complexes are frequent targets of lupus autoantibodies. Anti-U1 RNP antibodies are thought to preferentially recognize the protein components of the U1 RNP. However, several reports showed that coexistence of autoantibodies to U1 RNA are quite common.^33–36

van Venrooij and colleagues reported that 38% of anti-snRNP sera were positive for antibodies to U1 RNA and that they were always accompanied by anti-U1 RNP but not by anti-Sm.³⁶ Hoet and colleagues³⁵ reported that 9 patients with SLE overlap syndrome demonstrated changes in titers of anti-U1 RNA antibodies that were correlated with disease activity. More recently, Asano and colleagues³⁴ suggested that anti-U1 RNA antibodies are a serologic indicator for pulmonary fibrosis in systemic sclerosis patients with anti-U1 RNP antibodies. As discussed by Greidinger and Hoffman,³⁰ assays for anti-U1 RNA are poorly standardized and of limited availability to clinicians in most areas, making this observation primarily of research relevance.