To characterize a new serum autoantibody in patients with systemic sclerosis (SSc) directed against U11/U12 RNP and to identify the clinical features associated with this autoantibody.
To characterize a new serum autoantibody in patients with systemic sclerosis (SSc) directed against U11/U12 RNP and to identify the clinical features associated with this autoantibody.
We identified autoantibodies directed against the U11/U12 RNP complex in sera of patients with SSc and confirmed antibody specificity by immunoprecipitation, reverse transcriptase–polymerase chain reaction, and Southern blotting. We determined the prevalence of these antibodies in SSc and their specificity for SSc. We compared anti–U11/U12 RNP autoantibody–positive and negative SSc patients on demographic, disease classification, clinical variables, and survival.
We identified 33 patients with anti–U11/U12 RNP antibodies. In 2 consecutive series of SSc patients first seen at 10-year intervals (1994–1995 and 2004–2005), the prevalence of anti–U11/U12 RNP antibody–positive patients was 15 of 462 (3.2%). Seventeen (52%) of these 33 patients had limited cutaneous involvement. All patients had Raynaud's phenomenon and 82% had gastrointestinal (GI) involvement. None had “intrinsic” pulmonary arterial hypertension. The most significant clinical difference between anti–U11/U12 antibody–positive and negative cohorts was the prevalence of lung fibrosis, which occurred in 79% of the anti–U11/U12 RNP antibody–positive patients versus 37% of the anti–U11/U12 RNP antibody–negative patients (P < 0.0001). GI involvement was also significantly increased in the anti–U11/U12 RNP antibody–positive group. Patients with anti–U11/U12 RNP antibodies and pulmonary fibrosis had a 2.25-fold greater risk of death than anti–U11/U12 RNP negative patients with pulmonary fibrosis.
Anti–U11/U12 RNP antibodies are present in the sera of approximately 3% of patients with SSc and are a marker for lung fibrosis, which is often severe.
Systemic sclerosis (SSc; scleroderma) is a connective tissue disease of unknown etiology that commonly involves the skin and internal organs. A number of serum autoantibodies have been identified in SSc patients and they serve as biomarkers of clinical features. For example, anti–topoisomerase I (anti–topo I; anti–Scl 70) (1) and anti-Th/To (2) autoantibodies both are associated with an increased risk of interstitial lung disease. Anti–RNA polymerase III (anti–RNAP III) autoantibodies are associated with scleroderma renal crisis and are infrequently present in patients with significant lung disease (3, 4). Anticentromere (ACA), anti–U3 RNP, and anti-Th/To antibodies are more frequently detected in sera of patients with “intrinsic” pulmonary arterial hypertension (PAH) (2). Anti–U3 RNP antibody also is associated with scleroderma heart disease (5).
Serum autoantibodies to small nuclear RNP (snRNP) have been found in patients with SSc and other connective tissue diseases. Most of these antibodies are directed against the protein component of the complex. Some antibodies recognize individual RNP such as anti–U1 RNP or anti–U3 RNP, while others are directed against a complex of RNP, such as anti-Sm autoantibodies, which target the uridine (U) rich complexes of U1, U2, U5, and U4/U6 RNP (6). Of the anti–RNP antibodies, anti–U1 and anti–U3 are the most frequent in SSc patients, while anti–U5 and anti–U4/U6 are rare. Anti–U4/U6 autoantibodies were initially reported in the serum of a patient with SSc (7) and subsequently in a Japanese patient with primary Sjögren's syndrome (8). U4 and U6 RNA have been shown to coexist in a single small RNP particle (9), which explains their co-immunoprecipitation with antisera from patients with SSc (7). Anti–U5 RNP antibodies were identified in the serum of one Pittsburgh patient with SSc and polymyositis in overlap (10) and later in a Japanese patient with a similar overlap syndrome and large cell carcinoma of the lung (11).
U11/U12 RNP are found in low abundance in eukaryotic cells, are components of the spliceosome, and catalyze pre–messenger RNA (pre-mRNA) splicing of nuclear pre-mRNA introns (12). Gilliam and Steitz previously reported the presence of anti–U11/U12 RNP antibodies in one patient with diffuse cutaneous SSc (13), but this antibody may not have been specific to U11/U12 RNP since it also recognized the 5′ 2,2,7-trimethylguanosine (TMG) cap of small nuclear RNA (snRNA). Except for U6 RNA, all other U series RNA have a unique 5′ TMG cap that targets them to the nucleus (13, 14). Antibodies to the TMG cap have also been reported in patients with SSc (15). However, clinical features associated with anti–U11/U12 RNP antibodies have not been examined to date.
Our aim was to identify and characterize anti–U11/U12 RNP autoantibodies in 33 patients with SSc and describe their clinical features and disease course in comparison with SSc patients without these antibodies.
Serum samples were obtained with informed consent from patients seen by physicians in the Division of Rheumatology and Clinical Immunology at the University of Pittsburgh School of Medicine, and were stored at –80°C. All patients had received a physician-confirmed diagnosis of SSc between 1982 and 2005. To determine the prevalence of anti–U11/U12 RNP antibodies, consecutive patients first evaluated during 1994–1995 and 2004–2005 (a total of 4 calendar years) with serum samples available were tested for U11/U12 RNP antibodies. For demographic, clinical features, and survival comparisons, the 1982–2004 cohort of anti–U11/U12 RNP–positive patients was compared with the 1994–1995 patients who had no detectable U11/U12 RNP antibodies. This earlier cohort was chosen because followup was available for a longer period of time (a mean of 5.3 years after the first visit). The rationale for using 4 years of consecutive patients as a comparison group is that these patients had all 8 other SSc-associated serum autoantibodies determined.
Clinical and laboratory information obtained on first and followup visits on all SSc patients was prospectively collected using standardized data collection forms. The definitions for organ system involvement attributable to SSc used in this study have been published previously (2). Organ system involvement was considered present if it occurred at any time during the illness, including the followup period after first evaluation. Diffuse cutaneous involvement was defined as skin thickening proximal to the elbows or knees (upper arms, thighs, chest, or abdomen) at any time during the illness (16). Pulmonary fibrosis was defined as interstitial fibrosis or ground glass changes on chest radiograph or high-resolution computed tomography (CT) scan of the lungs, respectively. PAH was defined as “intrinsic” PAH, not secondary to lung fibrosis or heart disease, and required a mean PA pressure >25 mm Hg (right heart catheterization) or estimated PA systolic pressure >40 mm Hg detected on echocardiogram (17). Gastrointestinal (GI) involvement was defined as one of the following: distal esophageal hypomotility by cine esophagram or manometry, esophageal stricture by endoscopy, radiographic evidence of wide-mouth colon sacculation, radiographic evidence of duodenal or small intestinal dysmotility or hypomotility, malabsorption syndrome, administration of antibiotics for small bowel bacterial overgrowth, or GI scleroderma as the cause of death.
K562 and HeLa cells were cultured in RPMI 1640 supplemented with 10% fetal bovine serum, penicillin, and streptomycin (Invitrogen, Carlsbad, CA). Mouse monoclonal anti-5′ 2,2,7-TMG cap antibodies were purchased from Oncogene Science (Manhasset, NY).
Protein immunoprecipitation was done as previously described (10, 18). Briefly, K562 or HeLa cells were harvested, washed with 1× phosphate buffered saline, and resuspended in lysis buffer (500 mM NaCl, 10 mM Tris-HCl, 0.1% Nonidet P40, pH 8.0). Cells were sonicated on ice and cellular debris pelleted following centrifugation at 12,000g for 20 minutes. The resulting supernatants were used as the antigen source. Serum (10 μl) was incubated with 50 μl protein A– agarose beads (Invitrogen) for 16 hours at 4°C. The antibody-bound beads were washed 3 times in lysis buffer and incubated with cellular extracts for 2 hours at 4°C. Agarose-bound complexes were washed 3 times, resuspended in Laemmli buffer, and fractionated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE.)
For immunoprecipitation of RNP complexes, unlabelled K562 or HeLa cells were harvested and sonicated in NET-2 buffer (140 mM NaCl, 50 mM Tris-HCl, 0.05% Nonidet P40, pH 7.4) as previously described (15). Cellular debris was removed following centrifugation and cellular lysates incubated with antibody-bound protein A–agarose beads as described above. RNP complexes were extracted with phenol–chloroform–isoamyl alcohol (50:50:1), precipitated with ethanol and separated by electrophoresis on 7M urea polyacrylamide gels. The snRNA were visualized following silver staining.
Human serum immunoglobulins were crosslinked to protein A–agarose beads as previously described (18). Agarose-bound immunoglobulins were incubated with sonicated K562 or HeLa cell extracts for 2 hours at 4°C. The complex-bound beads were washed and samples were resolved by SDS-PAGE. Membranes were incubated with control or patient sera at a dilution of 1:250 followed by a secondary horseradish peroxidase-conjugated anti-human IgG antibody (Pierce Biotechnology, Rockford, IL). Signals were detected following chemiluminescence and autoradiography.
For reverse transcription of U11 and U12 snRNA, K562 proteins were immunoprecipitated with patient sera as described above. The RNA was reverse transcribed using Superscript II (Invitrogen). The primers were as follows: for U11, 5′-AAGGGCGCCGGGACCAA-3′; and for U12 snRNA 5′-GGCAGATCGCAACTCCCAGG-3′. Complementary DNA (cDNA) were amplified using the primers used for reverse transcription in combination with 5′-GGCTTCTGTCGTGAGTGG-3′ for U11 or 5′-TAACGATTCGGGGTGACGCC-3′ for U12. Amplification conditions were 20 cycles of 95°C for 1 minute, 68°C for 1 minute, and 72°C for 1 minute. DNA products were separated by electrophoresis on 12% nondenaturing acrylamide gels. The expected cDNA products were 127 bp and 116 bp for U11 and U12, respectively. The DNA products were cloned into the pGEM-T vector (Promega, Madison, WI) and sequences of cloned cDNA were confirmed using an ABI Prism 377 automated sequencer (Perkin-Elmer, Norwalk, CT). These cDNA were used as probes in the Southern blot analysis described below.
The cDNA, obtained by RT-PCR as described above, was resolved on a 1.5% agarose gel. The DNA was denatured and transferred to a supported nitrocellulose membrane. The membrane was hybridized to 32P-labeled U11 or U12 cDNA and the signals were visualized following autoradiography.
Baseline characteristics were compared by Student's t-test for continuous data, and chi-square, or Fisher's exact test where appropriate, for discrete data. All statistical analyses were 2-tailed and the results were considered significant if P values were less than 0.05. Survival was evaluated with the Kaplan-Meier method. Mortality risks in patients with pulmonary fibrosis were estimated using Cox proportional hazards modeling techniques.
Immunoprecipitation of RNP complexes in the course of an ongoing analysis of autoantibodies in SSc revealed the presence of an snRNA of comparable electrophoretic mobility to U11 snRNA in one patient (Figure 1A).
It has been previously shown that U11 and U12 snRNA have a TMG cap, similar to other TMG-capped snRNA. To confirm the presence of the TMG cap and identify proteins complexed with TMG-capped snRNA, cellular proteins were immunoprecipitated using resin-crosslinked monoclonal anti-TMG antibodies. Proteins were eluted from the resin, separated on SDS-PAGE, and analyzed by Western blot analysis using normal serum, anti-Sm, anti–U1 RNP, anti–U11/U12 RNP, or anti-TMG–positive sera. Anti–U11/U12 RNP autoantibodies recognized a 65–68-kD protein band precipitated by anti-TMG antibody (Figure 1B). A protein of similar size had been previously shown to be associated with U11 snRNA in eukaryotic cells (12, 13).
Sera from a healthy donor, patients with anti–U11/U12 RNP, anti-Sm, anti-TMG, and anti–U3 RNP antibodies, and a purified monoclonal anti-TMG antibody were used to immunoprecipitate their corresponding antigens. Immunoprecipitated snRNA were reverse transcribed using a U11-specific 3′ primer or a U12-specific 3′ primer followed by amplification using U11 or U12 forward and reverse primers, respectively. Amplified cDNA were resolved on a 12% acrylamide gel. Figure 2A shows a cDNA corresponding to the expected size of U11 (127 bp) or U12 (116 bp) amplified from RNP precipitated with sera positive for anti–U11/U12 RNP, anti-Sm, or anti-TMG antibodies and with monoclonal anti-TMG antibody but not sera from a healthy ANA–negative donor and a patient positive for anti–U3 RNP antibodies.
To further confirm the identity of the bands, amplified cDNA were cloned and sequenced. The nucleotide sequences of the cDNA completely matched the U11 and U12 cDNA sequences in GenBank. Sequenced cDNA were further used as radiolabeled probes in a Southern blot assay to confirm the presence of U11 and U12 cDNA in the RT-PCR reactions. Southern blot analysis using a U11 or U12 radiolabeled cDNA probe confirmed the presence of U11 and U12 cDNA in samples precipitated by anti–U11/U12 RNP, anti-Sm, and anti-TMG, and amplified by PCR, whereas no U11 or U12 cDNA was detected in samples precipitated with healthy donor serum or anti–U3 RNP–positive serum (Figure 2B).
Comparison of sera from other SSc patients to the index serum revealed that 32 additional SSc patients were positive for anti–U11/U12 RNP autoantibodies. Figure 3 shows immunoprecipitation results from representative patients whose sera were positive for anti–U11/U12 RNP antibodies. U11/U12 snRNA precipitated by SSc patient sera were compared with those precipitated by standard sera corresponding to U1, U2, U3, U4/U6, and U5 RNP, Sm, and TMG. Antibody reactivity to U11/U12 RNP complexes was also detected in serial serum samples from several patients (patients 1–3).
The prevalence of U11/U12 RNP antibodies was calculated in 2 consecutive 2-year periods. During 1994–1995, 4 of 244 consecutive new SSc patients (1.6%) with serum available for testing had anti–U11/U12 RNP antibodies and no other SSc-associated serum autoantibody. During 2004–2005, 11 of 218 consecutive new SSc patients (5.0%) had anti–U11/U12 RNP antibodies. Therefore, the overall prevalence of anti–U11/U12 RNP antibodies in these 2 time periods combined was 3.2%. The presence of anti–U11/U12 antibodies was not associated with the presence of any of the other SSc-associated antibodies. Although low levels of anti–U1 and U2 RNP antibodies were detected in some anti–U11/U12 RNP–positive sera by immunoprecipitation (Figure 3), less sensitive assays such as double immunodiffusion failed to detect the presence of these antibodies. Anti–U11/U12 RNP antibodies were specific to SSc patients and were not detected in sera of 272 patients with polymyositis/dermatomyositis, rheumatoid arthritis, systemic lupus erythematosus, or connective tissue diseases in overlap. Anti–U11/U12 RNP antibodies were also not detected in 24 patients with idiopathic pulmonary fibrosis.
The 15 anti–U11/U12 RNP–positive patients in the above 4 years were combined with 18 anti–U11/U12 RNP–positive patients identified from other years during 1982–2004. The demographic and clinical characteristics of the 33 anti–U11/U12 RNP–antibody positive SSc patients compared with the 240 consecutive U11/U12 RNP–negative SSc patients first evaluated in 1994–1995 are shown in Table 1. There were no differences in age or sex between the groups. A somewhat lower proportion of anti–U11/U12 RNP–positive patients were white (79% versus 90%, data not significant). Twenty-eight (85%) of the anti–U11/U12 RNP–antibody positive patients and 206 (86%) of the anti–U11/U12 RNP–antibody negative patients fulfilled the American College of Rheumatology (formerly the American Rheumatism Association) preliminary criteria for classification (16) as definite SSc. Half of the anti–U11 RNP–positive patients had limited cutaneous involvement (52%). The mean maximum total skin thickness score using the modified Rodnan skin thickness score (19) in anti–U11/U12 RNP patients was 4.6 for patients with limited cutaneous disease, and 30.2 for patients with diffuse cutaneous SSc. The comparison group had similar mean maximum skin thickness scores.
|Anti–U11/U12 RNP–antibody positive (n = 33)||Anti–U11/U12 RNP–antibody negative (n = 240)||P|
|Age at SSc onset, mean ± SD years||41.1 ± 16.9||43.2 ± 15.8||0.4760|
|White||26 (79)||215 (90)||0.1288|
|Female||24 (73)||186 (78)||0.6967|
|dcSSc||16 (48)||121 (50)||0.9821|
|lcSSc||17 (52)||117 (49)||0.9106|
|Organ system involvement|
|Peripheral vascular||33 (100)||234 (98)||0.7754|
|Skin, mean ± SD maximum TSS||17.0 ± 15.2||15.5 ± 13.7||0.5660|
|dcSSc||30.2 ± 10.7||26.0 ± 12.0||0.1830|
|lcSSc||4.6 ± 3.9||4.9 ± 3.1||0.6940|
|Joints/tendons||26 (79)||177 (74)||0.6827|
|Skeletal muscle||2 (6)||37 (15)||0.2401|
|Gastrointestinal tract||26/26 (100)||120/161 (75)||0.0110|
|Pulmonary fibrosis||23/29 (79)||74/202 (37)||0.0001|
|Pulmonary hypertension||0||27 (11)||0.0946|
|Heart||6/25 (24)||48/175 (27)||0.7180|
|Kidney (renal crisis)||2 (6)||27 (11)||0.5446|
All 33 anti–U11/U12 RNP–positive patients had Raynaud's phenomenon. Two of the anti–U11/U12 RNP–positive patients had skeletal muscle involvement, 2 had renal crisis, and none had “intrinsic” (i.e., without associated pulmonary fibrosis) PAH. All anti–U11/U12 RNP–antibody positive patients tested had GI involvement compared with 75% of the anti–U11/U12 RNP–negative patients (P = 0.0110).
Most patients in the anti–U11/U12 RNP–antibody positive group (29 of 33) and the 1994–1995 comparison group (202 of 240) were evaluated for pulmonary fibrosis by radiography or high-resolution CT scanning. The vast majority (>80%) of patients in both groups had only routine chest radiographs performed because high-resolution CT was commonly available only after 1995. Pulmonary fibrosis was significantly more frequent in anti–U11/U12 RNP–positive versus negative patients (23 of 29 [79%] versus 74 of 202 [37%]; P < 0.0001). The proportion of U11/U12 RNP–antibody positive patients with pulmonary fibrosis who developed pulmonary fibrosis–related dyspnea during the first 4 years of illness (83%) was significantly greater than the proportion of U11/U12 RNP–negative patients (55%; P = 0.0205). Pulmonary fibrosis was equally distributed among patients with diffuse cutaneous (90%) and limited cutaneous (81%) involvement. There was no difference in the proportion of anti–U11/U12 RNP–positive and anti–U11/U12 RNP–negative patients with pulmonary fibrosis whose dyspnea on exertion began prior to the first physician diagnosis of SSc (52% and 41%, respectively). There was also no significant difference in the lowest forced vital capacity percentage predicted between the 2 groups. There was no difference in the proportion of patients in the 2 groups who were treated with corticosteroids and/or immunosuppressive drugs for SSc or lung disease.
Table 2 shows the frequency of radiographic pulmonary fibrosis at any time during the course of the disease in all patients with SSc followed at the University of Pittsburgh from 1982–2004, according to the presence of different SSc-associated serum autoantibodies. Patients with anti–U11/U12 RNP autoantibodies had the highest frequency of pulmonary fibrosis among all patient groups.
|SSc-associated autoantibody||PF frequency||P vs. U11/U12|
|U11/U12 RNP||23/33 (70)|
|Centromere||59/436 (14)||< 0.0001|
|PM-Scl||27/65 (42)||< 0.009|
|RNA polymerase III||63/343 (18)||< 0.0001|
|Th/To||47/121 (39)||< 0.002|
|Topoisomerase I||204/425 (48)||< 0.02|
|U1 RNP||36/123 (29)||< 0.0001|
|U3 RNP||16/84 (19)||< 0.0001|
The 10-year cumulative survival rate was assessed from the first symptom attributable to scleroderma in all anti–U11/U12 RNP antibody–positive patients compared with 240 consecutive anti–U11/U12 RNP–negative patients seen between 1994 and 1995. There was no significant difference in the 10-year survival rate between the U11/U12 RNP–positive patients and the U11/U12 RNP–negative patients (66% and 73%, respectively; log-rank P = 0.69).
We next compared the proportion of patients with pulmonary fibrosis who died of this disease complication using patients with known causes of death as the denominator. The highest mortality from pulmonary fibrosis occurred in anti–U11/U12 RNP–positive patients. Eleven (38%) of the anti–U11/U12 RNP antibody patients died during followup (mean followup 6.9 years). All 11 deaths were attributed to SSc lung disease (9 patients) or its complications (1 patient with lung cancer and 1 with pneumonia). Mortality from pulmonary fibrosis or its complications in the anti–U11/U12 RNP–positive patients (11 of 23 [48%]) was significantly higher than that in patients with all other autoantibodies combined (68 of 415 [16%]; P < 0.0001). Death due to pulmonary fibrosis in the anti–U11/U12 RNP–positive patients also was significantly more frequent when compared with the other 2 major lung fibrosis-associated autoantibodies, anti–topo I and anti-Th/To, combined (P = 0.0002). Among the causes of death in the U11/U12 negative comparison group, “primary” PAH was the most frequent, but other causes were represented including renal, cardiac, and GI scleroderma, and nonscleroderma causes such as cancer and infection.
We then examined survival in anti–U11/U12 RNP–antibody positive patients with pulmonary fibrosis (n = 23) and all other anti–U11/U12 RNP–antibody negative SSc patients in our series with pulmonary fibrosis (n = 74). Survival was 39% among anti–U11/12 RNP–positive patients with a median followup of 6.9 years, and 34% among anti–U11/U12 RNP–negative patients with a longer median followup of 9.7 years (P = 0.47). However, after adjusting for age at symptom onset, sex, and tobacco use, the presence of anti–U11/12 RNP antibodies among SSc patients with pulmonary fibrosis was associated with a 2.25-fold greater risk of death or lung transplant compared with anti–U11/U12 RNP–antibody negative individuals at any point in time (Table 3).
|Variable estimate ± SE||HR||95% CI||P|
|Age at symptom onset, years||0.06 ± 0.01||1.06||1.04–1.09||< 0.0001|
|Male||0.26 ± 0.29||1.30||0.73–2.31||0.36|
|Tobacco use||−0.63 ± 0.48||0.53||0.21–1.36||0.18|
|Anti–U11/U12 positive||0.81 ± 0.32||2.25||1.20–4.24||0.01|
We have identified and characterized anti–U11/U12 RNP autoantibodies in a group of patients with SSc. To determine the prevalence of this antibody in SSc, we evaluated 2 series of consecutive SSc patients first evaluated during 1994–1995 and 2004–2005 (4 total years). Fifteen (3.2%) of these 462 SSc patients had anti–U11/U12 autoantibodies and no other SSc-associated autoantibodies. As a comparison, the prevalence of these antibodies in a pediatric cohort of patients with SSc was 7% (Medsger TA Jr: unpublished observations). These antibodies are specific to SSc.
Pulmonary fibrosis is currently a leading cause of death in patients with SSc. In more than 2,000 of the Pittsburgh SSc patients, pulmonary fibrosis accounted for 44% of the SSc-associated deaths (4). The presence of anti–U11/U12 RNP antibodies was highly associated with pulmonary fibrosis (79%). In our experience, 2 other SSc-associated antibodies are predictive of pulmonary fibrosis. Pulmonary fibrosis was present in 48% of 87 SSc patients with anti-Th/To antibodies and limited cutaneous involvement, and was found in 48% of anti–topo I antibody–positive SSc patients with diffuse cutaneous involvement (1, 2). Furthermore, we have shown that SSc patients positive for anti–U11/U12 RNP antibodies are at a higher risk of death from pulmonary fibrosis–related causes than patients with any of the other 8 SSc-associated autoantibodies. Therefore, screening for anti–U11/U12 RNP antibodies can serve as a marker for severe pulmonary fibrosis in SSc.
In published series of SSc patients, 90% had 1 of 8 serum autoantibodies, including anti–RNAP III, anti– topo I, ACA, anti-Th/To, anti-PM-Scl, anti–U1 RNP, anti–U3 RNP, and anti-Ku (ref.20, and Medsger TA Jr: unpublished observations). Anti–U11/U12 RNP antibody is detected in approximately 3% of SSc patients and is therefore a ninth SSc-associated serum autoantibody, which raises the proportion of all SSc patients with one of these 9 autoantibodies to nearly 95%.
In conclusion, we have identified and characterized a novel autoantibody targeted against U11/U12 RNP in SSc patients with a high frequency of pulmonary fibrosis, a condition that is often severe and fatal.
The authors would like to thank Judy Webb and Pat Ambrose for administrative assistance.