Prevalence in myositis of antibodies recognizing anti-U3 RNA probably in a novel complex with 22/25 kD protein and not fibrillarin

Abstract

New antibodies against a U3 snRNP, which were named anti-Myo 22/25 antibodies, were detected in four (8%) of 53 serum samples from patients with polymyositis/dermatomyositis (PM/DM) by RNA immunoprecipitation. In the protein immunoprecipitation analysis, all four serum samples precipitated 22 kDa and 25 kDa proteins, which were not precipitated by normal serum or serum positive for antifibrillarin antibodies. Three of the four PM/DM patients had other identified autoantibodies including anti-PL-12 antibodies, antihistone antibodies (AHA), anti-SS-A antibodies and anti-SS-B antibodies defined by double immunodiffusion, ELISA or RNA immunoprecipitation, although there were no significant correlations between anti-Myo 22/25 antibodies and clinical or laboratory findings. There may be a subgroup of PM/DM patients whose sera are positive for anti-Myo 22/25 antibodies.

Introduction

Polymyositis/dermatomyositis (PM/DM) is classified as a collagen disease and is an uncommon inflammatory condition characterized by subacute systemic skeletal muscle involvement and by typical or various cutaneous lesions in DM. PM/DM shares many clinical and laboratory features including autoantibodies with other collagen diseases.

Autoantibodies to aminoacyl transfer ribonucleic acid (tRNA) synthetases have been reported as specific antibodies to PM/DM [1]. Antibodies to the aminoacyl synthetases for tRNA^his (Jo-1), tRNA^thr (PL-7), tRNA^ala (PL-12), tRNA^gly (EJ) and tRNA^ile (OJ) are already known [2]. Furthermore, there are many other autoantibodies which are found in the sera from patients with PM/DM. Anti-signal recognition particle (SRP) antibodies [3] were also reported to be more prevalent in PM than in DM [4]. Anti-Mi-2 antibodies, which are antibodies to the Mi-2 protein complex (55 kDa and 235 kDa) with unknown functions [5], were reported to be found rarely in patients with PM, but were found specifically in patients with DM, especially idiopathic DM. Anti-Ku antibodies, which are prevalent in Japanese patients with systemic sclerosis (SSc)–PM overlapping syndrome, and anti-PM-Scl antibodies, which are prevalent in Caucasian patients, have been discovered [6,7]. In addition, rare antibodies including anti-KJ, anti-Fer, anti-Mas and anti-U2 small nuclear ribonucleoprotein (snRNP) antibodies have also been reported [2]. However, the presence of antibodies to any U3 snRNP complex have not been reported in patients with PM/DM.

We have reported recently the existence of antihistone antibodies (AHA) in patients with PM/DM [8]. In that report, we performed absorption testing with histones in indirect immunofluorescence studies. From the absorption test, we found residual nucleolar stainings after absorption. Thus, we suspected the existence of some autoantibodies which show nucleolar staining in indirect immunofluorescence studies, such as anti-Th/To antibodies and anti-U3 snRNP antibodies.

We have recently introduced an RNA immunoprecipitation technique to identify autoantibodies which show nucleolar staining in indirect immunofluorescence studies in the sera of PM/DM patients and found several patients with DM whose serum had new antibodies against a U3 snRNP, which was named as anti-Myo 22/25 antibodies in this study.

In the present study, we determined the frequency of anti-Myo 22/25 antibodies in the sera from patients with PM/DM by RNA immunoprecipitation and protein immuneoprecipitation. We also investigated the correlations between anti-Myo 22/25 antibodies and the clinical or serological features in patients with PM/DM.

Materials and methods

Patients and samples

Serum samples were collected from 53 Japanese patients with PM/DM (16 men and 37 women; age range, 13–68 years; mean age 45 years). They were all patients with PM/DM who had visited our clinic over the last 15 years.

Patients with PM/DM fulfilled the criteria of Bohan and Peter [9] or were diagnosed as having amyopathic DM by clinical appearance and histopathological examinations. The PM/DM patients were classified into four subgroups: primary idiopathic PM (six patients); primary idiopathic DM (33 patients); DM associated with neoplasia (seven patients); and juvenile DM (seven patients), according to the classification of Bohan and Peter [9]. A skin biopsy was performed in DM patients, and electromyography and muscle biopsies were performed at the time of diagnosis in all patients. Patients who had other overlapping autoimmune diseases, including systemic sclerosis (SSc) or SLE, were excluded from this study.

Serum samples were also obtained from 30 patients with SSc who met the ARA criteria of SSc [10] but did not overlap with other collagen diseases. Furthermore, we also collected sera from a patient who was previously reported to be positive for anti-U3 snRNP antibodies [11]. Ten serum samples from healthy volunteers and one SLE patient with anti-Sm antibodies were also collected. All serum samples were stored at − 80°C prior to use.

RNA immunoprecipitation

RNA immunoprecipitation was performed with a slight modification to the technique of Forman [12]. Ten microlitres of test sera were mixed with 10 µl of protein A-agarose (Gibco BRL, Gaithersburg, MD, USA) in 500 µl of an immunoprecipitation (IPP) buffer (10 mm Tris-HCl, 500 mm NaCl, 0·1% Np-40, pH 8·0) and rotated overnight at 4°C. The agarose particles, which now carried absorbed IgG, were washed three times in 750 µl of the IPP buffer, then resuspended in 500 µl of a NET-2 buffer (50 mm Tris-HCl, 150 mm NaCl, 0·05% NP-40, pH 7·4). The resuspended resins were mixed with HeLa cell extracts of 3·0 × 10⁶ cells in 100 µl of a NET-2 buffer per each sample for 2 h at 4°C.

To extract bound nucleic acids, the washed agarose particles were extracted with phenol/chloroform/isoamyl alcohol and precipitated with cold ethanol. The pellets were washed with cold 70% ethanol, vacuum dried and dissolved in 20 µl of an electrophoresis sample buffer. After denaturing in sample buffer (10 m urea, 90 mm Tris/borate 1 mm ethylenediaminetetraacetic acid, 0·025% bromophenol blue, 0·025% xylene cyanol) at 65°C for 3 min, 10 µl of the samples per lane were applied for electrophoresis in 7 m urea-10% acrylamid gels in TBE. The gels were stained with a Silver stain plus kit (Bio-Rad, Hercules, CA, USA).

We were able to determine the presence of anti-Jo-1 antibodies, anti-EJ antibodies, anti-OJ antibodies, anti-PL-7 antibodies, anti-PL-12 antibodies, anti-SRP antibodies, anti-SS-A antibodies, anti-SS-B antibodies and anti-Th/To antibodies, in addition to anti-U3 snRNP antibodies.

Protein immunoprecipitation assays

As described previously [13], HeLa cells were incubated in methionine-depleted RPMI 1640 medium, which contained ³⁵S-methionine (2 µCi/ml) (Amersham, Arlinton Heights, IL, USA). The cells were harvested, sonicated on ice and centrifuged, and the supernatant was used immediately for immunoprecipitation procedures. Immunoprecipitation of radiolabelled cell extracts was performed on serum samples from all 52 patients, as described elsewhere [13]. Serum samples were incubated with protein A-agarose. The antibody-coated beads were then washed and incubated with radiolabelled HeLa cell extract, followed by repeat washing. The beads were boiled in sample buffer, and the sample were resolved in 10–20% gradient sodium dodecyl sulphate polyacrylamide gels [14]. Gels were enhanced with 0·5 m sodium salicylated and labelled proteins were detected by autoradiography.

Double immunodiffusion tests

Autoantibodies including anti-Ku antibodies, anti-PM-Scl antibodies and anti-Mi-2 antibodies which we could not identify by RNA immunoprecipitation were identified using a double immunodiffusion technique [15].

Enzyme-linked immunosorbent assay (ELISA) for AHA

ELISA for AHA was carried out as described previously [16,17]. Briefly, microtitre plates (Corning, Corning, NY, USA) were coated with 5 µg/ml of total histones (Sigma, St Louis, MO, USA). The wells were then blocked with phosphate buffered saline (PBS) containing 10% fetal bovine serum (PBS-FBS), and serum samples diluted to 1 : 400 in PBS-FBS were added to each well. The bound antibodies were detected with alkaline phosphatase-conjugated goat antihuman IgG or IgM antibodies (Cappel, Durham, NC, USA), using 10 mg/ml p-nitrophenyl phosphate (Gibco BRL, Gaithersburg, MD, USA) in 10% diethanolamine and 0·5 mm MgCl₂ (pH 9·8) as a substrate. The plates were read at 405 nm in a Microplate Reader (Bio-Rad). Absorption values greater than the mean plus 2 SD values in the normal control sera were considered positive.

Indirect immunofluorescence study

Antinuclear antibodies (ANA) or anticytoplasmic antibodies were detected by indirect immunofluorescence on HEp-2 cells as a substrate, as described previously [8,18]. Serum samples diluted 1 : 400 in PBS were used for the analysis of antinuclear antibodies or anticytoplasmic antibodies.

Absorption test in the immunofluorescence study using histones

We carried out an absorption test using histones in the immunofluorescence study, as described previously [8], using sera which were considered positive for IgG or IgM AHA in ELISA. Serum samples diluted 1 : 400 with 3 mg/ml total histones (Sigma)/PBS were incubated for 2 h at room temperature. After centrifugation (12 000 g, 15 min, room temperature), the supernatant was used for indirect immunofluorescence study of ANA as described above.

Statistics

The significance of differences was assessed by Student's t-test and chi-squared test. Relationships between variables were evaluated by linear regression analysis. P-values less than 0·05 were considered significant.

Results

Prevalence of antibodies against U3 snRNP in patients with PM/DM

The RNA immunoprecipitation revealed the existence of antibodies recognizing U3RNA either directly or in a complex containing U3RNA in four serum samples (8%) of the 53 patients with PM/DM (Fig. 1) and in three serum samples (10%) of the 30 patients with SSc. The prevalence of antibodies recognizing U3 RNA in patients with PM/DM was similar to that of patients with SSc.

Fig. 1.

RNA immunoprecipitation. Lane 1, healthy control; lane 2, SLE patient with anti-Sm antibodies; lane 3, SSc patient with anti-U3 snRNP (fibrillarin) antibodies; lane 4, idiopathic DM patient without antibodies against U3 snRNP; lanes 5–8, PM/DM patients with antibodies recognizing U3 RNA; lane 9, total HeLa extract. A patient showed co-existence of the anti-SS-A antibody and anti-SS-B antibody with antibodies recognizing U3 RNA (lane 5). Another patient showed co-existence of the anti-PL-12 antibody with antibodies recognizing U3 RNA (lane 6).

Three patients with antibodies recognizing U3 RNA were diagnosed as having idiopathic DM and the other patient was diagnosed as having idiopathic PM.

Protein immunoprecipitation analysis

Protein immunoprecipitation analyses were performed using four serum samples recognizing U3RNA, one serum positive for antifibrillarin antibodies and 10 control sera. In the protein immunoprecipitation analysis, all four serum samples recognizing U3 RNA also precipitated 22 kDa and 25 kDa proteins, which were not precipitated by normal serum or serum positive for antifibrillarin antibodies (Fig. 2). One serum precipitated 36 kDa protein which was not precipitated in other sera or control sera. The 34 kDa protein (fibrillarin) was not precipitated in any of PM/DM sera. Therefore, the antibodies in PM/DM sera apparently able to detect U3 RNA and the two proteins were different from so-called anti-U3 snRNP antibodies (antifibrillarin antibodies) and were named as anti-Myo 22/25 antibodies.

Fig. 2.

Protein immunoprecipitation assay. Lane 1, healthy control; lane 2, SSc patient with anti-U3 snRNP antibodies; lanes 3–6, PM/DM patients with antibodies recognizing U3 RNA. None of these four PM/DM sera precipitated 34 kDa fibrillarin proteins, but all four precipitated 22 kDa, 25 kDa and 45 kDa proteins which were not precipitated in normal control serum or serum positive for antifibrillarin antibodies.

Co-existence of anti-Myo 22/25 antibodies with other antibodies

There were three patients with the anti-Jo-1 antibodies, two with the anti-EJ antibodies, one with the anti-PL-12 antibodies, one with the anti-SRP antibodies, one with the anti-U1 RNP antibodies, five with the anti-SS-A antibodies, one with the anti-SS-B antibodies and four with anti-Ku antibodies. There were no serum samples with anti-PL-7, anti-OJ, anti-Th/To, anti-Mi-2 or anti-PM-Scl antibodies. IgG AHA were found in seven sera from patients with PM/DM, and IgM AHA were found in two sera. Both IgG AHA and IgM AHA were detected in one serum sample and eight serum samples were considered positive for AHA.

Co-existence with anti-Myo 22/25 antibodies was found in one case with the anti-PL-12 antibodies, one case with the anti-SS-A antibodies, anti-SS-B antibodies and anti-Ku antibodies, and in one patient with the anti-Ku antibodies.

Indirect immunofluorescence studies

In immunofluorescence analysis, all four serum samples which were positive for anti-Myo 22/25 antibodies in RNA immunoprecipitation showed homogeneous nuclear staining (Fig. 3a). After incubation of the sera with 3 mg/ml total histones, most homogeneous nuclear staining was decreased markedly, leaving a trace of nucleolar and cytoplasmic staining in two cases (Fig. 3b). The two cases were found positive for AHA. These findings suggested that AHA may have played the major role in the homogeneous nuclear staining in the indirect immunofluorescence analysis of these patients. The cytoplasmic and nucleolar residual stainings after incubation with total histones may indicate the co-existence of anti-Myo 22/25 antibodies and other autoantibodies.

Fig. 3.

Indirect immunofluorescence analysis and absorption by histones. Homogeneous stainings are shown in (a). After incubation of the sera with total histones, the homogeneous nuclear stainings disappeared and there were cytoplasmic and nucleolar residual stainings (b).

Clinical correlations

We examined the correlations between anti-Myo 22/25 antibodies and the clinical and laboratory findings of the patients with PM/DM (Table 1). In this study, no patients with anti-Myo 22/25 antibodies showed internal malignancy, and only one patient with both anti-Myo 22/25 antibodies and the anti-PL-12 antibodies showed interstitial pneumonitis. However, there were no statistically significant differences between patients with or without anti-Myo 22/25 antibodies in any clinical or laboratory findings, including interstitial pneumonitis or internal malignancies or Raynaud's phenomenon. There were no patients with pulmonary hypertension or heart involvement. The co-existence of AHA or identified antibodies was seen more often in patients with anti-Myo 22/25 antibodies than those without. However, we could not find any correlations between other identified specific antibodies and the existence of anti-Myo 22/25 antibodies. This may be because of the small number of patients with each specific autoantibodies.

Table 1.

The clinical features of the PM/DM patients with and without anti-Myo 22/25 antibodies (anti-Myo 22/25)

	Anti-Myo 22/25(+) (n = 4)	Anti-Myo 22/25(-) (n = 49)
Male/female	1/3	15/34
Age (years; mean ± s.d.)	44 ± 19	46 ± 16
No. (%) with:
Proximal muscle weakness	4/4 (100%)	40/49 (82%)
Skin eruption	3/4 (75%)	44/49 (90%)
Gottron's papules	2/4 (50%)	29/49 (59%)
Heliotropic colouration	3/4 (75%)	23/49 (47%)
Raynaud's phenomenon	1/4 (25%)	5/48 (10%)
Interstitial pneumonitis	1/4 (25%)	15/49 (31%)
Internal malignancy	0/4 (0%)	7/49 (14%)
Heart involvement	0/4 (0%)	4/49 (8%)
Elevated ESR	3/4 (75%)	30/49 (61%)
Elevated CK	4/4 (100%)	36/49 (73%)
Positive AHA *	2/4 (50%)	6/49 (12%)
Positive ARS	1/4 (25%)	4/49 (8%)
Positive anti-SS-A	1/4 (25%)	4/49 (8%)
Positive defined antibodies *	3/4 (75%)	13/39 (27%)

AHA; antihistone antibodies: ARS; antitRNA synthetase antibodies (Jo-1, PL-7, PL-12, OJ, EJ).

^*P < 0·05.

Discussion

Small nuclear ribonucleoproteins are particles with both RNA and protein components that are found in the nucleus of all eukaryotic cells [19]. They contain one RNA molecule associated with one or several proteins. At present, a large number of snRNPs are known. U3, U8, U14 and U22 snRNP exist in the nucleolus and the other snRNP exist in the nucleoplasm. These particles have various functions, including premRNA processing (U1, U2, U4, U5 and U6 snRNP), histone mRNA 3′ formation (U7 snRNP) and rRNA maturation (U3, U8, U14 and U22 snRNP) [20].

U3 snRNP consists of U3 small nuclear RNA and at least six proteins, including 74, 59, 36, 30, 13 and 12·5 kDa proteins [21]. Of these proteins, only the common snRNP protein, fibrillarin, has been characterized. U3 snRNPs are located in the fibrillar centre of the nucleolus and anti-U3 snRNP antibodies from patients with scleroderma show clumpy nucleolar staining in indirect immunofluorescence studies [22]. In this study, no serum samples from patients with PM/DM showed obvious nucleolar staining in indirect immunofluorescence study. The residual nucleolar immunofluorescence in the absorption test with histones in indirect immunofluorescence studies suggests that one reason for the failure to detect nucleolar patterns in indirect immunofluorescence studies was the co-existence of other autoantibodies such as AHA.

Anti-U3 snRNP antibodies were first reported in the sera of the patients with systemic sclerosis [23]. Subsequent studies revealed several clinical correlations between anti-U3 snRNP antibodies and SSc. The prevalence of anti-U3 snRNP antibodies was higher in black SSc patients than white or oriental SSc patients [22,24], and in diffuse cutaneous SSc patients than limited cutaneous SSc patients [25‐27]. Furthermore, SSc patients with anti-U3 snRNP antibodies tended to be accompanied with primary pulmonary hypertension, renal involvement, intestinal involvement and muscle involvement [22].

In this study, we could not detect the precipitation of fibrillarin which is usually coprecipitated with U3 RNA, but we detected the precipitation of 22 kD and 25 kD proteins using all 4 serum samples that also recognized U3 RNA. We have concluded that our four sera recognize a novel U3 RNP particle containing U3 RNA and the 22 and 25–27 kD proteins, but not fibrillarin. However, this conclusion is preliminary and unsubstantiated, since we did not carry out RNA immunoprecipitations using deproteinized cell extracts. Thus, we have not excluded the possibility that our sera contain separate populations of antibodies, one set recognizing U3 RNA and another set detecting the 22 and 25 kD polypeptides.

Lee and Baserga reported about a 22-kD protein in yeast, named Imp3p U3 RNP [28]. The protein Imp3p has homology to yeast ribosomal protein S9 protein and S4 protein in Escherichia coli. The protein interacts in the yeast with Mpp10p under physiological conditions to process pre18S ribosomal RNA (rRNA). The protein is suggested to be a probable candidate for direct U3snoRNA binding in yeast [28]. The size of the precipitated proteins in our study are similar to yeast Imp3p. If there is a protein conserved between yeasts and human cells, this could be because of an important interaction with U3snoRNA [28].

It is of interest that previous studies of myositis have not reported antibodies to the 22/25 kDa proteins or U3 RNA. One possibility is that this is an immune response occurring more often in Japanese than western patients. Indeed, it is already known that anti-Ku antibodies occur in Japanese but not western patients with myositis–scleroderma overlap.

We could not find any correlations of anti-Myo 22/25 antibodies with clinical manifestations in this study. The most severe complications are internal malignancy and interstitial pneumonitis in PM/DM. However, no patients with anti-Myo 22/25 antibodies were accompanied with internal malignancy or interstitial pneumonitis, except for one patient with both anti-Myo 22/25 antibodies and anti-PL-7 antibodies, which are one of the disease marker of interstitial pneumonitis. Thus, with more patients it may be possible to reveal the clinical symptoms of PM/DM patients with anti-Myo 22/25 antibodies.

We found a significantly high prevalence of anti-Myo 22/25 antibodies associated in patients with well-known autoantibodies, including anti-SS-A antibodies, anti-RNA synthetase antibodies and AHA. This observation suggests a subgroup of PM/DM patients with various autoantibodies including anti-Myo 22/25 antibodies. We do not know whether the initiation of PM/DM is related to the appearance of any of these autoantibodies in these patients, but our findings emphasize that the pathogenesis of PM/DM is probably heterogeneous.