Anti-multiple nuclear dots (anti-MND) and anti-SP100 antibodies in hepatic and rheumatological disorders

Abstract

Multiple nuclear dots pattern has been described in primary biliary cirrhosis and, less often, in rheumatological disorders. Sp100 is the major antigen of multiple nuclear dots. We evaluated prevalence and diagnostic significance of multiple nuclear dots and anti-Sp100 reactivity both in hepatic and rheumatological diseases. A series of 283 consecutive liver patients (89 primary biliary cirrhosis, 12 primary sclerosing cholangitis, 85 autoimmune hepatitis, 97 hepatitis C virus-related chronic liver disease) and of 89 consecutive rheumatological cases were evaluated. Presence of multiple nuclear dots was assessed by indirect immunofluorescence on HEp-2 cells, anti-Sp100 reactivity by ELISA with recombinant protein. Multiple nuclear dots were detected in 20 patients (7%) with liver disease (of whom 15 with primary biliary cirrhosis), and in eight patients (9%) with rheumatological disorders. Anti-Sp100 was detected in 45 liver patients (16%), of whom 30 with primary biliary cirrhosis, but in only two with rheumatological disorders (2%) (P = 0·0004). The concordance between multiple nuclear dots and anti-Sp100 in liver and rheumatological patients was 90% and 25% (P = 0·0018), respectively. Among 89 consecutive patients with primary biliary cirrhosis, multiple nuclear dots and anti-Sp100 were present in 17% and 34%, respectively (P = 0·0152). Anti-Sp100 positivity was associated with older age and higher gamma-globulin levels. Multiple nuclear dots are similarly observed in liver and rheumatological patients. In contrast, anti-Sp100 is more frequent in liver patients and is significantly more often detected in primary biliary cirrhosis, of which it can be regarded as a highly specific serological marker. The antigenic target of multiple nuclear dots in most rheumatological patients is other than Sp100.

INTRODUCTION

Within the heterogeneous spectrum of antinuclear antibodies (ANA) revealed by indirect immunofluorescence (IFL), the multiple nuclear dots (MND) pattern is immunomorphologically characterized by the staining of 3–20 dots of variable size distributed all over the cell nucleus, but sparing the nucleoli, and – in contrast to the anticentromere pattern – MND reactivity does not stain the chromosomes in mitotic cells [1,2]. Early studies suggested that MND reactivity is highly specific for primary biliary cirrhosis (PBC) [3,4] irrespective of the antimitochondrial (AMA) status, but this is no longer accepted [5]. Sp100 is the main antigenic target of MND reactivity and consists of a 53-kDa nuclear protein [6] with transcription stimulating activity [7,8]. Anti-Sp100 antibodies were described in association with PBC, especially AMA negative PBC, and less often in rheumatological disorders [2]. Recently, the presence of anti-Sp100 antibodies in PBC patients was associated with an unfavourable course of the disease [9]. Besides Sp100, another nuclear protein called PML has been identified as an additional target of anti-MND positive sera of PBC patients [10]. PML colocalizes with Sp100 in the nuclear dot structures and is expressed abundantly in promyelocytic leukaemia cells [11,12]. Sp100 and PML were shown to be co-immunogenic in PBC patients, since anti-Sp100 and anti-PML usually co-exist in the same serum [9,10].

Given their reported specificity for PBC, in this study we evaluated the prevalence of anti-MND pattern and of anti-Sp100 reactivity in patients with different hepatic and rheumatological diseases, their correlation and their disease specificity.

PATIENTS AND METHODS

Patients and sera

Over the last nine years in our Autoimmune Serology Laboratory we evaluated nearly 30 000 sera, mainly from liver and rheumatological patients, for the detection of non-organ specific autoantibodies. The screening test is performed by indirect immunofluorescence on cryostat sections of rat liver, kidney and stomach. When ANA are present, their immunomorphological pattern is characterized further by indirect immunofluorescence on HEp-2 cells, a technique which unambiguously identifies anti-MND reactivity.

Since both anti-MND and anti-Sp100 have been detected mainly in hepatic and in rheumatological disorders [2], we focused our study on 283 consecutive patients with hepatic disease of different aetiology (89 with PBC, of whom 76 were AMA-positive and 13 AMA-negative PBC; 12 with primary sclerosing cholangitis, PSC; 85 with type 1 autoimmune hepatitis, AIH; and 97 with hepatitis C virus-related chronic hepatitis) and 89 consecutive patients with different rheumatological disorders (33 with Sjögren’s syndrome, 11 with rheumatoid arthritis, 43 with systemic lupus erythematosus, one with mixed connective tissue disease and one with vasculitis). The diagnosis of PBC was based on the major and minor criteria outlined by Taal and collaborators [13].

Indirect immunofluorescence (IFL)

Sera were tested by indirect IFL on commercially available HEp-2 cells (Kallestad, Chaska, MN, USA) at 1 : 40 dilution. Fluorescein-conjugated anti-human immunoglobulin (antihuman polyvalent immunoglobulin IgA, IgG, IgM FITC conjugate, Sigma Immunochemicals, St Louis, MO, USA) was used as secondary antibody at 1 : 100 dilution. The immunomorpho-logical patterns of reactivity were assessed under a fluorescence microscopy (Orthoplan, Leitz, Wetzlar, Germany).

Anti-Sp100 ELISA

Serum reactivity to recombinant Sp100 was assessed using a commercially available ELISA, assay according to the manufacturer’s instructions (Imtec, Berlin, Germany).

Statistical analysis

Comparison of categorical variables was performed using Fisher’s exact test. The Mann–Whitney test was used for comparison of continuous data. A probability (P) value less than 0·05 was considered significant. Statistical analysis was performed using GraphPad InStat version 3·0a for Macintosh, Graph-Pad Software, San Diego, CA, USA (www.graphpad.com) and StatView 5·0.1 for Macintosh, SAS Institute Inc., Cary, NC, USA (www.statview.com).

RESULTS

We identified 28 cases with anti-MND by IFL out of 283 consecutive liver and 89 consecutive rheumatological patients. Twenty of them were suffering from the following liver diseases: 15 PBC (of whom nine were AMA positive and six AMA negative), four HCV-related chronic hepatitis and one type 1 AIH (Table 1). All anti-MND positive liver patients but two (both with HCV-related chronic liver disease) had biochemical and/or histological evidence of pronounced cholestasis.

Table 1.

Prevalence of anti-MND and anti-Sp100 in hepatic and rheumatological disorders

	Anti-MND+	Anti-Sp100+
Primary biliary cirrhosis	15/89 (17%)1	30/89 (34%)A
Primary sclerosing cholangitis	0/12 (0%)2	1/12 (8%)B
Autoimmune hepatitis type 1	1/85 (1%)3	7/85 (8%)C
HCV related chronic hepatitis	4/97 (4%)4	7/97 (7%)D
Rheumatological diseases *	8/89 (9%)5	2/89 (2%)E

¹Versus

²:not significant;

³: P = 0·0003;

⁴: p = 0·0064;

⁵: not significant;

^Aversus

^B: not significant;

^C: P < 0·0001;

^D: P < 0·0001;

^E: P < 0·0001. Test used: Fisher’s exact test. Anti-MND: anti-multiple nuclear dots;

^*43 systemic lupus erythematosus, 33 Sjögren’s syndrome, 11 rheumatoid arthritis, one mixed connective tissue disease, one vasculitis.

The remaining eight anti-MND positive sera were observed in patients with the following connective tissue disorders: Sjögren’s syndrome (three cases), rheumatoid arthritis (two cases), systemic lupus erythematosus, vasculitis and mixed connective tissue disease (one case each). Two of these eight anti-MND positive patients, both with Sjögren’s syndrome, had high serum alanino-aminotransferase and aspartate-aminotransferase levels and their liver histology showed ‘mild hepatitis’ in both cases, but cholestasis was not present and neither of them reached the score of ‘probable’PBC [13]. The remaining six rheumatological patients lacked any clinical and biochemical evidence of liver involvement.

Anti-Sp100 was detected in 45 (16%) of 283 liver patients: 30 (34%) with PBC, seven (8%) with type 1 AIH, seven (7%) with HCV-related liver disease and one (8%) with PSC (Table 1).

Among 89 consecutive patients with PBC, anti-MND and anti-Sp100 were present in 17% and 34%, respectively (P = 0·0152). In comparison with all the other liver disease patients and the rheumatological patients as control cases, the specificity of anti-MND and anti-Sp100 for PBC diagnosis was similar (0·9541 and 0·9399, respectively), whereas anti-Sp100 had higher sensitivity (0·3371 versus 0·1685) and likelihood ratio (5·611 versus 3·669). Anti-Sp100 positivity was also associated with older age and higher gammaglobulin levels, whereas anti-MND positivity was not correlated to any peculiar clinical or biochemical profile.

Two (2%) of 89 consecutive patients with rheumatological disorders were positive for anti-Sp100, and both were also anti-MND positive.

The concordance between multiple nuclear dots and anti-Sp100 in liver and rheumatological patients was 90% and 25% (P = 0·0018), respectively. The detailed picture of anti-MND and anti-Sp100 status according to the different diseases is summarized in Table 2. Sera positive for both MND and anti-Sp100 had significantly higher levels of anti-Sp100 antibodies than MND-negative ones (P = 0·0053, Mann–Whitney test), irrespective of the underlying disease (Fig. 1).

Table 2.

Anti-MND and anti-Sp100 status in hepatic and rheumatological disorders

	MND+ anti-Sp100+	MND+ anti-Sp100−	MND− anti-Sp100+
Primary biliary cirrhosis(89 patients)	13 (15%)	2 (2%)	17 (19%)
Primary sclerosing cholangitis(12 patients)	0	0	1 (8%)
Autoimmune hepatitis type 1(85 patients)	1 (1%)	0	6 (7%)
HCV-related chronic hepatitis(97 patients)	4 (4%)	0	3 (3%)
Rheumatological diseases(89 patients)*	2 (2%)	6 (7%)	0

Anti-MND: anti-multiple nuclear dots;

^*43 systemic lupus erythematosus, 33 Sjögren’s syndrome, 11 rheumatoid arthritis, one mixed connective tissue disease, one vasculitis.

Fig. 1.

Bivariate scattergram showing levels of anti-Sp100 reactivity (Units/l) according to the immunofluorescence antimultiple nuclear dots (anti-MND) status in 47 anti-Sp100 positive sera. Anti-MND⁺ (20 cases; median 1153Ul, range 64–9798) versus anti-MND⁻ (27 cases; median 122U/l, range 45–4889): P = 0·0053 (Mann–Whitney test).

DISCUSSION

A previous study [14] reported that ANA with the MND pattern is detected essentially in PBC, especially in AMA-negative patients, but there is recent evidence suggesting that this is not always the case. Pawlotsky et al. [5] analysed 38 patients selected on the sole basis of anti-MND positivity, and within this series only 42% had PBC, whereas the remaining 22 patients (58%) were given a different diagnosis (hepatic as well as immuno-logical disorders, without liver involvement). In keeping with Pawlotsky’s study, in our experience anti-MND per se is neither specific for liver disease nor for PBC, since it is also present in a significant number of patients with different rheumatological disorders (Table 1).

The two rheumatological patients with Sjögren’s syndrome positive for both anti-MND and anti-Sp100 failed to meet the criteria for the diagnosis of PBC [13]. The possibility that these two patients will develop PBC in years to come cannot be completely excluded; however, their lack of cholestasis and their liver histology not consistent with PBC makes this possibility unlikely [15]. However, among liver patients, anti-MND is detected more often in PBC (17%) than in other diseases such as HCV-related chronic hepatitis (4%), type 1 AIH (1%) and PSC (0%).

In comparison with anti-MND positivity, anti-Sp100 appears to be a more sensitive serological marker of liver disease, and in particular it is more PBC-specific (Table 1). Our data indicate that Sp100 is the antigenic target of the MND reactivity in most liver patients, particularly in nearly all PBC cases, whereas an additional, still unknown antigen, different from Sp100, appears to be the target of the MND reactivity associated with connective tissue disorders.

Besides the enhanced likelihood ratio of anti-Sp100 in PBC, the quantitative and objective ELISA assay represents a further advantage over IFL, which requires an experienced observer to correctly identify unusual ANA patterns such as anti-MND. Anti-Sp100 therefore has all the prerequisites to replace anti-MND detection as an additional serological marker of PBC. Interestingly, its role would be extremely useful in the diagnostic work-up of AMA-negative PBC patients.

In 84% anti-MND negative/anti-Sp100 positive sera an ANA pattern other than MND was detected by IFL, therefore a latent anti-MND was likely to be obscured by other ANA specificities. It is not surprising that an ELISA assay is intrinsically more sensitive and objective than the ascertainment of an immunomorphological pattern by IFL on HEp-2 cells: the observation that anti-MND positive sera have higher levels of anti-Sp100 seems to support such an interpretation (Fig. 1).

Previous studies have shown that in PBC patients the most reactive fragment of Sp100 spans the 240–474 amino acid domain of the protein [10]. Since we tested all sera against the full-length Sp100 protein, it will be interesting to determine if the reactive epitope(s) are one and the same in hepatic diseases other than PBC and in non hepatic anti-Sp100 positive disorders.

In rheumatological patients, on the other hand, the main MND target antigen is likely to be other than Sp100. We cannot rule out that PML may account, at least partly, for anti-MND detected in connective tissue disorders, given the frequent co-existence of anti-Sp100 and anti-PML, at least in PBC patients [9,10]. Detection of MND pattern by conventional IFL will continue to prove useful in rheumatological patients, where anti-Sp100 specificity accounts for only few anti-MND positive cases.