Over-expression of TATA binding protein (TBP) and p53 and autoantibodies to these antigens are features of systemic sclerosis, systemic lupus erythematosus and overlap syndromes

Abstract

The aim of this study was to determine the expression levels of p53 and TATA binding protein (TBP) and the presence of autoantibodies to these antigens in Asian Indian patients with systemic sclerosis (SSc), overlap syndromes (OS) and systemic lupus erythematosus (SLE). Fifty patients with SSc, 20 with OS, including mixed connective tissue diseases (MCTD), 20 with SLE, 10 disease controls (DC) and 25 controls (C) were studied. The over-expression of p53 and TBP antigen was determined quantitatively by sandwich enzyme-linked immunosorbent assay (ELISA), varies between four- and sevenfold higher in patients with SSc, OS and SLE, in comparison to DC and C. The expressed protein antigens were not present as free antigens but as immune-complexes. Autoantibodies to p53 were detected by ELISA in 78% subjects with SSc, 100% with OS and 80% with SLE. Autoantibodies to TBP were observed in 28% patients with SSc, 25% with OS and 15% with SLE. In comparison to healthy controls, the titre of antibodies to p53 was significantly higher in patients with SSc (P = 0·00001) than the patients with OS (P = 0·00279) and SLE (P = 0·00289), whereas the titre of antibodies to TBP was higher in patients with OS (P = 0·00185) than the SLE (P = 0·00673) and the SSc (P = 0·00986) patients. Autoantibodies to p53 and TBP were detected in all these patients and the levels of these two autoantibodies showed weak negative correlation with each other. We propose that the over-expression of these antigens might be due to hyperactive regulatory regions in the p53 and TBP gene.

INTRODUCTION

Antibodies to DNA [1,2], RNA and various nuclear proteins [3,4–12] are a common feature of autoimmune disorders. The p53 tumour suppressor gene, a nuclear transcription factor, plays an important role in the control of cell proliferation and apoptosis. In white patients, over-expression of p53 has been detected in the synovial tissue of patients with long-standing rheumatoid arthritis (RA) [13,14], whereas the somatic mutations in the p53 gene have been reported in RA synovial tissue [15–17]. The expression of p53 antigen was found to be higher in synovial tissue from patients with long-standing RA when it was compared with osteoarthritis and reactive arthritis [18]. Further, serum anti-p53 antibodies had been detected in patients with type I diabetes [19], systemic lupus erythematosus (SLE) [20] and autoimmune thyroid disease [21], but were rarely detected in RA and Sjögren's syndrome [22]. An absence of anti-p53 antibodies was observed in Chinese patients with RA and SLE [23] and it was suggested that ethnic differences in Asian patients with autoimmune diseases are responsible for the lack of antibodies against p53 in that population [24]. The functional interaction between p53 and TATA binding protein (TBP) has been shown in vivo[25] and mice inoculated with T antigen-expressing plasmids produced antibodies to TBP [26].

The differential expression profile of p53 antigen and the detection of autoantibodies to p53 in different racial phenotypes led us to investigate the presence of serum anti-p53 and anti-TBP antibodies and to determine, quantitatively, the expression level of p53 and TBP antigens in heterogeneous Asian Indian patients with systemic sclerosis, SLE and overlap syndromes. In this report, we have shown that an over-expression of p53 and TBP antigens, that varies between four- and sevenfold higher in comparison to controls, was a common feature in patients with SSc, OS and SLE, and the over-expressed antigens were present in the sera as immune complexes. More than 80% of the patients with SSc, OS and SLE have autoantibodies to p53 whereas 25% of the patients have autoantibodies to TBP. The individual antibodies titre to p53 was higher in patients with SSc than the patients with overlaps and SLE. However, antibodies titre to TBP is higher in OS than SLE and SSc. We have proposed that the over-expression of these antigens might be due to a hyperactive promoter of these genes.

MATERIALS AND METHODS

Selection of patients

Of 90 patients (mean age 38·6 years ± s.d. = 0·08; mean disease duration 8 years), 50 SSc patients with localized and systemic scleroderma, 20 patients with OS, of whom nine were identified with mixed connective tissue diseases (MCTD), and 20 patients with SLE were selected on the basis of clinical investigation at the All India Medical Institute of Sciences, New Delhi. These patients fulfilled the American College of Rheumatology criteria for scleroderma [27] and revised 1997 criteria for SLE [28,29]. The term ‘overlap syndromes’ was used when patients had combined features of more than one rheumatic diseases. MCTD was diagnosed according to criteria proposed by Alarcon-Segovia and Cardiel [30]. Our study protocol included the detailed analysis of clinical and biochemical indices. All patients gave their informed consent and this study was approved by the Indian Council of Medical Research Committee on ethics in research involving humans. Additionally, 25 normal sera from healthy blood donors as well as 10 disease control sera from patients with chronic tuberculosis chest (TBC) were obtained.

Storage of sera

Blood samples without anticoagulant were centrifuged at 800 g at 4°C for 10 min to separate sera. Sera were stored at −70°C in aliquots for further analysis. Sera in use were maintained strictly at 4°C.

Purification of recombinant p53 protein

Sf21 cells were grown in Grace's insect cell culture medium (Biological Industries, Kibbutz Beit Haemek, Israel) supplemented with 3·33% lactalbumin, 10% fetal calf serum (FCS), 1× penicillin streptomycin mix (Sigma, St Louis, MO, USA); 1 × 10⁶ Sf21 cells were seeded in 90 mm plates. After 12 h, the media were removed and 2 × 10⁵ pfu (wtp53 baculovirus expression vector) was added to the cells. The virus was removed after 1 h of incubation and complete media was added. After 48 h of incubation, cells were washed with cold phosphate buffered saline (PBS) and lysed with 1·6 ml of lysis buffer (50 mm Tris-HCl, pH 8·0, 150 mm NaCl, 1% NP-40, 1 mm DTT and 0·35 mm phenyl methyl sulfonyl fluoride (PMSF)). The cell lysate was subjected to Western analysis for authenticity of protein. After 30 min of incubation on ice, cells were pelleted at 20 000 r.p.m. for 30 min; 2 µg of mouse monoclonal PAb421 anti-p53 antibody (p53 Ab-1, Oncogene Research Products, Boston, MA, USA) was added to the supernatant and incubated further for 1 h with continuous rocking at 4°C. One hundred µl of swollen protein A-Sepharose was added to it and incubated for 1 h at 4°C on a rocking platform. Tertiary complexes were collected at 12 000 g for 20 s at 4°C and resuspended in wash buffer (10 mm Tris-HCl, pH 8·0, 150 mm NaCl, 1% NP-40, 1 mm DTT, 0·35 mm PMSF). Beads were incubated for 20 min at 4°C on a rocking platform and washed three times with RIPA buffer. The final wash was performed with 10 mm Tris-HCl, pH 7·5, 0·1% NP40 and centrifuged at 12 000 g for 20 s at 4°C. The purified protein was detected with Coomassie and silver staining and further confirmed by Western blot using mouse monoclonal PAb421 anti-p53 antibody (p53 Ab-1, Oncogene Research Products).

Purification of recombinant TBP

The coding region of TATA-box binding protein TBP (kindly received from R. G. Roeder, Rockefeller University) was cloned in pET11d expression vector and was utilized for protein purification. Briefly, an overnight-grown inoculum BL21 (DE3 pLys S) strain transformed with 6-His-hTBP plasmid construct was used for culture in luria broth (LB) medium at 200 r.p.m. at 37°C. The culture of 0·7–0·8 OD₆₀₀ was induced with 0·4 mmβ-D isopropyl-thiogalactopyranoside (IPTG) for 3 h at 200 r.p.m. at 30°C. Induced cells were sonicated at 30 s alternative blasts for 5 min, 20% output in lysis buffer (10 mm Tris-HCl pH 7·9, 10% glycerol, 0·5 m NaCl, 0·1% NP40, 5 mm DTT, 0·5 mm PMSF). The expressed protein was analysed by Western analysis using polyclonal anti-TBP antibody as well as monoclonal Anti-His antibody (Novagen, Darmstadt, Germany). The cleared lysate, along with 1 mm immidazole HCl (pH 7·9), was loaded on top of a Ni-NTA column (QIAexpress®, Hilden, Qiagen) at a flow rate of 5–10 column vols per hour. The column was washed with 20 column vols of BC100 (20 mm Tris-HCl pH 7·9, 20% glycerol, 100 mm KCl, 5 mm DTT, 0·5 mm PMSF) containing 20 mm imidazole.HCl followed by elution with BC100 containing 100 mm imidazole.HCl. The purified protein was detected with Coomassie and silver staining and confirmed further by Western blot utilizing anti-TBP polyclonal antibodies.

ELISA

The purified proteins p53 and TBP were used as test antigens in ELISA antibody assay. The proteins at 10 µg/ml in 0·05 m carbonate–bicarbonate buffer were coated onto alternate rows of maxisorp immunoplates (Nunc™, Nalge Nunc International, Dusseldorf, Germany) and left overnight at 37°C in humid chamber. The plates were blocked for 6 h with 1% bovine serum albumin (BSA) in carbonate–bicarbonate buffer and washed with 1× tris borate saline tween (TBST) buffer (10 mm Tris, pH 8·0, 150 mm NaCl with 0·05% Tween 20). The duplicate samples for each test serum were diluted 1 : 333 (determined from percentage reactivity over a dilution range of 1/100–1/50 000) in 1× TBST with 0·1% BSA, added to the wells with and without antigen, left for 6 h at 37°C and rewashed. Antibody binding was detected with 100 µl of antihuman IgG, H&L-specific, alkaline phosphate conjugated (Calbiochem, San Diego, CA, USA), diluted 1 : 5000 in TBST with 0·1% BSA. Plates were left for 2 h at 37°C and rewashed. The reactivity was visualized using 100 µl 1 mg/ml para nitro phenyl phosphate (PNPP) (Sigma) in 10 mm diethanolamine (pH 9·5) having 0·5 mm MgCl₂; after 30 min 0·1 m ethylenediaminetetraacetic acid (EDTA) was used and the colour development was measured in a microplate reader at 405 nm (Model 550, Bio-Rad, Hercules, CA, USA). The mean optical density (OD) of the background duplicates in the absence of antigen was subtracted from the OD with antigen to obtain the final reading. The cut-offs for positivity for anti-p53 and anti-TBP were derived for normal human serum samples tested concurrently in each essay run.

Sandwich ELISA

A 96-well U-bottomed Maxisorp immunoplate was coated with 100 µl of mouse monoclonal antibody PAb421 to p53 in 1 : 5000 dilution (diluted in 50 mm carbonate buffer, pH 9·6) by incubating overnight at 4°C. Excess antibody was removed by washing three times with TBST. Blocking was performed by 200 µl of 2% BSA in TBS (pH 7·4) and allowed to incubate for 2 h at room temperature and washed three times with TBST. The recombinant p53 protein was added in a twofold serial dilution to obtain a standard plot. Sera were diluted to 1 : 40 times in TBS and 15 µl of each were added to the wells in duplicate. The plate was incubated at 37°C for 2 h and washed three times with TBST. One hundred µl of a second mouse monoclonal antibody PAb1801 (p53 Ab-2, Oncogene Research Products) was diluted (1 : 5000) with 1% BSA in TBST and then added to each well. The plate was incubated for 2 h at 37°C in humid chamber and then washed. In a similar manner, two antibodies against TBP (ITBP20, CRP Inc., Cambridge, VA, USA) and 1TBP18 (Abcam, UK) were utilized. Antimouse IgG, alkaline phosphate conjugated (Calbiochem) was diluted by 1 : 5000 in 1% BSA with TBST and then added to each well. The plate was incubated for 2 h at 37°C in a humid chamber and washed as described earlier. Finally, 100 µl of alkaline phosphate substrate with 50 mm Na₂CO₃, 1 mm MgCl₂ and 1 mg/ml PNPP was added and the enzymatic reaction was allowed to take place for 10–30 min at room temperature. The reaction was terminated by adding 50 µl of 0·1 m EDTA pH 8·0. The optical density at 405 nm was determined using a Microplate reader. The mean OD of antigen was subtracted from the blank OD to detect the background effect of the reagents used. Thus the measured density of antigen in patients was plotted along with that of the control to obtain a cut-off value for the positive reaction. Calculation of the sera antigen titres was performed by interpolating the OD405 values by drawing a line down to the x-axis, corresponding to the antigen concentration.

ELISA inhibition assay

The purified proteins of p53 and TBP were utilized to sensitize a 96-well microtitre plate in an ELISA inhibition assay following the above-mentioned protocol for antigen coating. Another unsensitized microtitre plate was used and marked with permanent marker to identify the intended location of all samples to be tested and to prepare serial dilutions of test sera. The outermost wells of the plate were left unused, as these were not sensitized in the previous plate to avoid evaporation. The first column of these wells contained undiluted aliquots of the sera samples to be tested and the rest were in 2% BSA in TBST (pH 7·4). The serum samples of patients that were to be tested were mixed thoroughly to achieve 1 : 2 dilution by pipetting up and down several times. Each sample was subjected to further several-fold serial dilutions in order to achieve all the desired dilutions with no intermixing. The antibody against p53, i.e. PAb1801 (p53 Ab-2, Oncogene Research Products) and antibody against TBP, i.e. (ITBP20, CRP Inc.) were diluted individually (1 : 5000) with 1% BSA in TBST and used in all wells (experimental and control). The sensitized and blocked microtitre plate was retrieved and the positive and negative controls (the positive control in this assay was a purified sample of the antigen of interest, the same antigen with which the previous plate has been sensitized; the negative control was PBS) were located. One hundred µl antisera–antigen solutions from the non-sensitized plate were transferred to their corresponding wells on the sensitized and blocked plate. The plate was incubated at 37°C for 1·5 h after covering with parafilm to minimize evaporation. After discarding the solution, all the wells were washed three times with TBS/Tween 20. To each well, 100 µl of conjugate antibody antimouse IgG, alkaline phosphate conjugated (Calbiochem), was added after diluting 1 : 5000 in 1% BSA with TBST. The plate was incubated for 2 h at 37°C in a humid chamber and washed as described earlier. For alkaline phosphatase conjugate, we used 1 mg/ml p-nitrophenyl phosphate disodium (Sigma) dissolved in substrate buffer with 50 mm Na₂CO₃ and 1 mm MgCl_2. The enzymatic reaction took place for 10–30 min at room temperature. The reaction was terminated by adding 50 µl of 0·1 m EDTA pH 8·0. The OD at 405 nm was determined using a Microplate reader.

Titre determination for anti-p53 and anti-TBP antibodies

In order to show maximum binding with test antigens, the dose-dependency of the capacity of anti-p53 and anti-TBP in sera was tested by ELISA. A dosage range from 1/100 to 1/50 000 was used at twofold dilutions and other mean dilutions to reach a titre of maximum cut-off values with respect to the control normal sera.

Immunoblot analysis of autoantibodies

Affinity purified TBP and p53 were used to test sera anti-p53 and anti-TBP by Western blots. Antigens were subjected to 12% sodium dodecyl sulphate-polyacrylamide gel (SDS-PAGE) in vertical gel electrophoresis (Mini-Protean II Cell, Bio-Rad) and electrotransferred (Trans-Blot SD Semi-Dry Transfer Cell, Bio-Rad) at 15 V for 90 min. Nitrocellulose (Hybond C⁺, Amersham, San Diego, CA, USA) was blocked with 5% skimmed milk in 1× TBST overnight. Blots were exposed to sera at a dilution of 1 : 100 in antibody dilution buffer with 2% BSA for 2 h, washed with 1× TBST, incubated with AP-conjugated antihuman immunoglobulin (Calbiochem) at 1 : 5000 dilution, developed with BCIP and NBT substrates (Bio-Rad AP colour reagents A and B). The sera of TBC patients were used as negative disease control and commercial antibody specific to blotted antigen was used as positive control to show the true binding of experimental sera antibodies with purified antigen.

Western blot analysis of antigens

The amount of p53 and TBP was detected by resolving various sera from patients and healthy donors on 12% SDS-PAGE at a dilution of 1 : 40. The blots of these samples were developed with anti-p53 monoclonal antibody (Oncogene Research Products) and anti-TBP polyclonal antibody as primary antibodies in 1 : 5000 using alkaline phosphatase substrates. These Western blots were scanned for density measurement using Quantity One software of the Gel Documentation 2000 System (Bio-Rad).

Radiolabelling of p53 and TBP

BL21 DE3 cells containing either the TBP plasmid construct or wild-type p53 gene construct were grown in M9 methionine-deficient minimal media and induced with 1 mm IPTG for 2 h at 30°C. The cells were incubated with 200 µg/ml rifampicin labelled with 15 µCi/ml cells [³⁵S]-methionine (specific activity 1000 Ci/mmol, Board of Radiation and Isotope Technology, India) and sonicated in lysis buffer. The expressed proteins were analysed on the basis of molecular weight on SDS-PAGE and autoradiography.

Immunoprecipitation

The immunoprecipitation assay for anti-TBP and anti-p53 was performed using [³⁵S]-methionine-labelled protein. Briefly, 10 µl of sera was mixed with 2 mg of protein A-Sepharose CL-4B (Pharmacia Inc., Piscataway, NJ, USA) in 500 µl of immunoprecipitation (IPP) buffer (10 mm Tris-Cl, pH 8·0; 500 mm NaCl; 0·1% NP40) and incubated with end-over-end rotation for 2 h at 4°C. The Sepharose particles with adsorbed IgG were washed four times in 500 µl of IPP buffer using a 10-s spin and resuspended in NET-2 buffer (50 mm Tris-Cl, pH 7·5; 150 mm NaCl; 0·05% NP40). Antibody-coated Sepharose beads were mixed with 400 µl of [³⁵S]-methionine-labelled extracts and rotated at 4°C for 2 h. After 4 washes with NET-2 buffer, the Sepharose beads were resuspended in SDS sample buffer. After boiling (90°C for 5 min), the proteins were fractionated on SDS-PAGE and dried, and labelled proteins were analysed by autoradiography.

Detection of free and immunocomplex forms of antigens

The crude sera of patients and control were adjusted at pH 8·0 by adding 1/10 vol of 1·0 m Tris (pH 8·0). Sera were passed through a protein-A agarose column with binding capacity of 10–20 mg of antibodies per millilitre of wet beads. Beads were washed with 10 column vols of 100 mm Tris (pH 8·0) and 10 mm Tris (pH 8·0). The fractions of flow-through were collected to detect the presence of free-form antigens. The columns were eluted with 100 mm glycine (pH 3·0) stepwise and eluates were collected in tubes containing 1/10 vols of 1 m Tris (pH 8·0) to neutralize the pH. These fractions, along with fractions of flow-through, were subjected to Western blot analysis to compare the free and immunocomplex forms of p53 and TBP antigens in patients and in control sera.

Statistical analysis

Statistical analysis was performed using the two-tailed test for difference between the two samples of normal versus patient with different means and variance. The difference between experimental groups was tested for significance using a t-test. The comparison of densitometrically assessed band areas above the threshold was performed for overexpressed p53 and TBP using Quantity One software (Bio-Rad). Results are expressed as mean density and standard deviation (s.d.). Simple regression analyses were performed using Sigma Plot software version 8·0.2. Correlation coefficients were calculated for anti-p53 and anti-TBP in different experimental groups of SSc, overlapping syndromes and SLE using Karl Pearson's correlation analysis.

RESULTS

Analysis of clinical features

The serum samples of 50 patients with SSc, 20 patients with SLE, 20 patients with overlapping syndromes, 25 normal donors and 10 disease controls with TBC were selected for the study. The subjects were 20 male (∼ 22%) and 70 female (∼ 77%) patients with a mean age of 38·6 years (range 21–60, s.d. ± 0·08). The mean disease duration was 8 years (range 4–17) at the time of blood collection. Eleven of 50 patients with SSc were male (22%) and the remaining 39 were female (78%). All 20 patients with OS were female (100%). Two of 20 patients with SLE (10%) were male and 18 of 20 patients were female (90%). Six of 50 patients (SSc) had localized symptoms, three were affected with morphea scleroderma (6%) and the remaining three had linear scleroderma (6%). Forty-four of 50 patients were progressively affected; 13 patients (26%) showed crest syndrome/limited scleroderma; 31 (62%) showed diffuse scleroderma. The diffuse scleroderma might overlap with other autoimmune diseases, including SLE and polymyositis. In these cases the disorder was referred to as MCTD. Of 20 patients with overlapping syndromes, nine were confirmed with MCTD (45%) utilizing the Varelisa RNP antibodies enzyme immunoassay kit (Pharmacia & Upjohn Diagnostics, Freiburg, Germany) for U1 snRNP (uridine rich small nuclear ribonucleoproteins) antibody. Among the control samples, there were 10 male and 15 female patients in the same age group with no identified autoimmune disease, whereas the disease control group comprised three male and seven female subjects. The data showing patient age, diagnoses, the corresponding levels of antigen and antibodies in serum are shown in Table 1.

Table 1.

The comprehensive table shows patient diagnoses, sex ratio, mean age, the levels of TBP and p53 antigens in sera and the concentration of autoantibodies

				p53 in sera			TBP in sera			Anti-p53 antibodies		Anti-TBP antibodies
Type	Population (n)	Sex ratio (m : f)	Mean age ± s.d. Mean disease duration (years)	Mean density ± s.d. Software	ELISA (mean)	Population	Mean density ± s.d. Software	ELISA (mean)	Population	Mean OD ± s.d.	Population	Mean OD ± s.d.	Population
SSc	50	14 (28%)	37 ± 0 (8)	1761·5 ± 6·8	31·5	14%	1023 ± 12·6	21·6	30%	1·702 ± 1·2	78%	1·622 ± 1·4	28%
		36 (72%)		(50×↑)			(10×↑)
SLE	20	1 (5%)	39 ± 0·008 (7·5)	421·5 ± 8·5	36·0	15%	421·5 ± 8·5	30·3	20%	1·581 ± 1·9	80%	1·451 ± 1·3	15%
		19 (95%)		(27×↑)			(4×↑)
OS/MCTD	20	0 (0%)	43·5 ± 0·006 (7·0)	241·9 ± 9·1	23·5	40%	940·9 ± 8·7	20·5	40%	1·414 ± 1·6	100%	1·467 ± 1·3	25%
		20 (100%)		(6×↑)			(9×↑)
Control	25	10 (40%)	40·6 ± 0·08	35 ± 2	4·9		96·3 ± 4	4·7		0·163 ± 0·4		0·152 ± 0·9
		15 (60%)
Disease control	10	3 (30%)	33·6 ± 17·4	36 ± 5	5·0		90·1 ± 3·6	4·8		0·158 ± 0·6		0·155 ± 0·8
		7 (70%)

The present study included several systemic autoimmune diseases such as SSc, OS and SLE. The assessment of disease activity in SSc and OS, including MCTD, is not well understood or validated. Hence, we did not attempt a correlation with disease activity. In view of the small number of patients in each category, it was considered that correlation with disease expression or organ involvement would not be statistically correct.

Over-expression of p53 and TBP antigens in patients with SSc, SLS and OS

The expression of p53 and TBP antigens in patients with SSc, OS, SLE and controls was studied by Western blotting and the expression levels of these antigens were determined by sandwich ELISA. The predominant bands of p53 and TBP were observed in sera of all patients who were scanned using the Bio-Rad Gel Documentation 2000 System. The density was measured in terms of intensity × area and calculated as mean values with standard deviations for all categories of patients, i.e. SSc, OS, SLE and controls (Fig. 1a). Almost all the patient sera under study showed the presence of TBP, whereas the control sera were negative. The relative comparisons of densitometrically assessed band areas showed a density value (intensity × mm²) for immunoreactive TBP in a range of 1081·5–4709·6 for SSc, 640·6–1603·4 for SLE and 122·8–161·7 for OS. These data are suggestive of a wide range of over-expression of TBP in autoimmune patients.

Fig. 1.

(a) Western blot analysis of p53 and TBP antigens in sera of C, DC and in patients with SSc, SLE and OS. The p53 and TBP bands from Western blot analysis were scanned by densitometry using the Bio-Rad Gel Documentation 2000 System. The density was measured in terms of intensity × area and calculated as mean values with standard deviation for patients and controls. Results are expressed as the mean density (intensity × mm²) ± s.d. (b) Quantitative determination of p53 and TBP antigen levels by sandwich ELISA. The serum p53 and TBP levels were determined by sandwich ELISA utilizing two different antibodies against each antigen and are presented in mg/dl. The data for the same six patients and controls as in (a) for p53 (a–f) and TBP (g–i) are shown. The mean values of antigen levels of six patients in each category of diseases are presented inside the box. When compared with C and DC, the mean p53 level was 6·4-fold higher in SSc, 7·3 in SLE and 4·7 in OS, whereas the mean TBP level was 4·9-fold higher in SSc, 6·4 in SLE and 4·4 in OS.

The densitometric-assessed areas for p53 showed a density value (intensity × mm²) range of 476–1904·9 for SSc, 115·1–741·7 for SLE and 477·3–1767·7 for OS. Most controls were negative for p53 and TBP expression, although a few showed low amounts of p53 in a range of 58–75·2. The p53 expression was higher in 14% patients with SSc (×50), 15% with SLE (×27) and in 40% with OS (×6). In a similar manner, expression of TBP was higher in 30% patients with SSc (×10), 20% with SLE (×4) and 40% with OS (×9).

Further, the levels of TBP and p53 antigens in sera were determined quantitatively (in mg/dl) by sandwich ELISA utilizing two different antibodies against each of the antigens. The antigen preparations were of high quality and they were free of IgG, as it was tested with antihuman IgG (Fig. 2a). The densitometric analysis (Fig. 1a) of the Western blots of different categories of patients and control sera was compared with the result obtained in sandwich ELISA (Fig. 1b). The mean serum antigen levels of TBP and p53 were 4·7 and 4·9 mg/dl in the C and DC populations, respectively. The mean p53 levels were 31·5 mg/dl in SSc, 36·0 mg/dl in SLE and 23·5 mg/dl in OS patients. The mean TBP levels were 21·6 mg/dl in SSc, 30·3 mg/dl in SLE, 20·5 mg/dl in OS patients and 4·7 mg/dl in controls. When compared to controls, the mean p53 levels were 6·4-fold higher in SSc, 7·3 in SLE and 4·7 in OS patients, whereas the mean TBP levels were 4·9-fold higher in SSc, 6·4 in SLE and 4·4 in OS. Although the densitometric assessment of the Western blots in patients was several-fold higher than controls, we considered that the sandwich ELISA method in detecting the antigen levels was more authentic.

Fig. 2.

(a) Immunoblots showing anti-p53 antibodies in sera. Lane 1, MW markers; lane 2, recombinant p53 as positive control as detected by Ab421; lane 3, p53 antigen preparation incubated with antihuman IgG; lane 4, C serum; lane 5, DC serum; lane 6, OS serum; lane 7, SLE serum; lane 8, SSc serum. (b) Autoradiogram of immunoprecipitated [³⁵S]-methionine-labelled antigens. An autoradiogram strip showing immunoprecipitated [³⁵S]-methionine-labelled p53 and TBP with normal and patient sera. Lane 1, MW marker; lane 2, p53 as positive control; lane 3, TBP as positive control; lane 4, C sera; lanes 5–8, patient sera; lane 9, DC sera; lanes 10–13, patient sera.

In order to demonstrate the true binding of serum antigens to antibodies, inhibition experiments were performed using both the antigen and the secondary antibody as the inhibitors. In the first experiment, the serum proteins were resolved upon SDS-PAGE and transferred onto nitrocellulose (NC). The primary antibody Ab421 against p53 antigen was preincubated with recombinant p53 and the generated complex was probed with the NC. Although a P53 band was detected in the serum of the SSc patient utilizing Ab421, no band was detected by probing the NC with the pregenerated complex of Ab421-rec p53, thus confirming the binding specificity. In the second inhibition experiment, after being transferred to the NC the serum proteins were incubated with the secondary antibody, which was not conjugated with horse radish peroxidase (HRP). By probing this NC with secondary antibody now conjugated with HRP, no colour was developed (Fig. 3a). Moreover, an ELISA inhibition assay confirmed the absolute binding between the antibody and sera antigens of interest.

Fig. 3.

(a) Inhibition experiments showing true binding on immunoblot. An inhibition experiment was performed utilizing both the antigen and the secondary antibody as inhibitors. Lane 1, MW markers; lane 2, positive control showing the binding of recombinant p53 to Ab421; lane 3, serum C; lane 4, serum DC; lane 5, serum SSc; lane 6, no band was detected when the SSc serum blot was incubated with pregenerated complex of Ab421-rec p53 complex; lane 7, no band was detected when the blot was incubated with non-conjugated secondary antibody followed by incubation with HRP-conjugated antibody. (b) Overexpressed p53 antigen is present as immune complexes. Western blot analysis showing an absence of free p53 antigen in the flow-through fractions from the protein-A-agarose column incubated with of control serum (lane 2) and in patient serum (lane 3). Released p53 antigen from the immune complexes under low pH in control (lane 4) as well as in patient (lane 5). Lane 1 shows recombinant p53 as positive control.

In order to determine whether the overexpressed p53 and TBP were present as free antigens or immune complexes, the serum was passed through protein-A agarose and the flow-through fractions were analysed on SDS-PAGE. No free antigens were visualized on the gel. However, under low pH conditions the antigens were released from the protein-A agarose column, thus suggesting that the antigens were present as immune complexes. In patients’ serum, the released antigens from the immunocomplexes was three to four times higher than the control serum (Fig. 3b).

Autoantibodies to p53 and TBP antigens are common features of SSc, SLS and OS

All sera of SSc, OS, SLE and controls were studied at a dosage range from 1/100 to 1/50 000 with twofold dilutions and other mean dilutions to determine the titre of autoantibodies keeping the test antigens constant. This study of dose-dependent maximum reactivity of anti-p53 and anti-TBP by ELISA showed the most stable binding pattern at a dilution range of sera between 1/100 and 1/500 (Fig. 4). We reached a titre of 1/333 with maximum cut-off values to detect autoantibodies in various samples. This dilution was used further to compare the patients’ samples with control.

Fig. 4.

Dose-dependent reactivity of anti53 and anti-TBP antibodies. ELISA showing the pattern of binding capacity of autoantibodies in sera over a dosage dilution range of 1/100, 1/500, 1/1000, 1/3000, 1/5000, 1/10 000, 1/20 000, 1/40 000 and 1/50 000 (1–8) (as shown by dots) for anti-p53 antibodies and over a range of 1/100, 1/500, 1/1000, 1/5000, 1/10 000, 1/20 000 and 1/40 000 (1–7) for anti-TBP antibodies. Each series represents Os, SSc, SLE and C. The maximum reactivity of anti-p53 and anti-TBP antibodies by ELISA showed the most stable binding pattern at a dilution range of sera between 1/100 and 1/500.

The presence of antibodies to p53 was detected by ELISA in 39 of 50 (78%) subjects with SSc, in 20 of 20 (100%) subjects with OS and in 16 of 20 subjects (80%) with SLE. The presence of antibodies to TBP was observed in 14 of 50 (28%) subjects with SSc, in five of 20 (25%) subjects with OS and in three of 20 (15%) in subjects with SLE. The individual antibodies level to p53 was comparable in patients with SLE (P = 0·00289) and OS (P = 0·00279), but both were much higher than control samples. Most of the OS patients showed a tendency to remain at the highest side of the OD graph thus indicating, statistically, the common presence of high anti-p53. However, the SSc group sera manifested significantly higher titres of IgG antibodies to p53 (P = 0·00001). The antibody levels of anti-TBP was higher in patients with OS (0·00185) than patients with SLE (0·00673) and SSc (0·00986) (Fig. 5). SSc and SLE samples showed considerably higher amounts of anti-TBP than controls. The cut-off values at OD units > 0·8 for p53 and OD units > 0·9 for TBP were considered as positive samples in comparison to both controls and disease controls. The serum level of anti-p53 showed a low degree of negative correlation with anti-TBP (Fig. 6). The values of Pearson's r were 0·10 for patients with SSc, 0·21 for patients with OS and 0·17 for patients with SLE.

Fig. 5.

Determination of levels of autoantibodies by ELISA. The levels of anti-p53 and anti-TBP antibodies were determined by ELISA in C (n = 25), DC (n = 10), SSc (n = 50), OS (n = 20) and SLE (n = 20). The cut-off for positivity is shown as a dotted line. Optical density (OD) units above 0·8 (anti-p53) and above 0·9 (anti-TBP) were considered positive.

Fig. 6.

Scatter-graphs comparing levels of anti TBP and anti-p53 antibodies. The scatter-plots were shown for 50 SLE, 20 OS and 20 SSc sera. The correlation coefficients (r) were depicted for each experimental group: SSc (0·10), OS (0·21) and SLE (0·17).

The baculovirus-expressed, antibody-purified recombinant p53 was utilized as antigen to confirm the presence of anti-p53 antibodies in the sera of patients and controls by Western blotting. The same amount of protein was also checked with commercial anti-p53 antibodies as positive control. The result showed the presence of anti-p53 antibodies in patients’ sera in comparison to controls, in which no anti-p53 antibodies were detected in controls (Fig. 2a). To confirm there were no immunoglobulins in the antigen preparation, the p53 antigen preparation was mixed with antihuman (IgG/IgM/IgA) conjugated with HRP and was subjected to colour development. No colour was detected, thus suggesting that there were no immunoglobulins in the antigen preparation (Fig. 2a). This result was confirmed further by immunoprecipitation of [³⁵S]-methionine-labelled antigens by the patients’ sera. Eighty-three per cent showed strong positive signals for anti-p53 antibodies and 22% for anti-TBP antibodies, as represented by four patient sera and a single control (Fig. 2b). Anti-p53 antibodies present in the sera immunoprecipitated the radiolabelled p53 protein. In a similar manner, anti-TBP antibodies in sera could immunoprecipitate the radiolabelled TBP protein. In contrast to the findings with anti-p53 and anti-TBP, the control sera did not immunoprecipitate these labelled antigens. The bands of 53 and 40 kDa were confirmed further utilizing commercial anti-p53 and anti-TBP antibodies which migrated at the same position.

DISCUSSION

In this study, we have investigated the expression levels of p53 and TBP antigens in patients with SSc, SLE and overlap syndromes as well as in healthy and disease controls. We preferred Western blot analysis over immunohistological techniques and flow cytometry because of the high sensitivity and reliability of this method [14]. The expression levels of TBP and p53 antigens in patients with SSc, SLE and OS, as determined by sandwich ELISA, varied between four- and sevenfold higher than the controls. However, antigen levels varied between four and 50 times higher by densitometric analysis of the Western blots in 40% of the patients with OS, 14–30% of SSc patients and in 15–20% of SLE patients. We assumed that the detected antigens in serum using sandwich ELISA might be present as free antigens. The observed levels of antigens in higher ratios, by densitometric analysis of the Western blots, could be due to the presence of antigens both as free and as immune complexes. The expression of p53 has been reported previously in the synovial tissue of patients with early and long-standing rheumatoid arthritis [18,31]. It was suggested that this phenomenon was probably secondary to the increased production of wild-type p53 protein in response to DNA damage and secondary to somatic mutations caused by the genotoxic local environment.

SSc is characterized by activation of fibroblast endothaliocytes and lymphocytes and the expression of p53 was observed in the skin of a limited number of SSc patients [32], although an absence of p53 gene mutations in skin fibroblasts was observed in SSc patients [33]. Unlike normal skin, skin lesions of cutaneous lupus (CLE) and dermatomyositis showed the expression of p53 protein [34]. Our results showed that both the nuclear antigens p53 and TBP in the patient sera were not present as free antigens. The flow-through fractions from the protein-A agarose column did not show either of the free antigens, thus suggesting that IgG type antibodies were present in an excess amount in the patients’ serum.

We have shown by ELISA and radio-immunoprecipitation that anti-p53 and anti-TBP antibodies were present only in the serum of patients and not in healthy controls. In ELISA, the OD units of anti-p53 and anti-TBP autoantibodies were more than 2·5 times higher than the OD units of the control individuals. Radio-immunoprecipitation showed that the presence of an excess amount of autoantibodies in the patient's serum could form immunocomplexes with the radiolabelled antigens. The autoantibodies to p53 were titrated to reach an optimal dilution of sera, which was utilized to detect the contrasting presence of anti-p53 in OS than in controls. The patients with OS showed the presence of high amounts of autoantibodies to p53. Similarly, the level of anti-p53 antibodies was high in SLE patients and higher in SSc patients than the controls of the same age groups, including both sexes. The patients with OS showed higher titres of anti-TBP antibodies than patients with SLE and SSc in our studies. The majority of sera which gave positive ELISA for anti-p53 and anti-TBP also immunoprecipitated the radio-labelled antigens and bind to the recombinant antigens in immunoblots. The Pearson coefficient of correlation showed a low degree of negative correlation between the two autoantibodies.

Elevated levels of anti-p53 autoantibodies has been detected in patients with Graves’ disease and immune vasculitis, including Wegener's granulomatosis [35,36], and the overall prevalence of its level (23%) was comparable to that of various cancers, although patients with autoimmune disease face an increased risk for malignancies. It is interesting to note that anti-p53 antibodies were detected rarely in the serum and synovial fluid of patients with rheumatoid arthritis and Sjögren's syndrome [22], although over-expression of p53 was noted in these patients [18,31]. Further, an absence of anti-p53 antibodies was noted in Chinese patients with RA and SLE [23], although antibodies against p53 were observed in Caucasian SLE patients [20].

We have demonstrated that an over-expression of p53 and TBP as well as the presence of high levels of autoantibodies to these antigens were a common feature in heterogeneous Asian Indian patients with SSc, OS and SLE. At the same time, we failed to detect, by protein-A agarose affinity analysis, any of the free p53 and TBP antigens in the patient sera. The over-expression of these antigens and the presence of autoantibodies to these antigens might be two different phenomena independent of each other. We propose that the over-expression of these antigens could be due to hyperactive regulatory regions of TBP and p53 gene that might arise due to somatic mutations. Abundant somatic mutations have been observed in the coding region of the p53 gene in RA synovial tissue and they are probably the result of intense local chronic inflammation [15,16,32], although an absence of the p53 mutation was observed in systemic sclerosis [33].

It was suggested that the over-expression of p53 along with its transcription target p21 might be a defensive mechanism to repair DNA damage and prevent apoptosis [37]. Excessive keratinocyte apoptosis occurs in CLE and dermatomyositis [34] that could release the fragmented autoantigens into dermis. The over-expression of nuclear DNA binding proteins such as TBP and p53 and RNA polymerases [38] of the transcription machinery in patients with autoimmune disorders is interesting, as these proteins were also found to be physically associated [25]. We are currently involved in establishing a possible link between the over-expression of these antigens and the high level of antibodies against these antigens.