Identification Of Antibodies Against HPAEC And Its Significance In Vascular Damage

ObjectiveSystemic connective tissue diseases (CTD) such as systemic lupus erythematosus (SLE), systemic sclerosis (SSc) and polymyositis (PM) are frequently complicated with pulmonary vascular damage. Part of the patients may eventually develop into pulmonary arterial hypertension (PAH). Several reports strongly suggest that PAH is frequently observed in patients with mixed connective tissue disease (MCTD), which has the clinical features of the overlap of SLE, SSc, and PM, along with high titers of anti-U1 ribonucleoprotein antibodies (anti-U1RNP Ab). It is reported that nearly half of the death of patients with MCTD is due to PAH. PAH is one of the most serious complications in patients with MCTD, which shows remarkable vascular damages from large to small vessels in the lungs. In patients with MCTD or anti-U1RNP Ab-positive patients, development of PAH has been recognized as one of the important life-threatening factors. In vitro experiments have demonstrated that autoantibodies derived from patients with connective tissue diseases could induce up-regulation of adhesion molecules on human pulmonary artery endothelial cell (HPAEC), suggesting that immune-mediated inflammatory process may be involved in PAH. Whereas the studies focusing on the pathogenesis of PAH in patients with MCTD are increasing, the mechanism of autoantibody-mediated pulmonary artery damages remains unknown.The snRNPs (small nuclear ribonucleoproteins) are a group of intranuclear particles that are comprised of a small RNA and several polypeptides. All snRNPs are located within the nucleoplasm in a uniform speckled pattern and are mainly involved in the splicing process of premessenger RNA. The most abundant snRNP is U1RNP. All snRNPs share a common SmRNP core and are immunoprecipitable with anti-Sm antibodies. Antibodies against U1RNP are frequently high in titres for patients with MCTD and can up-regulate intracellular adhension molecule-1 (ICAM-1), endothelial leucocyte adhension molecule-1 (ELAM-1) and classâ…¡MHC molecule expression on HPAEC and play important roles in the immunopathological processes leading to pulmonary arterial vasculopathy in CTD patients with pulmonary hypertension.Apart from anti-RNP Ab, anti-endothelial cell antibodies (AECA) are frequently found in sera from patients with systemic autoimmune diseases. They have also been implicated as a cause of arterial cell injury. The titers of AECA in sera derived from patient with MCTD are increased, especially in those in active stage and those with PAH. The binding of antibodies to HPAEC may induce the activation of HPAEC by up-regulation of adhesion molecules and the synthesis of cytokines and chemokines. RANTES (regulated upon activation, normal T cell expressed and secreted) is an important chemoattractant for monocyte and T-cells. RANTES may play a critical role in PAH through the induction of endothelin converting enzyme-1 and endothelin-1, a potent endothelium-derived factor with strong vasoconstrictive and mitogenic action. Exprimnets showed that mRNA of RANTES was more increased in the lungs of PAH patients than controls, and endothelial cells were the major source of RANTES. Endothelial cells have significant proinflammatory activities, amplifying and perpetuating the inflammatory process and contributing to vessel regeneration and repair.It is, therefore, reasonable to suppose that autoantibodies to HPAEC might play a role in the pathogenesis of PAH in patients with MCTD and might be an indicator of the prognosis of the disease. To investigate the prevalence of anti-HPAEC Ab in anti-U1RNP Ab-positive sera from patients with systemic autoimmune diseases and elucidate the possible anti-HPAEC Ab-mediated pulmonary artery endothelial dysfunction, we are to identify antibodies against HPAEC in sera from patients with positive anti-U1RNP antibodies and investigate the correlation of anti-HPAEC Ab with anti-U1RNP Ab. Further, by detection of the level of RANTES secreted by HPAEC upon the addition of anti-HPAEC Ab, the possible role of the antibodies in the mechanisms of pulmonary vascular damage is to be analysed.Methods1. Patients and seraSerum samples were collected with informed consent from patients with systemic autoimmune diseases, who had undergone Kyoto University Hospital. A total of 64 anti-U1RNP Ab-positive serum samples and 2 anti-U1RNP Ab-negative sera were collected: 31 from patients with SLE, 14 from patients with MCTD, 3 from SSc, 2 from rheumatoid arthritis (RA), 2 from Sjogren's syndrom, 1 from PM, 1 from undifferentiated connective tissue disease, 1 from discoid lupus erythematosus, and the other 9 from patients with no definite diagnosis. Sera from two healthy donors and two of anti-U1RNP Ab-negative patients with PAH were used as controls. All sera were kept frozen at-20℃until use.2. RNA-immunoprecipitation assay (RNA-IPP)Anti-U1RNP Ab in the collected sera was determined by RNA-IPP. Initially, 10μl of patient sera were incubated with 2 mg of protein A cepharose CL-4B preswollen in 500μl of IPP buffer (10 mM Tris-HCl, 500 mM NaCl, and 0.1% Nonidet P-40, pH 8.0) for 2h at 4℃in a microcentrifuge tube with end-over-end rotation. The antibody-coated beads were then washed three times with 500μl of IPP buffer. Antigens were prepared from HeLa cells (6×10⁶ cells for each sample). After sonication of HeLa cells in NET-2 buffer (0.3 ml for each sample) for 40 seconds for three times, supernatant of cell extract from cell lysate were combined with antibody-coated beads and 400μl of NET-2 buffer and incubated for 2h at 4℃. Following three washes with NET-2 buffer, the beads were resuspended in 270μl of NET-2 buffer, and 30μl of 3M sodium acetate, 15μl of 20% SDS. Then, the bound nucleic acids were extracted with phenol/chloroform/isoamyl alcohol (50:50:1) containing 0.1%8-hydroxyquinoline, precipitated with 3 volume of ethanol, and separated electrophoretically on a 10% polyacrylamide 7M urea gel. After electrophoresis and silver stain, the gels were photographed.3. Preparation of cell lysates and immunoblottingHPAEC were cultured with medium for vascular endothelial cells (HuMedia-EG2) until use according to the manufacturer's protocol. Trypsin/EDTA solution was used to detach HPAEC from the inner wall of culture flasks, and neutralizing solution was used to decrease the effect of trypsin on the separated cells. After three washes with phosphate buffer solution (PBS), HPAEC were lysised with lysis buffer (10mM Tris-Hcl, 150mM NaCl, 1mM EDTA, 1mM EGTA, 1% Nonidet P-40, 0.2%SDS, 10mM NaF, 2mM Na₂VO₄) and disposed with SDS sample buffer. Cell lysate includes 40,000 per 20μl. HeLa cells were also prepared as control. Cell lysates were stored at-80℃until use.Cell lysates dissolved in SDS sample buffer were boiled and fractionated electrophoretically in a 10% SDS-polyacrylamide gel and transferred to a nitrocellulose filter. Both HPAEC and HeLa cell lysates were used for screening of autoantibodies in each serum sample. For each well, 20μl (cell number was 40,000) of lysates were loaded. For blocking, the nitrocellulose filter was incubated in 5% skim milk/PBS at 4℃overnight. The filter was washed three times with PBS-NP40 (PBS with 0.05% NP40) and incubated with patient serum (diluted to 1:100 in PBS) for 3 h at room temperature. After three washes with PBS-NP40, the filters were incubated with goat anti-human IgG (H+L) in PBS for 2h. After three washes with PBS-NP40, IgG-bound proteins were visualized by BCIP/NBT color development substrate in alkaline phosphatase buffer. All incubation steps involving filters were performed in sealable polyethylene bags.In elution blot, autoantibodies bound to antigens of HPAEC on NC filters were eluted by using ImmunoPure IgG elution buffer. 3 ml of elution buffer was added to a cut of filter with desirable band. After voltex for 5 min, the eluate was collected. Repeat the same step with another 3 ml of elution buffer, and 800μl of 1M Tris-HCl (pH 7.5) were added to adjust pH to 7. The solution containing eluted IgG was used for reaction with NC filters with both HPAEC and HeLa cell lysates.4. Flow cytometric analysisTo clarify whether the autoantigens were expressed on the cell surface or not, flow cytometric analysis was performed. HPAEC were detached from culture flasks using 0.01%-EDTA/PBS without trypsin. Following three washes with 2% FCS/PBS, cells were incubated with 2μl of each patient serum for at 4℃for 1h. After washing with 2%-FCS/PBS, 2μl of FITC-conjugated anti-human IgG were added into each sample and incubate at 4℃for lh. After three washes, the samples were analyzed by CellQuest software. HeLa cells were treated as controls in the same way.5. Measurement of chemokine RANTES secreted from HPAECPatient IgG was purified from sera by ImmunoPure IgG purification kit. Purified IgG (final concentration=1mg/dl) was added to culture medium of HPAEC. After 48 hs, the culture supernatant was collected and concentration of RANTES was detected by ELISA.Results1. Identification of autoantibodies against HPAECAmong 64 of anti-U1-RNP Ab-positive sera determined by RNA-IPP, IgG antibodies reacting with 31kDa protein (HPA31) of HPAEC lysates were detected in 37 cases (57.8%). When HeLa cell lysates were used, anti-HPA31 Ab was seen in only 6 cases (9.4%). In 37 sera with anti-HPA31 Ab in HPAEC lysates, however, autoantibodies against two proteins of 30/32kDa in HeLa cell lysates were observed in 33 cases (89%). Among the 37 sera with anti-HPA31 Ab, anti-U2RNP and anti-Sm Ab were positive in 11 (30%) and 7 (19%) sera, respectively. For patients with PAH, 3 of the 4 anti-U1RNP Ab-positive sera (75%) and none of the 2 anti-U1RNP Ab-negative sera had anti-HPA31 Ab. In all of the serum samples with anti-UlRNP Ab, anti-UlRNP-B7B Ab were positive in 27 out of 37 cases with anti-HPA31 Ab (73.0%) and in 6 of 27 without anti-HPA31 Ab (22.2%) (P＜0.005). In anti-HPA31 Ab-positive group, autoantibodies to U1RNP-A and U1RNP-70K were detected in 25 (67.6%) and 14 (37.8%), respectively. The frequency of anti-HPA31 Ab recognized together with anti-UlRNP-A Ab was markedly higher than that of anti-HPA31 Ab-positive sera without anti-UlRNP-A Ab (P＜0.01). As for both anti-U1RNP Ab-negative sera and healthy donors, no antibodies to U1RNP-B'/B, A and 70K were detected. Moreover, there is another antoantibodies against 90kDa (anti-HPA90 Ab) that could be seen in HPAEC lysates in 14 cases out of 37 cases with positive anti-HPA31 Ab, but anti-HPA90 Ab was not detected in sera with negative anti-HPA31 Ab (P＜0.001).2. Elution blotting of anti-HPA31 Ab and anti-HPA90 AbAntibodies against HPA31 and HPA90 were eluted from NC filters and elution blotting was performed. Anti-HPA31 Ab recognized not only HPA31 but also HPA90 in HPAEC lysates, without reacting with any other antigens in both lysates. Eluted anti-HPA90 Ab failed to react with 90kDa antigen either in HPAEC or in Hela cell lysates, but react with 31kDa protein in HPAEC lysates and 50kDa, 32kDa and 30kDa in HeLa cell lysates, which indicated the cross-reaction between Anti-HPA31 Ab and Anti-HPA90 Ab, and that there might be some correlation among HPA90, 50kDa, 32kDa, 31kDa and 30kDa autoantigens.3. Flow cytometric analysisAnti-HPA31 Ab appeared to bind antigens on the surface of HPAEC. In contrast, such antibodies were not detected in sera derived from anti-HPA31 antibody-negative patients. On the other hand, anti-HPA31 Ab-positive sera failed to bind the surface of HeLa cells.4. Increased RANTES secretion from HPAEC by anti-HPA31 AbThe average concentration of RANTES in culture supernatant of HPAEC with the addition of IgG purified from anti-HPA31 Ab-positive sera was 23.1pg/ml, and for HPA31-negative cases is 13.1pg/ml. This result indicates that secretion of RANTES may be stimulated by anti-HPA31 Ab (P＜0.05).Conclusions1.IgG antibodies in sera derived from patients with connective tissue diseases recognize antigens of 31kDa (HPA31) and 90kDa (HPA90) in HPAEC lysates.2.The positive detection of anti-HPA31 Ab is associated with that of anti-U1RNP Ab,anti-U1RNP-B'/B Ab and anti-U1RNP-A Ab, but not with anti-U1RNP-C Ab,anti-U2RNP Ab and anti-SmRNP Ab.3.Anti-HPA31 Ab and anti-HPA 90 Ab react with antigens of each other by cross reaction, but both do not recognize U1RNP-B'/B,U1RNP-A and U1RNP-C Ag. Anti-HPAEC Ab is not associated with anti-U1RNP Ab in structure.4.The epitope of HPA31 Ag is located on the surface of HPAEC.5.Anti-HPAEC Abs can promote the secretion of chemokine RANTES from HPAEC by binding with antigens on HPAEC surface.6.Increased production of RANTES is a possible mechanism of AECA-induced pulmonary vascular injury.