Identification Of Serum Biomarkers For Systemic Lupus Erythematosus Using Human Proteome Microarrays

BackgroundSystemic lupus erythematosus(SLE)is a complicated chronic autoimmune disease with multiple organs damage and unpredictable disease progress.Its diversity of clinical features is caused by the interaction of genetic,hormonal,and environmental factors,and the array of autoantibodies with which it is associated not only helped to diagnose,but they are also associated with certain clinical manifestations and subsets of the disease.Only anti-nuclear(ANA),anti-Sm,anti-double-stranded DNA(anti-dsDNA)and antiphospholipid antibodies were included in the American college of rheumatology(ACR)classification criteria for SLE,besides,anti-Ro,anti-La,and antiribonucleoprotein(RNP)are always frequent elements of the serologic profile in SLE.About 20%SLE patients may lose ANA positivity over time and anti-dsDNA and complements fluctuate with disease activity and treatment.Renal injury was the major cause for morbidity and mortality which commonly named lupus nephritis(LN)and renal biopsy was still the golden standard of renal injury,but the technic was less repeatable and with unpredicted damage.SLE patients those who complicated with severe thrombocytopenia may die of intracranial hemorrhage or gastrointestinal bleeding and it was an independent risk factor for the poor prognosis.Serum antibody can be tested before any specific clinical manifestations and/or iconography abnormity,well antibody biomarkers are still absent and the monitoring plan for early suspected serum changes of having SLE was never recommended.The complexity of SLE lay in the variety of serology and immunology indexes,also,kinds of clinical manifestations which leading following atypical SLE symptoms always made it difficult to give an accurate diagnosis.Because of the concealment and uncommon of SLE situation,even parts of SLE subjects were irresponsible to traditional immunization therapy,these all were big challenges in the practical job whether in treatment or diagnosis.There were over 100 autoantibodies have been founded associated with SLE,but we are still exploring appropriate autoantibodies with both sensitivity and specificity.Protein microarrays,an emerging class of proteomic technologies,can analysis a large amount of antigen-antibody reactions at the same time with very few sample even can guarantee the quality of sensitivity and specificity,they are fast becoming critical tools in biochemistry and molecular biology and have been applied in various researches for serum biomarkers exploration.HuProt~TMM protein microarray is equipped with the most numerous protein containing and this study would like to analysis the profile of SLE serum antibody with this array.ObjectivesThe study was aimed to screen antibodies with abnormal expression level and give the further validation for the candidate antibodies.To evaluate the proper antibodies individually and identify a panel of antibodies that may serve as effective biomarkers for auxiliary diagnosis of SLE.MethodsThe project was conducted with two stages to explore the antibody variation of serum in SLE patients from the molecular epidemiology dimension using case-control study pattern.Stage one:A exploration experiment with small samples.We chose 15 SLE patients and 5 healthy controls and one serum matched with one protein microarray composed of about 20000 full-length unique proteins to screen the different antibodies between two groups preliminary.Stage two:A validation test with large samples.We costumed the microarray basing on the potential biomarkers list calculated from the results of the first stage.110 SLE patients,30 rheumatoid arthritis(RA),30 primary Sjogren's syndrome(pSS)and 106healthy controls were included in the second stage.Receiver operating characteristic(ROC)curves analysis was performed on training phase,testing phase and combined phase.A discrimination model of antibodies panel was deduced to predict the risk of diagnosis with SLE by stepwise Logistic regression based on the training set and validated the formula in the testing set.Finally,for a best model with the help of ROC curves.ResultsThe first stage:The study firstly established the profiles of differential expression antibodies in the serum of SLE patients by the HuProt~TMM protein microarrays that incorporating about 20000 purified proteins.81 proteins were included into the next stage according to the P value and fold change(P<0.05 and fold change>1.5)eventually.The second stage:35 IgG(RPLP2,RPLP1,RPLP0,SNRPC,PARP1,HK1,TROVE2,SRSF7,HIST1H1C,SRSF6,H1F0,MAK16,MAK16,RPL23A,JHU04032,USF1,MAK16,RPL14,RPL35,NFATC2,RPL7A,SP5,RPL14,FOXC2,RELL1,TBX1,PIK3CG,JUND,CDK19,GABRA4,DBP,NOC3L,LLPH,DGCR8,MAST3)were significantly increased in SLE group compared with healthy controls(all P<0.05).The area under curve(AUC)of RPLP0-IgG,RPLP1-IgG and RPLP2-IgG antibody exceeded 0.9 respectively for differentiating SLE patients and healthy subjects.14 IgG(RPLP2,RPLP1,RPLP0,MAK16,RPL23A,MAK16,JHU04032,PARP1,RPL35,MAK16,RPL7A,RPL14,DCX,SRSF6)were significantly increased in SLE patients compared with disease controls(all P<0.05).Totally,if we choose a single protein as a biomarker,it always shows a high sensitivity and a common specificity.The discrimination model was constructed to predict the risk of diagnosis with SLE by stepwise Logistic regression as follows:logit(P=SLE)=-1.826+0.014RPLP2-IgG+0.027PARP1-IgG+0.564MAK16-IgG-0.881RPL7A-IgG,and the AUC of these four-antibody panel was 0.979 with a nice predictive ability.Conclusions1.Compared with healthy controls,the serum IgG levels of RPLP2?RPLP1?RPLP0?SNRPC?PARP1?HK1?TROVE2?SRSF7?HIST1H1C?SRSF6?H1F0?MAK16?MAK16?RPL23A?JHU04032?USF1?MAK16?RPL14?RPL35?NFATC2?RPL7A?SP5?RPL14?FOXC2?RELL1?TBX1?PIK3CG?JUND?CDK19?GABRA4?DBP?NOC3L?LLPH?DGCR8?MAST3 were significantly increased.The serum IgG of RPLP0?RPLP1?RPLP2?SNRPC?PARP1?HK1?TROVE2?HIST1H1C?SRSF6?MAK16 were all equipped with good power for prognosing risk and disease discrimination.2.The four-antibody(RPLP2-IgG,PARP1-IgG,MAK16-IgG and RPL7A-IgG)in serum may serve as potential novel biomarkers for SLE.3.A further ELISA experiment based on the antibodies mentioned above can validate the result with microarray proteins and the appropriate bioinformatics analysis would provide some useful clues for SLE mechanism research.