Serum Proteomic Fingerprints Of Severe Hepatitis

In China, because of the high prevalence of hepatitis B, severe hepatitis is the most common cause for liver failure. Severe hepatitis is characterized by massive denaturation, necrosis and dysfunction of the liver, resulting in a serious condition with a high rate of mortality.It is known that the clinical manifestations of severe hepatitis are associated with impaired synthesis of protein, metabolic disturbance and accumulation of toxins. Regarding the development of hepatic encephalopathy, accumulation of ammonia, aromatic amino acid, short-chain fatty acid, mercaptans, endogenous benzodiazepine, phenols, false neurotransmitters, r-aminobutyric acid, et al have all been implicated. Furthermore, the elevation of some abnormal peaks was observed by chromatogram technology in the serum of patients with severe hepatitis about 30 years ago;and it was suspected that they were peptides with molecule weight between 500 Da to 5000 Da (also named middle molecules) but its exact nature is unknown.The treatment of severe hepatitis can be classified into three categories: routine medical therapy, liver support system and liver transplantation. The goals of liver support system are the removal of toxins and to replace the functions of the liver, and it has become one of the main therapies for severe hepatitis in China. To date, liver support system has not replaced the main functions of the liver;further refinements in the technology will undoubtedly facilitate significant improvement in the outcomes of severe hepatitis. Because of donor shortage, liver transplantation is mainly performed for patients whose prognosisis considered very poor. Accurate prediction of prognosis is very important for the management of patients.SELDI-TOF-MS (Surface Enhanced Laser Desorption Ionization-Time Of Flight-Mass Spectrometry) is one of the recently developed sophisticated technologies, which, based on capturing protein/peptides by chemically modified surface, is specifically powerfully for analyzing the complex biological samples.SELDI-TOF-MS offers a sensitive and high-throughput technology for the profiling of low molecular weight proteins. The SELDI approach has been successfully used to identify serum biomarkers in ovarian cancer, hepatocellular carcinoma, and infectious diseases et al. These results suggest that the low molecular weight serum proteome contains an unexplored archive of histological information and provide useful biomarkers for disease detection.SELDI-TOF-MS presents a new opportunity to examine the serum proteomic fingerprints of severe hepatitis, thus may provide new insights into the pathophysiology, diagnosis and treatment of severe hepatitis.Materials and MethodsTo determine the best conditions for identifying the most discriminating serum protein profiles between control group and severe hepatitis group, serum samples were compared to plasma samples and various chips were initially evaluated.54 patients with chronic hepatitis B (28 of 54 were severe hepatitis patients) and 22 healthy controls were enrolled in this study. Serum samples were analyzed with the SELDI-TOF-MS (Ciphergen Biosystems, USA) to obtain a quantitative proteomic fingerprints with molecular masses ranging from 1 KDa to 10 KDa.The correlation between serum proteomic fingerprints and liverfunction impairment was examined. The discriminating peaks of severe hepatitis were identified. Proteomic fingerprinting models for predicting prognosis in patients with severe hepatitis were developed with bioinformatics.ResultsCompare to plasma, serum of severe hepatitis patients can provide more discriminating profiles. Various chip chemistries (hydrophobic, anionic, cationic, and metal binding) were evaluated to which affinity chemistry provided the best serum profiles in terms of number and resolution of proteins. The Q10 chip (a kind of strong anion exchange chip) was observed to give the best results.Serum proteins from 54 patients with chronic hepatitis B were profiled on SELDI-TOF-MS, and the liver function impairments were evaluated by model for end-stage liver disease (MELD). We analyzed the correlation between peak intensities and MELD scores by use of the Spearman rank-order correlation test. 57 peaks including M/Z 1065n 1224% 1237 were correlated with MELD score (p<0. 000001).By comparing proteomic fingerprints of severe hepatitis serum to control serum, 59 discriminating peaks including M/Z 1713% 6634% 1537 were identified(/K1X10*5) ? Application of Principal components analysis (PCA) to SELDI-TOF-MS data distinguished control samples and severe hepatitis samples clearly. The discriminating peaks were also used to construct discriminant partial least squares (DPLS) and Artificial Neural Networks (ANN) models. Cross validation showed that both DPLS and ANN models correctly classified the control samples and severe hepatitis samples.28 cases of severe hepatitis were divided into 2 subgroups: survival group and die group. Comparing proteomic fingerprints of severe hepatitis serum to control serum, 38 proteomic features were observed to besignificantly different between two groups (p<0. 05). DPLS and ANN were used to construct a model for predicting prognosis. Cross-validation showed that the accuracy for prediction of prognosis was 85. 7~89. 3%. The relative importance of each input variable in the predictive model was calculated, and the 5 most important variables were M/Z 4099, 5199, 3963, 5358 and 4969.ConclusionOur study applied for the first time SELDI-TOF-MS to the study of severe hepatitis (liver failure). A unique serum proteomic fingerprint was present in the sera of patients with severe hepatitis. Predictive models derived from the serum proteomic fingerprint could predict the prognosis of severe hepatitis with the accuracy of 85.7—89.3%.The exact nature of the discriminating proteins or peptides is under investigation. Increasing number of cases will be required to further evaluate the predictive models for severe hepatitis.