| Hepatocellular carcinoma (HCC) was apt to invade portal vein to cause tumor thrombus (PVTT), while PVTT could cause the intrahepatic dissemination and metastasis of HCC. According to the development of tumor thrombus, PVTT could be divided into two stages: microscopic portal vein tumor thrombus (Mi-PVTT) and macroscopic portal vein tumor thrombus (Ma-PVTT). To our knowledge, there were a few reports available about Mi-PVTT, the biological characteristics of Mi-PVTT was not well-known, whether could it be an indicator of metastasis was not clear, till now there were no reliable approach to early detect and predict the occurence of PVTT, especially serum biomarker. Because the formation of PVTT was a multistep and multivariate process, researches based on univariate and single gene level could not elucidate the molecular mechanisms of PVTT totally, which must be based on the research of high-throughput genomics and proteomics.Due to the attributes of surface enhanced laser desorption /ionization time-of-flight mass spectrometry (SELDI-TOF-MS) in providing rapid proteomic profiling of proteins/peptides from different biological specimens, the technology had been applied extensively in cancer research. The aim of this study was to screen biomarkers with relation to the formation of PVTT in tumor tissue and serum, also investigate the relationship between tissue and serum proteomics for elucidating the molecular mechanisms of the formation of PVTT, Meanwhile, the effects of Clinicopathologic variables on serum proteome were evaluated, in particular the comparison of serum proteome in HCC with different stage of PVTT, based on above findings, a clinical diagnostic and predictive proteomic pattern of PVTT was established.Part One: Establishment of the platform of protein chip SELDI-TOF- MSThe objective of this part of research was to establish a reliable protein chip SELDI-TOF-MS platform and to optimize the technological conditions in the experiments. Serum and cancer cell lysate samples from 8 cases of HCC patients were collected respectively, special protein or peptide spectra was determined by SELDI-TOF-MS measurement after treating the sample onto four types of ProteinChips(WCX2, H^ SAX2, IMAC3) for each case. According to the protein profiling results, suitable parameters and protein chips in serum and cell lysate were screened respectively, and the reliability and stability of the technology were assessed.RESULTS:1. Most of the proteins detected by protein chip range from 1100 to 30000 at the Mass/charge (m/z) value.2. Analysis of serum protein fingerprints showed WCX2 Chip combined the most proteins, followed by IMAC3 Chip, and less in H4 and SAX2 Chips.3. Analysis of cell lysate protein fingerprints showed WCX2 Chip combined the most proteins, followed by H4 Chip, and less in IMAC3 and SAX2 Chips.4. The coefficient variation (CV) of protein intensity and m/z value was 0.184731 and 0.000242, respectively, within the same protein chip.5. The coefficient variation (CV) of protein intensity and m/z value was 0.214244 and 0.000055, respectively, between the protein chips.Part Two: Screen tissue biomarkers with relation to the formation of PVTTinHCCThe objective of this part of research was to screen tissue biomarkers with relation to the formation of PVTT in HCC and supply the research basis of the molecular mechanisms involved in formation of PVTT. A cohort of tissue samples were obtained from 13 HCC patients with macroscopic PVTT including primary tumor(P-A) and tumor thrombus(P-C), also 13 HCC patients without PVTT including only primary tumor(NP-A). After tumor cells extracted from the tumor tissue were lysated, special protein or peptide spectra was determined by SELDI-TOF-MS measurement after treating the sample onto two types of Protein Chips(WCX2> H4) for each case. The protein fingerprints between groups of P-A and NP-A as well as P-A and P-C were compared by Biomarker Wizard software package.RESULTS:1. According to the protein fingerprints in WCX2 protein chip, 34 significant differential proteins were found between P-A and NP-A group. Among the P-Agroup, 22 proteins were upregulated with the m/z value of 2309, 2365, 3013, 2093, 2496, 2991, 3113, 3247, 5995, 2326, 2064, 4232, 1528, 2760, 2973, 2530, 2948, 1335, 1791, 3274, 2401 and 4131, 12 proteins were downregulated with the m/z value of 1207, 6333, 1440, 1484, 6172, 12325, 1663, 1214, 1192, 1137, 1172 and 7708. Only one significant differential protein was found between groups of P-A and P-C which was downregulated with the m/z value of 2365 in P-C group.2. According to protein fingerprints in H4 protein chip, 21 significant differential proteins were found between P-A and NP-A group. Among the P-A group, 9 proteins were upregulated with the m/z value of 3254, 2998, 3244, 1628, 2947, 4555, 5994, 1623 and 3852, 12 proteins were downregulated with the m/z value of 2735, 6177, 1269, 1245, 3666, 13801, 2746, 2718, 1157, 8405, 15130 and 1225. No significant differential protein was found between P-C and P-A group.3. Four differential proteins were detected by WCX2 and H4 protein chip together.Part Three: Screen serum biomarkers with relation to the formation of PVTT in HCC and establish the serum diagnostic and predictive proteomic patternThe objective of this part of research was to investigate the effects of Clinicopathologic variable on serum proteome of HCC and to screen serum biomarkers related to formation of PVTT and therefore establish the serum diagnostic and predictive proteomic pattern. Serum samples were collected from 152 HCC patients. Special serum protein or peptide spectra was determined by SELDI-TOF-MS measurement after treating the sample onto WCX2 protein chip for each case. The serum protein profiles were compared by Biomarker Wizard software among the patients stratified according to gender, age, AFP, presence of PVTT, tumor size, tumor number, presence of cirrhosis, respectively, while decision tree classification algorithm of Ma-PVTT and blind validity were determined by Biomarker Patterns Software.RESULTS:1. According to serum protein fingerprints of 112 HCC patients, a total of 100 protein peaks were identified at the m/z value ranging from 1100 to 30000, 16 significant differential proteins were found between the groups of HCC with singletumor and those with multiple tumors (PO.01) .2. Only one significant differential protein was found between the groups of HCC with tumor size >3cm and those with tumor size ^ 3cm, and 4 significant differential proteins between the groups of HCC with tumor size >5 cm and those with tumor size ^ 5cm, while 3 significant differential proteins between the groups of HCC with tumor size > 10 cm and those with tumor size ^10 cm.3. Four significant differential proteins were found between age groups of HCC patients >55 years and those ^55 years, while no significant differences when patients stratified according to other age standards.4. No significant differential protein was found when patients stratified according to gender, presence of cirrhosis and AFP.5. Sixteen significant differential proteins were found between the groups of Ma-PVTT and those without PVTT (N-PVTT) (PO.01). Among Ma-PVTT group, 6 proteins were upregulated with the m/z value of 3478,1314,1744,1725,2022 and 3380, 10 proteins were downregulated with the m/z value of 8901,9353,9415,8773,2766,2745,8697,7773,8569 and 1373 respectively. Only 2 significant differential protein were found between the groups of Mi-PVTT and N-PVTT ( PO.01). Among Mi-PVTT group, 2 proteins were upregulated with the m/z value of 15942 and 7978. 8 significant differential proteins were found between the groups of Ma-PVTT and Mi-PVTT ( PO.01), Among Ma-PVTT group, 3 proteins were upregulated with the m/z value of 1246, 1725 and 2937, 5 proteins were downregulated with the m/z value of 5319, 7783, 7978, 9296 and 15944.6. Serum protein profiles obtained from 95 HCC patients were used to generate Ma-PVTT decision tree by Biomarker Patterns Software, which combined with 7 candidate proteins with the m/z value of 3478,2022,8901,9415,8773,2766 and 2745, with a sensitivity of 75.8% (25/33) and specificity of 82.3% (51/62).7. The validity of the decision tree was then challenged with a test set of 40 samples. As a result of blind assessment, an accuracy of 87.5% (35/40) , sensitivity of 100% (18/18), specificity of 77.3% (17/22 ) , positive predictive value of 78.3% (18/23) and negative predictive value of 100 % (17/17) were obtained.CONCLUSIONS1. SELDI-TOF MS was an ideal technological platform for proteomic research...
|
PDF Full Text Request