Study On The Role Of Bone Marrow X - Related Non - Receptor Picornine Kinase In Assessing The Prognosis Of Human Glioblastoma Patients

Part OneThe research on relations between expression of BMX and CD133in glioma cell lines and analysis of significance of BMX expression indicating glioma malignancyObjective:Detect the expression of BMX different glioma cell lines, compared with CD133, classic biomarker of glioma stem cells. An analysis of co-expression of BMX and CD133is carried out on the GBM frozen sections and glioma cell lines. Further step is taken to observe the BMX expression in glioma with various grades, and the relation between BMX expression and the glioma malignancy is analysized. BMX working as a chemoresistant factor is also explored. Method:Select by random the, glioma samples of different grade, including WHO I II III IV,3cases from each grade, which all belong to operation samples owned by the Department of Neurosurgery. The technique of western blot is adopted to observe the expression of BMX in glioma of different grades. Additionally, U138, U87, U373and T98are selected for cell culture in vitro as four glioma cell lines with specific genotype. The immunof luorescence staining is adopted to have BMX and CD133as a further step. BMX and CD133expression can be observed via microscope. Meanwhile,10samples of glioblastoma are selected to go for OCT embedding right after tumor resection, then for cut sections via the frozen cutting sections machine, into slice with a thickness of4um. The immnofluorescence staining are used to detect BMX and CD133expression, and the fluorescence intensity can be observed under the fluorescence microscope. The cell U87is cultured in vitro, co-stained with BMX and p-gp. Meanwhile, temozolomide of two different concentration of10uM and100uM are added in culture medium, to observe the expression change of BMX and p-gp. Result:A western blot result shows that the BMX expression is directly proportional to the GBM malignancy, with expression difference from point of statistics on various grades (p=0.0000), reaching the maximum in the glioblastoma (BMX/GAPDH=0.982±0.099). The expression of BMX is found in U87\U87\U373and T98, and co-expression of BMX and CD133is found in U87and U138. Moreover, the positive cells of CD133are found with BMX expression, directly proportional to each other. However, not all the positive cells of BMX express CD133, and we find through the frozen sections CD133positive area is companied with common BMX expression, especially around the blood vessel of tumor tissue, with BMX/CD133showing stronger positive and more BMX/CD133++cells accumulated. The co-expression of BMX and p-gp is found in the U87cells, showing fluorescence intensity elevated while the concentration of temozolomide goes up. Conclusion:BMX is expressed in various glioma cell lines, and can be expressed together with CD133. The expression levels are relevant, goes as two subgroups called BMX/CD133++and BMX/CD133+-, while there is no BMX/CD133-+. The same result is also shown in the frozen sections of glioblatoma, a further conclusion is obtained that increased expression is found around the blood vessel of tumor tissue, and its expression is relevant with the tumor malignancy. The higher gliom grade diagnosis results in, the more expression is shown by BMX. Moreover, BMX can be regarded as one of the chemoresistant biomarker. Part Two Analysis on expression profiling of human glioblastoma samples with differential prognosis and study on mRNA expression level of BMX within tumor samples.Objective:The existence of glioma stem cells is one of the main factors to impact the prognosis of glioblastoma patients. We are trying to find the genes regulating glioma stem cells, through an analysis on gene expression profiling of the patients who behave in greater different ways in prognosis. Meanwhile, a study is carried out to observe if BMX is a gene showing different expression in glioblatoma patients with different prognosis. Method:On retrospectively review of glioblastoma patients who received operation in neurosurgery department from2009to2012,10cases are selected in which great difference is shown in prognosis. Among them,4are showing sound prognosis while6with poor outcome. All the H&E sections were reviewed by pathologist to make sure the diagnosis is glioblastoma, and matched frozen samples are found in tissue bank which have been RNAlater processed. A microarray is carried out via Agilent Gene Chip after RNA is extracted. The result is then analysized by GeneSpring GX software, version11.5(Agilent Technologies), after the normalization, the raw data was analyzed by T-test, the fold change≥2with P value≤0.05are regarded having statistical significance. The top10differential expressed genes are selected (ranking from both highest and lowest respectively) to go for qRT-PCR validation. Result:The10patients underwent operation with total tumor resection. Among them, the average survival time for4patients with sound prognosis is21±3.4months while the other patients is6.5±1.22months in average.332differential expressed genes are found in microarray among which206are upregulated and126are downregulated. The top10of upregulated genes are GPR26、 PDYN、TAC1、SST、PNOC、SOHLH1、SCGN、SV2B、BMX、NEUROD2, while the top10of downregulated genes are APO4、CXCL10、UTS2、ISG15、RSAD2、NR4A3、 IFI44L、AGR3、C12orf55、C9orf171. The top10terms in GO function analysis are synaptic transmission、transmission of nerve impulse、cell-cell signaling、neurological system process、learning or memory、gated channel activity、potassium ion binding、metal ion transmembrane transporter activity、cation channel activity、regulation of transmission of nerve impulse. The BMX is selected as target gene, qRT-PCR result is shown lower expression in sound prognosis group than in the poor one (4.018±0.832VS7.522±0.523, p=0.0017). Conclusion:The average survival time for glioblastoma of different prognosis are distinguishingly different, clearly shown in gene expression. BMX is shown with the highest expression as the marker of glioma stem cells in patients with poor prognosis, while largely upregulation in the sound outcome group. Part Three The study and analysis on the IDH mutation, MGMT promoter methylation and chromosomal1p19q co-deletion in glioblastoma patients correlated with BMX protein expression and significance in predicting prognosisObjective:BMX is regarded as a glioma stem cell marker and correlated with glioma malignancy. We aim to study on whether it is also relevant with the prognosis of patients diagnosed with glioblastoma compared with classical glioblastoma biomarker MGMT promoter methylation, IDH mutation and chromosomal1p19q co-deletion. Meanwhile, we combined these four biomarkers in order to find the best biological marker to identify the prognosis of glioblastoma patients. Method:In retrospective review of106consecutive glioblastoma cases who receiving standard treatment modulation in neurosurgery department from2002to2009, the clinic medical histories are reviewed and summed up by gender, age, tumor location, management options, tumor resection extent, general status before and after the operation. Meanwhile, clinical follow-up are conducted for progression free survival time and overall survival time. The excluded criterion is as followed:underwent second operation when tumor recurrence, death caused by clinical complication or concomitant symptom, loss to follow-up. The paraffin section with thickness of4um are well prepared for these samples, H&E slides were reviewed by pathologist to make sure the diagnosis of glioblastoma (WHO IV). The microdissection is carried out to get pure tumor content, followed by DNA extraction. MGMT promoter methylation was detected by methylation specific PCR, direct sequencing for the detection of IDH mutation. Fluorescence in situ hybridization for1p19q co-deletion test, and BMX expression is screened by immunohistochemistry. Meanwhile,67cases from the external hospital are selected as the validation cohort. IRS system is adopted for scoring BMX expression. The relations between the four markers and the overall survival time is checked by the log rank test, P value≤0.05indicate being statistical significant. The survival curve was made by Kaplan-Merier method. Result:Among106cases finally selected, male patients account for63and female43. There are3cases under age of30(2.8%),12cases within age30-40(11.36%),24cases within age40-50(22.64%),34cases within50-60(32.08%), and33cases above age60(31.13%).47cases have tumor on frontal lobe(44.34%),33cases(33.13%),6cases (5.66%),16cases(15.09%) on temporal lobe, parietal lobe and occipital lobe respectively, other4are located on other locations(3.77%).92patients got total tumor resection(86.79%), while12patients have partial tumor resection(11.32%),2for biopsy(1.89%). The average survival time is19.9+12.87months. Gender, side and tumor resection extent have no impact on prognosis. Patients under the age of60share sound prognosis than those above60years old (with average median survival time of18months VS14months, p=0.0175). Patients bearing tumor on occipital lobe have better survival time(median survival time is18months, p=0.0017). Among all the patients, the occurrence rate of MGMT promoter methylation accounts for37.74%, the patients with MGMT promoter methylation have better prognosis than those with MGMT promoter unmethylation(median survival time is20.5months VS15.0months, p=0.0028).13cases are present with IDH1mutation, these patients have better prognosis than IDHwild type cases but no statistical significance is shown (median survival time is22months VS17months, p=0.1278). We adopt IHC scoring point4as a cut-off point, find that patients with BMX expression over4live shorter than those under4(median survival time is22months VS.12months, p=0.0000). The same result is shown by the validation cohort(median survival time is14months VS10months, p=0.0054). We take further step to make combination of MGMT promoter methylation, IDH mutation and BMX expression to find out the most significant biomarker. The result shows that MGMT promoter methylation combined with BMX IHC score=≤4has the most obvious significance in predicting the prognosis of GBM patients (median survival time is28months VS9.0months, p=0.0000), Ip19q test shows polysomy in those MGMT promoter methylation/BMX IHC≤4, no co-deletion was detected. In the primary-recurrent paired cases, we found8out of10BMX negative in primary cases but positive in recurrent cases, EGFRvIII is no longer expressed in these paired cases. Conclusion:BMX expression level can be considered as a good biomarker to prodict the prognosis of the glioblastoma. A combination of BMX and MGMT promoter methylation is even more distinct. IDH mutation is not so significant in predicting GBM patients prognosis. MGMT promoter methylation/BMX lower expression cases show no correlation with1p19q co-deletion. And BMX is probably the key factor in reoccurrence of the glioma and in its process.