| BackgroundIn recent years,liquid biopsy is more and more applied in early diagnosis of cancer prognosis,prediction of recurrence risk assessment,monitoring disease progression and prediction of response to treatment through continuous sampling.Compared with traditional biopsy,liquid biopsy has the advantages such as non-invasiveness,homogenization and continuous sampling.At present,liquid biopsy technology is mainly based on CTCs,cfDNA and ctDNA in peripheral blood,of which CNVs gene mutation and methylation pattern in ctDNA is mainly detected in clinical.However,there are still some patients cannot get effective information through the above detection technologies.Cell free DNA(cfDNA)was mainly released into the bloodstream from the apoptotic cells due to nucleosome protection.The nucleosome footprint profile in genome is closely related to gene transcription and expression level,especially in the gene expression of transcription start site(TSSs),where nucleosome combined significantly decreased.Since gene expression techniques based on tissue biopsy have been widely used in breast cancer,such as Oncotype DX21 gene detection,Mamma Print 70 gene detection,EndoPredict 11 gene detection,Prosigna PAM50 gene detection and Breast Cancer Index(BCI)11 gene detection,we performed this study by whole genome sequencing of cfDNA in plasma and analysis of the depth coverage of TSSs area and the difference between benign breast tumor and malignant breast cancer,different molecular subtypes,lymph node metastasis and the prediction of the efficacy of neoadjuvant chemotherapy,and it will provide a broader and more reliable data source for the noninvasive detection and treatment of breast cancer,with a greater prospect.Research PurposeThis study aimed to establish a noninvasive classification method which could distinguish benign breast tumor and malignant breast cancer,different molecular subtypes,lymph node metastasis and predict the efficacy of neoadjuvant chemotherapy,based on the whole genome sequencing of cfDNA in plasma and analysis of the depth coverage of TSSs area that could predict gene expression.Research MethodsIn chapter 1,we used two breast cancer cell lines of MDA-MB-231 and T-47D and 50 cases of breast cancer patients' plasma,and studied the relationship of cfDNA distribution characteristics and gene expression.Our research and analysis was carried out from the following three aspects:1.the relationship of cfDNA distribution characteristics of cell supernatant and intracellular nucleosome footprint profile;2.the relationship of cfDNA distribution characteristics of cell supernatant and intracellular gene expression;3.the relationship of cfDNA distribution characteristics of breast cancer patients plasma and breast cancer cell lines,and the specificity of cfDNA distribution characteristics of breast cancer patients plasma.In order to determine the above relations,we first analyzed the depth coverage of cfDNA TSSs ± 1 k bp area of breast cancer cell lines' supernatant and breast cancer patients plasma,the depth coverage of intracellular nuclesome DNA TSSs±1 k bp area,intracellular different gene expression,and high expressed genes in breast cancer from TCGA database,and then performed spearman rank correlation analysis and permutation tests to study the correlation between per two groups.In chapter 2,we researched clinical application in the diagnosis of breast cancer classification,molecular subtypes,and lymph node metastasis prediction of cfDNA distribution characteristics.First we assessed the whole and local chromatin variables in breast cancer,and performed analysis from the four following different aspects:5-megabase(Mb)windows,sub-compartments and the copy number of the mitochondrial genome,and the depth coverage of TSSs±1 kb area.And finally we used the depth coverage of TSSs ± 1 kb area and logistic regression model,and developed the classification method the development respectively for distinguish of benign breast tumor and malignant breast cancer(benign breast tumor,197 cases;breast cancer,263 cases),molecular subtypes classification(Luminal A,41 cases;Luminal B,82 cases,Her2,26 cases;Triple negative,24 cases)and lymph node metastasis prediction(lymph node metastasis,53 cases;lymph node non-metastasislymph,51 cases).And we evaluated effect of the classifier through the ROC curve.In chapter 3,we studied the relationship of cfDNA distribution characteristics and the treatment effect of neoadjuvant chemotherapy in breast cancer.We first analyzed cfDNA distribution characteristics and the changes of a total of 20 patients'plasma(12 responders and 10 nonresponders)after different cycles of neoadjuvant chemotherapy treatment(post-1 cycle,post-3/4 cycles,and post-8 cycles of treatment)and carried out unsupervised cluster analysis of the two groups,and finally performed the related gene function analysis.Subsequently,we tried to analyze the difference of cfDNA distribution in plasma samples of 38 patients(28 responders and 10 non-responders)receiving neoadjuvant chemotherapy before chemotherapy,to explore the feasibility of predicting the treatment efficacy before chemotherapy.ResultsIn chapter 1,Spearman rank correlation analysis showed that cfDNA distribution characteristics of cell supernatant were positively correlated with intracellular nucleosome footprint profiles,and negatively correlated with intarcellular gene expression.Permutation tests showed that the overlaps of HTSSs(high coverage depth around TSSs)or LTSSs(low coverage depth around)of cell supernatant cfDNA and intracellular nucleosome footprint profiles were significantly enriched(p<10-22).The overlapsof LTSSs of cell supernatant cfDNA and intracellular HEGs(highly expressed genes)(T-47D:p=9.116e-06;MDA-MB-231:p=1.981e-05),and of HTSSs of cell supernatant cfDNA and intracellular LEGs(lowly expressed genes)(T-47D:p=2.987e-06;MDA-MB-231:p=3.156e-10)were significantly enriched.The overlaps of lower coverage depth near TSSs in breast cancer patients compared to healthy donors and highly expressed genes in primary tumor tissue compared to the adjacent breast tissue(TCGA;p=0.014)were significantly enriched.In chapter 2,through the whole and local fragments of cfDNA study,it was found that the relative coverage of promoter region of cfDNA is suitable for establishing PPCET classifier.The accurate and specific of established classifiers for the prediction of malignant breast cancer and benign breast tumor,molecular subtypes classification and lymph node metastasis were all about 80%.In chapter 3,through the relationship of the series of plasma specimens and the treatment efficacy of neoadjuvant chemotherapy study,the results showed that the TSS regions of 321 genes were significantly differentially covered in responders,including 205 genes post-1 cycle:93 of these genes' TSS regions were downregulated and 112 were upregulated during early treatment,and 112 genes post-3/4 cycles:66 of these genes' TSS regions were downregulated and 50 were upregulated during mid-treatment,with stable coverage after treatment(post-8 cycles).The genes changed after chemotherapy were mainly related to the response to DNA damage and immune response by inhibiting cell proliferation.CfDNA distribution characteristics of breast cancer plasma samples before neoadjuvant chemotherapy were analyzed,and there were 232 genes with significantly different depth coverage in responders and nonresponders:100 high TSSs coverage genes and 132 low TSSs coverage genes in responder.Conclusion1.CfDNA distribution characteristics were positively correlated with intracellular nuclesome footprint profile and negatively correlated with intracellular gene expression;2.CfDNA distribution characteristics of breast cancer patients were breast cancer specific in-TCGA database;3.Three classifiers were made for breast cancer diagnosis,the molecular subtypes classification and lymph node prediction based on the relative coverage of promoter region of cfDNA,and they could separate each group into different categories,which has provided a feasible liquid biopsy technology for the non-invasive diagnosis of breast cancer;4.Through the study the correlation between the effect of neoadjuvant chemotherapy and a series of plasma samples,the genes changed after chemotherapy in the responders were mainly related to the response to DNA damage and immune response by inhibiting cell proliferation;5.Only the distribution characteristics of cfDNA in breast cancer plasma samples before neoadjuvant chemotherapy were analyzed,and differences between responders and non-responders were found,which is expected to develop new biomarkers to predict the efficacy of neoadjuvant chemotherapy for breast cancer before chemotherapy,and achieve non-invasive prediction. |
PDF Full Text Request