Tzu-Hung Hsiao

Distinct signaling pathways after higher or lower doses of radiation in three closely related human lymphoblast cell lines.

PURPOSEnThe tumor suppressor p53 plays an essential role in cellular responses to DNA damage caused by ionizing radiation; therefore, this study aims to further explore the role that p53 plays at different doses of radiation.nnnMATERIALS AND METHODSnThe global cellular responses to higher-dose (10 Gy) and lower dose (iso-survival dose, i.e., the respective D0 levels) radiation were analyzed using microarrays in three human lymphoblast cell lines with different p53 status: TK6 (wild-type p53), NH32 (p53-null), and WTK1 (mutant p53). Total RNAs were extracted from cells harvested at 0, 1, 3, 6, 9, and 24 h after higher and lower dose radiation exposures. Template-based clustering, hierarchical clustering, and principle component analysis were applied to examine the transcriptional profiles.nnnRESULTSnDifferential expression profiles between 10 Gy and iso-survival radiation in cells with different p53 status were observed. Moreover, distinct gene expression patterns were exhibited among these three cells after 10 Gy radiation treatment, but similar transcriptional responses were observed in TK6 and NH32 cells treated with iso-survival radiation.nnnCONCLUSIONSnAfter 10 Gy radiation exposure, the p53 signaling pathway played an important role in TK6, whereas the NFkB signaling pathway appeared to replace the role of p53 in WTK1. In contrast, after iso-survival radiation treatment, E2F4 seemed to play a dominant role independent of p53 status. This study dissected the impacts of p53, NFkB and E2F4 in response to higher or lower doses of gamma-irradiation.

Applying gene set analysis to characterize the activities of immune cells in estrogen receptor positive breast cancer

Background: Estrogen receptor (ER) is a crucial biomarker for subtyping breast cancer. The present study aimed to understand the influence of infiltrated immune cells to patients’ outcome in estrogen receptor positive (ER+) breast cancer. n Methods: Gene expression profiles of three breast cancer cohorts downloaded from Gene Expression Omnibus (GEO) were used in this study. We utilized gene set enrichment analysis (GSEA) to estimate the activities of immune cell infiltration based on 31 published immune gene sets. Each gene set was tested for ER+ associated prognostic value. GSEA was applied to identify biological functions associated with prognostic immune gene sets. n Results: Nine subtypes of immune cells showed ER+ specific association with patient survival; seven of them formed two co-activation clusters, including: (I) activated CD4, CD8, effector memory CD4, and (II) regulatory T cell, dendritic cell, eosinophil, and mast cell, substantially representing innate and adaptive immunity. Among them, activated CD8 and mast cell were independent prognostic factors in multivariate Cox regression. Functional annotation analysis revealed their involvement in breast cancer subtyping, relapse, and metastasis. n Conclusions: We devised a gene set analysis to comprehensively investigate the involvement of ER specific immune cell activities and prognosis in breast cancer. Our work provides hints of the interaction between infiltrated immune cells and activated oncogene in ER+ breast cancer and may contribute to the biological basis for the development of immunotherapy.

Analyzing Differential Regulatory Networks Modulated by Continuous-State Genomic Features in Glioblastoma Multiforme

Gene regulatory networks are a global representation of complex interactions between molecules that dictate cellular behavior. Study of a regulatory network modulated by single or multiple modulators expression levels, including microRNAs (miRNAs) and transcription factors (TFs), in different conditions can further reveal the modulators’ roles in diseases such as cancers. Existing computational methods for identifying such modulated regulatory networks are typically carried out by comparing groups of samples dichotomized with respect to the modulator status, ignoring the fact that most biological features are intrinsically continuous variables. Here, we devised a sliding window-based regression scheme and proposed the Regression-based Inference of Modulation (RIM) algorithm to infer the dynamic gene regulation modulated by continuous-state modulators. We demonstrated the improvement in performance as well as computation efficiency achieved by RIM. Applying RIM to genome-wide expression profiles of 520 glioblastoma multiforme (GBM) tumors, we investigated miRNA- and TF-modulated gene regulatory networks and showed their association with dynamic cellular processes and brain-related functions in GBM. Overall, the proposed algorithm provides an efficient and robust scheme for comprehensively studying modulated gene regulatory networks.

Abstract 2265: An integrated analysis of gene mutations and gene sets for predicting paclitaxel response in lung adenocarcinoma

Lung cancer is the leading cause of cancer death worldwide. A prevalent histological subtype of lung cancer is adenocarcinoma. Chemotherapies and targeted therapies have been developed to treat such malignancy. However, due to the heterogeneity of cancer genomes, drug responses vary considerably among patients and the average survival rate remains quite unsatisfactory. Therefore, integrated biomarkers for predicting drug responses are greatly needed. Addressing this, in the present study we aimed to develop a prediction model based on an integrated analysis of gene mutations and gene sets. Briefly, the two-tailed Student9s t-test was performed to identify the gene mutations and gene sets of which activities were associated with drug sensitivity, and classification trees were derived from these genomic features. We applied the analysis to genomic datasets and drug sensitivity data from the Cancer Cell Line Encyclopedia (CCLE) and gene sets defined in the Molecular Signatures Database (MSigDB), and constructed a prediction model for response to paclitaxel, a widely used drug for cancers, in lung adenocarcinoma. Taking KRAS mutation as an example, we identified 20 and 15 drug response-associated gene sets in KRAS-mutant and KRAS-wild type cell lines, respectively. The two lists of gene sets were mutually exclusive, suggesting the need of building individual prediction models for groups of cancer subtypes. We then built a classification tree for each of the two groups and tested their prediction performance by leave one out cross-validation tests; ∼64% and ∼81% accuracy was achieved for KRAS-mutant and KRAS-wild type cell lines, respectively. Gene sets of “SIG_CHEMOTAXIS” and “PID_ERB_GENOMIC_PATHWAY” served as crucial nodes for the trees of KRAS-mutant and KRAS-wild type cells, respectively. In conclusion, we developed a novel method that integrates gene mutations and gene sets for predicting drug responses and demonstrated its high performance in lung adenocarcinoma. Our model is widely applicable to identify potent biomarkers for anticancer drugs in cancers and accelerate the realization of precision medicine. Citation Format: Chia-Yu Huang, Yu-Chiao Chiu, Tzu-Pin Lu, Liang-Chuan Lai, Mong-Hsun Tsai, Tzu-Hung Hsiao, Eric Y. Chuang. An integrated analysis of gene mutations and gene sets for predicting paclitaxel response in lung adenocarcinoma. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2265.

Abstract 1507: Investigation of estrogen receptor-modulated association between immune activities and patient survival in breast cancer

In breast cancer, estrogen receptor (ER) is a crucial biomarker for subtyping and predicting patients’ prognosis. Nonetheless, its modulation of immune-related genes and survival was rarely discussed. Addressing this, in the present study we devised a comprehensive analysis to compare the survival associations of immune genes between ER+ and ER- breast tumors. Specifically, we modeled immune activities by a gene set enrichment analysis of 31 immunologic gene sets defined by a recent study. Whole-genome expression profiles and ER status of 279 breast cancer patients (245 ER+ and 34 ER-) were downloaded from the Gene Expression Omnibus (GSE4922); another dataset was incorporated for validation (GSE2034). Nine immunologic gene sets were significantly predictive of relapse-free survival (RFS) in a ER+ specific manner (Cox P-values 0.05 in ER-). In order to investigate the modulation of ER in gene set interactions, we calculated the Pearson correlation coefficients between gene sets in ER+ and ER- cohorts, respectively. The gene sets formed two coexpression clusters in ER+ patients (all correlation coefficients>0.7); one cluster was composed of protective gene sets (with hazard ratios 0.6 and In conclusion, using a gene set framework we comprehensively investigated the involvement of ER in modulating between immune activities and prognosis. Further biological investigations into our findings are warranted. Citation Format: Yi-Hsuan Chang, Yu-Chiao Chiu, Tzu-Pin Lu, Liang-Chuan Lai, Mong-Hsun Tsai, Tzu-Hung Hsiao, Eric Y Chuang. Investigation of estrogen receptor-modulated association between immune activities and patient survival in breast cancer. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1507.

Abstract 5284: Characterizing the differences of allelic imbalance between tumor and normal tissues by next-generation sequencing

Genetic variation has proved to contribute to tumor formation. One of the phenomenon caused by genetic variation, allelic imbalance, means different expression abundance between two alleles, recently has become a special landmark on cancer biology, with high prevalence in recent cancer studies. However, the present of allelic imbalance has not been fully understood yet. Taking the advantages of next generation sequencing, allelic imbalance can be identified and quantified through bioinformatics algorithms. Here, we developed a mathematical model to identify with allelic imbalance by concurrently analyzing RNA-Seq and DNA-Seq. First we called heterozygous SNP variants from both RNA-Seq and Exome-Seq data. The allele ratio is counted by the ratio of read depth between two alleles after integrating SNPs into gene level. Next, we identify allelic imbalance genes by differences of RNA and DNA allele ratios between tumor and normal samples. As results, a total of 91143 SNPs were identified in Exome-Seq, and about 3% (2586 out of 91143) SNPs were heterozygous in both tumor and normal tissues. We found that about 118 allelic imbalance genes in tumor tissue are significantly different from those in normal tissue. Furthermore, we investigate the enriched function of these allelic imbalance genes. These result showed that these genes are significantly enriched in DNA repair pathway. We speculated that allelic imbalance in tumor tissue may be involved in the dysregulation of DNA repair. Also other enriched functions were identified, such as z-finger protein pathway. Our goal is to investigate the landscape of allelic imbalance between tumor sample and normal sample. Our results demonstrate the role of allelic imbalance of dysregulation of tumor. Moreover, our method can be applied to uncover more basic mechanisms on allelic imbalance and tumorigenesis. Citation Format: Yao-Wei Jheng, Yu-Chiao Chiu, Tzu-Pin Lu, Liang-Chuan Lai, Mong-Hsun Tsai, Tzu-Hung Hsiao, Eric Y Chuang. Characterizing the differences of allelic imbalance between tumor and normal tissues by next-generation sequencing. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 5284.

Characterization of conditions for competing endogenous RNA regulation in GBM

Summary form only given. MicroRNAs (miRNAs) are short non-coding RNAs with the average length of 22 nucleotides. They are known to induce mRNA degradation or suppression of translation by complementarily binding to 3 untranslated regions (3 UTRs) of target mRNA transcripts. Recently, an alternative mechanism through which miRNAs participate in gene regulation was postulated and experimentally validated, namely the competing endogenous RNAs (ceRNAs). By competing for a limited pool of common targeting miRNAs (miRNA programs; miRP), pairs of genes (ceRNAs) sharing, fully or partially, identical miRNAs binding sites can “talk” to each other: when one ceRNA is up-regulated (or down-regulated) in cells, it attracts (or releases) the targeting miRNAs away from (or toward) the other ceRNA, and in turn have protective (or harmful) effects on expression of the other ceRNA. Based on in silico and in vitro analysis, recent reports suggested the dynamic and condition-specific properties of ceRNA regulation. The essential factors involved in ceRNA regulation include size of miRP, number of miRP binding sites, expression level of miRP, and expression level of ceRNAs. For better characterizing the optimal conditions for ceRNA regulation, in the present study we aim to confer how essential factors determine strength of ceRNA regulation in vivo, by analyzing TCGA datasets of glioblastoma multiforme (GBM) patients with 491 tumor samples profiled with paired miRNA and gene expression. Based on the definition that two genes sharing any number of common targeting miRNAs as a putative ceRNA pair, and by utilizing TargetScan algorithm, we identified 47,451,423 putative ceRNA pairs, involving 10,872 ceRNAs (genes). Pairwise correlation coefficients of gene expression profiles were then computed for each of the putative ceRNA pairs, and then the CDF. Varying size of miRP, for example, generated multiple CDFs, and then the goodness-of-fit was performed for pinpointing the essential factors and optimal conditions for intensified ceRNA activity. Our analysis results demonstrated that increased size of miRPs as well as the abundance of miRP binding sites stabilize ceRNA activity and strengthen coexpression of ceRNA pairs. Furthermore, the expression levels of both miRPs and ceRNAs affect ceRNA activity and lead to statistically significant differences in distributions of correlation coefficients. Taken together, the results indicated that ceRNA regulation depends on states of the essential factors and thus may involve complex and dynamic processes in vivo. Our findings bring biological insights into complex ceRNA crosstalk in glioblastoma multiforme and contribute to further unveiling complex mechanism governing ceRNA regulation.

Abstract 1993: Higher activity of cell proliferation is associated with poor survival in lung adenocarcinoma

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DCnnThe number of patients with lung adenocarcinoma, a major histological type of lung cancer, has soared recently. Although some lung adenocarcinoma patients could be diagnosed at early stages and received surgical resection, nearly 50% of patients still died from recurrence. This phenomenon indicates that using only traditional prognostic factors, like TMN staging, is not sufficient for regimen decision. Here we proposed a bioinformatic method to evaluate the activity of cell cycle from expression profiles and proved that the activity of cell cycle can be applied as prognostic factor of lung adenocarcinoma, especially for early stages.nnA gene set named as cell cycle signature (CCS), which correlated with cell cycling, was used to measure the activity of tumor proliferation. The E2F1 target genes and E2F3 target genes were also applied to measure the E2F signaling pathway activity, which was regulated by the cell cycle. The prognostic ability of these gene sets was evaluated by applying Cox hazard proportional analysis to three microarray datasets of lung adenocarcinoma.nnOur result revealed that the association between high activities of cell cycle and poor survival outcomes existed in lung adenocarcinoma patients. The result of Cox hazard proportional analysis showed that the CCS and E2F1/E2F3 target genes are significant contributors to survival risk. The p-values of CCS, E2F1/E2F3 target genes were all showed significant (p<0.05), and these three gene sets still gave statistical significances after the survival analysis was applied on only early stage lung adenocarcinoma. The hazard ratios of CCS, E2F1 target genes and E2F3 target genes were 5.69, 3.32, and 3.96, respectively; all p-values were smaller than 0.002.nnWe demonstrated that the three gene sets, which accountable for the activity of cell cycle, can be applied as prognostic factors for lung adenocarcinoma, including early stage patients. Our data inferred that tumor with higher activity of proliferation is malignant eventhough its size may still be small. This finding is valuable in predicting the survival of early stage lung adenocarcinoma patients and might be useful for making treatment decisions.nnCitation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 1993.

Abstract 1326: Differential cytokine-related signaling pathways responded to ionizing radiation in two lung adenocarcinoma cell lines

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DCnnLung cancer is the leading cause of cancer-related mortality in the world. Metastasis is the leading cause of death and an enormous therapeutic challenge for non-small cell lung cancer. Radiation therapy is a treatment of unresectable lung cancer and may palliate the symptoms, locally control tumor growth, and provide higher survival rate. Yet, the molecular mechanism of radio-sensitivity in lung cancer still remains unclear. Two lung adenocarcinoma cell lines (CL1-0 and CL1-5) with different metastatic potentials were irradiated with 10Gy γ-radiation. CL1-5 showed more radio-sensitive than CL1-0 since the surviving fractions in CL1-5 were ten times lower than CL1-0 by clonogenic assays. Gene expression profiles were also analyzed using microarray in order to better understand differential gene expression between these two cell lines after radiation. Total RNAs for whole genome gene expression analysis were extracted from these two cell lines at 0, 1, 4, 9 and 24 h after 10Gy irradiation. Gene Set Enrichment Analysis (GSEA) revealed the pathways of death, cytokine, cell adhesion, IL1R, NF-κB, and TNFR had differences between these two cell lines following 10Gy radiation exposure. Tumor necrosis factor (TNF) cytokine family and interleukin-1 (IL1) was up-regulated in CL1-5 after irradiation. The expression level of TNF-α validated by quantitative real time PCR was increased in CL1-5 and was not detected in CL1-0. Therefore, TNF-α released into the culture medium in CL1-5 after irradiation might be a death signal to activate the NF-κB pathway and cause more cell death. Moreover, the different responses of NF-κB-related pathways induced by radiation between the two cell lines might contribute the different survival. These cytokine modulations might provide promising targets for further investigation regarding radiation treatment in lung cancer.nnCitation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 1326.