Disrupting the interaction between a p53 gain-of-function mutant and the transcriptional co-activator PC4 reverses drug resistance in cancer cells.

PC4 is a chromatin-associated protein and transcriptional coactivator whose role in gene regulation by wild-type p53 is now well known. Little is known about the roles of PC4 in tumor cells bearing mutant p53 genes. We show that PC4 associates with one of the tumor-associated gain-of-function p53 mutants, R273H. This association drives its recruitment to two promoters, UBE2C and MDR1, known to be responsible for imparting aggressive growth and resistance to many drugs. Here, we introduced a peptide that disrupts the PC4-R273Hp53 interaction to tumor cells bearing the R273HTP53 gene, which led to a lowering of MDR1 expression and abrogation of drug resistance in a mutant-specific manner. The results suggest that the PC4-R273Hp53 interaction may be a promising target for reducing proliferation and drug resistance in tumors.

Utilizing coordination chemistry through formation of a CuII-quinalizarin complex to manipulate cell biology: An in vitro, in silico approach.

Quinalizarin, an analogue of anthracycline anticancer agents, is an anticancer agent itself. A CuII complex was prepared and characterized by elemental analysis, UV-Vis & IR spectroscopy, mass spectrometry, EPR and DFT. The intention behind the preparation of the complex was to increase cellular uptake, compare its binding with DNA against that of quinalizarin, modulation of semiquinone formation, realization of human DNA topoisomerase I & human DNA topoisomerase II inhibition and observation of anticancer activity. While the first two attributes of complex formation lead to increased efficacy, decrease in semiquinone generation could results in a compromise with efficacy. Inhibition of human DNA topoisomerase makes up this envisaged compromise in free radical activity since the complex shows remarkable ability to disrupt activities of human DNA topoisomerase I and II. The complex unlike quinalizarin, does not catalyze flow of electrons from NADH to O2 to the extent known for quinalizarin. Hence, decrease in semiquinone or superoxide radical anion could make modified quinalizarin [as CuII complex] less efficient in free radical pathway. However, it would be less cardiotoxic and that would be advantageous to qualify it as a better anticancer agent. Although binding to calf thymus DNA was comparable to quinalizarin, it was weaker than anthracyclines. Low cost of quinalizarin could justify consideration as a substitute for anthracyclines but the study revealed IC50 of quinalizarin/CuII-quinalizarin was much higher than anthracyclines or their complexes. Even then, there is a possibility that CuII-quinalizarin could be an improved and less costly form of quinalizarin as anticancer agent.

Ets1 facilitates EMT/invasion through Drp1-mediated mitochondrial fragmentation in ovarian cancer.

Ovarian cancer has sustained as a major cause of cancer-related female mortality owing to its aggressive nature and a dearth of early detection markers. Ets1 oncoprotein, a transcription factor belonging to the Ets family, is a well-established promoter of epithelial to mesenchymal transition (EMT) and a prospective malignancy marker in ovarian cancer. Our study establishes Ets1 as a regulator of mitochondrial fission-fusion dynamics through Drp1 augmentation via direct binding at DNM1L (DRP1) promoter. Ets1 overexpression-mediated Drp1 increment resulted in mitochondrial load reduction and compromised OXPHOS Complex 5 (ATP synthase) expression, facilitating a greater reliance on glycolysis over OXPHOS. Furthermore, our work demonstrates that inhibition of mitochondrial fission through molecular or pharmacological inhibition of Drp1 successfully mitigates Ets1-associated EMT in both in vitro and in vivo syngeneic mice model. Collectively, our data highlight the role of Drp1-mediated mitochondrial fragmentation in driving Ets1-mediated bioenergetic alterations and EMT/invasion in ovarian cancer.

In-silico analysis of TCGA data showing multiple POLE-like favourable subgroups overlapping with TP53 mutated endometrial cancer: Implications for clinical practice in low and middle-income countries.

Introduction: The Cancer Genome Atlas cohort of endometrial carcinoma (TCGA-UCEC) includes almost 40% TP53-mutants encompassing missense and truncated variants. TCGA revealed 'POLE', characterized by POLE gene bearing exonuclease domain mutation (EDM), as the prognostically best molecular profile. The worst profile was characterized by TP53-mutated Type 2 cancer requiring adjuvant therapy having cost implications in low-resource settings. We aimed to find more 'POLE-like' favourable subgroups by searching TCGA cohort, especially within TP53 mutated risk group, that could eventually avoid adjuvant treatment in resource-poor settings. Method: Our study was an in-silico survival analysis performed on the TCGA-UCEC dataset using SPSS statistical package. TP53 and POLE mutations, microsatellite instability (MSI), time-to-event and clinicopathological parameters were compared among 512 endometrial cancer cases. Deleterious POLE-mutations were identified by Polyphen2. Progression free survival was studied using Kaplan-Meier plots keeping original 'POLE' as comparator. Result: In presence of wild type (WT)-TP53, other deleterious POLE-mutations behaved like POLE-EDM. Only truncated and not missense TP53 benefitted from POLE/MSI overlap. However, TP53 missense mutation, Y220C, was found to be as favourable as 'POLE'. Overlapping POLE, MSI and WT-TP53 also performed favourably. Truncated TP53 overlapped with POLE and/or MSI, TP53 Y220C alone and, WT-TP53 overlapped with POLE and MSI both, were named 'POLE-like' for prognostically behaving like the comparator 'POLE'. Conclusion: Obesity being a lesser frequent event in low and middle-income countries (LMICs), relative proportion of women with lower BMI and Type 2 endometrial cancers may be high. Identification of 'POLE-like' groups may facilitate therapeutic de-escalation in some TP53-mutated cases - a novel option. Instead of 5% (POLE-EDM), potential beneficiary would then comprise 10% (POLE-like) of TCGA-UCEC.

Promoter methylation and enhanced SKP2 are associated with the downregulation of CDKN1C in cervical squamous cell carcinoma.

PURPOSE: Cervical Squamous Cell Carcinoma (CSCC) is one of the significant causes of cancer deaths among women. Distinct genetic and epigenetic-altered loci, including chromosomal 11p15.5-15.4, have been identified. CDKN1C (Cyclin-Dependent Kinase Inhibitor 1C, p57KIP2), a member of the CIP/KIP family of cyclin-dependent kinase inhibitors (CDKIs), located at 11p15.4, is a putative tumor suppressor. Apart from transcriptional control, S-Phase Kinase Associated Protein 2 (SKP2), an oncogenic E3 ubiquitin ligase, regulates the protein turnover of CDKN1C. But the molecular status of CDKN1C in CSCC and the underlying mechanistic underpinnings have yet to be explored. METHODS: TCGA and other publicly available datasets were analyzed to evaluate the expression of CDKN1C and SKP2. The expression (transcript/protein) was validated in independent CSCC tumors (n = 155). Copy number alteration and promoter methylation were correlated with the expression. Finally, in vitro functional validation was performed. RESULTS: CDKN1C was down-regulated, and SKP2 was up-regulated at the transcript and protein levels in CSCC tumors and the SiHa cell line. Notably, promoter methylation (50%) was associated with the downregulation of the CDKN1C transcript. However, high expression of SKP2 was found to be associated with the decreased expression of CDKN1C protein. Independent treatments with 5-aza-dC, MG132, and SKP2i (SKPin C1) in SiHa cells led to an enhanced expression of CDKN1C protein, validating the mechanism of down-regulation in CSCC. CONCLUSION: Collectively, CDKN1C was down-regulated due to the synergistic effect of promoter hyper-methylation and SKP2 over-expression in CSCC tumors, paving the way for further studies of its role in the pathogenesis of the disease.

Identifying polymorphic cis-regulatory variants as risk markers for lung carcinogenesis and chemotherapy responses in tobacco smokers from eastern India.

Aberrant expression of xenobiotic metabolism and DNA repair genes is critical to lung cancer pathogenesis. This study aims to identify the cis-regulatory variants of the genes modulating lung cancer risk among tobacco smokers and altering their chemotherapy responses. From a list of 2984 SNVs, prioritization and functional annotation revealed 22 cis-eQTLs of 14 genes within the gene expression-correlated DNase I hypersensitive sites using lung tissue-specific ENCODE, GTEx, Roadmap Epigenomics, and TCGA datasets. The 22 cis-regulatory variants predictably alter the binding of 44 transcription factors (TFs) expressed in lung tissue. Interestingly, 6 reported lung cancer-associated variants were found in linkage disequilibrium (LD) with 5 prioritized cis-eQTLs from our study. A case-control study with 3 promoter cis-eQTLs (p < 0.01) on 101 lung cancer patients and 401 healthy controls from eastern India with confirmed smoking history revealed an association of rs3764821 (ALDH3B1) (OR = 2.53, 95% CI = 1.57-4.07, p = 0.00014) and rs3748523 (RAD52) (OR = 1.69, 95% CI = 1.17-2.47, p = 0.006) with lung cancer risk. The effect of different chemotherapy regimens on the overall survival of lung cancer patients to the associated variants showed that the risk alleles of both variants significantly decreased (p < 0.05) patient survival.

No association of the common Asian mitochondrial DNA haplogroups with lung cancer in East Indian population.

OBJECTIVES: Mitochondrial dysfunction has long been associated with the pathogenesis of lung cancer (LC). Mitochondrial DNA (mtDNA) haplogroups have been reported to modify the risk of LC in a few different populations; however, no study has been done among the Indians. Here, we explore the relationship between mtDNA haplogroups and LC in a representative eastern Indian sample set. METHODS: Different combinations of six mtDNA SNPs, which define the major Asian mtDNA haplogroups M and N, and their sub-haplogroups D, G, M7, R, and F were genotyped via polymerase chain reaction (PCR) - restriction fragment length polymorphism (RFLP) - sequencing approach in 94 smoker LC patients and 100 healthy smoker controls from an eastern Indian cohort. RESULTS: The distribution of 7 mtDNA haplogroups did not show any significant differences between patients and controls (p<0.05). We did not find sub-haplogroup M7 in our study population. CONCLUSIONS: Our study is the first to indicate that the major Asian mtDNA haplogroups have no significant (p<0.05) association with LC in East Indian population.

Transition of amyloid/mutant p53 from tumor suppressor to an oncogene and therapeutic approaches to ameliorate metastasis and cancer stemness.

The tumor suppressor p53 when undergoes amyloid formation confers several gain-of-function (GOF) activities that affect molecular pathways crucial for tumorigenesis and progression like some of the p53 mutants. Even after successful cancer treatment, metastasis and recurrence can result in poor survival rates. The major cause of recurrence is mainly the remnant cancer cells with stem cell-like properties, which are resistant to any chemotherapy treatment. Several studies have demonstrated the role of p53 mutants in exacerbating cancer stemness properties and epithelial-mesenchymal transition in these remnant cancer cells. Analyzing the amyloid/mutant p53-mediated signaling pathways that trigger metastasis, relapse or chemoresistance may be helpful for the development of novel or improved individualized treatment plans. In this review, we discuss the changes in the metabolic pathways such as mevalonate pathway and different signaling pathways such as TGF-ß, PI3K/AKT/mTOR, NF-κB and Wnt due to p53 amyloid formation, or mutation. In addition to this, we have discussed the role of the regulatory microRNAs and lncRNAs linked with the mutant or amyloid p53 in human malignancies. Such changes promote tumor spread, potential recurrence, and stemness. Importantly, this review discusses the cancer therapies that target either mutant or amyloid p53, restore wild-type functions, and exploit the synthetic lethal interactions with mutant p53.

Deep sequencing reveals the spectrum of BCR-ABL1 mutations upon front-line therapy resistance in chronic myeloid leukemia: An Eastern-Indian cohort study.

The course of clinical management in chronic myeloid leukemia (CML) often faces a road-block in the form of front-line (imatinib) therapy resistance. Subsequently, several hotspot mutations were clinically validated in the kinase domain (KD) of BCR-ABL1, in deterring imatinib sensitivity and further, made targeted by next-generation tyrosine-kinase-inhibitor (TKI) drugs. Identifying KD mutations, occurring even at low frequencies, became pertinent here. Globally, cohorts from different origins were tested and the mutational spectra were mapped to categorize clinical management as well as related pathological features of CML. Moreover, targeted deep sequencing could reveal the mutational landscape more efficiently than the less sensitive Sanger sequencing method. However, no such efforts were reported from Eastern Indian cohorts of imatinib-resistant CML-sufferers. This study assessed a prospective study cohort of imatinib-resistant CML cases from Eastern India. Following dissecting the molecular and clinical parameters, the mutational spectrum was comparatively examined using conventional Sanger and next-generation deep sequencing method. This cohort showed a prevalence of e14a2-p210 variant of BCR-ABL1 and acquired resistance against imatinib, while the disease was mostly confined in its chronic phase. Together with a few common hotspot mutations identified in this cohort, deep sequencing revealed cases with a candidate mutation, otherwise undetermined by Sanger method. Also, cases with a second low frequency mutation were identified upon applying deep sequencing. Along with highlighting a few aspects of CML biology employing an Eastern-Indian cohort, this data could mark the immense importance of deep sequencing to contribute in the clinical management of CML upon front-line therapy resistance.

PTPRJ is downregulated in cervical squamous cell carcinoma.

Squamous cell carcinoma of the uterine cervix (CSCC) is one of the leading causes of death in Indian women. Protein tyrosine phosphatase receptor (PTPR) type J (also known as DEP1) is a recently reported tumour suppressor receptor phosphatase. Critical molecular analysis of PTPRJ/DEP1 (11p11.2) has not performed in CSCC to date. Here, we observed frequent downregulation of cancer samples (n=31) at the transcriptional level. Immunohistochemistry revealed concordant low expression of PTPRJ protein with a few samples showing intermediate expression. To probe for the cause of such downregulation of the gene in CSCC (n=155), we analysed the copy number and promoter methylation of PTPRJ. The genetic locus showed deletion (14.8%) and the promoter showed methylation (33.5%) of PTPRJ. To the best of our knowledge, for the first time we explored the molecular status of PTPRJ although we observed no statistically significant association with the prognosis of Indian CSCC patients (n=76). However, we observed enhanced expression of PTPRJ protein levels that contributes to effective cisplatin chemotherapy in the SiHa cell line. Thus, the present study paves the way for further research into the plausible mechanisms of downregulation of PTPRJ in cervical cancer.

Molecular crosstalk between CUEDC2 and ERα influences the clinical outcome by regulating mitosis in breast cancer.

Development of endocrine resistance in hormone-receptor-positive (HR+ve) subtype and lack of definitive target in triple-negative subtype challenge breast cancer management. Contributing to such endocrine resistance is a protein called CUEDC2. It degrades hormone receptors, estrogen receptor-α (ERα) and progesterone receptor. Higher level of CUEDC2 in ERα+ve breast cancer corresponded to poorer disease prognosis. It additionally influences mitotic progression. However, the crosstalk of these two CUEDC2-driven functions in the outcome of breast cancer remained elusive. We showed that CUEDC2 degrades ERα during mitosis, utilising the mitotic-ubiquitination-machinery. We elucidated the importance of mitosis-specific phosphorylation of CUEDC2 in this process. Furthermore, upregulated CUEDC2 overrode mitotic arrest, increasing aneuploidy. Finally, recruiting a prospective cohort of breast cancer, we found significantly upregulated CUEDC2 in HR-ve cases. Moreover, individuals with higher CUEDC2 levels showed a poorer progression-free-survival. Together, our data confirmed that CUEDC2 up-regulation renders ERα+ve malignancies to behave essentially as HR-ve tumors with the prevalence of aneuploidy. This study finds CUEDC2 as a potential prognostic marker and a therapeutic target in the clinical management of breast cancer.

MiRNA-146a/AKT/ß-Catenin Activation Regulates Cancer Stem Cell Phenotype in Oral Squamous Cell Carcinoma by Targeting CD24.

CD44highCD24low population has been previously reported as cancer stem cells (CSCs) in Oral Squamous Cell Carcinoma (OSCC). Increasing evidence suggests potential involvement of microRNA (miRNA) network in modulation of CSC properties. MiRNAs have thus emerged as crucial players in tumor development and maintenance. However, their role in maintenance of OSCC stem cells remains unclear. Here we report an elevated expression of miR-146a in the CD44highCD24low population within OSCC cells and primary HNSCC tumors. Moreover, over-expression of miR-146a results in enhanced stemness phenotype by augmenting the CD44highCD24low population. We demonstrate that miR-146a stabilizes ß-catenin with concomitant loss of E-cadherin and CD24. Interestingly, CD24 is identified as a novel functional target of miR-146a and ectopic expression of CD24 abrogates miR-146a driven potential CSC phenotype. Mechanistic analysis reveals that higher CD24 levels inhibit AKT phosphorylation leading to ß-catenin degradation. Using stably expressing miR-146a/CD24 OSCC cell lines, we also validate that the miR-146a/CD24/AKT loop significantly alters tumorigenic ability in vivo. Furthermore, we confirmed that ß-catenin trans-activates miR-146a, thereby forming a positive feedback loop contributing to stem cell maintenance. Collectively, our study demonstrates that miR-146a regulates CSCs in OSCC through CD24-AKT-ß-catenin axis.

A comprehensive meta-analysis and a case-control study give insights into genetic susceptibility of lung cancer and subgroups.

Reports of genetic association of polymorphisms with lung cancer in the Indian subcontinent are often conflicting. To summarise and replicate published evidence for association with lung cancer and its subgroups. We performed a meta-analysis of candidate associations on lung cancer, its histological subtypes and smoking status in the Indian subcontinent following PRISMA guidelines. Multiple testing corrections were done by the Benjamini-Hochberg method through assessment of significance at a false discovery rate of 10%. We genotyped and investigated rs1048943/CYP1A1 in a case-control sample from eastern India, followed by its global meta-analysis using a similar protocol. Meta-analysis of 18 variants of 11 genes reported in 39 studies (7630 cases and 8169 controls) showed significant association of rs1048943/CYP1A1 [2.07(1.49-2.87)] and rs4646903/CYP1A1 [1.48(1.93-1.95)] with overall lung cancer risk at 10% FDR, while nominal association (p < 0.05) was observed for del1/GSTT1, del2/GSTM1, rs1695/GSTP1 and rs17037102/ DKK2. Subtype analysis showed a significant association of del1/GSTT1 with adenocarcinoma, rs4646903/CYP1A1 with squamous carcinoma, and rs1048943/CYP1A1 with both. Association of rs4646903/CYP1A1 in smokers and effect modification by meta-regression analysis was observed. Genotyping of rs1048943/CYP1A1 that presented significant heterogeneity (p < 0.1) revealed an association with adenocarcinoma among eastern Indian smokers, while a global meta-analysis in 10458 cases and 10871 controls showed association with lung cancer and its subgroups. This study identified the susceptibility loci for lung cancer and its covariate-subgroups.

MicroRNA-324-5p-CUEDC2 Axis Mediates Gain-of-Function Mutant p53-Driven Cancer Stemness.

Regulation of cancer stemness has recently emerged as a new gain-of-function (GOF) property of mutant p53. In this study, we identify miR-324-5p as a critical epigenetic regulator of cancer stemness and demonstrate its role in mediating GOF-mutant p53-driven stemness phenotypes. We report that miR-324-5p is upregulated in human cancer cell lines and non-small cell lung carcinoma (NSCLC) tumors carrying TP53 GOF mutations. Mechanistically, we show that GOF mutant p53 upregulates miR-324-5p expression via c-Myc, an oncogenic transcription factor in cancer cells. Our experimental results suggest that miR-324-5p-induced CSC phenotypes stem from the downregulation of CUEDC2, a downstream target gene of miR-324-5p. Accordingly, CUEDC2 complementation diminishes elevated CSC marker expression in miR-324-5p-overexpressing cancer cells. We further demonstrate that mutant p53 cancer cells maintain a low level of CUEDC2 that is rescued upon miR-324-5p inhibition. Importantly, we identify CUEDC2 downregulation as a novel characteristic feature of TP53-mutated human cancers. We show that activation of NF-κB due to downregulation of CUEDC2 by miR-324-5p imparts stemness in GOF mutant p53 cancer cells. Finally, we provide evidence that TP53 mutations coupled with high miR-324-5p expression predict poor prognosis in patients with lung adenocarcinoma. Thus, our study delineates an altered miR-324-5p-CUEDC2-NF-κB pathway as a novel regulator of GOF mutant p53-driven cancer stemness. IMPLICATIONS: Our findings implicate miRNA-324-5p as a novel epigenetic modifier of human cancer stemness.

Differences in the bacteriome of swab, saliva, and tissue biopsies in oral cancer.

Microbial dysbiosis has been implicated in the pathogenesis of oral cancer. We analyzed the compositional and metabolic profile of the bacteriome in three specific niches in oral cancer patients along with controls using 16SrRNA sequencing (Illumina Miseq) and DADA2 software. We found major differences between patients and control subjects. Bacterial communities associated with the tumor surface and deep paired tumor tissue differed significantly. Tumor surfaces carried elevated abundances of taxa belonging to genera Porphyromonas, Enterobacteriae, Neisseria, Streptococcus and Fusobacteria, whereas Prevotella, Treponema, Sphingomonas, Meiothermus and Mycoplasma genera were significantly more abundant in deep tissue. The most abundant microbial metabolic pathways were those related to fatty-acid biosynthesis, carbon metabolism and amino-acid metabolism on the tumor surface: carbohydrate metabolism and organic polymer degradation were elevated in tumor tissues. The bacteriome of saliva from patients with oral cancer differed significantly from paired tumor tissue in terms of community structure, however remained similar at taxonomic and metabolic levels except for elevated abundances of Streptococcus, Lactobacillus and Bacteroides, and acetoin-biosynthesis, respectively. These shifts to a pro-inflammatory profile are consistent with other studies suggesting oncogenic properties. Importantly, selection of the principal source of microbial DNA is key to ensure reliable, reproducible and comparable results in microbiome studies.

Integrative genomics and pathway analysis identified prevalent FA-BRCA pathway alterations in arsenic-associated urinary bladder carcinoma: Chronic arsenic accumulation in cancer tissues hampers the FA-BRCA pathway.

Arsenic in drinking water is one of the major etiological factors in urinary bladder carcinoma (BlCa). Here, high-resolution CGH-SNP microarray analysis in arsenic accumulated BlCa tissues showed significant (p < 0.05) association of chromosomal alterations with high arsenic (≥112 ng/g) accumulation, further corroborated by high γH2AX nuclear expression. Cytobands 5q11-35, 9p24.3-21.5, 18q11.1-25, etc. showed deletion, whereas 12q was amplified in high arsenic samples (AsH). Consecutively, IPA® found FA-BRCA pathway to be exclusively altered in AsH group. Validation of several key regulatory genes (RAD50, BRIP1, UIMC1, FANCD2, BRCA2 and BRCA1) of the pathway, were performed in independent BlCa cases (n = 81). UIMC1, RAD50 and BRIP1 were differentially deleted and associated with poor survival of AsH samples. Moreover, reduced nuclear expression with diffused cytoplasmic expression of FANCD2 was higher in AsH samples. Collectively, frequent deregulation of RAD50, UIMC1 and BRIP1 may result in reduced nuclear translocation of FANCD2, which may cause more chromosomal aberrations among AsH samples.

Functional Landscape of Dysregulated MicroRNAs in Oral Squamous Cell Carcinoma: Clinical Implications.

MicroRNA (miRNA) dysregulation is associated with the pathogenesis of oral squamous cell carcinoma (OSCC), and its elucidation could potentially provide information on patient outcome. A growing body of translational research on miRNA biology is focusing on precision oncology, aiming to decode the miRNA regulatory network in the development and progression of cancer. Tissue-specific expression and stable presence in all body fluids are unique features of miRNAs, which could be potentially exploited in the clinical setting. Recent understanding of miRNA properties has led them to be useful, attractive, and potential tools either as biomarkers (distinct miRNA expression signature) for diagnosis and prognostic outcomes or as targets for novel therapeutic entities, enabling personalized treatment for OSCC. In this review, we discuss recent research on different aspects of alterations in miRNA profiles along with their clinical significance and strive to identify probable potential miRNA biomarkers for diagnosis and prognosis of OSCC. We also discuss the current understanding and scope of development of miRNA-based therapeutics against OSCC.

Downregulation of beta-catenin in chemo-tolerant TNBC through changes in receptor and antagonist profiles of the WNT pathway: Clinical and prognostic implications.

PURPOSE: In approximately 30% of triple-negative breast cancer (TNBC) patients a complete pathological response is achieved. However, after neo-adjuvant chemotherapy treatment (NACT) residual tumour cells can be intrinsically resistant to chemotherapy. In this study, associations of the WNT/beta-catenin pathway with chemo-tolerance of NACT treated TNBC patients were compared to that of pre-treatment TNBC patients. METHODS: Expression analyses were performed in both pre-treatment and NACT treated TNBC samples using immunohistochemistry and qRT-PCR, along with DNA copy number variation (CNV) and promoter methylation analyses to elucidate the mechanism(s) underlying chemo-tolerance. In addition, in vitro validation experiments were performed in TNBC cells followed by in vivo clinicopathological correlation analyses. RESULTS: A reduced expression (41.1%) of nuclear beta-catenin together with a low proliferation index was observed in NACT samples, whereas a high expression (59.0%) was observed in pre-treatment samples. The reduced nuclear expression of beta-catenin in the NACT samples showed concordance with reduced expression levels (47-52.9%) of its associated receptors (FZD7 and LRP6) and increased expression levels (35.2-41.1%) of its antagonists (SFRP1, SFRP2, DKK1) compared to those in the pre-treatment samples. The expression levels of the receptors showed no concordance with its respective gene copy number/mRNA expression statuses, regardless treatment. Interestingly, however, significant increases in promoter hypomethylation of the antagonists were observed in the NACT samples compared to the pre-treatment samples. Similar expression patterns of the antagonists, receptors and beta-catenin were observed in the TNBC-derived cell line MDA-MB-231 using the anthracyclines doxorubicin and nogalamycin, suggesting the importance of promoter hypomethylation in chemotolerance. NACT patients showing reduced receptor and/or beta-catenin expression levels and high antagonist expression levels exhibited a comparatively better prognosis than the pre-treatment patients. CONCLUSIONS: Our data suggest that reduced nuclear expression of beta-catenin in NACT TNBC samples, due to downregulation of its receptors and upregulation of its antagonists through promoter hypomethylation of the WNT pathway, plays an important role in chemo-tolerance.

sigFeature: Novel Significant Feature Selection Method for Classification of Gene Expression Data Using Support Vector Machine and t Statistic.

Biological data are accumulating at a faster rate, but interpreting them still remains a problem. Classifying biological data into distinct groups is the first step in understanding them. Data classification in response to a certain treatment is an extremely important aspect for differentially expressed genes in making present/absent calls. Many feature selection algorithms have been developed including the support vector machine recursive feature elimination procedure (SVM-RFE) and its variants. Support vector machine RFEs are greedy methods that attempt to find superlative possible combinations leading to binary classification, which may not be biologically significant. To overcome this limitation of SVM-RFE, we propose a novel feature selection algorithm, termed as "sigFeature" (https://bioconductor.org/packages/sigFeature/), based on SVM and t statistic to discover the differentially significant features along with good performance in classification. The "sigFeature" R package is centered around a function called "sigFeature," which provides automatic selection of features for the binary classification. Using six publicly available microarray data sets (downloaded from Gene Expression Omnibus) with different biological attributes, we further compared the performance of "sigFeature" to three other feature selection algorithms. A small number of selected features (by "sigFeature") also show higher classification accuracy. For further downstream evaluation of its biological signature, we conducted gene set enrichment analysis with the selected features (genes) from "sigFeature" and compared it with the outputs of other algorithms. We observed that "sigFeature" is able to predict the signature of four out of six microarray data sets accurately, whereas the other algorithms predict less data set signatures. Thus, "sigFeature" is considerably better than related algorithms in discovering differentially significant features from microarray data sets.