Adenomatous Polyposis Coli loss controls cell cycle regulators and response to paclitaxel in MDA-MB-157 metaplastic breast cancer cells.

Adenomatous Polyposis Coli (APC) is lost in approximately 70% of sporadic breast cancers, with an inclination towards triple negative breast cancer (TNBC). TNBC is treated with traditional chemotherapy, such as paclitaxel (PTX); however, tumors often develop drug resistance. We previously created APC knockdown cells (APC shRNA1) using the human TNBC cells, MDA-MB-157, and showed that APC loss induces PTX resistance. To understand the mechanisms behind APC-mediated PTX response, we performed cell cycle analysis and analyzed cell cycle related proteins. Cell cycle analysis indicated increased G2/M population in both PTX-treated APC shRNA1 and parental cells, suggesting that APC expression does not alter PTX-induced G2/M arrest. We further studied the subcellular localization of the G2/M transition proteins, cyclin B1 and CDK1. The APC shRNA1 cells had increased CDK1, which was preferentially localized to the cytoplasm, and increased baseline CDK6. RNA-sequencing was performed to gain a global understanding of changes downstream of APC loss and identified a broad mis-regulation of cell cycle-related genes in APC shRNA1 cells. Our studies are the first to show an interaction between APC and taxane response in breast cancer. The implications include designing combination therapy to re-sensitize APC-mutant breast cancers to taxanes using the specific cell cycle alterations.

Neonatal thyroxine activation modifies epigenetic programming of the liver.

The type 2 deiodinase (D2) in the neonatal liver accelerates local thyroid hormone triiodothyronine (T3) production and expression of T3-responsive genes. Here we show that this surge in T3 permanently modifies hepatic gene expression. Liver-specific Dio2 inactivation (Alb-D2KO) transiently increases H3K9me3 levels during post-natal days 1-5 (P1-P5), and results in methylation of 1,508 DNA sites (H-sites) in the adult mouse liver. These sites are associated with 1,551 areas of reduced chromatin accessibility (RCA) within core promoters and 2,426 within intergenic regions, with reduction in the expression of 1,363 genes. There is strong spatial correlation between density of H-sites and RCA sites. Chromosome conformation capture (Hi-C) data reveals a set of 81 repressed genes with a promoter RCA in contact with an intergenic RCA ~300 Kbp apart, within the same topologically associating domain (χ2 = 777; p < 0.00001). These data explain how the systemic hormone T3 acts locally during development to define future expression of hepatic genes.

Building and Interpreting Artificial Neural Network Models for Biological Systems.

Biology has become a data driven science largely due to the technological advances that have generated large volumes of data. To extract meaningful information from these data sets requires the use of sophisticated modeling approaches. Toward that, artificial neural network (ANN) based modeling is increasingly playing a very important role. The "black box" nature of ANNs acts as a barrier in providing biological interpretation of the model. Here, the basic steps toward building models for biological systems and interpreting them using calliper randomization approach to capture complex information are described.

Statistical and artificial neural network-based analysis to understand complexity and heterogeneity in preeclampsia.

Preeclampsia is a pregnancy associated disease. It is characterized by high blood pressure and symptoms that are indicative of damage to other organ systems, most often involving the liver and kidneys. If left untreated, the condition could be fatal to mother and baby. This makes it important to delineate the complexities associated with the disease at a molecular level that would help develop methods for early diagnosis. In microarray-based studies, Textoris et al. and Mirzakhani et al. have analyzed the transcriptome with a view to identify biomarkers for preeclampsia. The current study has extensively analyzed these microarray data sets to understand the complexity and heterogeneity associated with preeclampsia. A statistical multiple comparisons-based approach has been used to identify features capable of distinguishing preeclampsia from normotensive cases. These features were then used to build an artificial neural network-based machine learning model that successfully classified the samples. Further, the machine learning model was used to delineate features critical for its internal representation by extending the calliper randomization approach to the analysis of microarray data. Functional analysis of the features identified by the calliper randomization approach revealed pathways that could be crucially involved in the mechanism of the underlying disease. Biological processes associated with the features identified have revealed among others, genes involved in reproductive processes to be differentially expressed.

Analysis of isoform expression from splicing array using multiple comparisons.

There is a high prevalence of alternatively spliced genes (isoforms) in the human genome. Studies toward understanding aberrantly spliced genes and their association with diseases have lead researchers to profile the expression of alternatively spliced products. High-throughput profiling of isoforms has been done using microarray technology. Expression of isoforms reflects regulation both at transcriptional and posttranscriptional levels. This chapter details the methods to perform exhaustive comparison of isoforms using the R statistical framework.

Sequence periodicity in nucleosomal DNA and intrinsic curvature.

BACKGROUND: Most eukaryotic DNA contained in the nucleus is packaged by wrapping DNA around histone octamers. Histones are ubiquitous and bind most regions of chromosomal DNA. In order to achieve smooth wrapping of the DNA around the histone octamer, the DNA duplex should be able to deform and should possess intrinsic curvature. The deformability of DNA is a result of the non-parallelness of base pair stacks. The stacking interaction between base pairs is sequence dependent. The higher the stacking energy the more rigid the DNA helix, thus it is natural to expect that sequences that are involved in wrapping around the histone octamer should be unstacked and possess intrinsic curvature. Intrinsic curvature has been shown to be dictated by the periodic recurrence of certain dinucleotides. Several genome-wide studies directed towards mapping of nucleosome positions have revealed periodicity associated with certain stretches of sequences. In the current study, these sequences have been analyzed with a view to understand their sequence-dependent structures. RESULTS: Higher order DNA structures and the distribution of molecular bend loci associated with 146 base nucleosome core DNA sequence from C. elegans and chicken have been analyzed using the theoretical model for DNA curvature. The curvature dispersion calculated by cyclically permuting the sequences revealed that the molecular bend loci were delocalized throughout the nucleosome core region and had varying degrees of intrinsic curvature. CONCLUSIONS: The higher order structures associated with nucleosomes of C.elegans and chicken calculated from the sequences revealed heterogeneity with respect to the deviation of the DNA axis. The results points to the possibility of context dependent curvature of varying degrees to be associated with nucleosomal DNA.

On selecting mRNA isoform features for profiling prostate cancer.

Alternative splicing of human pre-mRNA is a very common phenomenon and is a major contributor to proteome diversity. mRNA isoforms that arise as a result of alternative splicing also provide a more complete picture of the transcriptome as they reflect the additional processing a pre-mRNA undergoes before being translated into a functional product. It has been reported that molecular alterations of cells can occur as a result of the differential expression of mRNA isoforms, resulting in cancerous or normal tissue. Quantification of mRNA isoforms can thus be used as a better indicator in distinguishing a normal tissue from a cancerous tissue. In our earlier study we had used mRNA isoforms expression to identify biomarkers for prostate cancer (Li et. al, 2006. Cancer Res. 66 (8) 4079-4088). Here we have used statistical methods of multiple comparison and have developed a simple scoring scheme to extract isoform features. Further, we have rigorously analyzed the isoform expression data to understand the variability and heterogeneity associated with the expression levels between (i) prostate cancer cell lines and non-prostate cancer cell lines and (ii) normal prostate tissue and prostate cancer tissue. We found that there were several isoforms that showed significant difference in expression within the same class. We were also able to successfully identify isoforms with similar changes in expression levels, that when used as features for classification was able to provide robust class separation. The features selected using the multiple comparison methods had subsets that were common and disparate with those that were selected using statistical t-tests. This reveals the importance of selecting features using a combination of complementary methods.

Two-dimensional transcriptome profiling: identification of messenger RNA isoform signatures in prostate cancer from archived paraffin-embedded cancer specimens.

The expression of specific mRNA isoforms may uniquely reflect the biological state of a cell because it reflects the integrated outcome of both transcriptional and posttranscriptional regulation. In this study, we constructed a splicing array to examine approximately 1,500 mRNA isoforms from a panel of genes previously implicated in prostate cancer and identified a large number of cell type-specific mRNA isoforms. We also developed a novel "two-dimensional" profiling strategy to simultaneously quantify changes in splicing and transcript abundance; the results revealed extensive covariation between transcription and splicing in prostate cancer cells. Taking advantage of the ability of our technology to analyze RNA from formalin-fixed, paraffin-embedded tissues, we derived a specific set of mRNA isoform biomarkers for prostate cancer using independent panels of tissue samples for feature selection and cross-analysis. A number of cancer-specific splicing switch events were further validated by laser capture microdissection. Quantitative changes in transcription/RNA stability and qualitative differences in splicing ratio may thus be combined to characterize tumorigenic programs and signature mRNA isoforms may serve as unique biomarkers for tumor diagnosis and prognosis.

MAASE: an alternative splicing database designed for supporting splicing microarray applications.

Alternative splicing is a prominent feature of higher eukaryotes. Understanding of the function of mRNA isoforms and the regulation of alternative splicing is a major challenge in the post-genomic era. The development of mRNA isoform sensitive microarrays, which requires precise splice-junction sequence information, is a promising approach. Despite the availability of a large number of mRNAs and ESTs in various databases and the efforts made to align transcript sequences to genomic sequences, existing alternative splicing databases do not offer adequate information in an appropriate format to aid in splicing array design. Here we describe our effort in constructing the Manually Annotated Alternatively Spliced Events (MAASE) database system, which is specifically designed to support splicing microarray applications. MAASE comprises two components: (1) a manual/computational annotation tool for the efficient extraction of critical sequence and functional information for alternative splicing events and (2) a user-friendly database of annotated events that allows convenient export of information to aid in microarray design and data analysis. We provide a detailed introduction and a step-by-step user guide to the MAASE database system to facilitate future large-scale annotation efforts, integration with other alternative splicing databases, and splicing array fabrication.

Characterization of the yeast ionome: a genome-wide analysis of nutrient mineral and trace element homeostasis in Saccharomyces cerevisiae.

BACKGROUND: Nutrient minerals are essential yet potentially toxic, and homeostatic mechanisms are required to regulate their intracellular levels. We describe here a genome-wide screen for genes involved in the homeostasis of minerals in Saccharomyces cerevisiae. Using inductively coupled plasma-atomic emission spectroscopy (ICP-AES), we assayed 4,385 mutant strains for the accumulation of 13 elements (calcium, cobalt, copper, iron, potassium, magnesium, manganese, nickel, phosphorus, selenium, sodium, sulfur, and zinc). We refer to the resulting accumulation profile as the yeast 'ionome'. RESULTS: We identified 212 strains that showed altered ionome profiles when grown on a rich growth medium. Surprisingly few of these mutants (four strains) were affected for only one element. Rather, levels of multiple elements were altered in most mutants. It was also remarkable that only six genes previously shown to be involved in the uptake and utilization of minerals were identified here, indicating that homeostasis is robust under these replete conditions. Many mutants identified affected either mitochondrial or vacuolar function and these groups showed similar effects on the accumulation of many different elements. In addition, intriguing positive and negative correlations among different elements were observed. Finally, ionome profile data allowed us to correctly predict a function for a previously uncharacterized gene, YDR065W. We show that this gene is required for vacuolar acidification. CONCLUSION: Our results indicate the power of ionomics to identify new aspects of mineral homeostasis and how these data can be used to develop hypotheses regarding the functions of previously uncharacterized genes.

Rival penalized competitive learning (RPCL): a topology-determining algorithm for analyzing gene expression data.

DNA arrays have become the immediate choice in the analysis of large-scale expression measurements. Understanding the expression pattern of genes provide functional information on newly identified genes by computational approaches. Gene expression pattern is an indicator of the state of the cell, and abnormal cellular states can be inferred by comparing expression profiles. Since co-regulated genes, and genes involved in a particular pathway, tend to show similar expression patterns, clustering expression patterns has become the natural method of choice to differentiate groups. However, most methods based on cluster analysis suffer from the usual problems (i) dead units, and (ii) the problem of determining the correct number of clusters (k) needed to classify the data. Selecting the k has been an open problem of pattern recognition and statistics for decades. Since clustering reveals similar patterns present in the data, fixing this number strongly influences the quality of the result. While there is no theoretical solution to this problem, the number of clusters can be decided by a heuristic clustering algorithm called rival penalized competitive learning (RPCL). We present a novel implementation of RPCL that transforms the correct number of clusters problem to the tractable problem of clustering based on the degree of similarity. This is biologically significant since our implementation clusters functionally co-regulated genes and genes that present similar patterns of expression. This new approach reveals potential genes that are co-involved in a biological process. This implementation of the RPCL algorithm is useful in differentiating groups involved in concerted functional regulation and helps to progressively home into patterns, which are closely similar.

The PlantsP and PlantsT Functional Genomics Databases.

PlantsP and PlantsT allow users to quickly gain a global understanding of plant phosphoproteins and plant membrane transporters, respectively, from evolutionary relationships to biochemical function as well as a deep understanding of the molecular biology of individual genes and their products. As one database with two functionally different web interfaces, PlantsP and PlantsT are curated plant-specific databases that combine sequence-derived information with experimental functional-genomics data. PlantsP focuses on proteins involved in the phosphorylation process (i.e., kinases and phosphatases), whereas PlantsT focuses on membrane transport proteins. Experimentally, PlantsP provides a resource for information on a collection of T-DNA insertion mutants (knockouts) in each kinase and phosphatase, primarily in Arabidopsis thaliana, and PlantsT uniquely combines experimental data regarding mineral composition (derived from inductively coupled plasma atomic emission spectroscopy) of mutant and wild-type strains. Both databases provide extensive information on motifs and domains, detailed information contributed by individual experts in their respective fields, and descriptive information drawn directly from the literature. The databases incorporate a unique user annotation and review feature aimed at acquiring expert annotation directly from the plant biology community. PlantsP is available at http://plantsp.sdsc.edu and PlantsT is available at http://plantst.sdsc.edu.

On selecting features from splice junctions: an analysis using information theoretic and machine learning approaches.

The computational recognition of precise splice junctions is a challenge faced in the analysis of newly sequenced genomes. This is challenging due to the fact that the distribution of sequence patterns in these regions is not always distinct. Our objective is to understand the sequence signatures at the splice junctions, not simply to create an artificial recognition system. We use a combination of a neural network based calliper randomization approach and an information theoretic based feature selection approach for this purpose. This has been done in an effort to understand regions that harbor information content and to extract features relevant for the prediction of splice junctions. The analysis using the neural network based calliper randomization approach revealed regions important in the internal representation of the network model. The calliper approach captured both correlated as well as independently important features. The feature selection approach captures features that are independently informative. The two different methods can capture features with different properties. Comparative analysis of the results using both the methods help to infer about the kind of information present in the region.