Introduction and methods: Hepatocellular carcinoma (HCC) remains a difficult disease to study even after a decade of genomic analysis. Metabolic and cell-cycle perturbations are known, large changes in tumors that add little to our understanding of the development of tumors, but generate “noise” that obscures potentially important smaller scale expression changes in “driver genes”. Recently, some researchers have suggested that HCC shares pathways involving the master regulators of embryonic development. Here, we investigated the involvement and specificity of developmental genes in HCV-cirrhosis and HCV-HCC. We obtained microarray studies from 30 patients with HCV-cirrhosis and 49 patients with HCV-HCC and compared to 12 normal livers. Differential gene expression is specific to liver development genes: 86 of 202 (43%) genes specific to liver development had differential expression between normal and cirrhotic or HCC samples. Of 60 genes with paralogous function, which are specific to development of other organs and have known associations with other cancer types, none were expressed in either adult normal liver or tumor tissue. Developmental genes are widely differentially expressed in both cirrhosis and early HCC, but not late HCC: 69 liver development genes were differentially expressed in cirrhosis, and 58 of these (84%) were also dysregulated in early HCC. 19/58 (33%) had larger-magnitude changes in cirrhosis and 5 (9%) had larger-magnitude changes in early HCC. 16 (9%) genes were uniquely altered in early tumors, while only 2 genes were uniquely changed in late-stage (T3 and T4) HCC. Together, these results suggest that the involvement of the master regulators of liver development are active in the pre-cancerous cirrhotic liver and in cirrhotic livers with emerging tumors but play a limited role in the transition from early to late stage HCC. Common patterns of coordinated developmental gene expression include: (1) Dysregulation of BMP2 signaling in cirrhosis followed by overexpression of BMP inhibitors in HCC. BMP inhibitor GPC3 was overexpressed in nearly all tumors, while GREM1 was associated specifically with recurrence-free survival after ablation and transplant. (2) Cirrhosis tissues acquire a progenitor-like signature including high expression of Vimentin, EPCAM, and KRT19, and these markers remain over-expressed to a lesser extent in HCC. (3) Hepatocyte proliferation inhibitors (HPI) E-cadherin (CDH1), BMP2, and MST1 were highly expressed in cirrhosis and remained over-expressed in 16 HCC patients who were transplanted with excellent recurrence-free survival (94% survival after 2 years; mean recurrence-free survival = 5.6 yrs), while loss in early HCC was associated with early recurrence and (2 year). Loss of HPI overexpression was also correlated with overexpression of c-MET and loss of STAT3, LAMA2, FGFR2, CITED2, KIT, SMAD7, GATA6, ERBB2, and NOTCH2.

In spite of many efforts in the past, inference or reverse engineering of regulatory networks from microarray data remains an unsolved problem in the area of systems biology. Such regulatory networks play a critical role in cellular function and organization and are of interest in the study of a variety of disease areas and ecotoxicology to name a few. This dissertation proposes information theoretic methods/algorithms for inferring regulatory networks from microarray data. Most of the algorithms proposed in this dissertation can be implemented both on time series and multifactorial microarray data sets. The work proposed here infers regulatory networks considering the following six factors: (i) computational efficiency to infer genome-scale networks, (ii) incorporation of prior biological knowledge, (iii) choosing the optimal network that minimizes the joint network entropy, (iv) impact of higher order structures (specifically 3-node structures) on network inference (v) effects of the time sensitivity of regulatory interactions and (vi) exploiting the benefits of existing/proposed metrics and algorithms for reverse engineering using the concept of consensus of consensus networks. Specifically, this dissertation presents an approach towards incorporating knock-out data sets. The proposed method for incorporating knock-out data sets is flexible so that it can be easily adapted in existing/new approaches. While most of the information theoretic approaches infer networks based on pair-wise interactions this dissertation discusses inference methods that consider scoring edges from complex structures. A new inference method for building consensus networks based on networks inferred by multiple popular information theoretic approaches is also proposed here. For time-series datasets, new information theoretic metrics were proposed considering the time-lags of regulatory interactions estimated from microarray datasets. Finally, based on the scores predicted for each possible edge in the network, a probabilistic minimum description length based approach was proposed to identify the optimal network (minimizing the joint network entropy). Comparison analysis on in-silico and/or real time data sets have shown that the proposed algorithms achieve better inference accuracy and/or higher computational efficiency as compared with other state-of-the-art schemes such as ARACNE, CLR and Relevance Networks. Most of the methods proposed in this dissertation are generalized and can be easily incorporated into new methods/algorithms for network inference.

Alcoholism is a complex neurological disorder characterized by loss of control in limiting intake, compulsion to seek and imbibe ethanol, and chronic craving and relapse. It is suggested that the characteristic behaviors associated with the escalation of drug use are caused by long-term molecular adaptations precipitated by the drug’s continual administration. These lasting activity-dependent changes that underlie addiction-associated behavior are thought, in part, to depend on new protein synthesis and remodeling at the synapses. It is well established that mRNA can be transported to neuronal distal processes, where it can undergo localized translation that is regulated in a spatially restricted manner in response to stimulation. Through two avenues of investigation, the research herein demonstrates that behavioral responses to ethanol result, at least in part, from alterations in the synaptic transcriptome which contribute to synaptic remodeling and plasticity. The synaptoneurosome preparation was utilized to enrich for RNAs trafficked to the synapse. Two complementary methods of genomic profiling, microarrays and RNA-Seq, were used to survey the synaptic transcriptome of DBA/2J mice subjected to ethanol-induced behavioral sensitization. A habituating expression profile, characteristic of glucocorticoid-responsive genes, was observed for a portion of synaptically targeted genes determined to be sensitive to repeated ethanol exposure. Other ethanol-responsive genes significantly enriched for at the synapse were related to biological functions such as protein folding and extra-cellular matrix components, suggesting a role for local regulation of synaptic functioning by ethanol. In a separate series of experiments, it was shown that altered trafficking of Bdnf, an ethanol-responsive gene, resulted in aberrant ethanol behavioral phenotypes. In particular, mice lacking dendritically targeted Bdnf mRNA exhibited enhanced sensitivity to low, activating doses and high, sedating doses of ethanol. Together these experiments suggest that ethanol has local regulatory effects at the synapse and lays the foundation for further investigations into the role of the synaptic transcriptome in ethanol-responsive behaviors. Supported by NIAA grants R01AA014717, U01 AA016667 and P20AA017828 to MFM, F31AA021035 to MAO, and NIDA T32DA007027 to WLD.

The analysis of Affymetrix GeneChip® data is a complex, multistep process. Most often, methodscondense the multiple probe level intensities into single probeset level measures (such as RobustMulti-chip Average (RMA), dChip and Microarray Suite version 5.0 (MAS5)), which are thenfollowed by application of statistical tests to determine which genes are differentially expressed. An alternative approach is a probe-level analysis, which tests for differential expression directly using the probe-level data. Probe-level models offer the potential advantage of more accurately capturing sources of variation in microarray experiments. However, this has not been thoroughly investigated, since current research efforts have largely focused on the development of improved expression summary methods. This research project will review current approaches to analysis of probe-level data and discuss extensions of two examples, the S-Score and the Random Variance Model (RVM). The S-Score is a probe-level algorithm based on an error model in which the detected signal is proportional to the probe pair signal for highly expressed genes, but approaches a background level (rather than 0) for genes with low levels of expression. Initial results with the S-Score have been promising, but the method has been limited to two-chip comparisons. This project presents extensions to the S-Score that permit comparisons of multiple chips and "borrowing" of information across probes to increase statistical power. The RVM is a probeset-level algorithm that models the variance of the probeset intensities as a random sample from a common distribution to "borrow" information across genes. This project presents extensions to the RVM for probe-level data, using multivariate statistical theory to model the covariance among probes in a probeset. Both of these methods show the advantages of probe-level, rather than probeset-level, analysis in detecting differential gene expression for Afymetrix GeneChip data. Future research will focus on refining the probe-level models of both the S-Score and RVM algorithms to increase the sensitivity and specificity of microarray experiments.

RNA degradation plays a fundamental role in maintaining cellular homeostasis, along with being a part of normal regulatory mechanisms, whether it occurs as a surveillance mechanism eliminating aberrant mRNAs or during RNA processing to generate mature transcripts. 3’-5’ exoribonucleases are essential mediators of RNA decay pathways, and one such evolutionarily conserved enzyme is polynucleotide phosphorylase (PNPase). The human homologue of this fascinating enzymatic protein (hPNPaseold-35) was cloned a decade ago in the context of terminal differentiation and senescence through a novel ‘overlapping pathway screening’ approach. Since then, significant insights have been garnered about this exoribonuclease and its repertoire of expanding functions. hPNPaseold-35 has progressed a long way from being just a 3’-5’ exoribonuclease to a functionally relevant molecule implicated in a multitude of diverse and important biological effects. hPNPaseold-35 plays central roles in diverse physiological processes including growth inhibition, senescence, mtRNA import, mitochondrial homeostasis, and RNA degradation, all while primarily being localized in the mitochondrial IMS (inter membrane space). hPNPaseold-35 also holds immense promise as a therapeutic agent due to its ability to degrade specific miRNA (miR-221) and mRNA (c-myc) species, and this property can be exploited in treating malignancies that are characterized by upregulation of harmful miRNA or mRNA molecules. But apart from these two targets, little is known about any other targets hPNPaseold-35 may degrade. Thus, the primary objective of this dissertation research was to identify targets other than c-myc or miR-221 that hPNPaseold-35 could directly degrade to discover newer and biologically relevant therapeutic targets for the treatment of hPNPaseold-35 –associated disease states. In order to do this we performed extensive microarray analyses following hPNPaseold-35 overexpression and depletion in mammalian cell lines, and were able to identify transcripts that could be potentially regulated by hPNPaseold-35 directly or indirectly. Apart from this we also analyzed the 3’UTR of c-myc in order to identify any specific sequence or secondary structural elements necessary for hPNPaseold-35 mediated degradation. Lastly, we identified certain residues in hPNPaseold-35 that have been under positive natural selection through evolution.

DNA microarrays represent an ultra-high throughput gene expression assay employed to study the transcriptomic profiles of biological tissues. These devices are increasingly being used to study many aspects of gene regulation, and there is growing interest in the biotechnology and pharmaceutical industries for developing such devices in efforts toward rational product/drug design. The DNA microarray also provides a unique and objective means for diagnosis and prognosis of human diseases based on patterns of gene expression. This is especially important in cancer research and the thrust toward personalized medicine. This dissertation details the design and development of oligonucleotide microarrays and the design and execution of a gene expression study conducted using human glioma specimines. Chapter 2 details the design and development a ~10,000 gene human oligonucleotide microarray. This device consisted of a 21,168 features, each composed of a particular human gene-probe and was applied to the challenge of diagnostic and prognostic estimation for human gliomas (chapter 3). Gliomas are the most frequent and deadly neoplasms of the human brain characterized by a high misdiagnosis rate and low survival. The study in chapter 3 demonstrated that the specified design and development parameters were appropriate for conducting gene expression analysis and that this platform can be used successfully to predict malignancy grade and survival for glioma patients.

ABSTRACT MOLECULAR MECHANISMS AND GENE SIGNATURES INVOVLED IN CALCINEURIN INHIBITOR NEPHROTOXICITY IN KIDNEY ALLOGRAFT By Huong Le Diem Nguyen, M.S. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Physiology at Virginia Commonwealth University. Virginia Commonwealth University, 2011. Major Director: Valeria Mas, Ph.D. Associate Professor, Department of Surgery and Pathology Director of Molecular Transplant Research Laboratory, Division of Transplant Calcineurin inhibitors (CNI), cyclosporin A and tacrolimus, are potent immunosuppressive agents but induce toxicities causing damages and graft dysfunction, and have been suggested to contribute to late-term loss of graft in kidney transplant recipients. Even though insights on mechanism of CNI nephrotoxicity have been uncovered, prevention and treatment of these toxicities remain a major challenge in the clinical administration of CNI due to low dose-toxicity correlation, difficulty in establishing a differential patho-histological diagnosis, and varying individual susceptibility. We hypothesize that CNI nephrotoxicity follows distinct disease pathways and is characterized by significant gene signatures that differentiate it from other conditions such as acute rejection and chronic allograft dysfunction. Moreover, we postulate that CNI-induced toxicity profiles contribute to the IF/TA signatures. Microarray analysis and gene annotation were done on the study database included of tissues diagnosed with CNI nephrotoxicity (n = 9), interstitial fibrosis/tubular atrophy (IF/TA, n=10), and normal allografts (NA, n = 8). All samples were histologically classified based on the revised Banff ‘07 criteria for renal allograft pathology. Top-scored biological networks in CNI tissues were related to metabolic disease, cellular development, renal necrosis, apoptosis cell-death, immunological disease, inflammatory disease, and many others. Canonical pathway analysis emphasized oxidative stress response mediated by NRF2 and various cell-death signaling pathways including 14-3-3 signaling pathway, p53 signaling pathway, and TGF-β signaling pathway. Profiling of differentially expressed genes was done based on their statistical significance and biological relevance to the unique pathology of CNI nephrotoxicity. Among these, three genes RGS1, CXCR4, and TGIF1 were further quantitatively evaluated using real time-PCR. Between CNI group and normal allograft, t-test results showed only RGS1 gene expression level was statistically significant. Between IF/TA group in normal allograft, both RGS1 and CXCR4 showed statistical significance. The calculated relative fold changes revealed an up-regulated pattern of RGS1 and CXCR4 expression in association with pathological groups (CNI and IF/TA). We did not, however, find any association between the expression of TGIF1 in either CNI group or IF/TA group.

Phosphate (Pi) deficiency is a global problem for food production. Plants have evolved complex mechanisms to adapt to low Pi. We focused on the initial aspects of adaptation to low Pi - perception and immediate-early responses to changes in external Pi. To examine whether a labile repressor controls expression of the high affinity Pi transporter, Pht1;1, we performed electrophoretic mobility shift assays (EMSA) but observed only weak protein-DNA binding activity using extracts from Arabidopsis suspension cultures or seedlings. The regulatory role of different regions in Pht1;1 promoter was dissected by promoter deletion analysis, using uidA as a reporter. We identified two domains important for regulation: sequences between -1898 bp and - 932 bp are important for induction of Pht1;1 in low Pi; the intron in the 5’UTR impacts Pht1;1 expression in the young part of both primary and lateral root apices. A complementary approach to identify repressors of Pi starvation responses was pursued: We identified ZAT18, a putative transcription factor, as a candidate repressor. ZAT18 contains an EAR motif, a repressor domain in plants; the expression of ZAT18 responds to Pi starvation. Using transgenic lines with promoter::ZAT18-VENUS constructs, we studied its expression, localization and abundance in different levels of Pi availability: ZAT18 is mainly expressed in the nucleus of Arabidopsis root hair cells. Its accumulation was induced by 4 day Pi starvation. We also performed a microarray analysis to examine global gene expression levels during Pi starvation and rapid recovery. Our data indicated that 258 genes were induced and 188 genes were suppressed during Pi starvation. For most of these genes, responses were reversed after 4 hour Pi recovery. Further study of these genes will help to define targets of the early Pi starvation-signalling pathway.

Genome-wide microarray technology has facilitated the systematic discovery of diagnostic biomarkers of cancers and other pathologies. However, meta-analyses of published arrays using melanoma as a test cancer has uncovered significant inconsistences that hinder advances in clinical practice. In this study a computational model for the integrated analysis of microarray datasets is proposed in order to provide a robust ranking of genes in terms of their relative significance; both genome-wide relative significance (GWRS) and genome-wide global significance (GWGS). When applied to five melanoma microarray datasets published between 2000 and 2011, a new 12-gene diagnostic biomarker signature for melanoma was defined (i.e., EGFR, FGFR2, FGFR3, IL8, PTPRF, TNC, CXCL13, COL11A1, CHP2, SHC4, PPP2R2C, and WNT4). Of these, CXCL13, COL11A1, PTPRF and SHC4 are components of the MAPK pathway and were validated by immunocyto- and immunohisto-chemistry. These proteins were found to be overexpressed in metastatic and primary melanoma cells in vitro and in melanoma tissue in situ compared to melanocytes cultured from healthy skin epidermis and normal healthy human skin. One challenge for the integrated analysis of microarray data is that the microarray data are produced using different platforms and bio-samples, e.g. including both cell line- and biopsy-based microarray datasets. In order to address these challenges, the computational model was further enhanced the stratification of datasets into either biopsy or cell line derived datasets, and via the weighting of microarray data based on quality criteria of data. The methods enhancement was applied to 14 microarray datasets of three cancers (breast, prostate, and melanoma) based on classification accuracy and on the capability to identify predictive biomarkers. Four novel measures for evaluating the capability to identify predictive biomarkers are proposed: (1) classifying independent testing data using wrapper feature selection with machine leaning, (2) assessing the number of common genes with the genes retrieved in independent testing data, (3) assessing the number of common genes with the genes retrieved in across multiple training datasets, (4) assessing the number of common genes with the genes validated in the literature. This enhancement of computational approach (i) achieved reliable classification performance across multiple datasets, (ii) recognized more significant genes into the top-ranked genes as compared to the genes detected by the independent test data, and (iii) detected more meaningful genes than were validated in previous melanoma studies in the literature.

There has been an increased global interest in biofuels which provide a renewable and sustainable alternative to fossil fuels. Isobutanol is an attractive and superior alternative to the currently produced bioethanol possessing several key advantages. Previous work focuses on strategies for metabolic optimisation of carbon utilisation. However, existing solutions reach a stage where the amount of alcohol produced reaches toxic thresholds for bacteria. This inhibits growth and reduces carbohydrate consumption resulting in lower product yields rendering the biofuel production process uneconomical. In this project, a novel strategy has been adopted which uses a whole system integrated genomics approach consisting of expression profiling, selection to create isobutanol-adapted lineages, next generation sequencing, and comparative behavioural genomics to interrogate the system thoroughly and identify critical determinants of resistance to isobutanol. These were used in the highly-defined model species, E. coli K12 to deliver results of the adaptive mechanisms which take place across the entire genome. 41 gene candidates (4 previously identified in literature) were identified to play a role in isobutanol tolerance. These candidates belong to a range of functional groups such as carbohydrate metabolism, oxidative stress response, osmotic stress response; but also identified novel membrane-associated functions such as the Tol-Pal system, BAM complex and colanic acid production. The results also identify critical genes with unknown functions. The results support previous notions that central carbon metabolism shifts from aerobic to anaerobic metabolism in the presence of isobutanol, but also shows there is a transitionary phase where mixed acid fermentation pathways are utilised. This shift was previously thought to be mediated by the ArcA-ArcB two-component system. However, these results suggest the inactive 2Fe-2S core of the anaerobic-regulator Fnr is re-activated by Fe2+ to form the 4Fe-4S core transported by the FeoAB ferrous iron transport system. The strategy also identified the Tol-Pal system and show it is essential to grow in the presence of isobutanol, which is responsible for the maintaining the integrity of the cell envelope structure and increasing the rate of cell division. The BAM complex is responsible for folding and assembly of outer membrane proteins (OMP) and OMP membrane permeability- this system was found to be important for growth in isobutanol, and SurA, which is the primary OMP assembly pathway provided tolerance which was specific to isobutanol. Colanic acid, an extracellular polysaccharide is produced when the cell experiences stress, and provides protection by forming a physical barrier around the cell. The results show that the presence of colanic acid plays a large role in allowing E. coli to grow in presence of isobutanol, and its role becomes essential at critical concentrations. The results also show deletion of the negative regulator of the colanic acid gene cluster improves growth at critical and growth-inhibiting concentrations. When consolidated, these results facilitated knowledge-led based design and subsequently led to the identification of components for a synthetic design schedule, which lists the genetic manipulations proposed to exploit E. coli to enhance isobutanol tolerance.