One of the key research topics of post-genome study is annotation of the gene with regards to specific function and biological processes. This can help us to understand the precise role that a gene or a group of genes carries. In this thesis, I developed techniques to automatically annotate genes on single gene and a group of genes levels. It is shown that these techniques improve our understanding of biological systems/diseases, and will aid drug discovery. In the first project, I attempted to achieve precise annotation for single genes. In the second and third projects, I performed annotations of a group of genes using pathway knowledge. I examined this problem from supervised and unsupervised learning aspects, respectively. The main contributions of the work are organized as follows: In gene annotation project, I built up an automated scheme to reconcile the term differences arising from the different automated annotation services. The method leaves less than 20% of the annotations for manual work. The generalization performance across other species is of a similar standard, again leaving less than 20% of the annotations for manual inspection. In addition, less than 10% of the results have different functions from EcoCyc results in E.coli genome annotation task. Overall, this method can significantly reduce human effort involvement (6 months’ work by several biologists for a bacterial genome) to resolve inconsistent gene annotations. Then I started from the current limitations of pathway analysis and presented a novel approach for pathway discovery. Enrichment analysis is the most popular approach to map gene expression profiling from genes to biological pathways. It is a powerful tool to identify pathways enriching of differentially expressed gene; however, it is unable to discover active/inhibitive pathways. In this study, I attempted to resolve this issue by integrative classification of KEGG and TF gene sets. I assumed that the pathways with good classification performance should be considered as the active/inhibitive pathways. Based on this hypothesis, I built up a generic approach to incorporate two types of biological data for active pathway discovery. The experimental results show that integration of transcription factor data boosts classification performance. In addition, this method identified relevant biological pathways, which are highly associated with tumour genesis and development. But they are ignored by Gene Set Enrichment Analysis, such as cancer pathway, inflammation and metabolic pathways. Furthermore, this method achieves comparable classification performance with the best-reported results. Lastly, I performed subtyping analysis of Rheumatoid Arthritis patients based on gene expression profiling. I revalidated the two clusters of patients based on two independent cohorts. The experimental results indicate that the subgroup structure does not correspond to the drug response status. In addition, I developed a pathway subtyping approach and achieved the same number of clusters as gene-level clustering results. The pathway clustering results show that one group of the patients has high proliferation and low inflammation response, while the other group has the reverse trend. It suggests that designing drugs with better trade-off between anti-inflammation and anti-proliferation for specific subgroup of patients may achieve better clinical outcomes.

The traditional view of the genome is based on the dogma that genetic information flows from DNA to RNA to protein. Genes have essentially been synonymous with proteins, with RNA viewed primarily as an intermediate template for protein translation. Intriguingly, only ~2% of the human genome encodes proteins, and the number of protein-coding genes (~20,000) is similar between humans and the simple nematode worm.I have been involved with the analysis of a large-scale transcriptome study, intiated by the RIKEN Genomic Sciences Centre in Japan. As part of this work, it was discovered that the genome carries instructions for tens of thousands of “non protein-coding RNAs” (ncRNAs)(please refer to Appendix A to view the original articles that appeared in Science). The significance of these ncRNAs remains a matter of intense interest and debate. Some have argued that these ncRNAs are simply transcriptional noise, while others have suggested that they comprise a critical regulatory system, which directs the complex patterns of gene expression that underlie differentiation and development.To investigate this further, I have conducted a series of studies that explore the expression and evolution of ncRNAs in mammals. Firstly, I established a comprehensive, on-line database of ncRNAs. This collection provides information on more than twenty thousand ncRNAs, and has proven a valuable resource for ncRNA studies. Secondly, I have systematically analyzed the conservation of known functional ncRNA subsets. I found that small ncRNAs (microRNAs and snoRNAs) were well-conserved similar to protein-coding sequences, whereas longer functional ncRNAs were not. These results indicate that long ncRNAs are evolving more rapidly than other functional genomic elements, and suggest that many of the recently-discovered ncRNAs – most of which are long and of unknown significance – might still be functional, despite having poor sequence conservation. Thirdly, I have shown that many ncRNAs are derived from genuine transcripts, whose expression appears regulated in a biologically-relevant manner. Fourthly, I helped develop a computational strategy to identify extremely large ncRNAs and discovered >60 novel candidates, several of which were characterized experimentally. Prior to this work, only a handful of extremely large ncRNAs had been previously described, and these play critical roles in processes such as genomic imprinting and X chromosome inactivation. This study represented the first systematic discovery of extremely large ncRNAs. Finally, I designed custom microarrays and profiled ncRNA expression across the development of CD8+ T cells. CD8+ T cells serve an important role in immunity by killing virus-infected and tumour cells, and transit through a series of functionally-distinct developmental stages. I found that ~200 novel ncRNAs are dynamically expressed during CD8+ T cell development.Taken together, my findings indicate that ncRNAs are a major, regulated output of the mammalian genome, and are consistent with the notion that ncRNAs represent an important, previously-unrecognised biological control system.

Background: Meat tenderness is the consumer's most preferred sensory attribute. This trait is affected by a number of factors, including genotype, age, animal sex, and pre-and post-slaughter management. In view of the high percentage of Zebu genes in the Brazilian cattle population, mainly Nellore cattle, the improvement of meat tenderness is important since the increasing proportion of Zebu genes in the population reduces meat tenderness. However, the measurement of this trait is difficult once it can only be made after animal slaughtering. New technologies such as RNA-Seq have been used to increase our understanding of the genetic processes regulating quantitative traits phenotypes. The objective of this study was to identify differentially expressed genes related to meat tenderness, in Nellore cattle in order to elucidate the genetic factors associated with meat quality. Samples were collected 24 h postmortem and the meat was not aged. Results: We found 40 differentially expressed genes related to meat tenderness, 17 with known functions. Fourteen genes were up-regulated and 3 were down-regulated in the tender meat group. Genes related to ubiquitin metabolism, transport of molecules such as calcium and oxygen, acid-base balance, collagen production, actin, myosin, and fat were identified. The PCP4L1 (Purkinje cell protein 4 like 1) and BoLA-DQB (major histocompatibility complex, class II, DQ beta) genes were validated by qRT-PCR. The results showed relative expression values similar to those obtained by RNA-Seq, with the same direction of expression (i.e., the two techniques revealed higher expression of PCP4L1 in tender meat samples and of BoLA-DQB in tough meat samples). Conclusions: This study revealed the differential expression of genes and functions in Nellore cattle muscle tissue, which may contain potential biomarkers involved in meat tenderness.

Tropical Theileriosis is a parasitic disease of calves with a profound economic impact caused by Theileria annulata, an apicomplexan parasite of the genus Theileria. Transmitted by Hyalomma ticks, T. annulata infects and transforms bovine lymphocytes and macrophages into a cancer-like phenotype characterized by all six hallmarks of cancer. In the current study we investigate the transcriptional landscape of T. annulata-infected lymphocytes to define genes and miRNAs regulated by host cell transformation using next generation sequencing. We also define genes and miRNAs differentially expressed as a result of the attenuation of a T.annulata-infected macrophage cell line used as a vaccine. By comparing the transcriptional landscape of one attenuated and two transformed cell lines we identify four genes that we propose as key factors in transformation and virulence of the T. annulata host cells. We also identify miR- 126-5p as a key regulator of infected cells proliferation, adhesion, survival and invasiveness. In addition to the host cell trascriptome we studied T. annulata transcriptome and identified the role of ROS and TGF-β2 in controlling parasite gene expression. Moreover, we have used the deep parasite ssRNA-seq data to refine the available T. annulata annotation. Taken together, this study provides the full list of host cell’s genes and miRNAs transcriptionally perturbed after infection with T. annulata and after attenuation and describes genes and miRNAs never identified before as players in this type of host cell transformation. Moreover, this study provides the first database for the transcriptome of T. annulata and its host cells using next generation sequencing.

In the present study, I used transcriptomic data to elucidate the role of selection in maintaining genetic cohesion and promoting divergence among wild populations of the annual sunflower Helianthus annuus. I observed that nearly half of the loci displaying high levels of population structure (44%) show allele frequency spectrum skew consistent with recent exposure to natural selection. Among transcriptomic regions at which allele frequency divergence is lowest, fewer loci maintain strong signals of selection (34%). Additionally, I find evidence supporting the maintenance of evolutionarily complex haplotype structure within populations at loci showing high levels of among-population allele frequency divergence.

Societies where workers sacrifice their own reproduction and cooperatively nurture the offspring of a reproductive queen caste have originated repeatedly across the Tree of Life. The attainment of such reproductive division of labor enabled the evolution of remarkable diversity in development, behavior, and social organization in the Hymenoptera (ants, bees, and wasps). Wasps of the family Vespidae exhibit a gamut of social levels, ranging from solitary to highly social behavior. The highly social yellowjackets and hornets (Vespinae) have well developed differences in form and function between queens and workers, large colony sizes, and intricate nest architecture. Moreover, certain socially parasitic species in the Vespinae have secondarily lost the worker caste and rely entirely on the workers of a host species to ensure the survival of parasitic offspring. Understanding the evolution of behavioral traits in the Vespinae over long periods of time would be greatly enhanced by a robust hypothesis of historical relationships. In this study, I analyze targeted genes and transcriptomes to address three goals. First, infer phylogenetic relationships within yellowjackets (Vespula and Dolichovespula) and hornets (Vespa and Provespa). Second, test the hypothesis that social parasites are more closely related to their hosts than to any other species (Emery's rule). Third, test the protein evolution hypothesis, which states that accelerated evolution of protein coding genes and positive selection operated in the transition to highly eusocial behavior. The findings of this study challenge the predominant understanding of evolutionary relationships in the Vespinae. I show that yellowjacket genera are not sister lineages, instead recovering Dolichovespula as more closely related to the hornets, and placing Vespula as sister to all other vespine genera. This implies that traits such as large colony size and high paternity are mostly restricted to a particular evolutionary trajectory (Vespula) from an early split in the Vespinae. I demonstrate that obligate and facultative social parasites do not share immediate common ancestry with their hosts, indicating that socially parasitic behavior likely evolved independently of host species. Moreover, obligate social parasites share a unique evolutionary history, suggesting that their parasitic behavior might have a genetic component. Lastly, I analyze transcriptomic data to infer a phylogeny of vespid wasps and use this phylogeny to discover lineage-specific signatures of positive selection. I identify more than two hundred genes showing signatures of positive selection on the branch leading to the highly eusocial yellowjackets and hornets. These positively selected genes involve functions related mainly to carbohydrate metabolism and mitochondrial activity, in agreement with insights from studies of bees and ants. Parallels of functional categories for genes under positive selection suggests that at the molecular level the evolution of highly eusocial behavior across the Hymenoptera might have followed similar and narrow paths.

Tobacco smoke remains a major public health concern and a factor contributing to the development and progression of various lung diseases world-wide. Smoking cessation can significantly reduce the risk of developing smoking-related diseases, although some smokers remain at an elevated risk despite quitting. Here, I used high-throughput genomic technologies to pave the way for understanding the transcriptomic response in airway epithelium to tobacco exposure, smoking cessation and potentially reduced exposure products (PREP). First, using a longitudinal dataset of nasal airway epithelial cells obtained from active smokers enrolled in smoking cessation programs over 24 weeks, I demonstrated that tobacco-related alterations in the airway gene expression are rapidly reversed within 4 to 8 weeks following smoking cessation. Genes with different biological functions revert towards baseline with different dynamics following smoking cessation. These findings suggest that the nasal-epithelium can serve as a minimally-invasive site to measure the reversible impact of smoking. Secondly, using a dataset of nasal airway epithelial cells from active smokers who switched to potentially reduced exposure products (PREP) for 6 weeks, I showed that gene expression differences induced by switching to PREP may only constitute a partially reduced exposure relative to smoking cessation. My results demonstrate that the nasal-epithelium can also serve as a minimally-invasive site to measure the responses to PREP and may ultimately yield biomarkers to evaluate the potential disease risks associated with these products. Lastly, using small RNA-seq data from bronchial epithelial cells of smokers, I found alterations in airway micro-RNA expression that associate with time since quitting in former smokers. These studies have provided novel insights into the physiologic responses of the airway epithelium to tobacco smoke and PREP and may ultimately serve as a useful approach for evaluating the disease risks associated with changes in smoking behavior. This work sets the stage for additional studies aimed at identifying prevention strategies that decrease the persistent risk of smoking-related lung disease in former smokers and identify biomarkers to assess lung disease risk in former smokers. / 2016-12-31T00:00:00Z

Olfactory ensheathing cells (OECs), the glia of the olfactory nerve, are promising candidates for patient-specific cell-mediated repair of both peripheral nerves and the spinal cord. The recent discovery that OECs originate from the neural crest, rather than the olfactory epithelium as previously thought, potentially means that homogeneous populations of OECs for repair could be expanded in culture from neural crest stem cells persisting in the patient's own skin and hair follicles. The first step towards this long-term goal is to understand the molecular mechanisms underlying neural crest differentiation into OECs, as opposed to Schwann cells (the glia of all other peripheral nerves), which are less effective in spinal cord repair. To identify transcription factors and signalling pathways that might be involved in OEC versus Schwann cell differentiation, I took an unbiased transcriptome profiling approach. Taking advantage of Sox10 expression throughout both OEC and Schwann cell development, I used laser-capture microdissection on cryosections of mouse embryos carrying a Sox10:H2BVenus transgene, to isolate OEC subpopulations (olfactory mucosal OECs, from the olfactory nerve, and olfactory nerve layer OECs, from the olfactory nerve layer surrounding the olfactory bulb) at different stages of development, and Schwann cells from trigeminal nerve branches on the same sections, for RNA-seq and cross-wise comparison of transcriptomes. Validation of candidate genes by in situ hybridisation revealed some contamination with adjacent cells from mesenchyme, olfactory epithelium or olfactory bulb, but also identified the expression in developing OECs of various genes previously reported to be expressed in adult OECs, and of over 20 genes previously unknown in OECs. Some of these genes are expressed by OECs but not Schwann cells; some are expressed by olfactory nerve layer OECs but not olfactory mucosal OECs, while some are expressed by olfactory mucosal OECs and Schwann cells but not olfactory nerve layer OECs. For a subset of the genes, I was also able to analyse OEC differentiation in mouse mutants. I also collected transcriptome data from neural crest-derived cells that persist on the olfactory nerve in Sox10-null embryos (in which neural crest-derived cells colonise the olfactory nerve, but normal OEC differentiation is disrupted). Comparison with wild-type OEC transcriptome data from the same embryonic stage identified genes whose expression is likely either downregulated or up-regulated in the absence of Sox10, supporting a role in normal OEC differentiation. Overall, these various transcriptomic comparisons (between OECs at different developmental stages, different OEC subpopulations, OECs versus Schwann cells, and OECs versus Sox10-null neural crest-derived cells on the olfactory nerve) have identified multiple transcription factor and signalling pathway genes, amongst others, that are expressed during OEC development in vivo (including some specific to different OEC subpopulations) and that may be important for OEC differentiation. Furthermore, some of these genes are not expressed by embryonic Schwann cells. This work provides a foundation for understanding how to promote OEC rather than Schwann cell differentiation from neural crest stem cells in culture, with the potential for clinical application in the future.

O splicing alternativo é um processo pelo qual os exons de um transcrito primário são ligados de diferentes maneiras durante o processamento do RNA, levando à síntese de proteínas distintas. Compreende um importante mecanismo na expressão gênica de eucariotos, responsável pelo aumento da diversidade proteômica e, portanto da capacidade codificante do genoma. Diferentes mecanismos parecem afetar a regulação do splicing alternativo, incluindo estresse metabólico. No presente estudo, foi realizada uma análise detalhada de sequências ORESTES de tecidos de cabeça e pescoço. Essa análise revelou que o ganho de sequências exônicas é mais freqüente que sua perda, e que a regra GT/AG é predominante em sítios de splicing. Nós observamos que o splicing alternativo geralmente não altera a matriz de leitura, mas pode afetar um domínio protéico e remover ou adicionar novos sítios de fosforilação e glicosilação. Elementos reguladores potenciais e elementos repetitivos foram freqüentes nas sequências alternativas e nas suas vizinhas. A expressão de isoformas de splicing potenciais foi investigada em diferentes tecidos, incluindo sob condições de estresse. Foram validados cerca de 50 eventos de splicing novos em células normais e tumorais. Diversas variantes, tais como aquelas dos genes HNRNPK, ACTN1, BAT3, CEP192, MPV17, PDK1, PRKAR1A, RAG1AP1 e TRIP6 mostraram padrões de expressão distintos em diferentes tipos celulares, em amostras normais e tumorais de pacientes com carcinoma de cabeça e pescoço e, em alguns casos, em diferentes estágios do tumor. Também foi validado um transcrito novo do gene RIPK2, responsável por codificar uma quinase de serine/treonine que ativa a via de sinalização NF-kB, e foi observada uma mudança na expressão dessa variante em resposta ao estresse térmico in vitro. Ainda não está claramente definido se o splicing alternativo é causa ou conseqüência do processo neoplásico. Nossos dados adicionam informações novas a esse tópico e fornecem alguns exemplos que evidenciam a importância do processamento do RNA na regulação da expressão gênica, tanto em condições normais como de doença. / Alternative splicing is a process by which the exons of the primary gene transcript are linked in different ways during RNA processing resulting in distinctive proteins. It is an important mechanism in eukaryotic gene expression that enhances proteome diversity and, therefore, the coding capacity of the genome. Different mechanisms seem affect alternative splicing regulation, including metabolic stresses. In the present study, a detailed informatics analysis of ORESTES sequences from head and neck tissues was performed. This in silico analysis revealed that gain of exon sequences is more frequent than exon skipping and GT/AG rule is predominant in splice sites. We observed that alternative splicing usually does not alter the reading frame but may disrupt a protein domain and remove or add new phosphorylation and glycosylation sites. Repetitive and potential regulator elements were frequent in the alternative sequences or in their neighbors. The expression of putative splicing isoforms was investigated in different tissues, including upon stress conditions. We validated approximately 50 new splicing events in normal and tumor cells. Several variants, such as those from HNRNPK, ACTN1, BAT3, CEP192, MPV17, PDK1, PRKAR1A, RAG1AP1 and TRIP6 genes showed distinctive expression pattern in different cell types, in normal and cancer samples from head and neck carcinoma patients and, in some cases, in different tumor stages. We also validated a new transcript of RIPK2 gene, which codes a serine/threonine kinase that activates the NF-kB pathway, and observed a shift in the expression of this variant as a response to temperature stress in vitro. It is currently not clear whether alternative splicing is the cause or the consequence of the neoplastic process. Our data add new information to this topic and provide some examples on the importance of RNA processing in gene expression regulation, both in normal and disease conditions.

The late blight pathogen, Phytophthora infestans, is the most destructive pathogen of its solanaceous hosts potato and tomato. It is a threat to global food security and it is therefore important to understand the cellular and molecular dynamics underlying colonisation of its host plants. This greater understanding will inform strategies to improve host plant resistance. In addition to studying the cell biology of the interaction, it is important to understand the temporal changes in gene expression and regulation during host-pathogen interactions at the earliest infection time points. Previously published transcriptomic studies of P. infestans used two days post infection (dpi) as the earliest sampling time point. Expression of a marker gene (Hmp1) for biotrophy and a selection of effector coding genes has been reported as early as 12 hours post inoculation (hpi), suggesting that infection was initiated before then. Transcriptomic studies of P. infestans have focussed mostly on leaf tissue, and there is still a lack of research on the transcriptome of P. infestans grown in alternative plant tissues such as tubers, or in host cell-free apoplastic fluid. This thesis explores transcriptomic studies of the early, biotrophic stages of potato infection by Phytophthora infestans, which is critical for understanding which genes are involved at what stages of infection development. By using the latest sensitive microarray technology to study the P. infestans transcriptome in an infection time course that remained biotrophic for its duration, a list of 1,707 transcripts of P. infestans were discovered to be differentially expressed. This list included 114 transcripts for RxLR effectors, out of which 26 were detected from 12 hours post infection, including: Avr2, Avr3a, Avrblb1 (ipi01), Avrblb2, and the recently characterised RD2. Also of interest was that transcripts encoding a PAMP (CBEL) detected at 12 hours, were suppressed in the pathogen by 24 hours. Transcripts encoding 55 RxLR effectors were co-expressed (with >95 % correlation coefficient) with the biotrophy marker gene Hmp1, suggesting that these effectors are important throughout the biotrophic stages of infection. QRT-PCR and cell biology data supported the expression of the biotrophy marker gene Hmp1 as early as 12 hours after infection and this was further supported by the co-expression of avirulence genes such as Avr2 and Avr3a. A set of 17 transcripts, including six cytoplasmic effectors (RxLR effectors), as well as a transcript encoding an apoplastic effector (glucanase inhibitor), was found to be infection-specific, supporting the hypothesis that these genes might have roles in establishing biotrophy. By examining pathogen behaviour in tuber tissue, clear cell biology evidence of functional haustoria was found. Gene expression analysis of a selection of leaf infection-related genes suggested that effectors are used to promote infection also in host tuber tissue. However, some cytoplasmic RXLR effector proteins such as PITG_05146 and PITG_15128, which were up-regulated during biotrophic infection of leaf tissue, were not detected during tuber infection, indicating potential differences in pathogenic requirements. A microarray experiment was conducted on in vitro stages of zoospores, and mycelium grown in apoplastic fluid of N. benthamiana, nutrient rich pea broth, and sterile water. This revealed 13,819 transcripts that were differentially expressed between any two conditions. This list included transcripts encoding 322 RxLR effectors, of which avirulence effectors such as Avr2, Avr3a, and RD2 were highly up-regulated during hyphal growth in apoplastic fluid compared to other in vitro stages. This provides evidence that the apoplast contains chemical signals that induce expression of infection-related genes in P. infestans. Curiously, the leaf infection-specific genes identified in Chapter 3 were not expressed when P. infestans was grown in apoplastic fluid, revealing that additional stimuli are required for induction of all necessary pathogen genes during infection. Future research, building upon the findings from this project, should be focused on the following areas: 1) Explore whether haustoria are produced only in order to deliver effectors or if there are other purposes as well, such as nutrient uptake; 2) The continued exploration of differences between genes co-expressed with Hmp1 during leaf infection, tuber infection, and in apoplastic fluid to further dissect the transcriptional regulation of these genes; 3) Identify whether Hmp1-co-expressed genes of unknown function may play a role in haustorium formation; 4) Investigate, using molecular transformation and cell biology, whether secreted proteins co-expressed with Hmp1 are secreted from haustoria; 5) Investigate the role(s) of infection-specific genes in establishing disease. 6) Transcriptomic studies of P. infestans biotrophic infection of tuber tissue to determine the differences in pathogenic adaptation in this tissue type, compared to leaf infection.