Alx4 is a paired-type homeodomain transcription factor that plays a key role in development, strongly expressed in the first branchial arch and craniofacial region. Pax3 also belongs to this family, and it displays a similar pattern of expression to that of Alx4. When Pax3 or Alx4 activity is lost individually, defects arise in an overlapping set of embryonic structures. In addition to their expression patterns, this suggests that these two factors may interact to play a role in normal murine development. We demonstrate an overlapping pattern of expression of Pax3 and Alx4 in the developing embryo and that Pax3 and Alx4 physically interact in vivo and in vitro. Pax3 and Alx4 can activate transcription from a P3 homeodomain consensus site, and preliminary analysis of mice null for both Pax3 and Alx4 show a novel mutant phenotype. We have therefore demonstrated a physical and genetic interaction between Pax3 and Alx4.

The currently known network of human protein-protein interactions (PPIs) is providing new insights into diseases and helping to identify potential therapies. However, according to several estimates, the known interaction network may represent only 10% of the entire interactome - indicating that more comprehensive knowledge of the interactome could have a major impact on understanding and treating diseases. The primary aim of this thesis was to develop computational methods to provide increased coverage of the interactome. A secondary aim was to gain a better understanding of the link between networks and phenotype, by analyzing essential mouse genes. Two algorithms were developed to predict PPIs and provide increased coverage of the interactome: FpClass and mixed co-expression. FpClass differs from previous PPI prediction methods in two key ways: it integrates both positive and negative evidence for protein interactions, and it identifies synergies between predictive features. Through these approaches FpClass provides interaction networks with significantly improved reliability and interactome coverage. Compared to previous predicted human PPI networks, FpClass provides a network with over 10 times more interactions, about 2 times more proteins and a lower false discovery rate. This network includes 595 disease related proteins from OMIM and Cancer Gene Census which have no previously known interactions. The second method, mixed co-expression, aims to predict transient PPIs, which have proven difficult to detect by computational and experimental methods. Mixed co-expression makes predictions using gene co-expression and performs significantly better (p < 0.05) than the previous method for predicting PPIs from co-expression. It is especially effective for identifying interactions of transferases and signal transduction proteins. For the second aim of the thesis, we investigated the relationship between gene essentiality and diverse gene/protein features based on gene expression, PPI and gene co-expression networks, gene/protein sequence, Gene Ontology, and orthology. We identified non-redundant features closely associated with essentiality, including centrality in PPI and gene co-expression networks. We found that no single predictive feature was effective for all essential genes; most features, including centrality, were less effective for genes associated with postnatal lethality and infertility. These results suggest that understanding phenotype will require integrating measures of network topology with information about the biology of the network’s nodes and edges.

Alx4 is a paired-type homeodomain transcription factor that plays a key role in development, strongly expressed in the first branchial arch and craniofacial region. Pax3 also belongs to this family, and it displays a similar pattern of expression to that of Alx4. When Pax3 or Alx4 activity is lost individually, defects arise in an overlapping set of embryonic structures. In addition to their expression patterns, this suggests that these two factors may interact to play a role in normal murine development. We demonstrate an overlapping pattern of expression of Pax3 and Alx4 in the developing embryo and that Pax3 and Alx4 physically interact in vivo and in vitro. Pax3 and Alx4 can activate transcription from a P3 homeodomain consensus site, and preliminary analysis of mice null for both Pax3 and Alx4 show a novel mutant phenotype. We have therefore demonstrated a physical and genetic interaction between Pax3 and Alx4.

The currently known network of human protein-protein interactions (PPIs) is providing new insights into diseases and helping to identify potential therapies. However, according to several estimates, the known interaction network may represent only 10% of the entire interactome - indicating that more comprehensive knowledge of the interactome could have a major impact on understanding and treating diseases. The primary aim of this thesis was to develop computational methods to provide increased coverage of the interactome. A secondary aim was to gain a better understanding of the link between networks and phenotype, by analyzing essential mouse genes. Two algorithms were developed to predict PPIs and provide increased coverage of the interactome: FpClass and mixed co-expression. FpClass differs from previous PPI prediction methods in two key ways: it integrates both positive and negative evidence for protein interactions, and it identifies synergies between predictive features. Through these approaches FpClass provides interaction networks with significantly improved reliability and interactome coverage. Compared to previous predicted human PPI networks, FpClass provides a network with over 10 times more interactions, about 2 times more proteins and a lower false discovery rate. This network includes 595 disease related proteins from OMIM and Cancer Gene Census which have no previously known interactions. The second method, mixed co-expression, aims to predict transient PPIs, which have proven difficult to detect by computational and experimental methods. Mixed co-expression makes predictions using gene co-expression and performs significantly better (p < 0.05) than the previous method for predicting PPIs from co-expression. It is especially effective for identifying interactions of transferases and signal transduction proteins. For the second aim of the thesis, we investigated the relationship between gene essentiality and diverse gene/protein features based on gene expression, PPI and gene co-expression networks, gene/protein sequence, Gene Ontology, and orthology. We identified non-redundant features closely associated with essentiality, including centrality in PPI and gene co-expression networks. We found that no single predictive feature was effective for all essential genes; most features, including centrality, were less effective for genes associated with postnatal lethality and infertility. These results suggest that understanding phenotype will require integrating measures of network topology with information about the biology of the network’s nodes and edges.

Protein-protein interactions regulate and drive biological processes and understanding the assembly of these interactions is important. The LysR-Type Transcriptional Regulators (LTTRs) are a large family of transcriptional regulators found in prokaryotes. I have used the LTTRs as a model for protein specificity. In order to understand a residue’s contribution to oligomerization, alanine-scanning mutagenesis was used to probe the contribution of residues identified from in silico analysis of two proteins: OxyR and CynR. The contribution of the residues to oligomerization was characterized using lcI repressor fusions. In OxyR, seven residues were identified as hot spots. Moreover, these hot spots are not especially conserved. The interaction surface of OxyR was mapped onto a multiple sequence alignment of the LTTR family. This mapping identified putative contacts in the CynR regulatory domain dimer interface. Combined with the in vivo testing, three residues were identified as hot spots. The residues identified in OxyR and CynR do not overlap. To investigate the assembly of the LTTRs I used a negative-dominance assay with lcI repressor fusions. Taken together, I show that the LTTRs in E. coli K-12 are mostly specific in their interactions.

published_or_final_version / Medicine / Doctoral / Doctor of Philosophy

published_or_final_version / Chemistry / Doctoral / Doctor of Philosophy

Understanding normal and cancer cell biology requires the development and application of systems biology approaches capable of probing the functional human proteome, and the protein-protein interactions (PPIs) within it. Such technologies will facilitate our understanding of how molecular events drive phenotypic outcomes, and how these processes are perturbed in disease conditions. In this thesis, I first describe the development of a mammalian, Gateway compatible, lentivirus-based protein-fragment complementation assay (magical-PCA), for the in vivo high-throughput identification of PPIs in mammalian cells. This technology provides a vast improvement over current PCA methodologies by allowing for pooled, proteome-scale mapping of PPIs in any mammalian cell line of interest, using any bait protein of interest. A proof-of-concept pooled genome-scale screen using the magical-PCA approach was performed using the human mitochondrial protein TOMM22 as a bait, providing evidence that this technology is amenable to proteome-wide screens. Moreover, the TOMM22 screens offered novel insight into links between TOMM22 and proteins involved in mitochondrial organization, apoptosis, and cell cycle dynamics. Second, I performed a pooled genome-scale magical-PCA screen with the oncoprotein BMI1, a component of the E3 ubiquitin ligase complex involved in histone H2A mono-ubiquitination and gene silencing, to identify novel BMI1 protein interactors. Consequently, I have uncovered a novel physical and functional association between BMI1 and components of the mammalian splicing machinery. I further discovered that BMI1 knockdown influenced the alternative splicing of a number of cellular pre-mRNAs in colon cancer cell lines, suggesting that the association between BMI1 and cellular splicing factors impinges on pre-mRNA processing. Importantly, BMI1 expression was shown to influence the alternative splicing of the SS18 oncoprotein towards an exon 8-excluded isoform, which was shown in this study to promote cell proliferation when assessed in an anchorage-independent growth assay. Together, these studies highlight the development of a new methodology for the detection and proteome-scale screening of mammalian PPIs. A proof-of-concept screen with human TOMM22 highlighted the utility of the approach, as I was able to detect both strong and weak or transient PPIs. Application of my screening methodology to BMI1 provided crucial insight into the function of this oncoprotein, and BMI1-driven tumorigenesis.

Thesis (Ph. D.)--University of Hawaii at Manoa, 2005. / Includes bibliographical references (leaves 162-188). / Also available by subscription via World Wide Web / xvi, 188 leaves, bound ill. (some col.) 29 cm

Thesis (Ph. D.)--West Virginia University, 2007. / Title from document title page. Document formatted into pages; contains xii, 156 p. : ill. (some col.). Vita. Includes abstract. Includes bibliographical references.