Membrane proteins are involved in vital cellular functions and have important implications in disease processes, drug design and therapy. However, it is difficult to obtain diffraction quality crystals to study transmembrane protein structure. Transmembrane protein topology prediction tools try to fill in the gap between abundant number of transmembrane proteins and scarce number of known membrane protein structures (3D structure and biochemically characterized topology). However, at present, the prediction accuracy is still far from perfect. TMHMM is the current state-of- the-art method for membrane protein topology prediction. In order to improve the prediction accuracy of TMHMM, based upon the method of GenomeScan, the author implemented AHMM (augmented HMM) by incorporating functional domain information externally to TMHMM. Results show that AHMM is better than TMHMM on both helix and sidedness prediction. This improvement is verified by both statistical tests as well as sensitivity and specificity studies. It is expected that when more and more functional domain predictors are available, the prediction accuracy will be further improved.

Membrane proteins are involved in vital cellular functions and have important implications in disease processes, drug design and therapy. However, it is difficult to obtain diffraction quality crystals to study transmembrane protein structure. Transmembrane protein topology prediction tools try to fill in the gap between abundant number of transmembrane proteins and scarce number of known membrane protein structures (3D structure and biochemically characterized topology). However, at present, the prediction accuracy is still far from perfect. TMHMM is the current state-of- the-art method for membrane protein topology prediction. In order to improve the prediction accuracy of TMHMM, based upon the method of GenomeScan, the author implemented AHMM (augmented HMM) by incorporating functional domain information externally to TMHMM. Results show that AHMM is better than TMHMM on both helix and sidedness prediction. This improvement is verified by both statistical tests as well as sensitivity and specificity studies. It is expected that when more and more functional domain predictors are available, the prediction accuracy will be further improved.

<p>Membrane proteins are of broad interest since they constitute a large fraction of the proteome in all organisms, up to 20-30%. They play a crucial role in many cellular processes mediating information flow and molecular transport across otherwise nearly impermeable membranes. Traditional three-dimensional structural analyses of membrane proteins are difficult to perform, which makes studies of other structural aspects important. The topology of an α-helical membrane protein is a two-dimensional description of how the protein is embedded in the membrane and gives valuable information on both structure and function.</p><p>This thesis is focused on predicting the topology of α-helical membrane proteins and on assessing and improving the prediction accuracy. Reliability scores have been derived for a number of prediction methods, and have been integrated into the widely used TMHMM predictor. The reliability score makes it possible to estimate the trustworthiness of a prediction.</p><p>Mapping the full topology of a membrane protein experimentally is time-consuming and cannot be done on a genome-wide scale. However, determination of the location of one part of a membrane protein relative to the membrane is feasible. We have analyzed the impact of incorporating such experimental information <i>a priori </i>into TMHMM predictions and show that the accuracy increases significantly. We further show that the C-terminal location of a membrane protein (inside or outside) is the optimal information to use as a constraint in the predictions.</p><p>By combining experimental techniques for determining the C-terminal location of membrane proteins with topology predictions, we have produced reliable topology models for the majority of all membrane proteins in the model organisms <i>E. coli </i>and <i>S. cerevisiae</i>. The results were further expanded to ~15,000 homologous proteins in 38 fully sequenced eukaryotic genomes. This large set of reliable topology models should be useful, in particular as the structural data for eukaryotic membrane proteins is very limited.</p>

Membrane proteins are of broad interest since they constitute a large fraction of the proteome in all organisms, up to 20-30%. They play a crucial role in many cellular processes mediating information flow and molecular transport across otherwise nearly impermeable membranes. Traditional three-dimensional structural analyses of membrane proteins are difficult to perform, which makes studies of other structural aspects important. The topology of an α-helical membrane protein is a two-dimensional description of how the protein is embedded in the membrane and gives valuable information on both structure and function. This thesis is focused on predicting the topology of α-helical membrane proteins and on assessing and improving the prediction accuracy. Reliability scores have been derived for a number of prediction methods, and have been integrated into the widely used TMHMM predictor. The reliability score makes it possible to estimate the trustworthiness of a prediction. Mapping the full topology of a membrane protein experimentally is time-consuming and cannot be done on a genome-wide scale. However, determination of the location of one part of a membrane protein relative to the membrane is feasible. We have analyzed the impact of incorporating such experimental information a priori into TMHMM predictions and show that the accuracy increases significantly. We further show that the C-terminal location of a membrane protein (inside or outside) is the optimal information to use as a constraint in the predictions. By combining experimental techniques for determining the C-terminal location of membrane proteins with topology predictions, we have produced reliable topology models for the majority of all membrane proteins in the model organisms E. coli and S. cerevisiae. The results were further expanded to ~15,000 homologous proteins in 38 fully sequenced eukaryotic genomes. This large set of reliable topology models should be useful, in particular as the structural data for eukaryotic membrane proteins is very limited.

Membrane proteins comprise around 20-30% of a typical proteome and play crucial roles in a wide variety of biochemical pathways. Apart from their general biological significance, membrane proteins are of particular interest to the pharmaceutical industry, being targets for more than half of all available drugs. This thesis focuses on prediction methods for membrane proteins that ultimately rely on their amino acid sequence only. By identifying soluble protein domains in membrane protein sequences, we were able to constrain and improve prediction of membrane protein topology, i.e. what parts of the sequence span the membrane and what parts are located on the cytoplasmic and extra-cytoplasmic sides. Using predicted topology as input to a profile-profile based alignment protocol, we managed to increase sensitivity to detect distant membrane protein homologs. Finally, experimental measurements of the level of membrane integration of systematically designed transmembrane helices in vitro were used to derive a scale of position-specific contributions to helix insertion efficiency for all 20 naturally occurring amino acids. Notably, position within the helix was found to be an important factor for the contribution to helix insertion efficiency for polar and charged amino acids, reflecting the highly anisotropic environment of the membrane. Using the scale to predict natural transmembrane helices in protein sequences revealed that, whereas helices in single-spanning proteins are typically hydrophobic enough to insert by themselves, a large part of the helices in multi-spanning proteins seem to require stabilizing helix-helix interactions for proper membrane integration. Implementing the scale to predict full transmembrane topologies yielded results comparable to the best statistics-based topology prediction methods.

Heteropolymeric O antigen (O-Ag)-capped lipopolysaccharide is the principal constituent of the Gram-negative bacterial cell surface. It is assembled via the integral inner membrane (IM) Wzx/Wzy-dependent pathway. In Pseudomonas aeruginosa, Wzx translocates lipid-linked anionic O-Ag subunits from the cytoplasmic to the periplasmic leaflets of the IM, where Wzy polymerizes the subunits to lengths regulated by Wzz1/2. The Wzx and Wzy IM topologies were mapped using random C-terminal-truncation fusions to PhoALacZα, which displays PhoA/LacZ activity dependent upon its subcellular localization. Twelve transmembrane segments (TMS) containing charged residues were identified for Wzx. Fourteen TMS, two sizeable cytoplasmic loops (CL), and two large periplasmic loops (PL3 and PL5 of comparable size) were characterized for Wzy. Despite Wzy PL3–PL5 sequence homology, these loops were distinguished by respective cationic and anionic charge properties. Site-directed mutagenesis identified functionally-essential Arg residues in both loops. These results led to the proposition of a “catch-and-release” mechanism for Wzy function. The abovementioned Arg residues and intra-Wzy PL3–PL5 sequence homology were conserved among phylogenetically diverse Wzy homologues, indicating widespread potential for the proposed mechanism. Unexpectedly, Wzy CL6 mutations disrupted Wzz1-mediated regulation of shorter O-Ag chains, providing the first evidence for direct Wzy–Wzz interaction. Mutagenesis studies identified functionally-important charged and aromatic TMS residues localized to either the interior vestibule or TMS bundles in a 3D homology model constructed for Wzx. Substrate-binding or energy-coupling roles were proposed for these residues, respectively. The Wzx interior was found to be cationic, consistent with translocation of anionic O-Ag subunits. To test these hypotheses, Wzx was overexpressed, purified, and reconstituted in proteoliposomes loaded with I−. Common transport coupling ions were introduced to “open” the protein and allow detection of I− flux via reconstituted Wzx. Extraliposomal changes in H+ induced I− flux, while Na+ addition had no effect, suggesting H+-dependent Wzx gating. Putative energy-coupling residue mutants demonstrated defective H+-dependent halide flux. Wzx also mediated H+ uptake as detected through fluorescence shifts from proteoliposomes loaded with pH-sensitive dye. Consequently, Wzx was proposed to function via H+-coupled antiport. In summary, this research has contributed structural and functional knowledge leading to novel mechanistic understandings for O-Ag biosynthesis in bacteria. / Bookmarks within the document have been provided for ease of access to a particular section in the body of the thesis. Each entry in the Table of Contents, List of Tables, and List of Figures has been "linked" to its respective position and as such can be clicked for direct access to the entry. Similarly, each in-text Figure or Table reference has been "linked" to its respective figure/table for direct access to the entry. / 1.) Canadian Institutes of Health Research (CIHR) Frederick Banting and Charles Best Canada Graduate Scholarship doctoral award, 2.) CIHR Michael Smith Foreign Study Award, 3.) Cystic Fibrosis Canada (CFC) doctoral studentship, 4.) University of Guelph Dean's Tri-Council Scholarship, 5.) Ontario Graduate Scholarship in Science and Technology, 6.) Operating grants to Dr. Joseph S. Lam from CIHR (MOP-14687) and CFC