Search

ATP-Binding-Cassette (ABC) transporters are primary active pumps that typically couple the binding and hydrolysis of ATP to the translocation of compounds across cellular membranes. Some, like ABCB1, ABCC1 and ABCC3, are polyspecific and can efflux clinically important drugs which may contribute to their therapeutic failure. In this study I have investigated (1) the mechanism of ABC transporter function, (2) studied the potential for regulation by ubiquitin ligases (both using ABCB1 as a model), and (3) tested the involvement of ABCC1 and ABCC3 in autocrine signalling in cancer. (1) In 1966, Jardetzky et. al [1] proposed that membrane pumps function by exposing their ligand-binding pocket alternately on different sides of the membrane. For ABC transporters, this coupling of the aspect and affinity of the ligand-binding cavities of the two transmembrane domains (TMDs) to the ATP catalytic cycle of the two nucleotide-binding domains (NBDs) is fundamental to the transport mechanism but is poorly understood at the molecular level. Structure data suggest signals are transduced through intracellular loops of the TMDs which slot into grooves on the top surface of the NBDs. At the base of these grooves is the Q-loop. By analysing the function of Q-loop mutants in combination with ligand binding cavity mutants I have discovered that the Q-loops are crucial to the transport cycle and that they are required to couple ligand binding to conformational changes at the NBDs necessary to drive the transporter into an inward closed state. 4 (2) ABCB1 is known to be a key component of chemical barrier separating the circulation from the cerebrospinal fluid. It has also been reported to transport β-amyloid across the lumenal membrane and into the circulation. Loss of ABCB1 from the barrier with age has therefore been suggested to play a role in Alzheimer’s Disease. The ubiquitin ligase Nedd4-1 has been implicated in the post-translational regulation of ABCB1 abundance in cells. Here, I report that ABCB1 can be ubiquitinated by Nedd4-1 in vitro and identify the residues modified (by mass spectrometry). (3) Lysophosphatidylinositol (LPI) is an autocrine metabolite produced by cancer cells that binds to the G-protein coupled transmembrane receptor GPR55 on the surface of cells. Stimulation of GPR55 activates a signalling cascade that induces proliferation and metastases of the cancer cells. How LPI is released from the cells was not known. In this study I show that ABCC1 and ABCC3, which are known to be expressed in ovarian and pancreatic cancers, can transport LPI into inside-out vesicles suggesting a new role for these “drug resistance” transporters in cancer biology.

An alcohol and anesthetic binding pocket is hypothesized to exist among transmembrane domains of the α1 glycine receptor (GlyR). Prior work has shown that amino acid residue serine-267 plays a significant role in the enhancing effects of alcohol and anesthetics and is theorized to form part of an alcohol and anesthetic binding cavity among subunit transmembrane domains. Propyl methanethiosulfonate (PMTS), an alcohol-like thiol, was previously shown to bind to a cysteine residue introduced at position 267 (S267C) and this resulted in permanent enhancement of GlyR function. If ethanol is binding to residue 267 in wildtype GlyR to potentiate receptor function then we hypothesized that covalent thiol labeling would produce receptor enhancement by the same mechanisms as ethanol. Using outside-out patch single channel electrophysiology we determined the open and closed dwell-times and burst properties of S267C GlyR in the absence and presence of PMTS. The primary consequence of PMTS binding to S267C GlyR was an increase in the lengths of burst durations, paralleling the main effect of ethanol on wildtype GlyR. Our findings thus provide a new line of evidence suggesting that ethanol is exerting its enhancing effects on the GlyR through its interactions with amino acid residue 267 in the second transmembrane domain.

In this dissertation, we develop statistical and machine learning methods for problems in biological systems and processes. In particular, we are interested in two problems–predicting structural properties for membrane proteins and clustering genes based on microarray experiments. In the membrane protein problem, we introduce a compact representation for amino acids, and build a neural network predictor based on it to identify transmembrane domains for membrane proteins. Membrane proteins are divided into two classes based on the secondary structure of the parts spanning the bilayer lipids: alpha-helical and beta-barrel membrane proteins. We further build a support regression model to predict the lipid exposed levels for the amino acids within the transmembrane domains in alpha-helical membrane proteins. We also develop methods to predict pore-forming residues for beta-barrel membrane proteins. In the other problem, we apply a context-specific Bayesian clustering model to cluster genes based on their expression levels and cDNA copy numbers. This dissertation is organized as follows. Chapter 1 introduces the most relevant biology and statistical and machine learning methods. Chapters 2 and 3 focus on prediction of transmembrane domains for the alpha-helix and the beta-barrel, respectively. Chapter 4 discusses the prediction of relative lipid accessibility, a different structural property for membrane proteins. The final chapter addresses the gene clustering approach.