Termination of protein synthesis involves the recognition of one of three stop codons (UAG, UAA or UGA) and hydrolysis of the nascent polypeptide chain from the peptidyl-tRNA on the ribosome. Unlike sense codons, which are decoded by aminoacyl-tRNAs, stop codons are decoded by proteins known as release factors. The decoding release factors occupy the same site as aminoacyl-tRNA, interacting directly with the stop codon at the decoding centre and inducing peptidyl-tRNA hydrolysis at the peptidyl transferase centre. Eubacteria have two codon-specific decoding release factors - RF1, which recognizes UAG and UAA, and RF2, which recognizes UGA and UAA. Biochemical studies identified two tripeptide 'anticodon' motifs, PXT in RF1 and SPF in RF2, which structural studies have shown occur in exposed loops (anticodon loops) on the surface of the proteins. Structures of isolated release factors show a compact 'closed' conformation whereas structures of release factors bound to the ribosome show them to be in a highly extended 'open' conformation. This suggests that a large conformational change in the release factor must take place upon or before binding to the ribosome. This transition has been invoked as a mechanism for how translational fidelity is maintained (Rawat et al, 2003), however, small angle X-ray scattering data from E. coli RF1 suggest the decoding release factors are also in the open conformation in solution challenging this mechanism. Mora et al. (2003a) presented evidence that swapping the anticodon loop of RF2 with that of RF1 switched the stop codon specificity of the release factor. Recent structures of the decoding release factors bound to the ribosome showed that there was a second structural element of the release factor, the tip of helix α5, involved in recognition of the first base of the stop codon. The objectives of this thesis were to investigate both the anticodon loop and the helix α5 region for their roles in stop codon recognition, and to investigate whether there is a conformational change in the release factors on binding to the ribosome. The anticodon loop was investigated using chimeras of E. coli RF1/RF2 and E. coli RF1/C. elegans mitochondrial RF1 (MRF1) within the anticodon loop. An RF1 variant containing the RF2-specific SPF tripeptide motif did not switch stop codon specificity showing that the tripeptide motifs are not sufficient determinants for the codon specificity of RF1 and RF2 as was originally proposed. Surprisingly repeating the complete swap of the RF1 anticodon loop to that of RF2 did not switch the stop codon specificity as claimed in Mora et al. (2003a). The studies in this thesis identified additional regions of the anticodon loop of the release factor that are important for stop codon recognition. Two of the RF1/RF2 anticodon loop variants produced showed altered codon specificity recognizing all three standard stop codons and the sense codon UGG. These variants provided unexpected insights into the mechanism of stop codon recognition and can explain why there are two release factors in eubacteria. The C. elegans MRF1 contains a novel anticodon loop that is shorter and lacks the classical PXT motif. E. coli RF1/C. elegans MRF1 chimeras showed that this anticodon loop could function in E. coli RF1 and maintain the same codon specificity. While size and sequence within the loop together are important for recognition clearly there is more than one way RF1-type release factors can recognize the UAG and UAA stop codons. Vertebrate mitochondria use four stop codons, two of the standard stop codons, UAA and UAG, and the reassigned arginine codons AGA and AGG. Two vertebrate mitochondrial release factors have been identified, mtRF1a and mtRF1 (renamed here mRF1[Canonical] and mRF1[Noncanonical]). Bioinformatic studies showed mRF1[C] had similar helix α5 and anticodon loop regions to classical RF1s. mRF1[NC] had different helix α5 and anticodon loop regions and was hypothesized to recognize the non-standard stop codons AGA and AGG. E. coli RF1/Human mRF1[NC] chimeras were constructed that showed that the helix α5 and anticodon loop regions are important for stop codon recognition. Nevertheless the chimeras showed poor activity at the AGA and AGG stop codons on E. coli 70S ribosomes suggesting that mRF1[NC] has evolved to function exclusively on 55S mitoribosomes. A release factor variant of RF2 was designed that had the potential to trap this E. coli factor in the closed conformation in solution by disulphide bond formation. The RF2 double cysteine variant was successfully expressed and purified. The disulphide bond between the two cysteines was detected directly by mass spectrometry in a high proportion of molecules, showing the closed form of RF2 exists in solution. The RF2 closed form variant was shown to have release activity concomitant with the proportion of the open form in the RF preparation showing that the conformational change is required for normal release factor function. Preliminary binding studies have suggested that the RF2 closed form variant can bind to the ribosome. The ability of the closed form of RF2 to bind to the ribosome allowed a mechanism of translational fidelity to be proposed from the studies in this thesis; the release factor would recognize the stop codon in the decoding centre and, if cognate, the conformational change would occur allowing peptidyl-tRNA hydrolysis.

PCR-based codon-specific random mutagenesis and site-specific mutagenesis were performed to construct a library of 18 amino acid changes at Arg276 in the conserved leucine-loop of the core catalytic domain of human uracil-DNA glycosylase (UNG). Each Arg276 mutant was then overproduced in E. coli cells and purified to apparent homogeneity by conventional chromatography. All of the R276 mutant proteins formed a stable complex with the uracil-DNA glycosylase inhibitor protein (Ugi) in vitro, suggesting that the active site structure of the mutant enzymes was not perturbed. The catalytic activity of all mutant proteins was reduced; the least active mutant, R276E, exhibited 0.6% of wild-type UNG activity, whereas the most active mutant, R276H, exhibited 43%. Equilibrium binding measurements utilizing a 2- aminopurine-deoxypseudouridine DNA substrate showed that all mutant proteins displayed greatly reduced base flipping/DNA binding. However, the efficiency of UV-catalyzed cross-linking of the R276 mutants to single-stranded DNA was much less compromised. Using a concatemeric [³²P]U·A DNA polynucleotide substrate to assess enzyme processivity, UNG was shown to use a processive search mechanism to locate successive uracil residues, and Arg276 mutations did not alter this attribute. A transient kinetics approach was used to study six different amino acid substitutions at Arg276 (R276C, R276E, R276H, R276L, R276W, and R276Y). When reacted with double-stranded uracil-DNA, these mutations resulted in a significant reduction in the rate of base flipping and enzyme conformational change, and in catalytic activity. However, these mutational effects were not observed when the mutant proteins were reacted with single-stranded uracil-DNA. Thus, mutations at Arg276 effectively transformed the enzyme into a single-strand-specific uracil-DNA glycosylase. The nuclear form of human uracil-DNA glycosylase (LTNG2) was overproduced in E. coli cells and purified to apparent homogeneity. While UNG2 retained ~9 % of UNG activity, it did form a stable complex with Ugi. Paradoxically, low concentrations of NaC1 and MgC1₂ stimulated UNG2 catalytic activity as well as the rate of rapid fluorescence quenching; however, the rate of uracil flipping was reduced. When UNG2 bound pseudouracil-containing DNA, conformational change was not detected. / Graduation date: 2005

Graduation date: 2003

Ion channels are proteins that facilitate ion diffusion across cell membrane. Nevertheless, various groups of ion channels can act as surface receptors and play important roles in signal transduction. Transient Receptor Potential Melastatin 7 (TRPM7) ion channel has been implicated in diverse cellular functions including actomyosin cytoskeletal remodeling and anoxic neuronal death. However the mechanisms behind TRPM7’s physiological roles remain undetermined. TRPM7 possesses unusually long intracellular domains and a functional C-terminal alpha kinase domain that may contribute to regulation of channel activity and signal transduction. We therefore identified proteins that interact with TRPM7 C-terminus. Pull-down assays coupled with LC-MS/MS revealed that cytoskeletal proteins (actin and tubulin) and synaptic vesicle proteins (VAMP2 and SNAP25) associate with the TRPM7. In addition, we further found that TRPM7 does not directly bind microtubules or single dimeric tubulin subunits. Thus one or more microtubule binding proteins is involved in the association between TRPM7 and microtubules.

Ion channels are proteins that facilitate ion diffusion across cell membrane. Nevertheless, various groups of ion channels can act as surface receptors and play important roles in signal transduction. Transient Receptor Potential Melastatin 7 (TRPM7) ion channel has been implicated in diverse cellular functions including actomyosin cytoskeletal remodeling and anoxic neuronal death. However the mechanisms behind TRPM7’s physiological roles remain undetermined. TRPM7 possesses unusually long intracellular domains and a functional C-terminal alpha kinase domain that may contribute to regulation of channel activity and signal transduction. We therefore identified proteins that interact with TRPM7 C-terminus. Pull-down assays coupled with LC-MS/MS revealed that cytoskeletal proteins (actin and tubulin) and synaptic vesicle proteins (VAMP2 and SNAP25) associate with the TRPM7. In addition, we further found that TRPM7 does not directly bind microtubules or single dimeric tubulin subunits. Thus one or more microtubule binding proteins is involved in the association between TRPM7 and microtubules.

Thesis (Ph. D.)--University of Toledo, 2008. / Typescript. "Submitted as partial fulfillment of the requirements for the Doctor of Philosophy in Chemistry." Includes bibliographical references (leaves 245-252).

Thesis (Ph. D.)--University of Texas at Austin, 2002. / Vita. Includes bibliographical references. Available also from UMI Company.

published_or_final_version / Biochemistry / Doctoral / Doctor of Philosophy

published_or_final_version / Biochemistry / Doctoral / Doctor of Philosophy

The ability of a cell to respond precisely to environmental stress depends on the expression of a large number of genes in a finely coordinated manner. One of such genes is CHOP that encodes the CCAAT/Enhancer-Binding Protein Homologous Protein. CHOP is usually expressed to mediate apoptosis under the condition of excessive stress. The expression of CHOP therefore has to be stringently regulated as its expression will determine the fate of a cell under stress. The expression of many genes is regulated at the posttranscriptional level through the metabolism of their mRNA, such as maturation, transport, storage, and degradation of mRNA. Many metabolic processes of mRNA are known to be mediated by RNA-binding proteins that specifically interact with the mRNA. RNA-binding proteins that interact with the CHOP mRNA have until present not been identified. The aim of this study is to investigate what proteins may bind specifically to CHOP mRNA. The study will enable further understanding regarding how the expression of CHOP is regulated in cellular stress response. Proteins extracted from HeLa cells were incubated with a 335bp [3H]-labelled CHOP RNA probe that spans over a part of the coding region and the 3’UTR of CHOP mRNA. Sucrose density gradient ultracentrifugation revealed that after incubation with proteins extracted from HeLa cells, the sedimentation rate of the [3H]-CHOP RNA probe was significantly higher than that of the free [3H]-RNA probe. The formation of heavy molecular complexes involving the [3H]-CHOP RNA probe was therefore suggested. However, no increase in sedimentation rate of the [3H]-CHOP RNA probe was observed in the presence of an excess of unlabelled CHOP RNA probe. Similar observations were made when the experiments were performed using proteins isolated from cells treated with As2O3. Two putative sequence elements, the Adenylate-Uridylate-Rich Element (ARE) and the Putative Regulatory Element (PRE) located respectively in the 3’UTR and coding region of the CHOP mRNA were then examined for their involvement in RNA-protein interaction. The deletion of ARE and/or PRE, from the [3H]-CHOP RNA probe had little effect on the binding of the RNA probe to the HeLa cell proteins. Consistently, unlabelled CHOP RNA probes with the same deletions were only slightly weaker in competing with the intact [3H]-CHOP RNA probe to bind to HeLa cell proteins. Human Antigen R (HuR) was identified by Western blot analysis to be present in the proteins that were obtained by pull-down assays using biotinylated CHOP RNA as a probe. The deletion of ARE and/or PRE resulted in a slight reduction of HuR obtained by pull down assays. This study provides the first evidence that physical binding interaction occurs between intracellular RNA-binding proteins and CHOP mRNA. More importantly, one such protein is HuR. Data suggest that HuR binding to the CHOP mRNA is mediated by sequences in the CHOP mRNA other than ARE and PRE. / published_or_final_version / Biochemistry / Master / Master of Philosophy