A Dissertation submitted to the Faculty of Science, University of Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science. 2015. / Prdm1/Blimp1 is a transcription factor whose mechanism of action is mainly repression; however it has been identified as an activator in some cases. As a transcriptional repressor, it plays multiple roles during embryonic development, including neural crest specification. Prdm1 acts by repressing large sets of genes via sequence specific recruitment of co-repressors, many of which are epigenetic modifiers. Neural crest is a transient, migrating cell population that gives rise to a number of diverse cell lineages that form important structures in the vertebrate embryo. Examples of these include peripheral nervous system, melanocytes and cranial cartilage. Prdm1 is expressed during neural crest specification in Xenopus, zebrafish and lamprey. The expression of Prdm1 had not yet been investigated in the neural crest during chick embryonic development. The mechanism of Prdm1 action or the nature of possible binding partners that mediate its effects in the neural crest had not yet been addressed. Prdm1 binding partners are known to play important roles during embryonic development, yet in many cases no spatiotemporal expression analysis during early vertebrate development has been performed. Single and double in situ hybridization for Prdm1 and all the binding partners was performed to determine localization of mRNA during early stages of chick embryonic development. We report the expression patterns of Prdm1 and seven of its known or putative binding partners (Hdac1, Hdac2, Tle1, Tle3, G9a, Prmt5 and Lsd1) during early stages (HH4-HH10) of chicken embryogenesis. Prdm1 expression was observed in the neural plate border and pre-migratory neural crest during chick development. Six Prdm1 binding partners (except Tle1) are co- expressed with Prdm1 in the prospective neural plate border at HH4-HH6, and all seven show strong and specific expression in the neural plate border at HH7-HH8, suggesting all of them co-operate with Prdm1 during neural crest development in chick embryos. Future work will focus on protein interaction studies in order to directly demonstrate the association between Prdm1 and the binding partners it co-localizes with.

The G-protein coupled receptors (GPCRs) are seven-transmembrane receptors that transduce signals from the cell surface to intracellular effectors. There are more than 1000 GPCRs in metazoans, while no GPCR has been definitively identified in plants. The most promising plant GPCR candidate, Arabidopsis G-protein coupled receptor 1 (GCR1), physically couples to the G-protein < subunit GPA1 and is involved in cell cycle regulation, blue light and phytohormone responses, but its signalling network remains largely unknown. This project aimed to achieve a better understanding of GCR1 signalling by identifying its interactors using a novel yeast two hybrid system – the Ras Recruitment System (RRS). Screening of an Arabidopsis cDNA library using a bait comprising intracellular loop 1 (i1) and 2 (i2) of GCR1 resulted in the isolation of 20 potential interactors. Extensive reconfirmation screening demonstrated that three of these interactors: Thioredoxin h3 (TRX3), Thioredoxin h4 (TRX4) and a DHHC type zinc finger family protein (zf-DHHC1) interact specifically with both i1 and i2 of GCR1. This was supported by the reverse RRS (rRRS) and 6xHis-pull-down assays. It is speculated that TRX3 and TRX4, which can reduce disulfide bridges of target proteins and act as powerful antioxidants, may regulate GCR1-mediated signalling events in response to oxidative stress. Alternatively, they may modulate GCR1 targeting or signalling through their chaperone activities. zf-DHHC1 has a predicted membrane topography that is shared by most DHHC domain-containing palmitoyl acyl transferases. It may modify GCR1 activity through palmitoylation of the two cysteines located at the cytoplasmic end of the first transmembrane domain. Together, these findings contribute to the growing understanding of the GCR1 signalling network, and provide valuable starting points for further investigation.

Le terme de sulfoprotéomique est utilisé pour désigner l’étude de la sulfatation des protéines. Bien que la sulfatation soit depuis peu considérée comme une MPT d’une importance majeure, il y a toujours peu de travaux scientifiques qui y sont consacrés en comparaison avec ce qui se fait sur la phosphorylation notamment. Ce retard s’explique notamment par la difficulté à analyser les espèces protéiques sulfatées dans les conditions classiques utilisées en protéomique, notamment par spectrométrie de masse. Ces travaux de thèse visent justement à développer des méthodes d’analyses par spectrométrie de masse dédiées à l’étude de la sulfatation des protéines, afin d’augmenter le champ des connaissances de cette MPT. Pour cela, nous avons largement utilisé le mode d’ionisation négatif, très peu, voire jamais utilisé en protéomique, avec deux techniques de fragmentation pour réaliser des spectres MS/MS, à savoir les fragmentations CID et HCD. Les résultats obtenus nous ont permis de mettre en évidence une méthode d’analyse permettant la formation d’ions spécifiques de la sulfatation et de la phosphorylation (qui sont isobariques), permettant ainsi une identification certaine de chacune des deux MPTs. Nous avons également entrepris d’étudier le rôle de la sulfatation d’un récepteur cellulaire, CXCR4, dans son interaction avec son ligand naturel, la chimiokine SDF-1/CXCL12. Cette étude a été menée par électrophorèse capillaire, et pourra constituer une base de travail solide pour des futures analyses mettant en œuvre le couplage entre l’électrophorèse capillaire et la spectrométrie de masse pour une meilleure caractérisation des complexes formés entre les partenaires protéiques. / Sulfoproteomics term designs protein sulfation studies. It appears during the 2000’s, when the interest for others Post-Translational Modifications (PTMs) than phosphorylation and glycosylation was growing up. Even though sulfation is thought to be an important PTM, a weak number of publications has emerged about it, notably if we compare with the huge quantity of phosphorylation papers. This difference is mainly due to the difficulty to correctly analyze sulfated proteins and peptides in the classical ways of proteomics, as in mass spectrometry for example. The goal of this thesis is to develop mass spectrometry methods dedicated to the characterization of sulfated species, in order to improve the knowledge of this PTM. To do that, we have mainly used negative ion mode, which is almost never used, with two fragmentations techniques for the MS/MS spectra, which are CID and HCD. Results obtained allow us to pinpoint an analytical method allowing the differentiation between sulfation and phosphorylation (they are isobaric), based on the presence of specific ion for each PTM in MS/MS. In another part of the project, we have investigated the role of sulfation in the interaction between a cellular receptor, CXCR4, and its in vivo ligand, the chemokine SDF-1/CXCL12. We used capillary electrophoresis for this work, and it could be a good basis for future analyses using capillary electrophoresis coupled with mass spectrometry, in order to have a better characterization of the observed complexes.

前體信使核糖核酸（pre‐mRNA）的剪接是在RNA成熟與蛋白質多樣性發生中所必需的一類高度動態的過程。作為一類特定的非小核糖核蛋白剪接因子，絲氨酸精氨酸（SR）蛋白在mRNA的組成型剪接及選擇性剪接，mRNA的轉運與翻譯中均扮演關鍵角色。SR蛋白在其氮端含有1個或2個RNA識別基序（RRMs），其碳端的RS結構域含有連續排列且可被高度磷酸化的精氨酸絲氨酸（RS）二肽。SR蛋白的磷酸化水平可調節其亞細胞定位與生理功能，而屬於蛋白激酶超家族的SR蛋白激酶（SRPK）家族負責SR蛋白的磷酸化修飾。 / 在此項課題中，我們著重於SRPK2獨特的底物特異性及其磷酸化機制的研究。課題選用兩個代表不同類型的底物：人類絲氨酸精氨酸剪接因子1（SRSF1）和人類細胞凋亡染色質聚縮引導因子S（acinusS）。研究結果顯示，氮端非激酶區為SRPK2對SRSF1和acinusS的激酶活力所必需。另外，雖然兩種底物類型一級結構迥異，但一個位於SRPK2的大葉且保守的docking groove，負責對它們的識別與結合。 / SRPK1以processive機制催化SRSF1中8‐10個位點，而我們的實驗結果顯示SRPK2以processive機制磷酸化SRSF1的約5‐6個位點。我們證明，SRPK2的docking groove對processive機制的磷酸化有著重要作用，而且位於dockinggroove中的組氨酸601決定了SRPK2較低的processvity。有趣的是，SRPK2的docking groove也在acinusS絲氨酸422的位點特異性磷酸化中起關鍵作用。我們證明該位點特異的磷酸化機制主要是由SRPK2的docking groove與位於acinusS磷酸化位點氮端推定的docking motif之間的離子型相互作用，及其隨之與一個同樣位於acinusS的磷酸化位點N端負的電荷區域之間的離子型排斥作用所調節。 / 這些結果顯示，SRPK2的docking groove採取了兩種不同的磷酸化機制，因而其底物可以或者processive機制，或者高度位點特異的機制被磷酸化修飾。此外，為闡明此兩種迥異的磷酸化機制的分子基礎，蛋白質晶體學研究正在進行之中。 / Pre‐mRNA splicing is a highly dynamic process that plays an essential role in mRNA maturation and protein diversity generation. One particular family of non‐small nuclear ribonucleoproteins (snRNPs) splicing factors, the serinearginine (SR) proteins, play critical roles in both constitutive and alternative mRNA splicing, mRNA transport, and translation. N‐terminus of SR proteins consists one or two RNA recognition motifs (RRMs), and the C‐terminal RS domain contains continuous RS dipeptides that could be extensively phosphorylated. The phosphorylation states of SR proteins regulate their subcellular localization and physiological functions. SR protein kinase (SRPK) family is a member of the kinase superfamily that accounts for SR protein phosphorylation. / In this study, we focused on the distinct substrate specificity and phosphorylation mechanism of SRPK2. Two substrates representing different classes are selected: human serine/arginine splicing factor 1 (SRSF1) and human apoptotic chromatin condensation inducer in the nucleus S (acinusS). Our results showed that the N‐terminal non‐kinase region of SRPK2 is required for the full catalytic activity towards both SRSF1 and acinusS. Besides, a conserved docking groove in the large lobe of SRPK2 was shown responsible for the recognition and binding of both substrate classes despite the significant difference in their primary structures. / While SRPK1 modifies SRSF1 for 8‐10 sites in a processive manner, our results show that SRPK2 processively phosphorylates SRSF1 for approximately 5‐6 sites. We provided evidence that the docking groove of SRPK2 is important for the processive phosphorylation mechanism and His601 within the groove accounts for the lower processivity. Interestingly, the docking groove also plays a critical role in the site‐specific phosphorylation of acinusS at Ser422. We demonstrated that the single site phosphorylation mechanism of SRPK2 is mainly regulated by ionic interaction with a putative docking motif, and the following ionic repulsion between the docking groove and an electronegative region N‐terminal to the P‐site of acinusS. / These results suggest that the docking groove of SRPK2 adopts two distinct phosphorylation mechanisms so that different RS domains can be phosphorylated in either processive or highly site‐specific manner. Protein crystallography studies are undergoing to provide the molecular basis of the two distinct phosphorylation mechanisms. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Liang, Ning. / Thesis (Ph.D.) Chinese University of Hong Kong, 2014. / Includes bibliographical references (leaves 151-170). / Abstracts also in Chinese.

氫化酶作為一種催化劑，能催化氫分子成為質子及電子的相互轉換。 [鎳鐵]- 氫化酶散播最廣的一種氫化酶，從古菌到細菌都能找到 [鎳鐵]- 氫化酶。完整成熟的 [鎳鐵]-氫化酶需要插入鐵、氰化物、一氧化碳以及鎳到它的催化核心。這複雜的過程需要其它由若干 hyp 基因編譯的輔助蛋白酶的幫助，其中蛋白HypA 與 HypB 負責將鎳運送到[鎳鐵] -氫化酶的催化核心。敲除了 hypA 或hypB 基因的細菌株缺失[鎳鐵] -氫化酶的活性，如在生長介質裡添補鎳可恢復部份[鎳鐵] -氫化酶的活性。當HypB 與鳥嘌呤核苷酸結合時會變成蛋白二聚體。對比HypB 脫輔基蛋白及與HypB 與鳥嘌呤三核苷酸類似物的蛋白複合物的晶體結構可發現，HypB 透過一個保守賴氨酸殘基( Archaeoglobus fulgidus HypB 的殘基 148 )組成分子間鹽橋以構成蛋白二聚體。Escherichia coli 的體內實驗顯示，此保守賴氨酸殘基對活性氫化酶的製造起必要的作用，反映由此殘基所構成的鹽橋對HypB 功能的重要性。此外，本研究展示了A. fulgidusHypA 及 HypB 蛋白之間的相互作用。通過在A. fulgidus HypB 上進行系統性的突變，發現HypB 利用其GTP 酶域上的一段氨基端區域與HypA 相互作用。跟據這個結果，我們進而在E. coli HypB 上發現了兩個保守的非極性殘基與HypA 相互作用。當以丙氨酸取代在HypB 上的這兩個非極性殘基時，HypB 無法激活E. coli 中的氫化酶，導置降低的氫化酶活性，這表明了HypA 和HypB 的相互作用對[鎳鐵] -氫化酶成熟過程的必要性。 / Hydrogenases catalyze the inter-conversion of molecular hydrogen into protons and electrons. [NiFe]-hydrogenase is the most widely distributed hydrogenases, which is found in organisms ranging from archaea to bacteria. Maturation of [NiFe]-hydrogenase requires the insertion of iron, cyanide and carbon monoxide, followed by nickel, to the catalytic core of the enzyme. The maturation process of hydrogenase is a complicated procedure, which requires many accessory proteins encoded by hyp genes. HypA and HypB participate in the nickel delivery step to the catalytic core of hydrogenase, which is supported by the fact that strain deficient in hypA or hypB gene lack hydrogenase activity which can be recovered partially by elevating nickel content in the medium. HypB is capable to form dimer in solution upon guanine nucleotide binding. By comparing the crystal structures of HypB in dimer and monomer form, an important lysine residue (residue 148 in A. fulgidus HypB) which is required to form an intermolecular salt bridge during GTP-dependent dimerization, has been identified. Substitution of this lysine resiue with alanie would break HypB dimer in vitro. In vivo complementation study in E. coli showed that the corresponding lysine residue in E. coli HypB is required for active hydrogenase production indicating the importance of this intermolecular salt bridge to the biological function of HypB. Besides, interaction between A. fulgidus HypA and HypB are demonstrated in this work. By making systematic mutation to A. fulgidus HypB, the N‐terminal region of the GTPase‐domain has been identified to be important for its interaction with HypA. Further mutagenesis study has been done on E. coli HypB and two conserved non‐polar residues responsible for interaction with HypA have been identified. Alanine substitution of these conserved non‐polar residues result in HypB mutants which failed to rescue hydrogenase activity in vivo in E. coli showing that HypA/HypB interaction is required for hydrogenase maturation. / Detailed summary in vernacular field only. / Chan, Kwok Ho. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 88-95). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese. / Chapter Chapter 1 --- Introduction: Hydrogenase biosynthesis requires insertion of nickel facilitated by protein HypA and HypB --- p.1 / Chapter 1.1 --- What is hydrogenase? --- p.1 / Chapter 1.2 --- [NiFe] hydrogenase contains a complex catalytic core composed of metal atoms and diatomic ligands --- p.2 / Chapter 1.3 --- The [NiFe] catalytic core --- p.4 / Chapter 1.4 --- Building the catalytic [NiFe] core --- p.4 / Chapter 1.5 --- Nickel insertion into the hydrogenase precursor involves the proteins HypB, HypA and SlyD --- p.7 / Chapter 1.5.1 --- Protein HypB --- p.7 / Chapter 1.5.2 --- Protein HypA --- p.11 / Chapter 1.5.3 --- Protein SlyD --- p.12 / Chapter 1.6 --- Objectives - How HypB dimerization and HypA/HypB interaction are involved in hydrogenase maturation process? --- p.13 / Chapter Chapter 2 --- A conserved Lys residue is required for GTP-dependent dimerization and hydrogenase maturation --- p.17 / Chapter 2.1 --- Introduction --- p.17 / Chapter 2.2 --- Materials and Methods --- p.22 / Chapter 2.2.1 --- Recombinant Plasmid Construction --- p.22 / Chapter 2.2.2 --- HypB mutant construction by site-directed Mutagenesis --- p.22 / Chapter 2.2.3 --- Protein Expression and purification --- p.23 / Chapter 2.2.4 --- HypB protein purification --- p.23 / Chapter 2.2.5 --- Analytical gel filtration chromatography coupled with Light Scattering (SEC/LS) --- p.24 / Chapter 2.2.6 --- Nucleotide binding affinity determination --- p.25 / Chapter 2.2.7 --- GTPase activity determination --- p.26 / Chapter 2.2.8 --- Sample preparation for hydrogenase activity assay --- p.26 / Chapter 2.2.9 --- Hydrogenase activity determination --- p.27 / Chapter 2.3 --- Results --- p.29 / Chapter 2.3.1 --- AfHypB undergoes GTP-dependent dimerization --- p.29 / Chapter 2.3.2 --- Analysis of Structural difference between the apo form and GTP S-bound form suggests a mechanism of GTP-dependent dimerization for HypB --- p.30 / Chapter 2.3.3 --- Lys-148 is essential for GTP-dependent dimerization --- p.31 / Chapter 2.3.4 --- Disruption of dimerization by K148 mutation did not affect nucleotide binding and GTP hydrolysis activity significantly --- p.32 / Chapter 2.3.5 --- The conserved lysine residue is required for hydrogenase maturation in E. coli --- p.33 / Chapter 2.4 --- Discussion --- p.45 / Chapter 2.4.1 --- A conserved intermolecular salt‐bridge is required for GTP-dependent dimerization of HypB and hydrogenase maturation --- p.45 / Chapter 2.4.2 --- The extra metal binding site at the dimeric interface of HypB may provide a mechanism of why GTP-dependent dimerization is essential to Ni insertion --- p.46 / Chapter Chapter 3 --- N-terminal region of GTPase‐domain of HypB is required for interaction with HypA --- p.51 / Chapter 3.1 --- Introduction --- p.51 / Chapter 3.2 --- Methods and materials --- p.53 / Chapter 3.2.1 --- Recombinant Plasmid Construction --- p.53 / Chapter 3.2.2 --- HypB variant construction by site‐directed Mutagenesis --- p.53 / Chapter 3.2.3 --- Protein Expression --- p.54 / Chapter 3.2.4 --- Tag‐free AfHypA and AfHypB purification --- p.54 / Chapter 3.2.5 --- Analytical size exclusion chromatography coupled with Light Scattering --- p.54 / Chapter 3.2.6 --- GST pull‐down of GST‐AfHypA and AfHypB --- p.55 / Chapter 3.2.7 --- Tandem affinity pull‐down of GST‐EcHypA and His‐SUMO‐EcHypB --- p.55 / Chapter 3.2.8 --- GST pull‐down of GST‐EcHypA and His‐SUMO‐EcHypB --- p.56 / Chapter 3.2.9 --- Hydrogenase activity determination --- p.57 / Chapter 3.3 --- Results --- p.58 / Chapter 3.3.1 --- HypA and HypB from A. fulgidus form 1:1 heterodimer in solution --- p.58 / Chapter 3.3.2 --- The N‐terminal regions upstream of the first helix of A. fulgidus HypB is required for HypA-HypB interaction --- p.59 / Chapter 3.3.3 --- Two conserved hydrophobic residues on HypB from E. coli are required to interact with HypA --- p.60 / Chapter 3.3.4 --- HypA-HypB interaction is required for hydrogenase maturation in E. coli --- p.62 / Chapter 3.4 --- Discussion --- p.73 / Chapter 3.4.1 --- The N‐terminal region of the GTPase domain is required for interaction with HypA and hydrogenase maturation in E. coli --- p.73 / Chapter 3.4.2 --- Location of interaction site on HypB reveals possible role for HypA/HypB interaction --- p.74 / Chapter 3.4.3 --- Mode of specific interaction with HypA: Interaction via a disordered region implies a coupled folding and binding process --- p.75 / Chapter Chapter 4 --- Conclusion and Future Perspectives --- p.80 / Chapter A1.1 --- Summary of findings in this work --- p.80 / Chapter A1.2 --- Implications in hydrogenase maturation --- p.81 / Chapter A1.3 --- Questions unresolved --- p.82 / Chapter 4.3.1 --- Factors that activate GTPase activity of HypB are still elusive --- p.82 / Chapter 4.3.2 --- How nickel delivery is regulated by HypA/HypB complex is still unclear --- p.83 / References --- p.88 / Chapter Appendix 1 --- Preliminary results of HypA/HypB protein complex structural study --- p.96 / Chapter A1.1 --- Structural study may provide invaluable insights to the role of HypA‐HypB interaction --- p.96 / Chapter A1.2 --- X‐ray crystallography as an approach to determine HypA/HypB complex structure --- p.96 / Chapter A1.3 --- Initial crystal hits were obtained with purified AfHypA/HypB complex --- p.97 / Chapter Appendix 2 --- Publications associated to the thesis --- p.100 / Chapter Appendix 3 --- Constructs and Primers used --- p.101

Protein deposition in the form of amyloid fibrils is the hallmark of more than 40 human pathologies, including Alzheimer's disease (AD) and Parkinson's disease (PD). Misfolded protein oligomers formed as intermediates during the aggregation process have been strongly implicated in the onset and progression of these diseases. In this thesis, I describe our efforts to uncover molecular agents that can reduce the toxicity caused by protein aggregation via targeting the generation, the physiochemical properties or the membrane affinity of oligomeric species. We employed an integrative approach combining in vitro techniques, including chemical kinetics, atomic force microscopy, and biophysical measurements, and in vivo methods, including neuroblastoma cells and C. elegans models of AD and PD, to identify a range of small molecules and antibodies that can suppress the toxicity related to protein aggregation through a variety of mechanisms. In Chapter 3, we show that the deleterious effects of protein aggregation can be suppressed in AD and PD worms by interfering with the aggregation rates of the amyloid-β peptide (Aβ) and the α-synuclein protein (αS). In Chapter 4, we resolve the mechanism of action for a molecule that enhances the rate of Aβ42 aggregation in AD worms with the result that toxicity is reduced, and find that it potentiates the secondary nucleation microscopic step in vitro. In Chapter 5, we characterize molecules and antibodies that modify the physiochemical properties and self-association of oligomers comprised of several proteins into clusters with reduced diffusibility. In Chapter 6, we classify a family of molecules that protect the cell by displacing several types of oligomeric species from the membrane through a generic mechanism. These results demonstrate strategies by which one can target the aggregation process to alter its resulting toxicity, provide insight into modifying the properties of the most deleterious species associated with protein aggregation and suggest that the protection of the cell from the oligomer-induced cytotoxicity associated with numerous protein misfolding diseases is a promising strategy to combat protein misfolding diseases.

The transition period for dairy cows is defined as the three weeks pre and postpartum. During the transition period, dairy cows experience a myriad of metabolic, managerial, and nutritional requirement changes. These changes lead to stress and increased susceptibility to diseases which can negatively affect lactational performance in the short and long term. However, dietary amino acid availability can have a dramatic impact on the health and performance of dairy cows around parturition. Thus, the objective of the thesis was to evaluate the effects of supplementing yeast-derived microbial protein, as an alternative protein source for dairy cows during the transition period. This was accomplished by using visual observations and precision dairy monitoring technologies to record disease, feeding behavior, and performance of dairy cows from 21 days prepartum to 150 days postpartum. Yeast-derived microbial protein was found to decrease dry matter intake but not negatively affect milk production or health of the animals. Yeast-derived microbial protein may be used as an alternative protein source for transition dairy cows as it did not negatively affect milk production or health of the animals.

The implication of the transmembrane signalling Receptor Tyrosine Kinases (RTKs) in cancer has accelerated the pursuit for drugs to target these molecules. In the process our understanding of how these membrane bound molecules are entangled in cell signalling has significantly expanded. There is now evidence that RTKs can facilitate the formation of a lattice-type network of signalling molecules to elicit whole cell responses to external ligand stimuli. Although beginning to be unravelled, knowledge pertaining to the mechanisms of molecular control that initiate these signalling pathways is still in its infancy. In this thesis, a random mutagenesis approach allowed the identification of the crucial interaction surfaces between membrane-bound EphA3 and its preferential binding partner ephrinA5, that are required to induce the formation of higher-order Eph signalling complexes. Modelling and experimental dissection of this co-ordinated receptor aggregation has provided detailed insights into the molecular mechanisms of Eph receptor activation, which in some aspects may also apply to other members of the RTK family. In particular, the importance of certain molecular interfaces in determining preferential and non-preferential Eph/ephrin interactions, suggests their role in the selection of biologically important binding partners. In addition to the assignment of the ephrin-interaction surfaces, the random mutagenesis strategy also identified a continuous conformational epitope as binding site for an anti-EphA3 monoclonal antibody. Fortuitously, antibody binding to this site functionally mimics ephrin stimulation of EphA3 positive cells, and in particular together with divalent ephrinA5, yields synergistically enhanced EphA3 activation. Elucidation of the underlying mechanism has provided opportunities to develop an efficient EphA3 targeting mechanism that is based on increased affinity and accelerated ephrinA5 uptake as consequence of this unique activation mechanism. On a genetic level, novel oligonucleotide analogues known as Peptide Nucleic Acids (PNAs) were analysed for their ability to sterically inhibit EphA3 DNA transcription and suggest a dosedependent downregulation of EphA3 expression, in malignant melanoma cells. Combined, ephrinA5, the anti-EphA3 MAb (IIIA4) and PNA, offer the possibility to investigate the specific machinery involved in Eph receptor expression and signalling for the specific targeting of EphA3 expressing tumour cells.

Termination of protein synthesis occurs in response to the translocation of a stop codon (UAA, UAG or UGA) into the A site of the ribosome. Unlike sense codons, stop signals in the mRNA are recognized by two classes of specialized proteins called release factors (RFs): the class I or decoding RF, which recognizes the stop codon and promotes peptidyl-tRNA hydrolysis and class II RF, a G-protein that promotes the dissociation of the decoding RF from the ribosome. The discovery that stop codons are decoded by a protein factor rather than a specific tRNA opened up the possibility that the signal for termination of protein synthesis might extend beyond the stop codon itself. Biochemical and genetic experiments in prokaryotes confirmed that bias in nucleotide usage around stop codons correlates with translation termination efficiency. The objective of the current investigation was to define the eukaryotic termination signal by determining the bias in the nucleotide sequence surrounding eukaryotic stop codons and to identify whether this was a determinant of translation termination efficiency. Bioinformatic analysis of five diverse eukaryotic genomes was undertaken to identify potential eukaryotic translation termination signal elements. Significant nucleotide bias was identified both 5� and 3� of the stop codon in all the genomes investigated. Correlations were identified between nucleotide bias and gene expression levels, and between nucleotide bias and natural recoding sites predicting that nucleotides 5� and 3� of the stop codon affect termination efficiency. These correlations were common to all organisms investigated and suggested the existence of a eukaryotic termination signal. Termination signals identified from the bioinformatic analysis were assayed to determine the efficiency of termination in an in vitro dual luciferase reporter assay. Results indicated that nucleotides both 5� and 3� of the stop codon could significantly alter termination signal efficiency, although readthrough did not vary by greater than 1%. The effect of nucleotides 3� to the stop codon on termination efficiency was investigated further in mammalian cultured cells using the dual luciferase reporter assay. Results showed a significant relationship between the identity of these nucleotides and observed termination efficiencies with nucleotides at positions +4 and +8 giving the strongest correlation. Termination sequence elements of the form UGA CUN NCN mediated up to 5% readthrough in cultured cells. Investigations into the underlying mechanisms that were responsible for the variation in termination efficiency were also undertaken. Co-transfection of specific suppressor tRNAs enhanced but did not change the pattern of observed termination efficiency, indicating that the mechanisms mediated by the termination signal element was not mediated through suppressor tRNA binding. Alignments of 18S rRNA sequences indicated potential extensive interactions between the rRNA and the mRNA termination signal element. Experiments that assessed the effect of eRF1 levels on termination at inefficient termination signals in vitro revealed that increased levels of eRF1 could improve termination efficiency. These results indicate that, as in prokaryotes, specific nucleotides beyond the stop codon modulate translation termination efficiency in eukaryotes, and that the translation termination signal should be considered a sequence element.

Protein:protein interactions are central to the regulation of the intrinsic programmed cell death (apoptosis) pathway. Opposing members of the Bcl-2 family of proteins, which have distinct sequence features, interact with each other on the outer mitochondrial membrane to regulate apoptosis. Pro-survival proteins such as Bcl-2, Bcl-x[L], Bcl-w, Mcl-1 and A1 protect cells from apoptosis and contain up to four regions of homology to Bcl-2 (Bcl-2 homology domains 1 - 4, BH1-4). Pro-apoptotic BH3-only proteins such as Bim, Puma, Noxa, Bad, Bmf, and Bid promote apoptosis by interacting with and inactivating pro-survival proteins, and contain just the BH3-domain. The pro-apoptotic proteins Bax and Bak are essential for apoptosis and contain three regions of homology to Bcl-2 (the BH1-, BH2- and BH3-domains). In this study, two different sets of interactions involving pro-survival proteins were investigated. Initially, the pro-apoptotic protein Bnip3 was examined to determine if it was a mitochondrial anchor for the pro-survival protein Bcl-w. Secondly, to characterise the interactions between a pro-survival protein and different BH3-domains, structures were solved of the pro-survival protein A1 in complex with four different BH3-domains. In the structure of Bcl-w, the hydrophobic C-terminus is bound to its own BH3-domain binding groove. This location of the C-terminus is consistent with the observation that Bcl-w is only loosely associated with the outer mitochondrial membrane in healthy cells. Upon interaction of Bcl-w with a BH3-domain, Bcl-w becomes tightly associated with the mitochondrial membrane, presumably due to displacement of the C-terminal residues by the BH3-only protein. In healthy cells it has been suggested that Bcl-w is associated with the membrane due to an interaction with an unidentified membrane protein, which preliminary experiments suggested may be Bnip3. Protein interaction experiments performed in vitro and in vivo did not reveal an interaction between Bnip3 and Bcl-w. It was originally thought that each pro-apoptotic BH3-only protein could interact with all pro-survival proteins. However, it has recently become clear that there is selectivity within the pathway suggesting functional groupings. Bim and Puma behave as originally predicted and can interact with all pro-survival proteins and are potent killers. In contrast, Noxa and Bad interact with distinct subsets of pro-survival proteins. Noxa only binds Mcl-1 and A1, while Bad binds Bcl-2, Bcl-x[L] and Bcl-w. As a result, either Noxa or Bad acting alone is a weak killer, but together they are potent. Other BH3-only proteins bind tightly to some pro-survival proteins and weakly to others. The diversity that exists between BH3-domain sequences precludes sequence-based identification of the determinants of specificity. In this study, crystal structures of A1:Puma BH3-domain, A1:Bmf BH3-domain, A1:Bak BH3-domain and A1:Bid BH3-domain complexes have been solved. Differences identified between these structures explain some of the variation in affinities observed in pro-survival protein:BH3-domain complexes. These observations, in combination with published data, suggest that BH3-domains bind weakly when the optimal interactions with conserved residues cannot be formed. Additionally, differences were observed in the A1:Bak BH3-domain structure that may be functionally important for the regulation of Bak.