Résumé: Les récepteurs du glutamate de type NMDA (NMDAR) sont des canaux ioniques impliqués dans les phénomènes de plasticité de la transmission synaptique dans le système nerveux central, des mécanismes supposés être à la base du développement neuronal, de l’apprentissage et de la formation de la mémoire. Les NMDAR forment des tétramères à la membrane plasmique, constitués de deux sous-unités obligatoires GluN1 et deux sous-unités variables GluN2 (GluN2A-D) ou GluN3. Dans le prosencéphale, les récepteurs comportant les sous-unités GluN2A (GluN2A-NMDAR) et GluN2B (GluN2B-NMDAR) sont les plus abondants et présentent des profils d’expression différents au cours du développement, les GluN2B-NMDAR étant fortement exprimés aux stades précoces tandis que l’expression des GluN2A-NMDAR augmente progressivement au cours du développement postnatal. Des contributions relatives de ces deux sous-types majoritaires de NMDAR aux propriétés de signalisation distinctes dépendent directement les phénomènes de plasticité neuronale, tels que l’adaptation des synapses glutamatergiques et des circuits neuronaux excitateurs. Bien que la régulation moléculaire des NMDAR ait fait l’objet d’intenses recherches ces dernières décennies, la localisation précise de ces deux sous-types de récepteurs dans la membrane postsynaptique demeurait méconnue. Pour répondre à cette question, nous avons étudié la distribution des NMDAR à la surface de neurones d’hippocampe de rats en combinant deux techniques de microscopie de super-résolution - la microscopie de reconstruction optique stochastique directe (dSTORM) et la déplétion d’émission stimulée (STED) - permettant de dépasser la limite de résolution inhérente à la diffraction de la lumière. Ces techniques nous ont permis de mettre en évidence que les sous-types de récepteurs GluN2A- et GluN2B-NMDAR présentent une nano-organisation différente à la surface neuronale. En effet, ils sont organisés en structures nanoscopiques (nanodomaines) qui diffèrent en nombre, en surface et en morphologie, notamment au niveau des synapses. Au cours du développement, l’organisation membranaire des deux sous-types de NMDAR évolue, avec en particulier de profonds changements de distribution des GluN2A-NMDAR. De plus, cette organisation nanoscopique est impactée différemment par des modulations de l’interaction avec les protéines d’échafaudage à domaine PDZ ou de l’activité de la kinase CaMKII suivant le sous-type de NMDAR considéré. En effet, la réorganisation des GluN2A-NMDAR implique principalement des changements de nombre de récepteurs dans les nanodomaines sans modification de leur localisation, tandis que la réorganisation des GluN2B-NMDAR passe essentiellement par des modifications de localisation des nanodomaines sans changements du nombre de récepteurs qu’ils contiennent. Ainsi, les GluN2A- et GluN2B-NMDAR présentent des nano-organisations différentes dans la membrane postsynaptique, reposant vraisemblablement sur des voies de régulation et des complexes de signalisation distincts. / NMDA-type glutamate receptors (NMDARs) are a type of ion permeable channels playing critical roles in excitatory neurotransmission in the central nervous system by mediating different forms of synaptic plasticity, a mechanism thought to be the molecular basis of neuronal development, learning and memory formation. NMDARs form tetramers in the postsynaptic membrane, most generally associating two obligatory GluN1 subunits and two modulatory GluN2 (GluN2A-D) or GluN3 (GluN3A-B) subunits. In the hippocampus, the dominant GluN2 subunits are GluN2A and GluN2B, displaying different expression patterns, with GluN2B being highly expressed in early development while GluN2A levels increase gradually during postnatal development. In the forebrain, the plastic processes mediated by NMDARs, such as the adaptation of glutamate synapses and excitatory neuronal networks, mostly rely on the relative implication of GluN2A- and GluN2B-containing NMDARs that have different signaling properties. Although the molecular regulation of synaptic NMDARs has been under intense investigation over the last decades, the exact topology of these two subtypes within the postsynaptic membrane has remained elusive. Here we used a combination of super-resolution microscopy techniques such as direct stochastic optical reconstruction microscopy (dSTORM) and stimulated emission depletion (STED) microscopy to characterize the surface distribution of GluN2A- or GluN2B-containing NMDARs. Both dSTORM and STED microscopy, based on different principles, enable to overcome the resolution barrier due to the diffraction limit of light. Using these techniques, we here unveil a differential nanoscale organization of native GluN2A- and GluN2B-NMDARs in rat hippocampal neurons. Both NMDAR subtypes are organized in nanoscale structures (termed nanodomains) that differ in their number, area, and shape. These observed differences are also maintained in synaptic structures. During development of hippocampal cultures, the membrane organization of both NMDAR subtypes evolves, with marked changes for the topology of GluN2A-NMDARs. Furthermore, GluN2A- and GluN2B-NMDAR nanoscale organizations are differentially affected by alterations of either interactions with PDZ scaffold proteins or CaMKII activity. The regulation of GluN2A-NMDARs mostly implicates changes in the number of receptors in fixed nanodomains, whereas the regulation of GluN2B-NMDARs mostly implicates changes in the nanodomain topography with fixed numbers of receptors. Thus, GluN2A- and GluN2B-NMDARs have distinct organizations in the postsynaptic membrane, likely implicating different regulatory pathways and signaling complexes.

Voltage-gated sodium channels auxiliary subunits evolutionary emerged nearly 500 million years ago during the Cambrian explosion. These subunits alter one the most important ion channels to electrical signaling, the voltage-gated sodium channels support the propagation of electric impulses in animals. The mechanism for the auxiliary subunits effects on the channels is poorly understand, as is the stoichiometry between the auxiliary subunit and the channel. The focus of my thesis is to generate assays and to use these approaches to understand the interactions different types of voltage-gated channels and their auxiliary subunits. A biochemical approach was taken to identify novel interactions between the eukaryotic sodium channel auxiliary subunits and a prokaryotic voltage-gated sodium channel, a protein that diverged from the eukaryotic voltage-gated sodium channels billions of years ago. These interactions between the auxiliary subunits and channels were probed with chemical and photochemical crosslinkers in search of interaction surfaces and similarity to explain the mechanisms of interaction. The work in this thesis identified novel interactions between the voltage-gated sodium channel auxiliary subunits and voltage-gated channels that are distantly related to the voltage-gated sodium channels principally thought to be modulated by the auxiliary subunits. From this work a rudimentary concept can be theorized that the voltage-gated sodium channel β-subunits and not only β1 have a more primary role in electrophysiology by associating with multiple different types of ion channels.

The work in this thesis addresses the general problem of how ribosomal subunits are exported from the nucleus to mature in the cytoplasm. There are three parts in this dissertation. In the first part, I asked questions about the specificity for export receptors in the nuclear export of the large (60S) ribosomal subunit in yeast. In principle, I tethered different export receptors that are known to work in various unrelated export pathways to the ribosome by fusing them to the trans-acting factor Nmd3. Interestingly, all the chimeric receptors were able to support export, although to different degrees. Moreover, 60S export driven by these chimeric receptors was independent of Crm1, an export receptor that is essential for 60S export in wild-type cells. The second question I addressed in this project was whether or not a nuclear export signal could be provided in cis on ribosomal proteins (Rpls) rather than in trans by a transacting factor. The nuclear export signal (NES) of Nmd3 was fused to different ribosomal proteins and tested for support of 60S export. Several Rpl-NES fusion constructs worked to promote 60S export. Rpl3 gave the best efficiency. In conclusion, these results imply unexpected flexibility in the 60S export pathway. This may help explain how different export receptors could have evolved in different eukaryotic lineages. In the second part of my thesis, I identified the assembly pathway for the base of the ribosome stalk. The stalk is an important functional domain of the large ribosomal subunit because of its requirement for interaction with translation factors. Mrt4 is a nuclear paralog of P0, which is an essential part of the stalk. Here, I identified Yvh1 a novel ribosome biogenesis factor that is required for the release of Mrt4. Yvh1 is a conserved dual phosphatase, but the C-terminal zinc-binding domain rather than the phosphatase function was required for its activity to release Mrt4. Mrt4 localizes in the nucleus and nucleolus in the wild-type cells, but was persistent on cytoplasmic 60S subunits in yvh1[Delta] cells. The persistence of Mrt4 on the 60S subunits blocked the loading of P0 and assembly of the stalk. I also found the binding of Yvh1 depended on Rpl12, a protein that binds together with P0 to form the base of the stalk. Deletion of Rpl12 phenocopied yvh1[Delta]. These data identified the function of Yvh1 as a release factor of Mrt4. I also showed that the function of Yvh1 is conserved in human cells. In my final project, I analyzed the interdependence and order of the known cytoplasmic maturation events of the 60S subunit. 60S subunits require several maturation steps in the cytoplasm before they become competent in translation. There are four major steps involving two ATPases, Drg1 and Ssa1, and two GTPases, Efl1 and Lsg1. In my study, I ordered these steps into one serial pathway. Drg1 releases Rlp24 in the earliest step of 60S maturation in the cytoplasm. Truncation of the C-terminus of Rlp24 blocked cytoplasmic maturation of the large subunit by preventing the recruitment of Drg1 and led to a secondary defect in the release of Arx1 because of a failure to recruit Rei1. Deletion of REI1 mislocalized Tif6 from the nucleus and nucleolus to the cytoplasm and deletion of ARX1 suppressed the Tif6 mislocalization, indicating that the release of Arx1 was required for Tif6 release downstream. I found that mutation of efl1 or sdo1, the known release factors for Tif6, also blocked Nmd3 release. Tif6-V192F, which could bypass the growth defects of efl1 or sdo1 mutants, suppressed the defect of Nmd3 recycling. These results showed that the release of Tif6 was a prerequisite for Nmd3 release. Thus, the release of Nmd3 is downstream of the Tif6 release step. In conclusion, I have ordered the events of cytoplasmic maturation with Drg1 as the first step after ribosome export, followed by Rei1/Jji1 and then Sdo1/Efl1. The release of Nmd3 by Lsg1 appears to be the last step of ribosome maturation in the cytoplasm. Thus, the two ATPases Drg1 and Ssa work first and then the two GTPases Efl1 and Lsg1 work in a linear pathway of 60S maturation in the cytoplasm. / text

Thesis (Ph. D.)--University of Virginia, 2008. / Includes bibliographical references. Also available via the Internet as viewed 10 July 2008.

Diss. (sammanfattning) Umeå : Univ., 2004. / Härtill 3 uppsatser.

Thesis (Ph. D.) -- University of Texas Southwestern Medical Center at Dallas, 2004. / Vita. Bibliography: 154-163.

Thesis (Ph.D.) -- University of Texas Southwestern Medical Center at Dallas, 2005. / Embargoed. Vita. Bibliography: 91-96.

Thesis (Ph. D.)--University of Florida, 2004. / Typescript. Title from title page of source document. Document formatted into pages; contains 258 pages. Includes vita. Includes bibliographical references.

A hemoglobina de Glossoscolex paulistus (HbGp) é caracterizada por uma massa molecular de 3,6 MDa, alta estabilidade oligomérica, resistência a auto-oxidação, e alta afinidade em ligar oxigênio. A estrutura quaternária desta macromolécula apresenta 144 cadeias com grupo heme (globinas) e 36 cadeias sem grupo heme (linkers), dispostos em duas camadas hexagonais. No presente trabalho estudos de caracterização das subunidades da HbGp, bem como da estabilidade da HbGp em diferentes formas, em função do pH, e em diferentes concentrações de ureia, por diferentes técnicas biofísicas, foram realizados. Os estudos de caracterização por eletroforese SDS-PAGE, MALDI-TOF-MS e ultracentrifugação analítica (AUC) das subunidades isoladas mostraram que apenas o monômero d obtido da cromatografia de exclusão por tamanho (SEC) tem alto grau de pureza. Para as demais frações mais de uma contribuição foi observada em solução. Assim, para a fração trimérica, duas espécies estão presentes em solução, a espécie predominante (87 %) é atribuída ao trímero abc e a outra espécie (13 %) pode ser associada ao complexo (abc + L). Os dados espectroscópicos e de AUC mostraram que a estabilidade da HbGp depende fortemente do estado de oxidação do heme, do ligante coordenado no centro metálico e da concentração de proteína. Assim, a forma oxidada, a meta-HbGp, mostrou-se menos estável em meio alcalino e na presença de ureia, seguida pelas formas oxi- e cianometa-HbGp. Desta forma, no pH 8,0, a meta-HbGp está totalmente dissociada em trímero abc e monômero d, enquanto a oxi-HbGp está apenas parcialmente dissociada com uma contribuição de 88 % de proteína íntegra em solução e a cianometa-HbGp não sofre dissociação oligomérica. Os valores de coeficiente de sedimentação s20,w e massa molecular (MM) determinados para as espécies em solução são similares aos observados para as correspondentes espécies isoladas por SEC. Na presença de ureia a mesma tendência foi observada para as três formas da HbGp. Porém, para uma caracterização melhor de processo de desnaturação, os dados espectroscópicos foram analisados usando modelos de dois e três estados para obter informações sobre os parâmetros termodinâmicos do sistema. Assim, bons ajustes foram obtidos usando ambos os modelos, no entanto, o modelo de três estados foi mais adequado para descrever o processo. Por este modelo o processo de desnaturação da HbGp pode ser descrito por duas etapas. A primeira etapa, na faixa de 1,0 - 3,0 mol/L de ureia, está associada a transição do estado nativo para o estado intermediário (N → I), e é caracterizada pela dissociação do oligômero nas diferentes subunidades da HbGp. O estado intermediário apresenta propriedades físico-químicas similares ao estado nativo, sugerindo que o processo de dissociação oligomérica não induz mudanças significativas na estrutura secundária e na região do grupo heme da proteína. Os parâmetros termodinâmicos associados à primeira transição apresentaram erros consideráveis, que podem ser atribuídos à complexidade do estado intermediário com diferentes espécies em solução bem como à semelhança ao estado nativo. A segunda etapa (I → U) com transição bem definida entre 4,5 - 5,0 mol/L de desnaturante é caracterizada pela desnaturação das subunidades dissociadas. Os dados de AUC e SAXS são consistentes com os dados obtidos por espectroscopia, onde a primeira etapa do processo foi caracterizada pela dissociação oligomérica do oligômero em dodecâmero (abcd)3, tetrâmero abcd, trímero abc e monômero d. Para concentrações acima de 4,0 e 5,0 mol/L de ureia, para oxi-HbGp e cianometa-HbGp, respectivamente, aumentos significativos nos valores de I(0), Dmax e Rg sugerem que as subunidades da HbGp estão desnaturadas em solução. As massas moleculares (MM) obtidas por espectrometria de massas e AUC, e os coeficientes de sedimentação s20,w são consistentes com outros resultados reportados para hemoglobinas ortólogas. Além disso, os resultados aqui apresentados representam um avanço importante na caracterização do processo de desnaturação de proteínas oligoméricas complexas. / Glossoscolex paulistus hemoglobin (HbGp) is characterized by a molecular mass of 3.6 MDa, a high oligomeric stability, a high resistance to oxidation and a high affinity to oxygen. The quaternary structure of this macromolecule consists of 144 globin chains, and 36 additional chains lacking the heme group, named linkers, organized in a double-layered hexagonal structure. In this current work the characterization of the HbGp subunits and the effect of pH and urea upon the oligomeric stability were studied by several biophysical techniques. Our results obtained by electrophoresis SDS-PAGE, MALDI-TOF-MS and analytical ultracentrifugation (AUC) showed that only the monomer d isolated by size exclusion chromatography (SEC) presented high purity. For the other fractions various species were observed in the solution. Thus, for the trimeric fraction, two species are present in the equilibrium, the main species with percentage contribution of 87 % is assigned to the trimer abc and the species with 13 % in the solution is associated to the complex (abc + L). Additionally, the data obtained by several spectroscopic techniques and AUC show clearly that the oligomeric stability of HbGp depends on the iron oxidation state, the specific ligand coordinated to the iron and the protein concentration. Therefore, our results show that the met-HbGp form is the less stable one in the alkaline medium and in the presence of urea, followed by the oxy- and cyanomet- forms. In this way, at pH 8.0, the met- form is fully dissociated into smaller subunits, such as, trimer abc and monomer d, while the oxy-HbGp is partially dissociated with a significant percentage contribution (88 %) of undissociated protein, and the cyanomet-HbGp does not undergo oligomeric dissociation. The sedimentation coefficients (s20,w) and molecular masses (MM) values for species present in the solution, at different pH, are very close to the values obtained for isolated species. In the presence of urea the same behavior was observed for the three HbGp forms as compared to the alkaline medium. However, for a full characterization of the unfolding process the thermodynamic parameters were obtained by spectroscopic data analysis using models of two and three states. Adequate fits were obtained for both models, but the three states model was very appropriate to describe the HbGp denaturation process. Thus, the denaturation process of HbGp is defined by two phases. The first phase between 1.0 and 3.0 mol/L, of urea is assigned to the transition of native state to an intermediate state (N → I), and is characterized by dissociation of the oligomer in several subunits. The strong similarity of the intermediate state to the native one suggests that oligomeric dissociation induces little changes in the secondary structure and the region of heme group of the protein. As a consequence, the thermodynamic parameters associated to the first transition have large errors due to the complexity of the intermediate state with different species in the solution, as well as its great similarity to the native state. The second phase (I → U), associated with a cooperative transition at 4,5 - 5,0 mol/L of denaturant agent, is attributed to the unfolding of the dissociated subunits. Our AUC and SAXS data are very consistent with spectroscopic data. Thus, in the first phase the oligomeric dissociation of whole protein in dodecamer (abcd)3, tetramer abcd, trimer abc and monomer d was observed. For urea concentrations above 4.0 - 5,0 mol/L, for oxy-HbGp and cyanomet-HbGp, respectively, the significant increase in I(0), Dmax and Rg values suggests that the HbGp subunits are denatured in the solution. The molecular masses values (MM) obtained by mass spectrometry and AUC, and the sedimentation coefficients (s20,w) are consistent with others results reported in the literature for orthologous hemoglobins. In addition, the results of this work correspond to an important advance in the characterization of the denaturation process of this complex oligomeric protein.

Abstract DNA polymerase ε from HeLa cells has been purified as a heterodimer of a 261 kDa catalytic subunit and a tightly associated smaller polypeptide, the B-subunit. The cDNAs encoding the B-subunits of both human and mouse Pol ε were cloned and shown to encode proteins with a predicted molecular weight of 59 kDa. These subunits are 90 % identical and share 22 % identity with the 80 kDa B-subunit of Saccharomyces cerevisiae Pol ε. The gene for the human Pol ε B-subunit was localized to chromosome 14q21-q22 by fluorescence in situ hybridization. Primary structure analysis of the Pol ε B-subunits demonstrated that they are similar to the B-subunits of Pol α, Pol δ and archaeal DNA polymerases, and comprise a novel protein family of DNA polymerase associated-B-subunits. The family members have 12 conserved motifs distributed in the C-terminal parts, which apparently form crucial structural and functional sites. Secondary structure predictions indicate that the B-subunits share a similar fold, and phylogenetic analysis demonstrated that the B-subunits of Pol α and ε form one subfamily, while the B-subunits of Pol δ and the archaeal proteins form a second subfamily. The corresponding eukaryotic and archaeal catalytic subunits are not related, but all have the characteristics of replicative DNA polymerases. This indicates that the B-subunits of replicative DNA polymerases from archaea to eukaryotes belong to the same protein family and perform similar functions. In S. cerevisiae, Pol ε associates with the checkpoint protein Dpb11. In this study, a human protein, TopBP1, with structural similarity to the budding yeast Dpb11, fission yeast Cut5 and the breast cancer susceptibility gene product Brca1 was identified. The human TOPBP1 gene localizes to chromosome 3q21-q23 and encodes a phosphoprotein of 180 kDa. TopBP1 has eight BRCT domains and is also closely related to the recently identified Drosophila melanogaster Mus101. TopBP1 expression is induced at the G1/S boundary and it performs an important role in DNA replication, as evidenced by inhibition of DNA synthesis by TopBP1 antiserum in isolated nuclei. TopBP1 also associates with Pol ε and localizes, together with Brca1 to distinct foci in S-phase, but not to sites of ongoing DNA replication. Inhibition of DNA replication leads to re-localization of TopBP1 and Brca1 to stalled replication forks. DNA damage induces formation of distinct TopBP1 foci that co-localize with Brca1 in S-phase, but not in G1-phase. The role of TopBP1 in the DNA damage response is also supported by the interaction between TopBP1 and the human checkpoint protein hRad9. These results implicate TopBP1 in replication and checkpoint functions.