Overexpression of oncoprotein MDM2 and mutations of tumor suppressor p53 are frequently observed in human cancers. The NFκB pathway is one of the deregulated pathways in oncogenesis. The overall goal of the project was to study the regulation of NFκB pathway by MDM2 in lung cancer. Our first effort was to determine the frequency of MDM2 overexpression in human lung tumor samples and to identify co-occurring abnormal gene expression by studying the levels of MDM2 and members of NFκB pathway with respect to p53 status. Higher than normal levels of MDM2 were found in approximately 30% of the cancer samples harboring wild-type (WT) and mutant p53. Expression of NFκB2, a mutant p53 inducible gene showed significant statistical correlation with MDM2 in cancer samples that harbored WT p53. A downstream target gene of NFκB2, Bcl2, showed a significant correlation to MDM2 levels, independent of p53 status. Lung cancer samples harboring mutant p53 exhibited elevated levels of NFκB2 though not statistically significant. Our next step was to corroborate findings from lung tumor samples with data from lung cancer cell line harboring WT p53-H460. Consistent with lung tumor samples, ectopic overexpression of MDM2 in H460, showed elevated expression of NFκB2 and Bcl2 with promoter upregulation of NFB2. Silencing of MDM2 proportionally downregulated NFκB2 and Bcl2 in H460 cells. Domain analysis of MDM2 suggested that increase in the NFκB2 promoter activity was not confined to the p53 binding domain of MDM2 suggesting their interaction via p53-dependent and p53-independent mechanisms. A functional cell growth assay showed retarded cell proliferation with downregulation of MDM2. Data from human lung tumor samples and lung cancer cell line suggest that overexpression of MDM2 mediates NFB2 upregulation to confer growth advantage, thus favoring oncogenesis.

Mutations in the breast cancer susceptibility 1 (BRCA1) gene are linked to breast as well as ovarian cancers. However, most cancer-causing mutations within the BRCA1 gene have been found in the N’ and C’ terminal regions of the BRCA1 protein, both believed to be important for DNA double-strand break (DSB) repair. The BRCA1 C’ terminal (BRCT) repeats have been implicated in phospho-serine protein binding whereas the N’ terminal RING domain interacts with the BARD1 protein to form a hetero-dimeric complex with E3 ubiquitin ligase activity. The BRCA1 BRCT domain binds CtIP, BACH1, and RAP80, all of which have been directly implicated in homologous recombination repair (HRR). Lysine 1702 (K1702) of BRCA1 resides within the phospho-serine binding pocket of the first BRCT repeat of BRCA1. To determine the effect of manipulating the ability of BRCA1 to bind CtIP and other phospho-proteins binding to the BRCA1 BRCT domain on DSB repair, and specifically HRR, we introduced a K1702M mutation into BRCA1 known to impair BRCT binding to a pSer-X-X-Phe peptide representing BACH1. Surprisingly, instead of impairing HRR, we found that BRCA1 K1702M resulted in hyper-recombination with > 3-fold higher levels of HRR compared to wild-type BRCA1 using an HRR assay based on GFP expression in BRCA1-defective HCC1937 cells. This hyper-recombinogenic phenotype coincided with cell-cycle arrest in S/G2 suggesting that the potential lack of binding of critical proteins to the BRCA1 BRCT domain results in abnormal HRR by priming cells to undergo more HRR which is enhanced during the S and G2 phases of the cell-cycle. In line with the increased HRR seen with the HRR/GFP assay, HCC1937 cells expressing BRCA1 K1702M showed increased levels of RAD51 foci and nuclear staining suggesting that HRR was highly elevated. Interestingly, the hyper-recombinogenic phenotype of BRCA1 K1702M could be reduced to normal levels with a second mutation (I26A) in BRCA1 that affects BRCA1 and CtIP ubiquitination. These results reveal a hierarchal regulation of HRR with ubiquitination having a dominate role in DSB repair by BRCA1 and suggests that targeted disruption of BRCT-CtIP binding increases HRR that is in turn controlled by ubiquitination. In addition, we provide evidence that BRCA1 serine 1387 phosphorylation within the SQ cluster region of BRCA1 is involved in the cell survival and DNA damage response to IR. The BRCA1 S1387A mutant only partially increased the radiosurvival of HCC1937 cells compared to cells expressing wild-type BRCA1 and immunocytochemical analysis revealed wild-type BRCA1 was located in the nucleus whereas the S1387A mutant was cytoplasmic in response to IR. We also show that BRCA1 SQ cluster serine phosphorylation in addition to serine 1387 is involved in HRR. Altogether, these findings reveal the importance of various regions of BRCA1 in DSB repair and may lead to multiple strategies of modulating BRCA1 function in response to DNA damage.

Ceramides are bioactive lipids that can promote splicing of apoptosis-related genes, including caspase 9 and BCL-x. A recent study demonstrated that expression of neutral sphingomyelinase (NSMase), an enzyme that hydrolyzes sphingomyelins to generate ceramide, is regulated by Group VIA phospholipase A2 (iPLA2β)-dependent mechanism during β-cell apoptosis. This prompted us to hypothesize that iPLA2 is upstream of ceramide generation in the process regulating splicing of apoptotic genes. To test this, Jurkat T cells were treated with the selective inhibitor of iPLA2β, bromoenol lactone (BEL), RNA was isolated and converted to cDNA, and caspase 9 and BCL-x mRNA viii species were amplified using RT-PCR. Inhibition of iPLA2β activity with BEL caused a significant shift in splicing favoring variants encoding the anti-apoptotic forms of caspase 9 (caspase 9b) and BCL-x (BCL-x(L)). This shift was consistent with previously reported effects of ceramide and suggested that iPLA2β regulates splicing of these pre-mRNAs. We next determined whether iPLA2β regulates splicing events during a biological response. Caspase-9 and BCL-x splice variants were compared in human and mouse islets, mouse islet cell lines, and in rat insulinoma (INS1) cells. INS-1 insulinoma cells were treated with thapsigargin to induce ER stress, which can eventually lead to apoptosis. Thapsigarin-treated INS-1 cells exhibited an increase in the ratio of BCL-x(s) (pro-apoptotic) to BCL-x(L) (anti-apoptotic) but BEL prevented this shift in splicing. Splicing data obtained from genetically modified rodent mice (iPLA2β knockouts and transgenics) also demonstrated the involvement of iPLA2β in alternative splicing. Together, these observations indicate that iPLA2β plays an important role in the regulation of pre-mRNA splicing of key apoptotic factors. Our findings therefore suggest a novel role for iPLA2β in determining whether cells survive or undergo apoptosis.

The pro-apoptotic, caspase 9a, and the anti-apoptotic, caspase 9b, are derived from the caspase 9 gene by alternative splicing. This study demonstrates that the alternative splicing of caspase 9 is dysregulated in a large percentage of non-small cell lung cancer (NSCLC) tumors of the adenocarcinoma type. Furthermore, modulation of the levels of splice variants of caspase 9 had dramatic effects on the anchorage-independent growth and tumorigenic capacity of NSCLC cells. Due to these findings, the molecular mechanisms regulating the post-transcriptional processing of caspase 9 were therefore examined and an exonic splicing silencer (ESS) regulating the pre-mRNA processing of caspase 9 was identified. To study the possible RNA trans-factors interacting with this RNA sequence, we utilized an electromobility shift assay (EMSA) coupled with competitor studies and demonstrated three specific protein:RNA complexes for this ESS. Affinity purification and mass spectrometry analysis identified hnRNP L as part of these protein:RNA complexes. Downregulation of hnRNP L induced a significant increase in caspase 9a/caspase 9b mRNA ratio, which translated to the protein level. Expression of hnRNP L verified the siRNA specificity lowering the caspase 9a/9b ratio, but expression of hnRNP L produced the contrasting effect in non-transformed cells suggesting a post-translational modification specific for NSCLC cells. Indeed, the phospho-status of hnRNP L was significantly increased in NSCLC cells, and mutagenesis studies identified Ser52 as a critical residue regulating the ability of hnRNP L to repress the inclusion of the exon 3,4,5,6 cassette into the mature caspase 9 mRNA. The biological relevance of this mechanism was demonstrated by stable downregulation of hnRNP L in NSCLC cells, which induced a complete loss of both anchorage-independent growth and tumorigenic capacity. This effect of hnRNP L downregulation was due to distal modulation of the alternative splicing of caspase 9 as the loss of both phenotypes was “rescued” by ectopic expression of caspase 9b. Therefore, this study identifies cancer-specific mechanism of hnRNP L phosphorylation and subsequent lowering of the caspase 9a/9b ratio, which is required for the tumorigenic capacity of NSCLC cells.

The KsgA enzymes comprise an ancient family of methyltransferases that are intimately involved in ribosome biogenesis. Ribosome biogenesis is a complicated process, involving numerous cleavage, base modification and assembly steps. All ribosomes share the same general architecture, with small and large subunits made up of roughly similar rRNA species and a variety of ribosomal proteins. However, the fundamental assembly process differs significantly between eukaryotes and eubacteria, not only in distribution and mechanism of modifications but also in organization of assembly steps. Despite these differences, members of the KsgA/Dim1 methyltransferase family and their resultant modification of small-subunit rRNA are found throughout evolution, and therefore were present in the last common ancestor. The first member of the family to be described, KsgA from Escherichia coli, was initially shown to be the determining factor for resistance/sensitivity to the antibiotic kasugamycin and was subsequently found to dimethylate two adenosines in 16S rRNA during maturation of the 30S subunit. Since then, numerous other members of the family have been characterized in eubacteria, eukaryotes, archaea and in eukaryotic organelles. The eukaryotic ortholog, Dim1, is essential for proper processing of the pre-rRNA, in addition to and separate from its methyltransferase function. The KsgA/Dim1 family bears sequence and structural similarity to a larger group of S-adenosyl-L-methionine dependent methyltransferases, which includes both DNA and RNA methyltransferases. In this document we report that KsgA orthologs from archaea and eukaryotes are able to complement for KsgA function in bacteria, both in vivo and in vitro. This indicates that all of these enzymes can recognize a common ribosomal substrate, and that the recognition elements must be largely unchanged since the evolutionary split between the three domains of life. We have characterized KsgA structurally, and discuss aspects of KsgA's activity in light of the structural data. We also propose a model for KsgA binding to the 30S subunit, based on solution probing data. This model sheds light on KsgA's unusual regulation and on the dual function of the Dim1 enzymes.

Calcineurin is a Ser/Thr phosphatase whose function is implicated in critical physiological processes such as immune system activation, fetal heart development, and long-term depression in neurons. Calcineurin has been implicated in the progression of Alzheimer’s disease and cardiac hypertrophy. It is not well understood how calcineurin is activated on a molecular level by Ca2+ and its activating protein calmodulin. Previous data from our lab show that calmodulin interaction induces the folding of the intrinsically disordered regulatory domain of calcineurin in two discrete and distant regions into α-helical conformations and that this folding is critical for complete activation of calcineurin. It was also discovered that one of the helical elements which we call the “distal helix” was unstable at a human body temperature of 37°C in dilute buffer. This raises the question; how can a structure critical for the complete activation of calcineurin be unstable at average human body temperature? Proteins do not exist in solutions of the dilute buffer, but rather in a crowded cosmos that ranges between 200 and 400 g/L of macromolecules such as proteins, DNA, and other cellular components. We show here that phenomenon known as macromolecular crowding can stabilize the distal helix and that stabilization increases the activity of calcineurin at human body temperature. Much about intrinsically disordered proteins (IDPs) remains a mystery, especially what influences the rate at which they interact with their target molecules. IDPs lack any sort of stable three-dimensional structure because of their lack of sufficient hydrophobic or aromatic amino acids while having a large proportion of polar and charged amino acids. Because of the high degree of charged amino acids, electrostatic forces play a significant role in their interaction other proteins. This is known to be the case for calmodulin which is net negatively charged protein that has over 300 binding targets of which are usually basic amphipathic alpha-helices. The calmodulin-binding site located in the intrinsically disordered regulatory domain of calcineurin is net positively charged, and, interestingly, is flanked by acidic patches on either side. These acidic patches might perturb attractive electrostatic forces between the calmodulin-binding site and calmodulin. Using fluorescence spectroscopy in conjunction with a stopped-flow apparatus to measure the kinetics between calmodulin and calcineurin we seek to characterize the influence of the steric and electrostatic forces between the two proteins. Also, we present data on RCAN1-4 (Regulator of Calcineurin Isoform 1-4) which has been shown to be an inhibitor in some contexts and an activator of calcineurin in other. RCAN1-4 is expressed in the heart and its upregulation has been shown to prevent calcineurin-mediated pathological cardiac hypertrophy suggesting that it plays an inhibitory role in this context. The work shown demonstrates that RCAN1-4 is a competitive inhibitor of calcineurin and whose binding affinity is modulated by Ca2+/calmodulin. These data unveil a binding site utilized by RCAN1-4 which is commonly used among other calcineurin substrates.

<p>Three aspects of the electron transport chain have been investigated in the present studies: (<b>i</b>) structure-function studies of the central proton-electron conducting cytochrome <i>b</i>₆<i>f</i>complex, focusing on the effect of lipids in structural stabilization and electron transfer function; (<b>ii</b>) transmembrane electron transfer pathways in the cytochrome <i>b</i>₆<i>f</i>and mitochondrial cytochrome <i>bc</i>₁complexes, determined by heterogeneity in the internal polarity of the membrane protein complexes; and (<b>iii</b>) purification and characterization of a novel ~1 MDa supercomplex, dominated by the presence of photosystem I (PSI), ATP-synthase and ferredoxin-NADP⁺reductase (FNR) from higher plant system <i>Spinacea</i>.<b></b></p>

Thesis advisor: Mary F. Roberts / Myo-inositol and its phosphorylated derivatives are found across all domains of life, and these molecules play crucial roles in a wide variety of cellular processes. While the biosynthesis of inositol is an evolutionarily conserved pathway, there are a wide variety of enzymes that use inositol and its derivatives as substrates. This thesis explores two such enzymes; a phosphatidylinositol- specific phospholipase C (PI-PLC) produced by <i>Staphylococcus aureus</i>, and AF2372, a dual action inositol monophosphatase/ fructose bisphosphatase produced by the <i>Archaeoglobus fulgidus</i>. At the outset of this work, the structure of the <i>S. aureus</i> PI-PLC was unknown, but some interesting biochemical properties about the enzyme had been observed. The structure of AF2372 had been reported, but a structure had not yet been solved in the presence of osmolytes known to thermoprotect the enzyme. Both the <i>S. aureus</i> PI-PLC and AF2372 catalyze the cleavage of phosphorylated inositol compounds, but share no mechanistic, structural, or taxonomical similarities. Protein crystallography is a powerful tool, and with it I have been able to study these two enzymes at a molecular level, providing insight into complex biological questions about each enzyme. / Thesis (PhD) — Boston College, 2012. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Chemistry.

A família das septinas caracteriza-se pela capacidade de ligar nucleotídeos de guanina e de se associarem formando filamentos. Diversas funções biológicas têm sido reportadas para esses filamentos e sua dissociação pode estar relacionada a patologias. A septina 12 humana expressa especificamente em testículos, foi identificada em filamentos que compõem o annulus do espermatozoide, cuja integridade está relacionada com a morfologia deste. Embora muitos estudos tenham sido reportados, vários aspectos das bases moleculares e fisiológicas de sua função e automontagem permanecem desconhecidos. Neste trabalho, procurou-se obter informações estruturais para o domínio de ligação ao nucleotídeo da SEPT12 (SEPT12G), do mutante SEPT12GT89M e do heterodímero SEPT7-SEPT12. A expressão destas proteínas foi realizada em células de E. coli da linhagem Rosetta(DE3) pelos vetores de expressão pET28a(+) e pETDuet-1. As etapas de purificação foram cromatografia de afinidade e exclusão molecular. A proteína SEPT12G foi submetida à avaliação do estado oligomérico, fluorescência intrínseca, ensaios de conteúdo de nucleotídeo, atividade GTPásica e transição térmica. O heterodímero SEPT7-SEPT12 foi submetido à avaliação do estado oligomérico e conteúdo de nucleotídeo. Ensaios de cristalização foram realizados para todas as proteínas. A coleta de dados realizada na linha I24 do Diamond Light Source (Didcot, Inglaterra) resultou em conjuntos de dados de alta resolução para a SET12G, SEPT12GT89M e baixa resolução para a SEPT7NGc. Os estudos biofísicos mostraram que a SEPT12G foi obtida em sua forma nativa ou, seja, capaz de ligar e hidrolisar o nucleotídeo GTP e que o heterodímero obtido apresentou ambas as proteínas. As estruturas cristalográficas foram resolvidas e permitiram realizar observações importantes para o grupo I das septinas humanas (SEPT3, SEPT9 e SEPT12). Para a SEPT12 pôde-se observar como a mudança que ocorre no motivo G4 pode alterar a estabilidade da interface G. No contexto do grupo I esta estrutura permitiu concluir que todas as proteínas deste subgrupo podem formar duas interfaces NC, dos tipos aberta e fechada. Além disso, reforçou a observação da orientação diferencial da hélice &alpha;5\', cuja função ainda não está esclarecida, mas sem dúvidas é um diferencial que caracteriza este grupo, possivelmente relacionado com a ancoragem da região polibásica na conformação aberta. A estrutura cristalográfica do mutante SEPT12T89M permitiu observar que a mutação levou a uma alteração na primeira esfera de coordenação do íon Mg2+, mudança que interrompe o mecanismo do switch universal e deixa a proteína catalíticamente inativa. Por fim, o estudo cristalográfica do complexo entre a SEPT12 e SEPT7 não foi possível, uma vez que todas as tentativas resultaram em cristais contendo apenas a SEPT7, o que pode ser consequência da precipitação da SEPT12 ou da condição de cristalização utilizada, que desestabiliza o heterodímero. / The septin family of proteins is characterized by their ability to bind guanine nucleotides and associate into filaments. Several biological functions have been reported for these filaments and their dissociation may be related to pathologies. Human septin is 12 specifically expressed in testes and has been identified in filaments that form the sperm annulus, whose integrity is related to its morphology. Although many studies have been reported, the molecular and physiological bases of septin filament function and self-assembly have yet to be completely elucidated. This study aims to obtain structural information for the nucleotide binding domain of SEPT12 (SEPT12G), the SEPT12GT89M mutant and the SEPT7-SEPT12 heterodimer. Expression of these proteins was performed in E. coli Rosetta(DE3) strain using the pET28a (+) and pETDuet-1 expression vectors. Purification was performed by affinity and size exclusion chromatography. The SEPT12G protein was submitted to an evaluation of its oligomeric state, intrinsic fluorescence, nucleotide content, GTPase activity and thermal transition. The oligomeric state and nucleotide content of SEPT7-SEPT12 was also evaluated. Crystallization assays were performed for all proteins. Data collection on line I24 of the Diamond Light Source (Didcot, England) resulted in high-resolution data sets for SET12G and SEPT12GT89M but only low resolution data for the SEPT7NGc. Biophysical studies showed that SEPT12G was obtained in its native form or, in other words, capable of binding and hydrolyzing GTP and that the purified heterodimer presented both proteins. The crystallographic structures were solved by molecular replacement allowing the identification of features characteristics of the group I septins (SEPT3, SEPT9 and SEPT12). The structure also confirmed that all the proteins of this group are able to form two different NC interfaces: open and closed. In addition, it reinforced the observation that the &alpha;5\' helix assumes a different orientation, whose function has not yet been clarified, but without doubt is a characteristic of this group which may be related to anchoring the polybasic regions whilst in the open conformation. The SEPT12T89M mutant crystal structure shows that the first shell coordination of the Mg2+ ion is altered, leading to an interruption of the universal switch mechanism and a consequent lack of catalytic activity. Finally, structural studies of the interaction between SEPT12 and SEPT7 were not possible, since all attempts resulted in crystals containing only SEPT7. This may be a consequence of SEPT12 precipitation or the crystallization condition used, destabilizing the heterodimeric interface.

Chez les organismes multicellulaires la diversité protéique dans chaque cellule et chaque tissu est obtenue initialement en régulant l’expression d’une partie des gènes d’un génome. Ces gènes sélectionnés peuvent ensuite être soumis à un épissage alternatif de sorte que certains exons sont retenus ou exclus dans l’ARNm final. Nous étudions les détails moléculaires de la protéine MEC-8, un facteur d’épissage tissu spécifique chez Caenorhabditis elegans. Les mutants MEC-8 sont responsables d’un phénotype insensible au touché chez Caenorhabditis elegans. Plus précisément, MEC-8 lie le pré-ARNm de mec-2 un composant des récepteurs mécanosensoriels afin de réguler la production d’un isoforme particulier nécessaire pour la transduction du signal mécanosensoriel. Des études portant sur le motif conservé de reconnaissance à l’ARN (RRM) chez des orthologues des vertébrés (RBPMS) et des insectes (couch potato, CPO) ont démontré la présence d’un motif d’homodimérisation dans le domaine RRM1 de MEC-8. Cependant MEC-8 contient aussi un second domaine RRM dans sa partie C-terminale, domaine qui n’est pas retrouvé dans les protéines RBPMS et CPO. Nous avons donc exprimé chaque domaine RRM de MEC-8 indépendamment ainsi que la protéine entière et ces constructions ont été utilisées pour diverses expériences biophysiques. Nous avons ainsi identifié la séquence de liaison optimale pour les deux domaines RRM1 et RRM2. Ces analyses ont aussi été menées sur les domaines homologues issus des protéines RBPMS et CPO qui présentent une forte affinité pour la même séquence d’ARN. Nous avons donc découvert que malgré des différences de fonction et de localisation les membres de la famille RBPMS lient tous le même motif d’ARN. Les détails atomiques des deux RRM en complexe avec leur motif de liaison ont été obtenus en utilisant de la spectroscopie RMN et de la cristallographique des rayons X. Les deux complexes RRM-ligand de MEC-8 présentent de surprenantes similarités dans leur architecture. / In multicellular organisms, proteomic diversity in each cell and tissue is provided initially by selective expression of gene subsets from the total genome, which are further subjected to alternative splicing, such that a different pattern of exons can be retained or excluded in the final protein coding mRNA. We are investigating the molecular details of the tissue-specific splicing factor protein MEC-8 from the worm Caenorhabditis elegans. The MEC-8 mutant protein is responsible for a touch insensitive phenotype in Caenorhabditis elegans, relating to its role as an alternative splicing factor. More precisely, MEC-8 can bind to the mec-2 pre-mRNA, a component of mechanosensory receptor, to regulate the production of a certain isoform required for transducing the touch signal. Previous studies of the conserved RNA Recognition Motif (RRM) domain in orthologues from vertebrate (RBPMS) and insect (couch potato; CPO) have demonstrated a homodimerization motif in MEC-8 RRM1. However, MEC-8 also contains a second RRM domain in the C-terminus that is not found in the characterized RBPMS and CPO proteins. We have therefore expressed the independent RNA-binding domains of MEC-8 as well as the full-length protein and have used these constructs in a variety of biophysical assays. We identified the optimal RNA binding sequence for both the RRM1 and RRM2, and quantified the penalty of sequence variations. The investigation has also been extended to the homologous domains from human RBPMS and Drosophila CPO, which show a high affinity to the same RNA sequence. We therefore find that despite differences in function and localization, the members of the RBPMS protein family all bind to the same RNA motif. Atomic details of binding have also been obtained by using a combination of NMR spectroscopy and X-ray crystallography. The ligand-bound complexes reveal a surprising similarity in the architecture of the bound ligand for the first and second RRM domains from MEC-8.