Thèses sur le sujet « Protein oligomer »

Proteins are composed of a unique sequence of amino acids, whose order guides a protein to adopt its particular fold and perform a specific function. It has been shown that a protein's 3-dimensional structure is embedded within its primary sequence. The problem that remains elusive to biochemists is how a protein's primary sequence directs the folding to adopt such a specific conformation. In an attempt to gain a better understanding of protein folding, my research tests a novel model of protein packing using protein design. The model defines the knob-socket construct as the fundamental unit of packing within protein structure. The knob-socket model characterizes packing specificity in terms of amino acid preferences for sockets in different environments: sockets filled with a knob are involved in inter-helical interactions and free sockets are involved in intra-helical interactions. Equipped with this knowledge, I sought to design a unique protein, Ksα1.1, completely de novo. The sequence was selected to induce helix formation with a predefined tertiary packing interface. Circular dichroism showed that Ksα1.1 formed α-helical secondary structure as intended. The nuclear magnetic resonance studies demonstrated formation of a high order oligomer with increased protein concentration. These results and analysis prove that the knob-socket model is a predictive model for all α-helical protein packing. More importantly, the knob-socket model introduces a new protein design method that can potentially hold a solution to the folding problem.

Proteins are composed of a unique sequence of amino acids, whose order guides a protein to adopt its particular fold and perform a specific function. It has been shown that a protein's 3-dimensional structure is embedded within its primary sequence. The problem that remains elusive to biochemists is how a protein's primary sequence directs the folding to adopt such a specific conformation. In an attempt to gain a better understanding of protein folding, my research tests a novel model of protein packing using protein design. The model defines the knob-socket construct as the fundamental unit of packing within protein structure. The knob-socket model characterizes packing specificity in terms of amino acid preferences for sockets in different environments: sockets filled with a knob are involved in inter-helical interactions and free sockets are involved in intra-helical interactions. Equipped with this knowledge, I sought to design a unique protein, Ksα1.1, completely de novo. The sequence was selected to induce helix formation with a predefined tertiary packing interface. Circular dichroism showed that Ksα1.1 formed α-helical secondary structure as intended. The nuclear magnetic resonance studies demonstrated formation of a high order oligomer with increased protein concentration. These results and analysis prove that the knob-socket model is a predictive model for all α-helical protein packing. More importantly, the knob-socket model introduces a new protein design method that can potentially hold a solution to the folding problem.

Beta2-microglobulin (β2m) is a 99-residue globular protein that represents the light chain of the major histocompatibility complex class I (MHCI). β2m is responsible for two types of human amyloid diseases: dialysis-related amyloidosis (DRA) caused by wt β2m, and hereditary systemic amyloidosis due to the D76N β2m mutant. Two independent projects were carried out during my PhD studies addressing both types of β2m-related amyloidosis: (i) The human cells adopt a sophisticated unfolded protein response (UPR) system for targeting misfolded/aggregated polypeptides. However, the D76N variant, which is unstable and aggregation prone, bypasses the UPR system and reaches the extracellular space forming amyloids. To understand the mechanism(s) that allow the D76N variant to escape the UPR system and to characterize its effect on MHCI, we performed a complete structural and biophysical study on a MHCI bearing the D76N variant. Our results show that MHCI acts as a chaperone that stabilize and hide the amyloidogenic variant from the UPR system, and transfers it to cell membrane, where during its normal turnover, the D76N β2m variant dissociates into blood serum and aggregates causing human pathologies. (ii) The early oligomeric species are responsible for cellular cytotoxicity in amyloidosis. Previous data showed a favourable association interface (between two facing D β-strands) that may involve in β2m oligomerization. To further elucidate its role in aggregation, we created a S-S linked dimer of β2m (DimC33) that interacts via the DD interface. Our data show that DimC33 is highly amyloidogenic compared to wt β2m, suggesting that dimerization through DD interface is instrumental for enhancing its aggregation propensity. Furthermore, DimC33 was co-crystallized in complex with Thioflavin-T (ThT), a well-known amyloid-specific dye, showing unique ThT binding site that may indicate a second key interface involved in β2m oligomerization.

Thesis (Ph. D.)--West Virginia University, 2010.
Title from document title page. Document formatted into pages; contains xiv, 180 p. : ill. (some col.). Includes abstract. Includes bibliographical references.

Voltage gated potassium channels (Kv) play important roles in different biological process such as generation and propagation of the nerve pulse and the cardiac action potential, promotion of insulin secretion, cell volume control, induction of cell proliferation, apoptosis, migration and initiation of many signaling pathways. Kv channels can homo- or hetero- tetramerize. The composition of the channel modulates their surface expression and serves as a mechanism for regulating channel activity. Kv channel interaction with accessory subunits provides mechanisms for channels to respond to stimuli beyond changes in membrane potential. The present dissertation is focused in the analysis of the effect of one regulatory subunits family (KCNEs) on different Kv channels. The first channel analyzed is Kv7.1, which is one of the most well-known channels to interact with all the KCNE family members. In fact KCNE1-Kv7.1 complex is focus of a huge number of studies, due to its important role in heart. Most of the studies though are focused to their electrophysiological properties and molecular determinants involved in the interaction. We performed traffic analysis experiments of Kv7.1 in the present of KCNE1-5 and demonstrated that Kv7.1 membrane surface localization is modified by some of them. Next, analysis was expanded to another channel from the same family, less characterized: Kv7.5. We demonstrated that from the five KCNEs members, only KCNE1 and KCNE3 modulate Kv7.5 activity. Furthermore, we demonstrated that Kv7.5 association to KCNE3 modifies the targeting of the regulatory subunit. Next, we moved to a non-related Kv channel such as Kv1.3, which plays a crucial role in the immune system. We first focused into characterizing the modulation of Kv1.3. We demonstrated that KCNE4, but not KCNE2, functions as an inhibitory Kv1.3 partner. Kv1.3 trafficking, targeting and activity are altered by the presence of KCNE4. Furthermore, by the combination of a plethora of approaches such as electrophysiological experiments from chimeric proteins and GFP single bleaching counting steps methodology we deciphered the stoichiometry of the Kv1.3-KCNE4 complex. Next, by immunoprecipitation experiments, traffic analysis and electrophysiological experiments, we analyzed the molecular determinants involved in the association between Kv1.3 and KCNE4. We have map a domain of Kv1.3 and a specific motif of KCNE4 involved in the formation of Kv1.3-KCNE4 complex, but not in the modulation of the channel. We also proposed a 3D docking model of Kv1.3 and KCNE4. Finally, due to the importance of Kv1.3 in the immune system, the expression of all the KCNE family has been analyzed in several cell lines of leukocytes. We have demonstrated that all KCNEs suffer a differential regulation among proliferation of leukocytes. Furthermore, a different regulation can be observed, depend on the mode of leukocytes’ activation. Our results further suggest a new and yet unidentified physiological role for KCNE subunits in the immune system. Putative associations of these ancillary proteins with Kv channels would yield a wide variety of biophysically and pharmacologically distinct channels that fine-tune the immunological response.
Els canals de potassi dependents de voltatge (Kv) juguen un paper molt important tant en cèl•lules excitables com no excitables. La possibilitat de formar hetero-oligomers i la d’associació amb subunitats reguladores són uns dels mecanismes que existeixen per tal de proveir de diferents mecanismes per a respondre de manera diferent enfront a canvis en el potencial de membrana. La composició del canalosoma modula tant la seva expressió a superfície com l’activitat d’aquests. Aquesta tesi es centra en l’estudi de l’efecte del la família de les subunitats reguladores KCNE sobre diferents Kv. Primerament s’estudià l’efecte que causaven en el tràfic del canal Kv7.1 (canal model per a l’estudi dels KCNEs) i posteriorment s’amplià l’estudi a un altre membre de la mateixa família, Kv7.5. A continuació s’estudià un canal d’elevada importància per a l’activació i proliferació leucocitària: Kv1.3, centrant-nos sobretot en l’efecte causat per un dels KCNEs: KCNE4. Aquesta subunitat no només inhibeix dràsticament el corrent del canal Kv1.3, sinó que a més a més, modifica el seu tràfic i localització. Aquests canvis són deguts a una interacció directa entre ambdues proteïnes. A continuació s’estudià en detall el complex Kv1.3-KCNE4. Mitjançant la combinació d’experiments d’electrofisiologia i monitorització de fluorescència de molècules individuals en la membrana, es va poder establir l’estequiometria d’aquest complex. Posteriorment, mitjançant l’anàlisi de diverses proteïnes quimèriques i mutants, tant del canal com de la subunitat reguladora, es van cercar els determinants moleculars implicats en l’associació entre ambdues proteïnes. S’han pogut determinar els motius claus en KCNE4 i Kv1.3 implicats en la formació del complex, però no en la modulació del canal. Finalment, degut a la importància de Kv1.3 en el sistema immunitari, s’han analitzat els nivells d’expressió dels KCNEs en diferents línies leucocitàries. S’ha observat que aquestes subunitats pateixen una regulació diferencial en funció de la manera d’activació i al llarg de la proliferació del leucòcits, suggerint un possible paper en la regulació precisa de la resposta immunològica.

Alzheimer’s disease (AD) is the most common cause of senile dementia worldwide. AD is a neurodegenerative disorder characterized by the loss of memory and language skill, collapse of the cognitive function, and distortion of social behavior. As of today, the onset mechanisms of AD and cure are unknown; however, three hallmarks are commonly encountered: extra and intracellular accumulation of amyloid beta (A!) peptide plaques, formation of intracellular neurofibrillary tangles, and inevitable neuronal death. Hypothetically, a possible scenario provoking or involved in the onset of AD is a cascade effect that starts with an imbalance in the production and clearance of Aß peptide that consequently leads to its accumulation, formation of tau protein tangles and neuronal death. This work studied and characterized the mechanisms governing A! peptide aggregation and the effects of using anti-Aß monoclonal antibodies to modify this process. These mechanisms play an important role in the formation of AD plaques and are critical in the search for therapies involving Aß peptide plaque clearance. Yet, antibody-based therapies for plaque clearance are not well understood, adding to the existing concerns about side effects in humans, hence there is a necessity of knowledge in this matter. In this work different Nterminus, C-terminus, and Mid-domain antibodies were used against Aß peptide species (monomers, oligomers, and fibrils) to probe peptide aggregates modification and disruption. Additionally, construction of a soft supported lipid bilayer membrane was proposed to study the adhesion mechanisms of Aß peptide and interactions with antibodies, mimicking the neuronal cell surface. The main characterization techniques used in this work were: atomic force microscopy (AFM) and transmission electron microscopy that allowed the physical exploration and visualization of the different processes of aggregation in terms of adhesion, size evolution, and distribution of the peptide; and attenuated total reflectance Fourier spectroscopy (ATR/FTIR) which allowed monitoring the change of secondary structures for the peptide during the processes studied. It is endeavored that this work will help to elucidate the effects attributed to the molecular interactions between A! peptide species and antibodies to target Aß plaque’s clearance in the brain of AD patients. Ultimately, this study provides novel information critical for the formulation of effective therapies to prevent and treat AD with less collateral effects. It also represents a contribution to the basic scientific knowledge regarding peptide-antibody interactions with application to other diseases related to protein misfolding.

This dissertation focuses on rationalised DNA design as a tool for the discovery and development of new therapeutic entities, as well as understanding the biological function of DNA beyond the storage of genetic information. The study is comprised of two main areas of study: (i) the use of DNA as a coding unit to illustrate the relationship between code-diversity and dynamics of self-assembly; and (ii) the use of DNA as an active unit that interacts and regulates a target protein. In the study of DNA as a coding unit in code-diversity and dynamics of self-assembly, we developed the DNA-Based Diversity Modelling and Analysis (DDMA) method. Using Polymerase Chain Reaction (PCR) and Real Time Polymerase Chain Reaction (RT-PCR), we studied the diversity and evolution of synthetic oligonucleotide populations. The manipulation of critical conditions, with monitoring and interpretation of their effects, lead to understanding how PCR amplification unfolding could reshape a population. This new take on an old technology has great value for the study of: (a) code-diversity, convenient in a DNA-based selection method, so semi-quantitation can evaluate a selection development and the population\'s behaviour can indicate the quality; (b) self-assembly dynamics, for the simulation of a real evolution, emulating a society where selective pressures direct the population's adaptation; and (c) development of high-entropy DNA structures, in order to understand how similar unspecific DNA structures are formed in certain pathologies, such as in auto-immune diseases. To explore DNA as an active unit in Tumour Necrosis Factor α (TNF-α) interaction and activity modulation, we investigate DNA's influence on its spatial conformation by physical environment regulation. Active TNF-α is a trimer and the protein-protein interactions between its monomers are a promising target for drug development. It has been hypothesised that TNF-α forms a very intricate network after its activation between its subunits and receptors, but the mechanism is still not completely clear. During our research, we estimate the non-specific DNA binding to TNF-α in the low micro-molar range. Cell toxicity assays confirm this interaction, where DNA consistently enhances TNF-α's cytotoxic effect. Further binding and structural studies lead to the same conclusion that DNA binds and interferes with TNF-α structure. From this protein-DNA interaction study, a new set of tools to regulate TNF-α's biological activity can be developed and its own biology can be unveiled.

Orientador: Goran Neshich
Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Biologia
Made available in DSpace on 2018-08-20T22:35:51Z (GMT). No. of bitstreams: 1 Moraes_FabioRogeriode_D.pdf: 15399723 bytes, checksum: 4f1315f86b2c74d078c5105b299a9750 (MD5) Previous issue date: 2012
Resumo: Dentro do ambiente celular, há uma variedade de moléculas e a interação entre si regulam praticamente todos os processos necessários e essenciais para a manutenção da vida. Interações entre proteínas estão envolvidas no controle de vários processos intra e intercelulares, como regulação metabólica e da expressão gênica, reconhecimento antígeno-anticorpo etc. que definem as características biológicas do funcionamento da vida entre os diversos organismos. Ao conhecer a interface de interação de uma proteína chave para desenvolvimento de casos patológicos, é possível desenhar drogas com alta especificidade com o sítio de ligação. Para avançar nessa frente, o conhecimento da estrutura proteica é fundamental, porém não suficiente. É necessário conhecermos o sítio de ligação alvo para cada parceiro de interação. Este estudo visa entender as características do nano-ambiente das interfaces proteicas - área através da qual as macromoléculas se comunicam e exercem sua funcionalidade. Propomos utilizar uma abordagem de estudo das características físico-químicas e estruturais dos resíduos formadores de interfaces de complexos conhecidos e com estrutura quaternária resolvida experimentalmente, utilizando um conjunto de dados sem redundância sequencial, extraindo os parâmetros/descritores que descrevem de forma objetiva as diferentes classes de complexos, revelando as características principais sobre interações proteína-proteína. A finalidade deste trabalho é de conhecer os detalhes que definem uma área como interface e aplicá-lo em uma ferramenta preditiva para todas as proteínas com arranjo estrutural conhecido e/ou modelado. Propomos de forma pioneira, o uso de classificadores específicos para cada tipo de aminoácido e independente do uso de descritores sobre conservação de aminoácidos. Resultados obtidos com classificador linear e por ensemble de redes neurais destacam a nossa abordagem, desenhada e aplicada nesta tese, como uma com os melhores indicadores de desempenho na predição precisa dos resíduos de aminoácido na interface entre as abordagens descritas recentemente na literatura. Ainda, enquanto os outros métodos dependem de descritores sobre conservação de aminoácidos, é mostrado aqui que nenhum ganho de desempenho é obtido com a incorporação de tais descritores em nosso modelo classificador. Esse resultado indica que o uso de descritores puramente físico-químicos e estruturais é suficiente para explicar o grau de conservação dos aminoácidos
Abstract: Inside cells, there is a variety of molecules and their interactions regulate virtually all necessary and essential processes to the maintenance of life. Interactions among proteins are involved in the control of several processes within and out of the cell, such as, metabolic and gene expression regulation, anti-body and antigen recognition, etc. that defines biological characteristics of life among many organisms. If the protein interface amino acids of a key protein related to a given pathologic phenomenon are known, it is possible to rationally design drugs with high specificity for a specific binding site. To gain insight in this field, the knowledge of the protein three-dimensional structure is mandatory, but not sufficient. It is also necessary to know the interface between the target protein and its partners. This study focuses in understanding the characteristics of the area through which the macromolecules communicate to each other and exercise their function. Here, it is proposed an approach to study the physicochemical and structural characteristics of the interface forming residues with known quaternary structure (experimentally solved). It was selected a sequence non-redundant dataset and by extracting parameters/descriptors, that objectively describe different complex classes, it was possible to unravel the basic characteristics of protein-protein binding. The goal of this study is to unravel the details that outline a specific area as interface and apply it in a form of a predictive tool for all proteins with known atomic structure. It is proposed by the first time, the use of amino acid specific classifiers regarding amino acid type and free of amino acid conservation attributes. The results obtained here by employing linear and ensemble of neural network classifiers show that, based on purely physicochemical and structural descriptors, it is possible to get precise predictions about interface forming residues in protein-protein assemblies. Comparatively, the method described here retains better performance indicators than the ones recently described in the literature. In addition, we showed that, for our method, adding "conservation" attributes does not induce any performance gain, which is a major difference if compared to other described methods. This result indicates the purely physicochemical and structural descriptors are sufficient to explain how conserved amino acids are
Doutorado
Bioinformatica
Doutor em Genetica e Biologia Molecular

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.

Alzheimer's disease (AD), a progressive, neurodegenerative brain disorder, is an impending socio-economic crisis due to the aging global population. Soluble oligomers of the amyloid-β (Aβ) peptide cause neurotoxicity, synaptic dysfunction and memory impairments which underlie AD, but their mechanisms of action are poorly understood. Recently, the cellular prion protein (Prpc) was identified as a high affinity receptor which mediates the neuronal binding and memory impairments of AI3 oligomers. In this study, the cellular biology of the interaction between AI3 oligomers and Prpc was investigated. We report that fibrillar Aβ oligomers recognised by the QC antibody, which have been shown to correlate with the onset and severity of AD, bind preferentially to neuronal cells expressing Prpc. The green tea polyphenol (-)-epigallocatechin gallate (EGCG) and the red wine extract resveratrol both re-modelled the fibrillar conformation of AI3 oligomers. The resulting non-fibrillar oligomers displayed significantly reduced binding to Prpc-expressing cells and were no longer cytotoxic. Fluorescence microscopy and co-localisation with sub-cellular markers revealed that the AI3 oligomers co-internalised with Prpc, accumulated in endosomes and subsequently trafficked to Iysosomes. We report that the cell surface binding, internalisation and downstream toxicity of Aβ oligomers is dependent on the transmembrane low density lipoprotein receptor- related protein-1 (LRP1). Western blotting revealed that the binding of AI3 oligomers to cell surface Prpc impaired its ability to inhibit the activity of the l3-secretase BACE1 which cleaves the amyloid precursor protein to produce AI3. Using an siRNA approach, AI3 oligomers were found to activate the signalling molecule STEP61 in a Prpc-dependent manner. AI3 oligomers also stimulated caspase-3 activation via Prpc. These data indicate that soluble, fibrillar Aβ oligomers bind to Prpc in a conformation-dependent manner and require the transmembrane LRP1 for their cytotoxicity, thus revealing potential targets for therapeutic intervention in AD.

SgrAI is a restriction endonuclease (ENase) that cuts a long recognition sequence and exhibits self-modulation of cleavage activity and sequence specificity. Previous research has shown that SgrAI forms large oligomers when bound to particular DNA sequences and under the same conditions where SgrAI exhibits accelerated DNA cleavage kinetics. However, the detailed structure and stoichiometry of SgrAI:DNA as well as the basic building block of the oligomers, has not been fully characterized. Ion mobility mass spectrometry (IM-MS) was employed to analyze SgrAI/DNA complexes and show that the basic building block of the oligomers is the DNA-bound SgrAI dimer (DBD). The oligomers are heterogeneous containing a mixture of species with variable numbers of DBD. The collision cross sections (CCS) of the oligomers were found to have a linear relationship with the number of DBD. Models of the SgrAI/DNA oligomers were constructed and a head-to-tail arrangement was most consistent with the experimental CCS.

The unfolding and refolding of a number of oligomeric enzymes have been studied. These were: fumarase from pig heart, the NAD+ -dependent isocitrate dehydrogenase from yeast, the citrate synthases from pig hean, Acinetobacter anitratum and Thermoplasma acidophilum and the chaperone protein GroEL from Escherichia coli. In each case the unfolding by guanidinium chloride (GdnHCI) was monitored by enzyme activity (to detect possible perturbations at the active site), protein fluorescence (to detect changes in tertiary structure) and far U.v. circular dichroism (to detect changes in protein secondary structure). In general the losses in secondary and tertiary structure were found to run broadly in parallel, whereas the enzyme activity was lost at much lower concentrations of GdnHCl. This sensitivity to mild, denaturing conditions may reflect the greater flexibility of the active site compared with the molecule as a whole. Interestingly) the bacterial citrate synthases were activated in the presence of low concentrations of GdnHCl. Following denaturation) refolding was initiated by lowering the concentration of GdnHCI by dilution or dialysis. Only the dimeric citrate synthases (from pig heart and Thermoplasma acidophilum) could be reactivated to a moderate extent using the dilution procedure; less than 5% reactivation was observed for the other enzymes. In the cases of fumarase, NAD+ -dependent isocitrate dehydrogenase and the dimeric citrate synthases the degrees of reactivation following dialysis were significantly greater (approximately 50-75% of the native enzymes) than those obtained following the dilution procedure. Factors such as protein concentration and the inclusion of dithiothreitol in the dialysis or dilution buffer were found to influence significantly the extent of reactivation. The greater yield of reactivation of unfolded protein using the dialysis procedure probably reflects the ability of the enzyme to make the correct structural adjustments between intermediates when the concentration of GdnHCI is lowered gradually. In the case of Thermoplasma acidophilum the recovery of citrate synthase activity was much greater at 20 ·C than at 55 ·C (the optimal temperature for growth of this organism). This has implications for the folding process in vivo under the extreme growth conditions of thermophiles and possibly other extremophiles. The hexameric citrate synthase from Acinetobacter anitratum and the tetradecameric chaperonin, GroEL could not be reactivated following denaturation. Far u.v. circular dichroism measurements on GroEL indicated that the native secondary structure of this protein was regained to a large extent. In vivo a number of the proteins studied (fumarase and citrate synthase from pig hean and yeast NAD+ -dependent isocitrate dehydrogenase)are translocated into mitochondria as precursors in a non-native state prior to processing, folding and assembly. The lack of complete refolding of the proteins studied in this work points to the existence of specialised mechanisms in vivo to promote efficient folding. Chaperone proteins have been implicated in the assistance of protein folding in vivo. Intriguingly. the studies on the inefficient refolding of the chaperonin GroEL support the proposal that this protein may fold in vivo by way of a "self chaperoning" mechanism.

The regulation of protein function through oligomerization is a common theme in biological systems. In this work, I have focused on the effects of the oligomeric states of the human Rad52 protein on activities related to DNA binding. HsRad52, a member of the RAD52 epistasis group, is thought to play an important and as yet undefined role in homologous recombination. HsRad52 preferentially binds to ssDNA over dsDNA and stimulates HsRad51-mediated strand exchange (Benson et al., 1998). In either the presence or absence of DNA, HsRad52 has been observed to form both 10 nm ring-like structures as well as higher order oligomers consisting of multiple 10 nm rings (Van Dyck et al., 1998; Van Dyck et al., 1999). Earlier protein-protein interaction studies mapped the domain responsible for HsRad52 self-association in the N-terminus (residues 85-159) (Shen et al., 1996). Data presented here identifies a novel self-association domain in the C-terminus of HsRad52 that is responsible for the formation of higher order oligomers. VanDyck et al. observed DNA ending binding complexes consisting of multiple rings (Van Dyck et al., 1999). They proposed that these higher order oligomers may be functionally relevant. In this work, we demonstrate that DNA binding depends on neither ring shaped oligomers nor higher order oligomers but that activities of HsRad52 that require simultaneous interaction with more than one DNA molecule depend on the formation of higher order oligomers consisting of multiple HsRad52 rings. Early studies of HsRad52 proposed that the DNA binding domain resides in the highly conserved N-terminus of the protein (Park et al., 1996). A series of studies using truncation mutants of HsRad52 have provided evidence that supports this hypothesis. For example, we demonstrated that a truncation mutant containing only the first 85 residues of the protein is still able to bind DNA (Lloyd, submitted 2002). In this study, we demonstrate that aromatic (Y65, F79 and Y81) and hydrophobic (L43, I52 and I66) residues within the N-terminus contribute to DNA binding by either directly contacting the DNA or by stabilizing the structure of the protein. In summary, through the work presented in this dissertation, we have determined that the formation of 10 nm rings is mediated by a self-association domain in the N-terminus and that the formation of higher order oligomers consisting of multiple HsRad52 rings is mediated by an additional self-association domain in the C-terminus. We have correlated the oligomeric properties of HsRad52 with its biochemical functions related to DNA binding. Additionally, we have demonstrated that aromatic and hydrophobic residues contribute to DNA binding. Further studies will differentiate between the contribution of these residues to the DNA binding by stabilizing the overall structure of the protein versus making specific DNA contacts. Additional studies will also address how the oligomeric state of HsRad52 contributes to its role in HsRad51-mediated strand exchange.

Small angle X-ray scattering (SAXS) is a powerful tool for quaternary structure determination of proteins in solutions. Even if data acquisition with this technique is relatively simple from the experimental point of view, the data interpretation is not so straightforward. Several methods and programs have been described for the ab initio determination of quaternary structure from SAXS measurements, even at relatively high resolution. However, none of these programs utilize some known structural characteristics of the proteins used in this study, hemocyanins. These characteristics are the presence of Dn type symmetries and the geometry of the structural subunits that form oligomeric proteins. This work describes a new method for the quaternary structure determination of oligomeric proteins from experimental SAXS curves of such whole proteins and simplified models of structural subunits. Structural subunit models are generated from biochemical data, electron microscopy or high resolution structures. These models, composed of spheres with varying radii, were used to reconstruct the quaternary structure of whole proteins by fitting their orientation and position to the experimental SAXS data. The fits were obtained with custom made programs and can be used to model any oligomeric protein with similar symmetry constraints. The method has been verified with arthropod hemocyanins, where it succeeded to reconstruct the quaternary structure of whole molecules. Consequently, the same method has been applied to much more complex molluscan hemocyanins. In this case, only the positions of the structural subunits could be obtained at a satisfactory level while it was impossible to determine their spatial orientations. The reason is the almost spherical form of structural subunit monomers used for the fits. The method predicts the creation of higher resolution monomer models, but it would require a lot more computing time.
La diffusione a piccoli angoli di raggi X (SAXS) è una tecnica molto efficace nello studio della struttura quaternaria delle proteine in soluzione. Anche se l'acquisizione di tali misure è relativamente facile dal punto di vista sperimentale, la loro interpretazione non è immediata. In letteratura sono descritti diversi programmi con i quali si possono ottenere modelli strutturali direttamente dalle curve SAXS, anche a risoluzione relativamente elevata. Per contro nessun programma tiene conto di alcune caratteristiche note delle proteine oggetto di questo lavoro (emocianine), quali per esempio la presenza delle simmetrie Dn e la struttura delle subunità strutturali che le costituiscono. Questa tesi riguarda lo sviluppo di un metodo di determinazione della struttura quaternaria delle proteine oligomeriche a partire da curve SAXS di molecole intere e modelli semplificati di subunità strutturali. I modelli delle subunità strutturali sono costruiti partendo da informazioni derivanti da tecniche biochimiche, microscopia elettronica o strutture ad alta risoluzione. Questi modelli, composti da sfere a raggio diverso, sono stati usati per ricostruire la struttura delle molecole oligomeriche, fittando i dati su curve SAXS tramite programmi realizzati ad hoc. Tali programmi possono essere utilizzati per modellare qualsiasi proteina con caratteristiche di simmetria simili. Il metodo è stato verificato sulle emocianine di artropodi dove è riuscito a ricostruire in maniera soddisfacente la struttura quaternaria delle molecole intere. A questo punto, l'analisi è stata estesa anche sulle emocianine di molluschi, proteine con struttura molto più complessa. In questo caso, si è riusciti ad ottenere, in maniera soddisfacente, solamente le posizioni delle subunità strutturali che compongono la proteina intera, mentre non era possibile determinarne il corretto orientamento nello spazio. Questo risultato sarebbe da attribuire alla forma pressoché sferica dei monomeri strutturali usati nei fit. La generazione di monomeri ad una risoluzione più elevata è prevista dal metodo ma comporterebbe un tempo di calcolo maggiore.

Les interactions protéine-protéine (IPPs) représentent une classe importante de cibles thérapeutiques. En effet, la plupart des processus cellulaires sont composées ou influencés par les IPPs et leur disfonctionnement peut entraîner de nombreuses maladies. Même si les surfaces protéiques sont généralement larges, des éléments clés sont responsables de l'interaction et de la reconnaissance protéique (hot spots). La structure secondaire la plus commune dans les protéines naturelles est l’hélice alpha et les petites molécules semblent être des candidats intéressants pour perturber ces interactions protéine-protéine. Nous nous sommes particulièrement intéressés aux protéines de la famille Bcl-2 qui sont des régulateurs importants de l'apoptose. Elles sont surexprimées dans de nombreux cancers et contribuent à l'initiation tumorale, la progression et la résistance à la thérapie anticancéreuse. Le but de notre étude est de concevoir des petites molécules capables d'interagir avec les membres anti-apoptotiques et ainsi restaurer le mécanisme de l'apoptose mitochondriale. Nous avons synthétisé une chimiothèque d'oligomères de structure thiénylpyridine. Ces foldamères d’un genre nouveau imiteraient les éléments clés d'une surface protéique en mimant l'orientation spatiale des différents substituants d’une hélice alpha. La plupart de ces composés ont été engagée dans des étude. S de modélisation moléculaire associées à des résultats de radiocristallographie afin d'évaluer la capacité de mimes d’hélice alpha. Par ailleurs, l’évaluation biologique de ces composés a été réalisée et les premiers résultats montrent que nos composés sont capables de rompre les interactions protéine-protéine
Protein-protein interactions (PPIs) are attractive targets for drug discovery. Indeed, most of cellular processes are composed or influenced by PPIs and their misregulation can result in numerous disease states. Even if proteins surfaces are generally large, some key elements are responsible of the interaction and the recognition (hot spots). The most common secondary structure in natural proteins is an alpha helix. Small molecules seem to be attractive candidates for stabilizing or disrupting protein-protein interactions based on alpha helices. We are particularly interested in Bcl-2 family proteins which are central regulators of apoptosis. They are overexpressed in many cancers and contribute to tumour initiation, progression and resistance to therapy. The aim of our study is to design small molecules able to interact with anti-apoptotic members and restore the intrinsic apoptosis mechanism. These foldamers would mimic only the key elements of a protein surface and potentially lead to small molecules having almost the full activity of a protein domain. For this, we synthesized a library of thienylpyridyl oligomers. Most of these compounds underwent molecular modeling studies based on radiocrystallography in order to evaluate the ability to mimic an alpha helix and the spatial orientation of different substituents. Moreover, biological tests have been done to evaluate the ability to induce apoptosis in cancer cells

Bone bruise lesions (BBL) are documented on MRIs diagnosing acute knee ligament injury (AKLI). Recent evidence has indicated that a majority of patients that sustain an AKLI, especially anterior cruciate ligament (ACL) knee injury, will develop post-traumatic osteoarthritis (PTOA) 10-20 years following injury. It has been proposed that the initial damage sustained to the articular cartilage overlying BBL causes a cascade of events that may result in PTOA. Researchers have proposed a modification to treatment protocols for more severe BBL, or have stressed the need for the development of protective therapies to protect the articular cartilage. However, there are limited tools available to evaluate the clinical outcome of articular cartilage overlying BBL. Furthermore, damage to the cartilage overlying BBL may be different according to differing BBL severities. Therefore, the use of a cartilage degradation biomarker, serum cartilage oligomeric matrix protein (sCOMP) and the use of a BBL severity classification system may be useful to determine if differences exist between patients with and without BBL, and with differing BBL severities. The purpose of this dissertation was to investigate the utility of sCOMP as a biomarker for acute articular cartilage damage. The purposes of these studies were to determine the inter and intraday reliability of this marker, to document sCOMP longitudinally in collegiate athletes and following AKLI, and to determine if differences in sCOMP and self-reported pain and function exist for patients with and without BBL, and differing BBL following AKLI. The results of these studies indicated sCOMP measures had strong inter and intraday reliability. Additionally, exercise does seem to influence sCOMP levels; however, these elevations may not be clinically meaningful. Furthermore, sCOMP levels were not different between patients with BBL and without, and between differing BBL severities. The results of these studies support the use of a BBL severity classification for future research studies in order to further elucidate the outcomes of these lesions.

Thesis (Ph.D.)--Boston University
Alzheimer's disease (AD) is characterized by progressive dementia and accumulation of a cleavage product of the amyloid precursor protein, amyloid-β (Aβ) peptide, in the brain. Several lines of evidence suggest that soluble, oligomeric intermediates of Aβ are primarily responsible for synaptic dysfunction and the cognitive deficit observed in AD. The cellular prion protein (PrP^c), a cell surface glycoprotein involved in transmissible spongiform encephalopathies, was previously identified as a high affinity receptor for Aβ oligomers. It has been suggested that binding of Aβ oligomers to PrP^c transduces the synaptotoxic events seen in AD. The two reported binding sites of Aβ oligomers are located on the unstructured N-terminal tail of PrP^c. We show here that the soluble physiological cleavage fragment of PrP^c, N1, was necessary and sufficient for binding Aβ oligomers. This binding interaction was influenced by positively charged residues in the two binding sites and is dependent on the length ofthe sequence between them. Importantly, the addition of synthetic N1 peptide suppressed Aβ oligomer toxicity in cultured murine hippocampal neurons and in a mouse model of Aβ-induced memory dysfunction. Collectively, these data suggest that N1, or small peptides derived from it, could be potent inhibitors of Aβ oligomer toxicity by targeting Aβ oligomers and represent an entirely new class of therapeutic agents for AD. To directly target PrP^c as a toxicity receptor, in silica screening and molecular dynamics were used to generate small molecule ligands. We screened these ligands using biochemical and biophysical assays to identify high affinity ligands for PrP^c that block the binding of Aβ oligomers. We found one compound, called DS26 that bound to PrP^c with sub-micromolar affinity. Further, DS26 inhibited Aβ-dependent suppression of long-term potentiation in mouse hippocampal slices. Interestingly, we show that DS26 operated by an unexpected allosteric mechanism in which ligand binding to a site in the structured C-terminal half of PrP^c induced an intramolecular interaction with the N-terminal tail, thereby preventing Aβ binding. Together, these data demonstrate that pharmacologically targeting PrP^c can suppress Aβ toxicity. Additionally, this study clarifies previous conflicting studies regarding the role of PrP^c in AD.

Les fibres amyloïdes sont souvent à l’origine de nombreuses maladies dégénératives telles que la maladie d’Alzheimer ou la maladie de Parkinson. La formation de ces plaques insolubles est due à une agrégation anormale de protéines. Les études structurales et biologiques des amyloïdes sont hautement complexes du fait de leur organisation sous forme de superstructures unidirectionnelles composées d’une infinité d’unités peptidiques ou protéiques, mais aussi à cause de leur hétérogénéité conformationnelle et polymorphique. Au cours de ces différents travaux de thèse en collaboration avec différents laboratoires d’analyses structurales, nous avons développé plusieurs outils de synthèse tant pour la formation de différents polymorphes de fibres amyloïdes que pour la formation d’espèces oligomériques de tailles conséquentes qui sont un challenge du point de vue synthétique et méthodologique mais aussi pour leur caractérisation. Ces différentes avancées permettront de mieux comprendre les mécanismes de formation de fibres amyloïdes et de préparer des échantillons homogènes pour les analyses structurales et biologiques. L’étude de modifications chimiques telles que la N-méthylation ou les polypeptides D est également un enjeu important pour l’élucidation des interactions protéine-protéine vis-à-vis des structures amyloïdogéniques et ainsi permettre l’élaboration de nouveaux composés inhibant la formation de plaques amyloïdes
Amyloid fibrils are associated with many human disorders including Alzheimer’s or Parkinson’s diseases. The formation of insoluble plaques is the result of protein misfolding and aggregation due to abnormal conformational isomerization of the involved protein. The structural and biological studies of amyloids are highly complex. In this thesis, we report on the development of different synthetic methodologies for the preparation of distinct amyloid fibril polymorphs as homogeneous samples for structural and biological studies. We also synthesized covalently-tethered oligomers composed of nine copies of an amyloidogenic peptide segment, where we were able to control the self-assembly of the structure by insertion of N-methylated amino-acids and to obtain monomeric oligomers mimicking a cross section of an amyloid fibril. We also report on the chiral recognition of L-peptides and L-proteins towards corresponding D-enantiomers during amyloid formation. Moreover, we studied various N-methylated peptide analogues to suppress amyloid growth. Overall, the results obtained in this thesis pave the way towards rational design of peptide-based inhibitors and diagnostics against amyloid propagation

Doutoramento em Biologia - Instituto Superior de Agronomia
Blad is an edible, Lupinus cotyledon-derived 173 amino acid residue polypeptide which occurs naturally and undissociated as part of a soluble 210 kDa oligomer, hereby termed Bladcontaining oligomer (BCO). BCO is isolated during growth of Lupinus seedlings and is a breakdown product of β-conglutin catabolism. In this work, BCO inherent, potent and wide spectrum antifungal activity was demonstrated under in vitro and greenhouse conditions. Several field trials showed that BCO exhibits an equal or better performance than the conventional chemicals over a large variety of crops/fungal pathogens. BCO antifungal activity was also showed to extend to human pathogens, including both yeasts and filamentous fungi. In vitro tests demonstrated a higher efficacy on a molar basis when compared to the best chemical fungicides tested, with an absence of mammal toxicity or genotoxicity. In addition to the lectin and catalytic activities which allow BCO to target the fungal cell wall, this work revealed other highly complex and multitarget mode(s) of action. BCO enters fungal cells causing several morphological alterations and seems to essentially act through inhibition of metal ion homeostasis resulting in an apoptotic cell death. In an attempt to provide a structural characterization of Blad, data from electronic microscopy, dynamic light scattering, SAXS and X-ray crystallography were obtained from heterologous expressed Blad, however the results were not conclusive. Last but not least, this work also aimed at the development of a “super- Blad”: a fusion protein consisting of Blad and two selected antibacterial peptides, SP10-5 and Sub5. The resulting chimeras not only maintained a strong antibacterial activity but were also able to retain the ability to inhibit the growth of both yeast and filamentous fungi. In summary, the overall approach of this thesis considered phytopathogens and human pathogens as belonging to the same scientific area and therefore aimed at producing a uniform natural multitarget, step-stone antifungal solution
N/A

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemistry, 2004.
Vita.
Includes bibliographical references.
Oligomeric mini-proteins, short peptides with protein-like features, constitute valuable minimal models for the study of oligomeric proteins. Oligomerization is a common feature of cellular proteins that may confer structural and functional advantages. Oligomerization is proposed to have arisen by several evolutionary pathways. The structural characterization of peptide 1, a stable oligomeric mini-protein previously developed in the Imperiali group, was undertaken by X-ray crystallography, as knowledge of the structure would enable the rational design of subsequent generations of BBA oligomers varying in packing and stoichiometry. The structure of peptide 1 could not be solved by direct methods, by molecular replacement with search models derived from the monomeric precursor, or by the introduction of heavy atoms. Two selenomethionine mutants having solution-phase properties comparable to the native were identified. The structures of these two peptides were independently solved via MAD phasing experiments, and the refined structures employed as search models for a molecular replacement solution of 1. The structures of the three peptides are homologous, and constitute the first reported structures of a mixed act/ oligomeric mini-protein. The X-ray crystal structures reveal that the oligomeric BBA motif has a domain-swapped architecture that supports a protein-like and water-exclusive core. The structures elucidate the unique role of unnatural amino acids in conferring native secondary structure in a short peptide sequence (21 residues per monomer).
(cont.) Furthermore, the crystal structures reveal that the stoichiometry of the oligomer is tetrameric, rather than trimeric, as originally proposed. A tetrameric solution-phase stoichiometry for this mini-protein family was confirmed by rigorous analytical ultracentrifugation experiments. Heterooligomeric BBA peptides have been designed and characterized in collaboration with Christina M. Taylor and Professor Amy E. Keating of the MIT Biology Department. Acidic and basic residues were substituted along the inter-monomer interface and specific steric interactions were designed in order to disfavor homoassociation and favor heteroassociation. Heterotetramers comparable to peptide 1 in terms of structure and stoichiometry, and approaching the native homotetramer in terms of stability, have been characterized by a variety of biophysical techniques.
by Mayssam H. Ali.
Ph.D.

Une des caractéristiques histopathologiques de la maladie d'Alzheimer (AD) est la présence de plaques amyloïdes formées par les peptides amyloïdes β (Aβ) de 40 et 42 résidus, qui sont les produits de clivage par des protéases de l'APP. Afin de comprendre le rôle des variations structurelles du TM dans le traitement de l'APP, les peptides APP_TM4K ont été étudiés dans la bicouche lipidique en utilisant l’ATR-FTIR et ssNMR. Tandis que la structure secondaire globale du peptide APP_TM4K est hélicoidale, hétérogénéité de conformation et d'orientation a été observée pour le site de clivage γ et , que peuvent avoir des implications dans le mécanisme de clivage et donc dans la production d’Aβ. Les peptides Aβ s'agrègent pour produire des fibrilles et aussi de manière transitoire d'oligomères neurotoxiques. Nous avons constaté qu'en présence de Ca2+, l’Aβ (1-40) forme de préférence des oligomères, tandis qu'en absence de Ca2+ l'Aβ (1-40) s’agrège sous forme de fibrilles. Dans les échantillons sans Ca2+, l’ATR-FTIR révèle la conversion des oligomères en feuillets β antiparallèles en la conformation caractéristique des fibrilles en feuillets β parallèles. Ces résultats nous ont amené à conclure que les Ca2+ stimulent la formation d'oligomères d'Aβ (1-40), qui sont impliqués dans l’AD. Les positions et une orientation précise de deux nouveaux médicaments puissants modulateurs de la γ-sécrétase - le benzyl-carprofen et le sulfonyl-carprofen  dans la bicouche lipidique, ont été obtenus à partir des expériences des ssNMR. Ces résultats indiquent que le mécanisme probable de modulation du clivage par la y-sécrétase est une interaction directe avec le domaine TM de l’APP
A histopathological characteristic of Alzheimer’s disease (AD) is the presence of amyloid plaques formed by amyloid β(A) peptides of 40 and 42 residues-long, which are the cleavage products of APP by proteases. To understand the role of structural changes in the TM domain of APP, APP_TM4K peptides were studied in the lipid bilayer using ATR-FTIR and ssNMR. While the overall secondary structure of the APP_TM4K peptide is helical, conformational and orientational heterogeneity was observed for the y- and for the -cleavage sites, which may have implications for the cleavage mechanism and therefore the production of Aβ. Starting from its monomeric form, Aβ peptides aggregate into fibrils and / or oligomers, the latter being the most neurotoxic. We found that in the presence of Ca2 +, Aβ (1-40) preferably forms oligomers, whereas in the absence of a2 + Aβ (1-40) aggregates into fibrils. In samples without Ca2 +, ATR-FTIR shows conversion from antiparallel β sheet conformation of oligomers into parallel β sheets, characteristic of fibrils. These results led us to conclude that Ca2 +stimulates the formation of oligomers of Aβ (1-40), that have been implicated in the pathogenesis of AD. Position and precise orientation of two new drugs  powerful modulators of γ-secretase  benzyl-carprofen and carprofen sulfonyl  in the lipid bilayer were obtained from neutron scattering and ssNMR experiments. These results indicate that carprofen-derivatives can directly interact with APP. Such interaction would interfere with proper APP-dimer formation, which is necessary for the sequential cleavage by β -secretase, diminishing or greatly reducing Aβ42 production

Knee pain can alter lower-extremity neuromechanics and often results in functional disability. The relationship between lower-extremity neuromechanical alterations, due to anterior knee pain, and articular cartilage condition is unclear. The purpose of this study was to determine the independent effect of anterior knee pain during running on articular cartilage condition, as reflected by serum cartilage oligomeric matrix protein concentrations and muscle cocontraction duration. Seven men and five women completed a 30-min run in three different sessions: control (no infusion), sham (isotonic saline infusion), and pain (hypertonic saline infusion). Saline was infused into the right infrapatellar fat pad for the duration of the run. Subject-perceived pain was recorded every 3 min on a 100-mm visual analog scale. During the run, bilateral electromyography was recorded for five leg muscles, and heel and toe markers were used to track foot position. During the 30-min run of the pain session average subject-perceived pain was 27.8 (SD = 2.3 mm) and 19.7 (SD = 1.9) mm greater than during the control (0.0 mm) and sham (8.1 mm) session, respectively (p < 0.01). Knee pain while running did not result in changes in muscular cocontraction duration (p = 0.13). Blood samples were drawn prior to the run, immediately following the run, and 60 min following the run. Samples were analyzed using enzyme-linked immunosortbent assay to determine serum cartilage oligomeric matrix protein concentration. Average serum cartilage oligomeric matrix protein concentration was 14% greater at immediate post run (132.19 ± 158.61 ng/ml; Range = 22.61-290.81 ng/ml) relative to pre run (116.02 ± 118.87 ng/ml; Range = 19.81-234.89 ng/ml) (p < 0.01), and 18% less at 60 min post run (108.45 ± 171.78 ng/ml; Range = 20.84-280.23 ng/ml) relative to immediate post run (Figure 4; p < 0.01). Serum cartilage oligomeric matrix protein did not significantly differ between baseline and 60 min post-exercise (p = 0.29). There was not a difference in cartilage oligomeric matrix protein concentration between sessions. Knee pain while running does not cause an increase in serum cartilage oligomeric matrix protein concentration (p = 0.29). There are two important findings from this study. First, anterior knee pain during a 30 min running session does not appear to independently affect cartilage oligomeric matrix protein concentrations. This implies other factors, aside from anterior knee pain alone, influence articular cartilage degradation during movement that occurs while individuals are experiencing anterior knee pain. Second, the present experimental anterior knee pain model can be used to evaluate the independent effects of anterior knee pain over an extended duration while subjects perform a dynamic activity like running.

Alzheimer’s disease (AD) is the most common form of dementia. It was first described in 1906 by Alois Alzheimer. Later on, in 1984 George Glenner and Colin Masters isolated the amyloid-beta (Aβ) peptide from a human brain and associated it to the disease. Since then the amyloid hypothesis has been a rather controversial matter discussed among the scientific community. This is because although Aβ has been targeted by the majority of the drugs in clinical trials not even one has been approved up to date: 13 have been discontinued and 10 are in phase 3 clinical trials. A possible explanation for these devastating numbers is the high complexity of the target due to the variety of aggregation forms that Aβ can adopt. Therefore, understanding the links between protein aggregation and neurotoxicity, and specially obtaining the 3D structures of the aggregates responsible for neurotoxicity is key to design effective diagnostic and therapeutic strategies. Unfortunately, this remains one of the most important unresolved issues in the field. The group of Dr. Carulla has been working on the hypothesis that Aβ interacts with the cell membrane leading to ionic dyshomeostasis. In order to study this scenario, the group has changed the paradigm and treated Aβ as a membrane protein and applied well known methodologies used to characterize this family of proteins to study Aβ. By doing so, the group has proved that Aβ is able to form a type of oligomer in the presence of detergent micelles which adopts a very specific and defined structure with characteristics of a β-barrel assembly and functions as a pore. They refer to these types of oligomer as β-Barrel Pore-Forming Oligomer (βPFO). Hereby we present the work carried out to identify by using different biophysical techniques, the 3D structure of βPFO. As a starting point, we have used detergents to study the oligomerization process in a membrane mimetic environment. Micelles compared to other more native-like biomimetics environments based on lipids, will enable the application of novel mass spectrometry (MS) strategies and well-established solution NMR techniques thus providing high-resolution structural information. Since the accumulation of different amounts of Aβ in the membrane is a plausible scenario in the context of the disease, we have used different Aβ to detergent micelle ratios ([Aβ]:[M]) to study the role of this variable in the oligomerization process of Aβ. Throughout the work done we have optimized not only the ratio but also other conditions such as the buffer and the pH to modulate the preparation of samples enriched in defined oligomer populations. To study the stoichiometry of βPFO, we used with Native Mass Spectrometry which proved to be an adequate technique to preserve the non-covalent interactions of our samples analyse them in the gas phase. One of the key parts of the project consisted in the screening of a wide range of non-ionic detergents compatible with MS. After this work as we were able to identify Pentaethylene Glycol Monooctyl Ether (C8E5) as the best candidate for our samples. To continue working with the different samples we implemented a new approach based on coupling size exclusion chromatography (SEC) directly to a SYNAPT G2. This approach has allowed us to establish that higher molecular weight oligomers are better preserved and therefore better detected as we increase the signal to noise ratio. This enabled us to study different points of the SEC chromatogram and therefore understand better the composition of our samples and our system. For the standard βPFO samples, we reported specific charge states for the octamer and tetramer species. In parallel to complement the native-MS results, we have also worked to develop a method to analyse chemically cross-linked βPFOs by MALDI-MS. After a process of trials and optimizations we established a zero-length cross-linker (DMTMM) which allowed us to cross-link the βPFOs and detect again tetramer and octamer such as in the native-MS approach. In order to assess the relevance and to potentially validate the standard βPFO preparation as a target for AD’s it is crucial to characterize the binding of the Nanobodies to the oligomer. This work will also give us the opportunity to generate Nanobodies that could recognize their specific structures in brain tissue and thus assess whether the oligomers proposed are related to AD’s and if so, evaluate them as new targets for AD. Moreover we are very interested in the potential use of these Nanobodies as novel diagnostics or therapeutics tools.
La malaltia d'Alzheimer (AD) és la forma més comuna de demència. Va ser descrita per primera vegada el 1906 per Alois Alzheimer. Més endavant, al 1984, George Glenner i Colin Masters van aïllar el pèptid amiloide-beta (Aβ) d'un cervell humà i el van associar a la malaltia. Des de llavors, la hipòtesi amiloide ha estat un tema bastant controvertit discutit entre la comunitat científica. Una possible explicació és l’alta complexitat del sistema a causa de la varietat de formes d’agregació que Aβ pot adoptar. Per tant, entendre els vincles entre l'agregació de proteïnes i la neurotoxicitat, i especialment l'obtenció de les estructures 3D dels agregats responsables de la neurotoxicitat, és clau per dissenyar estratègies diagnòstiques i terapèutiques efectives. Malauradament, aquest tema continua sent un dels problemes pendents més importants. El grup de la Dra. Carulla ha estat treballant en la hipòtesi que l'Aβ interactua amb la membrana cel·lular que condueix a una deshomeostasi iònica. Per estudiar aquest escenari, el grup ha canviat el paradigma i ha tractat Aβ com a proteïna de membrana i aplicant tècniques biofísiques ben establertes per a caracteritzar proteïnes de membrana per tal d’estudiar Aβ. D'aquesta manera, el grup ha demostrat que Aβ és capaç de formar un tipus d'oligòmers en presència de micel·les de detergent que adopten una estructura molt específica i definida amb capacitat de formar porus a través de membranes lipídiques. Es refereixen a aquest tipus d’oligòmers com a oligòmers formadors de porus barril β (βPFO). A la present tesi doctoral, presentem l’estudi realitzat per identificar mitjançant diferents tècniques biofísiques, l'estructura 3D de βPFO. Hem utilitzat detergents per estudiar el procés d’oligomerització en un entorn mimètic de membrana. Les micel·les en comparació amb altres entorns biomimètics basats en lípids, permeten l'aplicació d’estratègies d'espectrometria de masses (MS) i de ressonància magnètica nuclear (RMN) ben establertes, proporcionant així informació estructural d'alta resolució. Atès que l’acumulació de diferents quantitats d’Aβ a la membrana és un escenari plausible en el context de la malaltia, hem utilitzat diferents relacions de micel·les Aβ a detergents ([Aβ]: [M]) per estudiar el paper d’aquesta variable en l’oligomerització. procés d'Aβ.

The topic of this PhD project concerns aspects of the general problem of the protein folding and unfolding, in line with the research conducted in the laboratory of Protein Chemistry at CRIBI, where the activities were mostly performed. The mechanism of acquisition of the three-dimensional structure of proteins (folding) is an important biological event. It is generally a multi-stage process involving the formation of intermediates, which are partly folded states having some structural features of the native protein, but not the final side chains interactions that allow the protein to exert its specific function. The failure in achieving the correct folding (misfolding) may cause protein aggregation. In fact, partly folded proteins can easily self-assembly in regular insoluble aggregates (amyloid fibrils), which are associated with serious diseases, the so called amyloidosis (Chiti & Dobson, 2006). In particular, my PhD research was focused on ?-lactalbumin (?-LA), a milk metalloprotein, widely used as a model to study protein folding. A new wave of interest in this protein appeared in the last decade after the discovery of a variant of the protein that, besides its physiological role in lactation, is able to induce apoptosis of tumor cells (Svensson et al., 1999). The cytotoxic activity resides in the formation of a complex with a milk fatty acid, oleic acid (OA), named HAMLET (Human Alpha-lactalbumin Made LEthal to Tumour cells) (Svensson et al., 2000). The cell events caused by HAMLET were extensively studied (Kohler et al., 2001; Duringer et al., 2003). On the contrary, the mechanism and the nature of the interaction between ?-LA and OA and the physico-chemical properties of the complex are not completely understood yet. The preparation of the complex is a controversial aspect. Indeed, despite the effective association of OA with the protein in solution (Polverino de Laureto et al., 2002), the resulting ?-LA/OA complex was believed to possess lower antitumor activity than HAMLET, obtained using a chromatographic procedure (Svensson et al., 2000). Other debated matters focus on the protein/fatty acid stoichiometry and the monomeric/oligomeric state of ?-LA in the complex. Moreover, a systematic investigation of the effect of the protein on the phase behavior of OA is still lacking. The aim of this PhD Thesis was the investigation of the OA binding propensity and cytotoxic activity of three proteolytic derivatives of bovine ?-LA, obtained by limited proteolysis with pepsin at acidic pH. The use of proteolytic dissection of a protein has been widely employed to study the folding of several proteins (Wetlaufer, 1981; Gegg et al., 1997; Llinás & Marqusee, 1998). The characterization of protein fragments containing specific structural elements of the intact protein (?-helix, ?-sheet) and able of autonomous folding was successfully used to unravel folding features of ?-LA (Polverino de Laureto et al., 1999; 2001). Indeed, fragments corresponding to structural domains in multidomain proteins were shown in some cases to acquire in solution a native-like conformation (Fontana et al., 2004). The ?-LA fragments investigated are: species 1–40/53–123, lacking the ?-subdomain of the native protein; 1–40/104–123, given by the N-terminal fragment 1–40 covalently linked by two disulfide bridges to the C-terminal fragment 104–123 and containing three of the four ?-LA helices; 53–103, containing the C-helix and the calcium binding loop of intact ?-LA (Polverino de Laureto et al., 1999; 2001). The conformational differences between the three fragments were used as the rationale for a study of their efficacy to bind OA, thus giving an insight into the mechanism of this binding. This study has also some physiological relevance: pepsin is an enzyme of the stomach and therefore these species might be generated in vivo, under the same conditions of low pH in which the complex HAMLET was hypothesized to form, thus contributing to the apoptotic activity of the complex formed by the intact protein. The complexes between the three fragments and OA were prepared first by the chromatographic procedure described for HAMLET, then by directly mixing in solution the two components. The conformational properties of the complexes were characterized by circular dichroism, showing that the complexes prepared by both procedures display similar conformations and acquire ?-helix. The effect of calcium on the conformation of the complexes was then investigated by circular dichroism. Fluorescence spectroscopy was used to study the involvement of Trp residues in the interaction with OA. Moreover, the ability of the complexes to induce apoptosis-like cell death was evaluated, in line with the cytotoxic activity displayed by HAMLET. In order to verify the physical state of OA involved in the interaction with ?-LA, the OA aggregation behavior was investigated using different techniques, such as transmission electron microscopy (TEM), titration with a fluorescent dye and turbidimetric analyses, and an effect of solubilization of OA was observed. A manuscript with these results will be soon submitted to an international journal and is herewith attached. In the second part of the PhD, the research was focused on the aggregation propensity of ?-LA and its three proteolytic fragments. ?-LA is able to form amyloid fibrils in vitro, even if not disease-related. ?-LA fibrils are formed when the structure of the protein is partially destabilized, e.g. at low pH, upon reduction of three disulphide bridges at neutral pH (Goers et al., 2002), or after proteolytic cleavage (Polverino de Laureto et al., 2005). Lipids may act as an effective catalyst of fibrillogenesis, providing a generic environment where protein molecules adopt conformation and orientation promoting their assembly into fibrillar structures (Thirumalai et al., 2003; Stefani, 2004; Sparr et al., 2004; Zhao et al., 2004). In particular, since fatty acid–protein interactions are known to modulate the process of fibrillation (Kim & Takahashi 2006), aggregation processes of ?-LA and its fragments were investigated to elucidate whether and how OA might affect fibrils formation. Firstly, aggregation was followed at pH 2.0, as a comparison with the known fibrillogenic behavior of intact ?-LA. Secondly, the aggregation was investigated at pH 7.4, since in this physiological condition the conformational changes induced by OA on the fragments structure were studied and their cytotoxic activity was analyzed, in the first part of the Thesis. The formation of fibrils was followed by thioflavin T (ThT) fluorescence assay, circular dichroism and TEM. All three fragments were able to produce amyloid fibrils at pH 2.0, similarly to ?-LA. In the used range of protein concentration and protein/lipid ratio, OA seems to accelerate the rate of fibrillation for ?-LA and the three fragments at pH 2.0. At pH 7.4, ?-LA is not able to form fibrils both in the absence and in the presence of OA. Also the three fragments did not form amyloid fibrils at neutral pH, while in the presence of OA they underwent a conformational change to ?-sheet structure, were able to bind ThT and to form aggregates with the typical amyloid morphology. During the PhD course, I also collaborated to an ongoing project of the laboratory at CRIBI, focused on the characterization of oligomeric species on the aggregation pathway of human lysozyme, which belongs to the so called “lysozyme/lactalbumin superfamily”. Soluble oligomers of lysozyme were produced at low pH and high temperature, and then analyzed by a range of techniques including binding to fluorescent probes, Fourier-transform infrared (FTIR) spectroscopy and limited proteolysis. Oligomers have solvent-exposed hydrophobic patches, and FTIR spectra are indicative of highly misfolded species. Moreover, the oligomeric lysozyme aggregates were found to be more susceptible to proteolysis than both the monomeric protein and the mature fibrils, indicating their lack of organized structure. This study showed that the soluble lysozyme oligomers are structurally flexible species present at low concentration during the initial phases of aggregation. The results of this study were accepted for publication in an international journal and the ‘in press’ manuscript is attached to the Thesis. During the third year of PhD, I spent a six months period at the University of Cambridge (UK), in the Cambridge Centre for Proteomics, under the supervision of Dr. Kathryn Lilley. The aim of this period was to learn proteomic methodologies for large scale identification of proteins, improving my background in mass spectrometry. I was involved in an ongoing project of the laboratory, focused on a parallel affinity purification method coupled to mass spectrometry for identifying proteins and their binding partners in Drosophila melanogaster embryos (Veraksa et al., 2005). Triple tagged proteins were generated (Spradling et al., 1999) and isolated from a variety of tissues in embryos. Affinity purification using two tags in parallel allowed the isolation of intact native protein complexes. The high sensitivity and high mass accuracy of the hybrid LTQ-Orbitrap instrument ensured maximal coverage of low abundance complex components, generating high confident data with low false discovery rates. The software ProteinCenter™ (Proxeon) was utilized for the visualization and the statistical comparison of the datasets. The data were compared with datasets published in public databases, which validate the data and increased the certainty of the detected interaction sets. Also novel protein interactions not previously reported were mapped. These high confidence in vivo protein datasets add high confidence data to the currently incomplete D. melanogaster proteome and interactome. Here, the results of five different size proteins from different cellular localizations are reported in detail to show the workflow and the efficiency of the methodology. Summing up, this PhD Thesis is composed of a major part dealing with the characterization of the interaction of ?-LA and oleic acid and the effects on the protein aggregation, and a minor part dealing with the mass spectrometry analysis of Drosophila protein complexes, besides the publication on the oligomeric species in the aggregation pathway of lysozyme.
L’argomento di questa Tesi di Dottorato riguarda in generale aspetti del problema del folding e unfolding proteico, in linea con la tematica di ricerca condotta nel laboratorio di Chimica delle Proteine al CRIBI, dove le attività sono state per la maggior parte svolte. Il meccanismo di acquisizione della struttura tridimensionale di una proteina (folding) è un evento biologico importante. In generale, è un processo multi-stadio che coinvolge la formazione di intermedi, che sono stati parzialmente strutturati contenenti alcune caratteristiche strutturali della proteina nativa, ma non le interazioni finali tra le catene laterali che permettono alla proteina di esercitare la sua funzione specifica. Il fallimento nell’acquisizione del corretto folding (misfolding) può causare aggregazione proteica. Infatti, proteine parzialmente strutturate possono facilmente auto-assemblarsi in aggregati regolari insolubili (fibrille amiloidi), associati a gravi malattie, le cosiddette amiloidosi (Chiti & Dobson, 2006). In particolare, il mio progetto di Dottorato è focalizzato sull’?-lattalbumina (?-LA), una metallo-proteina del latte, ampiamente utilizzata come modello di studio del folding proteico. Nell’ultimo decennio, questa proteina ha suscitato un nuovo interesse per la scoperta di una variante che, oltre al ruolo fisiologico nella lattazione, è in grado di indurre apoptosi nelle cellule tumorali (Svensson et al., 1999). L’attività citotossica risiede nella formazione di un complesso con un acido grasso del latte, l’acido oleico (OA), denominato HAMLET (Human Alpha-lactalbumin Made LEthal to Tumor cells) (Svensson et al., 2000). Gli eventi cellulari causati da HAMLET sono stati estesamente studiati (Kohler et al., 2001; Duringer et al., 2003). Al contrario, il meccanismo e la natura dell’interazione tra ?-LA e OA e le proprietà fisico-chimiche del complesso non sono ancora completamente chiariti. Infatti, nonostante l’effettiva associazione dell’OA con la proteina in soluzione (Polverino de Laureto et al., 2002), si ritiene che il complesso così ottenuto possieda attività antitumorale minore dell’HAMLET, preparato secondo una procedura cromatografica (Svensson et al., 2000). Altre questioni dibattute riguardano la stechiometria proteina/acido grasso e lo stato monomerico/oligomerico dell’?-LA nel complesso. Inoltre, manca un’indagine sistematica dell’effetto della proteina sul comportamento di fase dell’OA. Lo scopo di questa Tesi di Dottorato è lo studio della capacità di legare OA ed esprimere attività citotossica di tre derivati proteolitici di ?-LA bovina, ottenuti per proteolisi limitata con pepsina a pH acido. L’uso di un approccio di proteolytic dissection di una proteina è stato largamente impiegato per studiare il folding di molte proteine (Wetlaufer, 1981; Gegg et al., 1997; Llinás & Marqusee, 1998). La caratterizzazione di frammenti proteici, contenenti elementi strutturali specifici della proteina intera (?-elica, ?-sheet) e in grado di assumere struttura in modo autonomo, è stata usata con successo per chiarire caratteristiche strutturali di ?-LA (Polverino de Laureto et al., 1999; 2001). Infatti, frammenti corrispondenti a domini strutturali in proteine multi-dominio hanno mostrato in alcuni casi la capacità di acquisire in soluzione una conformazione simile a quella nativa (Fontana et al., 2004). I frammenti di ?-LA studiati sono: la specie 1–40/53–123, priva di parte del dominio ? della?proteina nativa; 1–40/104–123, formato dal frammento N-terminale 1-40 legato covalentemente al frammento C-terminale 104-123 mediante due ponti disolfuro e contenente tre delle quattro ?-eliche di ?-LA; 53-103, contenente l’elica C e il sito di legame al calcio (Polverino de Laureto et al., 1999; 2001). Le differenze conformazionali dei tre frammenti sono state utilizzate come razionale per studiare la loro efficacia di legame all’OA, e per approfondire quindi il meccanismo di questo legame. Questo studio ha anche rilevanza fisiologica: la pepsina è un enzima dello stomaco e quindi queste specie potrebbero essere generate in vivo, nelle stesse condizioni di pH acido in cui si è ipotizzata la formazione del complesso HAMLET, contribuendo così all’attività apoptotica del complesso formato dalla proteina intera. I complessi dei tre frammenti con OA sono stati preparati prima seguendo la procedura cromatografica descritta per l’HAMLET, poi per diretto miscelamento in soluzione dei due componenti. Le proprietà conformazionali dei complessi sono state caratterizzate mediante dicroismo circolare, mostrando che i complessi preparati attraverso entrambe le procedure presentano conformazioni simili e acquisizione di ?-elica. Inoltre, è stato valutato l’effetto del calcio sulla conformazione dei complessi mediante dicroismo circolare. La spettroscopia di fluorescenza è stata utilizzata per analizzare il coinvolgimento dei residui di Trp nell’interazione con OA. Inoltre, è stata studiata la capacità dei complessi di indurre morte cellulare per apoptosi, in linea con l’attività citotossica mostrata dall’HAMLET. Per analizzare lo stato fisico dell’OA coinvolto nell’interazione con l’?-LA, il comportamento di aggregazione di OA è stato studiato con diverse tecniche, quali microscopia elettronica a trasmissione (TEM), titolazione con un colorante fluorescente e analisi turbidimetriche, ed è stato osservato un effetto di solubilizzazione dell’OA. Questi risultati hanno permesso di preparare un articolo che sarà presto spedito ad una rivista internazionale e che è allegato alla Tesi. Nella seconda parte del Dottorato, la ricerca è stata focalizzata sulla tendenza di ?-LA e dei suoi tre frammenti proteolitici ad aggregare. ?-LA è in grado di formare fibrille amiloidi in vitro, sebbene non sia associata a patologie. Fibrille di ?-LA si producono quando la struttura della proteina è parzialmente destabilizzata, ad esempio a pH acido, per riduzione di tre ponti disolfuro a pH neutro (Goers et al., 2002), o per taglio proteolitico (Polverino de Laureto et al., 2005). I lipidi possono agire come efficaci catalizzatori della fibrillogenesi, creando un ambiente in cui le molecole proteiche adottano una conformazione e un’orientazione che promuove il loro assemblaggio in strutture fibrillari (Thirumalai et al., 2003; Stefani, 2004; Sparr et al., 2004; Zhao et al., 2004). In particolare, poiché è noto che le interazioni proteina-acido grasso modulano il processo di fibrillogenesi (Kim & Takahashi 2006), i processi di aggregazione di ?-LA e dei suoi frammenti sono stati studiati per capire se e come l’OA possa influenzare la formazione di fibrille. In primo luogo, l’aggregazione è stata seguita a pH 2.0, in parallelo con il già noto comportamento fibrillogenico di ?-LA intera. In secondo luogo, l’aggregazione è stata studiata a pH 7.4, poiché in questa condizione fisiologica nella prima parte della Tesi sono stati studiati i cambiamenti conformazionali indotti dall’OA sulla struttura dei frammenti e la loro attività citotossica. La formazione di fibrille è stata seguita mediante saggi di fluorescenza con la tioflavina T (ThT), dicroismo circolare e TEM. Tutti e tre i frammenti sono in grado di produrre fibrille amiloidi a pH 2.0, analogamente all’?-LA. Ai valori di concentrazione proteica e rapporto proteina/lipide utilizzati, OA sembra accelerare la velocità di formazione di fibrille di ?-LA e dei frammenti a pH 2.0. A pH 7.4, ?-LA non forma fibrille amiloidi sia in assenza sia in presenza di OA. Anche i tre frammenti non sono stati in grado di formare fibrille a pH neutro, mentre in presenza di OA hanno mostrato un cambiamento conformazionale verso una struttura ?-sheet, hanno legato ThT e formato aggregati con la tipica morfologia amiloide. Durante il Dottorato, ho inoltre collaborato ad un progetto in corso nel laboratorio al CRIBI, relativo alla caratterizzazione di specie oligomeriche nel processo di aggregazione del lisozima umano, che appartiene alla cosiddetta “superfamiglia lisozima/lattalbumina”. Oligomeri solubili di lisozima sono stati prodotti a pH acido e alta temperatura, e quindi analizzati con varie tecniche, quali legame a molecole fluorescenti, spettroscopia infrarosso in trasformata di Fourier (FTIR) e proteolisi limitata. Gli oligomeri presentano superfici idrofobiche esposte al solvente, e gli spettri FTIR sono indicativi di specie altamente destrutturate. Inoltre, gli aggregati oligomerici di lisozima si sono rivelati più suscettibili alla proteolisi rispetto sia alla proteina monomerica sia alle fibrille mature, indicando la mancanza di una struttura organizzata. Questo studio ha dimostrato che gli oligomeri solubili di lisozima sono specie strutturalmente flessibili presenti a bassa concentrazione durante le fasi iniziali dell’aggregazione. I risultati di questo studio sono stati accettati per la pubblicazione in una rivista internazionale e il manoscritto in press è allegato alla Tesi. Durante il terzo anno di Dottorato, ho trascorso un periodo di sei mesi all’Università di Cambridge (UK), presso il Cambridge Centre for Proteomics, sotto la supervisione della Dr.ssa Kathryn Lilley. L’obiettivo di questo periodo è stato di apprendere metodologie di proteomica per l’identificazione di proteine su larga scala, ottimizzando le mie conoscenze di spettrometria di massa. Ho collaborato ad un progetto in corso nel laboratorio, incentrato su un metodo di purificazione di affinità in parallelo accoppiata a spettrometria di massa per identificare complessi proteici in embrioni di Drosophila melanogaster (Veraksa et al., 2005). Proteine con tre tags sono state generate (Spradling et al., 1999) e isolate da vari tessuti embrionali. La purificazione di affinità usando due tags in parallelo ha permesso di isolare complessi proteici nativi intatti. L’alta sensibilità e l’alta accuratezza di massa dello strumento ibrido LTQ-Orbitrap ha assicurato la massima copertura di componenti di complessi poco abbondanti, generando dati ad alta confidenza con basse false discovery rates. È stato utilizzato il software ProteinCenter™ (Proxeon) per la visualizzazione e l’analisi statistica dei set di dati. I risultati sono stati confrontati con dati presenti in database pubblici, confermandone la validità e aumentando l’autenticità delle interazioni individuate. Sono state inoltre mappate nuove interazioni proteiche non riportate in precedenza. Questi set di dati in vivo aggiungono alta confidenza al proteoma e ‘interattoma’ di D. melanogaster attualmente incompleto. In questa Tesi sono riportati in dettaglio i risultati di cinque proteine di diverse dimensioni e localizzazione cellulare, per presentare la procedura e l’efficienza della metodologia. In sintesi, questa Tesi di Dottorato è composta da una parte principale che riguarda la caratterizzazione dell’interazione tra ?-LA e acido oleico e gli effetti sull’aggregazione proteica, e una parte minore relativa all’analisi di spettrometria di massa di complessi proteici in Drosophila, oltre alla pubblicazione sulle specie oligomeriche studiate nel processo di aggregazione del lisozima.