Each one teaches one: characterizing active forms of proteins by single molecule force spectroscopy

This Ph.D. candidate thesis collects the research work I conducted under the supervision of Prof.Bruno Samor´ı in 2005,2006 and 2007. Some parts of this work included in the Part III have been begun by myself during my undergraduate thesis in the same laboratory and then completed during the initial part of my Ph.D. thesis: the whole results have been included for the sake of understanding and completeness. During my graduate studies I worked on two very different protein systems. The theorical trait d’union between these studies, at the biological level, is the acknowledgement that protein biophysical and structural studies must, in many cases, take into account the dynamical states of protein conformational equilibria and of local physico-chemical conditions where the system studied actually performs its function. This is introducted in the introductory part in Chapter 2. Two different examples of this are presented: the structural significance deriving from the action of mechanical forces in vivo (Chapter 3) and the complexity of conformational equilibria in intrinsically unstructured proteins and amyloid formation (Chapter 4). My experimental work investigated both these examples by using in both cases the single molecule force spectroscopy technique (described in Chapter 5 and Chapter 6). The work conducted on angiostatin focused on the characterization of the relationships between the mechanochemical properties and the mechanism of action of the angiostatin protein, and most importantly their intertwining with the further layer of complexity due to disulfide redox equilibria (Part III). These studies were accompanied concurrently by the elaboration of a theorical model for a novel signalling pathway that may be relevant in the extracellular space, detailed in Chapter 7.2. The work conducted on -synuclein (Part IV) instead brought a whole new twist to the single molecule force spectroscopy methodology, applying it as a structural technique to elucidate the conformational equilibria present in intrinsically unstructured proteins. These equilibria are of utmost interest from a biophysical point of view, but most importantly because of their direct relationship with amyloid aggregation and, consequently, the aetiology of relevant pathologies like Parkinson’s disease. The work characterized, for the first time, conformational equilibria in an intrinsically unstructured protein at the single molecule level and, again for the first time, identified a monomeric folded conformation that is correlated with conditions leading to -synuclein and, ultimately, Parkinson’s disease. Also, during the research work, I found myself in the need of a generalpurpose data analysis application for single molecule force spectroscopy data analysis that could solve some common logistic and data analysis problems that are common in this technique. I developed an application that addresses some of these problems, herein presented (Part V), and that aims to be publicly released soon.

Neuroprotective activity of guanosine in an in vitro model of Alzheimer's disease

The β-Amyloid (βA) peptide is the major component of senile plaques that are one of the hallmarks of Alzheimer’s Disease (AD). It is well recognized that Aβ exists in multiple assembly states, such as soluble oligomers or insoluble fibrils, which affect neuronal viability and may contribute to disease progression. In particular, common βA-neurotoxic mechanisms are Ca2+ dyshomeostasis, reactive oxygen species (ROS) formation, altered signaling, mitochondrial dysfunction and neuronal death such as necrosis and apoptosis. Recent study shows that the ubiquitin-proteasome pathway play a crucial role in the degradation of short-lived and regulatory proteins that are important in a variety of basic and pathological cellular processes including apoptosis. Guanosine (Guo) is a purine nucleoside present extracellularly in brain that shows a spectrum of biological activities, both under physiological and pathological conditions. Recently it has become recognized that both neurons and glia also release guanine-based purines. However, the role of Guo in AD is still not well established. In this study, we investigated the machanism basis of neuroprotective effects of GUO against Aβ peptide-induced toxicity in neuronal (SH-SY5Y), in terms of mitochondrial dysfunction and translocation of phosphatidylserine (PS), a marker of apoptosis, using MTT and Annexin-V assay, respectively. In particular, treatment of SH-SY5Y cells with GUO (12,5-75 μM) in presence of monomeric βA25-35 (neurotoxic core of Aβ), oligomeric and fibrillar βA1-42 peptides showed a strong dose-dependent inhibitory effects on βA-induced toxic events. The maximum inhibition of mitochondrial function loss and PS translocation was observed with 75 μM of Guo. Subsequently, to investigate whether neuroprotection of Guo can be ascribed to its ability to modulate proteasome activity levels, we used lactacystin, a specific inhibitor of proteasome. We found that the antiapoptotic effects of Guo were completely abolished by lactacystin. To rule out the possibility that this effects resulted from an increase in proteasome activity by Guo, the chymotrypsin-like activity was assessed employing the fluorogenic substrate Z-LLL-AMC. The treatment of SH-SY5Y with Guo (75 μM for 0-6 h) induced a strong increase, in a time-dependent manner, of proteasome activity. In parallel, no increase of ubiquitinated protein levels was observed at similar experimental conditions adopted. We then evaluated an involvement of anti and pro-apoptotic proteins such as Bcl-2, Bad and Bax by western blot analysis. Interestingly, Bax levels decreased after 2 h treatment of SH-SY5Y with Guo. Taken together, these results demonstrate that Guo neuroprotective effects against βA-induced apoptosis are mediated, at least partly, via proteasome activation. In particular, these findings suggest a novel neuroprotective pathway mediated by Guo, which involves a rapid degradation of pro-apoptotic proteins by the proteasome. In conclusion, the present data, raise the possibility that Guo could be used as an agent for the treatment of AD.

Machine learning methods for prediction of disulphide bonding states of cysteine residues in proteins

The goal of this thesis work is to develop a computational method based on machine learning techniques for predicting disulfide-bonding states of cysteine residues in proteins, which is a sub-problem of a bigger and yet unsolved problem of protein structure prediction. Improvement in the prediction of disulfide bonding states of cysteine residues will help in putting a constraint in the three dimensional (3D) space of the respective protein structure, and thus will eventually help in the prediction of 3D structure of proteins. Results of this work will have direct implications in site-directed mutational studies of proteins, proteins engineering and the problem of protein folding. We have used a combination of Artificial Neural Network (ANN) and Hidden Markov Model (HMM), the so-called Hidden Neural Network (HNN) as a machine learning technique to develop our prediction method. By using different global and local features of proteins (specifically profiles, parity of cysteine residues, average cysteine conservation, correlated mutation, sub-cellular localization, and signal peptide) as inputs and considering Eukaryotes and Prokaryotes separately we have reached to a remarkable accuracy of 94% on cysteine basis for both Eukaryotic and Prokaryotic datasets, and an accuracy of 90% and 93% on protein basis for Eukaryotic dataset and Prokaryotic dataset respectively. These accuracies are best so far ever reached by any existing prediction methods, and thus our prediction method has outperformed all the previously developed approaches and therefore is more reliable. Most interesting part of this thesis work is the differences in the prediction performances of Eukaryotes and Prokaryotes at the basic level of input coding when ‘profile’ information was given as input to our prediction method. And one of the reasons for this we discover is the difference in the amino acid composition of the local environment of bonded and free cysteine residues in Eukaryotes and Prokaryotes. Eukaryotic bonded cysteine examples have a ‘symmetric-cysteine-rich’ environment, where as Prokaryotic bonded examples lack it.

Pyrolysis of peptides and proteins. Analytical study and environmental applications

Analytical pyrolysis was used to investigate the formation of diketopiperazines (DKPs) which are cyclic dipeptides formed from the thermal degradation of proteins. A quali/quantitative procedure was developed combining microscale flash pyrolysis at 500 °C with gas chromatography-mass spectrometry (GC-MS) of DKPs trapped onto an adsorbent phase. Polar DKPs were silylated prior to GC-MS. Particular attention was paid to the identification of proline (Pro) containing DKPs due to their greater facility of formation. The GC-MS characteristics of more than 80 original and silylated DKPs were collected from the pyrolysis of sixteen linear dipeptides and four model proteins (e.g. bovine serum albumin, BSA). The structure of a novel DKP, cyclo(pyroglutamic-Pro) was established by NMR and ESI-MS analysis, while the structures of other novel DKPs remained tentative. DKPs resulted rather specific markers of amino acid sequence in proteins, even though the thermal degradation of DKPs should be taken into account. Structural information of DKPs gathered from the pyrolysis of model compounds was employed to the identification of these compounds in the pyrolysate of proteinaceous samples, including intrinsecally unfolded protein (IUP). Analysis of the liquid fraction (bio-oil) obtained from the pyrolysis of microalgae Nannochloropsis gaditana, Scenedesmus spp with a bench scale reactor showed that DKPs constituted an important pool of nitrogen-containing compounds. Conversely, the level of DKPs was rather low in the bio-oil of Botryococcus braunii. The developed micropyrolysis procedure was applied in combination with thermogravimetry (TGA) and infrared spectroscopy (FT-IR) to investigate surface interaction between BSA and synthetic chrysotile. The results showed that the thermal behavior of BSA (e.g. DKPs formation) was affected by the different form of doped synthetic chrysotile. The typical DKPs evolved from collagen were quantified in the pyrolysates of archaeological bones from Vicenne Necropolis in order to evaluate their conservation status in combination with TGA, FTIR and XRD analysis.

Machine-learning methods for structure prediction of β-barrel membrane proteins

Different types of proteins exist with diverse functions that are essential for living organisms. An important class of proteins is represented by transmembrane proteins which are specifically designed to be inserted into biological membranes and devised to perform very important functions in the cell such as cell communication and active transport across the membrane. Transmembrane β-barrels (TMBBs) are a sub-class of membrane proteins largely under-represented in structure databases because of the extreme difficulty in experimental structure determination. For this reason, computational tools that are able to predict the structure of TMBBs are needed. In this thesis, two computational problems related to TMBBs were addressed: the detection of TMBBs in large datasets of proteins and the prediction of the topology of TMBB proteins. Firstly, a method for TMBB detection was presented based on a novel neural network framework for variable-length sequence classification. The proposed approach was validated on a non-redundant dataset of proteins. Furthermore, we carried-out genome-wide detection using the entire Escherichia coli proteome. In both experiments, the method significantly outperformed other existing state-of-the-art approaches, reaching very high PPV (92%) and MCC (0.82). Secondly, a method was also introduced for TMBB topology prediction. The proposed approach is based on grammatical modelling and probabilistic discriminative models for sequence data labeling. The method was evaluated using a newly generated dataset of 38 TMBB proteins obtained from high-resolution data in the PDB. Results have shown that the model is able to correctly predict topologies of 25 out of 38 protein chains in the dataset. When tested on previously released datasets, the performances of the proposed approach were measured as comparable or superior to the current state-of-the-art of TMBB topology prediction.

Conjugation of Toll Like Receptor 7 agonist through conjugation to immunogenic proteins. Synthesis, characterization, formulation and pre-clinical evaluation

The next generation of vaccine adjuvant are represented by a wide ranging set of molecules called Toll like agonists (TLR’s). Although many of these molecules are complex structures extracted from microorganisms, small molecule TLR agonists have also been identified. However, delivery systems have not been optimized to allow their effective delivery in conjunction with antigens. Here we describe a novel approach in which a small molecule TLR agonist has been conjugated directly to antigens to ensure effective co delivery. We describe the conjugation of a relevant protein, a recombinant protective antigen from S.pneumoniae (RrgB), which is linked to a TLR7 agonist. Following thorough characterization to ensure there was no aggregation, the conjugate was evaluated in a murine infection model. Results showed that the conjugate extended animals’ survival after lethal challenge with S.pneumoniae. Comparable results were obtained with a 10 fold lower dose than that of the native unconjugated antigen. Notably, the animals immunized with the same dose of unconjugated TLR7 agonist and antigen showed no adjuvant effect. The increased immunogenicity was likely a consequence of the co-localization of TLR7 agonist and antigen by chemical binding and is was more effective than simple co-administration. Likely, this approach can be adopted to reduce the dose of antigen required to induce protective immunity, and potentially increase the safety of a broad variety of vaccine candidates

Machine learning tools for protein annotation: the cases of transmembrane β-barrel and myristoylated proteins

Biology is now a “Big Data Science” thanks to technological advancements allowing the characterization of the whole macromolecular content of a cell or a collection of cells. This opens interesting perspectives, but only a small portion of this data may be experimentally characterized. From this derives the demand of accurate and efficient computational tools for automatic annotation of biological molecules. This is even more true when dealing with membrane proteins, on which my research project is focused leading to the development of two machine learning-based methods: BetAware-Deep and SVMyr. BetAware-Deep is a tool for the detection and topology prediction of transmembrane beta-barrel proteins found in Gram-negative bacteria. These proteins are involved in many biological processes and primary candidates as drug targets. BetAware-Deep exploits the combination of a deep learning framework (bidirectional long short-term memory) and a probabilistic graphical model (grammatical-restrained hidden conditional random field). Moreover, it introduced a modified formulation of the hydrophobic moment, designed to include the evolutionary information. BetAware-Deep outperformed all the available methods in topology prediction and reported high scores in the detection task. Glycine myristoylation in Eukaryotes is the binding of a myristic acid on an N-terminal glycine. SVMyr is a fast method based on support vector machines designed to predict this modification in dataset of proteomic scale. It uses as input octapeptides and exploits computational scores derived from experimental examples and mean physicochemical features. SVMyr outperformed all the available methods for co-translational myristoylation prediction. In addition, it allows (as a unique feature) the prediction of post-translational myristoylation. Both the tools here described are designed having in mind best practices for the development of machine learning-based tools outlined by the bioinformatics community. Moreover, they are made available via user-friendly web servers. All this make them valuable tools for filling the gap between sequential and annotated data.

Integrating omics in prion diseases as a model to explore the strain paradigm in neurodegenerative diseases

This project aims at deepening the understanding of the molecular basis of the phenotypic heterogeneity of prion diseases. Prion diseases represent the first and clearest example of “protein misfolding diseases”, that are all the neurodegenerative diseases caused by the accumulation of misfolded proteins in the central nervous system. In the field of protein misfolding diseases, the term “strain” describes the heterogeneity observed among the same disease in the clinical and pathologic progression, biochemical features of the aggregated protein, conformational memory and pattern of lesions. In this work, the two most common strains of Creutzfeldt-Jakob Disease (CJD), named MM1 and VV2, were analyzed. This thesis investigates the strain paradigm with the production of new multi omic data, and, on such data, appropriate computational analysis combining bioinformatics, data science and statistical approaches was performed. In this work, genomic and transcriptomic profiling allowed an improved characterization of the molecular features of the two most common strains of CJD, identifying multiple possible genetic contributors to the disease and finding several shared impaired pathways between the VV2 strain and Parkinson Disease. On the epigenomic level, the tridimensional chromatin folding in peripheral immune cells of CJD patients at onset and of healthy controls was investigated with Hi-C. While being the first application of this very advanced technology in prion diseases and one of the first in general in neurobiology, this work found a significant and diffuse loss of genomic interactions in immune cells of CJD patients at disease onset, particularly in the PRNP locus, suggesting a possible impairment of chromatin conformation in the disease. The results of this project represent a novelty in the state of the art in this field, both from a biomedical and technological point of view.

Evaluation of the transcriptomic response of an Alzheimer's disease murine model induced at different ages: searching for predictors

Alzheimer’s disease (AD) is a chronic, progressive neurodegenerative disease, characterized by the impairment of mnesic and cognitive functions, that represents the most frequent type of dementia in older people worldwide. Aging is the most important risk factor for the sporadic form of the pathology and it is associated to the progressive impairment of the proteostasis network. The endoplasmic reticulum (ER), the main cellular actor involved in proteostasis, appears significantly compromised in AD due to the accumulation of β-amyloid (Aβ) protein and phosphorylated-tau protein. Increasing proteins misfolding activates a specific cellular response known as Unfolded Protein response (UPR) which orchestrates the recovery of ER function. The aim of the present study was to investigate the role of UPR and aging process in a murine model of AD induced by intracerebroventricular (i.c.v.) injection of Aβ1-42 oligomers at 3 or 18 months. The oligomers injection in aged animals caused the increased of memory impairment, oxidative stress, and the depletion of glutathione reserve. Furthermore, the RNA-sequencing analysis was performed and the bioinformatic analysis showed the enrichment of several pathways involved in neurodegeneration and protein regulations. The following analysis highlighted the significant dysregulation of the three branches of the UPR, the protein kinase RNA-like ER kinase (PERK), inositol-requiring protein 1α (IRE1α) and activating transcription factor 6 (ATF-6). In turn, ER stress affected the PI3K/Akt/Gsk3β and MAPK/ERK pathways, highlighting Mapkapk5 as a potential marker of the neurodegenerative process, which regulation could lead to the definition of new pharmacological and neuroprotective strategies to counteract AD.

A 3D biomimetic in vitro model: a novel strategy to investigate glioma biology

Gliomas are one of the most frequent primary malignant brain tumors. Acquisition of stem-like features likely contributes to the malignant nature of high-grade gliomas and may be responsible for the initiation, growth, and recurrence of these tumors. In this regard, although the traditional 2D cell culture system has been widely used in cancer research, it shows limitations in maintaining the stemness properties of cancer and in mimicking the in vivo microenvironment. In order to overcome these limitations, different three-dimensional (3D) culture systems have been developed to mimic better the tumor microenvironment. Cancer cells cultured in 3D structures may represent a more reliable in vitro model due to increased cell-cell and cell-extracellular matrix (ECM) interaction. Several attempts to recreate brain cancer tissue in vitro are described in literature. However, to date, it is still unclear which main characteristics the ideal model should reproduce. The overall goal of this project was the development of a 3D in vitro model able to reproduce the brain ECM microenvironment and to recapitulate pathological condition for the study of tumor stroma interactions, tumor invasion ability, and molecular phenotype of glioma cells. We performed an in silico bioinformatic analysis using GEPIA2 Software to compare the expression level of seven matrix protein in the LGG tumors with healthy tissues. Then, we carried out a FFPE retrospective study in order to evaluate the percentage of expression of selected proteins. Thus, we developed a 3D scaffold composed by Hyaluronic Acid and Collagen IV in a ratio of 50:50. We used two astrocytoma cell lines, HTB-12 and HTB-13. In conclusion, we developed an in vitro 3D model able to reproduce the composition of brain tumor ECM, demonstrating that it is a feasible platform to investigate the interaction between tumor cells and the matrix.