Browsing by Issue Date, starting with "2025-02-27"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
- From production to purification: Towards an integrative process for recombinant pre-miRNA-29b as biopharmaceuticalPublication . Carapito, Ana Rita Mugeiro; Sousa, Fani Pereira de; Martins, Mara Guadalupe Freire; Sponchioni, MattiaRecent advances in RNA research have greatly demonstrated the potential of RNA-based therapies, offering innovative ways to target a variety of diseases with enhanced specificity. Unlike traditional small-molecule drugs, RNA therapeutics like small-interfering RNA and microRNA (miRNA) can precisely regulate gene expression and target specific biological pathways. For instance, recent studies demonstrate that miRNA-29 regulates some pathological routes associated with Alzheimer's disease (AD), a neurodegenerative disorder affecting millions of people around the world. MiRNA-29 plays a crucial role in processes like amyloid-β peptides (Aβ) formation, which contributes to memory loss and neuronal cell damage. Low levels of miRNA-29 are linked to increased production and activity of the enzyme β-secretase (BACE1), which leads to higher production of Aβ and, consequently, β-plaque formation. Given its important functions, restoring or increasing miRNA-29 levels in AD patients can be a promising strategy for AD treatment. Current research is investigating the use of recombinant miRNA-29b precursor (pre-miRNA-29b) to silence BACE1 expression and decrease Aβ levels, aiming to develop novel approaches to slow the progression of AD. Given this, biopharmaceuticals production and subsequent purification constitute an important process that needs to be in full accordance with criteria established by regulatory entities. Regarding RNA production, the standard technique is through chemical synthesis, however this strategy comes with some disadvantages, namely the biomolecules length that can be correctly produced and the low production levels. Recombinant production is the alternative method that is more cost-effective and applicable for large-scale production. Escherichia coli is the most widely used and studied host, nonetheless, Rhodovulum sulfidophilum (R. sulfidophilum) presents interesting characteristics considering nucleic acids production, that comprise the ability to secrete nucleic acids to the extracellular medium, without secreting RNases. The recovery of recombinantly produced nucleic acids from the extracellular medium might be a great advantage for the further downstream processing because the contaminants, such as cell debris and endotoxins, are not present, as usually are in the intracellular samples. The downstream process is the most expensive stage of the whole bioprocess, and usually, several chromatographic steps are required to achieve the intended purity and quality. Given this, there is a high demand for specific and efficient purification strategies. Multimodal chromatography is currently under thorough research as it can lead to the same specificity for target compounds as it is observed for affinity chromatography, without using biological ligands that greatly increase the process cost. Ionic Liquids (ILs) are molten salts that can present this multimodal character when used as ligands immobilized onto the stationary phase, being recently studied for nucleic acids purification. The different moieties of the cation in an IL can allow the exploitation of different types of interactions with the target molecule. Therefore, this Doctoral Thesis explores a promising and alternative recombinant host for the production of pre-miRNA-29b, as well as its purification using newly synthesized resins, aiming for the development of a whole bioprocess. Initially, a DNA vector was designed to produce the target pre-miRNA-29b in R. sulfidophilum. In this study, the impact of the plasmid on bacterial growth was analysed and compared to the non-transformed strain. The transformed strain has shown a global growth about 5 times lower than the non-transformed strain, but this did not impact negatively the target RNA production. An optimization of the extracellular extraction protocol was also conducted during this study, comparing a protocol with ethanol or isopropanol as precipitation agents. The results proven that the protocol using isopropanol as precipitation agent was more efficient reaching a concentration of 0.7 μg of pre-miRNA-29b per liter of medium. After successfully developing an efficient pre-miRNA-29b production system, it became important to develop a purification strategy. For this, four different silica-based Ionic Liquids (SILs) were synthesized and evaluated on their ability for nucleic acids separation. An initial screening of binding and elution conditions with a low molecular weight RNA sample was performed by using both ionic and hydrophobic conditions, to select the more promising support for further purification assays. The support SSi[C3C3NH2Im]Cl was selected, proving to be highly efficient in separating different species of DNA (genomic and plasmidic) from RNA. Then, it became important to test this support regarding its ability for pre-miRNA-29b purification from other types of RNAs. Given that this approach is more challenging due to the high physical and chemical similarity among small RNAs, four different ILs were employed in this stage to act as competition agents aiming to enhance target selectivity. From the four tested ILs, the 1-ethylimidazolium chloride was proven to have a higher impact on the pre-miRNA-29b selectivity improvement, achieving 76% purity. Since the aim is to develop a method suitable for a biopharmaceutical production and purification, it would be needed to achieve higher purity levels than what was verified with SSi[C3C3NH2Im]Cl. Therefore, a commercial multimodal resin, Capto Q ImpRes was chosen to analyse and compare its performance to the previous newly synthesized support. Both intra- and extracellular RNA samples were tested regarding pre-miRNA-29b purification, and a Design of Experiments (DoE) was established for each sample. Screening assays with intracellular RNA sample were made to identify the factors to be implemented in the DoE. Sodium chloride concentration and pH were defined, and it was possible to establish conditions that uncover a balance between the recovery and purity of the target. For intracellular RNA samples, the DoE effectively identified optimal purification parameters, attaining recovery rates up to 73% and purity levels around 78%. However, these two responses shown to be almost inversely proportional. Nonetheless, it was possible to achieve the optimal point with a recovery of 48.21% and purity of 51.15%. On the other hand, purifying extracellular pre-miRNA-29b presented substantial challenges, exhibiting significant inconsistencies in purity and difficulties in identifying the target molecule, probably due to the low concentration of pre-miRNA in the extracellular extract and the difficulty in detection. To search for an alternative that could lead to an improved selectivity for the biopharmaceutical under study, a set of several oligonucleotides were designed to interact with the target. Initially, 13 oligonucleotides were designed to interact through base complementarity with different regions of the target. Each oligonucleotide was linked to a carbon chain (6 or 12 carbons) with an amino group, resulting in 26 different ligands. After data analysis, some ligands showed higher specificity for their target sites, while others demonstrated versatility in recognizing multiple sites. Based on this, 4 oligonucleotides were identified as the most promising for further experimental testing. Overall, in this doctoral thesis, a bioprocess was developed starting with the upstream stage with recombinant production of the pre-miRNA-29b, followed by the downstream stage exploiting different approaches applicable to target purification. The work opens the route for the development of an integrated production and purification process for pre-miRNA-29b, with great potential for large-scale translation.
- Towards Code Smell Detection: A Multiphase Methodology Using Fuzzy Logic-based Feature SelectionPublication . Monteiro, Bruno Miguel Gonçalves; Pombo, Nuno Gonçalo Coelho CostaEnsuring the quality of software code remains a fundamental challenge in software engineering. Code smells are characteristics in the source code that could infer a need for refactoring and imply potential problems with its design and implementation, making the code harder to understand, maintain, and evolve. If left unresolved, these smells can lead to the appearance of bugs, making them as critical to address as software defects themselves. Recent research has explored various Machine Learning (ML) techniques for detecting code smells. However, misclassification continues to be a significant challenge, often leading to inconsistent outcomes and compromising the reliability of the model’s predictions. Feature selection has emerged as a key technique to improve the accuracy of these models by reducing dimensionality and removing irrelevant or redundant features, thereby allowing for more precise predictions in code smell detection. Among these, the feature selection method based on Modified Fuzzy C-Means algorithm with supervision (MFCMS), which integrates both clustering and supervised learning, shows significant promise. By capturing complex relationships between features, MFCMS has the potential to improve the reliability of code smell detection, leading to more consistent predictions. This dissertation introduces a multiphase approach to enhance the detection of the long method smell using ML techniques. The method integrates data preprocessing, feature extraction, and machine learning modeling to improve both accuracy and efficiency. Although experimental results revealed that the proposed approach provides some improvements over the baseline model, its main contribution lies in applying MFCMS for dimensionality reduction, especially when combined with PCA. This combination reduced feature correlation and enhanced the feature selection process. Despite these advancements, some misclassifications remain, indicating that further refinement is needed. This work lays the foundation for future research, particularly in exploring feature selection methods for code smell detection. It unlocks the potential for developing more robust, automated tools that can improve software quality and maintainability.
- Circulating microRNA biomarkers for lung cancer detectionPublication . Pereira, Maurício Barbosa; Cruz, Carla Patrícia Alves Freire Madeira daOnce considered a rarity, lung cancer surged to the forefront as the leading cause of cancerrelated deaths, quickly becoming one of the most pressing public health challenges of the modern era. Its lethal reputation stems from its aggressive progressive nature and the frequent delay in diagnosis until advanced stages, where treatment efficacy diminishes, leaving patients with poorer outcomes and a heightened risk of recurrence. This glaringly escalating issue highlights the need for innovative approaches in both diagnosis and treatment. One promising frontier in lung cancer research is the exploration of microRNAs, small noncoding RNAs with the potential to revolutionise diagnostics and therapy. Let-7e-5p, a member of the Let-7 family, has emerged as an undeniably promising biomarker and therapeutic target due to its role in regulating gene expression and influencing cancer progression. In this study Let-7e-5p expression was quantified by RT-qPCR in lung cancer cell lines (A549 and H1299) and peripheral blood mononuclear cells (PBMCs), revealing consistent underexpression. Additionally, cell viability assays (MTT) assessed the effects of natural mineral waters (NMWs), while the influence of TMPyP4, a G-quadruplex ligand, and NMWs on miRNA regulation was tested through specific treatments and subsequent quantification of its expression. TMPyP4 induced a time-dependent increase in Let-7e-5p expression, whereas NMWs, in contrast, appeared to promote cellular stress rather than enhancing basal miRNA levels. Collectively, these results establish Let-7e-5p as a pivotal biomarker and a compelling target for therapeutic innovation in lung cancer. While challenges remain in fully harnessing its potential, these findings open the door to future explorations of Let-7e-5p, offering a promising glimpse into a future where Let-7e-5p could play an impactful role in reshaping strategies against this relentless disease.
- A relação entre os maus-tratos infantis e as estratégias de regulação emocional em adultos: uma Revisão Sistemática da LiteraturaPublication . Morais, Ana Cláudia Pereira; Ramos, Ludovina Maria de AlmeidaOs maus-tratos infantis afetam diversas áreas da vida de uma criança, assim como impactam o seu desenvolvimento normativo. A presente revisão sistemática da literatura tem como objetivo compreender a relação entre maus-tratos infantis e as estratégias de regulação emocional em adultos. Para conduzir a revisão, utilizou-se o modelo PRISMA (Preferred Reporting Items for Systematic Reviews and MetaAnalyses). A pesquisa do artigos científicos foi realizada nas bases de dados Scopus, PubMed e Web of Science, com um intervalo de tempo entre 2019 e 2024. No total foram incluídos 12 artigos científicos. Em termos gerais, os resultados suportam a existência de uma associação positiva entre maus-tratos infantis e adoção de estratégias de regulação emocional desadaptativas, como a supressão e a ruminação. Além disto, os maus-tratos infantis foram também associados a maiores dificuldades de regulação emocional e menor uso de estratégias adaptativas, como a reavaliação cognitiva. Salienta-se ainda a importância do papel dos cuidadores no desenvolvimento emocional das crianças.