A virosfera em expansão: sobre a descoberta, genômica e proteômica de entidades virais codificadoras de um vasto arsenal de proteínas desconhecidas

• Ano de defesa: 2022
• Autor(a) principal: Paulo Victor de Miranda Boratto
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ICB - DEPARTAMENTO DE MICROBIOLOGIA; Programa de Pós-Graduação em Microbiologia; UFMG
• Palavras-chave em Português: vírus gigantes; tupanvírus; pandoravírus; yaravírus; Microbiologia; Virologia; Vírus gigantes; Vírus de DNA
• Link de acesso: http://hdl.handle.net/1843/55415; https://orcid.org/ 0000-0001-7531-888X

Recently, intense debates have sought to contribute in the way viruses are classified and the way their evolutionary process is interpreted, especially after the discovery of the giant viruses of amoeba. This process represented one of the most important phases of modern virology, beginning in 2003 with the discovery of Acanthamoeba polyphaga mimivirus (APMV). After, driven by major improvements in the techniques of viral isolation and cultivation in protists, a variety of amoeba viruses were found in different environments around the globe. Phylogeny analyzes and phylogenomic reconstructions ended up classifying these viruses into a potentially monophyletic group, involving the participation of many viral families. This group, called nucleocytoplasmic large DNA viruses (NCLDVs), is currently formed by members of the families Poxviridae, Asfarviridae, Iridoviridae, Ascoviridae, Phycodnaviridae, Marseilleviridae and Mimiviridae, in addition to the strains represented by the pithovirus, pandoravirus, mollivirus, medusavirus, pacmanvirus, faustovirus, klosneuvirus, among others. Although they potentially shared a common ancestor, members of the NCLDV have important morphogenetic differences that are in full expansion, given the increasing number of isolated and/or detected viral entities by recently incorporated cultivationindependent methods. In this work, our aim is to characterize samples of viruses from the amoeba Acanthamoeba castellanii, obtained from different Brazilian environments. For that, we promoted the genomic, proteomic and/or biological characterizations of three evolutive viral strains (tupanvirus, pandoravirus and yaravirus) isolated from this amoebal host. Proteomic analysis of tupanvirus particles revealed the presence of more than 100 proteins encoded by viral genes, a great part of them without known function or origin. In this context, about 20% of the proteins are encoded by genes that were transferred from cellular organisms to tupanviruses throughout their evolutionary history. For pandoraviruses, the present thesis has contributed to a better understanding of the multiplication cycle of the members of this lineage. Our data indicates that all the analyzed isolates are capable of deeply modify the cytoplasmic environment of the infected amoebae, recruiting mitochondria and membranes that are used in the formation of viral factories. Various patterns of pandoravirus particle morphogenesis were observed, and the assembly seemed to be started by the apex or the base of the particle. By counting the viral particles during the infection cycle, we observed that pandoraviruses can undergo exocytosis shortly after morphogenesis, in a process that involves intense recruitment of membranes surrounding the newly formed particles. Treating the infected cells with brefeldin, the exocytosis of viral particles was affected in two of the three analyzed strains, indicating variability between the isolates. Despite occurrence of exocytosis, lysis of host cells proved to be an important contributor to the viral release. Finally, regarding the analyzes made for the yaravirus, our data report the discovery of a viral lineage never described before, presenting a puzzling origin and phylogeny. With particles of approximately 80nm and a genome of about 45kbp, the yaravirus now represents the smallest and simplest (in terms of number of genes) virus isolated from Acanthamoeba spp. Interestingly, more than 90% of its genome does not show sequence similarity in databases. Our data also describe theidentification of a divergent major capsid protein, with no significant homology to capsid proteins seen in other NCLDV members, but with a predicted structure for the double-jelly roll domain. In addition, only distant-related ATPase homologs were found in similarity searches for sequences deposited in more than 8,500 public metagenomes, highlighting the rarity of this virus in the natural environment and demonstrating how important are studies based on viral isolation, not just those focused on genome detection. In conclusion, our work not only allows us to expand the current knowledge about the diversity of eukaryotic viruses, but also complements it with information that may challenge the current classification of the DNA viruses.