The human gut microbiome is associated with health outcomes including gastrointestinal and metabolic health, autoimmune disease and cancer. However, the role of the microbiome in many disease processes, including in the preterm gastrointestinal tract and female genital tract, has yet to be defined. Further, the diverse community of viruses within the microbiome (the virome) is understudied compared to bacteria. Here, I examine the microbiome and virome in specific disease models that are poorly understood: necrotizing enterocolitis (NEC), discordant HIV shedding in women living with HIV (WHLIV), female genital tract inflammation and gammaherpesvirus infection. Specifically, I examined the gut virome longitudinally in a cohort of preterm infants at risk for NEC; the female genital tract (FGT) microbiome and virome longitudinally in a cohort of WLHIV from Lima, Peru; the FGT virome in women from Phoenix, Arizona with differing levels of genital inflammation and different microbiome compositions; and the gut microbiome in murine gammaherpesvirus 68 (MHV68) infection. Further, I contributed to research responding to the spread of SARS-CoV-2 in Arizona. I found that 1) gut virome beta diversity decreased before NEC onset in preterm infants, suggesting a role for the virome in NEC; 2) FGT microbiome instability was associated with discordant HIV shedding, while FGT virome composition changed in association with ART duration and immune recovery; 3) FGT virome composition was associated with inflammation and microbiome composition; and 4) MHV68 infection outcomes were independent of microbiome perturbation, which may reflect environmental influences. The results of this research advance understanding of the microbiome and virome in these specific disease processes, and support further investigation of the microbiome and virome in preterm infant gastrointestinal health and FGT health, as well as environmental effects in microbiome research.

Z-DNA binding protein 1 (ZBP1) is an interferon-inducible protein that plays a crucial role in antiviral defense by recognizing Z-form nucleic acid (Z-NA), a left-handed conformer of double-stranded DNA/RNA. When ZBP1 binds to Z-NA, it can trigger programmed cell death pathways, including apoptosis and necroptosis, in collaboration with receptor interacting protein kinases 1 and 3 (RIPK1 and RIPK3). Z-NA positive viruses including poxviruses and influenza A virus (IAV) activate ZBP1-dependent cell death during replication. Little is known whether ZBP1 plays any role during Z-NA negative virus infection. Doxycycline-inducible A549 ACE2 Tet-On cells were constructed to express ZBP1 and were infected with Z-NA negative viruses. ZBP1-expressing cells infected with Sindbis virus (SINV), La Crosse virus (LACV), Vesicular stomatitis virus (VSV) and human coronavirus OC43 (hCoV-OC43) underwent extensive cell death, which could be rescued by a caspase inhibitor but not by JAK1/2 or RIPK1 kinase inhibitors. However, cell death was not observed upon Zika virus (ZIKV), Encephalomyocarditis virus (EMCV), Chikungunya virus (CHKV) or human coronavirus 229E (hCoV-229E) infection. ZBP1 expression did not impact the replication of all tested viruses. In addition, ZBP1-mediated cell death during infection depends on the Zα2 and RHIM1 domains and partially on the C-terminal domain. These findings suggest that Z-NA can be detected by the Zα2 domain to initiate cell death pathways during infection with some Z-NA negative viruses and that the RHIM1/C-terminal domains are necessary for ZBP1-induced cell death. Further research is needed to determine the Z-NA ligand and the precise mechanism of ZBP1-mediated antiviral responses and how they can be exploited for the development of novel antiviral therapies.

The advent of CRISPR/Cas9 revolutionized the field of genetic engineering and gave rise to the development of new gene editing tools including prime editing. Prime editing is a versatile gene editing method that mediates precise insertions and deletions and can perform all 12 types of point mutations. In turn, prime editing represents great promise in the design of new gene therapies and disease models where editing was previously not possible using current gene editing techniques. Despite advancements in genome modification technologies, parallel enrichment strategies of edited cells remain lagging behind in development. To this end, this project aimed to enhance prime editing using transient reporter for editing enrichment (TREE) technology to develop a method for the rapid generation of clonal isogenic cell lines for disease modeling. TREE uses an engineered BFP variant that upon a C-to-T conversion will convert to GFP after target modification. Using flow cytometry, this BFP-to-GFP conversion assay enables the isolation of edited cell populations via a fluorescent reporter of editing. Prime induced nucleotide engineering using a transient reporter for editing enrichment (PINE-TREE), pairs prime editing with TREE technology to efficiently enrich for prime edited cells. This investigation revealed PINE-TREE as an efficient editing and enrichment method compared to a conventional reporter of transfection (RoT) enrichment strategy. Here, PINE-TREE exhibited a significant increase in editing efficiencies of single nucleotide conversions, small insertions, and small deletions in multiple human cell types. Additionally, PINE-TREE demonstrated improved clonal cell editing efficiency in human induced pluripotent stem cells (hiPSCs). Most notably, PINE-TREE efficiently generated clonal isogenic hiPSCs harboring a mutation in the APOE gene for in vitro modeling of Alzheimer’s Disease. Collectively, results gathered from this study exhibited PINE-TREE as a valuable new tool in genetic engineering to accelerate the generation of clonal isogenic cell lines for applications in developmental biology, disease modeling, and drug screening.

Monkeypox virus (MPXV) is an orthopoxvirus that causes smallpox-like disease and has up to a 10% mortality rate, depending on the infectious strain. The global eradication of the smallpox virus has led to the decrease in smallpox vaccinations, which has led to a drastic increase in the number of human MPXV cases. MPXV has been named the most important orthopoxvirus to infect humans since the eradication of smallpox and has been the causative agent of the 2022 world-wide MPXV outbreak. Despite being highly pathogenic, MPXV contains a natural truncation at the N-terminus of its E3 homologue. Vaccinia virus (VACV) E3 protein has two domains: an N- terminus Z-form nucleic acid binding domain (Z-BD) and a C-terminus double stranded RNA binding domain (dsRBD). Both domains are required for pathogenesis, interferon (IFN) resistance, and protein kinase R (PKR) inhibition. The N-terminus is required for evasion of Z-DNA binding protein 1 (ZBP1)-dependent necroptosis. ZBP1 binding to Z- form deoxyribonucleic acid/ribonucleic acid (Z-DNA/RNA) leads to activation of receptor-interacting protein kinase 3 (RIPK3) leading to mixed lineage kinase domain- like (MLKL) phosphorylation, aggregation and cell death. This study investigated how different cell lines combat MPXV infection and how MPXV has evolved ways to circumvent the host response. MPXV is shown to inhibit necroptosis in L929 cells by degrading RIPK3 through the viral inducer of RIPK3 degradation (vIRD) and by inhibiting MLKL aggregation. Additionally, the data shows that IFN treatment efficiently inhibits MPXV replication in a ZBP1-, RIPK3-, and MLKL- dependent manner, but independent of necroptosis. Also, the data suggests that an IFN inducer with a pancaspase or proteasome inhibitor could potentially be a beneficial treatment against MPXV infections. Furthermore, it reveals a link between PKR and pathogen-induced necroptosis that has not been previously described.

Predatory bacteria are a guild of heterotrophs that feed directly on other living bacteria. They belong to several bacterial lineages that evolved this mode of life independently and occur in many microbiomes and environments. Current knowledge of predatory bacteria is based on culture studies and simple detection in natural systems. The ecological consequences of their activity, unlike those of other populational loss factors like viral infection or grazing by protists, are yet to be assessed. During large-scale cultivation of biological soil crusts intended for arid soil rehabilitation, episodes of catastrophic failure were observed in cyanobacterial growth that could be ascribed to the action of an unknown predatory bacterium using bioassays. This predatory bacterium was also present in natural biocrust communities, where it formed clearings (plaques) up to 9 cm in diameter that were visible to the naked eye. Enrichment cultivation and purification by cell-sorting were used to obtain co-cultures of the predator with its cyanobacterial prey, as well as to identify and characterize it genomically, physiologically and ultrastructurally. A Bacteroidetes bacterium, unrelated to any known isolate at the family level, it is endobiotic, non-motile, obligately predatory, displays a complex life cycle and very unusual ultrastructure. Extracellular propagules are small (0.8-1.0 µm) Gram-negative cocci with internal two-membrane-bound compartmentalization. These gain entry to the prey likely using a suite of hydrolytic enzymes, localizing to the cyanobacterial cytoplasm, where growth begins into non-compartmentalized pseudofilaments that undergo secretion of vesicles and simultaneous multiple division to yield new propagules. I formally describe it as Candidatus Cyanoraptor togatus, hereafter Cyanoraptor. Its prey range is restricted to biocrust-forming, filamentous, non-heterocystous, gliding, bundle-making cyanobacteria. Molecular meta-analyses showed its worldwide distribution in biocrusts. Biogeochemical analyses of Cyanoraptor plaques revealed that it causes a complete loss of primary productivity, and significant decreases in other biocrusts properties such as water-retention and dust-trapping capacity. Extensive field surveys in the US Southwest revealed its ubiquity and its dispersal-limited, aggregated spatial distribution and incidence. Overall, its activity reduces biocrust productivity by 10% at the ecosystem scale. My research points to predatory bacteria as a significant, but overlooked, ecological force in shaping soil microbiomes.

Lung metastatic cancers represent a major challenge in both basic and clinical cancer research. The ability to treat lung metastases to date has been challenging, current treatment paradigms are a mix of classic radiotherapy, chemotherapies and tumor-targeted therapies, with no one treatment that is effective for all tumors. Oncolytic viruses (OVs) represent a new therapeutic modality for hard-to-treat tumors. However, major questions still exist in the field, especially around how to therapeutically arm and deliver OVs to sites of disseminated tumors. To address this need, oncolytic myxoma viruses (MYXV) that expresses TNF superfamily member transgenes (vMYX-hTNF or vMyx-mLIGHT) were tested in an immunocompetent syngeneic lung metastatic murine osteosarcoma model. Three versions of this model were used; 1-an early intervention model, 2-an established tumor model, defined by both average tumor burden and failure of anti-PD-L1 and vMyx-TNF monotherapies, and 3-a late-stage disease model, defined by the failure the combination of vMyx-hTNF/PBMCs and anti-PD-L1 therapy. These three models were designed to test different questions about therapeutic efficacy of armed MYXV and delivery of MYXV to lung metastases. In the early intervention model, vMyx-hTNF was found to be an effective therapy, especially when delivered by leukocyte carrier cells (either bone marrow or PBMCs). Next, the combination of immune checkpoint inhibitors, including anti-PD-L1, anti-PD-1 and anti-CTLA-4, with vMyx-TNF/PBMCs were found to increase efficacy in treated mice compared to monotherapies. The established model was used to test potential synergy of vMyx-hTNF with anti-PD-L1 therapy. This model was defined by the failure of the monotherapies, however, in combination, treated mice survived significantly longer, and had lower average tumor burden throughout. This model was also used to test tumor specific delivery using ex vivo loaded PBMCs as carrier cells. Using MYXV expressing Tdtomato, PBMCs were found to deliver MYXV to tumors more effectively than free virus. In the most stringent late-stage disease model, vMyx-mLIGHT/PBMCs and vMyx-mLIGHT/PBMCs plus anti-PD-1 were tested and found to be efficacious where combination vMyx-TNF/PBMCs plus PD-1 failed. These results taken together show that TNFSF arming of MYXV, especially when delivered by autologous PBMCs, represents a new potential treatment strategy for lung metastatic tumors.

Programmed cell death plays an important role in a variety of processes that promote the survival of the host organism. Necroptosis, a form of programmed cell death, occurs through a signaling pathway involving receptor-interacting serine-threonine protein kinase 3 (RIPK3). In response to vaccinia virus infection, necroptosis is induced through DNA-induced activator of interferon (DAI), which activates RIPK3, leading to death of the cell and thereby inhibiting further viral replication in host cells. DAI also localizes into stress granules, accumulations of mRNAs that have stalled in translation due to cellular stress. The toxin arsenite, a canonical inducer of stress granule formation, was used in this project to study necroptosis. By initiating necroptosis with arsenite and vaccinia virus, this research project investigated the roles of necroptosis proteins and their potential localization into stress granules. The two aims of this research project were to determine whether stress granules are important for arsenite- and virus-induced necroptosis, and whether the proteins DAI and RIPK3 localize into stress granules. The first aim was investigated by establishing a DAI and RIPK3 expression system in U2OS cells; arsenite treatment or vaccinia virus infection was then performed on the U2OS cells as well as on U2OSΔΔG3BP1/2 cells, which are not able to form stress granules. The second aim was carried out by designing fluorescent tagging for the necroptosis proteins in order to visualize protein localization with fluorescent microscopy. The results show that arsenite induces DAI-dependent necroptosis in U2OS cells and that this arsenite-induced necroptosis likely requires stress granules. In addition, the results show that vaccinia virus induces DAI-dependent necroptosis that also likely requires stress granules in U2OS cells. Furthermore, a fluorescent RIPK3 construct was created that will allowfor future studies on protein localization during necroptosis and can be used to answer questions regarding localization of necroptosis proteins into stress granules. This project therefore contributes to a greater understanding of the roles of DAI and RIPK3 in necroptosis, as well as the roles of stress granules in necroptosis, both of which are important in research regarding viral infection and cellular stress.