RSS 1.0Graduate Student Dissertations & Theses
Recent Submissions
-
Isoform-Specific Regulation of the Yeast V-ATPase a-subunitV-ATPase primarily acidifies organelles of the endocytic and secretory pathway. The lumen of each of these organelles maintains a distinct pH, but how this diverse pH range is established is not well understood. Different modes of regulation of the V ATPase may help fine-tune their activity. The a-subunit from the membrane bound Vo subcomplex is a regulatory hub of V-ATPase which harbors many regulatory interactions in its cytosolic N-terminal (aNT) domain. The aNT domains of all isoforms consist of a proximal and a distal globular domain connected by a coiled-coil. Vo a-subunit exhibits organelle- and tissue-specific isoforms. Vph1 and Stv1 are the two organelle-specific a subunit isoforms in yeast. V-ATPase containing the Vph1 isoform reside in the lysosome-like yeast vacuole whereas V-ATPase containing the Stv1 isoform reside predominantly in the Golgi apparatus. The RAVE (Regulator of H+ -ATPase of vacuoles and endosomes) complex and phosphoinositide phospholipids (PIP lipids) are two important factors previously implicated in isoform-specific V-ATPase regulation. The a subunit isoforms vary in their dependence on the RAVE and PIP lipid. But where the information of RAVE and PIP lipid recognition reside in the aNT domain and how these two regulatory inputs are integrated to control V-ATPase function are not well understood. We hypothesize that the aNT domain contains distinct sequences for RAVE and PIP lipid recognition, and that the differential interactions of the a-subunit isoforms with RAVE and PIP lipids result in isoform-specific function and regulation of the V ATPase. To better understand the regulatory information present in the aNT domain, we generated chimeric aNT constructs by combining parts of Vph1NT and Stv1NT. Vph1NT was previously shown to bind to a RAVE subunit and Vph1 V-ATPases require the RAVE complex for their assembly. Stv1-containing V-ATPases on the other hand assemble independent of RAVE and Stv1NT does not bind to RAVE subunits in vitro. In chapter 2, we have shown that replacing the proximal domain of Vph1NT with the proximal end of Stv1NT fully restores the RAVE interaction, implicating this region of Vph1 in RAVE-dependent assembly. The two isoforms also exhibit preferences for distinct PIP lipids enriched in their organelle of residence; Stv1NT binds tightly to Golgi PI(4)P and Vph1NT binds to vacuolar PI(3,5)P2. A PI4P binding site in the proximal domain of Stv1NT was previously reported. Our analysis with chimeric constructs suggests that a 6-amino acid sequence containing this site is sufficient to transfer PI(4)P binding to Vph1NT. Our results also suggest that both the proximal and the distal ends of Stv1NT contain sequences that promote PI4P binding. Interestingly, when expressed in yeast as a full-length a-subunit, the chimera containing both PI4P and PI(3,5)P2 binding sites has wild-type level activity and assembly in isolated vacuoles even though it lacks a RAVE binding site. Although V-ATPases in this chimeric strain is fully functional there are consequences of their altered regulatory properties. V-ATPases in this chimera disassemble during glucose deprivation but do not reassemble efficiently after glucose re addition, consistent with a lack of RAVE binding. We also observed a delayed growth in media containing raffinose suggesting they cannot readily adjust during a transition to a less preferred carbon source. While the lack of RAVE binding gives rise to the growth defect of this chimera in a poor carbon source, increased PI(3,5)P2 binding of this mutant, on the other hand, contributes to growth benefit in alkaline pH stress. Together these data reveal the interplay between two mechanisms of V-ATPase regulation and suggest that aNT domains can functionally integrate multiple regulatory inputs.
-
Context Matters: Using Genomic Knowledge to Improve Disorder Classification ModelsDespite heritability estimates that suggest a high ceiling for the classification of many complex genetic disorders, current models have only been moderately successful at accurately classifying cases and controls of these disorders. The knowledge base about the human genome is large and continuously growing, but disorder classification models rarely use any of that information beyond genetic associations. We use three different genomic context data granularities, 4 different machine learning models, and datasets of mood disorders, ADHD, and type 2 diabetes to test hypotheses on whether including genomic context can improve modelling of disorder risk. When predicting whether subjects had been diagnosed with any mood disorder, we found that using polygenic risk scores from other psychiatric disorders in logistic regression models improved classification performance as measured by the area under the receiver operating characteristic curve (AUC). In another study classifying cases of ADHD and controls, we found that the addition of summations of risk based on the genetic variants' inclusion in gene sets associated with ADHD improved AUCs in random forest modelling. The random forest importance scores of those gene set polygenic risk scores showed biological relevance through the correlation of importance scores with relative gene set expression in the brain. In the final study classifying type 2 diabetes cases and controls, for each genetic variant, we attached several types of functional genomic annotations to genotype data. These genomic context informed genotype data were used in convolutional neural networks and significantly improved AUC compared to polygenic risk score models while using a within-model adversarial ancestry task to adjust for potential confounding due to ancestry. In these models, we found that some risk features developed by context informed data overlapped with features developed with standard genotype input while other risk features were unique to the input type. Together, these studies provide evidence that context matters when looking at the disorder risk conferred by genetic variants in complex genetic disorders.
-
Coping with Stress: Rewiring of Host Stress Responses by Human Cytomegalovirus to Redirect Protein TranslationHuman cytomegalovirus (HCMV) is a highly prevent pathogen with seropositivity ranging from 40-90% in North America, and in excess of 100% in developing nations. Although HCMV infections are often mild to asymptomatic among immunocompetent hosts, HCMV represents a significant cause of morbidity and mortality for those with compromised immune systems. HCMV's systemic dissemination within the host is facilitated by infected blood monocytes. Within monocytes, HCMV establishes a quiescent infection, thus acting as "Trojan horses", delivering the virus to distant end organs all while remaining hidden from innate immune detection. Blood monocytes are inherently short-lived however, with an average lifespan of only 48 h. To circumvent this shortcoming HCMV aberrantly regulates the Akt/mTORC1 signaling axis. Our lab has previously shown that HCMV binding and entering into target monocytes elicits a unique action of Akt, characterized by the preferential phosphorylation at Serine 473 (S473). Critically, downstream substrate specificity of Akt signaling is dictated by the ratio of S473 to threonine 308 phosphorylation, suggesting this unique Akt activation would have biologic significance during infection. The studies of this thesis reveal that a major consequence of HCMV-mediated aberrantly regulated Akt is a robust increase in protein translation through the activation of mTORC1 and its downstream signaling substrates. Moreover, HCMV uniquely usurps the activity of heat shock factor 1 (HSF1) to bypass cellular stress response traditionally intended to limit protein synthesis through the direct interaction between HSF1 and mTOR. The tightly regulated signaling events downstream of Akt and mTORC1 lead to a reshaping of the host translatome to redirect protein synthesis towards antiapoptotic factors necessary for the induction of monocyte survival beyond their canonical 48 h lifespan. Elucidating the molecular mechanisms behind HCMV's control over protein translation may allow for the development of novel therapies that specifically target HCMV-infected monocytes, thus preventing viral dissemination and the establishment of disease.
-
The molecular basis of IP3R recognition by the ubiquitin-proteasome pathwayInositol 1,4,5-trisphosphate (IP3) receptors (IP3Rs) form ~1.2MDa tetrameric Ca2+ channels in the endoplasmic reticulum (ER) membrane of mammalian cells. IP3R1 is the most ubiquitously expressed among all three IP3R isoforms. Upon activation by the second messenger, IP3, IP3Rs undergo a conformational change that leads to channel opening and allows Ca2+ ions to flow from the ER stores into the cytosol. IP3R-dependent Ca2+ signaling is crucial to many cellular events. The Wojcikiewicz laboratory has found that active IP3Rs are quickly processed by the ubiquitin-proteasome pathway (UPP), which is initiated by their association with the erlin1/2 complex. The association also recruits the E3 ligaseRNF170 to active IP3Rs. However, how activated IP3Rs are recognized by the erlin1/2 complex remains unclear. Using IP3R mutants, we discovered that the erlin1/2 complex binding site is on the third intraluminal loop (IL3) of IP3R and also found that a region at the N-terminus of IL3 is critical to IP3R channel activity. We also used UPP inhibitor TAK-243 to confirm the sequence of events that leads to IP3R processing by the UPP: the erlin1/2 complex association is prior to IP3R ubiquitination and degradation. Surprisingly, we found that long-term treatment with UPP inhibitors can inhibit IP3R-mediated Ca2+ signaling and affect other aspects of Ca2+ handling in cells. Overall, these results help us understand how the large ion channels are deconstructed and further our knowledge of substrate processing by the UPP.
-
Characterizing actin assembly mechanisms in ALS pathologyAmyotrophic lateral sclerosis (ALS) is a neurodegenerative disease linked to >sixty genes. However, the mechanistic details of onset are unknown and there is no known cure or effective treatment. Disease mechanisms linked to the dysfunction of the neuronal cytoskeleton are arguably the least explored, despite being involved in important cell processes. Eight cytoskeleton associated proteins are genetically linked to ALS onset, including tubulin-4A, spastin, kinesin-5A, dynactin subunit-1, neurofilament, peripherin, alsin, and profilin-1 (PFN1). Most of these genes are linked to microtubule dynamics, yet only PFN1 directly regulates actin assembly. Profilin is essential for many neuronal cell processes mediating the dynamics of lipids, nuclear signals, and the cytoskeleton in neurons. We performed a quantitative biochemical comparison of the impact of the eight ALS-associated profilin variants on actin assembly using single-molecule microscopy assays. The binding affinity of each ALS-related profilin for actin monomers was loosely correlated with the actin filament nucleation strength. The A20T, R136W, Q139L, and C71G variants failed to activate formin-based actin assembly, mostly explained by deficiencies in binding to poly-L-proline stretches in formin. In addition, chemical denaturation experiments suggest that the folding stability of some ALS profilin proteins impacts on actin assembly. Finally, to better understand the connection between the cytoskeleton and ALS we investigated the role of profilin and its direct binding ligand TDP-43 (TAR DNA-binding protein 43) on actin assembly. TDP-43 misregulation is a hallmark of nearly all forms of neurodegeneration, including an estimated 40% of familial ALS. Using advanced microscopy assays, we visualized purified TDP-43 forming biomolecular condensates. Actin was sequestered within TDP-43 condensates tempering total filament assembly. These observations were further exacerbated in the presence of profilin. These results indicate disruptions to actin assembly contribute to ALS and suggest therapeutic interventions targeting actin assembly may be a useful in treating ALS.
-
Engineering modifiable monobody-based biosensor scaffolds for use in vitro and in cellsProtein-based fluorescent biosensors are powerful tools for analyte recognition in vitro and in cells. Many protein-based binding scaffolds have been developed that recognize ligands with affinity and specificity comparable to those of conventional antibodies, but are smaller, readily overexpressed, and more amenable to engineering. Like antibodies, these binding domains are useful as recognition modules in protein switches and biosensors, but they are not capable of reporting on the binding event by themselves. Here, we engineer two adaptable monobody-based biosensor scaffolds. The first, termed FN3-AFF, is composed of a small binding scaffold-a consensus-designed fibronectin 3 monobody-that it engineered so it undergoes a conformational change upon ligand binding. This change is detected by Förster resonance energy transfer using chemical dyes or cyan and yellow fluorescent proteins as donor/acceptor groups. By grafting substrate recognition residues from different monobodies onto this scaffold, we create fluorescent biosensors for c-Abl Src homology 2 (SH2) domain, WD40-repeat protein 5 (WDR5), small ubiquitin-like modifier-1 (SUMO), and h-Ras. The biosensors bind their cognate ligands reversibly, with affinities consistent with those of the parent monobodies, and with half times of seconds to minutes. We modify the rates of the FN3-AFF switch using a combination of computational strategies and experimental validation to identify rate altering mutations. The resulting mutant results in a 2-fold improvement of kinetics (0.88 s-1 to 1.65 s-1). The second design is an intensiometric biosensor, termed YFP-FN3, created by circularly permuting a naturally occurring fluorescent protein and fusing FN3 of three targets (WDR5, SH2, and HRAS) to it. Ligand binding to the FN3 greatly enhances the fluorescence of the fused fluorescent protein by causing its chromophore to mature. We illustrate the utility of these biosensors for imaging proteins in cells, recognizing endogenous levels of one of the targets, WDR5. These designs serve as generalizable platforms for creating genetically-encoded, ratiometric or intensiometric biosensors by swapping binding residues from known monobodies, with minimal modification.
-
The dynamic role of the androgen receptor (AR) and ABL interactor-1 (ABI1) axis in prostate cancerThe nuclear hormone receptor and transcription factor, the androgen receptor (AR), is the main driver of prostate cancer growth and disease progression. Hallmark target genes of AR are used as biomarkers to indicate disease progression and treatment response. Furthermore, current clinical treatments for prostate cancer include androgen deprivation treatment and anti-AR which deplete ligand availability or directly target the androgen receptor, respectively. Transcription regulates key functions of living organisms in normal and diseases states, including cell growth and development, embryonic and adult tissue organization, and tumor progression. Here we identify a novel mechanism of transcriptional regulation by an actin regulatory and signaling protein, ABI1. As established by ChIP sequencing and DNA binding assays, ABI1 binds to chromatin through its intrinsically disordered DNA binding domain. Furthermore, ABI1 interacts with AR in vitro and in vivo and targets its activity to specific subset of genes. ABI1-AR driven transcription is dysregulated during prostate tumor progression. Additionally, anti-androgen and anti-AR treatments induce alterations in AR-mediated transcription which leads to downregulation of ABI1 expression and induces disruption of epithelial integrity. The results from this study indicate that ABI1 controls tumor plasticity through connecting actin cytoskeleton and cellular signaling to transcriptional regulation. We propose that ABI1 is a regulator of transcriptional homeostasis in prostate cancer.
-
All roads lead to Akt: how HCMV controls Akt activity to generate a microenvironment favoring viral disseminationInfection with human cytomegalovirus (HCMV) is highly prevalent, with seropositivity reaching 80% in developed countries and upwards of 100% in developing countries. Because of the obligate intracellular nature, similar to other viruses, HCMV modulates cellular environment to promote infection. The Phosphatidylinositol-3-kinase (PI3K)/Akt pathway is modulated by many viruses because of its central importance in multiple cellular processes. HCMV has been reported to regulate Akt activity during all three stages of its lifecycle: quiescent, latent, and lytic infection. Previously, our lab showed that HCMV promotes Akt phosphorylation preferentially at serine 473 (S473) during quiescent infection to promote monocyte survival and hematogenous dissemination of the virus. This is in contrast to growth factor-mediated survival of monocytes which involves Akt phosphorylation at both S473 and threonine 308 (T308). The exact mechanism how HCMV induces an atypical Akt activation in monocytes, and its biological significance remains unclear. The studies in this thesis reveal that HCMV glycoproteins gB and gH work in concert to initiate a HCMV-specific signalosome responsible for a robust Akt activation through phosphorylation at S473 in monocytes. Moreover, we demonstrate that virion-associated US28-mediated signaling is required to inhibit Akt phosphorylation at T308. Induction of the atypical Akt activation is essential to establish HCMV quiescent infection and monocyte survival which are required for successful dissemination of the virus. We found that HCMV phosphorylates Akt at both S473 and T308 during viral entry into fibroblasts. However, with the progression of infection, HCMV suppresses Akt phosphorylation at both residues for efficient viral genome replication. Overall, our data demonstrate that HCMV regulates Akt activity in a multifaceted approach which depends on the type of infection to generate a microenvironment favoring viral infection and dissemination.
-
The microtranscriptome of Parkinson's disease: human and animal studiesParkinson's disease (PD) is a progressive debilitating neurodegenerative disorder affecting 2% of subjects over age 65. The major motor symptoms of PD result from loss of midbrain dopamine-synthesizing neurons and include resting tremor, rigidity, slowed movements (bradykinesia), and postural instability. Although there are known genetic causes, most PD cases have undetermined etiology. Thus, there is considerable interest in identifying biomarkers of early-stage PD to develop effective intervention strategies that might modify the disease course. Studies completed in this dissertation were designed to identify and validate such biomarkers through comprehensive analysis of the oral microtranscriptome in human subjects, as well as in animal and cellular PD models. A total of 300 human subjects were recruited, including 178 with PD and 122 controls. Both groups contained subjects with comorbid or isolated conditions frequently associated with PD, including restless legs syndrome, dystonia, and essential tremor. Subjects completed questionnaires to assess risk factors, medication use, and motor and non-motor symptoms, and underwent formal neurological examination. Approximately half of the subjects also completed olfactory testing and computerized neuromotor, cognitive, and postural testing. Saliva samples were collected using a vial or swab. Levels of microRNA and microbiota were quantified using next generation sequencing. Saliva miRNA levels were also compared in mice expressing wildtype (WT) or A53T mutant copies of human alpha synuclein (SNCA), a known cause of PD. Performance of the mice was also assessed in motor and cognitive tasks during pre- and early-symptomatic stages of the model. Lastly, measurements of both extracellular miRNA and cellular mRNA, as well as oxidative stress and DNA damage, were completed in human iPSC-derived dopaminergic neurons expressing WT or A53T SNCA, combined with exposure to media or Paraquat, an environmental risk factor for PD. All PD subjects fell within the mild-moderate early stage category. Significant differences were observed for olfactory, postural, and more challenging cognitive tests. Profiling of salivary miRNAs revealed subsets highly-changed in early stage PD. The mouse and dopaminergic neuron data strongly supported the findings for two specific miRNAs - miR-103a-3p and miR-107 - and indicate a number of key cellular pathways are altered across the disease and models.
-
Investigating the Paxillin and Hic-5 InteractomesFocal adhesions are macromolecular structures that connect the cellular actin cytoskeleton to the surrounding extracellular matrix through transmembrane integrin receptors and the action of numerous proteins, including enzymes, signaling proteins, and scaffolding proteins. Paxillin and Hic-5 are two scaffolding adapter proteins that primarily localize to focal adhesions, as well as other intracellular regions including the nucleus and centrosome. Their signaling partners at focal adhesions and effects on the actin cytoskeleton have been well-characterized over the course of decades through rigorous biochemical studies, but broader analysis and comparison of the interactomes of these two closely related proteins has not yet been thoroughly pursued. In the introduction, recently described roles for paxillin and Hic-5 in regulating cell shape and invadopodia through actin dynamics will be described, in addition to recent discoveries regarding their interaction with the microtubule and intermediate filament cytoskeletons. This background will provide context to data presented in chapter two, in which paxillin and Hic-5 were used as baits for proximity labeling to compare their interactomes, to identify numerous potentially novel interactors for both proteins, and to confirm many previously known partners. Two interesting results of this analysis include a possible proximity interaction between both paxillin and Hic-5 with septin-7, and confirmation of a robust proximity reaction between paxillin and ponsin. In chapter three, another newly discovered interactor with paxillin will be characterized: the formin mDia1. These proteins are shown to bind in vitro and co-localize in vivo. Most interestingly, paxillin relieves mDia1 auto-inhibition to accelerate actin polymerization in in vitro TIRF microscopy assays. Finally, the significance of these results and avenues for future investigation will be discussed, including further confirmation of proximity interactors and experiments to better understand the mechanism by which paxillin interacts with mDia1.
-
Regulation of fibrosis through the ubiquitin pathwayScarring in the cornea obstructs the refraction of incoming light onto the retina causing visual disability. Both acute scarring and chronic fibrosis are characterized by an accumulation of disorganized extracellular matrix (ECM). Disorganized ECM is deposited into the wound by specialized cells termed myofibroblasts. Pathological myofibroblasts are characterized by the expression of the highly contractile alpha smooth muscle actin (a-SMA) and the av-family of integrins (avb1,b3, b5, b6). The persistence of myofibroblasts in a healing wound promotes an autocrine loop of TGFb activity, over contraction of tissue, deposition of fibrotic ECM proteins, and ultimately the generation of scar tissue. My work is focused on the relative contribute of the deubiquitinase, USP10 to scarring in the cornea. I found that after wounding an increase in the expression of USP10 leads to deubiquitination of integrins and a subsequent increase in integrin recycling and matrix deposition. Knockdown of USP10 in vivo after corneal wounding significantly reduced the presence of myofibroblasts and immune cells in the healing wound, and corneal scarring. Through a yeast 2-hybrid screen I also identified a novel USP10 interacting protein, the formin Daam1. I found that Daam1 sequesters USP10 on actin stress fibers inhibiting its activity. Under pathological conditions, the expression of both USP10 and Daam1 are increased. My data suggest that Daam1 acts as a cellular reservoir, adding a layer of homeostatic control over USP10 activity and integrin function. Although defects in protein degradation have been identified as a major contributor to many diseases, together, my studies indicate that protein degradation (ubiquitin) pathways need to be considered in the context of integrin biology and in the pathogenesis of fibrotic healing.
-
Characterization of the effects of steroid-resistant nephrotic syndrome associated MYO1E mutations on myosin 1e activity and podocyte functionsMYO1E mutations are associated with familial pediatric nephrotic syndrome, a disease in which pharmaceutical treatment is limited to immunosuppressive drugs such as steroids and cyclosporine but still often fails to prevent progression to end-stage renal disease. While Myo1e deficiency leads to glomerular filtration abnormality, including podocyte foot process effacement and glomerular basement membrane thickening, developing a precise treatment for this disease is hindered, as the underlying mechanisms linking MYO1E mutations and nephrotic syndrome are unclear. With the power of whole genome and exome sequencing, multiple novel Myo1e variants have been rapidly identified from SRNS cohorts. To determine whether a MYO1E mutation identified in patients is likely to be pathogenic or benign, we have characterized the functional effects of novel sequence variants in cultured podocytes. Differential protein degradation, localization, endocytic and motor activities of Myo1eT119I and Myo1eD388H have been discovered (Chapter 2). Specifically, even when expressed as a full-length protein, Myo1eT119I is deficient in localization to podocyte junctions and clathrin-coated vesicles (CCVs). Consistent with the hypothesis that Myo1eT119I is a loss-of-function mutant, cells expressing Myo1eT119I exhibit decreased CCV density and prolonged CCV lifetimes. The junctional and CCV localization of Myo1eD388H is not affected but it exhibits increased association with structures in the membrane-actin interface. Most importantly, Myo1eD388H is deficient in ATP hydrolysis and actin filament translocation. We have also characterized other MYO1E variants to provide cell-based evidence to assist in the curation of variants of uncertain significance (Chapter 3). Unexpectedly, while Myo1eD185G may be considered as a likely benign variant based on the population and computational predictions, it exhibits prolonged association with the podocyte junctions, while no junctional localization and dissociation abnormality was found in a likely deleterious variant, Myo1eR523W. We also discovered that localization of Myo1edel3094-7 to the podocyte junctions can be partially restored with a proteasomal inhibitor treatment, which may be considered as a potential treatment for patients with this variant. Finally, to follow up on the questions derived from Chapter 2, we examined our hypothesis that Myo1eD388H tail is constitutively active, we demonstrated the similar junctional exchange but differential endocytic activity of Myo1eD388H and Myo1eTAIL, revealing the critical regulation of Myo1e motor to tail domain when it comes to clathrin-mediated endocytosis (Appendix 1). Overall, the studies documented here have uncovered the differential molecular defects of steroid resistant nephrotic syndrome (SRNS)-associated Myo1e variants and further elucidated the underlying mechanisms of podocyte disease.
-
Examination of the effects of Myosin-1e expression on tumor behavior and computational analysis of MYO1E mutations associated with kidney diseaseActin and actin cytoskeleton associated proteins are important for maintenance of cellular homeostasis and cellular processes such as cell polarization, cell movement, and endocytosis. The long-tailed class one myosin, Myosin-1e (Myo1e), is an actin dependent molecular motor that is found to be expressed in and contribute to the function of epithelial cells. Specific mutations of Myo1e are associated with dysfunction of specialized kidney epithelial cells known as podocytes and increased incidence of kidney disease. We found that disease associated mutation of Myo1e affects the conformation of key functional domains within the Myo1e protein structure using molecular dynamic simulations (Chapter 2). This finding leads to a more mechanistic understanding of dysfunctional Myo1e protein conformational pathways that may be associated with kidney disease. We also find that Myosin-1e expression influences the behavior of mammary tumors in mice (Chapter 3). Specifically, mammary epithelial cell oncogenic transformation induced using the MMTV-PyMT oncogene (mouse mammary tumor virus, polyoma middle T-antigen) was phenotypically altered in mice lacking Myo1e expression compared to Myo1e expressing mice. The tumor cells in mice lacking Myo1e were more differentiated and expressed transcriptomic profiles associated with tumor suppression compared to the more oncogenic profiles of Myo1e expressing cells. In-vitro analysis revealed a cell autonomous effect of Myo1e expression on cell-cell junctional strength and gene expression associated with differentiation (Chapter 3). Appendix chapters 1 and 2 discuss experiments examining Myo1e expression on in-vitro cell migration and epithelial to mesenchymal transition (EMT). Overall, experiments discussed in this thesis have uncovered insight into the effects of Myo1e expression on breast cancer progression, cell differentiation and motility. Moreover, experiments in this thesis also contribute to further understanding of the dynamics of protein conformation associated with disease mutations of a class one myosin.
-
The impact of mosquito salivary factors on Chikungunya virus transmissionVector-borne diseases affect an estimated 3.9 billion people in over 128 countries annually (WHO, 2017b). Mosquito-borne viruses, such as Chikungunya virus (CHIKV), are transmitted to human hosts through the saliva of female blood-feeding mosquitoes. During blood-feeding, salivary molecules enhance blood-feeding efficiency, dampen specific inflammatory signals, and enhance virus infection and dissemination. Skin is the interface where mosquito salivary molecules and viruses interact with the human immune system. Our long-term goal is to understand the impact of mosquito saliva on the skin and the role of salivary molecules in potentiating virus infection. This thesis aims to model the human skin bite site, identify mosquito salivary molecules present during virus infection, and investigate their impact on virus replication. We developed a human skin arbovirus infection model that was viable for four days ex vivo (Esterly et al., 2022a). Arbovirus infection in human skin ex vivo allowed flavivirus and alphavirus replication and dissemination. Using this model, we will investigate mosquito salivary factors and their impact on virus transmission. Next, we identified mosquito salivary factors highly expressed in Ae. aegypti salivary glands during CHIKV infection. We hypothesize that salivary proteins that contain secretion signals and are found to be highly expressed during CHIKV infection are delivered to the bite site during mosquito feeding. Through a combination of RNA-seq and miRNA-seq analysis, we identified several salivary factors, protein and small non-coding RNAs, which warrant further investigation. Lastly, we assessed the impact of identified and recombinantly expressed proteins or specific small RNA inhibitors on CHIKV replication in vitro relative to Ae. aegypti salivary gland extract (SGE). We found that both protein and small RNAs can influence CHIKV replication in a significant manner; however, the effect observed by mosquito saliva is a combination of multiple factors and will require more investigation to characterize fully. In summary, mosquito salivary factors work in concert to enhance virus replication and dissemination. Although individual factors may not recapitulate the total effect of SGE, the discovery and isolation of these factors further our understanding of arbovirus transmission. These studies highlight the complexity of mosquito salivary secretion and further our knowledge of mosquito-borne virus transmission. The following thesis will describe our most recent findings and contribution to understanding mosquito saliva's role during arbovirus transmission.
-
Biomimetic extracellular matrix hydrogels to model and investigate conventional outflow cell biology under normal and simulated glaucomatous conditionsDysfunction of the conventional outflow pathway (comprised of the trabecular meshwork (TM) and adjacent Schlemm's canal (SC)) is the principal cause of elevated intraocular pressure in primary open-angle glaucoma. Other in vitro TM model systems cannot accurately mimic the cell-extracellular matrix (ECM) interface, limiting their use for investigating glaucoma pathology. In this dissertation, we report a novel biomimetic hydrogel by mixing donor-derived human TM (HTM) cells with ECM proteins found in the native tissue. We demonstrated that this HTM hydrogel system allowed for investigation of actin arrangement, ECM remodeling, cell contractility, and HTM stiffness on a simulated tissue level (Chapter 2). Furthermore, we showed that TGFβ2-induced ERK signaling negatively regulates Rho-associated kinase-mediated phospho-myosin light chain expression and HTM cell contractility when cultured on soft ECM hydrogels but not on glass (Chapter 3). YAP and TAZ are important mechanotransducers implicated in glaucoma pathogenesis. We demonstrated that YAP/TAZ activity was upregulated by transforming growth factor beta 2 (TGFβ2) in both HTM and HSC cells cultured on/in ECM hydrogels (Chapters 4 and 5). It is widely accepted that the glaucomatous TM/SC interface is stiffer. To mimic the stiffness difference between diseased and healthy tissue, we utilized two different methods. In Chapter 4, riboflavin was used to facilitate secondary UV crosslinking of collagen fibrils and stiffen the matrix. We showed that ECM stiffening elevated YAP/TAZ activity in HTM cells through modulating focal adhesions and cytoskeletal rearrangement. In Chapter 6, we developed an ECM-alginate hybrid hydrogel system, which allowed for on-demand control over matrix stiffness during the culture of cells. We found that the stiffened matrix increased nuclear YAP and filamentous-actin fibers in HSC cells, which was completely reversed by matrix softening. We further demonstrated that YAP/TAZ inhibition could rescue HTM/HSC cell dysfunction induced by either TGFβ2 or stiff matrix (Chapters 4, 5, and 6). Finally, we showed that pharmacologic YAP/TAZ inhibition had promising potential to improve outflow facility in an ex vivo mouse eye perfusion model (Chapter 6). Collectively, we have developed bioengineered ECM hydrogels for modeling and investigation of conventional outflow cell-ECM interactions under normal and simulated glaucomatous conditions.
-
A dual to the death: using novel host-directed antivirals to promote death of HCMV-infected myeloid cells through apoptosis and necroptosisHuman cytomegalovirus (HCMV) is a ubiquitous member of the betaherpesvirus family, with seroprevalence rates ranging from 40-100% worldwide. Although primary infection is asymptomatic in most immunocompetent patients, HCMV is a significant cause of morbidity and mortality in the immunosuppressed and immunonaїve, including transplant recipients, patients with AIDS, and developing fetuses in utero. The diverse clinical presentations of HCMV are attributable to the pervasive systemic dissemination and extensive cellular tropism of the virus. Peripheral blood monocytes are believed to be the key cells responsible for HCMV dissemination from the initial site of infection to distant organ systems. Monocytes are normally short-lived, surviving for only 48 h in circulation before undergoing apoptosis. Previous work from our lab has shown that HCMV circumvents the short lifespan of monocytes by inducing a noncanonical activation of Akt to upregulate the expression of antiapoptotic proteins, thereby prolonging survival of infected monocytes. HCMV promotes survival in the absence of viral replication and lytic gene expression, rendering current direct-acting antivirals ineffective against quiescently infected monocytes. There are currently no antivirals that target quiescent or latent HCMV infection. We hypothesize that targeting host proteins that are essential for HCMV's induction of monocyte survival mechanisms will reduce viability of infected monocytes, ultimately reducing systemic viral dissemination. The studies in this thesis investigate novel host-directed antivirals targeted at two different cellular factors and their efficacy against quiescent HCMV infection in monocytes. The first host protein under scrutiny as an antiviral target is Sirtuin 2 (Sirt2), an NAD+-dependent deacetylase. Treatment with novel Sirt2 inhibitors promoted death of HCMV-infected monocytes as a cellular antiviral defense response through two concurrent regulated death pathways: apoptosis and necroptosis. HCMV has developed mechanisms to impede both death pathways, but inhibition of Sirt2 relieves the viral obstructions on both pathways by disrupting HCMV's unique phosphorylation on Akt. The second host protein targeted as a potential antiviral strategy is Mcl-1, an antiapoptotic member of the Bcl-2 family of proteins. HCMV-infected monocytes are dependent on Akt-dependent upregulation of Mcl-1 for survival during early infection. Treatment with Mcl-1 inhibitors blocked interaction between Mcl-1 and proapoptotic protein Bak, reducing viability of infected monocytes. Subsequent testing of Mcl-1 inhibitors in an ex vivo skin organ culture system resulted in a decrease in HCMV-infected cells that crawled out of the skin tissue, suggesting that Mcl-1 inhibition may reduce viral dissemination. Our studies lay down the groundwork for the investigation of novel host-directed antivirals, an approach that successfully targets quiescent HCMV infection in peripheral blood monocytes for the first time. Expanding the study of host-directed antivirals may bring the field one step closer to the possibility of a comprehensive antiviral regimen that is effective against all stages of viral infection.
-
Rational design of a genetically encoded fluorescent protein color switch using a modular, entropy-driven mechanismEngineered protein conformational switches have applications in cellular and in vitro biosensing, molecular diagnostics and artificial signaling systems in synthetic biology. They broadly consist of an input module and an output module that communicate via a conformational change. The overarching goal of this thesis is to tackle two major challenges in protein switch design - signal transduction, by coupling a target recognition domain to an output domain to produce a robust change in signal in addition to modularity, which allows the facile creation of sensors binding novel targets. Here, we attempted to test a rational design strategy that exploits two key protein engineering principles (1) loop entropy, by which long insertions into a loop of a host protein destabilizes the host due to an entropic cost associated with loop closure unless the inserted sequence adopts a folded structure; and (2) alternate frame folding (AFF), which allows a protein - green fluorescent protein variants(GFP), in this case - to switch between two mutually exclusive folds. Toward this goal, we first studied the effect of loop entropy at two different insertion sites in a GFP variant (chapter 2) using a well-characterized ribose binding protein as the input domain. We provide stability measurements using circular dichroism and fluorescence data to support our hypothesis of the application of the loop entropy principle in a GFP beta barrel scaffold. To provide a proof-of-concept of the combination of loop entropy and the AFF mechanism in a genetically encodable GFP scaffold, we chose an unstable, circularly permuted FK506-binding protein (cpFKBP) as the input recognition domain and inserted it in one of the two mutually exclusive folds of the GFP-AFF fusion protein (chapter 3). Upon addition of ligand, binding induced folding of the cpFKBP domain effects a conformational change in which the tenth beta strand of GFP exchanges, replacing Thr203 (green state) with Tyr203 (yellow state). We confirmed this mechanism in vitro by a ratiometric change in fluorescence output and observed that the process is slow and irreversible. We elucidate the biophysical principles underlying this mechanism by using denaturant and temperature to modulate the relative populations of the two folds in vitro. We also observed a faster and higher intensiometric response in mammalian cells which may be attributed to an alternate mechanism. We then harnessed this intensiometric response in a single fold of the fluorescent protein combined with a previously engineered monobody scaffold capable of binding a variety of targets (chapter 4). Altogether this work may have the potential to create a novel class of fluorescent protein biosensors comparable to existing single fluorescent protein-based biosensors currently available.
-
Targeting SHIP paralogs to promote microglial effector function in the CNSThe two SH2-containing inositol 5'-phosphatases , SHIP1 (INPP5D) and SHIP2 (INPPL1), play an essential role in modulation of cellular signaling by transforming the PI3K product PI(3,4,5)P3 into PI(3,4)P2. PI3K signaling triggers activation of downstream signaling cascades that drive survival, effector functions, differentiation, and proliferation. SHIP1 can also mask the cytoplasmic tails of key receptors or their adaptor proteins such as DAP12, thus preventing PI3K recruitment to Trem2, a critical receptor for microglial function. Several GWAS studies correlated single nucleotide polymorphisms (SNPs) in INPP5D with Alzheimer's Disease (AD). However, it remains unclear whether these SNPs are deleterious or protective in AD and how they alter SHIP1 protein expression. SHIP2 overexpression has also been correlated with AD, suggesting that both SHIP1 and SHIP2 might be therapeutic targets. To study how SHIP1 and SHIP2 modulate microglial functions we used small molecule inhibitors and agonists of these enzymes. In our initial study we found that both SHIP paralogs are expressed in murine microglia and that Pan- SHIP1/2 inhibition increases lysosomal size and enhances microglial phagocytosis of Ab1-42 fibrils and dead neurons using both flow cytometry and confocal microscopy. Our lead Pan-SHIP1/2 inhibitor, K161, showed Blood Brain Barrier penetration as detected in the cerebral cortex of treated mice with mass spectrometry. K161 treatment of WT mice showed no difference in microglial frequency or lysosomal content in vivo; however, we observed a significant 2 increase in Ab1-42 and dead neurons phagocytosis ex vivo in microglia in K161- treated mice versus controls. Subsequently, we discovered a novel and highly potent SHIP1 selective agonist (K306) via artificial intelligence guided computational screening. We found that K306 can reduce the release of inflammatory cytokines in macrophages and microglial cells stimulated with LPS or Ab1-42. Interestingly, K306 didn't alter microglial phagocytic uptake of cargo, but did promote degradation of phagocytosed lipid-laden cargo - defining a novel role of SHIP1 in degradation of lipid cargo in microglia. These results highlight the importance of SHIP1 and SHIP2 in microglial biology and their modulation as therapeutics in different stages of neurodegenerative disease where microglia play a major role, such as AD.
