Alzheimer's disease (AD) is the most common cause of dementia, a general term for memory loss and other cognitive abilities. Although this disease has been a major research focus since the 1980s, the pathological mechanisms are still not well understood, and therapeutic interventions have been ineffective. The most definitive method for classifying AD is through identifying accumulations of toxic amyloid-beta (Aβ) and tau proteins in post-mortem brain tissue. Dr. Su-in Lee’s lab has developed a machine learning method that integrates the pathological protein phenotypes with gene expression levels in the same brain tissue. They have highlighted 25 genes with expression level changes that correlate with the tau and Aβ protein aggregation phenotypes. For this proposal, I have integrated these human neuropathology-based phenotypes with the genetic power of Caenorhabditis elegans (C. elegans) to directly test the impact of these candidate genes on the cellular pathology. Previously, all C. elegans tau models had neuronal specific expression. However, neurons are resistant to RNAi. Therefore, I generated a novel transgenic C. elegans tau AD model that has been codon-optimized to express tau in body wall muscles instead of neurons. I measured the animal’s health with age in a series of phenotypic assays: egg-laying, growth, movement, paralysis, and lifespan analysis. This line exhibits premature paralysis and decreased crawling speeds, providing an easy to score phenotype. This new model allows for high-throughput RNAi screening to test the identified 25 genes’ effects on worm health by utilizing the automated worm-movement technology developed in the Matt Kaeberlein lab that can simultaneously determine the rate of paralysis of thousands of worms. The results of my genetic screening will lead to a better understanding of the human genes that are dysregulated in human AD brains, provide a basis for genetically-dissecting the pathways influencing tau toxicity, and suggest new therapeutic targets.

Alzheimer’s Disease (AD) is a neurodegenerative disorder characterized by the formation of senile plaques and neurofibrillary tangles through the accumulation of toxic amyloid-beta and Tau protein. There is growing recognition that extracellular vesicles (EVs) can package and transport toxic peptides associated with neurodegenerative disorders – such as AD – to other cells in the brain. Researchers in the Kaeberlein lab have designed methods to isolate these type of vesicles from C. elegans nematodes, a popular invertebrate genetic model. However, current nematode EV purification methods do not permit the following of EV signals from specific tissues when they are under AD proteotoxic-stress. I have generated a transgenic C. elegans AD model that has muscle specific expression of the pathogenic human Tau protein. The protein coding sequence was designed to use optimized codons to ensure high expression of the transgene. I have also generated transgenic nematode lines that express versions of known transmembrane proteins with small affinity tags in a tissue specific manner. The small affinity tags on the proteins make it possible to specifically pull down the EVs from designated tissues through standard immunohistochemistry techniques. The abundance of tissue-specific EV protein and RNA cargos from transgenic lines with or without human Tau have then been quantified using LC-MS-MS and RNAseq analyses, and parsed and condensed into a MySQL database via a C# program. The database allows for simple searching through large amounts of MS data, making data analysis more efficient and effective. Thus the methodology and tools I develop in this project could become a promising new approach for identifying novel therapeutic gene targets and biomarkers of AD stress.

Hydra vulgaris are some of the simplest animals with neurons. They only have two thin, near transparent layers of tissue: myo-endodermal and myo-ectodermal layers. Interspersed in each layer is a network of neurons known as the nerve net. All cells in the animal are constantly renewed, which allows Hydra to regenerate after being cut in pieces or dissociated into single cells, though how the nerve net regenerates has not been well studied. Each cell in the animal can be examined simultaneously due to the animals’ small size and simple, translucent body pattern. Hydra also exhibit stereotypical (regular and defined) behaviors. This makes Hydra great models for examining simple signaling pathways from which the complex pathways in vertebrates derive. The Hydra nerve net is composed of circuits that coordinate the behavior of the animal. The most obvious are the contractile burst (CB) and rhythmic potential (RP) circuits. I selectively blocked these circuits to understand how they drive behavior. It has been found that N-[1-(2-phenylethyl)-3-piperidinyl]-1-benzofuran-2-carboxamide (E9), affects Hydra behavior, appearing to block the CB circuit but leaving others, including the RP circuit, uninhibited (unpublished data, Woods Hole MA). I worked to understand the affinity and response rate of this molecule to Hydra by establishing a dose response curve for E9 on Hydra. To determine effective concentrations of E9, I imaged animals in serial concentrations ranging from 3uM-300uM. I then identified the response of neural circuit firing patterns to varying concentrations of E9 by applying this technique to animals that express GCaMP, a protein that fluoresces when bound to calcium, in neurons. I found that 30uM is the lowest concentration of E9 sufficient to block the CB circuit. This research provides a tool for studying the link between circuits and behavior, and allows us to characterize how behaviors depend on identified circuits.

Alzheimer’s disease is a neurodegenerative disease that results in deterioration of memory and cognitive function. One of the hallmarks of Alzheimer’s disease (AD) is the formation of tangled fibrils of the Tau protein. Tau is a microtubule-associated protein found in healthy neurons; in the disease state, it can aggregate and impair normal neuronal functions. Caenorhabditis elegans is a powerful genetic model that has been used to elucidate the cellular and genetic pathways that are impacted by AD-associated proteotoxic stress. All previous C. elegans Tau models have neuronal specific expression. However, neurons are resistant to RNAi. Therefore, we generated a novel transgenic C. elegans model of Tau hyperphosphorylation that has been codon-optimized to express Tau in body wall muscles instead of neurons. Two models were developed: an overexpression (OE) line and a single copy insert (SCI) line. We measured the animal’s health with age in a series of phenotypic assays: egg-laying, growth, movement, paralysis, and lifespan analysis. The OE Tau line displayed a significantly lower egg laying rate, developmental delay by approximately 1 day, and significantly reduced speed in comparison to synchronized N2 populations. The SCI Tau line displayed a significantly lower egg laying rate, smaller adults, and no significant reductions in speed in comparison to synchronized N2 populations. These phenotypic characteristics provide a quick, robust metric by which to measure Tau toxicity with age. The muscle expression opens up the possibility of genome wide RNAi screening to identify the genetic pathways underlying cellular responses to Tau toxicity. We will be screening candidate genetic suppressors of Tau toxicity using feeding RNAi. These experiments could point to genetic targets for future genetic therapies for AD.

With the aim of better understanding developmental disorders, researchers have expended great energy to elucidate the mechanisms that regulate the development of the peripheral nervous system (PNS). Larval Drosophila melanogaster presents a highly useful model organism for studying PNS development due to the larva’s rapid development and neurological parallels to mammals. Thus, to investigate the development of pain-sensing neurons in the PNS, we studied nociceptive class IV dendrite arborization (C4da) neurons in larval D. melanogaster. Randomly inducing mutations with the mutagen ethyl methanesulfonate ultimately identified a miRNA, miR-14, that when absent perturbs C4da neuron development. Confocal microscopy revealed that miR-14 knockout results in greater detachment of neurons from the extracellular matrix (ECM) than those in wildtype specimens. Furthermore, imaging showed that miR-14 knockout leads to far greater ensheathment of c4da dendrites in the epithelial cell-cell junctions than observed in wildtype specimens. These observations suggest that miR-14 plays a critical role in directing proper C4da neuron development. Additionally, relative to wildtype specimens, miR-14 mutants demonstrated significantly increased responsiveness to numerous stimuli including noxious touch, noxious chemicals, and blue light but not noxious heat. These findings indicate that miR-14 influences pain sensitivity in a modality-specific manner. Finally, analysis of miR-14 mutant transcriptomes by RNA-seq demonstrated decreased expression of innexin and integrin—proteins associated with epithelial cell-cell junction formation and ECM adhesion, respectively. These results suggest that miR-14 regulates proteins that enable proper attachment of C4da neurons to the ECM and prevent abnormal C4da ensheathment in epithelial cell-cell junctions. Taken together, these data support the hypothesis that miR-14 regulates sensitivity to pain by controlling C4da neuron-epithelial interactions.

Self-assembling protein nanocages are hollow macromolecular structures that can be loaded with a desired payload. With proper functionalization they are readily endocytosed by specific cells and employable as targeted drug delivery vehicles. The delivery of a given therapeutic into a cell cytoplasm requires simultaneous disassembly of the nanocage and endosomal lysis. To this end, scientists at the Baker lab have engineered icosahedral nanocages using pH sensitive components that disassemble at low pH and disrupt endosomal membranes. However, current designs appear to be forming hyper-stable, partially assembled nanocage intermediates. We seek to improve the assembly kinetics of our 2 component nanocage by using the Rosetta macromolecular modelling suit to redesign current nanocage components. We hypothesize that weaker interactions between nanocage subunits will enhance assembly cooperativity and avoid formation of hyperstable intermediates. To test this hypothesis, we decreased the affinity between components by truncating portions of cage subunits which directly interact. We also modified charge distribution, and altered connections (protein loops) within subunits to simplify assembly. Designs are screened in varying expression conditions by transforming plasmids into Escherichia coli (E. coli) and purifying through immobilized metal affinity chromatography (IMAC) and Size Exclusion Chromatography (SEC). Finally, in vitro assembly of nanocage components is conducted to test potentially ideal stoichiometric equivalents for assembly. We expect that upon decreasing the affinity between individual cage components, we will observe efficient assembly of the complete nanocage. Successful design of a self-assembling nanocage with pH sensitive components will provide a secure intracellular delivery mechanism for targeted delivery of a wide range of therapeutics. Additionally, it will increase our understanding of the kinetics and thermodynamics of multiprotein complex assembly.