Search

Additional file 1: Figure S1. Quality control of total mRNA-seq datasets. (a, b, c) Scatter plot showing the Pearson correlation between two biological replicates in total mRNA-seq datasets in NSCs. (d) Principal component analysis of total mRNA-seq datasets. Figure S2. Validation of differentially expressed genes (DEGs). (a-g) RT-qPCR was performed to validate the expression of DEGs identified by RNA-seq. (*

Additional file 8. Original uncropped blot images.

Connexins are transmembrane proteins that form hemichannels allowing the exchange of molecules between the extracellular space and the cell interior. Two hemichannels from adjacent cells dock and form a continuous gap junction pore, thereby permitting direct intercellular communication. Connexin 36 (Cx36), expressed primarily in neurons, is involved in the synchronous activity of neurons and may play a role in aberrant synchronous firing, as seen in seizures. To understand the reciprocal interactions between Cx36 and seizure-like neural activity, we examined three questions: a) does Cx36 deficiency affect seizure susceptibility, b) does seizure-like activity affect Cx36 expression patterns, and c) does acute blockade of Cx36 conductance increase seizure susceptibility. We utilize the zebrafish pentylenetetrazol (PTZ; a GABA(A) receptor antagonist) induced seizure model, taking advantage of the compact size and optical translucency of the larval zebrafish brain to assess how PTZ affects brain-wide neuronal activity and Cx36 protein expression. We exposed wild-type and genetic Cx36-deficient (cx35.5-/-) zebrafish larvae to PTZ and subsequently mapped neuronal activity across the whole brain, using phosphorylated extracellular-signal-regulated kinase (pERK) as a proxy for neuronal activity. We found that cx35.5-/- fish exhibited region-specific susceptibility and resistance to PTZ-induced hyperactivity compared to wild-type controls, suggesting that genetic Cx36 deficiency may affect seizure susceptibility in a region-specific manner. Regions that showed increased PTZ sensitivity include the dorsal telencephalon, which is implicated in human epilepsy, and the lateral hypothalamus, which has been underexplored. We also found that PTZ-induced neuronal hyperactivity resulted in a rapid reduction of Cx36 protein levels within. 30 minutes and one-hour exposure to 20 mM PTZ significantly reduced the expression of Cx36. This Cx36 reduction persists after one-hour of recovery but recovered after 3-6 hours. This acute downregulation of Cx36 by PTZ is likely maladaptive, as acute pharmacological blockade of Cx36 by mefloquine results in increased susceptibility to PTZ-induced neuronal hyperactivity. Together, these results demonstrate a reciprocal relationship between Cx36 and seizure-associated neuronal hyperactivity: Cx36 deficiency contributes region-specific susceptibility to neuronal hyperactivity, while neuronal hyperactivity-induced downregulation of Cx36 may increase the risk of future epileptic events. Doctor of Philosophy Within the brain, cells (neurons) communicate with each other to pass along information. This communication is important for normal functions of the brain such as learning and memory, muscle movement, etc. Epilepsy is a disease of the brain that is caused by rapid over synchronized communication between cells. This leads to seizures which can include convulsions, loss of attention, and much more. Currently, 30% of patients suffering from epilepsy do not have a treatment option that works for them, it is, therefore, imperative to investigate new targets for treatment in this disease. Connexin36 is a protein in the brain that directly connects cells so they can pass information quickly between them. Connexin36, therefore, might make a good target for treatment. Previous work has aimed to understand this relationship but has been limited in their ability to look at the entire brain at any one time. The goal of this study was to understand the relationship between connexin 36 and brain hyperactivity across the whole brain simultaneously. To understand this relationship, we first determined what happened to brain activity if the protein was missing entirely after exposure to a seizure causing drug. We were asking: How does connexin 36 affect hyperactivity. We found that different regions of the brain responded differently without the connexin 36 protein. This suggests that one size does not fit all, and one must look at the whole brain to understand the effects of the connexin 36 protein. Next, we asked a similar question, but in the opposite direction, how does hyperactivity affect connexin 36? We found, in the short-term, hyperactivity reduced the amount of connexin 36 present in certain regions of the brain. This continued until 3 hours following exposure to the seizure causing drug Pentylenetetrazol (PTZ). Lastly, to determine if this short-term reduction in connexin 36 meant that an individual was more likely to experience hyperactivity. To do this, we used a connexin 36 blocking drug, then introduced the seizure causing drug at different concentrations. We found, at all concentrations, the connexin 36 blocking drug caused significant changes in neuronal activity, depending on the brain regions. Overall, our results showed that connexin 36 plays an important role in hyperactivity and that a short-term reduction in connexin 36 is detrimental, and may contribute to an increase in the possibility of subsequent hyperactivity.

Von Hippel-Lindau (VHL) syndrome is an autosomal dominant predisposition to cancer in neurological tissues, the kidneys, adrenal glands, pancreas, and liver, including neurological hemangioblastoma (HB), pheochromocytoma (PCC), pancreatic neuroendocrine tumors (PNET), pancreatic and renal cysts, and clear cell renal cell carcinoma (ccRCC). The disease process follows Knudson's two-hit model, requiring spontaneous loss or mutation of a normal VHL tumor suppressor allele to induce expression of the disease. VHL syndrome principally involves dysregulation of oxygen sensing pathways including the Hypoxia Inducible Factor (HIF)-Vascular Endothelial Growth Factor-A (VEGF-A) and HIF-Erythropoietin (EPO) pathways. RNA sequencing (RNA-Seq) data from our previously published experiments revealed a potentially novel VEGF-A splice variant with excision of the VEGF Receptor-1 (VEGFR-1)/Flt-1 binding domain, rendering this isoform resistant to native down-regulation. Additionally, phenotypic changes were observed in adult VHL mutant mice, specifically very red appearing extremities with prominently visible vasculature. In order to determine the etiology of this phenotype, we observed red blood cell count, Epo gene expression levels, and arterialization of the blood vessels in these experimental mice as compared to littermate controls. Current research into the VEGF-A isoform is ongoing in the lab, and preliminary evidence for the etiology of the apparent chronic erythema phenotype is inconclusive due to lack of experimental replicates due to COVID-19 quarantine orders. Master of Science Von Hippel-Lindau (VHL) syndrome is characterized by cancer development primarily in the brain and spinal cord, kidneys, adrenal glands, pancreas, and liver. VHL syndrome involves mutations which render the VHL gene dysfunctional. Since the VHL gene's normal role is one of preventing cancer development, sensing oxygen levels, and impacting blood vessel development, it follows that the loss of this gene results in tumor development with a rich blood vessel network. One of the downstream effectors of this process is a signaling molecule called Vascular Endothelial Growth Factor-A (VEGF-A). Our lab found a unique variant of VEGF-A, which may be overactive in the body in the setting of VHL disease. Additionally, we noted that our VHL mutant mice turned very red, and we sought to identify the biological cause of this phenomenon. In order to determine the cause of this redness, we studied red blood cell counts and their regulatory gene (Erythropoietin, EPO), as well as potential blood vessel abnormalities using high-power microscopy.

Gap junctions, comprised of connexin proteins, are essential for direct intercellular electrical, metabolic, and immunological coupling. Connexin43 (Cx43, gene name GJA1) is the most ubiquitously expressed gap junction protein, and Cx43 gap junctions are altered in pathological states including cardiac disease and cancer. The GJA1 mRNA undergoes alternative translation initiation to yield a truncated Cx43 isoform, GJA1-20k, that can regulate gap junction formation. Using epithelial-mesenchymal transition (EMT) as a cellular model of gap junction remodeling, we have demonstrated altered translation initiation of Gja1 as a mechanism by which cellular Cx43 gap junctions can be dynamically regulated. Suppression of Gja1 alternative translation is necessary for Cx43 gap junction loss, and stable expression of GJA1-20k rescues gap junction formation during EMT. To identify regulatory factors acting on the Gja1 mRNA, an MS2 RNA aptamer tagging system was adapted to isolate Gja1 with associated RNA binding proteins. We find the RNA binding protein IMP1 is sensitive to hypoxic stress and complexes with Gja1 mRNA, where it is necessary for alternative translation to generate GJA1-20k. We have demonstrated alterations in translation initiation of the Gja1 mRNA as a critical mechanism by which cells modulate Cx43 gap junctional coupling in changing conditions and identified a novel regulator of this process in mammalian cells. Doctor of Philosophy Communication between cells is necessary for healthy function of organs throughout the body. Gap junctions form conduits through which signals can pass directly between neighboring cells. Many diseases, including cancer and heart disease, involve disturbances in gap junction communication. Connexin proteins are the building blocks of gap junctions, and it was recently demonstrated that smaller fragments of connexins are synthesized by cells by a poorly understood process called alternative translation. Importantly, levels of these connexins fragments can alter gap junction formation. We have used mammalian cells to delineate the mechanism by which this alternative protein translation regulates gap junction formation and generated insight into how such protein synthesis is dynamically regulated. Harnessing this knowledge will inform development of new therapeutics inducing alternative translation to rescue gap junctions, and restore normal communication in pathological conditions.

Despite decades of research and thousands of studies on cardiac electrophysiology, cardiovascular disease remains among the leading causes of death in the United States today. Despite substantially beneficial advances, we have largely shifted cardiovascular disease from an acute to a chronic issue. It is therefore clear that our current understanding of the heart's functions remain inadequate and we must search for untapped therapeutic approaches to eliminate these deadly and costly ailments once and for all. This thesis will focus on the electrophysiology of the heart, specifically the mechanisms of cell-to-cell conduction. Canonically, the understood mechanism of cardiac conduction is through gap junctions (GJ) following a cable-like conduction model. While both experimentally and mathematically, this understanding of conduction has explained cardiac electrical behavior, it is also incomplete, as evidenced by recent conflicting modeling and experimental data. The overall goal of this thesis is to explore a structure modulating an ephaptic, or electric field, cellular coupling mechanism: the GJ-adjacent perinexus, with three specific aims. First, I identified the perinexus – a recently-established structure in rodent myocardium – in human atrial tissue. I also observed a significant tendency for open-heart surgery patients with pre-operative atrial fibrillation to have wider perinexi, indicating a possibly targetable mechanism of atrial fibrillation, one of the costliest, and most poorly-understood cardiac diseases. Next, I developed a high-throughput, high-resolution method for quantifying the perinexus. Finally, I sought to reconcile a major controversy in the field: whether cardiac edema could either be beneficial or harmful to cardiac conduction. Using a Langendorff perfusion model, I added osmotic agents of various sizes to guinea pig hearts and measured electrical and structural parameters. My findings suggest that while cardiac conduction is multifaceted and influenced by several parameters, the strongest correlation is an inverse relationship between conduction velocity and the width of the perinexus. This study is the first to osmotically expand and narrow the perinexus and show an inverse correlation with conduction. Importantly, my conduction data cannot be explained by factors consistent with a cable-like conduction mechanism, indicating once again that the perinexus could be a therapeutic target for a myriad of cardiac conduction diseases. Doctor of Philosophy The ways by which cells in the heart communicate have been studied extensively and are thought to be well-understood. However, despite decades of research, cardiovascular disease is a major problem in the developed world today and we remain unable to develop treatments to truly cure many major cardiac diseases. Because of this lack of clinical success in preventing or treating conditions such as atrial fibrillation, Brugada syndrome and sudden cardiac death, all of which are associated with disruptions in the heart’s electrical communication systems, I have sought to better understand the ways by which cellular communication is achieved. Currently, we think of cardiac tissue to propagate electrical signals as if it was a series of cables, just like the electrical wires over our streets and in our homes. However, we have seen experimental evidence, along with computer simulations, that supports the idea of a second mechanism of cellular electrical conduction. This second mechanism is called ephaptic, or electric field, coupling and relies on changes in charges inside and outside the cell to trigger the action potential – the electrical signal which tells the cell to contract. In order for ephaptic coupling to occur, two main conditions must be met. First, there must be a suitably-sized cleft, or ephapse, between adjacent cells. Models have estimated this space to be between 10-100 nm wide. Second, there must be a large concentration of sodium channels, as sodium ions are primarily used to set off the action potential. The region in which I am most interested is the cardiac perinexus, which is the space immediately adjacent to plaques of connexin proteins which link adjacent cells. The perinexus is both of an appropriate size (we’ve measured it between 10 and 25 nm on average) and rich in sodium channels, making it an ideal candidate to be a cardiac ephapse. In recent years, our lab has shown experimentally that expanding this space can disrupt cardiac conduction and my first study showed that clinically, patients with chronic atrial fibrillation (a-fib) prior to open-heart surgery have wider perinexi than patients without chronic a-fib. No one, however, has been able to demonstrate that narrowing the perinexus would be therapeutic by making it easier for cells to communicate via this ephaptic mechanism. Knowing I would need a better method for measuring the width of huge numbers of perinexi, I then developed a faster, more precise measurement program. Finally, I perfused several osmotic agents – substances which would theoretically draw fluid into or out of various compartments of cardiac tissue – into guinea pig hearts and observed changes to both their electrical behavior and tissue structure. Using my new perinexal measurement program, I found that changing the perinexus was the only factor that could explain the conduction changes I observed with each osmotic agent and that parameters associated with cable theory, such as gap junctional protein expression or interstitial resistance, could not explain conduction changes. Therefore, I have indicated, along with my clinical study, that the cardiac perinexus could be a therapeutic target for preventing, managing, or possibly even curing cardiac conduction diseases.

Rotaviruses (RVs) are medically significant gastrointestinal pathogens and are a leading cause of childhood mortality in many countries. The RV RNA-dependent RNA polymerase, VP1, synthesizes RNA during viral replication only in the presence of another RV protein, VP2, which comprises the innermost core shell layer of the virion. Though these VP1-VP2 interactions are essential for RV replication, the mechanism by which the core shell regulates polymerase activity remains incompletely understood. Here, we sought to identify and characterize specific regions of both VP1 and VP2 that are required for core shell dependent polymerase activity. First, we used bioinformatics approaches to analyze VP1 and VP2 sequence diversity across many RV strains and identify positional locations of critical amino acid changes within the context of known structural domains and motifs. We next tested how the identified sequence differences influenced VP2-dependent VP1 activity in vitro. These data revealed that VP1 and VP2 protein diversity correlates with functional differences between avian and mammalian RV strains. Then, we used these sequential and functional incompatibilities to map key regions of VP1 important for mediating RNA synthesis. To pinpoint critical interacting regions of VP1 and VP2, we used site directed mutagenesis to engineer several modified VP1 and VP2 proteins. Then, we employed an in vitro RNA synthesis assay to test how the introduced mutations influenced VP2-dependent VP1 activity. Altogether, our results revealed several functionally important VP1 residues critical for in vitro VP2-dependent VP1 activity, either individually or in combination with neighboring residues, including E265/L267, R614, and D971/S978/I980. Structural analyses show VP2 interactions at these surface-exposed VP1 sites, which altogether supports a direct contact model of core shell dependent RV polymerase activity. Moreover, recombinant VP1 proteins containing multiple mutations at buried residues were incapable of facilitating RNA synthesis in vitro under the assay conditions, indicating that an extensive intramolecular signaling network exists to mediate VP1 activity. Taken together, these results suggest that VP2 binding at the VP1 surface may induce intramolecular interactions critical for VP1 activity. Overall, results from these studies provide important insight into VP1-VP2 binding interface(s) that are necessary for RV replication. Ph. D. Rotaviruses (RVs) are clinically-significant gastrointestinal pathogens that cause severe diarrhea and dehydration in children. RVs encode a specialized polymerase enzyme, called VP1, which functions to synthesize RNA during viral replication. RNA synthesis activities of VP1 are tightly regulated by another RV protein, VP2, which comprises the innermost core shell layer of the virion. Though these VP1-VP2 interactions are essential for viral replication, the mechanism by which the core shell supports polymerase activity remains poorly understood. Here, we sought to identify and characterize specific regions of both VP1 and VP2 that are essential for polymerase activity in a test tube (i.e., in vitro). First, we analyzed VP1 and VP2 sequence diversity across many RV strains. Then, we tested how the identified sequence differences influenced VP2-dependent VP1 activation in vitro. To pinpoint critical regions of VP1 and VP2, we next engineered and assayed several mutant proteins. Altogether, our results revealed several functionally important residues of VP1 and VP2, which raises new ideas about VP1-VP2 binding interface(s) that are important for viral replication. Moreover, results from these studies may provide a scientific platform for the rational design of next-generation RV vaccines or antiviral therapeutics.

Reproducibility of results in biomedical research is an area of concern that should be paramount for all researchers. Importantly, this issue has been examined for experiments concerning cardiac electrophysiology. Specifically, multiple labs have found differences in results when comparing cardiac conduction velocity (CV) between healthy mice and mice that were heterozygous null for the gap junction (GJ) forming protein, Connexin 43. While the results of the comparison study showed differing extracellular ionic concentrations of the perfusates, specifically sodium, potassium, and calcium ([Na+]o, [K+]o, and [Ca2+]o), there was a lack of understanding why certain combinations of the aforementioned ions led to specific CV changes. However, more research from our lab indicates that these changes can predict modifications to a secondary form of cardiac coupling known as ephaptic coupling (EpC). Therefore the work in this dissertation was twofold, 1) to examine the effects of modulating EpC through perfusate ionic concentrations while also modulating GJC and 2) to investigate the effects of modulating all three of the main ions contributed with cardiac conduction (Na+, K+, Ca2+) and the interplay between them. Firstly I designed and tested changes from the use of 3D printed bath for optical mapping procedures. After verification that the bath did not modify electrophysiological or contrile parameters, I studied the effects of physiologic changes to EpC determinants ([Na+]o and [K+]o) on CV during various states of GJ inhibition using the non-specific GJ uncoupler carbenoxolone (CBX). Multiple pacing rates were used to further modify EpC, as an increased pacing rate leads to a decrease in sodium channel availability through modification of the resting membrane potential. with no to low (0 and 15 µM CBX) GJ inhibition, physiologic changes in [Na+]o and [K+]o did not affect CV, however increasing pacing rate decreased CV as expected. When CBX was increased to 30 µM, a combination of decreasing [Na+]o and increasing [K+]o significantly decreased cardiac CV, specifically when pacing rate was increased. Next, the combinatory effects of cations associated with EpC (Na+, K+, and Ca2+) were tested in to examine how cardiac CV reacts to changes in perfusate solution and how this may explain differences in experimental outcomes between laboratories. Briefly, experiments were run where [K+]o was varied throughout an experiment and the values for [Na+]o and [Ca2+]o were at one of two specific values during an experiment. 30 µM CBX was added to half of the experiments to see the changes in the CV-[K+]o relationship with GJ inhibition. With unaltered GJ coupling, elevated [Na+]o maintains CV during hyperkalemia. Interestingly, both [Na+]o and [Ca2+]o must be increased to maintain normal CV during hyperkalemia with reduced GJ coupling. These data suggest that optimized fluids can sustain normal conduction under pathophysiologic conditions like hyperkalemia and GJ uncoupling. Taken as a whole, this dissertation attempts to shed light on the importance of ionic concentration balance in perfusate solutions on cardiac conduction. Ph. D.