The research of my thesis focused on the use of transcriptomic open discovery approaches to interrogate the molecular basis of two distinct yet related neurological disorders that are both associated with cognitive decline, Temporal Lobe Epilepsy and Alzheimer’s Disease. Interestingly, a potential role for compromised synaptogenesis early in disease was common to both, as was the direct role that neurons may play in brain inflammatory processes involving glia. Temporal lobe epilepsy (TLE) is a progressive disorder mediated by pathological changes in molecular cascades and hippocampal neural circuit remodeling that results in spontaneous seizures and cognitive dysfunction. Targeting these cascades may provide disease-modifying treatments for TLE patients. Janus Kinase/Signal Transducer and Activator of Transcription (JAK/STAT) inhibitors have emerged as potential disease-modifying therapies; however, a more detailed understanding of the contribution of JAK/STAT signaling to epileptogenesis is required to increase the potential therapeutic efficacy and reduce adverse effects associated with un-targeted JAK/STAT inhibition. With our collaborators, my lab developed a mouse line in which tamoxifen treatment conditionally abolishes STAT3 signaling from forebrain excitatory neurons (nSTAT3KO). Seizure frequency (continuous in vivo electroencephalography) and memory (contextual fear conditioning and motor learning) were analyzed in wildtype (Wt) and nSTAT3KO mice after intrahippocampal kainate (IHKA) injection as a model of TLE. Selective STAT3 KO in excitatory neurons reduced seizure progression and hippocampal memory deficits without reducing the extent of cell death or mossy fiber sprouting induced by IHKA injection. In my thesis, RNA was extracted from harvested hippocampi 24 h after IHKA and libraries were prepared for bulk RNA-sequencing (70–80 million reads/sample) using the NextSeq 500 Illumina system. 3190 genes were differentially expressed in Wt mice injected with KA vs saline (fold change |1.5|, FDR=<0.05). Ingenuity Pathway Analysis (IPA) revealed significant enrichment in 2 overarching sets of pathways: 1) those related to synaptic signaling and 2) those related to inflammation. As expected, many of the IHKA-induced genes were previously associated with epilepsy or seizure disorders (260 for Seizure Disorder, 267 for Epilepsy or Neurodevelopmental Disorder), and Seizure Disorder had the highest activation score in Neurological Disease based on gene expression patterns. Interestingly, a closer analysis of the IHKA-induced gene set revealed an enrichment of STAT3-associated genes (216), most of which were upregulated by IHKA. Compared to the 3190 Differentially Expressed Genes (DEGs) between IHKA and saline-injected Wt mice 24 hours after SE, more than half of these DEGs (1609) were rescued when comparing IHKA-injected nSTAT3KO mice and saline-injected Wt mice, indicating a significant rescue of gene expression when nSTAT3 is absent in excitatory neurons. While nSTAT3 KO influences the expression of genes in many different pathways, including the reversal of genes whose expression was inhibited in pathways of learning and memory by IHKA, the greatest surprise came from the predicted regulatory control over microglial function. nSTAT3KO mice displayed the greatest number of rescued DEGs compared to IHKA-injected WT mice in pathways that regulate inflammation and ion transport, and while inflammation was an expected response to IHKA, we were surprised to find evidence for its rescue in nSTAT3 KO mice. We also interrogated the expression of the Alzheimer’s disease genome as modeled using a rat model (TgF344-AD ) of familial AD that allows for behavioral and molecular characterization of AD, and expresses an endogenous pathogenic form of tau in addition to Abeta oligomers and plaques. AD is a neuropsychiatric disorder characterized initially by short term memory loss and disorientation, followed by declining cognitive functioning, and eventually, death. Widespread failure of 99% of AD drugs that make it to clinical trials has led to renewed interest in early signatures of disease in hopes of altering disease trajectory through early intervention. Key to such efforts is capturing a molecular window into AD at its earliest stages. The TgF344-AD rat shows overt pathology (including Aβ plaques, frank neuronal loss, and endogenous tau pathology) at 16 and 26 mo, but only to a very limited extent at 6 mo (Towne, 2013). Thus, in my thesis research, we set out to uncover any cell-type specific transcriptomic alterations that may be present in advance of major behavioral deficits or appearance of pathology, given that a strong body of literature suggests a long pre-symptomatic stage of illness in which subtle abnormalities may be present. 10x Genomics’ v3 gene expression assays were used to perform snRNA-seq on freshly dissected hippocampi from 6 mos, 9 mos and 19 mos littermate pairs of Tg and Wt rats (n=16 for 6 months and 9 months, with 8 for 19 months). ~2000 cells/subject were collected, and cDNA libraries were sequenced to a depth of ~120k reads/nuclei. Interestingly, data analysis revealed wide-scale gene changes in dentate granule cells (DGCs) and non-DGC excitatory neurons (Excit Ns) at 6 mos, suggestive of a significant decrease in synaptogenesis in Tg vs their Wt littermates, as well as small increases in cholesterol biosynthesis in the Tg rats in these cell types. By 9 months, some differentially expressed genes were observed across genotype in classes of glial cells, but the strongest impact on gene expression could still be seen in Excit Ns and DGCs, which continued to display evidence of decreased synaptogenesis, though to a lesser extent than at 6 mos. Interestingly, 9 mos Tg rats displayed an even stronger upregulation in genes related to cholesterol biosynthesis than 6 mos for both DGCs and Excit Ns. At 19 months, cholesterol and steroid biosynthesis were amongst the top biological pathways enriched for in Excit Ns and Inhibitory neurons of the Tg, to an even greater extent than changes in synaptogenesis. Altogether, our results suggest the transcriptional basis for a profound suppression of synapse formation or maintenance during early stages of illness in the TgF344-AD rat model, as well as abnormalities in neuronal cholesterol biosynthesis. Given that cholesterol is a key component of plasma membranes and lipid rafts, structures needed for the generation of new synapses and the stability of their receptor populations, it may be that deficiencies in the available cholesterol of Tg neuronal cells is leading to the impaired synaptogenesis in these cell types. Future work will focus on identifying whether these transcriptional alterations can be detected at even earlier time points, whether they are prescient for changes at the membrane in vivo that are correlated with memory impairment, and whether they are related to the alterations in the genome seen in our acquired epilepsy models, suggesting a common theme for the brain’s genomic response to injury of the hippocampus. / 2025-02-12T00:00:00Z

Long non-coding RNAs (lncRNAs) are key regulators of gene expression, playing crucial roles in biological processes across many species, tissues and diseases. The liver is a highly responsive organ in which large changes in gene expression are perpetuated by a myriad of internal and external stimuli; as such, the liver makes an ideal system in which to study lncRNAs. Global patterns of expression, maturation and localization were established for both lncRNA and protein-coding gene (PCG) transcripts across five subcellular compartments in male and female mouse liver, both with and without exposure to TCPOBOP, a direct agonist of the nuclear receptor CAR. In contrast to PCGs, lncRNAs showed very strong enrichment for tight chromatin binding, which increased the sensitivity for lncRNA detection and facilitated discovery of many novel sex-biased and xenobiotic-responsive lncRNAs. These findings helped identify candidate regulatory lncRNAs based on their co-localization within topologically associating domains, or their transcription divergent or antisense to PCGs associated with pathways linked to liver physiology and disease. The liver cell type-specific expression of lncRNAs and PCGs was assessed by single nucleus RNA-seq (snRNA-seq). Liver sexual dimorphism was largely restricted to hepatocyte populations, where many sex-biased genes exhibited zonated expression. Changes in lncRNA and PCG expression following exposure to endogenous hormones (growth hormone) and exogenous chemicals (TCPOBOP) was assessed, identifying cell cluster-specific perturbations to native sex-bias and hepatocyte zonation-dependent gene expression, and highlighting the interconnectedness between liver sexual dimorphism and zonation of the hepatic lobule at the single nuclei level. Finally, an in vivo method for epigenetic reprogramming of lncRNAs using a dual adeno-associated virus delivery system was utilized to knockdown two TCPOBOP-inducible lncRNAs in mouse liver. The knockdown phenotype of one of these lncRNAs, established by snRNA-seq, suggests it plays a functional role in regulating cholesterol metabolism and transport, triglyceride catabolism, and pyruvate metabolism in mouse liver. Together, these studies characterize hepatic lncRNA expression patterns, on both the sub-cellular and single cell levels, and present a strategy for interrogating the roles of specific lncRNAs in liver tissue in vivo. / 2023-01-18T00:00:00Z