Friedreich ataxia (FRDA) is a neurodegenerative disorder that is inherited in an autosomal recessive pattern. The most common FRDA mutation is hyperexpansion of a GAA triplet repeat sequence in the first intron of the affected gene, frataxin (FXN), resulting in decreased frataxin protein expression. The hyperexpanded GAA repeats can adopt unusual DNA structures and induce aberrant epigenetic changes leading to heterochromatin mediated gene silencing. Several epigenetic changes, including increased levels of DNA methylation, histone modifications, repressive chromatin formation and elevated levels of non-coding RNA have been reported in FRDA. It has been reported that a novel FXN antisense transcript (FAST-1), is present at higher levels in FRDA patient-derived fibroblasts and its overexpression is associated with the depletion of CTCF, a chromatin insulator protein, and heterochromatin formation involving the critical +1 nucleosome. Previously, characteristics of FAST-1 were investigated in our lab and a full-length FAST-1 transcript containing a poly (A) tail was identified. To investigate any possible effects of FAST-1 on FXN expression, I first overexpressed this FAST-1 transcript in three different non-FRDA cell lines and a consistent decrease of FXN expression was observed in each cell type compared to control cells. I also identified that FAST-1 copy number is positively correlated with increased FAST-1 expression, which in turn is negatively correlated with FXN expression in FAST-1 overexpressing cells. Additionally, we found that FAST-1 overexpression is associated with increased levels of DNA methylation at CpG sites U6 and U11 of the FXN upstream GAA repeat region, together with CTCF depletion and heterochromatin formation at the 5'UTR of the FXN gene. To further investigate the role of FAST-1 in FXN gene silencing, I used a small hairpin RNA (shRNA) strategy to knock down FAST-1 expression in FRDA fibroblast cells. I found that knocking down FAST-1 increases FXN expression, but not to the level of control cells. Lastly, I investigated the pattern of FAST-1 expression and histone modifications at the FXN transgene in our new FRDA mouse model, designated YG8LR. The YG8LR mice showed decreased levels of FXN expression and H3K9ac and increased levels of FAST-1 expression and H3K9me3. Our data suggest that since FAST-1 is associated with FXN gene silencing, inhibition of FAST-1 may be an approach for FRDA therapy.

Thesis (Ph. D.)--University of Washington, 1999. / Vita. Includes bibliographical references (leaves 107-120).

Abstract Fragile X syndrome, the most common inherited form of mental retardation syndrome, is caused by an expansion of the CGG trinucleotide repeat in the 5' UTR of the FMR1 gene, with concurrent hypermethylation of the region, which represses FMR1 expression. The syndrome is associated with the folate-sensitive chromosomal fragile site at Xq27.3 (FRAXA), where the gene responsible for the syndrome was first localized by linkage analysis using RFLP markers. In this study the linkage relationships of the RFLP markersat Xq27-28 and the characteristics of the CGG repeat expansion were investigated in northern Finnish fragile X families and molecular diagnostic methods were applied in order to improve diagnosis of the syndrome. Furthermore, the origin of fragile X mutations in the northern part of Finland was studied by haplotype analysis. Linkage studies were performed in 34 northern Finnish fragile X families/pedigrees using a total of 15 RFLPs (defining 11 loci). A refined genetic map around FRAXA including five RFLP markers having recombination fractions of 0.04 or less with FRAXA was obtained in an international study of 112 affected families, containing linkage data on twelve northern Finnish families. Linkage analysis significantly improved carrier detection in fragile X families compared with previous cytogenetic methods used in diagnosis. The most efficient RFLP-based protocol for carrier detection was proposed, which is based on use of the most adjacent markers and a minimum number of restriction enzymes. CGG repeat expansion of the FMR1 gene was investigated in original families collected for linkage studies and additional new ones. Large CGG repeat expansions (Δ > 500 bp) with concomitant methylation of the adjacent CpG island, i.e. full mutations, were found to be associated with mental retardation completely in males, but only 50% of the females having a full mutation were mentally impaired. Premutations (Δ < 700 bp) were found in healthy carriers. There was a size range of Δ = 500 to 700 bp, where the expansions could be either abnormally methylated or non-methylated, and it appeared that methylation is more important in determining the phenotype than the exact size of an expansion. Instability of the enlarged CGG repeats was detected, leading preferentially to size increases in successive generations. The instability of premutations was found to be stronger and the size increases larger in maternal than in paternal transmissions, and transition to a full mutation occurred only in female transmissions. In addition, the size of a maternal premutation was shown to have an important influence on the risk of its transition to a full mutation when transmitted. The critical premutation size leading invariably to full mutation in the offspring was found to be between Δ = 175 to 200 bp. In one of the studied families a rare contraction of a paternal premutation to a normal CGG repeat number in one of the daughters and further in her son was detected. Direct mutation analysis including measurement of the CGG repeat size and hypermethylation allowed unambiguous diagnosis of carriers and affected individuals in most cases. Haplotype analysis using two tightly linked microsatellite markers flanking the CGG repeat mutation was performed in 60 unrelated northern and eastern Finnish fragile X families. A significant difference was found in allelic and haplotypic distributions between normal X and fragile X chromosomes. A single haplotype, which was present only in 8% of the normal X chromosomes, accounted for 80% of the fragile X chromosomes. This enrichment of one fra(X) mutation in the Finnish population suggests founder effect.

Microsatellite repeats are a phenomenon found in DNA where a short sequence, usually 1-6bps, is repeated dozens to hundreds of times. Microsatellite repeats that are able to be transcribed are termed expanded tandem repeat-containing RNA (xtrRNA) [1]. xtrRNA have been associated with many diseases, such as Amyotrophic Lateral Sclerosis (ALS) and Frontotemporal Dementia (FTD), which are both caused by a repeat in the C9ORF72 gene. Recent research has been focused on trying to provide treatments for patients with these diseases. This study focuses on creating a drug screening process for therapeutics targeting transcription by stopping or slowing the transcription of C9ORF72 repeat expansions. One project has focused on interrupting the interaction of two transcription factors, SUPT5H and SUPT4H1, to slow transcription. Another project has focused on slowing transcription by using transcriptional inhibitors or nucleoside analogs at low concentrations. Our hypothesis is that if transcription rates are slowed enough, pausing or arrest of RNA polymerase will be induced at complex sequences, including GC-rich regions and repeats. This should reduce synthesis of xtrRNA and provide a starting point for therapeutic development.

More than 40 neurological diseases are known to be caused by large expansions oftandem repeat sequences scattered throughout the human genome in introns, exons and untranslated regions. The GGGGCC (G4C2) repeat expansion located in the first intron of the C9ORF72 gene is the most common genetic cause of amyotrophic lateral sclerosis and frontotemporal dementia (ALS/FTD). In C9 FTD/ALS, expanded transcripts are known to aggregate and accumulate in the cell nucleus, sequestering RNA binding proteins. Other expanded RNA species are exported to the cytoplasm to undergo a non-canonical form of translation termed ‘repeat-associated non-AUG (RAN) translation’. RAN translation leads to the production of toxic polydipeptide repeat proteins in the absence of a canonical AUG start codon. This dissertation will highlight new mechanistic features of translation across the G4C2 repeat expansion, identify a potential therapeutic for C9 FTD/ALS using RNAi and develop a cellular system to explore the G4C2 repeat RNA lifecycle. First, we demonstrate that increasing G4C2 repeat expansion size results in suppression of translation from both canonical and non-canonical start codons, suggesting that large polydipeptide repeats are rarely fully translated. We further find that initiation does not occur from within the repeat expansion, relying on upstream sequence for initiation. However, some reading frames are prone to substantial frameshifting, such as poly-GA. We also show that a bias in ii codon usage efficiency contributes to previously observed variations in the levels of each polydipeptide. Our results support and extend previous studies by identifying two new mechanisms that bias production of poly-dipeptides toward poly-GA in C9 FTD/ALS. Further, we generated central mismatch-containing short hairpin RNAs (shRNAs) targeting the G4C2 repeat expansion to reduce aggregation or block translation of repeatcontaining transcripts. Iterative design was able to improve shRNA processing efficiency and cellular abundance, yet they were unable to reduce nuclear RNA foci in patient-derived cells. Despite this, we show preliminary data suggesting that these shRNAs are able to target cytoplasmic repeat-containing transcripts and resulting in a reduced translation of poly-GP. Finally, we optimized the previously published RNA-protein interaction detection (RaPID) technique, which uses proximity dependent labelling by a mutant biotin ligase and mass spectrometry for protein identification in living cells, to identify proteins interacting with the G4C2 repeat expansion. We embedded the box B RNA hairpin between G4C2 repeats and tested the ability for λN fused to a biotin ligase mutant, BASU, to specifically bind the box B hairpin in vitro. We show that 6 repeats each side of the hairpin combined with an extended hairpin stem promotes specific binding of the λN-BASU fusion protein and is likely to be successful in cells. C9 FTD/ALS is a currently incurable neurodegenerative disorder largely due to the limited understanding of disease mechanism. This dissertation demonstrates new mechanisms of translation across the G4C2 repeat expansion that results in toxic DPR production while also developing a nucleic acid therapeutic for long-term treatment of C9 FTD/ALS and further developing systems to explore RNA-mediated toxicity in cells.

Friedreich ataxia (FRDA) is an autosomal recessive neurodegenerative disorder, caused by a GAA repeat expansion mutation within intron 1 of the FXN gene, resulting in reduced level of frataxin protein. Normal individuals have 5 to 40 GAA repeat sequences, whereas affected individuals have approximately 70 to more than 1000 GAA triplets. Frataxin is a mitochondrial protein involved in iron-sulphur cluster and heme biosynthesis. The reduction in frataxin expression leads to oxidative stress, mitochondrial iron accumulation and consequential cell death with the primary sites of neurons of the dorsal root ganglia and the dentate nucleus of the cerebellum. FRDA, which is the most common inherited ataxia, affecting 1:50,000 Caucasians, is characterised by neurodegeneration, cardiomyopathy, diabetes mellitus and skeletal deformities. To investigate FRDA molecular disease mechanisms and therapy, several human FXN YAC transgenic mouse models have been established: Y47R, containing normal-sized (GAA)9 repeats; YG8R and YG22R, which initially contained expanded GAA repeats of 90-190 units and 190 units, respectively, but which have subsequently been bred to now contain expanded GAA repeats of 120-220 units and 170-260 units, respectively, and YG8sR (YG8R with a small GAA band) that was recently generated from YG8R breeding. To determine the FXN transgene copy number in the enhanced GAA repeat expansion-based FRDA mouse lines, a TaqMan qPCR assay was developed. The results demonstrated that the YG22R and Y47R lines had a single copy of the FXN transgene while the YG8R line had two copies. The YG8s lines showed less than one copy of the target gene, suggesting potential deletion of the FXN gene. Single integration sites of all transgenes were confirmed by fluorescence in situ hybridisation (FISH) analysis of metaphase and interphase chromosomes. However, in the YG8s line, at least 25% of the YG8s cells had no signals, while the remaining cells showed one signal corresponding to the transgenic FXN gene. In addition, the analysis of FXN exons in YG8s rescue mice by PCR confirmed the presence of all FXN exons in these lines, suggesting the incidence of somatic mosaicism in these lines. Extended functional analysis was carried out on these mice from 4 to 12 months of age. Coordination ability of YG8R, YG8sR and YG22R ‘FRDA-like’ mice, together with Y47R and C57BL6/J wild-type control mice, was assessed using accelerating rotarod analysis. The results indicated a progressive decrease in the motor coordination of YG8R, YG22R and YG8sR mice compared to Y47R or C57BL6/J controls. Locomotor activity was also assessed using an open field beam-breaker apparatus followed by four additional functional analyses including beam-walk, hang wire, grip strength and foot print tests. The results indicated significant functional deficits in the FRDA mouse models. Glucose and insulin tolerance tests were also conducted in the FRDA mouse models, indicating glucose intolerance and insulin hypersensitivity in the aforementioned lines. To investigate the correlation between the FRDA-like pathological phenotype and frataxin deficiency in the FRDA mouse models, frataxin mRNA and protein levels as well as somatic GAA repeat instability were examined. The results indicated that somatic GAA repeats increased in the cerebellum and brain of YG22R, YG8R and YG8sR mice, together with significantly reduced levels of FXN mRNA and protein in the liver of YG8R and YG22R compared to Y47R. However, YG8sR lines showed a significant decrease in FXN mRNA in all of the examined tissues compared to Y47R human FXN and C57BL6/J mouse Fxn mRNA. Protein expression levels were also considerably reduced in all the tissues of YG8sR mice compared to Y47R. Subsequently, the telomere length of human and mouse FRDA and control fibroblasts was assessed using qPCR and Q-FISH. The results indicated that the FRDA cells had chromosomes with relatively longer telomeric repeats in comparison to the controls. FRDA cells were screened for expression of telomerase activity using the TRAP assay and a quantitative assay for hTERT mRNA expression using TaqMan qRT-PCR. The results indicated that telomerase activity was not present in the FRDA cells. To investigate whether FRDA cells maintained their telomeres by ALT associated PML bodies (APBs), co-localisation of PML bodies with telomeres was assessed in these cells using combined immunofluorescence to PML and Q-FISH for telomere detection. The results demonstrated that the FRDA cells had significantly higher co-localised PML foci with telomeric DNA compared to the normal cells. Moreover, telomere sister chromatid exchange (T-SCE) frequencies were analysed in the human FRDA cell lines using chromosome orientation FISH (CO-FISH). The results indicated a significant increase in T-SCE levels of the FRDA cell lines relative to the controls. Furthermore, growth curve and population doubling analysis of the human FRDA and control fibroblasts was carried out. The results showed that the FRDA fibroblast cell cultures underwent growth arrest with higher cumulative population doubling compared to the controls. Though, further analysis of telomere length at different passage numbers revealed that the FRDA cells lost telomeres faster than the controls. Finally, the telomere dysfunction-induced foci (TIF) assay was performed to detect DNA damage in the human FRDA fibroblast cells using an antibody against DNA damage marker γ-H2AX and a synthetic PNA probe for telomeres. The frequency of γ-H2AX foci was significantly higher in the FRDA cells compared to the controls. Similarly, the FRDA cells had greater frequencies of TIFs in comparison to the controls, suggesting induced telomere dysfunction in the FRDA cells.

Microsatellites, or simple tandem repeat sequences, occur naturally in the human genome and have important roles in genome evolution and function. However, the expansion of microsatellites is associated with over two dozen neurological diseases. A common denominator among the majority of these disorders is the expression of expanded tandem repeat-containing RNA, referred to as xtrRNA, which can mediate molecular disease pathology in multiple ways. Frontotemporal Dementia (FTD) and Amyotrophic Lateral Sclerosis (ALS) are two fatal neurodegenerative diseases with significant clinical, neurological and genetic overlap thus referred to as C9FTD/ALS. Currently, gaps in the study of the underlying disease mechanisms persist, which can aid in the identification of promising therapeutic approaches. Access to simple models of neurological repeat expansion disease is critical for investigating biochemical mechanisms and for early therapeutic discovery. To better understand the molecular pathology of C9FTD/ALS repeat expansion disorder, we cloned GGGGCC repeats, which are the leading genetic cause of C9FTD/ALS. We employed a recursive directional ligation (RDL) technique to build multiple GGGGCC repeat-containing vectors and validated the cloning to facilitate step-by-step characterization of disease mechanisms at the cellular and molecular level using these vectors. In this study, we also differentiated C9FTD/ALS patient-derived induced pluripotent stem cells (iPSCs) to neural stem cells (NSCs) to be used as model systems. The use of iPSCs and NSCs to reveal important insights into the pathogenic mechanisms and to generate multiple neural cell types presents an excellent opportunity for researchers to model neurodegenerative diseases for cell therapy and drug discovery. We further investigated potential nuclear export mechanisms for C9FTD/ALS xtrRNA. The nuclear export mechanisms of xtrRNA in C9FTD/ALS are not well studied. ASOs and siRNAs were employed to knockdown genes of interest to study their involvement in the nuclear export of xtrRNA. We saw promising results on knockdown of TorsinA involved in nuclear export of xtrRNAs, corroborated by a substantial increase in the average number of xtrRNA foci in the nucleus. Our initial study provides evidence that TOR1A may be involved in the nuclear export of aberrant C9FTD/ALS repeat-containing RNAs. Due to the lack of reliable and robust assays to detect RAN translation products, the effect of the knockdown of TorsinA in these cell lines still remains to be explored. But the current study lays the groundwork for a deeper understanding of the less-studied nuclear export mechanisms in C9FTD/ALS and could reveal new therapeutic approaches to selectively block the nuclear export of xtrRNA through the use of RNAi and ASOs. The insights gained from this study will help us understand future events in the xtrRNA life cycle such as repeat translation mechanisms.

• Argonaute proteins are important effectors in RNA silencing pathways, but they must interact with other machinery to trigger silencing. Ago hooks have emerged as a conserved motif responsible for interaction with Argonaute proteins, but little is know about the sequence surrounding Ago hooks that must restrict or enable interaction with specific Argonautes. • Here we investigated the evolutionary dynamics of an Argonaute-binding platform in NRPE1, the largest subunit of RNA Polymerase V. We compared NRPE1 sequences from more than 50 species, including dense sampling of two plant lineages. • This study demonstrates that the Argonaute-binding platform of NRPE1 retains Ago-hooks, intrinsic disorder, and repetitive character while being highly labile at the sequence level. We reveal that loss of sequence conservation is due to relaxed selection and frequent expansions and contractions of tandem repeat arrays. These factors allow a complete restructuring of the Ago-binding platform over 50-60 million years. This evolutionary pattern is also detected in a second Ago-binding platform, suggesting it is a general mechanism. • The presence of labile repeat arrays in all analyzed NRPE1 Ago-binding platforms indicates that selection maintains repetitive character, potentially to retain the ability to rapidly restructure the Ago-binding platform.

Spinocerebellar ataxia type 7 (SCA7) is a hereditary neurodegenerative disorder affecting the cerebellum, pons and retina. SCA7 patients present with gait ataxia and visual impairment as the main symptoms. Anticipation, commonly observed in SCA7 families, is a phenomenon where an earlier age at onset and a more severe progression of disease is seen in successive generations. In order to identify the gene responsible for SCA7, we performed linkage analysis on a Swedish SCA7 kindred. Evidence for linkage of the SCA7 disease locus to a 32 cM region on chromosome 3p12-21.1, between markers D3S1547 and D3S1274, was established. A number of neurodegenerative disorders associated with anticipation are caused by expanded (CAG)n repeats in their respective disease genes. In order to isolate the SCA7 disease gene we, therefore, screened a human infant brain stem cDNA library for CAG repeat containing clones, mapping to chromosome 3. Four candidate clones were isolated and analysed, but could all be excluded as the SCA7 disease gene. In 1997, the SCA7 disease gene was identified and, as expected, shown to harbour a CAG repeat, expanded in SCA7 patients. Analysis of the SCA7 CAG repeat region in Swedish SCA7 patients demonstrated that CAG repeat size was negatively correlated to age at onset of disease. Furthermore, patients with larger repeats presented with visual impairment, whereas patients with smaller repeats presented with ataxia as the initial symptom. SCA7 is the most common autosomal dominant cerebellar ataxia in Sweden and Finland, but rare in other populations. In order to investigate if the relatively high frequency of SCA7 in these countries is the result of a founder effect in the region, a haplotype analysis was performed on all SCA7 families available. All 7 families shared a common haplotype of at least 1.9 cM surrounding the SCA7 locus. In addition, strong linkage disequilibrium was demonstrated for marker D3S1287 closely linked to the SCA7 gene, suggesting a founder effect for the SCA7 mutation in Sweden and Finland. The function of the SCA7 protein, ataxin-7, is not known and it does not show significant homologies to any previously known proteins. In order to gain insight into the function of ataxin-7 we analysed the expression of ataxin-7 in brain and peripheral tissue from SCA7 patients and controls. In brain, expression was found to be mainly neuronal with a nuclear subcellular localisation. Ataxin-7 expression was found throughout the CNS, not restricted to sites of pathology. We also confirmed previously reported findings of neuronal intranuclear inclusions (NIls) in the brains of SCA7 patients. Based on our findings, we conclude that the cell type specific neurodegeneration in SCA7 is not due to differences in expression pattern in affected and non-affected tissue or the distribution pattern of aggregated protein.

Trinucleotide repeat (TNR) instability including repeat expansions and repeat deletions is the cause of more than 40 inherited incurable neurodegenerative diseases and cancer. TNR instability is associated with DNA damage and base excision repair (BER). In this dissertation research, we explored the mechanisms of BER-mediated TNR instability via biochemical analysis of the BER protein activities, DNA structures, protein-protein interaction, and protein-DNA interaction by reconstructing BER in vitro using synthesized oligonucleotide TNR substrates and purified human proteins. First, we evaluated a germline DNA polymerase β (pol β) polymorphic variant, pol βR137Q, in leading TNR instability-mediated cancers or neurodegenerative diseases. We find that the pol βR137Q has slightly weaker DNA synthesis activity compared to that of wild-type (WT) pol β. Because of the similar abilities between pol βR137Q and WT pol β in bypassing a template loop structure, both pol βR137Q and WT pol β induces similar amount of repeat deletion. We conclude that the slightly weaker DNA synthesis activity of pol βR137Q does not alter the TNR instability compared to that of WT pol β, suggesting that the pol βR137Q carriers do not have an altered risk in developing TNR instability-mediated human diseases. We then investigated the role of DNA synthesis activities of DNA polymerases in modulating TNR instability. We find that pol βY265C and pol ν with very weak DNA synthesis activities predominantly promote TNR deletions. We identify that the sequences of TNRs may also affect DNA synthesis and alter the outcomes of TNR instability. By inhibiting the DNA synthesis activity of pol β using a pol β inhibitor, we find that the outcome of TNR instability is shifted toward repeat deletions. The results provide the direct evidence that DNA synthesis activity of DNA polymerases can be utilized as a potential therapeutic target for treating TNR expansion diseases. Finally, we explored the role of post-translational modification (PTM) of proliferating cell nuclear antigen (PCNA) on TNR instability. We find that ubiquitinated PCNA (ub-PCNA) stimulates Fanconi associated nuclease 1 (FAN1) 5’-3’ exonucleolytic activities directly on hairpin structures, coordinating flap endonuclease 1 (FEN1) in removing difficult secondary structures, thereby suppressing TNR expansions. The results suggest a role of mono-ubiquitination of PCNA in maintaining TNR stability by regulating nucleases switching. Our results suggest enzymatic activities of DNA polymerases and nucleases and the regulation of the activities by PTM play important roles in BER-mediated TNR instability. This research provides the molecular basis for future development of new therapeutic strategies for prevention and treatment of TNR-mediated neurodegenerative diseases.