Julie Secombe - US grants

DESCRIPTION (provided by applicant): KDM5 family transcriptional regulators are increasingly recognized as critical players at the interface of development and human disease. Mammalian cells encode four KDM5 paralogs (KDM5A-D), three of which are clinically significant: KDM5A or KDM5B are overexpressed in a number of cancers including breast, colorectal and melanoma, and mutations in KDM5C account for ~3% of X-linked intellectual disability patients. A lack of basic knowledge regarding the mechanisms of KDM5 action has hindered the development of therapies to treat these diseases. The long-term goal of these studies is therefore to dissect the gene- regulatory functions of KDM5 proteins using Drosophila, since the presence of a single KDM5 protein in this organism bypasses the issue of functional redundancy in mammalian cells. Previous data showed that KDM5 proteins can repress transcription via their histone demethylase activity and can activate gene expression by altering histone acetylation. In this proposal, new data demonstrate for the first time that KDM5 also influences gene expression by altering the recruitment of transcription factors to their target promoters. Specifically, KDM5 recruits the oncogenic transcription factor Myc to cell growth gene promoters, and this requires the chromatin-binding PHD motif of KDM5. KDM5 also recruits Foxo to oxidative stress resistance gene promoters, however this occurs in a PHD-independent manner. Based on these data, the first hypothesis of this proposal is that KDM5 affects transcription factor recruitment by more than one mechanism, and that this involves distinct domains of KDM5. Preliminary data also show that the expression of KDM5-Foxo co- regulated targets is reduced in a fly strain harboring an allele associated with severe intellectual disability in humans (kdm5L854F). Because the remaining 12 missense mutations in KDM5C associated with intellectual disability also occur in evolutionarily conserved residues, the second hypothesis is that the corresponding mutations in fly KDM5 will show transcriptional defects. These hypotheses will be tested by pursuing three specific aims: 1) Determine the mechanism by which KDM5 recruits Myc to cell growth genes; 2) Define the mechanism by which KDM5 recruits Foxo to oxidative stress target genes; and 3) Determine the transcriptional and phenotypic defects of kdm5 alleles analogous to human intellectual disability-associated mutations. These analyses are significant because defining how KDM5 functions in context-dependent manner will lead to new strategies for treating malignancies and cognitive phenotypes caused by dysregulation of KDM5 family proteins in humans. The proposal is innovative because it deviates from the current focus on the enzymatic function of lysine demethylase (KDM) proteins by describing two new mechanisms of gene activation by KDM5 that are independent of its enzymatic activity. An additional innovation is the analyses of kdm5 missense alleles in flies that are analogous to mutations found in patients with intellectual disability as a means to dissect the gene- regulatory functions of KDM5.

! A poorly explored potential contributor to aging is the mobilization of endogenous transposable elements (TEs), which can be highly mutagenic and promote genomic instability. The goal of this proposal is to determine the functional link between activation of endogenous TEs by the transcription factor Myc and changes to lifespan and aging. Previous data showed that increasing or decreasing levels of the oncoprotein Myc in the model organism Drosophila reduced or extended lifespan, respectively. New data that forms the basis of this proposal show that Myc activates the expression of a subset of endogenous TEs. Because mobilization of TEs can cause insertional mutagenesis, genome rearrangements and DNA damage, they have been proposed to contribute to tumorigenesis and other phenotypes associated with aging. However, a significant hurdle to understanding the effects of transposon mobilization has been a lack of methods to identify de novo TE insertions, since they are unique to individual somatic cells so are extremely difficult to identify by sequencing bulk tissue DNA. Only a single cell approach such as the one described in this proposal allows the number and distribution of de novo TE insertions that occur with age in somatic cells to be determined. Significantly, whole genome sequencing of multiple individual indirect flight muscle (IFM) nuclei from young and old wildtype flies revealed that the number of de novo transposon insertions increased with age. The three most frequently mobilized TEs were also found to be induced by Myc. However key questions remain unanswered regarding the link between Myc levels, TE activation and aging. For example, it is not known where TEs insert within the genome during aging or the frequency at which they mobilize. Nor is it known how Myc activates the expression of a subset of TEs, or whether their activation is functionally important for aging. The central hypothesis of this proposal is that is that Myc acts as a pro-aging gene by increasing and/or altering the distribution of de novo TE insertions in somatic cells, and that this adversely affects cell and tissue function leading to decreased lifespan. This hypothesis will be tested by pursuing three specific aims: Aim 1) Compare the frequency and genomic distribution of de novo TE insertions in wildtype flies and flies with increased or decreased Myc levels. Aim 2) Define the mechanism by which Myc activates the expression of specific TEs. Aim 3) Determine the functional link between Myc levels, TE activation and changes to lifespan. These analyses are significant because the activation of TEs by Myc could contribute to a range of age-related phenotypes, and suppressing their mobilization may provide a new therapeutic avenue to improve healthspan. The approach is methodologically innovative because it uses novel single cell genomic analyses to identify low abundance TE insertions that cannot be detected using whole tissue or organism approaches. It is conceptually innovative because interventions are tested that will for the first time define the functional contribution of de novo TE insertions to aging.

Abstract Mutations in the lysine demethylase 5 (KDM5) family of transcriptional regulators are found in patients with intellectual disabilities (ID) that show cognitive impairment ranging from mild (IQ 50-70) to severe (IQ < 30). However, the molecular mechanisms by which KDM5 proteins impact neuronal development and function remain unknown, leaving a large knowledge gap and preventing the identification of potential treatments for affected patients. Thus, the long-term goal of our research is to define at the molecular level how KDM5 regulates gene expression patterns necessary for neuronal development and function. We will achieve this using Drosophila because it is an established model organism used to define the molecular basis of human neurodevelopmental disorders. Studies described here use a powerful combination of genetic tools, cell biological analyses and cognitive behavioral assays to dissect KDM5?s gene regulatory activities in neuronal cells at distinct stages of development. In addition to classical loss of function analyses utilizing a newly generated kdm5 null allele and cell type specific inducible RNAi-mediated knockdown assays, we have generated a set of eight fly strains, each of which harbors a mutation in Drosophila kdm5 analogous to a human ID allele. This approach is possible because all disease-associated mutations occur in evolutionarily conserved amino acids. Data generated using these tools lead us to propose the central hypothesis that KDM5 regulates the expression of genes essential for neuronal development and function, and that this is affected by missense mutations associated with intellectual disability. This hypothesis will be tested in three specific aims. The first aim addresses the role of KDM5 in adult brain function by defining the mechanism by which KDM5 activates ribosomal protein (Rp) genes, as de novo translation has an evolutionarily conserved role in learning and memory. The second aim focusses on the role of KDM5 during development by determining the mechanism by which KDM5 functions in larval neuronal stem cells (neuroblasts) to facilitate the subsequent growth and guidance of the neurons required for learning and memory. The third aim defines the temporal and spatial requirements of KDM5 for learning and memory. This work is significant because we will define new mechanisms of gene regulation by KDM5 that are critical for neuronal development and activity in addition to providing insight into the underlying causes of neurodevelopmental disorders.

PROJECT SUMMARY/ABSTRACT ? RESEARCH PROJECT Mutations in the gene encoding the transcriptional regulator lysine demethylase 5C (KDM5C) are found in patients with intellectual disability (ID). While the direct link between loss of function mutations in KDM5C and ID is clear, how KDM5C functions to mediate critical neuronal processes, and therefore the consequence of mutations for mechanisms of IDD, remains unknown. The goal of this proposal is to understand the relationship between KDM5C-regulated gene expression programs and the occurrence of ID and additional comorbid features that are observed in patients. We will achieve this by bringing together a multi-disciplinary research team with expertise in complementary analytical tools and model systems to test two hypotheses. Aim 1 tests the hypothesis that the use of human iPSC-derived in vitro cell models of KDM5C-induced ID will result in the identification of clinically relevant gene expression changes and neuronal functional deficits. One key model system we will use is iPSC-derived cerebral organoids, which recapitulate structural and molecular aspects of fetal brain development and are a critical research tool used to define the underlying cause(s) of neurodevelopmental disorders. Indeed, molecular and cellular studies using in vitro organoid systems allow us to carry out studies in a human cell context that would simply not be possible in vivo. iPSCs and organoids will be generated from two sources: (1) Cells from patients with KDM5C-induced ID from a newly recruited cohort of individuals from which we are generating a genotype-phenotype database; (2) CRISPR-Cas9-mediated gene editing to generate a KDM5C null allele and a published ID allele (KDM5CA388P) that lacks histone demethylase activity using existing iPSC lines generated from typically developing controls. We will use this system to combine morphological, functional and multi-OMICS approaches to define the impact of patient- associated mutations in KDM5C. Aim 2 tests the hypothesis that the regulation of translation efficiency in neurons by the fly homolog of KDM5C is conserved in mammalian systems and that this function is important for cognition. Here we take advantage of fly and mouse animal model systems, both as discovery tools and to test hypotheses regarding possible contributors to the cognitive effects of mutations in KDM5C. Because other inherited forms of ID have altered translation and correcting this deficit has shown promise in mouse models of other ID disorders, we will test whether altered translation similarly plays a key role in a mouse model of KDM5C-induced ID. This work is significant because we will define the etiological links between mutations in human KDM5C and ID. The proposed studies are technologically innovative in the use of complementary model systems and state-of-the-art genomics techniques such as single cell transcriptomics (scRNA-seq). It is also conceptually innovative in proposing a role for translation in KDM5C-induced ID.