This competing renewal continues the work of our first R01 (awarded in 2019) to better understand why different mutations in the RNA-binding protein PUM1 cause two very different phenotypes. One is a mild, late- onset ataxia, the other is a severe neurodevelopmental syndrome, although both fall under the rubric of Spinocerebellar ataxia type 47 (SCA47). Interestingly, we were studying another SCA (SCA1) when we found that PUM1 regulates the levels of the relevant protein, ataxin1. This was the first indication that PUM1, previously known for its role in developing gametes, is important for neurological function in mice. PUM1 has several mechanisms of action that it can use to regulate different targets, but in the case of ATXN1 it directly binds to the 3'UTR to repress it. PUM1 heterozygous mice therefore develop an adult-onset ataxia very much like SCA1, due to overproduction of ATXN1 in the cerebellum. PUM1 knockout mice are sicker, being small from birth and born at lower Mendelian ratios. Crossing the PUM1 hets with SCA1 knockin mice exacerbates the SCA1 phenotype, while crossing PUM1 hets with ATXN1 hets (who have no phenotype) rescues the PUM1 loss-of-function ataxia. These observations prompted us to look for human patients bearing PUM1 variants, and we initially identified 15 individuals with either deletions or missense mutations. The most severe mutations (deletion or R1147W, reducing PUM1 protein levels by ~50%) cause PADDAS (PUM1-associated developmental delay and seizures); the mildest mutation (T1035S) reduces PUM1 by only 25% and causes adult-onset PRCA (PUM1-related cerebellar ataxia). Although the phenotypic severity tracks with protein dosage, the few PUM1 targets that were known at the time were upregulated to the same extent in cell lines derived from PRCA and PADDAS patients. We therefore hypothesized that whereas the mild PRCA mutation would primarily cause target dysregulation, the more severe PADDAS mutation disrupts PUM1's native interactors in addition to their downstream targets. To test this dual hypothesis, over the past four years we mapped PUM1 targets and interactors in the mouse brain (cortex, cerebellum, hippocampus), finding particularly strong interactions with ubiquitin ligases, mTOR, several RNA-binding proteins (FMRP, AGO2, CNOT1, RBFOX3), and a number of mitochondrial proteins. Our studies in patient-derived cell lines and in vitro experiments seem to bear out our dual hypothesis, but the crucial experiment now is to study PRCA and PADDAS mouse models (which we have in hand) in order to understand the distinct pathogenic pathways for each disease as well as to understand how PUM1 interactions support healthy brain function. We will therefore characterize the PRCA and PADDAS mouse models, create molecular profiles of the targets and interactors that are altered in each model, and determine the role of mitochondrial dysfunction in the broad range of symptoms that occur in PADDAS.