2026 Theses Doctoral

Mutations to the RNA-binding protein PUM1 play unique roles in various neurological disorders

Cabaj, Maximilian

PUM1 is an RNA-binding protein which represses mRNA targets bearing a specific 8-nt sequence in their 3’UTR. It is critical to stem cell fate, embryogenesis, and, as has been recently described, neuronal function. Deletion or mutation to this gene has been linked to several different mammalian neurological disorders. Pum1 knockout in mice causes motor incoordination, hyperactivity, and cerebellar degeneration, symptoms akin to spinocerebellar ataxia-type 1 (SCA1). Further, the SCA1 causal gene, Atxn1, is upregulated in these animals, suggesting that even changes to wildtype abundance of Atxn1 may be enough to drive disease. Pum1 has likewise been implicated in regulating the abundance of other neurological disease risk genes, like Snca, which codes for the Parkinson’s disease protein, α-Synuclein.

In people, single nucleotide polymorphisms (SNPs) to PUM1 cause either of two diseases: the first, PUM1-associated developmental delay and seizures (PADDAS), is an early onset, severe neurodevelopmental condition; the second, PUM1-related cerebellar ataxia (PRCA) is a late onset, pure ataxia. The most prevalent causal variant for PRCA (PUM1-T1035S) is located within the functional RNA binding domain, whereas the most common PADDAS-causing mutation (PUM1-R1147W) is located outside the domain. In fibroblasts, it was found that the R1147W isoform destabilizes native protein-protein interactors but maintains RNA-binding ability, whereas T1035S maintains interactors but loses mRNA binding ability. However, fibroblasts are not neurons, and neuropathogenicity cannot necessarily be extrapolated from these data.

Together, these data point towards two different biological processes affected in Pum1-related disorders: lost RNA-binding ability (Pum1-T1035S), lost protein-binding ability (Pum1-R1147W), or both (Pum1 deletion). In the present study, I expand our understanding of these Pum1 biological states by using three discrete mouse models of each. A Pum1 knockout model is used to investigate an established neurological disease risk gene, Snca. Next, novel mouse models of PADDAS and PRCA are used to investigate RNA- vs. protein-binding qualities of mutant Pum1 in each disease state.

In chapter 1, I introduce the landscape of RNA-binding protein biology vis-à-vis neurodevelopment or disease. Likewise, I describe key aspects of Pum1 biology, which include a potential regulatory role for the PD risk gene, Snca. In chapter 2, I recognize the molecular, computational, and neurobehavioral techniques used as part of these investigations. In chapter 3, I identify a role for Pum1 in regulating Snca in both human cells and the mouse brain, as well as describing a possible disease mechanism in miR-7-associated alpha-Synuclein regulation.

Further, I identify PD patients bearing PUM1 variants—some of which lose SNCA regulatory functions in vitro. In chapter 4, I robustly characterize two new mouse models of SCA47 disease subtypes, and use them to identify brain-specific changes to the transcriptome, proteome, and interactome. The findings presented in this dissertation describe new regulatory targets for Pum1 as well as identify novel biological functions for Pum1 in the brain.