Studies on the modulation of gene expression via 3´UTR editing to model and treat neurological disease

Abstract

It is increasingly clear that gene expression levels play a critical role in human health. For example, genome-wide association studies (GWAS) reveal that at least 93% of diseaseassociated changes are in non-coding genomes. Thus, it is the protein regulation – most likely the levels and timing and site of expression –, not their sequence, which mostly defines human and animal health. However, currently, we can delete genes but increasing their expression without destroying their natural expression site and levels is almost impossible. Yet from the above, it is easy to see that this ability carries major potential in understanding and hence defining treatment for human diseases.

In this thesis, I examine the potential of modifying the 3´ untranslated region (UTR) of endogenous genes to gain control over gene expression levels at the post-transcriptional level. As I and my colleagues find, this allows to retain natural spatial expression but enables modification of expression levels of at least some genes. My studies predominantly focus on glial cell line-derived neurotrophic factor (GDNF), a small secreted protein that elicits function-enhancing and survival-promoting effects on cells that carry receptors for GDNF, called GDNF family receptor α-1 (GFRα1) and RET. I find that modulation of GDNF 3´UTR allows conditional as well as constitutive upregulation of GDNF expression ranging from a few tens of % (muscle) to about 4-fold increase in the brain of the animals where both Gdnf alleles carry 3´UTR modification.

I then asked how GDNF levels vary in various human disease conditions which involve neurons that carry receptors for GDNF. Those are dopamine neurons in schizophrenia, where excessive dopamine in the brain structure called the striatum is believed to drive the disease. Together with colleagues, I found that a sub-group of patients indeed has abnormally high GDNF levels and that a similar increase in mice was sufficient to trigger the disease. I also assessed the effect of GDNF deletion and an increase in Parkinson’s disease (PD), where the problem is the opposite – dopamine neurons die and striatal dopamine is lost resulting in PD typical stooped gait, tremor, and rigidity of movement. Here I found that contrary to mine and our best expectations, GDNF played no role in modulating PD progression, at least not in the proteasome inhibition model of PD we used in our study. Encouraged by these new abilities with GDNF I then tested if the CRISPR-Cas9 system can be implemented to fast-track modulate gene expression via 3´UTR modulation directly in embryos. While the answer was in general “yes” it emerged that the outcome is not always easily predictable, underpinning the need for further basic research on 3´UTR biology.

Finally, I also analyzed how GDNF levels vary in the human spinal cord, where GDNF receptor-bearing motor neurons (MNs) die in amyotrophic lateral sclerosis (ALS), an uncurable disease that is usually fatal within a few years after diagnosis. I found that GDNF levels vary substantially between individuals, positively correlate with life expectancy in limb onset ALS sub-group, and that in a mouse model of limb onset ALS small, up to two-fold increase in endogenous GDNF expression gives an almost shockingly long 10-13 weeks extra health-span via preserving neuromuscular junctions. This result suggests that enhancing endogenous GDNF signaling may carry treatment potential for limb-onset ALS.

Taken together I feel that my work has paved the way to improved analysis of gene function by demonstrating that 3´UTR modulation carries major potential in understanding gene function in health and disease and in defining its therapeutic potential.