De rol en het therapeutisch potentieel van EGLN2/PHD1 in amyotrofische laterale sclerose.

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder characterized by progressive loss of upper and lower motor neurons in the motor cortex, the brainstem and the spinal cord. Motor neuron loss results in muscle wasting and paralysis leading to the death of the patient due to failure of the respiratory muscles, typically within three to five years after diagnosis. Nevertheless, no cure is available for ALS and this while the cumulative lifetime risk to get ALS is 1 in 400. In the last decades significant advances were made regarding the models to perform fundamental and translational research, the knowledge about the pathology and the role of both motor neuron-autonomous and non-cell-autonomous mechanisms and the available research technologies. In this thesis we aimed to apply these advances to study the role and therapeutic potential of egl-9 family hypoxia-inducible factor 2 (EGLN2) enzyme. EGLN2, also known as prolyl hydroxylase domain 1 (PHD1) enzyme, is a unique target that plays a central role in cellular energy metabolism, oxidative stress response and inflammation which are three of the main hallmarks of ALS. Given the urgent need for new therapeutic strategies for ALS, we used different ALS models, patient-derived cells and single-cell technology to study if and how inhibition of EGLN2 could be beneficial in ALS. First, we explored the effect of targeting EGLN2 in two zebrafish model for ALS. Zebrafish expressing two of the most common genetic defects in ALS, namely a mutation in human superoxide dismutase (SOD1G93A) and human chromosome 9 open reading frame 72 (C9ORF72) hexanucleotide repeat RNA, display motor axon abnormalities with a decreased axonal length and increased abnormal branching. This motor axonopathy was almost completely rescued when egln2 was downregulated using an antisense oligonucleotide morpholino (AMO). Hence, these data indicated that targeting Egln2 mitigates the ALS phenotype in zebrafish by acting on one or more mechanism that are involved in the motor neuron toxicity mediated by mutant SOD1 and C9ORF72 RNA toxicity. Second, we aimed to confirm this beneficial effect by genetically targeting Egln2 in two mouse models that closely resemble the ALS phenotype as seen in patients. Mice overexpressing human mutant SOD1 (SOD1G93A) and human wildtype fused in sarcoma (FUS+/+) both present an ALS phenotype with motor neuron degeneration, paralysis and premature death. Deletion of Egln2 mitigated the ALS phenotype in SOD1G93A mice, increasing the total motor neuron count, neuromuscular junction innervation and survival. However, deletion of Egln2 did not have the same motor neuron protective effect in FUS+/+ mice. This difference might be explained by the model or the mechanism, as these SOD1- and FUS- mice show differences in the presence and timing of pathological hallmarks of ALS. Third, to identify the cell-specific pathways that underly the beneficial effect of Egln2 deletion in SOD1G93A mice, we performed single-nuclei RNA sequencing (snRNA-seq) on the spinal cord of early symptomatic SOD1G93A mice with and without Egln2. This revealed that loss of EGLN2 induced transcriptomic changes in neuronal and non-neuronal cells. Notably, deletion of Egln2 downregulated the expression of translocator protein (Tspo) which is a proposed biomarker for neuroinflammation of ALS. In agreement, we found that the pro-inflammatory interferon response was specifically downregulated in astrocytes. This decreased interferon response was suggested to be the result of reduced stimulator of interferon genes (STING) activity, as its protein level was reduced in SOD1G93A/Egln2-/- spinal cord astrocytes. Together, this single-cell approach revealed that the motor neuron protective effect of EGLN2 ablation was linked to a decrease of the astrocytic interferon-dependent inflammatory axis. Finally, we aimed to investigate the direct effect of loss of EGLN2 on patient-derived astrocyte. For this, we used induced pluripotent stem cells (iPSCs) from an ALS patient carrying a SOD1A4V mutation. As till recently, SOD1A4V iPSC-derived astrocytes were not characterized, we started by investigating the effect of this SOD1 mutation on astrocyte homeostasis. Metabolic and transcriptomic analysis of these astrocytes revealed that they showed an increased glucose metabolism and a decreased interferon response. Especially the latter was of interest given the previously established link between the STING-induced astrocytic interferon response and EGLN2 in ALS mice. To investigate if targeting EGLN2 could rescue this aberrant response, we generated SOD1A4V iPSCs with an EGLN2 knock-out using CRISPR-cas9. Deletion of EGLN2 in SOD1A4V astrocytes restored the levels of interferon-stimulated genes back to isogenic control levels. As a result, these findings suggest that the astrocytic interferon response is subject to disease stage and environmental stimuli while strengthening the novel link between EGLN2 and the astrocytic interferon response. In conclusion, this thesis demonstrated how targeting EGLN2 mitigates motor neuron degeneration in three different in vivo ALS models. Moreover, a link between EGLN2 and the astrocytic STING-induced interferon response was established, using cell-specific tools such as snRNA-seq and patient-derived astrocytes. As a result, our data highlight EGLN2 as target that protects motor neurons and restores the interferon signaling pathway in astrocytes.