Understanding why some motor neurons survive and others do not
Using human stem cell–derived motor neurons (MNs) engineered with advanced CRISPR tools, we created fluorescent “reporter” cells that allowed us to measure the levels of the SMN protein in individual neurons over time. SMN is essential for MN health, and its deficiency causes SMA. We discovered that even in the same culture, MNs naturally produce different amounts of SMN. This variation is crucial: neurons with low SMN levels are consistently more fragile, while those with high SMN levels survive. This pattern is seen in healthy, SMA, and ALS models. Over time, the vulnerable low-SMN neurons die first, leaving behind a “survivor” population. These findings fundamentally change how we think about disease progression and help explain why some MNs persist even in advanced stages of motor neuron diseases.
We further found that low-SMN neurons are more electrically active - in other words, they fire more intensely - which may make them more susceptible to stress. This suggests that strategies that reduce excessive neuronal activity could benefit both SMA and ALS. To understand the molecular foundations of this vulnerability, we analyzed gene expression and protein changes in high-SMN versus low-SMN neurons. This work highlighted pathways related to neuronal activity, survival mechanisms, and cellular recycling systems, providing new targets now being validated.
These results were presented at multiple international scientific meetings (CureSMA, SMA Europe, CURE-ND, EMBO) and form the core of a completed PhD thesis.
Revealing a key role for the cell’s recycling system in SMA
A major part of the project focused on the cell’s “waste disposal and recycling” machinery - the lysosome–autophagy system - which is vital for clearing damaged proteins. Using our isogenic human SMA models, we discovered that SMA MNs have fewer lysosomes and that these organelles work poorly. As a result, toxic protein aggregates accumulate specifically in the neurons most at risk. We identified the molecular cause of this defect: an abnormal activation of the mTORC1 pathway that reduces the activity of TFEB, a master regulator of lysosomal function.
Importantly, restoring TFEB activity improved lysosome function, reduced protein aggregates, and significantly increased MN survival, both in human stem cell–derived cultures and in a zebrafish SMA model. Because TFEB-based therapies are already being explored for other neurodegenerative diseases, our findings open a promising new therapeutic avenue for SMA, potentially as a combination with existing SMN-boosting drugs.
This work is currently under review, contributed to a top-graded PhD thesis, and has been presented at international meetings (EMBO/EMBL, PROGENIE, RegenBell).
Identifying new genetic modifiers of SMA
We successfully developed a robust fruit fly model of SMA to search for genes that modify disease severity. Using a combination of fly genetics, human stem cell–derived neuromuscular organoids, and mouse data, we discovered that specific signaling pathways that govern early embryonic development are altered in SMA. Correcting these pathways improves disease features across models and thus identifies new potential therapeutic targets. These results are being prepared for publication.
Creation of advanced tools and technologies for the research community
A major achievement of the project was the development of a large set of genetically matched human stem cell lines to model SMA and ALS. These “isogenic” lines remove the variability that often complicates stem cell studies and are already being used as precision models in disease research. We also generated reporter lines that allow specific MN populations to be tracked and isolated.
Another key technological advance was the creation of a new spinal cord organoid model (CASCO) that accurately mimics the diversity of MNs found along the human spinal cord. This system produces both vulnerable and resistant MN types within the same organoid, enabling unprecedented studies of selective vulnerability. CASCO also incorporates blood vessel–like and muscle-like cells, making it one of the most physiologically relevant human spinal cord models available.
