A splicing defect in the survival of motor neuron gene (SMN) leads to a deficiency of protein function that results in spinal muscular atrophy (SMA). It is the second most common autosomal recessive disease, occurring with a prevalence of 1 to as little as 6,000 births.

Over the past decade, Tremendous biomedical scientific progress has been made in the development of therapeutics for SMA. A small molecule, an antisense and now most recently a gene therapy approach have all been shown to be effective, FDA-approved and working via the mechanism of increasing expression of the SMN protein.

The gene therapy approach has the unique advantage of needing injection only once in a patient's life. By contrast, the antisense approach requires spinal injections every 3-4 months for the rest of the patient's life. The small molecule can be delivered orally but is likely to have the least specificity.

Scientists working in the Center for Motor Neuron Biology and Disease at Columbia University reported in the April 1, 2021, issue of Nature Neuroscience on the discovery that there may be cause for long-term caution when using AAV gene therapy to treat neurological disease.

When researchers overexpressed SMN with an AAV9 vector, they observed protein aggregation with SMA-like pathogenesis, indicating gain-of-function toxicity.

Corresponding author Livio Pellizzoni, an associate professor of pathology and cell biology at Columbia University, told BioWorld Science, "if you make too much SMN, to really supra-physiological levels as is specifically reached with AAV gene therapy...this can cause...pathogenic protein aggregates."

Pellizoni pointed out that the small molecule and the antisense therapeutic do not have the same gain of toxicity due to overexpression concern. He stressed that this safety concern is specific to gene therapy and primarily expected to more potentially be an issue in post-mitotic cell types, like neurons.

Whether this will turn out to be a concern with SMA patients treated with AAV-SMN remains to be determined. The approach results in long-term expression in rodents throughout their lifespan, which is just over 2 years, but how long gene expression will last in humans is not yet clear.

Still, Pellizoni said, "this result suggests there should be more of an element of caution to be considered in evaluating and moving forward with AAV gene therapy approaches in humans."

Pellizoni's research has focused on understanding the normal biological function of SMN in RNA processing. Much of his research has focused on increasing the understanding of the mechanisms causing low levels of SMN with associated SMA disease and why SMN is essential to cellular life.

Pellizoni explained that this discovery was almost serendipitous in a sense. He and his team were using an AAV9 vector as a tool to introduce genes for studying SMN function in the mouse, but then they noticed this pathological consequence that resembled the disease itself, but at much later onset.

Mechanistically, the authors wrote in their paper, "aggregation of overexpressed SMN in the cytoplasm of motor circuit neurons sequesters components of small nuclear ribonucleoproteins, leading to splicing dysregulation and widespread transcriptome abnormalities with prominent signatures of neuroinflammation and the innate immune response."

Pellizoni explained that SMN works with other protein partners as one of the subunits of a larger molecular complex, but with gene therapy one member is expressed at much higher stoichiometric levels. Thus, there is a potential imbalance in this ratio, which can lead to the formation of aggregates. The complex loses its ability to bind its normal substrates correctly and that is why it exerts deleterious effects on pathways related to gene expression and RNA processing to ultimately damage the neurons over a long time of overexpression.

Gain-of-function toxicity appears to be restricted to neurons because of the way gene therapy works. Pellizoni explained that after cells take up the viral vector, they make many copies of the virus to express SMN. Then this is maintained for life because neurons are post-mitotic.

By contrast, the researchers showed that the expression of AAV-SMN in the normally regenerating liver cells of mice was diluted with time. AAV is not integrated into chromosome DNA and this is generally considered desirable as it is believed to be less oncogenic than other viral gene therapies that rely on chromosomal insertion, but this study suggests there still may be potential for adverse events, instead due to excessive overdose-like overexpression with AAV gene therapies.

The clinical outcomes for SMA are classified by a spectrum of severities. Type 1 is very severe, affecting newborns to have a very short life expectancy of around 2 years, but there is type 2, 3 even 4, which has fewer symptoms affecting the muscular system. SMA severity ranges from a paralyzed baby to an adult who has some motor issues, but can walk and has an almost normal life expectancy.

Pellizoni and his colleagues plan to continue studying the disease mechanisms of SMA with respect to pathway functions of SMN that may contribute to the disease pathogenesis from a mechanistic point of view.