Gene therapy delivered the Hsp70 heat shock protein gene to the brain and spinal cord of mice, where it restores proper folding to other proteins and clears their aggregated clumps, and extended survival with an ALS-like disease by a third

Gene therapy delivered the Hsp70 heat shock protein gene to the brain and spinal cord of mice, where it restores proper folding to other proteins and clears their aggregated clumps, and extended survival with an ALS-like disease by a third

Russian researchers gave mice carrying an ALS-like disease caused by a FUS gene mutation a single injection of a viral vector with the human Hsp70 gene. At the high dose, median survival rose from 128 to 170.5 days, and the overtly symptomatic stage lengthened by 30%. The age of symptom onset did not change.

In amyotrophic lateral sclerosis (ALS) the motor neurons, the nerve cells that control muscle movement, progressively fail. Some cases are caused by mutations in the FUS gene: the mutant protein loses the ability to fold correctly and clumps into toxic aggregates inside neurons, disrupting the cell's protein quality-control system. A similar breakdown was recently found in ordinary cellular aging as well: three quarters of insoluble clumps redissolved when energy was supplied, revealing that aggregation is a reversible process rather than irreversible wear. With age and in neurodegeneration, cellular production of Hsp70 declines.

In a preprint published on September 21, Russian researchers gave transgenic mice carrying the FUS mutation a single injection of a specially engineered adeno-associated virus (AAV) vector that crosses from the bloodstream into the brain and spinal cord without surgery, carrying the human HSPA1A gene. HSPA1A is the form of Hsp70 that surges sharply in response to cellular stress, unlike the related protein Hsc70, which operates in the cell constitutively. In a separate experiment the team confirmed that the vector remains active in the brain and spinal cord for at least 150 days after injection.

Median lifespan at the high dose increased by 23%, and the overtly symptomatic stage lengthened by 30% (18.7 days versus 14.4 in untreated mice); the low dose produced only a non-significant trend, making the effect dose-dependent. Treatment began on day 58 to 60, when misfolded FUS had already accumulated in the motor neurons: accordingly, the age at which symptoms first appeared did not shift, and motor performance tests remained at control levels. The therapy protects motor neurons from death but does not restore the function of cells already damaged. The authors note that earlier, neonatal treatment could potentially shift disease onset itself, because their data indicate that onset age is a stronger determinant of total lifespan than the duration of the symptomatic stage.

Immunofluorescence analysis of the spinal cord showed that the new Hsp70 accumulated in neurons, the very cells that die in ALS, rather than in the brain's immune cells (microglia) or in supporting cells (astrocytes). The protective effect comes from direct chaperone activity inside neurons, not from a secondary immune reaction to the virus itself.

The idea grew out of earlier work from the same group: in 2015, in PNAS, they showed that recombinant Hsp70 delivered intranasally to healthy aged mice extended their lifespan, improved spatial memory, and reduced anxiety. The problem lay in delivery: the protein penetrates cells poorly, yet the fight against FUS aggregation must happen inside the cell. AAV solves this by making the cell itself produce the chaperone continuously.

The only drug designed to make cells upregulate their own heat shock protein production, arimoclomol, showed no effect in a phase 3 trial in ALS patients in 2024. As the disease progresses, the cell's stress-response machinery becomes exhausted, whereas delivering a ready-made chaperone gene does not depend on that machinery's condition.

Hsp70 targets the aggregation process itself, a general mechanism not tied to any single gene. The authors suggest that the approach may be applicable to other diseases in which RNA-binding proteins form similar aggregates, in particular TDP-43, which is affected in the majority of ALS cases and in a subset of frontotemporal dementia cases.