MA-5 Improves Mitochondrial Function Across Multiple ALS Models

Amyotrophic lateral sclerosis is not one simple molecular disease. Some patients carry mutations in genes such as SOD1 or FUS; many others have sporadic ALS without a known inherited cause. Yet across these different backgrounds, one cellular problem repeatedly appears: impaired mitochondrial function. A Tohoku University-led team has now shown that Mitochonic acid 5, or MA-5, can improve several mitochondrial and disease-related phenotypes across fruit-fly ALS models, patient-derived skin fibroblasts and motor neurons generated from patient iPS cells.^[1][2]

The collaboration included researchers from Tohoku University, Kyoto University, the National Center of Neurology and Psychiatry and Keio University. The investigators tested MA-5 in three complementary systems: Drosophila expressing a human ALS-associated SOD1 mutation, fibroblasts obtained from people with ALS, and motor neurons differentiated from patient-derived induced pluripotent stem cells. The study was published in JCI Insight in September 2026.^[1]

Why mitochondria matter in ALS

Motor neurons are unusually demanding cells. Their axons can extend enormous distances relative to the size of the cell body, and maintaining ion gradients, synapses, axonal transport and muscle signaling consumes large amounts of energy. Mitochondria provide much of that ATP while also helping regulate calcium, redox balance and cell-death pathways.

ALS pathology has repeatedly been associated with mitochondrial fragmentation, impaired respiration, reduced ATP, oxidative stress and disrupted mitochondrial transport along axons. Roughly 10% of ALS cases are familial, and more than 20 genes have been implicated, including SOD1, TARDBP, FUS and C9orf72. Most cases are sporadic. A treatment that acts on a mitochondrial defect shared across these backgrounds could therefore, in principle, reach more patients than a therapy aimed at one mutation alone.^[2]

MA-5 began as a mitochondrial-disease compound

MA-5 was not originally developed for ALS. Tohoku University researchers screened derivatives of indole-3-acetic acid, a plant hormone, for compounds capable of increasing cellular ATP. In 2015 they reported Mitochonic acid 5—chemically, 4-(2,4-difluorophenyl)-2-(1H-indol-3-yl)-4-oxobutanoic acid—as a molecule that increased ATP and improved survival in fibroblasts from patients with several mitochondrial diseases under stress conditions.^[3]

Later work proposed a distinctive mechanism. MA-5 binds mitofilin/Mic60, a protein involved in organizing mitochondrial inner-membrane cristae, and promotes ATP synthase dimerization and higher-order supercomplex formation. The result is increased ATP production without simply driving electron-transport-chain complexes I through IV harder. That structural mechanism has made MA-5 different from conventional antioxidant strategies or drugs that primarily stimulate mitochondrial biogenesis.^[4][5]

The compound has also moved into human development for mitochondrial disease. A Phase I study launched in 2021 in 56 healthy adults to assess safety and pharmacokinetics. In 2025, Japan’s Agency for Medical Research and Development announced the start of a Phase II trial in mitochondrial disease. Those trials provide valuable human safety experience—but they are not ALS trials, and they do not establish efficacy in motor neuron disease.^[6][7]

The first test: an SOD1 ALS fruit-fly model

The researchers began with a Drosophila model in which motor neurons express the human SOD1 G85R mutation. SOD1 normally helps defend cells against reactive oxygen species, but certain mutant forms misfold and become toxic, causing a subset of familial ALS.^[8]

The mutant flies show impaired movement and mitochondrial abnormalities. MA-5 significantly improved locomotor performance. The treatment also restored ATP levels and mitochondrial membrane potential and produced a trend toward improvement in abnormal mitochondrial morphology in muscle.^[2]

That is encouraging but limited. A fruit fly is a powerful genetic disease model, not a miniature human with ALS. Its nervous system, lifespan, metabolism and drug distribution differ substantially from those of mammals. Improvement in fly climbing behavior cannot be assumed to predict improvement in human motor function.

Patient fibroblasts show a similar energy response

The team next tested skin fibroblasts obtained from people with ALS. These included cells from patients with SOD1 mutations and from patients with sporadic ALS without known ALS-associated mutations. In multiple samples, MA-5 increased ATP production, and isolated mitochondria also produced more ATP after treatment.^[2]

The researchers also challenged cells with oxidative stress. In fibroblasts from sporadic ALS patients, MA-5 reduced cell death and LDH release in a dose-dependent fashion up to 10 μM. At 30 μM the effect appeared weaker or potentially harmful. That detail is important: mitochondrial augmentation is not simply a case of “more drug is better,” and a therapeutic window would have to be defined carefully in later development.^[2]

Testing motor neurons made from patient iPS cells

The most disease-relevant human-cell experiments used induced pluripotent stem cells derived from ALS patients and differentiated into motor neurons. These cells retain the donor’s genetic background while allowing investigators to study the cell type that degenerates in ALS.

Motor neurons carrying SOD1 mutations showed impaired ATP production, which increased after MA-5 treatment. Similar mitochondrial improvements were also observed in FUS-mutant and sporadic ALS-derived cellular models. That breadth is central to the paper’s claim that mitochondrial dysfunction may be a therapeutic target shared across multiple ALS subtypes.^[1][2]

MA-5 also increased mitochondrial motility in neurons. That matters because mitochondria must travel along long axons to reach locations with high energy demand. In motor neurons, a mitochondrion that makes ATP efficiently but cannot reach distal axons or synapses may still fail to support the cell. Improving mitochondrial movement therefore points to a potentially broader effect on neuronal energy distribution, not just total ATP content.^[2]

A possible common target across genetically different ALS

The heterogeneity of ALS is one reason drug development has been so difficult. A therapy aimed specifically at mutant SOD1 can make biological sense for SOD1-ALS but does not directly address FUS-associated or sporadic disease. A treatment that improves a convergent cellular process could have broader reach.

The new study strengthens that hypothesis because MA-5 produced similar directional mitochondrial effects in cells representing SOD1-mutant, FUS-mutant and sporadic ALS. It does not prove that every ALS subtype shares the same mitochondrial defect, or that every patient would respond. The number of patient-derived lines was limited, and cellular response is not the same as clinical benefit.

Multiomics points beyond ATP to broader metabolic changes

The investigators combined transcriptomic and metabolomic analyses to ask what else changes after MA-5 treatment. The data suggested alterations in genes associated with the mitochondrial respiratory chain and suppression of the glycerophosphate shuttle, a pathway that transfers reducing equivalents from the cytosol into mitochondria and can contribute to reactive oxygen species generation.^[1][2]

The authors propose that downregulation of this shuttle may help reduce mitochondrial oxidative stress. That remains a mechanistic interpretation rather than proof that the glycerophosphate shuttle is the dominant therapeutic pathway. Earlier MA-5 work points to mitofilin/Mic60 and ATP-synthase organization as the compound’s primary molecular framework; the new omics data add possible downstream consequences.

C7orf31 emerges as a biomarker candidate

The study also identified C7orf31 and C3orf62 as candidate markers related to MA-5 response or ALS subtype. Plasma C7orf31 was markedly higher in the SOD1-ALS group than in sporadic ALS in the study cohort: 28.32 ± 12.37 ng/mL versus 7.20 ± 4.52 ng/mL, with P < 0.0001.^[2]

That is a potentially useful signal, but it is not yet a diagnostic test. Biomarkers need replication in independent cohorts, assessment across disease stages and comparison with other neurological conditions. It also remains unknown whether C7orf31 would reliably change with MA-5 treatment in people with ALS, because MA-5 has not yet been given to ALS patients in a clinical trial.

Human experience exists—but in mitochondrial disease, not ALS

MA-5 has one practical advantage over a completely new preclinical molecule: it has already entered human testing in another disease area. Its Phase I program assessed safety and pharmacokinetics in healthy adults, and Phase II development is underway for mitochondrial disease.^[6][7]

Repurposing it for ALS would still require substantial work. Investigators would need to establish dosing, penetration into the brain and spinal cord, long-term safety, interaction with existing ALS therapies and whether the biological effects differ by disease stage. Prior safety data can shorten some uncertainty, but they cannot substitute for an ALS efficacy trial.

Where MA-5 fits in the history of ALS drug development

ALS treatment has historically centered on modestly disease-modifying drugs and supportive care, while recent years have brought more genotype-specific approaches, including therapies aimed at SOD1. That precision-medicine direction is important, but it leaves a large population of sporadic ALS patients without an obvious gene-specific target.

Mitochondrial dysfunction offers a different strategy: target a cellular process that may lie downstream of several initiating causes. Yet this concept has disappointed before. Antioxidants, coenzyme Q-based approaches and strategies intended to improve mitochondrial metabolism have repeatedly produced promising laboratory findings without transforming ALS care.

MA-5 is scientifically interesting because its proposed mechanism is different. Rather than acting simply as an antioxidant or directly accelerating respiratory-chain complexes, it appears to influence mitofilin/Mic60, cristae organization and ATP-synthase assembly. Whether that molecular distinction is enough to produce meaningful clinical benefit is unknown.

The next bridge is mammalian ALS models—and then patients

Tohoku University’s own announcement emphasizes the present limits. The work was conducted mainly in fruit flies and cultured human cells. It remains unknown whether MA-5 can improve motor function, delay disease progression or extend survival in mammalian ALS models, much less in people with ALS.^[1]

The logical next steps include testing in ALS mouse models, measuring drug exposure in the central nervous system, defining chronic dosing and toxicity, and then designing early clinical studies in people with ALS. If C7orf31 or another marker proves reproducible, biomarker-guided enrollment might eventually help identify patients most likely to benefit.

Looking for a shared vulnerability

ALS is increasingly understood as a syndrome in which multiple genetic and molecular routes converge on motor-neuron degeneration. That makes it unlikely that one therapy will work equally well for every patient. But convergence also creates opportunities: if different forms of ALS damage the same cellular systems, those systems become shared therapeutic targets.

The new work adds weight to mitochondria as one such target. Across SOD1-mutant, FUS-mutant and sporadic ALS cellular backgrounds, one compound improved energy production and other mitochondrial measures in the same general direction. That is not a cure. It is a stronger preclinical rationale.

MA-5 has been studied for roughly a decade as a mitochondrial-disease compound. Its move into ALS research connects that history with one of neurology’s hardest therapeutic problems. Whether it will ultimately become an ALS drug remains unknown. What the study establishes is narrower but important: mitochondrial dysfunction is not only a feature of ALS pathology that can be observed—it is a process that can be experimentally manipulated across several models of the disease.