Submission Details

The AFG3L2 gene was first reported in relation to autosomal recessive spastic ataxia 5 (MIM#: 614487) in 2011 (PMID: 22022284) and in relation to autosomal dominant optic atrophy 12 (MIM#: 618977) in 2015 (PMID: 26539208). Cases diagnosed with optic atrophy 12 share the phenotype of optic atrophy and are characterized by other ocular features such as optic nerve pallor, progressive loss of central visual acuity, color vision defects, retinal nerve fiber thinning, and/or development of glaucoma. Although penetrance can be incomplete, with sub-clinically affected carriers, most published cases have pediatric onset and underlying abnormalities of mitochondrial morphology including fragmentation of the mitochondrial network. Subsequent reports have identified extraocular phenotypes in individuals diagnosed with optic atrophy, including hearing loss, progressive dystonia, and episodes of cerebellar ataxia, and have expanded the diversity of AFG3L2-related cases to include a diagnosis of optic atrophy with autosomal recessive inheritance (PMID: 32219868, PMID: 36974169). Families diagnosed with spastic ataxia 5 are characterized by cerebellar and neuromuscular features, ranging from cerebellar atrophy, myoclonic seizures, basal ganglia involvement, and progressive hearing loss to progressive spastic gait / gait ataxia, hyperkinetic movements, increased blood lactate, dystonia, dysarthria, dysmetria, and abnormalities of eye movement. These often occur in the presence of ocular features including optic atrophy, optic nerve pallor, progressive loss of central visual acuity, color vision defects, and retinal nerve fiber thinning. Abnormalities of mitochondrial morphology have also been identified, including fragmentation of the mitochondrial network. Per criteria outlined by the ClinGen Lumping & Splitting Working Group, the molecular mechanism (AFG3L2 loss of function) was found to be consistent among patients with optic atrophy 12 (MIM#: 618977) and spastic ataxia 5 (MIM#: 614487). The phenotypic variability between them appears to represent a spectrum of disease rather than separate disease entities, with underlying mitochondrial network dysfunction as a shared feature. Therefore, cases caused by inherited loss-of-function variants in AFG3L2 have been lumped into a single disease entity, referred to as AFG3L2-related optic atrophy and/or ataxia, with a semidominant mode of inheritance. Heterozygous missense variants in AFG3L2, particularly within the region encoding the protease domain, have also been linked to cases of spinocerebellar ataxia 28 (MIM#: 610246) as early as 2006 (PMID: 16251216). These patients can alternatively harbor heterozygous C-terminally truncating variants in AFG3L2, and the mechanism may be dominant negative (PMID: 24272953). Spinocerebellar ataxia 28 has neurodevelopmental features such as cerebellar atrophy, ataxia, dysarthria, hyperreflexia, and disordered eye movement, however, other features associated with AFG3L2 loss-of-function such as optic atrophy and mitochondrial network fragmentation have not yet been reported in affected individuals. Therefore, AFG3L2-related cases of spinocerebellar ataxia 28 have been recommended for a split curation for a separate disease entity.

Thirteen suspected disease-causing variants were scored as part of this curation (three nonsense and ten missense), which have been collectively reported in eleven probands in three publications (PMID: 22022284, PMID: 32219868, PMID: 32248051). Six of the probands scored in this curation harbored only one variant allele within the AFG3L2 locus, while the other five probands harbored biallelic AFG3L2 variants. The mechanism of pathogenicity appears to be monoallelic or biallelic loss of AFG3L2 function conferred by null and/or hypomorphic variants. Interestingly, some of the variants found in biallelic cases do not appear to cause disease in family members who harbor them in the heterozygous state (PMID: 32219868). Families with segregation evidence were not sufficiently large to contribute to the scoring of the gene-disease relationship (PMID: 32219868). The literature included more case-level evidence (PMID: 26539208, PMID: 29181157, PMID: 32600459), but its inclusion in this curation was not necessary to reach the maximum score for genetic evidence.

This gene-disease association is also supported by biochemical evidence that AFG3L2 encodes a subunit of an ATP-dependent metalloprotease complex (PMID: 17101804) that localizes to the inner mitochondrial membrane (PMID: 10395799) and regulates OPA1 processing. AFG3L2 physically interacts with paraplegin, another component of this complex encoded by the SPG7 gene, which is associated with spastic paraplegia cases that can include the optic atrophy phenotype (PMID: 35243150). AFG3L2 expression is ubiquitous but especially high in muscle, heart, cornea, and retinal pigment epithelial tissues (PMID: 30239781). Deletion of the budding yeast ortholog of AFG3L2 results in mitochondrial respiratory chain deficiency (PMID: 7926051). Naturally occurring and genetically engineered mouse models of Afg3l2 loss exhibit reduced motor function in limbs, postural instability, limb dysmetria, and complete paralysis leading to early death, as well as reduced activity of mitochondrial respiratory chain complexes (PMID: 18337413).

In summary, AFG3L2 is definitively associated with AFG3L2-related optic atrophy and/or ataxia. This has been repeatedly demonstrated in both research and diagnostic settings, and has been upheld over time without the emergence of contradictory evidence, leading to a Definitive classification. This classification was approved by the ClinGen Glaucoma and Neuro-Ophthalmology Gene Curation Expert Panel and the ClinGen Mitochondrial Diseases Gene Curation Expert Panel on May 16th, 2024 (SOP Version 10).