OPA1, a dynamin-related guanosine triphosphatase mutated in dominant optic atrophy, is required for the fusion of mitochondria. Proteolytic cleavage by the mitochondrial processing peptidase generates long isoforms from eight messenger RNA (mRNA) splice forms, whereas further cleavages at protease sites S1 and S2 generate short forms. Using OPA1-null cells, we developed a cellular system to study how individual OPA1 splice forms function in mitochondrial fusion. Only mRNA splice forms that generate a long isoform in addition to one or more short isoforms support substantial mitochondrial fusion activity. On their own, long and short OPA1 isoforms have little activity, but, when coexpressed, they functionally complement each other. Loss of mitochondrial membrane potential destabilizes the long isoforms and enhances the cleavage of OPA1 at S1 but not S2. Cleavage at S2 is regulated by the i-AAA protease Yme1L. Our results suggest that mammalian cells have multiple pathways to control mitochondrial fusion through regulation of the spectrum of OPA1 isoforms.

Autosomal dominant optic atrophy (ADOA) is the most prevalent hereditary optic neuropathy resulting in progressive loss of visual acuity, centrocoecal scotoma and bilateral temporal atrophy of the optic nerve with an onset within the first two decades of life. The predominant locus for this disorder (OPA1; MIM 165500) has been mapped to a 1.4-cM interval on chromosome 3q28-q29 flanked by markers D3S3669 and D3S3562 (ref. 3). We established a PAC contig covering the entire OPA1 candidate region of approximately 1 Mb and a sequence skimming approach allowed us to identify a gene encoding a polypeptide of 960 amino acids with homology to dynamin-related GTPases. The gene comprises 28 coding exons and spans more than 40 kb of genomic sequence. Upon sequence analysis, we identified mutations in seven independent families with ADOA. The mutations include missense and nonsense alterations, deletions and insertions, which all segregate with the disease in these families. Because most mutations probably represent null alleles, dominant inheritance of the disease may result from haploinsufficiency of OPA1. OPA1 is widely expressed and is most abundant in the retina. The presence of consensus signal peptide sequences suggests that the product of the gene OPA1 is targeted to mitochondria and may exert its function in mitochondrial biogenesis and stabilization of mitochondrial membrane integrity.

Mitochondria are dynamic organelles, the morphology of which results from an equilibrium between two opposing processes, fusion and fission. Mitochondrial fusion relies on dynamin-related GTPases, the mitofusins (MFN1 and 2) in the outer mitochondrial membrane and OPA1 (optic atrophy 1) in the inner mitochondrial membrane. Apart from a role in the maintenance of mitochondrial DNA, little is known about the physiological role of mitochondrial fusion. Here we report that mitochondria hyperfuse and form a highly interconnected network in cells exposed to selective stresses. This process precedes mitochondrial fission when it is triggered by apoptotic stimuli such as UV irradiation or actinomycin D. Stress-induced mitochondrial hyperfusion (SIMH) is independent of MFN2, BAX/ BAK, and prohibitins, but requires L-OPA1, MFN1, and the mitochondrial inner membrane protein SLP-2. In the absence of SLP-2, L-OPA1 is lost and SIMH is prevented. SIMH is accompanied by increased mitochondrial ATP production and represents a novel adaptive pro-survival response against stress.

We and others have shown recently that mutations in the OPA1 gene encoding a dynamin-related mitochondrial protein cause autosomal dominant optic atrophy (ADOA) linked to chromosome 3q28-q29. Here we report screening of the OPA1 gene in a sample of 78 independent ADOA families. OPA1 mutations were identified in 25 patients (detection rate 32.1%) including 16 novel mutations. We successfully amplified OPA1 cDNA prepared from leukocyte RNA of three patients, and found the amount of transcripts harboring the Arg366Stop mutation was significantly reduced compared with transcripts derived from the normal chromosome. Analysis of the distribution of OPA1 mutations in ADOA revealed that most missense mutations cluster within the putative GTPase domain, and that there is a preponderance of mutations, which result in premature translation termination. These observations support the notion that haploinsufficiency may represent a major pathomechanism for ADOA. In addition, we identified an ADOA patient who is a compound heterozygote for two OPA1 missense mutations. The fact that this patient is by far more severely affected than her simple heterozygotic parents and siblings implies that at least these OPA1 alleles behave semi-dominantly rather than purely dominantly. Clinical examination revealed considerable variability in disease expression among patients carrying OPA1 mutations and no strict correlation with either the position or the type of mutation.