|Home | About | Journals | Submit | Contact Us | Français|
3-Methylglutaconicaciduria type III (3-MGCA type III), caused by recessive mutations in the 2-exon gene OPA3, is characterized by early-onset bilateral optic atrophy, later-onset extrapyramidal dysfunction, and increased urinary excretion of 3-methylglutaconic acid and 3-methylglutaric acid. Here we report the identification of a novel third OPA3 coding exon, the apparent product of a segmental duplication event, resulting in two gene transcripts, OPA3A and OPA3B. OPA3A deficiency (as in optic atrophy type 3) causes up-regulation of OPA3B. OPA3 protein function remains unknown, but it contains a putative mitochondrial leader sequence, mitochondrial sorting signal and a peroxisomal sorting signal. Our green fluorescent protein tagged OPA3 expression studies found its localization to be predominantly mitochondrial. These findings thus place the cellular metabolic defect of 3-MGCA type III in the mitochondrion rather than the peroxisome and implicate loss of OPA3A rather than gain of OPA3B in disease etiology.
3-Methylglutaconicaciduria type III (3-MGCA type III; MIM 258501), also called Costeff optic atrophy syndrome, or Optic atrophy type 3 (OPA3), is an autosomal recessive neuro-ophthalmologic syndrome that consists of early-onset bilateral optic atrophy and late development of spasticity, extrapyramidal dysfunction and occasionally cognitive deficit [1-3]. Urinary excretion of the branched-chain organic acids 3-methylglutaconic acid (3-MGC) and 3-methylglutaric acid (3-MGR) are increased in 3-MGCA type III patients .
The occurrence of 3-MGCA type III in approximately 40 patients of Iraqi-Jewish origin  assisted in mapping the disease to 19q13.2-q13.3 . In 2001, we identified the causative gene, OPA3 . OPA3 consists of two exons and encodes for a 179-amino acid protein (OPA3) of unknown function, containing putative mitochondrial N-terminal and peroxisomal C-terminal sorting signals . All patients of Iraqi-Jewish origin are homozygous for a splice site founder mutation, c.143-1G>C (IVS1-1G>C), which abolishes mRNA expression in fibroblasts . We subsequently identified another novel OPA3 mutation, an in-frame 18-bp deletion in exon 2, c.320_337del (p.Q108_E113del), in a Kurdish-Turkish patient . Recently, a third OPA3 mutation was identified in a patient of Asian (Indian) origin; a nonsense c.415C>T (p.Q139X) mutation .
Two OPA3 mutations, G93S and Q105E, result in a rare dominant disorder (ADOAC; MIM 165300) involving optic atrophy, cataracts and extrapyramidal signs [9, 10]. The ADOAC phenotype may reflect a dominant negative effect, since heterozygous carriers of the Iraqi-Jewish loss of function founder mutation (c.143-1G>C) do not show a clinical phenotype. Similarly, a recently reported murine model harboring an L122P mutation in the heterozygous state appears normal .
The function of the OPA3 protein and how its deficiency causes the clinical symptoms and 3-methylglutaconic aciduria in 3-MGCA type III patients remain enigmatic. Here we report a comprehensive study of the OPA3 gene and its translated protein. We identified a third exon and an alternate transcript of OPA3 and determined its expression in various tissues and in 3-MGCA type III patients. We performed expression studies in fibroblasts with green fluorescent protein (GFP)-tagged OPA3 and explore mitochondrial and peroxisomal localization. We also imaged the localization, shape and inter-organellar interactions of mitochondria and peroxisomes in normal and 3-MGCA type III patients’ cells.
Patients samples were enrolled under the NIH protocol “Diagnosis and Treatment of Patients with Inborn Errors of Metabolism” (www.clinicaltrials.gov, trial NCT00369421), approved by the National Human Genome Research Institute’s Institutional Review Board. Each patient gave written informed consent, in accordance with the Declaration of Helsinki. Skin fibroblasts were grown in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum containing 100 U/ml penicillin and 0.1 mg/ml streptomycin.
RNA was isolated from cultured fibroblasts using the Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA), and reversely transcribed into cDNA using the SuperScript™ First-Strand Synthesis System for RT-PCR (Invitrogen Life Technologies, Carlsbad, CA). Tissue-specific human cDNAs were obtained from BD-Biosciences (Bedford, MA) as a multiple tissue cDNA panel. Primers were designed to distinguish between the splice forms; a common forward primer in exon 1 (5′-GCAAGGTTGCGCGTGCCCTGTGAG-3′) was combined with a reverse primer specific for exon 2 (5′-GGCCACGTTAGGTACATAGGCCATG-3′) to amplify OPA3A (625-bp fragment), or with a reverse primer specific for exon 3 (5′-GTTCCACCTGCAGGAGGCGGA-3′) to amplify OPA3B (795-bp fragment). PCR amplifications were performed under standard conditions.
For real-time quantitative PCR, exon-specific primer-probe assays were designed by the ABI Assay-by-Design service (Applied Biosystems, Foster City, CA; sequences available upon request). PCR amplifications on DNA-free RNA (DNA-free™ DNase, Ambion, Austin, TX) from patient and control fibroblasts were carried out on an ABI PRISM 7900 HT Sequence Detection System (Applied Biosystems, Foster City, CA). Results were analyzed with the comparative CT method as described .
GFP fusion proteins were created by PCR amplification of the OPA3A and OPA3B coding sequences from normal human fibroblast cDNA, followed by subcloning into pEGFP-C1 and pEGFP-N1 plasmids (Clontech, Mountain View, CA). Site-directed mutagenesis was performed employing the Quick Change Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). Specifically, the mitochondrial sorting signal (NRIKE) at residues 25-29 was replaced with the non-conserved amino acids AAAAA in the OPA3A and OPA3B containing pEGFP-N1 plasmids. The putative C-terminal peroxisomal targeting signals SKK at residues 177-179 of OPA3A and SEK at residues 178-180 of OPA3B were eliminated in the pEGFP-C1 constructs by exchanging the serine codon (TCC) at position 177 (OPA3A) or 178 (OPA3B) by a termination codon (TAA). To eliminate the mitochondrial leader sequence, the first seven residues of the leader sequence (MVVGAFP) were deleted, allowing the methionine at position 8 to function as the initiation codon in pEGFP-N1 plasmid constructs. The MitoProt (http://ihg.gsf.de/ihg/mitoprot.html) database did not predict this truncated N-terminus as a potential mitochondrial leader sequence. All plasmid constructs were verified by sequencing before use.
For localization studies, normal fibroblasts were electroporated with 3 μg of plasmid DNA in an Amaxa nucleofector electroporator (Amaxa GmbH, Walkersville, MD). Transfected fibroblasts were grown overnight on slides (Laboratory-Tek, Nalge Nunc Int., Rochester, NY), followed by fixation in 3% paraformaldehyde/phosphate-buffered saline (PBS). Some cells were treated with Mitotracker (Molecular Probes-Invitrogen, Carlsbad, CA) before fixation. Slides were then blocked/permeabilized in PBS containing 0.1% saponin, 100 μM glycine, 0.1% BSA and 2% donkey serum followed by incubation with either mouse mitochondrial antibodies (MAB 1273, Chemicon-Millipore, Billerica, MA) or rabbit PMP-70 peroxisomal antibodies (Zymed, San Francisco, CA). The cells were then washed and incubated with donkey anti-mouse secondary antibodies conjugated to cy3 or cy5 (Molecular Probes-Invitrogen, Carlsbad, CA), washed again, and mounted in VectaShield (Vector Laboratories, Burlingame, CA). Cells were imaged with a Zeiss LSM 510 META confocal laser-scanning microscope (Carl Zeiss, Microimaging Inc, Thornwood, NY) using a 488 nm Argon and a 543 nm HeNe laser. All images were acquired using a Plan NeoFluar 40X/1.3 oil DIC or a Plan Apochromat 63X/1.4 oil DIC objective. All images in Figs. 2A-C are collapsed Z-stacks (except for Fig. 2B [g], which is a single slice). For Fig. 2D and Supplemental Fig. S3, representative cells were imaged in a Z-stack, which were then rendered in 3-D and animated in shadow mode using AxioVision software (Carl Zeiss Vision GmbH, Thornwood, NY).
Previously believed to consist of two exons and a single transcript, we now demonstrate that the OPA3 gene (Fig. 1A) consists of 3 exons and is expressed in two transcripts, OPA3A (GenBank NM_025136) and OPA3B (GenBank NM_001017989). Both transcripts contain exon 1, which is spliced to exon 2 in OPA3A and exon 3 in OPA3B. Although cDNA studies indicate ubiquitous expression of OPA3A and OPA3B, OPA3A has low expression in brain and OPA3B has high expression in testis (Fig. 1B). The nucleotide sequences of exon 2 and exon 3, including their intron/exon boundaries, closely resemble each other, suggesting a segmental duplication event. Such duplications can be mediated by long interspersed element (LINE)-1 transposons in intronic sequences, and operate in pseudogene formation  and exon shuffling [14, 15]. A LINE-1 (L1MC4) transposon exists in OPA3 ~24-kb upstream of exon 2 (Fig. 1A). OPA3A is expressed and conserved from fungi to primates, while OPA3B seems to have arisen between fish and mammals (Supplemental Fig. S1).
The three known 3-MGCA type III-related mutations are all assigned to exon 2 of OPA3A [7, 8] and its adjacent splice site . We isolated fibroblast mRNA from patients with the c.143-1G>C (IVS1-1G>C) and c.320_337del (p.Q108_E113del) mutations for expression studies. Normal levels of OPA3A mRNA expression were found in cells with the in-frame c.320_337del deletion, while cells with the splice site mutation, c.143-1G>C, abolished fibroblast mRNA expression (Fig. 1C). In both cases, OPA3B expression was significantly up-regulated (Fig. 1C and D). Up-regulating one isoform to compensate for loss of another has also been described for other genes [16, 17].
Subsequently, we screened DNA of 39 patients with unexplained 3-methylglutaconic aciduria and found no mutations in OPA3, including exon 3 (unpublished results).
The OPA3A and OPA3B proteins share 47 N-terminal amino acids encoded by exon 1, and their C-terminal sequences are highly homologous (Fig. 1E). The OPA3A and OPA3B proteins have no homology to other proteins, but several lines of evidence support a mitochondrial function. First, most primary optic atrophies (as seen in 3-MGCA type III patients) result from defects in mitochondria-related proteins [18, 19]. Second, increased 3-MGC and 3-MGR is a common finding in mitochondrial respiratory chain disorders [20-22]. Third, mitochondrial degradation of leucine is a source of 3-MGC and 3-MGR [4, 20]. Fourth, other types of 3-MGCA, i.e., type I (MIM 250950) , type II (MIM 30260)  and type V (MIM 610198) , are associated with mitochondrial defects.
Moreover, the first 18 amino acids that occur both in OPA3A and OPA3B have an 87.1% predicted chance of comprising an N-terminal mitochondrial leader sequence (amino acids 1-18; MitoProt, http://ihg.gsf.de/ihg/mitoprot.html). A cleavable mitochondrial leader sequence is often followed by an additional sorting/cleavage signal [26, 27]; indeed, the NRIKE motif at positions 25-29 (Fig. 1E) conforms to such a prediction [26, 27] (PSORTII, http://psort.nibb.ac.jp/).
Despite the above evidence for mitochondrial localization and function of OPA3, increased excretion of 3-MGC and 3-MGR can also arise from defects in non-mitochondrial pathways . Regarding this possibility, it is noteworthy that OPA3A and OPA3B contain respective C-terminal tripeptides SKK and SEK (Fig. 1E), that resemble the peroxisomal targeting signal type 1 (PTS1: consensus sequence (S/A/C)(K/R/H)L) [29-31], but which have not been tested in cellular systems.
To determine the mitochondrial/peroxisomal targeting of OPA3A and OPA3B, we expressed green fluorescent protein (GFP)-tagged fusion proteins in normal fibroblasts. GFP was fused to either the C-terminus (OPA3A-GFP and OPA3B-GFP, leaving the putative mitochondrial signal free), or the N-terminus (GFP-OPA3A and GFP-OPA3B, leaving the potential peroxisomal signal free). Transfected OPA3A-GFP was distributed in a mitochondrial pattern and co-localized with a mitochondrial marker, but not a peroxisomal marker (Fig. 2A [a and b]). Even with a mutated NRIKE signal, the protein localized predominantly to mitochondria (Fig. 2A [c]), but with the leader sequence disrupted as well, a more diffuse cellular distribution was seen (Fig. 2A [d]). With expression of GFP-OPA3A, mitochondrial localization was lost (Fig. 2B [e]), but the GFP co-localized with a peroxisomal marker (Fig. 2B [f]). This was likely aided by weakening of the N-terminal mitochondrial signal by the added GFP group. This also indicated functionality of the SKK peroxisomal sorting signal, confirmed by loss of peroxisomal localization and resumed mitochondrial localization upon deletion of the SKK signal (Fig. 2B [g,h]). It is possible that OPA3A only traffics to mitochondria in vivo, since N-terminal targeting signals take priority over C-terminal signals .
Like OPA3A, OPA3B localized to mitochondria (Fig. 2C [i,k]), but it did not localize to peroxisomes, whether or not its SEK signal was available (Fig. 2C [j,l]), indicating that its SEK is not a functional peroxisomal targeting signal. When OPA3B-GFP was overexpressed, large intracellular aggregates formed (Fig. 2C), perhaps due to misfolding of the amino acid variants in OPA3B compared to OPA3A. Since OPA3A deficiency results in overexpression of OPA3B, these results raised the possibility that OPA3B-induced aggregate formation contributes to the pathology of 3-MGCA type III.
Evidence for a uniquely mitochondrial role of OPA3A is also comes from proteomic databases. For instance MitoRes (http://www2.ba.itb.cnr.it/MitoNuc/;) and MITOP (http://www.mips.biochem.mpg.de/proj/medgen/mitop/) both list OPA3A as an inner mitochondrial membrane protein . OPA3B was not identified in the mitochondrial proteome, perhaps due to low expression (Figure 1B) or perhaps the OPA3B transcript is not translated in vivo. Proteomic databases did not identify OPA3A or OPA3B as peroxisomal proteins (PeroxisomeDB, http://www.peroxisomeDB.org) .
Fluorescent labeling of mitochondria and peroxisomes revealed no differences in distribution, shape, size, or interactions of these organelles in fibroblasts from 3-MGCA type III patients compared with control fibroblasts (Supplemental Fig. S2). Confocal Z-series showed that peroxisomes localized uniformly throughout the cell, with a few attached to mitochondria, suggesting a conduit for metabolite exchange between the two organelles (Fig. 2D; Supplemental Fig. S3 for animated images).
Taken together, our findings indicate that the OPA3 gene produces two distinct RNA transcripts, OPA3A and OPA3B. OPA3B has lower expression levels than OPA3A, and may not yield a significant translation product in human cells, since OPA3B is not identified in proteomic databases and no human disease has been associated with mutations in the OPA3B-specific exon 3. In addition, OPA3A is expressed and conserved from fungi to primates, while OPA3B is uniquely found in mammals (Supplemental Fig. S1). Even though OPA3B gene expression was significantly up-regulated in OPA3A deficient cells, it is likely that loss of OPA3A rather than gain of OPA3B cause disease etiology.
A potential cause of increased 3-MGC and 3-MGR is excessive shunting of organic precursors from peroxisomes to mitochondria, a process whose mechanistic basis remains enigmatic . However, our cellular findings implicate that OPA3A does not localize to peroxisomes, based on its absence from peroxisomal proteomics databases , non-evolutionary conservation of the peroxisomal sorting signal SKK (e.g., the dog, cow, horse OPA3A proteins do not contain the C-terminal SKK sequence), and in vivo preference of N-terminal targeting signals over C-terminal signals .
We conclude that OPA3A predominantly localizes to mitochondria, which can explain the combination of elevated 3-MGC and 3-MGR  and optic maldevelopment and/or atrophy [18, 19]. Related phenotypes (e.g., some cases of 3-MGCA type IV [4, 35]) could also be caused by mitochondrial defects.
We thank Ian Nouvel for skillful laboratory assistance. This study was supported by the Intramural Research Program of the National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA and by the Costeff Support Group Foundation (Y.A.).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.