We describe the mapping and identification of the gene for hereditary myopathy with lactic acidosis (HML). HML is characterized by low physical performance, resulting in physical exertion that causes early exhaustion, dyspnoea and palpitations. Using an autosomal recessive mode of inheritance, we mapped the trait to chromosome 12q23.3–24.11, with a maximum lod score of 5.26. The 1.6-Mb disease-critical region contained one obvious candidate gene—ISCU—specifying a protein involved in iron–sulphur cluster assembly. IscU is produced in two isoforms; one cytosolic and one mitochondrial, coded for by different splice variants of the ISCU gene. Mutational analysis of all exon and intron sequences as well as 1000 bp of the promoter of the ISCU gene revealed one intron mutation that was specific for the disease haplotype. The mutation is located in a region with homology to the interferon-stimulated response element (ISRE), but we could not see any effect of the mutation on expression levels in vitro or in vivo. We did, however, observe a drastic difference in the splicing pattern between patients and controls. In controls the mRNA was, as expected, mainly in the mitochondrial form, while in the patients a larger mRNA transcript was predominant. Sequencing of the product revealed that the mutation activates cryptic splice sites in intron 5 resulting in aberrant mRNA containing 100 bp of the intron. To conclude, our data strongly suggest that an intron mutation in the ISCU gene, leading to incorrectly spliced mRNA, is the cause of myopathy with lactic acidosis in this family.
In 1964, Larsson et al. (1) described five families with 14 members affected by a hereditary myopathy characterized by poor physical performance since childhood.The patients showed a low exercise tolerance with markedly hyperkinetic circulation in working muscle, with release of elevated amounts of lactate and pyruvate. During exercise, the muscles became hard and tender, and developed cramp. In the most severe form of the disease, widespread weakness, myoglobinuria and severe acidosis were observed—to the extent that it could prove fatal (1,2). Biochemical and histochemical analysis showed low succinate dehydrogenase (SDH) activity in muscle cells but normal levels of 3-OH-acyl-CoA-dehydrogenase, phosphofructokinase, phosphorylase and lactate dehydrogenase, suggesting a defect in complex II of the respiratory chain (3). Electron microscopy studies of muscle biopsies showed abnormalities typical of mitochondrial myopathy, furthermore, the mitochondria contained electron-dense homogenous inclusions. After exercise fibres were markedly depleted of glycogen and contained large amounts of lipid droplets (3). By immunoblotting, it was shown that the patients had low levels of the 30-kD iron–sulphur protein and the 13.5-kD protein of complex II, but they had close to normal levels of the 70-kD protein (4). Furthermore, reduced levels of mitochondrial aconitase and additional abnormalities in the respiratory chain—affecting proteins with iron–sulphur centres in complexes I and III—were also observed (4,5). The fact that the levels of several iron–sulphur proteins were low (3–5) suggested that the patients suffered from a dysfunction in the synthesis, import, processing or assembly of iron–sulphur clusters (5).
To date, 19 individuals suffering from hereditary myopathy with lactic acidosis (HML) have been identified in nine families in northern Sweden. All the affected individuals are in single sib-ships thus indicating a recessive mode of inheritance (2). Furthermore, all but one of the families could be connected through genealogical analysis, suggesting a common ancestor (6). In this paper, we report mapping of the disease gene responsible for the mitochondrial myopathy in these families. Furthermore, we describe the identification of a disease-specific intron mutation in the iron–sulphur cluster assembly gene, ISCU, which results in aberrant splicing of the RNA.
HML maps to chromosome 12q23.3–24.1
This study deals with 9 families with 19 affected family members, of which 15 were available for analysis. All affected individuals (except in family I) have been investigated previously with respect to clinical and biochemical properties (1–3,7). The inheritance pattern in all families is consistent with autosomal recessive transmission of the disease (Fig. 1A) and genealogical analysis could connect all but one family to a common ancestor (6).
Based on the hypothesis of recessive inheritance, we performed a genome-wide screen for shared homozygosity regions in the affected individuals. Using an autosomal recessive mode of inheritance, we mapped the trait to chromosome 12q23.3–24.11. We achieved significant Lod scores for a number of markers in the region, with a maximum Lod score of 5.26 for marker D12S84 (Table 1). The disease-critical region was found to be restricted to a 1.6-Mb region between markers D12S1613 and D12S105. No other marker loci tested fulfilled the criteria of a recessive gene identical by descent. In order to restrict the disease-critical region further, a number of additional markers were analysed in the region and haplotypes were constructed (Fig. 1B). A haplotype common to all affected individuals spanning two markers (D12S1605 and D12S84) could be seen. However, several families shared a region identical by descent, spanning several markers.
Two-point Lod score table showing the markers most tightly linked to the disease locus using MLINK software. The maximum Lod score (5.26), obtained with marker D12S84 at θ = 0, is shown in bold.
The ISCU gene harbours a disease-specific mutation
The disease-critical region was scrutinized for functional candidate genes. This analysis revealed 16 known or putative genes located in this region (Fig. 2A). Because of the reported deficiency in several iron–sulphur proteins in the HML patients, the gene for iron–sulphur cluster assembly, ISCU, appeared to be an excellent candidate gene (8–10). ISCU is an iron–sulphur cluster scaffold protein that exists in both a mitochondrial isoform and a cytosolic isoform (8). ISCU has been shown to be important in iron homeostasis and to be, among other functions, essential for aconitase activity (11). We performed mutational analysis of all exon and intron sequences and also 1000 bp of the ISCU gene promoter. The mutational analysis revealed one disease-specific mutation, G→C, in intron 5, 382 bp downstream of the exon, (position 5044 in the Genbank sequence NC_000012) (Fig. 2B). Analysis of 177 population controls revealed only one individual heterozygous for the disease-specific mutation. Homology search revealed that the mutation is located in a region with strong homology to the interferon-stimulated response element (ISRE), suggesting that this region may be important for regulation of the gene.
Nuclear factors bind to both the normal and the mutant sequence but with different binding patterns
To investigate whether the sequence carrying the mutation may have a regulatory function we examined if any nuclear factors interact with this region. For this we performed gel-shift assays using nuclear extracts from RD4 cells which endogenously express IscU. Normal or mutant double-stranded oligomers covering 15 bp on either side of the site of the mutation were used in the assay. The gel-shift assay did indeed show binding of RD4 nuclear factors to the normal sequence, supporting the idea that the region has regulatory function (Fig. 3A). With the normal sequence, two strong interactions could be observed. When we used the mutant oligomer in the same assay, we could also see interaction with nuclear factors but with a different pattern of binding (Fig. 3A and B). A novel intermediate band appeared, while the high molecular weight interaction observed with the normal sequence disappeared. We did not see any significant difference in binding patterns in the competition experiment with the normal sequence using cold normal or mutant oligomer (Fig. 3A). When the reverse competition experiment was performed, there was a clear difference in the competition for the mutant-specific band. The normal oligomer was at least eight times less efficient in competing out this band than the mutant oligomer (Fig. 3B).
The region harbouring the mutation confers enhancer activity but no difference is observed between the normal and mutant sequence
In order to determine whether the region carrying the mutation might be involved in transcriptional regulation, a 172-bp fragment of intron 5, with or without the mutation, was inserted upstream of the SV40 promoter driving the luciferase reporter gene. The constructs were then introduced into IscU-producing RD4 cells and 45–49 h post-transfection, the cells were harvested and analysed for luciferase activity. We observed an approximate 2.2-fold increase in luciferase activity compared with the basal promoter alone (P-values of 0.0017 and 0.0023 for normal and mutant sequence, respectively) (Fig. 4). We could however not detect any difference in the enhancer activity between the normal and mutant sequences. Adding α-interferon to the cells also failed to reveal a difference between the normal and mutant sequence.
RT–PCR on muscle biopsies from patients and control reveal aberrant splicing of the ISCU mRNA due to activation of cryptic splice sites
Since we did not observe any difference between the normal and mutant sequence in the transcriptional assay, we examined if the mutation resulted in a lower expression of the ISCU gene in vivo. For this RT–PCR was performed on mRNA extracted from muscle biopsies from two patients and one control. The RT–PCR was performed using ISCU primers that could amplify both the cytosolic (c-IscU) and the mitochondrial (m-IscU) form of IscU. Location of the ISCU primers and the two alternative transcripts are shown in Figure 5A. The c-IscU mRNA would result in a 685 bp product and the m-IscU mRNA in a 589 bp product. GAPDH was used as an internal control (Fig. 5A). As expected, the majority of the IscU mRNA in the control was of the mitochondrial splice variant (Fig. 5A, lane 1). However, in the patients a larger product consistent with the size of the cytosolic form was predominant (Fig. 5A, lane 2–3). Sequencing of the products verified that the control mRNA corresponded to the mitochondrial isoform. However, the mRNA from the patients did not correspond to the cytosolic isoform but contained an extra 100 bp from intron 5, inserted between exons 4 and 5 (Fig. 5B). The aberrant splicing product is due to the activation of cryptic acceptor and donor splice sites in the intron sequence, with the acceptor splice site located 6 bp downstream of the mutation (Fig. 5B). The level of the alternatively spliced transcript is not lower than that of the normal transcript as shown in Figure 5A. This was also confirmed by QRT–PCR (data not shown). The added intron sequence results in an alternative C-terminal of 15 amino acids followed by a stop codon.
In this paper, we describe the mapping and identification of the gene responsible for HML in a family from Northern Sweden. Using microsatellite mapping, we identified a region on chromosome 12q that was common to all fifteen individuals affected, with a maximum Lod score of 5.26 for marker D12S84. No other marker loci tested fulfilled the criteria of a recessive gene identical by descent. In order to restrict and reduce the disease-critical region, a number of additional markers were analysed in the region and haplotypes were constructed. Recombination events allowed us to define a disease-critical region of 1.6 Mb restricted by markers D12S1613 and D12S105. The disease-critical region contained 16 genes or putative transcripts. Of these, one gene, the iron–sulphur cluster assembly gene ISCU, stood out as an excellent candidate gene.
Iron–sulphur clusters are important prosthetic groups involved in many cellular processes, including electron transfer. The ISCU gene is expressed as two isoforms, generated through different splicing. IscU1 is located in the cytosol, whereas IscU2 is mitochondrial (8). HML patients have been shown to have a defect in complex II in skeletal muscles, with low amounts of the 30-kD (iron–sulphur) and 13.5-kD proteins but with near-normal levels of the 70-kD protein of complex II (3–5). Reduced levels of mitochondrial aconitase, a known substrate of IscU, have also been observed (11). SDH and aconitase both harbour iron–sulphur centres. The selective deficiency of the 30-kD iron–sulphur polypeptide of SDH leads to the suggestion that a common defect related to the iron–sulphur centres is present in both enzymes. This indicates a generalized abnormality in the synthesis, import, processing or assembly of a group of proteins containing iron–sulphur clusters (5). In order to establish whether ISCU is indeed the gene responsible for the phenotype seen in these patients, mutational analysis of all exon and intron sequences—together with 1000 bp of the promoter—of the ISCU gene was performed. The analysis did not reveal any disease-specific mutations in exons, exon–intron junctions or the ISCU promoter. We did however find one disease-specific mutation 382 bp into intron 5. The mutation changed a G to a C in a region resembling the ISRE, suggesting that this region may be involved in the regulation of the gene. We could also show, by gel-shift assays, that this sequence interacts with nuclear factors, further supporting the regulatory function of the region. Using nuclear extracts from RD4 cells we detected interaction with both the normal and the mutant sequence, but with different binding patterns. Using the normal sequence, two strong interactions were observed. When we used the mutant sequence in the same assay, a novel mutant-specific factor/complex appeared, while one of the interactions observed with the normal sequence was lost. When put in front of the SV40 promoter this region did act as an enhancer, but there was no difference in enhancer activity between the normal and mutant sequence.
In order to analyse if a difference in expression pattern could be observed in vivo RT–PCR was performed on mRNA extracted from muscle biopsies from patients and controls. The RT–PCR did not reveal any significant difference in the amount of IscU mRNA between patients and controls. We could however see a drastic difference in the splicing pattern. In the muscle biopsy from the control the IscU mRNA was, as expected, almost exclusively of the mitochondrial form. In contrast to this, the mRNA from the patients contained an additional 100 bp corresponding to part of intron 5. The mutation activates cryptic acceptor and donor splice sites in the intron sequence, with the acceptor splice site located 6 bp downstream of the mutation. This indicates that the mutation leads to the formation of a splicing regulatory region affecting the proximal cryptic splice sites. In agreement with this, the mutant sequence observed in our patients perfectly matches a sequence—TTTCAT—that was recently reported to be associated with alternative splicing (12). The normal sequence has a G in the C-position of this element, and thereby differs from this consensus sequence.
Taken together, these data strongly suggest that the mutation observed in the HML patients results in aberrant splicing of the IscU RNA, thereby resulting in a dysfunctional IscU protein with an incorrect C-terminal.
MATERIALS AND METHODS
In total, 50 individuals—15 affected and 35 unaffected relatives—from nine families participated in the study. A clinical and biochemical analysis of the patients has been performed previously (1,2,7). Informed consent was obtained from all participating individuals and the study was approved by the local ethical committee.
Genome-wide scan and haplotype analysis
Genomic DNA was prepared from whole blood using standard salt methods. Affected and unaffected members of the families were analysed using the ABI PRISM Linkage Mapping Set version 2.5 (Applied Biosystems, Foster City, CA) with an average distance of 10 cM. The primers were amplified according to the manufacturer's instructions, using multiplex PCR reactions. PCR products were resolved through 36-cm capillary arrays using POP-4 polymer on an AB 3100 or 3730 DNA sequencer. Genotypes were analysed using AB Genescan 2.1 and GeneMapper 2.0 (Applied Biosystems, Foster City, CA). Lod score analysis was performed using MLINK software (Linkage Package) (13). The genotypes for each individual were ordered into whole-chromosome haplotypes and regions of homozygosity were assessed visually.
The genomic sequence of ISCU was retrieved from CHIP Bioinformatics Tools and primers were designed using the Whitehead Primer3 program, or manually and purchased from DNA technology (Aarhus, Denmark). PCR reactions were performed under standard conditions. PCR products were separated on agarose gels and purified using Microcon YM-100 columns (Millipore, Bedford, MA). Sequencing was performed using the BigDye 3.1 kit according to the manufacturer's instructions (Applied Biosystems, Foster City, CA). Labelled products were resolved through 36-cm capillary arrays using POP-4 polymer on an ABI 3730 DNA sequencer and analysed using Sequencing Analysis 3.7 software and Autoassembler 2.1 (Applied Biosystems, Foster City, CA).
For the DNA binding reactions, 10 to 12.5 fmol of 32P-end-labelled double-stranded oligonucleotide, 1 µg of poly(dI-dC) and various amounts of RD4 nuclear extract were incubated in a final buffer concentration of 20 mm HEPES (pH 7.9), 50 mm KCl, 1 mm dithiothreitol, 0.005% Triton X-100 and 10% glycerol. The following double-stranded oligonucleotides were used in the electrophoretic mobility shift assay: Normal seq: 5′-TAAGCTCCAATCTTTGATTTCAGAATCTGTGA-3′ Mut seq: 5′-TAAGCTCCAATCTTTCATTTCAGAATCTGTGA-3′.
After incubation at room temperature for 30 min, protein-DNA complexes were resolved on a non-denaturing 5% polyacrylamide gel and run for 3 h at 150–200 V. Gels were fixed in 20% methanol/10% acetic acid for 15 min, dried and visualized by autoradiography.
To construct the vectors, 172-bp ISCU intron fragments covering the site of the mutation were amplified from chromosomal DNA from patients or control individuals, using primers containing Nhe I and Bgl II sites, respectively. The primers used were: Forward: 5′-GCTAGCTAAGGCTTGGGTTGAGA-3′ Reverse: 5′-AGATCTACCATCCAGACAGGAAGTG-3′.
PCR fragments were cloned into the pGEM T easy vector (Promega, Madison, WI), excised with NheI and BglII, and then cloned into the NheI–BglII site of the pGL2-promoter vector (Promega, Madison, WI).
For the luciferase assay, pGL2-promoter vector or pGL2-promoter vector with mutant or normal sequence inserted in front of the SV40 promoter were used to transfect RD4 cells (50 000 cells/cm2) using Lipofectamine™ 2000 (Invitrogen, Carlsbad, CA) according to the manufacturer's recommendations. The RD4 cell line (a generous gift from Dr Tracey Rouault) was cultured in DMEM-complete medium as previously described (8). pCMV-β- galactosidase was used as an internal control in all transfections. After transfection, cells were incubated for 45–49 h at 37°C in an atmosphere of 5% CO2. For α-interferon stimulation, 1000 injection units per ml of leukocyte interferon (BioNative, Umea, Sweden) were added to the cells 15 h before harvesting. After harvesting, the cells were lysed with 1× cell lysis buffer (Roche, Basel, Switzerland) and luciferase activity was measured using the Promega Luciferase Assay System (Promega, Madison, WI). The luciferase activity from each sample was normalized to the β-galactosidase internal control to compensate for variations in transfection efficiency. β-galactosidase activity was measured at 420 nm using ONPG as a substrate. All reporter assay data have been summarized from five independent transfections with duplicate samples. The error bars represent standard deviations, and differences in reporter activation were assessed with Student's t-test.
Reverse transcriptase–polymerase chain reaction
Muscle biopsies (vastus lateralis) from two patients and one control were used for RNA extraction. Muscle biopsies were kept frozen in liquid nitrogen. For extraction the muscle biopsies were homogenized using a ball mill grinder. RNA was extracted using the RNeasy micro kit (Qiagen, Hilden, Germany) according to the manufacturer's conditions. cDNA synthesis was performed with the SuperScript first-strand synthesis system for RT–PCR (Invitrogen, Carlsbad, CA) according to the manufacturer's conditions. Standard PCR was performed on cDNA. PCR products were resolved on a 2% agarose gel. QRT–PCR was performed as previously described (14). ISCU and GAPDH primers used for RT–PCR and QRT–PCR were as follows: ISCU-F: 5′-GCGCAAGCCGGCAAGATG-3′ ISCU-R: 5′-TCTAAGGTGACTGCGCAGCA-3′ GAPDH-F: 5′-AGGACTCATGTCCATGCCAT-3′ GAPDH-R: 5′-ACCCTGTTGCTGTAGCCAAA-3′ ISCUQRTF: 5′-GGCCCGACTCTATCACAAGA-3′ ISCUQRTR: 5′-CACCAGTCCAGTTCCAACAT-3′ GAPDHQRTF: 5′-GACAACAGCCTCAAGATCATC-3′ GAPDHQRTR: 5′-ATGGCATGGACATGAGTCCT-3′.
This project was supported by the Swedish Research Council, the Oskar Foundation and the Marcus Borgström Foundation.
We thank the members of the families for their invaluable participation in the study. We thank Susann Haraldsson for excellent technical assistance.
Conflict of Interest statement. None declared.