Article Text

Download PDFPDF

Immunohistochemical analysis of the oxidative phosphorylation complexes in skeletal muscle from patients with mitochondrial DNA encoded tRNA gene defects
  1. B De Paepe1,
  2. J Smet1,
  3. M Lammens2,
  4. S Seneca3,
  5. J-J Martin4,
  6. J De Bleecker5,
  7. L De Meirleir1,
  8. W Lissens3,
  9. R Van Coster1
  1. 1
    Department of Pediatric Neurology & Metabolism, University Hospital, Ghent, Belgium
  2. 2
    Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands
  3. 3
    Center for Medical Genetics, University Hospital, Brussels, Belgium
  4. 4
    Department of Neuropathology & Born Bunge Institute, University Hospital, Antwerp, Belgium
  5. 5
    Department of Neurology, University Hospital, Ghent, Belgium
  1. Dr B De Paepe, Department of Pediatric Neurology and Metabolism, UZ 3K5, De Pintelaan 185, B9000 Ghent, Belgium; boel.depaepe{at}


Background: Mitochondrial diseases display a heterogeneous spectrum of clinical phenotypes and therefore the identification of the underlying gene defect is often a difficult task.

Aims: To develop an immunohistochemical approach to stain skeletal muscle for the five multi-protein complexes that organise the oxidative phosphorylation (OXPHOS) in order to improve the diagnostic workup of mitochondrial defects.

Methods: OXPHOS complexes were visualised in skeletal muscle tissue using antibodies directed against different subunits. The staining patterns of patients with heteroplasmic defects in mtDNA tRNA genes were compared with those of normal and disease controls.

Results: Normal skeletal muscle displayed a checkerboard staining pattern for complexes I to V due to the higher mitochondrial content of slow muscle fibres versus fast fibres. In patients with tRNA defects, a much more heterogeneous staining pattern was observed for complex I (all six patients) and complex IV (4 of 6 patients): a mosaic staining pattern in which individual fibres displayed staining intensities that ranged from strong to negative. Ragged red fibres (RRFs) in patients with MERRF (myoclonic epilepsy and ragged red fibres) were all complex I and IV negative, while in patient with MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes) the majority of RRFs were complex I negative and complex IV positive.

Conclusion: Immunohistochemical detection of OXPHOS complexes could represent a valuable additional diagnostic tool for the evaluation of mitochondrial cytopathy. The technique helps to detect heteroplasmic mtDNA defects. Staining for complex I in particular was able to identify two tRNA patients that stayed undetected with routine histochemical evaluation.

View Full Text

Statistics from

ATP production in mitochondria occurs through the process of oxidative phosphorylation (OXPHOS), which is accomplished by the concerted action of a set of five multiprotein complexes: NADH-ubiquinol oxidoreductase (complex I), succinate ubiquinone reductase (complex II), ubiquinol cytochrome c oxidoreductase (complex III), cytochrome c oxidase (complex IV) and ATP synthase (complex V). Genetic defects of OXPHOS have been recognised as important causes of human disease.1 The structural components of OXPHOS and the accessory proteins necessary for its functioning lie encoded within both the nuclear DNA (nuDNA) and mitochondrial DNA (mtDNA). The latter contains 37 genes, all of which are essential for maintaining a functional respiratory chain.2 The majority of pathogenic mtDNA defects are heteroplasmic point mutations localised in tRNA genes,3 mutations that affect the intramitochondrial protein translation. This group of diseases includes myoclonic epilepsy and ragged red fibres (MERRF) and mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes (MELAS).

A number of histochemical stains are routinely carried out when patients are suspected of having a mitochondrial disorder. Histochemical stainings of muscle provide OXPHOS-related information: the NADH tetrazolium reductase (NADH-TR) reaction, succinate dehydrogenase (SDH) and cytochrome c oxidase (COX) activity stainings, and the myosin ATPase stain (pH 4.3, 4.6, 10.0). The myological abnormalities found in muscle biopsies of patients with tRNA gene mutations are mostly non-specific, and include the presence of COX negative and SDH hyperactive fibres. Often fibres with massive mtDNA proliferation can be visualised by the modified Gomori trichrome stain as so-called ragged red fibres (RRFs).

Today, immunohistochemistry is not routinely included in the diagnostic investigations for mitochondrial disorders. Immunostaining in muscle for complex IV subunits however, has been used successfully in the past to differentiate between complex IV deficiencies that originate from mtDNA or nuDNA gene defects.4 We evaluated the potential of OXPHOS immunodetection to differentiate skeletal muscle of patients carrying mutations in mtDNA encoded tRNA genes from muscle of normal and diseased controls.


Patients and controls

Six patients (P1–P6) with characterised heteroplasmic defects in mtDNA encoded tRNA genes were evaluated. P1 carried the classical tRNALys A8344G mutation, with a mutation load exceeding 99%, and had MERRF. P2 had the same mutation with a mutation load of 98%, but had neither myoclonic epilepsy nor RRFs. Severe involvement of the basal ganglia and brainstem suggestive for Leigh syndrome was seen on cerebral MRI. He died before the age of 1 year.5 P3 and P4 carried the mtDNA tRNALeu A3251G mutation and had classical MELAS. P5 carried an adenine insertion in the mtDNA tRNATyr gene and a nucleotide alteration A14639G in the ND6 gene.6 P6 carried a new mutation in tRNACys (to be submitted for publication). Four patients with inclusion body myositis, one patient with homoplasmic mtDNA ND1 mutation and one with a defect in the nuDNA encoded SURF-1 gene were used as myopathic disease controls. Ten biopsies from patients that presented with various non-mitochondrial complaints were used as normal controls.


Monoclonal antibodies directed against subunits of the OXPHOS complexes, the mitochondrial protein porin, and against double stranded DNA (dsDNA) were commercially available and are listed in table 1.

Table 1 Mouse monoclonal antibodies

Enzymatic studies

Spectrophotometric assays in skeletal muscle were performed measuring complex I (NADH:ubiquinone oxidoreductase, rotenone sensitive),7 complex II (succinate:ubiquinone oxidoreductase, malonate sensitive),8 complex III (ubiquinone:cytochrome c oxidoreductase, antimycine sensitive),9 complex IV (cytochrome c oxidase),10 complex V (ATPsynthase, oligomycine sensitive)8 and citrate synthase11 enzyme activities. Blue native-polyacrylamide gel electrophoresis (BN-PAGE) followed by in-gel staining for the enzymatic activity of the OXPHOS complexes was performed according to methods described previously.12

Immunohistochemical protocols

Paraffin embedded tissues were deparaffinised in xylene and rehydrated in ethanol solutions. Paraffin and frozen sections (8 μm thick) were permeabilised with ice cold acetone for 2 min, and blocked with 2.5% BSA in phosphate buffered saline for 30 min. Primary antibodies were diluted in the same solution. Sections were incubated for two hours at room temperature. Peroxidase-labelled secondary antibodies were detected with the LSAB2 kit and DAB+ chromogen (Dako, Glostrup, Denmark), alkaline phosphatase labelled EnVision polymer with fast red chromogen (Dako). Nuclei were counterstained with haematoxylin (Sigma, St Louis, Misouri, USA) and slides were mounted with aquatex (Merck, Darmstadt, Germany).

Figure 1 Blue native-polyacrylamide gel electrophoresis for complex V shows normal amounts of holocomplex in P5 and reduced amounts of holocomplex in P1, compared to control skeletal muscle (C). Muscle of both patients contains a catalytically active complex V subcomplex.


In the patients, defects in mtDNA tRNA genes were characterised using PCR single-stranded conformational polymorphism analysis followed by direct sequencing.13 Standard spectrophotometric analysis of OXPHOS complexes showed a combined decrease of complex I and IV activity in P1 and normal levels in P2 and P5. BN-PAGE and subsequent activity staining of the OXPHOS complexes showed decreased complex I and IV activities in P1, and catalytically active subcomplexes of complex V in P1 and P5 (fig 1). The presence of complex V subcomplexes is considered as a marker for defects in the intramitochondrial protein translation. Table 2 summarises the results.

Table 2 Patient information and OXPHOS activity

An array of antibodies was tested for the specific detection of the five OXPHOS complexes. The antibodies best suited for OXPHOS immunodetection were selected, and finally five antibodies were retained. These were directed against four nuclearly encoded subunits (NDUFS7, SDHB, UQCRC2, ATP5A1), and one mtDNA encoded (MTCO1) OXPHOS gene product. The immunohistochemical pattern of OXPHOS staining in normal skeletal muscle displayed a checkerboard pattern of distribution due to the different mitochondrial load of type 1 and type 2 myofibres (fig 2A). The same pattern could be visualised using an antibody against porin (fig 2B). In the patients with tRNA mutations, a much more heterogeneous pattern for complex I and complex IV was seen. Staining intensities of individual fibres varied from strong to negative (fig 2C). All six patients displayed this mosaic staining pattern for complex I. Four of these six patients also had a mosaic staining for complex IV. The patient with homoplasmic mtDNA ND1 mutation and the patient with nuDNA SURF-1 gene mutation did not display this mosaic staining pattern. Inclusion body myositis patients and muscle from control subjects older than 65 years contained only occasional complex I and complex IV negative fibres. The OXPHOS immunostaining patterns were comparable in frozen and paraffin embedded muscle tissue.

Figure 2 Immunohistochemical staining of oxidative phosphorylation complexes. Normal control (A–B): frozen sections stained with antibodies against complex IV (A) and porin (B) show normal variation of staining in fast and slow muscle fibres (peroxidase-DAB). Patient 1 (C): frozen section of skeletal muscle of a patient with MERRF (myoclonic epilepsy and ragged red fibres) shows a heterogeneous staining pattern for complex I (peroxidase-DAB). Patient 3 (D–I): consecutive frozen sections of a patient with MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes) show a ragged red fibre (RRF) which is complex I negative (D), and complex II (E) and complex IV (F) positive (alkaline phosphatase–fast red). Consecutive paraffin sections stained for complex II (G), complex III (H) and complex IV (I) show both complex IV positive and negative RRFs (peroxidase–DAB).

RRFs were detected using antibodies against double stranded DNA and against porin in P1, P3 and P4, but RRFs could also be identified as OXPHOS hyper-reactive fibres due to their massive mitochondrial load. The results of the immunohistochemical analysis confirmed the results obtained by enzymatic staining and showed that in MERRF patients RRFs were complex I and IV negative, while in the MELAS patients the majority of RRFs were complex I negative and complex IV positive (fig 2D–I). In inclusion body myositis patients all RRFs were complex I and IV negative. Table 3 shows the results of histochemical and immunohistochemical stainings.

Table 3 Results of histochemical and immunohistochemical stainings


Muscle histological evaluation has been in use for decades to examine mitochondrial defects. Routine histochemistry includes modified Gomori trichrome staining for the detection of RRFs, and stains that assess the enzymatic activity of the OXPHOS complexes. Histology may provide clues to support a diagnosis of mitochondrial disorders, but rarely offers decisive evidence. Pathological features such as the presence of COX negative fibres can equally be seen in other neuromuscular disorders or in natural physiological processes such as aging. The diagnosis of a specific mitochondrial disease finally results from analysing all clinical, histological, biochemical and genetic investigations.14 Considerable progress has been made in recent years to further improve biochemical OXPHOS testing. A sensitive novel method to evaluate OXPHOS activities is BN-PAGE, a technique that has been further developed to fit the demands of mitochondrial disease diagnosis.12 The presence of catalytically active subcomplexes of complex V is a powerful indication for gene defects that compromise mitochondrial protein synthesis.15 16 In our cohort of mtDNA tRNA patients, both subjects tested displayed subcomplexes of complex V on BN-PAGE. We feel that, in view of better and faster diagnosis, the microscopic testing of muscle biopsies could also be improved. In this respect, the immunohistochemical staining protocol we described here could be added to the routine histochemical battery of tests.

The most challenging aspect of diagnosing mitochondrial disease is the dual genomic origin of OXPHOS protein complexes, further complicated by the heteroplasmy often associated with mtDNA mutations. mtDNA encoded tRNA defects lead to disturbed intra-mitochondrial protein synthesis. The OXPHOS complexes I and IV in particular are sensitive as they contain the most mtDNA encoded structural subunits. Routine spectrophotometric analysis of OXPHOS enzyme activities translates in a combined reduction of complex I and complex IV activity only if the amount of mutated mtDNA exceeds a threshold level. As a consequence, spectrophotometric analysis is unable to identify heteroplasmic mtDNA defects where the mutation load is low. However, this typical feature of many mitochondrial diseases may well be the key to further improve diagnostic testing. It has been shown that heteroplasmic mtDNA defects present in cultured skin fibroblasts translate in a mosaic staining pattern for complex I and IV,17 18 and that immunohistochemistry with antibodies against mtDNA and nuDNA encoded complex IV subunits could differentiate between defects of different genetic origin.4 Here, we showed that the immunohistochemical staining pattern of muscle, when stained for OXPHOS complexes I to V, can provide valuable indications as to the genetic origin of the defect. We conclude that immunohistochemistry of either frozen or paraffin embedded muscle could be a valuable asset in the diagnostic workup of mitochondrial disease. The mosaic staining pattern of OXPHOS complexes I and IV in particular can be used to discern defects caused by different heteroplasmic mutations in the mtDNA, and such patients should be screened for mutations in a tRNA or rRNA gene, or for a large-scale mtDNA deletion or depletion.

In skeletal muscle tissue with mitochondrial defects, a subset of muscle fibres with massive mitochondrial proliferation is often seen, in an attempt to compensate the failing OXPHOS capacity. Interestingly, these RRFs react differently when stained for COX activity when MERRF and MELAS patients are compared. COX positive RRFs are typically observed in MELAS, but are absent from MERRF tissues.19 Single fibre PCR studies have shown that in MELAS only high levels of mutated mtDNA lead to COX negative RRFs.20 21 In other words, MELAS mutations need to be in overload in order to compromise the COX activity of individual muscle fibres. Our experiments confirmed that, although they are complex IV positive, the RRFs of MELAS patients are invariably complex I immunonegative. Possibly mitochondria require higher amounts of wild-type mtDNA to build up complex I than is necessary to form complex IV. This effect may be due to the higher numbers of mtDNA encoded structural subunits in complex I (7 mtDNA encoded subunits versus 3 for complex IV). In analogy, it has been reported that cultured skin fibroblasts of MELAS patients have isolated complex I deficiency, while MERRF fibroblasts showed a combined I and IV defect.22

Take-home messages

  • Immunohistochemical analysis of oxidative phosphorylation (OXPHOS) complexes could become a valuable technique in the workup of skeletal muscle pathology, both on frozen and paraffin embedded material.

  • Immunohistochemical analysis of OXPHOS complexes I and IV is an effective tool to identify heteroplasmic mitochondrial DNA mutations through the mosaic staining pattern observed in muscle.

  • Ragged red fibres (RRFs) of patients with MERRF (myoclonic epilepsy and ragged red fibres) and normal aging muscle are complex I and IV negative, while RRFs of patients with MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes) are mostly complex I negative and complex IV positive.


View Abstract


  • Funding: This study was supported by grants from the Fund for Scientific Research Flanders (G.0666.06), Association belge contre les maladies neuromusculaires (ABMM) and KidAuQuai.

  • Competing interests: None.

  • Ethics approval: Ethics approval was obtained.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.