[Frontiers in Bioscience E3, 315-325, January 1, 2011]

Respiratory chain complex I is a mitochondrial tumor suppressor of oncocytic tumors

Franz A. Zimmermann1, Johannes A. Mayr1, Rene Feichtinger1, Daniel Neureiter2, Roman Lechner1, Christian Koegler3, Manfred Ratschek3, Husic Rusmir4, Karine Sargsyan4, Wolfgang Sperl1, Barbara Kofler1

1Department of Pediatrics, University Hospital Salzburg, Paracelsus Medical University, Austria, 2Department of Pathology, University Hospital Salzburg, Paracelsus Medical University, Austria, 3Institute of Pathology, Medical University of Graz, Austria, 4Biobank, Medical University of Graz, Austria


1. Abstract
2. Introduction
3. Materials and methods
3.1. Tumor samples
3.2. Immunohistochemical staining and analysis
3.3. Analysis of mtDNA 4. Results and review of the literature
4.1 .Renal oncocytoma
4.2. Thyroid oncocytoma
4.3. Parathyroid oncocytoma
4.4. Nasopharyngeal oncocytoma
4.5. Salivary gland oncocytoma
4.6. Adrenal gland oncocytoma
4.7. Pituitary gland oncocytoma
4.8. Eyelid oncocytoma
4.9. Oncocytic tumors of other tissue
5. Discussion
6. Acknowledgments
7. References


Oncocytic tumors, also called oxyphilic tumors, are characterized by hyperproliferation of mitochondria, which histologically presents as a fine granular eosinophilic cytoplasm. In accordance with the high mitochondrial density in oncocytomas, transcript levels of subunits of complexes of the oxidative phosphorylation (OXPHOS) system are increased. Hence, for a long time oncocytomas were presumed to have a highly active aerobic mitochondrial energy metabolism. Recently, detailed analysis of all OXPHOS complexes in a variety of oncocytomas revealed loss of complex I and compensatory up-regulation of the other complexes. In half of the oncocytoma cases examined the absence of complex I is caused by disruptive mutations in mitochondrial DNA encoding complex I subunits. The new data presented here on rare oncocytomas and the accompanying review of the literature clearly indicate that complex I deficiency in combination with up-regulation of mitochondria can be regarded as a hallmark of oncocytic tumor cells. Therefore, complex I of the respiratory chain has to be added to the growing list of mitochondrial tumor suppressors.


More than half a century ago, Nobel Prize laureate Otto Warburg posed the hypothesis of aerobic glycolysis, stating that normal cells tend to produce energy by glycolysis only under anaerobic conditions, whereas cancer cells do so even in the presence of oxygen (1). The first hint that a genetic defect of aerobic mitochondrial energy metabolism contributes to tumor formation was found in hereditary paraganglioma and familial pheochromocytoma, by demonstrating that mutations in the SDHB or SDHD genes, two nuclear encoded subunits of mitochondrial complex II, are associated with the development of these tumors (2) (3). Subsequently, germline mutations in the fumarate hydratase (FH) gene and reduced fumarate hydratase activity were shown to predispose to hereditary leiomyomatosis, renal cancer, and Leydig cell tumors (4) (5). Recently, glioblastoma multiforme, a very aggressive subtype of glioma, revealed specific point mutations in the isocitrate dehydrogenase 1 and 2 (IDH1 and IDH2) genes in 12-20% of the cases (6) (7). The same mutations were also found in >70% of astrocytomas, oligodendrogliomas and glioblastomas (8). These data indicate that the IDH1, IDH2, SDH and FH genes can be regarded as mitochondrial tumor suppressor genes (9) (10).

Besides the nuclear genes encoding SDH, IDH and FH, the mitochondrial genome (mtDNA) also has become of special interest in cancer research (11) (12) (13). The mtDNA, a circular DNA molecule within the mitochondrial matrix, encodes a total of 13 subunits of the respiratory chain, including seven subunits of complex I, one of complex III, three of complex IV and two of the F1F0-ATP synthase and 22 transfer RNAs. Over one hundred publications have reported the presence of somatic mtDNA mutations in a wide variety of tumor types (14) (11).

Although somatic mutations are frequent in cancers, their functional significance is uncertain for most tumor types (15). Coller et al. have shown by a mathematical approach that these mutations could also arise by chance without conferring any physiological advantage or tumorigenic effect (16) (17). This hypothesis does not apply equally to silent and non-silent mutations. In fact, most variants found in tumors are also present as polymorphic variants in the control population, whereas for those inducing amino acid changes, it is difficult to prove a role in tumor pathology. Therefore, functional investigations of specific alterations of aerobic energy metabolism might be used to identify isolated OXPHOS defects or at least should complement genetic analysis of mtDNA (15) (18).

Oncocyte is the general term applied to cells exhibiting a characteristic phenotype of a fine granular eosinophilic cytoplasm histologically and an increase in the number of mitochondria ultrastructurally (19). The abnormal increase in the amount of mitochondria was confirmed by electron microscopy in different studies (20) (21). The percentage of cytoplasm filled with mitochondria varies between 30% and 60%. This unique feature also led to the cytological designation of 'oxyphilic' tumor cells. In 1991, Ebner et al. were among the first to analyze the changes of the mitochondrial energy metabolism in more detail and found by enzyme histochemistry an up-regulation of all OXPHOS complexes in thyroid oncocytomas (22). In 1996, Heddi et al. analyzed the expression of nuclear and mitochondrial encoded genes of the respiratory chain in salivary gland oncocytoma and found a specific induction of OXPHOS gene expression as well as increased mtDNA content in oncocytomas, which was in accordance with the ultrastructural findings and the data of Ebner et al. (23). Based on transcriptional profiling data of 29 thyroid oncocytomas, Baris et al. hypothesized that, unlike most solid tumors, thyroid oncocytomas produce energy through an aerobic pathway (24). The reason for the extraordinary mitochondrial biogenesis in oncocytic cells remained unsolved until Savagner et al. reported in 2002 defective mitochondrial ATP synthesis in thyroid oncocytomas (25), suggesting a compensatory mechanism may be responsible for mitochondrial proliferation in response to a defective aerobic energy metabolism. The specific defect and genetic basis still remained unidentified but it became clear that the aberrant biogenesis of mitochondria in oncocytic cells bears intriguing similarities to mitochondrial encephalomyopathies. In the latter conditions, proliferation of abnormal mitochondria in muscle gives a phenotype termed ragged red fibers (RRF), which is a characteristic histochemical finding in skeletal muscle of patients with pathogenic mtDNA mutations (26).

In the present article we focus on recently published and new data regarding the specific alterations of aerobic energy metabolism at the biochemical and genetic levels in oncocytomas of different origintissues.


3.1. Tumor samples

The study population comprised salivary gland oncocytomas (n=13), parathyroid oncocytomas (n=6), pituitary gland oncocytomas (n=2), oncocytoma of the eyelid (n=1), and adrenal gland oncocytoma (n=1). Samples were obtained from the Biobank of the Medical University of Graz and the Department of Pathology, University Hospital of Salzburg. The study was approved by the ethics committee of the Medical University of Graz and was performed according to the Austrian Gene Technology Act and in accordance with the Helsinki Declaration of 1975 (revised 1983) and the guidelines of the Salzburg State Ethics Research Committee, being neither a clinical drug trial nor epidemiological investigation. All patients signed an informed consent concerning the surgical removal and therapy of the tumors. Furthermore, the study did not extend to examination of individual case records. The anonymity of the patients' samples has been ensured.

3.2. Immunohistochemical staining and analysis

The immunohistochemical staining of complex I (subunit NDUFS4, Abcam), complex II (subunit 70 kDa, Mitosciences), complex III (subunit core 2, Mitosciences), complex IV (subunit I, Mitosciences), complex V (subunit alpha, Mitosciences) and porin (31HL, Mitosciences) was performed as described previously (27), with the following change: for heat-induced epitope retrieval, EDTA-T buffer (1 mM EDTA, pH 8.0, 0.05% Tween 20) was used.

3.3. Analysis of mtDNA

Sequence analysis of the mitochondrially encoded subunits of complex I (ND1 to ND6 and ND4L), complex IV (COI to COIII) and mitochondrial tRNAs was performed as described previously (27).


4.1. Renal oncocytoma

Renal oncocytomas arise from the collecting duct and account for approximately 5% of all renal tumors (28). Renal oncocytomas are homogenous tumors with round cells which have lost their polarity (29). They are considered to be benign, in spite of rare cases of invasiveness.

The first data that complex I activity is decreased in oncocytomas were reported in 2003 by Simonnet et al. (30). A reduction of complex I activity to less than 65% compared to adjacent normal tissue has been demonstrated. Interestingly, complex I activity was also reduced in the corresponding normal tissue. Two-dimensional gel electrophoresis of mitochondrial proteins revealed decreased protein content of the nuclear encoded 75 kDa subunit of complex I in renal oncocytomas compared to the corresponding normal tissue (30). Two subsequent independent studies confirmed the complex I deficiency (31) (32). Complex I activity was either below the detection limit or <25% of the level in control kidney tissue (31) (32). Blue native gel electrophoresis of multisubunit enzyme complexes demonstrated a loss of assembled complex I, whereas complex V showed compensatory up-regulation (32). In accordance, deficiency of the complex I subunit NDUFS3 and upregulation of the complex IV subunits COX5A, COX5B and ATP5H from complex V was found by two dimensional gel electrophoresis in renal oncocytoma (33).

Two studies independently reported for the first time a genetic explanation for the striking loss of complex I in oncocytomas (31) (32) (Table 1): they found disruptive mutations (frame shift or nonsense) in subunits of complex I encoded by mtDNA (Table 1).

Gasparre et al. found in four of seven cases mutations, which generated a stop codon leading to truncation of the protein near the N-terminus. Additionally, one disruptive mutation was found in the cytochrome b gene, resulting in a loss of complex I and reduced complex II/III activity. We also detected in one sample a frame-shift mutation 14754_delC at the beginning of the cytochrome b gene, causing a loss of assembled complex III and in addition a deficiency of complex I (Table 1). It is rather astonishing that the absence of complex III causes a deficiency of complex I. Similar observations have been reported by Acin-Perez et al. in two cases of cytochrome b mutations (34). The authors argued that respiratory chain enzymes form supercomplexes and they found that complex I is formed de novo but is unstable in the absence of complex III. Finally, one case was reported to harbor a somatic mutation in the major H strand promoter, resulting in reduced activity of all OXPHOS complexes containing mitochondrial encoded subunits (31).

Mutation analysis of mtDNA by Mayr et al. revealed disruptive mutations in mitochondrially encoded complex I subunits in 11 of 15 tumor samples. Frame-shift mutations were found in nine cases, located in the first third of the gene and causing a truncated complex I subunit. A recurring frame-shift mutation in the ND1 gene at position 3571_3572insC seems to identify this site as a mutational hot spot, accounting for 18% of all mtDNA mutations detected in oncocytomas (Table 1). A leucine-to-proline substitution was detected in one case, possibly severe enough to cause a complex I deficiency. The A3243G transition, known as a frequent cause of the MELAS syndrome, was found in one renal oncocytoma. Interestingly, one patient developed two independent oncocytomas in the left and right kidneys. The oncocytoma from the right kidney harbored a frame-shift mutation in the ND5 gene, whereas the left kidney had a frame-shift mutation in the ND1 gene (32).

In oncocytomas where no potential pathogenic mtDNA mutations have been identified, it is most likely that nuclear components of complex I are affected. Nuclear gene sequences encoding and regulating the expression of complex I subunits and their assembly factors have, to our knowledge, not been analyzed in detail in renal oncocytomas. One reason might be the extensive number of candidate genes (>40), which would have to be sequenced.

The specific complex I deficiency in renal oncocytomas is in contrast to the alterations found in other renal tumors, where the Warburg effect is achieved by a significant reduction of all enzymes of the respiratory chain and accompanied by a massive reduction of mtDNA content. This mitochondrial phenotype in renal cell carcinoma can be explained by the loss of the von Hippel Lindau (VHL) gene, which has been shown to induce hypoxia-inducible factor 1 (HIF-1), a major factor contributing to reduced respiration in VHL-deficient cancer cells (35) (36). A general reduction of the OXPHOS complexes and reduced mtDNA copy number was also observed in neuroblastomas, but without alterations of the VHL protein (37).

4.2. Thyroid oncocytoma

Oncocytic thyroid tumors, also known as Huerthle cell tumors, are composed of at least 75% oncocytic tumor cells and account for 3% to 10% of thyroid tumors (38) (39). Oncocytic carcinoma of the thyroid appears to have a worse clinical outcome than non-oncocytic thyroid carcinoma, due to poor responsiveness of oncocytomas to radioiodine therapy (40).

In agreement with the mitochondrial hyperproliferation phenotype, coordinated overexpression of genes involved in mitochondrial biogenesis, such as nuclear respiratory factor 1 (NRF-1) and mitochondrial transcription factor A (TFAM), was observed in thyroid oncocytomas (41) (24). Also, death-associated protein 3 (DAP3), one of the constituents of the small subunit of the mitochondrial ribosome, was found to be up-regulated in thyroid oncocytic tumors (42). A first hint that thyroid oncocytomas exhibit a defect in aerobic energy metabolism came from Savagner et al., who showed a 50% lower level of ATP synthesis in thyroid oncocytomas in comparison with the basal rate of normal thyroid tissue (25). Interestingly, measurement of oxygen consumption by polarography of the same samples produced no evidence of OXPHOS defects.

The first nuclear gene mutation specific to oncocytic thyroid tumors was a heterozygous mutation of the nuclear encoded GRIM-19 gene, which encodes a complex I subunit necessary for assembly of the complex (43). Four of 26 cases of sporadic thyroid oncocytomas revealed heterozygous point mutations in the GRIM-19 gene. In one patient the mutation was also detected in non-cancerous thyroid tissue, peripheral blood, and in a blood sample from the patient's son, demonstrating the germline nature of the mutation. In none of six cases with known familial thyroid oncocytoma, nor in any of 20 non-oncocytic thyroid carcinomas, were mutations in the GRIM-19 gene detected (43).

The so-called common deletion (CD) of mtDNA, a 4977-bp deletion affecting the OXPHOS genes ATPase6, ATPase8, COIII, ND3, ND4L, ND4 and ND5, was detected with a higher prevalence in oncocytic thyroid tumors than in non-oncocytic thyroid tumors (44). Nevertheless, the percentage of mtDNA carrying the common deletion varied between 8% in oncocytic follicular carcinomas to 4% in oncocytic follicular adenomas and papillary carcinomas. The low level of mutational load could not explain the dramatic mitochondrial phenotype of oxyphilic cells.

Gasparre et al. found that 12 of 45 (27%) specimens showed disruptive mutations, either frame shift or nonsense, resulting in impaired protein synthesis (Table 1). In contrast to the mutations found in renal oncocytomas, not all mutations were homoplasmic; in some cases the mutation rate varied between 25% to 98% (45). XTC.UC1, a cell line derived from a metastasis of a Huerthle cell carcinoma, has a 3571_3572insC mutation in the ND1 gene, which was also found in three cases of thyroid oncocytomas. Interestingly, the XTC.UC1 cell line also carries a heteroplasmic (70%) point mutation (15557G>A) in the cytochrome b gene, which leads to an amino acid substitution at a non-conserved position (46). In a further 19 thyroid oncocytoma tissues, we used immunohistochemical analysis to verify that these mutations indeed can affect the level of complex I in the thyroid oncocytomas. We found a specific lack of complex I in all 19 cases, whereas other enzymes of mitochondrial energy metabolism showed compensatory up-regulation (27). Subsequent analysis of the mtDNA revealed disruptive mutations (six frame-shift and one nonsense) in complex I subunits in 7 of the 19 (37%) tumors, resulting in truncated subunits (27) (Table 1). Truncation of the ND subunits prohibits assembly of complex I and seems to lead to degradation of unassembled subunits even though the expression of complex I subunits is up-regulated to a similar extent as the expression of the other OXPHOS complexes in oncocytomas (25).

In contrast to our recent findings (27), Ebner et al. previously reported an increase in the enzymatic activity of complex I in nine thyroid oncocytomas (22). However, in this study the activity of complex I was determined by enzyme histochemistry of total NADH dehydrogenases, a method which also detects cytosolic NADH dehydrogenases and therefore can not be regarded as a measure of complex I activity.

4.3. Parathyroid oncocytoma

Oncocytic parathyroid adenomas account for 4% to 8% of all parathyroid adenomas (12). Analysis of the mtDNA in oncocytic parathyroid adenomas revealed nonsense mutations in mitochondrially encoded complex I subunits in 33% (4/12) of the samples studied (47).

Here we present data for six cases of oncocytic parathyroid adenomas (mean age: 63 years; range: 44 - 87; 33% male) for alterations of respiratory chain complexes by immunohistochemical staining and sequence analysis of the mtDNA. Oncocytic parathyroid tumor cells were negative for the complex I subunit NDUFS4 when compared to the adjacent non-cancerous tissue in all investigated specimens. Immunohistochemical staining of respiratory chain complex II subunit 70 kDa, complex III subunit core 2, complex IV subunit I, complex V subunit alpha, and the mitochondrial membrane protein porin, demonstrated an up-regulation of all of these proteins in all tumor samples compared to the adjacent normal tissue. Sequence analysis of the mitochondrially encoded subunits of complex I revealed disruptive mutations, either frame-shift or nonsense mutations, in the ND1 or ND5 gene. Three cases had frame-shift mutations (3571_3572insC, 13767delC, 14079delA) and one case exhibited a nonsense mutation (13031G>A) (Table 1).

In one of the cases, in addition to loss of complex I, a nodular loss of complex IV was found. Because <5% of the tumor cells were complex IV negative, no sequence analysis was possible (Figure 1). Combined defects of complex I and complex IV have also been described in other mitochondrial diseases, including MELAS (48) and MERRF (26) syndrome, both caused by mutations in mitochondrial tRNA genes.

4.4. Nasopharyngeal oncocytoma

Oncocytomas of the nasopharyngeal tract are extremely rare and have only been reported as single cases in the literature (49) (50). Previously, one case of a nasopharyngeal oncocytoma was described, which displayed a disruptive frame-shift mutation affecting the ND5 subunit of complex I (51). Associated loss of assembled complex I was demonstrated by immunohistochemistry and Western blot analysis. The disruptive ND5 mutation was homoplasmic in the tumor nodule. In the patient, pre-existing heteroplasmy of 2% was found in blood cells and 9% in the minor salivary gland. Two siblings of the patient also carried the ND5 mutation. One sibling had a heteroplasmy level of 12% mutated mtDNA in blood cells, whereas the other sibling carried only 2% mutated mtDNA in his leukocytes. Neither sibling developed cancer or any other disease. Three other siblings were deceased, one from cervical cancer, one from lung cancer, and the third from ischemic cardiopathy. Unfortunately, tissues from the deceased individuals were not available (51).

4.5. Salivary gland oncocytoma

Oncocytomas of the parotid gland account for <1% of salivary gland tumors and appear as oncocytic hyperplasia, benign oncocytoma, and rarely as oncocytic carcinoma (52) (53). Transcripts of mitochondrial genes in salivary gland oncocytomas were found to be increased up to 10-fold, which is in accordance with an observed 10-fold increase in mtDNA content. Accordingly, expression of nuclear genes involved in mitochondrial energy metabolism, ATP synthase subunit ß, and adenine nucleotide translocator isoform 2 (ANT2), was increased up to 30-fold in salivary gland oncocytomas, whereas only a 4-fold increase was observed in renal oncocytomas (23).

Here we report the analysis of 13 cases of salivary gland oncocytomas (mean age: 76 years; range: 63 - 89; 31% male). Ten cases of oncocytomas showed reduction of the complex I subunit NDUFS4 when compared to the corresponding normal tissue (data not shown). Two cases exhibited positive staining for complex I and another case showed heterogeneous staining. In these cases assembly of complex I seemed to be unaffected. However, by immunohistochemical methods it is impossible to detect enzymatic activity of complex I. In these cases it is possible that the enzymatic activity but not the assembly of complex I could be affected. Complex II to V and porin were up-regulated in all tumor samples compared to adjacent normal tissue (data not shown). MtDNA mutations were found in the complex I genes ND1, ND2, ND4 and ND6 (Table 1). In total, five (38%) of the tumors contained mutations (3571_3572insC, 5107delG, 14368delT, 11403G>A; 11559G>A) in mtDNA which explain the absence of assembled complex I (Table 1); all mutations were found in the first third of the affected proteins. In eight cases (62%) no pathogenic mutations were found.

4.6. Adrenal gland oncocytoma

Oncocytic neoplasms of the adrenal gland are extremely rare (54) (55) (56). To our knowledge, no analysis of mitochondrial functionality in adrenal gland oncocytomas has been reported in the literature. Therefore, we examined one case of an adrenal gland oncocytoma (female, 51 years old). A specific lack of complex I was also observed in this type of oncocytic tumor (Figure 2). As observed in all oncocytomas so far, up-regulation of OXPHOS complexes II to V and increased mitochondrial mass in the tumor sample were evident compared to the corresponding normal tissue (Figure 2). Sequence analysis revealed a deletion of one adenine in the ND4 gene, causing a frame shift at Lys93 (Table 1).

4.7. Pituitary gland oncocytoma

Pituitary gland oncocytomas are regarded as variants of null cell adenomas, as they present the same clinical features but with additional mitochondrial mass (56) (57). Because oncocytic changes occur in null cell adenomas, pituitary gland oncocytomas are suggested to illustrate the end stage of null cell adenomas (58). Nishioka et al. examined eight cases of pituitary gland oncocytomas and found a heterogeneous composition of cells staining positive for a 65-kDa mitochondrial protein and/or subunit II of complex IV. Because the identity of the 65-kDa mitochondrial protein is not known, it is unclear if the antibody used to recognize it is an appropriate tool to estimate mitochondrial mass.

We searched for respiratory chain deficiencies in two cases of oncocytic pituitary gland adenomas (case 1: male, 59 years old; case 2: female, 72 years old). Both cases showed loss of complex I (data not shown). In addition, one case exhibited a complex IV deficiency in about 80% of the tumor cells. Respiratory chain complex II subunit 70 kDa, complex III subunit core 2, complex V subunit alpha, and the mitochondrial membrane protein porin, were all up-regulated. In one case, insertion of a cytosine between positions 11872 and 11873 was found in the ND4 gene (Table 1). This mutation is located in the last third of the protein, causing a truncated ND4 subunit. In the second case, the one harboring the complex IV deficiency, additional mitochondrial tRNAs and complex IV subunits were sequenced but no pathogenic mutations were detected (data not shown).

4.8. Eyelid oncocytoma

Oncocytomas of the eyelid are extremely rare: only seven cases have been reported previously (59) and without any functional or genetic analysis of mitochondrial energy metabolism. For the present study only one case (female, 51 years old) was available for investigation of alterations of the respiratory chain complexes. As expected for an oncocytic tumor, a specific lack of complex I was also observed in the eyelid oncocytoma, and complex II, complex III, complex IV, complex V and porin all showed up-regulation, indicative of increased mitochondrial mass in the tumor sample compared to the corresponding normal tissue. Sequence analysis of mtDNA-encoded complex I subunits revealed no pathogenic mutation which could explain the loss of complex I.

4.9. Oncocytic tumors of other tissue

Mitochondrion-rich breast cancers are rare, although the prevalence of such tumors might be underestimated (60). One of five mitochondrion-rich breast tumors harbored a heteroplasmic disruptive mtDNA mutation and two cases had potentially pathogenic missense mutations (45) (Table 1).

Oncocytic tumors of the pancreas (61), the stomach (62) and the meninges (63) are very rare and have usually been described only as single case reports. Therefore, such types of oncocytomas were not available for the present study and information regarding alterations of mitochondrial metabolism in such tumors is nonexistent.


Oncocytomas from eight different organs are now described as a tumor subgroup which is characterized by hyperproliferation of mitochondria in response to complex I deficiency. Renal, thyroid and nasopharyngeal oncocytomas were previously reported to lack complex I (27, 30-32, 45, 51). Onococytomas of the parathyroid, the salivary gland, the pituitary gland, the eyelid and the adrenal gland were first described as complex I-deficient in this paper. Consistent with their phenotype of complex I deficiency a high proportion of oncocytomas harbor somatic mtDNA mutations, most of them affecting complex I-encoding subunits. Interestingly, disruptive mutations of mtDNA-encoded complex I subunits were generally homoplasmic. This strongly suggests that mtDNA mutations are selected during the process of cell division, further indicating that mtDNA mutations that disrupt complex I activity exert a positive selection pressure on the cell. Because complex I deficiency has been detected in adenoma as well as in carcinoma, we hypothesize that the metabolic switch is an early event in oxyphilic tumor progression. This hypothesis is further supported by the fact that complex I deficiency is associated with tumor cell proliferation in an animal cancer model (64).

Different methods have been used in the literature to identify complex I defects in either frozen or formalin-fixed paraffin-embedded (FFPE) tissue sections. Measurement of enzymatic activity gives the best information on the functional status of energy metabolism, being independent from assembly status. However, this method depends on either fresh or snap-frozen tumor specimens, which are not routinely available in most centers. Another problem of enzymatic analysis is the cellular composition of the samples. Only homogenous tumors such as oncocytomas of the kidney are appropriate for this kind of investigation because the analysis is performed using homogenized tissue lysates. Contamination with normal tissue will obscure the degree of an enzyme defect in a tumor. The use of paraffin-embedded tissues bypasses these problems as it permits the use of routinely fixed archival samples and analysis of respiratory defects at a single-cell level (27, 51). However, it has to be pointed out that this type of analysis only detects the loss of a protein and provides no information on the activity level in cases of detectable protein. Therefore, in the two cases of salivary gland oncocytomas, both with positive complex I staining, we cannot rule out that the enzymatic activity is lost.

In the past few years, evidence has emerged that complex I has a central role in apoptotic processes. It has been shown in cell culture experiments that the nuclear encoded complex I subunit NDUFS1 is disrupted by caspase-3 during staurosporine-induced apoptosis (65). Caspase-3 treatment turns complex I into a reactive oxygen species (ROS)-generating factor and causes loss of its NADH dehydrogenase activity. ROS production is followed by disruption of electron transport and transmembrane potential, loss of ATP production, and damage to mitochondria (Figure 3). Cells over-expressing a noncleavable variant of NDUFS1 maintained both transmembrane potential and ATP production, and ROS formation was not increased, mitochondrial morphology was unaffected, and plasma membrane integrity was stabilized (65). Therefore, loss of complex I could impair this apoptotic pathway, thereby giving complex I-deficient cells a survival advantage.

The combination of interferon-ß (IFN-ß) and retinoic-acid (RA) has synergistic effects on enhancing apoptosis and is used in cancer treatment. Application of IFN-ß/RA to a breast cancer cell line and to HeLa cells caused an up-regulation of various respiratory chain subunits, resulting in an increase of ROS production. Knockdown of the complex I subunit NDUFS3 or GRIM-19 in those same cell lines resulted in a reduction in complex I activity and less ROS production. Complex I knockdown cells showed increased survival and reduced apoptosis compared to wild-type cell lines (66). These data underline the fundamental role of complex I in cell death and its potential role in cancer treatment.

Another pathway that was shown to directly affect complex I involves granzyme A (GzmA) (67). Granzyme A is a serine protease released by cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells to trigger apoptosis via a caspase-independent apoptotic pathway. GzmA enters the mitochondrial matrix and cleaves the nuclear encoded complex I subunit NDUFS3, which also results in generation of ROS, disruption of the transmembrane potential and ATP production, and production of DNA damage (Figure 3). Cells over-expressing an uncleavable variant of NDUFS3 showed reduced ROS production and protection from both transmembrane potential disruption and nuclear DNA fragmentation (67). Granzyme B (GzmB), which is also released by CTLs and NK cells, kills target cells by activation of caspase-3 (68). Due to the facts that GzmA directly targets NDUFS3 and GzmB activates caspase-3, which cleaves NDUFS1, we hypothesize that complex I-deficient cells might contribute to tumor-cell resistance to attacks by CTLs and NK cells.

A recent study by Ishikawa et al. (64)showed that mtDNA mutations causing complex I deficiency are associated with tumor cell metastasis. Cytoplasmic hybrids (cybrids) were used to exchange the mtDNA between mouse tumor cell lines having low and high metastatic potential. Cybrids with high metastatic potential harbored mutations in the mitochondrially encoded complex I subunit ND6, causing a loss of complex I activity, whereas cybrids with wild-type mtDNA and normal complex I activity had low metastatic potential (64).

Activation of an oncogene or inhibition of a tumor suppressor gene usually leads to a diversity of changes in intracellular signaling pathways and cellular metabolism. Therefore, loss of complex I might also have consequences for intermediary metabolites, leading to different intracellular adaptations of the tumor cells. Complex II impairment leads to accumulation of succinate. Complex I deficiency might also result in accumulation of upstream metabolites within the citric acid cycle, including succinate and fumarate. These compounds have been shown to be used by tumor cells to generate building blocks for amino acid and lipid synthesis (69). Furthermore, accumulation of succinate and fumarate promotes stabilization of HIF-1. HIF-1 is a transcription factor that up-regulates a number of genes involved in aerobic glycolysis (e.g. glucose transporters, glycolytic enzymes, lactate dehydrogenase) (70). Currently, information on the metabolic changes induced by complex I impairment in tumor cells is lacking. In patients with complex I deficiency, accumulation of lactate is observed. Elevation of lactate is also a common feature of most solid tumors with increased aerobic glycolysis and is thought to contribute to tumor metastasis (71).

The fact that tumors with impaired aerobic energy metabolism have to rely on glycolysis opens therapeutic options. In particular, in cases with a clear, specific impairment of the respiratory chain such as oncocytomas, inhibition of glycolysis or deprivation of glucose should be considered for therapy. Glycolytic inhibitors serve as a classic example of cancer metabolism-targeting agents (72). Another option to deprive tumor cells of energy production by glycolysis is through a ketogenic diet. This is a high-fat, low-carbohydrate diet, which switches the primary source of energy metabolism from glucose to ketone bodies. Generation of ketone bodies is a physiological adaption to prolonged food restriction. The use of ketone bodies for ATP production is strictly dependent on the aerobic mitochondrial energy metabolism. Normal cells can use either glucose or ketone bodies for energy metabolism, whereas complex I-deficient cells cannot use ketone bodies at all. The first use of a ketogenic diet occurred in 1995 in the treatment of human malignant brain cancers. Both patients responded well to the ketogenic diet and no further chemo- or radiation therapy was needed (73). Because cells with impaired OXPHOS depend on glucose as a primary energy source, onocytic tumors, such as oncocytic thyroid carcinomas which have poor responsiveness to radioiodine, may be treated by a ketogenic diet or glycolysis inhibitors. This should even be effective in combination with classical tumor therapy.

In conclusion, data on different types of oncocytomas clearly demonstrate that impairment of respiratory chain complex I is a universal feature of oxyphilic tumors. Thus, oncocytoma could be classified as a mitochondrial disease. Given its potential role in cell proliferation and apoptosis, complex I can be considered a mitochondrial tumor suppressor.


Franz A. Zimmermann, and Johannes A. Mayr contributed equally to this work. This work was supported by the Salzburger Krebshilfe.


1. O. Warburg: On the origin of cancer cells. Science, 123(3191), 309-14 (1956)

2. B. E. Baysal, R. E. Ferrell, J. E. Willett-Brozick, E. C. Lawrence, D. Myssiorek, A. Bosch, A. van der Mey, P. E. Taschner, W. S. Rubinstein, E. N. Myers, C. W. Richard, 3rd, C. J. Cornelisse, P. Devilee and B. Devlin: Mutations in SDHD, a mitochondrial complex II gene, in hereditary paraganglioma. Science, 287(5454), 848-51 (2000)

3. D. Astuti, F. Latif, A. Dallol, P. L. Dahia, F. Douglas, E. George, F. Skoldberg, E. S. Husebye, C. Eng and E. R. Maher: Gene mutations in the succinate dehydrogenase subunit SDHB cause susceptibility to familial pheochromocytoma and to familial paraganglioma. Am J Hum Genet, 69(1), 49-54 (2001)

4. I. P. Tomlinson, N. A. Alam, A. J. Rowan, E. Barclay, E. E. Jaeger, D. Kelsell, I. Leigh, P. Gorman, H. Lamlum, S. Rahman, R. R. Roylance, S. Olpin, S. Bevan, K. Barker, N. Hearle, R. S. Houlston, M. Kiuru, R. Lehtonen, A. Karhu, S. Vilkki, P. Laiho, C. Eklund, O. Vierimaa, K. Aittomaki, M. Hietala, P. Sistonen, A. Paetau, R. Salovaara, R. Herva, V. Launonen and L. A. Aaltonen: Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet, 30(4), 406-10 (2002)

5. L. G. Carvajal-Carmona, N. A. Alam, P. J. Pollard, A. M. Jones, E. Barclay, N. Wortham, M. Pignatelli, A. Freeman, S. Pomplun, I. Ellis, R. Poulsom, M. A. El-Bahrawy, D. M. Berney and I. P. Tomlinson: Adult leydig cell tumors of the testis caused by germline fumarate hydratase mutations. J Clin Endocrinol Metab, 91(8), 3071-5 (2006)

6. F. E. Bleeker, S. Lamba, S. Leenstra, D. Troost, T. Hulsebos, W. P. Vandertop, M. Frattini, F. Molinari, M. Knowles, A. Cerrato, M. Rodolfo, A. Scarpa, L. Felicioni, F. Buttitta, S. Malatesta, A. Marchetti and A. Bardelli: IDH1 mutations at residue p.R132 (IDH1(R132)) occur frequently in high-grade gliomas but not in other solid tumors. Hum Mutat, 30(1), 7-11 (2009)

7. D. W. Parsons, S. Jones, X. Zhang, J. C. Lin, R. J. Leary, P. Angenendt, P. Mankoo, H. Carter, I. M. Siu, G. L. Gallia, A. Olivi, R. McLendon, B. A. Rasheed, S. Keir, T. Nikolskaya, Y. Nikolsky, D. A. Busam, H. Tekleab, L. A. Diaz, Jr., J. Hartigan, D. R. Smith, R. L. Strausberg, S. K. Marie, S. M. Shinjo, H. Yan, G. J. Riggins, D. D. Bigner, R. Karchin, N. Papadopoulos, G. Parmigiani, B. Vogelstein, V. E. Velculescu and K. W. Kinzler: An integrated genomic analysis of human glioblastoma multiforme. Science, 321(5897), 1807-12 (2008)
PMid:18772396    PMCid:2820389

8. H. Yan, D. W. Parsons, G. Jin, R. McLendon, B. A. Rasheed, W. Yuan, I. Kos, I. Batinic-Haberle, S. Jones, G. J. Riggins, H. Friedman, A. Friedman, D. Reardon, J. Herndon, K. W. Kinzler, V. E. Velculescu, B. Vogelstein and D. D. Bigner: IDH1 and IDH2 mutations in gliomas. N Engl J Med, 360(8), 765-73 (2009)
PMid:19228619    PMCid:2820383

9. E. Gottlieb and I. P. Tomlinson: Mitochondrial tumour suppressors: a genetic and biochemical update. Nat Rev Cancer, 5(11), 857-66 (2005)

10. C. B. Thompson: Metabolic enzymes as oncogenes or tumor suppressors. N Engl J Med, 360(8), 813-5 (2009)
PMid:19228626    PMCid:2848669

11. C. Santos, M. Martinez, M. Lima, Y. J. Hao, N. Simoes and R. Montiel: Mitochondrial DNA mutations in cancer: a review. Curr Top Med Chem, 8(15), 1351-66 (2008)

12. V. Maximo, J. Lima, P. Soares and M. Sobrinho-Simoes: Mitochondria and cancer. Virchows Arch, 454(5), 481-95 (2009)

13. P. Parrella, Y. Xiao, M. Fliss, M. Sanchez-Cespedes, P. Mazzarelli, M. Rinaldi, T. Nicol, E. Gabrielson, C. Cuomo, D. Cohen, S. Pandit, M. Spencer, C. Rabitti, V. M. Fazio and D. Sidransky: Detection of mitochondrial DNA mutations in primary breast cancer and fine-needle aspirates. Cancer Res, 61(20), 7623-6 (2001)

14. A. M. Czarnecka, W. Kukwa, T. Krawczyk, A. Scinska, A. Kukwa and F. Cappello: Mitochondrial DNA mutations in cancer--from bench to bedside. Front Biosci, 15437-60

15. A. M. Czarnecka, J. S. Czarnecki, W. Kukwa, F. Cappello, A. Scinska and A. Kukwa: Molecular oncology focus - Is carcinogenesis a 'mitochondriopathy'? J Biomed Sci, 17(1), 31

16. H. A. Coller, K. Khrapko, N. D. Bodyak, E. Nekhaeva, P. Herrero-Jimenez and W. G. Thilly: High frequency of homoplasmic mitochondrial DNA mutations in human tumors can be explained without selection. Nat Genet, 28(2), 147-50 (2001)

17. H. A. Coller, N. D. Bodyak and K. Khrapko: Frequent intracellular clonal expansions of somatic mtDNA mutations: significance and mechanisms. Ann N Y Acad Sci, 959434-47 (2002)

18. D. Meierhofer, J. A. Mayr, K. Fink, N. Schmeller, B. Kofler and W. Sperl: Mitochondrial DNA mutations in renal cell carcinomas revealed no general impact on energy metabolism. Br J Cancer, 94(2), 268-74 (2006)
PMid:16404428    PMCid:2361126

19. G. Tallini: Oncocytic tumours. Virchows Arch, 433(1), 5-12 (1998)

20. S. Jalalah, E. Horvath and K. Kovacs: Electron dense mitochondrial inclusions in a pituitary oncocytoma. J Submicrosc Cytol, 17(4), 667-71 (1985)

21. J. D. van der Walt, H. A. Reid, R. A. Risdon and J. H. Shaw: Renal oncocytoma. A review of the literature and report of an unusual multicentric case. Virchows Arch A Pathol Anat Histopathol, 398(3), 291-304 (1983)

22. D. Ebner, G. Rodel, I. Pavenstaedt and O. Haferkamp: Functional and molecular analysis of mitochondria in thyroid oncocytoma. Virchows Arch B Cell Pathol Incl Mol Pathol, 60(2), 139-44 (1991)

23. A. Heddi, H. Faure-Vigny, D. C. Wallace and G. Stepien: Coordinate expression of nuclear and mitochondrial genes involved in energy production in carcinoma and oncocytoma. Biochim Biophys Acta, 1316(3), 203-9 (1996)

24. O. Baris, F. Savagner, V. Nasser, B. Loriod, S. Granjeaud, S. Guyetant, B. Franc, P. Rodien, V. Rohmer, F. Bertucci, D. Birnbaum, Y. Malthiery, P. Reynier and R. Houlgatte: Transcriptional profiling reveals coordinated up-regulation of oxidative metabolism genes in thyroid oncocytic tumors. J Clin Endocrinol Metab, 89(2), 994-1005 (2004)

25. F. Savagner, B. Franc, S. Guyetant, P. Rodien, P. Reynier and Y. Malthiery: Defective mitochondrial ATP synthesis in oxyphilic thyroid tumors. J Clin Endocrinol Metab, 86(10), 4920-5 (2001)

26. D. C. Wallace, X. X. Zheng, M. T. Lott, J. M. Shoffner, J. A. Hodge, R. I. Kelley, C. M. Epstein and L. C. Hopkins: Familial mitochondrial encephalomyopathy (MERRF): genetic, pathophysiological, and biochemical characterization of a mitochondrial DNA disease. Cell, 55(4), 601-10 (1988)

27. F. A. Zimmermann, J. A. Mayr, D. Neureiter, R. Feichtinger, B. Alinger, N. D. Jones, W. Eder, W. Sperl and B. Kofler: Lack of complex I is associated with oncocytic thyroid tumours. Br J Cancer, 100(9), 1434-7 (2009)
PMid:19352385    PMCid:2694433

28. W. Thoenes, H. J. Rumpelt and S. Storkel: (Classification of renal cell carcinoma/tumors and their relationship to the nephron-collecting tubules system). Klin Wochenschr, 68(22), 1102-11 (1990)

29. N. Kuroda, M. Toi, M. Hiroi, T. Shuin and H. Enzan: Review of renal oncocytoma with focus on clinical and pathobiological aspects. Histol Histopathol, 18(3), 935-42 (2003)

30. H. Simonnet, J. Demont, K. Pfeiffer, L. Guenaneche, R. Bouvier, U. Brandt, H. Schagger and C. Godinot: Mitochondrial complex I is deficient in renal oncocytomas. Carcinogenesis, 24(9), 1461-6 (2003)

31. G. Gasparre, E. Hervouet, E. de Laplanche, J. Demont, L. F. Pennisi, M. Colombel, F. Mege-Lechevallier, J. Y. Scoazec, E. Bonora, R. Smeets, J. Smeitink, V. Lazar, J. Lespinasse, S. Giraud, C. Godinot, G. Romeo and H. Simonnet: Clonal expansion of mutated mitochondrial DNA is associated with tumor formation and complex I deficiency in the benign renal oncocytoma. Hum Mol Genet, 17(7), 986-95 (2008)

32. J. A. Mayr, D. Meierhofer, F. Zimmermann, R. Feichtinger, C. Kogler, M. Ratschek, N. Schmeller, W. Sperl and B. Kofler: Loss of complex I due to mitochondrial DNA mutations in renal oncocytoma. Clin Cancer Res, 14(8), 2270-5 (2008)

33. M. V. Yusenko, T. Ruppert and G. Kovacs: Analysis of differentially expressed mitochondrial proteins in chromophobe renal cell carcinomas and renal oncocytomas by 2-D gel electrophoresis. Int J Biol Sci, 6(3), 213-24
PMid:20440404    PMCid:2862395

34. R. Acin-Perez, M. P. Bayona-Bafaluy, P. Fernandez-Silva, R. Moreno-Loshuertos, A. Perez-Martos, C. Bruno, C. T. Moraes and J. A. Enriquez: Respiratory complex III is required to maintain complex I in mammalian mitochondria. Mol Cell, 13(6), 805-15 (2004)

35. D. Meierhofer, J. A. Mayr, U. Foetschl, A. Berger, K. Fink, N. Schmeller, G. W. Hacker, C. Hauser-Kronberger, B. Kofler and W. Sperl: Decrease of mitochondrial DNA content and energy metabolism in renal cell carcinoma. Carcinogenesis, 25(6), 1005-10 (2004)

36. H. Zhang, P. Gao, R. Fukuda, G. Kumar, B. Krishnamachary, K. I. Zeller, C. V. Dang and G. L. Semenza: HIF-1 inhibits mitochondrial biogenesis and cellular respiration in VHL-deficient renal cell carcinoma by repression of C-MYC activity. Cancer Cell, 11(5), 407-20 (2007)

37. R. G. Feichtinger, F. Zimmermann, J. A. Mayr, D. Neureiter, C. Hauser-Kronberger, F. H. Schilling, N. Jones, W. Sperl and B. Kofler: Low aerobic mitochondrial energy metabolism in poorly- or undifferentiated neuroblastoma. BMC Cancer, 10149

38. S. Sidhu: Hurthle cell neoplasm of the thyroid gland. ANZ J Surg, 78(3), 115 (2008)

39. M. R. Hanief, L. Igali and D. Grama: Hurthle cell carcinoma: diagnostic and therapeutic implications. World J Surg Oncol, 227 (2004)

40. Y. Kushchayeva, Q. Y. Duh, E. Kebebew and O. H. Clark: Prognostic indications for Hurthle cell cancer. World J Surg, 28(12), 1266-70 (2004)

41. F. Savagner, D. Mirebeau, C. Jacques, S. Guyetant, C. Morgan, B. Franc, P. Reynier and Y. Malthiery: PGC-1-related coactivator and targets are upregulated in thyroid oncocytoma. Biochem Biophys Res Commun, 310(3), 779-84 (2003)

42. C. Jacques, J. F. Fontaine, B. Franc, D. Mirebeau-Prunier, S. Triau, F. Savagner and Y. Malthiery: Death-associated protein 3 is overexpressed in human thyroid oncocytic tumours. Br J Cancer, 101(1), 132-8 (2009)

43. V. Maximo, T. Botelho, J. Capela, P. Soares, J. Lima, A. Taveira, T. Amaro, A. P. Barbosa, A. Preto, H. R. Harach, D. Williams and M. Sobrinho-Simoes: Somatic and germline mutation in GRIM-19, a dual function gene involved in mitochondrial metabolism and cell death, is linked to mitochondrion-rich (Hurthle cell) tumours of the thyroid. Br J Cancer, 92(10), 1892-8 (2005)
PMid:15841082    PMCid:2361763

44. V. Maximo, P. Soares, J. Lima, J. Cameselle-Teijeiro and M. Sobrinho-Simoes: Mitochondrial DNA somatic mutations (point mutations and large deletions) and mitochondrial DNA variants in human thyroid pathology: a study with emphasis on Hurthle cell tumors. Am J Pathol, 160(5), 1857-65 (2002)
PMid:12000737    PMCid:1850872

45. G. Gasparre, A. M. Porcelli, E. Bonora, L. F. Pennisi, M. Toller, L. Iommarini, A. Ghelli, M. Moretti, C. M. Betts, G. N. Martinelli, A. R. Ceroni, F. Curcio, V. Carelli, M. Rugolo, G. Tallini and G. Romeo: Disruptive mitochondrial DNA mutations in complex I subunits are markers of oncocytic phenotype in thyroid tumors. Proc Natl Acad Sci U S A, 104(21), 9001-6 (2007)
PMid:17517629    PMCid:1885617

46. E. Bonora, A. M. Porcelli, G. Gasparre, A. Biondi, A. Ghelli, V. Carelli, A. Baracca, G. Tallini, A. Martinuzzi, G. Lenaz, M. Rugolo and G. Romeo: Defective oxidative phosphorylation in thyroid oncocytic carcinoma is associated with pathogenic mitochondrial DNA mutations affecting complexes I and III. Cancer Res, 66(12), 6087-96 (2006)

47. J. Costa-Guda, T. Tokura, S. I. Roth and A. Arnold: Mitochondrial DNA mutations in oxyphilic and chief cell parathyroid adenomas. BMC Endocr Disord, 78 (2007)

48. Y. Goto, S. Horai, T. Matsuoka, Y. Koga, K. Nihei, M. Kobayashi and I. Nonaka: Mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes (MELAS): a correlative study of the clinical features and mitochondrial DNA mutation. Neurology, 42(3 Pt 1), 545-50 (1992)

49. D. K. Banerjee, K. R. Korzec, K. Kim and L. Fanelly: Nasopharyngeal oncocytoma. Otolaryngol Head Neck Surg, 113(1), 136-7 (1995)

50. F. W. Albers, C. A. Cuvelier, M. T. Renders and P. B. van Cauwenberge: Oncocytoma of the nasopharynx. Rhinology, 31(1), 41-3 (1993)

Key Words: Respiratory Chain Complex I, Oncocytic Tumors, Respiratory Chain, Warburg Effect, Mitochondria, Mitochondrial DNA, Tumor Suppressor

Send correspondence to: Barbara Kofler, Department of Pediatrics, University Hospital Salzburg, Paracelsus Medical University, Muellner Hauptstr. 48, A-5020 Salzburg, Austria, Tel: 43-662-4482, Fax: 43-662-4482, E-mail:b.kofler@salk.at