Defective Mitochondrial ATP Synthesis in Oxyphilic ... - Semantic Scholar

0013-7227/01/$03.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 86(10):4920 – 4925 Copyright © 2001 by The Endocrine Society

Defective Mitochondrial ATP Synthesis in Oxyphilic Thyroid Tumors F. SAVAGNER, B. FRANC, S. GUYETANT, P. RODIEN, P. REYNIER,

AND

Y. MALTHIERY

Inserm EMI-U 00-18 (F.S., P.R., P.R., Y.M.), Laboratoire de Biochimie et Biologie Mole´culaire, Angers F-49033; Laboratoire d’Anatomie Pathologique (B.F.), Hoˆpital Ambroise Pare´, Boulogne F-92104; Laboratoire d’Anatomie Pathologique (S.G.); and Service d’Endocrinologie (P.R.), Nutrition et Me´decine Interne, Angers F-49033, France Oxyphilic tumors (oncocytomas or Hu ¨ rthle cell tumors) form a rare subgroup of thyroid tumors characterized by cells containing abundant mitochondria. The relationship between the mitochondrial proliferation and the pathogenesis of these tumors is unknown. We have assessed the expression of the mitochondrial ND2 and ND5 (subunits of the nicotinamide adenine dinucleotide dehydrogenase complex) genes and the nuclear UCP2 (uncoupling protein 2) gene in 22 oxyphilic thyroid tumors and matched controls. The consumption of

O

XYPHILIC THYROID TUMORS, also known as oncocytomas or Hu¨rthle cell tumors, represent a rare subgroup of follicular thyroid neoplasms, characterized by cells with a distinctive eosinophilic cytoplasm (1). The cytoplasmic eosinophilia is owing to the abundance of morphologically altered mitochondria in the majority of tumor cells (1, 2). However, the relationship between mitochondrial proliferation and the histogenesis of oxyphilic tumors is unknown. The development of mitochondria involves the synthesis of proteins encoded by mitochondrial DNA (mtDNA), which carries the genes for the essential subunits of the respiratory chain complexes, as well as by nuclear DNA (nDNA). Electrons from the nicotinamide adenine dinucleotide generated by glycolysis in the cell are transported into the mitochondria in which the flow of electrons between the respiratory chain complexes supplies the energy used by ATP synthase to produce ATP. An alternative source of energy, independent of ATP synthase, is provided by the uncoupling protein (UCP), which plays an important role in energy homeostasis (3). The analysis of mtDNA in several oxyphilic tumors has shown that the abnormally high histochemical activity of the respiratory chain complexes is associated with a great increase in the amount of wild-type mtDNA (4). The frequency of aneuploid or polyploid cells in oxyphilic thyroid tumors suggests the presence of anomalies in nuclear genes (5). Mitochondrial proliferation has also been reported in mitochondrial diseases associated with respiratory chain defects or coupling defects between the respiratory chain and ATP production (6 –9). Among the UCPs that induce thermogenesis during mitochondrial respiration (3), UCP2 is the unique form expressed in many tissues and cell types. In particular, UCP2 is expressed in thyroid tissue, whereas the expression of UCP1 and UCP3 is limited to brown fat adipose

Abbreviations: mtDNA, Mitochondrial DNA; nDNA, nuclear DNA; rDNA, ribosomal DNA; UCP, uncoupling protein.

oxygen in mitochondria from tumors was determined by polarography. ATP assays were used to explore the mitochondrial respiratory chain activity and the oxidative phosphorylation coupling in seven fresh thyroid tumors and controls. Adenosine triphosphate synthesis was significantly lower in all the tumors, compared with controls, suggesting that a coupling defect in oxidative phosphorylation may be a cause of mitochondrial hyperplasia in oxyphilic thyroid tumors. (J Clin Endocrinol Metab 86: 4920 – 4925, 2001)

tissue and muscle/white fat tissues, respectively (10). However, the histochemical investigation of the key nuclear components involved in the oxidative phosphorylation process has revealed no coupling defects in oxyphilic thyroid tumors (11). The proliferation of mitochondria in oxyphilic tumors might result from the induction of genes involved in mitochondrial biogenesis. The patterns of mitochondrial transcripts (especially ND2 and ND5) as well as nuclear transcripts from certain genes coding for proteins involved in oxidative phosphorylation differ according to whether the tumor is a renal oxyphilic tumor or a salivary gland oxyphilic tumor (12). Changes in mtDNA transcription and mitochondrial mRNA stability were observed in the former but not in the latter, suggesting that the process of mitochondrial proliferation varies according to the origin of the tumor. We examined thyroidectomy specimens from 22 anonymous patients. In each case, tissue removed from the normal part of the thyroid gland served as a control for the oxyphilic thyroid tumor. For each set of paired specimens, we determined the gene expression profiles of the mitochondrial ND2 and ND5 genes as well as the nuclear UCP2 gene involved in energy production. Fresh tissue samples, obtained from 7 of the 22 patients, were analyzed by polarography to investigate the mitochondrial respiratory chain activity, the rate of oxidative phosphorylation (ADP/oxidation ratio) and ATP assays were used to determine the mitochondrial ATP synthesis. Materials and Methods Thyroid tissue samples Twenty-two benign or malignant oxyphilic thyroid tumors, diagnosed between 1992 and 2000 at the Ambroise Pare´ Hospital, Paris (15 cases) and the University Hospital, Angers (7 cases), were included in the study. All the samples used were rendered anonymous (i.e. all patient identifiers were deleted before the study). The cases were con-

4920

Savagner et al. • ATP Synthesis in Oxyphilic Thyroid Tumors

secutive and unselected apart from exclusions on account of insufficient material or association of the tumors with chronic thyroiditis. Nineteen of the tumors were follicular oxyphilic adenomas (six of which were trabecular) and three were follicular oxyphilic carcinomas. Five of the adenomas were associated with a multinodular goiter. The diagnoses were made according to the World Health Organization classification (1). Oxyphilic adenoma was distinguished from carcinoma on the basis of vascular or capsular invasion or metastasis. The patients were 2 men and 20 women, with a mean age of 53 yr (range 27– 82 yr). The average size of the tumor was 37.8 ⫾ 20.5 mm (mean ⫾ sd; range 15–90 mm). In addition to neoplastic thyroid samples, normal thyroid samples were taken sufficiently distant from the tumors to serve as controls. All the samples were immediately stored in liquid nitrogen until extraction of high-molecular-weight DNA and RNA. The 22 tumor samples and controls were fixed in formalin, embedded in paraffin, and stained with hematoxylin and eosin. Immunohistochemistry was performed on paraffin-embedded sections. Tissue samples from the seven fresh tumors (six adenomas and one carcinoma) and matched controls were kept in a preservative medium [100 mm sucrose, 1 mm EGTA, 20 mm 3-[N-morpholino]propanesulfonic acid (pH 7.4), 1 g/liter bovine albumin] to prepare the mitochondria for polarographic studies and ATP measurement. Tumoral and control tissues were compared using a nonparametric test for matched-pair samples (Wilcoxon test), the differences being considered statistically significant at P less than or equal to 0.05. All the numerical values below are expressed as means ⫾ sd.

Immunohistochemistry After morphological examination of hematoxylin- and eosin-stained sections, corresponding 3-␮m sections of the paraffin blocks were prepared for the detection of mitochondrial antigen expression as a semiquantitative index of mitochondrial biogenesis. A monoclonal antibody 113–1 was used, recognizing an unknown 60-kDa nonglycosylated protein component of human cell mitochondria (BioGenex Laboratories, Inc., San Ramon, CA). Immunostaining was performed with the standard avidin-biotin peroxidase technique with antigen retrieval. For negative control slides, the primary antibody was either omitted or replaced by a suitable concentration of normal IgG of the same species.

Polarographic studies Oxygen consumption was measured with a Clark electrode at 30 C in a 2-mL chamber (Oxygraph OROBOROS, Anton Paar, Innsbru¨ ck, Austria). The chamber was isolated from contact with the atmosphere by a close-fitting cap so that the electrode current was proportional to the partial pressure of oxygen in the sample. Substrates for the different complexes of the respiratory chain were introduced into the chamber and consumed by mitochondria in the oxyphilic tumors or matched controls. Mitochondria were isolated from seven fresh tissue samples and matched controls using the standard procedure (13). Oxygen uptake was measured after the addition of mitochondria (200 ␮g protein) to 2 ml incubation medium [300 mm mannitol, 10 mm KH2PO4 (pH 7.2), 10 mm KCl, and 5 mm MgCl2] to determine the basal respiratory activity. Substrates and inhibitors were introduced into the oxygraph chamber through the stopper port. First, 2 ␮l EDTA (20 mm) and 2 ␮l rotenone (5 mm) were added to inhibit exogenous ATPase and complex I (nicotinamide adenine dinucleotide ubiquinone oxidoreductase of the respiratory chain), respectively. Then, 20 ␮l succinate (1 m) and 10 ␮l ADP (30 mm) were successively added to the sample to determine the rate of oxidative phosphorylation (ADP/O ratio). Finally, 10 ␮l potassium cyanide (200 mm) were added to stop oxygen uptake by inhibiting complex IV (cytochrome c oxidase of the respiratory chain).

ATP measurement Mitochondrial ATP was measured by bioluminescence using the luciferin-luciferase reaction (Enliten, Promega Corp., Madison, WI) (14). Mitochondria were isolated, using the same methods as for the polarographic studies, from the seven fresh oxyphilic tumor samples and their matched controls. After incubation of mitochondria with 10 mm glutamate and malate for 10 min, the rate of ATP synthesis (expressed per

J Clin Endocrinol Metab, October 2001, 86(10):4920 – 4925 4921

milligram of mitochondrial protein) was determined for intact and permeabilized mitochondria.

DNA isolation and Southern blot analysis DNA was isolated using the phenol-chloroform procedure. The samples were digested overnight with RNase A (20 ␮g/ml) and proteinase K (20 mg/ml) at 37 C in Tris-HCl 10 mm, EDTA (pH 8) 0.1 m, and SDS 0.5%. The proteins were removed by organic extraction followed by ethanol precipitation with NaCl 0.2 m and centrifugation for 15 min at 10,000 g. Five micrograms of DNA were digested with the restriction enzyme XbaI (Biolabs, Beverly, MA). Southern blotting was performed according to standard methods. Probes labeled by digoxigenin were obtained by multirandom priming and were revealed with antidigoxigenin antibodies labeled by alkaline phosphatase (DigDNA labeling and detection kit, Roche, Basel, Switzerland). The mitochondrial and nDNA were detected by using the probes 12S ribosomal DNA (rDNA) (nt 592-1344) and 18S rDNA (nt 1201–1811), respectively. For each sample, the intensities of the mtDNA signals, and the corresponding nDNA signals, were quantified by densitometric analysis (Molecular Analyst, Bio-Rad Laboratories, Inc., Cambridge, MA).

RNA isolation and cDNA synthesis RNA was isolated using the guanidinium isothiocyanate procedure (Trizol Reagent, Life Technologies, Inc., Gaithersburg, MD). Residual DNA was removed by DNase treatment: 5 ␮g total RNA were incubated with 2 U RNase-free DNase I for 1 h at 37 C. To generate cDNA, 1 ␮g of RNA was first denatured at 70 C with 1 ␮m of oligodT (Promega Corp.) for 5 min before quenching on ice; then 0.5 mm of each of the 4 dNTPs, 10 mm dithiothreitol, 10 U RNase inhibitor, and 200 U superscript II (Life Technologies, Inc.) were added to the 5⫻ buffer to make up a final volume of 20 ␮l reaction mix. The reaction mix was incubated for 1 h at 42 C. The reverse transcriptase was inactivated at 70 C for 15 min.

Quantitative PCR analysis Real-time quantitative PCR with an external standard was used to determine the gene copy number (Lightcycler, Roche). Standard PCR products for each gene were generated by amplifying nuclear cDNA or mtDNA templates. PCR products were purified by the phenol-chloroform method and the copy number in the final sample was determined by two independent methods (i.e. spectrophotometry and gel analysis). For each gene tested, a sequence-specific standard curve was plotted using serial dilutions of the target gene standard PCR product, and the same primers were used to amplify the cDNA. The expression of two mitochondrial genes, ND2 and ND5, and two nuclear genes, UCP2 and ␤-ACTIN, was analyzed using the PCR primer sets indicated in Table 1. The amount of RNA determined for each sample was normalized by the quantification of the ␤-ACTIN transcripts. Two microliters of master mix containing Taq DNA polymerase, dNTPs, and SYBR green I (DNA Master SYBR Green I kit, Roche) were incubated for 5 min at room temperature with 0.16 ␮l of Taqstart antibody. The PCR reaction was then started by adding MgCl2 4 mm and forward and reverse primers (0.5 ␮m) to the capillary tubes of the Lightcycler apparatus containing the master mix and 2 ␮L of template (cDNA or a standard with a known copy number) in a final volume of 20 ␮l.

TABLE 1. Oligonucleotide pairs used for quantitative PCR Primer pairs

5⬘-GCACCCCTCTGACATCC-3⬘ 5⬘-CGGTCGGCGAACATCAGTGG-3⬘ 5⬘-GGGGATTGTGCGGTGTGTG-3⬘ 5⬘-CTTCTCCTATTTATGGGGGT-3⬘ 5⬘-CCAGTGCGCGCGCTACAGTCA-3⬘ 5⬘-GTGGTGCTGCCTGCTAGGAG-3⬘ 5⬘-CGACATGGAGAAAATCTGGC-3⬘ 5⬘-AGGTCCAAGACGCAGGATGG-3⬘

Genes

ND2 ND5 UCP2

␤-ACTIN

4922

J Clin Endocrinol Metab, October 2001, 86(10):4920 – 4925


Results

mitochondria, attesting to the functionality of ATP/ADP translocase in exporting ATP from intact mitochondria.

Immunohistochemistry

The 60-kDa mitochondrial protein was present in the cytoplasm of all the 22 oxyphilic thyroid tumor samples, both homogeneous (19 tumors) and heterogeneous (3 tumors) distributed (histochemistry score data not shown), confirming the increased mitochondrial biogenesis in the tumor samples. In contrast, the immunostaining of the matched control samples was extremely weak or undetectable. The homogeneous distribution of mitochondrial immunostaining with no difference between adenomas vs. carcinomas has been previously described for oxyphilic tumors (15). Polarographic analysis

Seven tumors and matched controls (n ⫽ 14) were used for the functional analysis of the respiratory chain. The mitochondria were prepared within 1 h after thyroidectomy so as to obtain interpretable results. Polarographic analysis of complexes I to IV produced no evidence of respiratory chain defects in the seven oxyphilic tumors vs. control tissue. The respiratory control indices were calculated by dividing the rate of oxygen uptake in state 3 (stimulated by ADP addition) by the rate of oxygen uptake in state 4 (when all the ADP is converted to ATP). Using succinate as substrate, the indices were 3.8 ⫾ 0.8 for tumors and controls (n ⫽ 14). The ADP/O ratio in the tumor samples was 1.2 ⫾ 0.3 (n ⫽ 7), whereas it was 1.6 ⫾ 0.4 in the matched controls (n ⫽ 7), but this difference is not significant (Table 2). ATP synthesis

Because the analysis of ATP synthesis can be usefully performed exclusively on mitochondria from fresh tissues, only the mitochondria from the seven tumors and their matched controls (n ⫽ 14), as used for the polarographic studies, were used for ATP measurement. The mitochondrial ATP synthesis, adjusted to the mitochondrial protein level after addition of the substrate, was 5.8 ⫾ 1.4 ␮mol/mg per 10 min in oxyphilic tumors samples, compared with 12.1 ⫾ 3.1 ␮mol/mg per 10 min in matched controls. This represents a low rate of mitochondrial ATP synthesis in comparison with the basal rate for normal thyroid tissue we had previously measured (13.5 ⫾ 2.8) (unpublished data). The Wilcoxon test showed that the decrease observed in ATP synthesis was highly significant (P ⫽ 0.018, Table 2). There was no difference in ATP levels between intact and permeabilized

MtDNA quantification

Twenty-two tumors and matched controls were explored for mtDNA quantification. DNA was preserved from rRNA contamination by buffered RNase during extraction. The ratio of mtDNA to nDNA in oxyphilic tumor samples and matched controls was determined by Southern blot analysis. Three hybridization bands were detected by the mitochondrial 12S rRNA probe (7.5 and 1.7 kb) and the nuclear 18S rRNA probe (1.5 kb). The densitometric analysis of the 1.7and 1.5-kb bands showed that the 12S rDNA/18S rDNA ratio was 1.67 ⫾ 0.09 in oxyphilic thyroid tumors, compared with 0.54 ⫾ 0.19 in controls. This increase in mtDNA content for tumors was highly significant, using the Wilcoxon test (P ⬍ 0.001). Fig. 1 shows the 1.7- and 1.5-kb bands of several oxyphilic tumors samples and matched controls. Deletions in mtDNA were explored by long PCR analysis using a standard procedure (16). The common mtDNA4977 deletion was found in two of the tumors as well as in the matched controls, with an identical level of heteroplasmy. All the other samples were free from this deletion. Mitochondrial and nuclear gene expression

Twenty-two tumors and matched controls were explored for mtDNA quantification. The expression of ND2 and ND5 mitochondrial genes was 12 times higher in oxyphilic thyroid tumor samples than in controls. When adjusted to the mtDNA/nDNA ratio, the relative mitochondrial transcript ratio was 3.8 times higher in the tumor samples than in controls. For the nuclear gene, we observed a 2-fold increase of UCP2 in oxyphilic tumors samples, compared with controls. Fig. 2 shows the different patterns of mitochondrial and nuclear gene expression in the tumor samples and controls. Table 3 summarizes the histology and the gene expression pattern of the different samples. Table 2 sums up the statistical analysis of the results. Discussion

Several authors have suggested that defective energyproducing mechanisms of oxyphilic cells may be responsible for mitochondrial proliferation (11, 17). This hypothesis stems from the observation that the active metabolism of

TABLE 2. Statistical analysis of mitochondrial (ND2, ND5) and nuclear (UCP2) gene expression and ATP synthesis in oxyphilic thyroid tumors, compared with matched controls

ND2b ND5b UCP2b ADP/O ratio ATP (␮mol/mg protein per 10 min)

Control tissuea (n ⫽ 22)

Oxyphilic tumorsa (n ⫽ 22)

658 ⫾ 290 612 ⫾ 302 272 ⫾ 119 n⫽7 1.6 ⫾ 0.4 12.1 ⫾ 3.1

6799 ⫾ 3215 7354 ⫾ 3208 503 ⫾ 191 n⫽7 1.2 ⫾ 0.3 5.8 ⫾ 1.4

Mean ⫾ SD. Copy number/␤-ACTIN copy number. ADP/O ratio, Rate of oxydative phosphorylation; NS, not significant.

a b

Wilcoxon test

P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001 NS P ⫽ 0.018



FIG. 1. Quantitation of mtDNA and nDNA for several oxyphilic thyroid tumors and matched controls by Southern blot analysis. Five micrograms of total DNA digested with XbaI was hybridized first with an 18S rRNA nuclear probe and then rinsed and hybridized with a 12S rRNA mtDNA probe. The intensities of the 1.5-kb (18S) and 1.7-kb (12S) bands were quantified by a RadioImager (Cyclone, Packard, Downers Grove, IL). C, Control tissue; O, oxyphilic thyroid tumor.

FIG. 2. Quantification of mitochondrial and nuclear gene transcripts from 22 oxyphilic thyroid tumors and matched controls. 2a, Mean (⫾ SD) of the ratio of the mitochondrial ND2 and ND5 cDNA copy numbers vs. the ␤-ACTIN cDNA copy number. 2b, Mean (⫾ SD) of the ratio of the nuclear UCP2 cDNA copy number vs. the ␤-ACTIN cDNA copy number. The cDNA copies were determined by real-time quantitative PCR analysis (Lightcycler, Roche, Basel, Switzerland) after reverse transcription of the total RNA of each tumor and control sample. (Table 1 shows the PCR primers used.)

oncocytic cells, with their high levels of oxidative enzymes, does not correspond to high thyroid cell function (18). However, in these histochemical studies, the respiratory enzymes were functional, and a protein uncoupling the oxidative phosphorylation process (UCP1) was not present in oxyphilic tumors (11). MtDNA alterations are associated with several mitochondrial degenerative diseases (19). Large mtDNA deletions, such as the most common mtDNA4977 deletion, result in a decline of the oxidative phosphorylation capacity and accumulate progressively in aging normal tissues (20). An increased frequency of the mtDNA4977 deletion in oncocytic tumors has been described (18). We were unable to identify

any mtDNA deletions that might provide a replicative advantage over wild-type mtDNA. The “common” mtDNA4977 deletion was found in 2 of the 22 oxyphilic tumors investigated, but in each case the same deletion was also detected in the corresponding controls and corresponded to two elderly patients (63 and 75 yr). Thus, the mtDNA4977 deletion might be associated with cellular aging rather than to the development of an oncocytic phenotype, as previously suggested (21, 22). The increase in mtDNA content observed in 22 thyroid tumors (3.10 ⫾ 0.29) was 25% lower than that indicated by other authors (4.31 ⫾ 1.09) (21). The analysis of mitochondrial gene expression showed that tumors had a 12-fold increase in ND2 and ND5 transcripts, compared with control tissue. The expression of these two mitochondrial genes has already been associated with abnormal mitochondrial biogenesis in oncocytic tumors (12). The gene expression ratio was adjusted to the mtDNA/ nDNA ratio calculated by Southern blot analysis to obtain a more accurate estimate of the real increase of mitochondrial gene expression. The mtRNA/mtDNA ratio thus determined for the oxyphilic thyroid tumors may be compared with that given for oxyphilic tumors in other tissues. We found an mtRNA/mtDNA ratio of about 4:1 in oxyphilic thyroid tumors, compared with controls, whereas this ratio was 1:1 in the case of oxyphilic salivary gland tumors and 1:5 in the case of oxyphilic renal tumors (12). In the study of a cell line derived from a thyroid oncocytoma, we found an mtRNA/mtDNA ratio as high as 2:1, compared with a control thyroid cell line (23). These large differences in the mtRNA/mtDNA ratio suggest tissue-specific regulation of mitochondrial transcription and replication. It might therefore be relevant to investigate the nuclear factors involved in this regulation in various tissues. Polarographic analysis produced no evidence of respiratory chain defects in oxyphilic thyroid tumors, compared with control tissue. The respiratory chain ratios in mitochondria isolated from seven oxyphilic tumors were consistent with the indices published for mitochondria in the normal thyroid (24). However, the ADP/O ratio was only 75% of the normal value. The oxidative phosphorylation coupling defect revealed by polarography might be related to the 2-fold increase in UCP2 expression observed in oxyphilic tumors, compared with controls. After verifying that UCP1 was not expressed in oxyphilic thyroid tumors (data not shown), we investigated the expression of UCP2, the role of which has been established in the uncoupling process (25).

4924

J Clin Endocrinol Metab, October 2001, 86(10):4920 – 4925


TABLE 3. Case reports Case

Histology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

OTA OFA OFC OFA OFA OFC OFA OTA OFA OFA OTA OTA OTA OTA OFA OFA OFC OFA OFA OFA OFA OFA

ND2#

ND5#

UCP2#

C

O

C

O

C

O

752 267 402 1,140 760 804 589 549 705 941 875 1,236 865 1,130 456 442 583 345 321 238 643 436

7,890 5,749 2,592 14,587 8,403 8,702 3,015 4,057 9,587 5,783 2,378 12,700 5,640 6,389 8,653 7,833 6,421 10,054 3,890 3,420 7,804 4,042

812 320 130 1,250 548 647 950 450 521 203 875 694 549 253 159 377 590 1,050 858 665 947 635

6,850 7,760 1,390 11,760 11,540 12,501 5,470 5,690 12,589 2,547 7,646 11,345 5,578 3,543 3,211 6,489 7,345 8,897 6,789 5,780 8,085 8,988

355 123 420 201 341 201 258 507 204 438 321 365 532 130 147 230 184 163 208 201 220 251

469 198 972 495 507 357 398 980 480 508 542 603 685 603 365 521 475 295 265 385 430 541

#, Copy number/␤-ACTIN copy number; OFA, oxyphilic follicular adenoma; OFC, oxyphilic follicular carcinoma; OTA, oxyphilic follicular adenoma (trabecular pattern); C, control surrounding tissue; O, oxyphilic thyroid tumor.

In our study of seven fresh oxyphilic tumors, the increased expression of UCP2 was probably responsible for the coupling defect reflected by a significant decrease in ATP synthesis. Mitochondrial proliferation could therefore be an adaptive response to a primary nuclear abnormality, namely the overexpression of UCP2. However, the overexpression of UCP2 might itself be a response to the proliferation of mitochondria compensating for the decreased mitochondrial ATP synthesis. In the latest case, the proliferation of mitochondria leads to overproduction of reactive oxygen species, which could be counteracted by an increase in UCP2 expression (26). Another line of reasoning suggests that the metabolism in oxyphilic tumors has switched to a glycolytic status (12). The decrease in mitochondrial ATP synthesis we noticed could lead to a shift toward anaerobic metabolism. Because the defect is measured in oxyphilic adenoma as well as in carcinoma, we suggest that the metabolism switch is an early event in the oxyphilic thyroid tumor progression. Thus, the oxyphilic cell might be early resistant to hypoxia, which could explain the aggressive clinical behavior of these tumors (22). In conclusion, the defective ATP synthesis we observed in seven oxyphilic thyroid tumors might explain the mitochondrial proliferation found in the tumor cells. Because the expression of UCP2 was higher in all the 22 oxyphilic thyroid tumors, compared with controls, we suggest that the oxidative phosphorylation coupling defect we detected may be associated with mitochondrial proliferation in oxyphilic thyroid tumors. It would therefore be of interest to further investigate the factors involved in the transcription and replication of mitochondrial DNA in oxyphilic tumors of the thyroid gland.

Acknowledgments We are grateful to C. Savagner for the statistical analysis and to K. Malkani for critical reading of the manuscript. We thank Anne, Dominique, and Florence for continuous support during the study. Received January 26, 2001. Accepted June 5, 2001. Address all correspondence and requests for reprints to: F. Savagner, Inserm EMI-U 00-18, Laboratoire de Biochimie et Biologie Mole´ culaire, Chu, 4 rue Larrey, F-49033 Angers cedex 01, France. E-mail: [email protected]. This work was supported by grants from l’Association pour la Recherche sur le Cancer.

References 1. Hedinger C, Williams ED, Sobin LH 1989 The WHO histological classification of thyroid tumors: a commentary on the ed. 2. Cancer 63:908 –911 2. Katoh R, Harach HR, Williams ED 1998 Solitary, multiple, and familial oxyphil tumours of the thyroid gland. J Pathol 186:292–299 3. Ricquier D, Bouillaud F 2000 The uncoupling protein homologues: UCP1, UCP2, UCP3, StUCP and AtUCP. Biochem J 345:161–179 4. Tallini G 1998 Oncocytic tumours. Virchows Arch 433:5–12 5. Bronner MP, Clevenger CV, Edmonds PR, Lowell DM, McFarland MM, LiVolsi VA 1988 Flow cytometric analysis of DNA content in Hu¨ rthle cell adenomas and carcinomas of the thyroid. Am J Clin Pathol 89:764 –769 6. Bouillaud F, Ricquier D, Thibault J, Weissenbach J 1985 Molecular approach to thermogenesis in brown adipose tissue: cDNA cloning of the mitochondrial uncoupling protein. Proc Natl Acad Sci USA 82:445– 448 7. DiMauro S, Bonilla E, Zeviani M, Nakagawa M, DeVivo DC 1985 Mitochondrial myopathies. Ann Neurol 17:521–538 8. Shoffner JM, Wallace DC 1990 Oxidative phosphorylation diseases. Disorders of two genomes. Adv Hum Genet 19:267–330 9. Heddi A, Lestienne P, Wallace DC, Stepien G 1993 Mitochondrial DNA expression in mitochondrial myopathies and coordinated expression of nuclear genes involved in ATP production. J Biol Chem 268:12156 –12163 10. Flier J, Lowell B 1997 Obesity research springs a proton leak. Nat Genet 15:223–224 11. Ebner D, Rodel G, Pavenstaedt I, Haferkamp O 1991 Functional and molecular analysis of mitochondria in thyroid oxyphil tumour. Virchows Arch B Cell Pathol Incl Mol Pathol 60:139 –144 12. Heddi A, Faure-Vigny H, Wallace DC, Stepien G 1996 Coordinate expression of nuclear and mitochondrial genes involved in energy production in carcinoma and oxyphil tumour. Biochim Biophys Acta 1316:203–209 13. Chretien D, Bourgeron T, Rotig A, Munnich A, Rustin P 1990 The measure-


14.

15. 16.

17. 18. 19.

ment of the rotenone-sensitive NADH cytochrome c reductase activity in mitochondria isolated from minute amount of human skeletal muscle. Biochem Biophys Res Commun 173:26 –33 Lemasters JJ, Hackenbrock CR 1976 Continuous measurement and rapid kinetics of ATP synthesis in rat liver mitochondria, mitoplasts and inner membrane vesicles determined by firefly-luciferase luminescence. Eur J Biochem 67:1–10 Tallini G, Carcangiu ML, Rosai J 1992 Oncocytic neoplasms of the thyroid gland. Acta Pathol Jpn 42:305–315 Reynier P, Chretien MF, Savagner F, et al. 1998 Long PCR analysis of human gamete mtDNA suggests defective mitochondrial maintenance in spermatozoa and supports the bottleneck theory for oocytes. Biochem Biophys Res Commun 254:373–377 Heddi A, Stepien G, Benke PJ, Wallace DC 1999 Coordinate induction of energy gene expression in tissues of mitochondrial disease patients. J Biol Chem 274:22968 –22976 Welter C, Kovacs G, Seitz G, Blin N 1989 Alteration of mitochondrial DNA in human oxyphil tumours. Genes Chromosomes Cancer 1:79 – 82 Wallace DC 1992 Diseases of the mitochondrial DNA. Annu Rev Biochem 61:1175–1212


20. Cortopassi G, Shibata D, Soong N, Arnheim N 1992 A pattern of accumulation of a somatic deletion of mitochondrial DNA in aging human tissues. Proc Natl Acad Sci USA 89:7370 –7374 21. Tallini G, Ladanyi M, Rosai J, Jhanwar SC 1994 Analysis of nuclear and mitochondrial DNA alterations in thyroid and renal oncocytic tumors. Cytogenet Cell Genet 66:253–259 22. Maximo V, Sobrinho-Simoes M 2000 Hu¨ rthle cell tumours of the thyroid. A review with emphasis on mitochondrial abnormalities with clinical relevance. Virchows Arch 437:107–115 23. Savagner F, Chevrollier A, Loiseau D, et al. 2001 Mitochondrial activity in XTC.UC1 cells derived from thyroid oncocytoma. Thyroid 11:327–333 24. Lamy FM, Rodesch FR, Dumont JE 1967 Action of thyrotropin on thyroid energetic metabolism. VI. Regulation of mitochondrial respiration. Exp Cell Res 46:518 –532 25. Ricquier D, Bouillaud F 2000 Mitochondrial uncoupling proteins: from mitochondria to the regulation of energy balance. J Physiol 529:3–10 26. Arsenijevic D, Onuma H, Pecqueur C, et al. 2000 Disruption of the uncoupling protein-2 gene in mice reveals a role in immunity and oxygen species production. Nat Genet 26:435– 439