Hypothyroid Phenotype Is Contributed by Mitochondrial Complex I Inactivation Due to Translocated Neuronal Nitric-oxide Synthase*

Although transcriptional effects of thyroid hormones have substantial influence on oxidative metabolism, how thyroid sets basal metabolic rate remains obscure. Compartmental localization of nitric-oxide synthases is important for nitric oxide signaling. We therefore examined liver neuronal nitric-oxide synthase-α (nNOS) subcellular distribution as a putative mechanism for thyroid effects on rat metabolic rate. At low 3,3′,5-triiodo-l-thyronine levels, nNOS mRNA increased by 3-fold, protein expression by one-fold, and nNOS was selectively translocated to mitochondria without changes in other isoforms. In contrast, under thyroid hormone administration, mRNA level did not change and nNOS remained predominantly localized in cytosol. In hypothyroidism, nNOS translocation resulted in enhanced mitochondrial nitric-oxide synthase activity with low O2 uptake. In this context, NO utilization increased active O2 species and peroxynitrite yields and tyrosine nitration of complex I proteins that reduced complex activity. Hypothyroidism was also associated to high phospho-p38 mitogen-activated protein kinase and decreased phospho-extracellular signal-regulated kinase 1/2 and cyclin D1 levels. Similarly to thyroid hormones, but without changing thyroid status, nitric-oxide synthase inhibitor Nω-nitro-l-arginine methyl ester increased basal metabolic rate, prevented mitochondrial nitration and complex I derangement, and turned mitogen-activated protein kinase signaling and cyclin D1 expression back to control pattern. We surmise that nNOS spatial confinement in mitochondria is a significant downstream effector of thyroid hormone and hypothyroid phenotype.

Hypothyroidism is a prevalent disorder associated to low oxygen utilization and low tissue proliferation rate (1). In addition to non-genomic effects (2), thyroid hormones influence transcription of a number of nuclear and mitochondrial-encoded respiratory genes (3). Although direct or transcriptional effects have considerable impact on oxidative metabolism and hemodynamic function, much is still unknown about how thyroid hormones set the metabolic rate of the body (4); consonant with slowness of transcriptional mechanisms, treatment of hypothyroidism may require weeks of hormone administration to normalize the altered functions (5). In the last decade, the effects of nitric oxide (NO) 2 expanded from the vascular system to the intracellular milieu. In this context, subcellular localization of nitric oxide-synthases (NOS) with effector molecules is an important regulatory mechanism for NO signaling (6,7). Accordingly, we are interested in the traffic of a posttranslationally modified variant of neuronal nitric-oxide synthase-␣ (nNOS) to mitochondria (formerly named mitochondrial nitric-oxide synthase or mtNOS), which vectorially directs NO to the matrix compartment (8,9). mtNOS could be low in adult rodents, but a modulated increase has been associated to thyroid status (10), release of cytochrome c (11), mitochondrial protein nitration (12), liver and brain development (13,14), adaptation to cold environment (15), and hypoxia (16).
Under physiological O 2 levels and in vivo respiratory dynamics, NO inhibits mitochondrial state 3 and 4 O 2 uptake (17). At 10 -20 nM, matrix NO tonically modulates O 2 uptake by reversible inhibition of cytochrome oxidase through heme iron nitrosylation at a-a 3 subunit (18) and promotes the formation of superoxide (O 2 . ) and hydrogen peroxide (H 2 O 2 ) (18,19). At complexes I and II-III, mitochondrial production of O 2 species depends on autooxidation of intermediary ubisemiquinone during electron transfer from reduced ubiquinol pools to cytochrome c 1 , with formation of O 2 . . This reaction is considerably amplified by NO inhibition of cytochrome oxidase or cytochrome b-c 1 segment that augments the reduction on the side of substrates (18,19). In mitochondria, O 2 . is either dismutated to H 2 O 2 by manganese superoxide dismutase (Mn-SOD) or reacts with NO to form peroxynitrite (ONOO Ϫ ). Considering that, variations in matrix NO differently change the proportion of these active species (19); the aim of this work was to search for nNOS traffic to mitochondria and its consequences in O 2 uptake, mitochondrial oxidant yield, and cell responses (13) as experimentally related to thyroid status. Animals and Treatments-Male Wistar rats (250 -300 g) were housed in a temperature-controlled room with food and water ad libitum; National Institutes of Health criteria for animal research were followed after approval by the Ethics Committee of the University Hospital. Rats were divided into five groups (n ϭ 12/group): hypothyroid, (0.02% methimazole (w/v) in drinking water for 28 days); hypothyroidϩT 3 , (after 25 days of methimazole treatment received 15 g of T 3 /kg of body weight for 3 days by intraperitoneal injection); hypothyroidϩL-NAME, (0.75 mg/ml L-NAME in drinking water in the last 21 days of methimazole treatment); hyperthyroid, (intraperitoneal injection of 60 g of T 3 /kg of body weight for 3 days); and control group. Blood samples were collected at the time the animals were killed for estimation of thyrotropin (TSH) level by radioimmunoassay (20).
Basal Metabolic Rate-BMR was measured at 22°C in a non-recirculating open flow system after 30 min of equilibration with an O 2 -CO 2 analyzer in standard temperature and pressure dry conditions (15).
Isolation and Purification of Liver Mitochondria-The livers were excised in ice-cold homogenization medium, and mitochondria were isolated and purified as described (13). Minimal contamination was found (2-3%) by comparing activities of lactate dehydrogenase (cytosolic marker) and succinate-cytochrome c reductase (mitochondrial marker).
Quantitative Real-time PCR-Real-time nested PCR for nNOS isoform was done with 0.5 l of a 1/10 dilution of PCR product in 25-l final volume with inner primers: nNOS, sense (5Ј-TTCAACTA-CATCTGTAACCA-3Ј) and antisense (5Ј-TGAACTGCACATTG-GCTGGA-3Ј). Real-time PCR reactions included 0.4 mM dNTPs, 1 M specific primers, 4 mM MgCl 2 , 2.5 units of Taq DNA polymerase, and 1:30,000 SYBR Green. Real-time PCR reactions were performed in DNA Engine Opticon (MJ Research, Inc.) and consisted of an initial denaturing step (94°C for 4 min), followed by 35 cycles (each of 94°C for 1 min, 55°C for 40 s, 72°C for 1 min). Sample quantification was normalized to endogenous ␤2-microglobulin that was also quantified by real-time PCR following the same protocol as nNOS isoform. Each experiment included a DNA minus control and a standard curve.
Immunoblotting for NOS, Hsp90, and Mitochondrial Protein Nitration-Proteins were electrophoresed on 7.5% SDS-polyacrylamide gel, electrotransferred to polyvinylidene difluoride membranes (13), incubated with anti-nNOS, anti-eNOS, anti-iNOS, anti-Hsp90, and anti-nitrotyrosine antibodies, and detected with the ECL system. Equal loading was controlled with the appropriated subcellular markers. Incu-bation of the anti-nitrotyrosine antibody with 10 mM nitrotyrosine prior to the membrane incubation was used to ensure the antibody specificity.
NOS Activity in Subcellular Fractions-NOS activity was determined in mitochondrial and cytosolic fractions by conversion of [ 3 H]L-arginine to [ 3 H]L-citrulline (13).
Immunoelectron Microscopy-Purified mitochondria were suspended in 4% paraformaldehyde and 0.5% glutaraldehyde, pH 7.4, for 2 h at 4°C, washed overnight with 0.32 M sucrose at 4°C, and then dehydrated in 70% ethanol and embedded in LR White (13). Ultrathin sections were obtained in 300-mesh nickel grids. Immunocytochemistry was performed using a primary mouse anti-C-terminal nNOS (1095-1289) at a dilution of 1:20 in phosphate-buffered saline, pH 7.4. Grids were washed in phosphate-buffered saline and counterstained with 1% uranyl acetate. Nonspecific background was blocked by incubation with 5% normal goat serum in phosphate-buffered saline at the beginning of the procedure. Positive control against 39-kDa subunit of complex I (inner membrane marker) and negative control in the absence of a primary antibody were included. Specimens were observed in a Zeiss EM-109-T transmission electron microscope at 80 kV.
Detection of Mitochondrial NO-Mitochondria (1 mg of protein per ml) were incubated in phosphate-buffered saline for 30 min at 37°C with 5% CO 2 , 10 M DAF-FM, and 0.5 M MitoTracker, and fluorescence was measured with an Ortho-Cytoron Absolute Cytometer (Johnson and Johnson) (24).
Mitochondrial O 2 Utilization and Electron Transfer Activity-O 2 uptake was measured polarographically with a Clark-type electrode (10). To assess NO effects, mitochondria were incubated with 0.3 mM L-arginine (L-Arg) alone or plus 3 mM L-NMMA for 5 min at 37°C (10). State 4 O 2 uptake was determined with 6 mM malate-glutamate as substrate of complex I and state 3 active respiration by the addition of 0.2 mM adenosine diphosphate (ADP). Complex I activity (NADH: ubiquinone reductase) was measured by the rotenone-sensitive reduction of 50 M 2,3-dimethoxy-6-methyl-1,4-benzoquinone with 1 mM KCN and 200 M NADH as electron donor at 340 nm with a Hitachi U3000 spectrophotometer at 30°C. Activity of complexes II-III was determined by cytochrome c reduction at 550 nm. Cytochrome oxidase activity (Complex IV) was determined by monitoring cytochrome c oxidation at 550 nm (⑀ 550 , 21 mM Ϫ1 ⅐cm Ϫ1 ); the reaction rate was measured as the pseudo-first order reaction constant (kЈ) and expressed as kЈ/min⅐mg of protein (13,14).
Mitochondrial Production of H 2 O 2 and O 2 . -H 2 O 2 production rate was monitored spectrofluorometrically at complexes I or II-III (6 mM malate-glutamate or 10 mM succinate as substrate) in an F-2000 spectrofluorometer (Hitachi, Tokyo, Japan) as described (13). To determine O 2 . at complexes I and II-III, mitochondria were subjected to three freeze/thaw cycles, and SOD-sensitive cytochrome c reduction was measured at 550 nm (0.1 mg of protein/ml and 10 M SOD to subtract unspecific reduction). Mn-SOD, catalase, and glutathione peroxidase activities were determined in 7,000 g supernatants as described (13). Liver Cell Isolation and Detection of Intracellular Oxidants-Hepatocytes were isolated by two-step collagenase perfusion (25). Intracellular oxidants and mitochondrial (O 2 . ) were detected by flow cytometry after incubating hepatocytes in phenol red-free Dulbecco's modified Eagle's medium with 5 M DCFH-DA or 5 M HE for 30 min at 37°C with 5% CO 2 .
Blue Native Polyacrylamide Gel Electrophoresis-To separate mitochondrial complexes, Blue Native-PAGE was performed according to Schägger (26). Gels of first dimension were stained with Coomassie Blue and membranes incubated with antibodies against 3-nitrotyrosine. For second-dimension analysis, gel bands, corresponding to the complex I mtNOS, Complex I Inactivation, and Hypothyroidism region derived from 5-mm-wide lanes, were excised and incubated for 2 h in cathode buffer (50 mM glycine and 7.5 mM imidazole, pH 7) supplemented with 1% SDS and 1% ␤-mercaptoethanol before electrophoresis on 10%-16.5% Tris/glycine gels (27). Membranes were incubated with antibodies against 3-nitrotyrosine.
Immunoprecipitation-For immunoprecipitations, 500 g of mitochondrial proteins were incubated with 4 g of antibodies against Complex I 39-kDa subunit or Complex IV VIc subunit and 30 l of Protein A/G PLUS-agarose (Santa Cruz) at 4°C; samples were blotted against polyclonal nNOS antibody. Protein loading was controlled by the respective mitochondrial complex antibodies.
Preparation of Whole Liver Homogenates and Immunoblotting for MAPKs and Cyclin D1-To study MAPKs and cyclin D1, liver was homogenized in lysis buffer as described (13). Proteins were separated on 12% SDS-PAGE, and cyclin D1 and MAPKs were detected with specific antibodies.
Matrix NO ([NO] ss ) was calculated by the percentage of liver mitochondria state 3 respiratory rate NO-dependent inhibition as previously described (15,19). [ONOO Ϫ ] production rate was calculated as shown in Equation 4.
Statistical Analysis-Data are mean Ϯ S.E. One-way analysis of variance was utilized with post hoc Dunnett test, and regression analysis and significance were accepted at p Ͻ0.05.

At low T 3 , Transcriptional Increase of nNOS Enhances mtNOS
Content-Thyroid status was discriminated by thyrotropin and BMR measurements (Table 1). In this context, nNOS mRNA, protein expression, and activity were clearly modulated by T 3 levels (Fig. 1, A-E). Likewise, nNOS mRNA expression quantified by RT real-time PCR was Ͼ3-fold increased in hypothyroid liver compared with controls (Fig.  1B), whereas eNOS or iNOS mRNA did not change (data not shown). In addition, nNOS became distinctively enhanced in mitochondria (mtNOS), indicating a subsequent import of the overexpressed enzyme to the organelles; these findings were reverted by hormone replacement (Fig. 1C). To corroborate subcellular distribution in the groups, liver fractions were compared. Western blotting confirmed the increase of total nNOS expression represented by liver homogenates and the enrichment of the mitochondrial fraction in the hypothyroid group. Differentially, at high T 3 , nNOS expression had results similar to controls, but this condition retained the protein predominantly localized in cytosol (Fig. 1C). This effect was parallel to the increased expression of heat shock protein 90 (Hsp90), one of the most important chaperones associated to nNOS (Fig. 1F). Alternatively, modulation of mtNOS content in the studied groups was validated as well by immunoelectron microscopy with monoclonal nNOS antibodies (Fig. 1D). In agreement, Ca 2ϩ -dependent NOS activity was significantly increased in hypothyroid mitochondria but was normal to slightly reduced after T 3 admin-

TABLE 1
Oxygen uptake rates depend on thyroid status and nitric oxide utilization Data are mean Ϯ S.E. of the different groups (n ϭ 5-6 animals/group). O 2 uptake rate and complex activities were measured in intact mitochondria and mitochondrial membranes, respectively. *, different from controls; †, different from L-NMMA, by analysis of variance and Dunnett post hoc test. To avoid effects of body weight on resting metabolic rate, O 2 uptake was corrected to lean body mass and exponentially related to mass 0.75 . ngat, nanogram atom. istration, with an opposite cytosolic pattern (Fig. 1E); Ca 2ϩ -independent NOS activity was not detected. Subcellular nNOS Localization Modulates Mitochondrial Respiration-According to BMR, organelles from hypo-and hyperthyroid groups had the lowest and highest O 2 uptake rates, respectively (Table  1). To discern the effects of mtNOS, mitochondria were supplemented with L-Arg alone or plus NOS inhibitor L-NMMA. The sum of the opposite effects of NOS substrate and inhibitor on basal O 2 utilization determines mtNOS functional activity on respiration (28). Likewise, mtNOS-dependent inhibition of state 3 O 2 uptake was increased ϳ39% in hypothyroid samples, 18% in controls, and negligible in T 3 -treated mitochondria. In agreement, matrix NO estimated from L-Arg inhibition of O 2 uptake (18) was augmented by 4-fold in hypothyroidism and decreased by a half in hyperthyroidism (Table 1). These results demonstrate that (a) translocated nNOS is functionally active and mitochondria retain ex vivo the cofactors for catalytic activity and (b) NOS confinement to the small mitochondrial compartment amplifies NO effects on O 2 uptake and BMR.

Control
We next examined the contribution of segmental activities to mitochondrial O 2 uptake. Electron transfer rate at complex I was markedly decreased at hypothyroid status solely (ϳ 60%), whereas cytochrome oxidase was less inhibited (ϳ 11%), and complex II-III activity was not modified (Table 1). No significant thyroid effects on antioxidant Mn-SOD, catalase, or glutathione peroxidase activities were detected.

Thyroid Status Defines Quality and Intensity of Mitochondrial
Oxidant Production-In connection with O 2 uptake rate, previous observations proposed a decreased mitochondrial H 2 O 2 yield in rat hypothyroidism and an increased yield in hyperthyroidism (29), though opposite results were reported as well (30,31). It is shown here that basal production of H 2 O 2 with substrates of complex I (malate-glutamate) or II (succinate) is not essentially modified by thyroid status (Fig. 2A). In contrast, L-Arg enhances the production of mitochondrial active oxygen species from hypothyroid rats, particularly when incubated with malate-glutamate, and L-NMMA prevents its effect. Furthermore, in the presence of L-NMMA, mitochondrial generation of H 2 O 2 in hypothyroidism is below that seen in control and hyperthyroid groups, probably because of its reduced number of functional respiratory chain units. Therefore, mtNOS functional activity on H 2 O 2 production (calculated by the ratio of NO-dependent to maximal H 2 O 2 production in the presence of rotenone (13)), was 33% in hypothyroid mitochondria and 9% in controls. Repercussion of this mitochondrial activity on total cell oxidants in vivo is shown in Fig. 2B. Hypothyroid-isolated hepatocytes exhibited 30% more HE fluorescence and 280% more DCFH fluorescence than T 3 -treated cells, and administration of L-NAME to hypothyroid animals turned oxidant levels down (Fig. 2, B and C).
Nitrative Stress in Hypothyroidism-In relation to expected NO, O 2 . , and ONOO Ϫ yields, mitochondria from hypothyroid animals had higher 3-nitrotyrosine content than organelles from the other groups (Fig. 3A). In particular, complex I was markedly nitrated on tyrosine, whereas complex IV was nitrated to a lesser extent (Fig. 3B). Twodimensional SDS-PAGE of complex I confirmed an increased nitration of different nuclear and mitochondrial-encoded subunits in hypothyroidism (MW suggests nitration in the region of 75, 50 (ND5), 39 (ND4), 30 (ND1), 22 (PDSW), and 17 kDa (B17;18). Control samples showed a constitutive nitration in some of these components, and hyperthyroid samples presented the lowest nitration. In addition, immunoprecipitation of complex I and IV proteins indicated a direct interaction with translocated nNOS, either in control or hypothyroid group (Fig. 3C).

NOS Inhibitor L-NAME Prevents Low T 3 -dependent Phenotypic
Changes-To test whether NOS inhibition prevents T 3 effects on mitochondria, L-NAME was administered to hypothyroid rats. L-NAME neither modified nNOS mRNA (not shown) nor affected nNOS expression or distribution (Fig. 4A). Mitochondrial DAF fluorescence (representa- . production rate was determined by following SOD-sensitive reduction of cytochrome c. *, p Ͻ0.05 with respect to control; †, idem to basal; ‡, idem to L-Arg. FEBRUARY 24, 2006 • VOLUME 281 • NUMBER 8 tive of matrix NO) rose by 2-fold at hypothyroidism but decreased to only 25% of this value after L-NAME treatment (Fig. 4B). Therefore, at similar thyrotropin levels (Ͼ80 ng/ml) and mtNOS content, L-NAME prevented NO increase and nitration of mitochondrial proteins (Fig. 4C) and decrease of complex I activity. Accordingly, L-NAME increased BMR of hypothyroid rats up to control values (Fig. 4D), whereas no detectable effects were seen in control animals. These results are consistent with L-NAME prevention of the increase of mitochondrial O 2 . and oxidants in vivo as shown in isolated hepatocytes (Fig. 2B).

mtNOS, Complex I Inactivation, and Hypothyroidism
Mitochondrial NOS Activity Contributes to Thyroid-dependent Cell Signaling-Thyroid status governs liver cell proliferation (33). This effect depends on activation of proliferating cascades and D-cyclins that stimulate the progression of cell cycle from G 0 to G 1 . We have previously described the modulation of mtNOS activity and the putative regulation of cell cycle through redox signaling in the sequence of proliferating to quiescent cell stages during rat liver development (13).
In the same way, at the hypothyroid condition, liver P-p38MAPK was markedly expressed, whereas P-ERK1/2 and cyclin D1 mRNA and protein expression were very low (Fig. 5, A and B); the opposite pattern was detected after administration of T 3 . In this context and regarding mtNOS-dependent redox changes, expression of cyclin D1 correlated with P-ERK1/2/P-p38 ratio that fitted well with experimentally measured mitochondrial d[ (Fig. 5C)

DISCUSSION
This study shows, for the first time, that thyroid hormones modulate mRNA expression and subcellular distribution of nNOS␣. As reported for eNOS subcellular traffic (34,35), the translocation of nNOS to mitochondria could be influenced by regulation of posttranslational changes found in mitochondrial nNOS, (N-acylation, Ser-1412 phosphorylation) or by the turnover of cytosol "anchoring" proteins, like Hsp90 (Fig.  1F), dystrophin (36), or caveolin-1 (37). . At high mtNOS, hypothyroidism promotes nitration of mitochondrial proteins. A, proteins from the different groups were separated by SDS-PAGE, and membranes were revealed with anti-3-nitrotyrosine antibody; nitrated bovine serum albumin was utilized as positive control. B, mitochondrial complexes were separated in Blue Native-PAGE stained with Coomassie Blue (left) or transferred to polyvinylidene difluoride membranes and blotted with a polyclonal anti-3-nitrotyrosine antibody (middle). Two-dimensional SDS-PAGE of complex I band was blotted against anti-3-nitrotyrosine antibody (right) (see "Experimental Procedures"). C, mitochondrial proteins from the hypothyroid group were immunoprecipitated with anti-39-kDa subunit (complex I) and anti-subunit VIc (complex IV) antibodies, and immunoprecipitates were subjected to Western blotting with nNOS antibodies. Input, purified mitochondria; Sn, supernatant; IP, immunoprecipitate; Cϩ, rat brain cytosol. . NOS inhibitor L-NAME prevents hypothyroid mitochondrial phenotype. A, similar mtNOS content of liver mitochondria from controls and hypothyroid rats treated or not with L-NAME is revealed with anti-nNOS antibody. B, representative flow cytometry histogram of 1 mg/ml of isolated mitochondria incubated with 10 M DAF-FM and 0.5 M MitoTracker (per duplicate) at 495 nm (excitation) and 515 nm (emission). C, a decrease of mitochondrial protein nitration by L-NAME is detected in the hypothyroid condition with anti-3-nitrotyrosine antibody. D, inhibition of NOS activity by L-NAME prevented in parallel the hypothyroid lowering of basal metabolic rate (BMR) and mitochondrial complex I activity (n ϭ 4). *, p Ͻ0.05 compared with control by analysis of variance. †, idem compared with hypothyroid without L-NAME.

mtNOS, Complex I Inactivation, and Hypothyroidism
The notion that spatial confinement is essential to NOS signaling (6,38) is supported here by differential modulation of O 2 uptake and nNOS distribution, depending on thyroid status; this notion could be extended to other cell adaptive responses (15,39). It is then surmised that mitochondrial import of nNOS contributes to the typical decrease of BMR in hypothyroidism; though liver represents 10 -12% of BMR, a similar response was found in skeletal muscle that provides 20 -50% of BMR (10). Otherwise, down-regulation of mtNOS enhances T 3 genomic and non-genomic effects on mitochondrial and systemic O 2 uptake (2, 3).
In the absence of T 3 , the interaction between nNOS and complexes I and IV provides steric advantage to further increase O 2 . and H 2 O 2 production rates. As shown here (Fig. 3B), and depending on NO and O 2 .
fluxes, ONOO Ϫ formation leads to selective nitration or nitrosylation of complex I (40). Although constitutive nitration is detected in controls, extensive nitration of different subunits in hypothyroidism is parallel to the markedly reduced electron transfer between NADH and ubiquinol acceptor. A similar causality could be applied to reduced complex IV activity (41). In accord, Persichini et al. (42) recently reported that the direct interaction shown here between cytochrome oxidase and mtNOS proceeds through PDZ domains. Likewise, complex inhibition contrib-utes to set down BMR (43) and to further increase production of O 2 species (Fig. 2D) and ONOO Ϫ and to perpetuate nitration itself. In addition, it was previously reported that nitrated proteins are prone to be degraded by the ubiquitin-proteasome system (44,45). Thus, restoring basal nitration level after T 3 replacement (Fig. 3, A and B) required de novo protein synthesis during mitochondrial biogenesis or a "denitration" process (46). In this context, NO itself may increase mitochondrial biogenesis by cGMP-dependent activation of peroxisome proliferator-activated receptor-␥ coactivator 1␣ that stimulates the transcription of nuclear-encoded mitochondrial proteins (47). It is worth noting that the administration of L-NAME completely prevents the phenotypic changes induced by hypothyroidism (low BMR, decreased complex I activity, increased protein nitration) without changing hormonal status. Considering that L-NAME did not modify nNOS expression or subcellular distribution, its effects likely rely on the inhibition of NOS activity within mitochondria.
In recent years, some reports indicated mitochondrial redox contribution to the activation of MAPK cascades (48). In agreement, higher levels of phosphorylated p38MAPK could be consistent with kinase activation by high oxidant levels or by ONOO Ϫ itself produced in hypothyroidism (13,49). Currently, P-p38 participates in cell cycle arrest and inhibition of cell proliferation, a hallmark of hypothyroidism (1), whereas low oxidative stress and low ONOO Ϫ in hyperthyroidism are associated here with liver ERK1/2 activation and cyclin D1 expression (13), a hallmark of tissue proliferation. A similar effect of L-NAME or T 3 in turning hypothyroid cell signaling back to control status indicates that differential P-ERK1/2/P-p38MAPK ratio and expression of cyclin D1 should depend not on thyroid hormones themselves but on the relative production of mitochondrial oxidants, NO and ONOO Ϫ , at the different T 3 levels.
T 3 -dependent targeting of nNOS to mitochondria provides a new insight into the pathophysiology of hypo-and hyperthyroid syndromes. The presented mechanisms may gain importance in other situations associated with complex I dysfunction, like Parkinson disease.