Ubiquitin Proteasome-dependent Degradation of the Transcriptional Coactivator PGC-1α via the N-terminal Pathway*

PGC-1α is a potent, inducible transcriptional coactivator that exerts control on mitochondrial biogenesis and multiple cellular energy metabolic pathways. PGC-1α levels are controlled in a highly dynamic manner reflecting regulation at both transcriptional and post-transcriptional levels. Here, we demonstrate that PGC-1α is rapidly degraded in the nucleus (t½ 0.3 h) via the ubiquitin proteasome system. An N-terminal deletion mutant of 182 residues, PGC182, as well as a lysine-less mutant form, are nuclear and rapidly degraded (t½ 0.5 h), consistent with degradation via the N terminus-dependent ubiquitin subpathway. Both PGC-1α and PGC182 degradation rates are increased in cells under low serum conditions. However, a naturally occurring N-terminal splice variant of 270 residues, NT-PGC-1α is cytoplasmic and stable (t½ >7 h), providing additional evidence that PGC-1α is degraded in the nucleus. These results strongly suggest that the nuclear N terminus-dependent ubiquitin proteasome pathway governs PGC-1α cellular degradation. In contrast, the cellular localization of NT-PCG-1α results in a longer-half-life and possible distinct temporal and potentially biological actions.

Members of the peroxisome proliferator-activated receptor ␥ (PPAR-␥) 2 coactivator-1 (PGC-1) family of transcriptional coregulators serve as inducible coactivators of nuclear receptor and nonnuclear receptor transcription factors involved in the control of mitochondrial biogenesis and cellular energy metabolic pathways. Dysregulation of PGC-1␣ has been implicated in the pathogenesis of diabetes and insulin resistance (1,2). PGC-1␣ was identified through its functional interaction with the nuclear receptor PPAR-␥ in brown adipose tissue (3). PGC-1␣ and its homologue, PGC-1␤, are preferentially expressed in tissues with high oxidative capacity such as heart and skeletal muscle and serve critical roles in the regulation of mitochondrial function and cellular energy metabolism, as does a naturally occurring 270-amino acid 3Ј splice variant (NT-PGC-1␣) recently described by Zhang et al. (4). The docking of PGC-1 coactivators to specific transcription factors provides a platform for the recruitment of regulatory protein complexes that exert powerful effects on gene transcription triggering biological responses that equip the cell to meet the energy demands of the changing environment (5). Multiple PGC-1␣ coactivation targets have now been identified including PPAR-␥, PPAR-␣, PPAR-␤, RXR␣, and glucocorticoid receptor. In addition, several nonnuclear receptor PGC-1 partners have been identified, including nuclear respiratory factors 1 and 2 (NRF-1 and 2) and Foxhead Box(01) (6,7). PGC-1␣ is inducible by a variety of physiological and dietary factors. For example, PGC-1␣ is stimulated by exercise and by fasting (1,5), and PGC-1␣ has recently been linked to regulation of angiogenesis (8). PGC-1␣ thus serves as an inducible regulator of metabolism and other key processes.
Cellular PGC-1␣ levels are controlled in a dynamic state dictated by the balance of its synthesis and degradation. Several studies have addressed determinants of PGC-1␣ synthesis. However, there is relatively little understanding of the mechanism(s) and regulation of PGC-1␣ degradation. Cytokines and phosphorylation have been implicated in the control of PGC-1␣ degradation rates (9). Puigserver's group has shown that acetylation is involved in the post-transcriptional control of PGC-1␣ (10). Recently, Sano et al. (11) and Anderson et al. (12) proposed that PGC-1␣ degradation involves the ubiquitin proteasome pathway. In this pathway, the conjugation of ubiquitin to the protein substrate proceeds via a three-step cascade mechanism. Initially, the ubiquitin-activating enzyme, E1, activates ubiquitin. One of several E2 enzymes (ubiquitin-conjugating enzymes) transfers the activated ubiquitin to the substrate that is specifically bound to a member of the ubiquitinprotein ligase, E3, family. Transfer of ubiquitin can be either directly to the E3-bound substrate or via an additional E3-ubiquitin intermediate. E3s catalyze the last step in the conjugation process, covalent attachment of ubiquitin to the substrate. In successive reactions, a polyubiquitin chain is synthesized by progressive transfer of additional activated ubiquitin moieties to the previously conjugated ubiquitin molecule. The polyubiquitin chain serves as a recognition marker for the 26 S proteasome. Following conjugation of ubiquitin, the target protein is degraded by the proteasome, and free ubiquitin is released and recycled (13).
Characterized by the initial site of ubiquitination, there are two subpathways of the ubiquitin system, the internal lysinedependent pathway and the N terminus-dependent pathway. In most cases, the first ubiquitin moiety is transferred to an ⑀-NH 2 group of an internal lysine residue of the target protein to generate an isopeptide bond. However, recent studies have suggested that the first ubiquitin may be covalently attached to the free and exposed N-terminal residue of the substrate (14). Substrates for N terminus-dependent ubiquitination include MyoD (15), the human papillomavirus 16 (HPV16) oncoprotein E7 (16), the Arf tumor suppressor (17), the Id1 and Id2 developmental regulators (18,19) as well as naturally occurring lysine-less proteins such as HPV58 E7 and p16 INK4␣ (20). The ubiquitination process may occur in the cytoplasm or in the nucleus (21).
Here, we investigated the ubiquitin proteasome-mediated degradation of PGC-1␣. We show that endogenous as well as exogenously expressed full-length PGC-1␣ are localized to the nucleus and degraded very rapidly (t1 ⁄ 2 Ͻ 0.5 h) in a ubiquitin proteasome-dependent manner. Interestingly, a recently reported N-terminal splice variant of 270 residues, NT-PGC-1␣, is cytoplasmic and stable (t1 ⁄ 2 Ͼ 7 h). However, PGC182 (amino acids 1-182), a C-terminally truncated mutant of PGC-1␣ containing four lysines, is nuclear and is very rapidly degraded (t1 ⁄ 2 Ͻ 0.5 h) via the ubiquitin proteasome pathway, similar to full-length PGC-1␣. A lysine-less mutant of PGC182 is also nuclear and rapidly degraded. Our findings thus suggest that PGC-1␣ is primarily degraded via the nuclear N terminusdependent ubiquitin proteasome pathway.

EXPERIMENTAL PROCEDURES
Cell Culture and Chemical Reagents-The C2C12 mouse myoblast and human HeLa and HepG2 cell lines were obtained from the American Type Culture Collection. The cells were propagated and differentiated as described earlier (18). Briefly, cells were propagated in growth medium, Dulbecco's modified Eagle's medium (Sigma) supplemented with 10% fetal bovine serum, 100 units/ml penicillin G, and 100 g/ml streptomycin (Invitrogen) and maintained in a humidified chamber at 37°C with 5% CO 2 . Myogenic differentiation of C2C12 cells was induced by changing the growth medium to differentiation medium, Dulbecco's modified Eagle's medium supplemented with 2% horse serum (HyClone), 100 units/ml penicillin G, and 100 g/ml streptomycin, when cells reached confluence. Cells were then maintained in differentiation medium for 6 days with medium being changed every 24 h. HL1 mouse atrial cardiac myocytes were obtained from Claycomb and propagated as described by Claycomb et al. (22).Transient transfections of HeLa, HL1, HepG2, or C2C12 myoblasts were performed using the FuGENE 6 reagent (Roche Applied Science) according to the manufacturer's instructions (23).
Plasmid Constructs-Full-length PGC-1␣ in the pSV-SPORT vector has been described previously (3). PGC182 was prepared by the addition of a stop codon following amino acid 182. Lysine and cysteine mutations were made using the QuikChange Site-directed Mutagenesis kit according to manufacturer's instructions (Stratagene). N-terminal modifications (6-X-LL-myc, EGFP, and p14 ARF ) were made using an adaptation of the QuikChange Site-directed Mutagenesis kit in which PCR products were used as the homologous primer pair either to insert or to delete sequence in lysine-less PGC182. FLAG-ubiquitin was kindly provided by Aaron Ciechanover (Technion Israel Institute of Technology, Haifa, Israel).
NT-PGC-K3R-myc was amplified by PCR using a forward primer containing an XhoI restriction site and a reverse primer containing the K3R-myc tag with NT-PGC-1␣ as a template. The 867-bp amplicon was subcloned into the pCR2.1-TOPO vector using the TA Cloning kit (Invitrogen). The insert and the pcDNA3.1(Ϫ) vector were then digested with BamHI and XhoI, gel-purified, and ligated.
Determination of Protein Half-life-As described previously (21,23), 16 -20 h after transfection, HeLa, HL1, HepG2, or C2C12 cells were incubated with cycloheximide (CHX) (100 g/ml; Sigma) to inhibit further protein synthesis. MG132 (20 M) was added along with CHX as noted. The cells were lysed after 0, 0.5, 1, 2, and 3 h in PBS containing 5% Igepal, 1 mM EDTA, 1 mM DTT, 2 mM PMSF, 2.5 g/ml leupeptin; and 1 M pepstatin for at least 30 min after which cells were sonicated and then centrifuged at 14,000 rpm for 10 min at 4°C in an Eppendorf microcentrifuge to remove cellular debris. The lysates were mixed with an equal amount of Laemmli sample buffer (Bio-Rad), and equal amounts of sample were run on a 10% Tris-HCl gel (Bio-Rad) followed by electroblotting onto nitrocellulose (Osmonics). For the detection of PGC182 and its mutants, the blots were probed with polyclonal anti-PGC-1␣ (1:1000 dilution) (24) followed by incubation with a secondary horseradish peroxidase-conjugated antibody and detection by chemiluminescence (GE Healthcare). The resulting bands were quantitated using the Kodak EDAS system, and the data were graphed using the Excel graphing program (Microsoft). The protein degradation rate is expressed as half-life (t1 ⁄ 2 ), the time for degradation of 50% of the protein. Half-life data reported were evaluated by Ն3 independent determinations.
Immunoprecipitation-HeLa cells were transfected with PGC182, its lysine-less mutants, and FLAG-ubiquitin in a 1:1 ratio. At 16 -20 h after transfection, cells were lysed in 20 mM Tris (pH 7.6) and 0.25% Triton X-100, with 1 mM DTT added where indicated, for 30 min at 4°C. Cells were scraped into an Eppendorf tube, sonicated, and spun at 14,000 rpm at 4°C for 10 min to remove cellular debris. Direct immunoprecipitation was performed by incubation of PureProteome Protein A Magnetic Beads (Millipore) with polyclonal anti-PGC-1␣ antibody (1:400 dilution) (24) and monoclonal anti-FLAG antibody (1:200 dilution; Sigma) followed by addition of cell lysates. The beads were washed, and following the addition of reducing and nonreducing Laemmli sample buffer (Bio-Rad) the immunoprecipitates were run on a 10% Tris-HCl gel and analyzed via Western blotting. The blots were probed with monoclonal anti-FLAG antibody (1:1000 dilution; Sigma) and polyclonal anti-PGC1␣ antibody (1:1000 dilution) (24) followed by incubation with a secondary horseradish peroxidase-conjugated antibody and detected by chemiluminescence (GE Healthcare).

RESULTS
As an initial step toward defining the pathway of PGC-1␣ degradation, we examined the half-life of endogenous (fulllength, 1-797 amino acids) PGC-1␣ in HL1 cells. As seen in Fig.  1A, endogenous PGC-1␣ is very rapidly degraded with t1 ⁄ 2 ϳ 0.3 h. Incubation of cells with MG132, a potent and selective inhibitor of the proteasome, markedly slowed the rate of PGC-1␣ degradation (Fig. 1A). Exogenously expressed PGC-1␣ behaves similarly. Following transfection of full-length PGC-1␣, the degradation rate was equally rapid (t1 ⁄ 2 ϳ 0.4 h) and was markedly slowed by proteasome inhibition (Fig. 1B). These observations suggest that PGC-1␣ is degraded via the ubiquitin proteasome pathway. In addition, subcellular localization of endogenously expressed PGC-1␣ as well as transfected PGC-1␣ demonstrates predominantly nuclear staining (Fig.  1C, see also Fig. 3, and data not shown). These latter observations raise questions as to where PGC-1␣ is rapidly degraded; within the nucleus or following export to the cytoplasm and via which subpathway of the ubiquitin proteasome system, the internal lysine-dependent pathway or the N terminus-dependent pathway.
To determine whether PGC182 is degraded via the N terminus-dependent subpathway, we systemically mutated all four internal lysine residues to arginine. The cellular localization of lysine-less PGC182 is the same as that seen for PGC182, nuclear with some cytoplasmic staining both in the absence or presence of MG132 (Fig. 3). As seen in Fig. 4, PGC182 K143R/K144R, PGC182 K77R/K143R/K144R, and PGC182 K54R/K77R/ K143R/K144R (i.e. lysine-less PGC182) are all degraded rapidly (t1 ⁄ 2 ϳ 0.5 h). Furthermore, the degradation of each of these species is markedly slowed by MG132 (Fig. 4). These results strongly suggest that PGC182 is degraded via the N terminusdependent subpathway. We also constructed mutants of lysineless PGC182 with 6 lysine-less myc tags (6-X-LL-myc lysineless PGC182), with EGFP or with p14 ARF on the N terminus as the addition of a bulky N-terminal tag to several substrates has been shown to force degradation through the lysine-dependent pathway by preventing ubiquitination at the N-terminal site. These constructs, however, were also degraded rapidly (data not shown).
To determine directly whether lysine-less PGC182 is, as predicted, conjugated to ubiquitin, we transfected cells with PGC182, lysine-less PGC182, or 6-X-LL-myc lysine-less PGC182 with FLAG-ubiquitin in the presence or absence of MG132. High molecular weight ubiquitin conjugates of PGC182, lysine-less PGC182, and 6-X-LL-myc lysine-less PGC182 accumulate in the presence of MG132 (Fig. 5). In addition, the analyses (cell lysis, immunoprecipitation, and SDS-PAGE) were performed in the absence or presence of reducing agents because two recent reports demonstrate degradation substrates that carry ubiquitin moieties attached via thiol-ester conjugates (27,28). As seen in Fig. 6A, when samples were immunoprecipitated with anti-PGC-1␣ and blotted with anti-FLAG, abundant high molecular weight ubiquitin conjugates are seen with PGC182 under both nonreducing and reducing conditions (first and second lanes). A similar pattern of high molecular weight ubiquitin conjugates is also seen with lysineless PGC182 and with 6-X-LL-myc lysine-less PGC182, under nonreducing and reducing conditions (lanes 3-6). Fig. 6B demonstrates the high molecular weight ubiquitin conjugates of PGC182, lysine-less PGC182 and 6-X-LL-myc lysine-less PGC182 following immunoprecipitation with anti-FLAG followed by blotting with anti-PGC-1␣. Furthermore, the ubiquitin-mediated degradation was unaltered in a lysine-less PGC182 mutant in which all five cysteine residues were mutated (i.e. lysine-less/cysteine-less PGC182) (t1 ⁄ 2 ϳ 0.5 h; Fig.  7). Taken together, these results demonstrate the conjugation of ubiquitin to PGC182 as well as to lysine-less PGC182 and 6-X-LL-myc lysine-less PGC182 whose only free amino acceptor is at the N terminus.
A 270-amino acid N terminus splice variant of PGC-1␣, termed NT-PGC-1␣, has been reported recently (4). To determine the degradation rate and localization of NT-PGC-1␣ we transfected HeLa cells with NT-PGC-1␣. The half-life of NT-PGC-1␣ was Ͼ7 h (Fig. 8A) as was the half-life of NT-PGC-K3R-myc (data not shown). As seen in Fig. 8B, NT-PGC-1␣ and NT-PGC-K3R-myc were localized to the cytoplasm. These results demonstrate that, unlike full-length PGC-1␣, the 270amino acid N-terminal variant is both very slowly degraded and localized to the cytoplasm. Taken together, these results further support the conclusion that PGC-1␣ is degraded rapidly in the nucleus via the N-terminal pathway.
To determine whether PGC-1␣ and PGC182 respond to metabolic perturbations, we switched HeLa or HepG2 cells transfected with PGC-1␣ or PGC182 to low serum (0.2%) media. 18 h in low serum media decreased the steady-state concentration of PGC-1␣ by 40% (five determinations) and of PGC182 by 29% (four determinations). We then determined whether the decreases in PGC-1␣ or PGC182 concentrations were a result of altered protein degradation rates. As seen in Fig.  9, whereas at 10% serum PGC-1␣ is degraded rapidly (t1 ⁄ 2 ϭ 0.4h) (A), the rate of degradation is increased (t1 ⁄ 2 ϭ 0.3 h) at 0.2% serum (B). Actin content did not vary. Similarly, PGC182 at 10% serum is degraded at t1 ⁄ 2 ϭ 2 h (C), and the rate of degradation is enhanced (t1 ⁄ 2 ϭ 0.7 h) at 0.2% serum (D). Again, the actin content did not vary. The accelerated rate of PGC182 and PGC-1␣ degradation observed upon low serum exposure is ubiquitin proteasome-dependent. PGC182 is degraded rapidly in HepG2 cells in 10% serum (t1 ⁄ 2 ϭ 1.3 h), and this degradation is completely abrogated by MG132. In 0.2% serum, the rate of degradation is increased (t1 ⁄ 2 ϭ 0.4 h) and similarly is completely abrogated by MG132 (Fig. 10). Similar results were seen with PGC-1␣ (data not shown). These observations demonstrate that PGC-1␣ and PGC182 are able to respond similarly to altered metabolic conditions. However, the molecular regulatory events and pathways involved are yet to be defined.

DISCUSSION
PGC-1␣ is an inducible master regulator of cellular energy metabolism. As with most key regulatory molecules, cellular PGC-1␣ concentration is tightly controlled through a dynamic balance of its synthesis and degradation. Our investigation of the molecular mechanisms underlying PGC-1␣ degradation

Ubiquitin-mediated Degradation of PGC-1␣
DECEMBER 17, 2010 • VOLUME 285 • NUMBER 51 revealed a key role for the nuclear ubiquitin proteasomal pathway system. The following lines of evidence support this conclusion: (i) steady-state, endogenous (or exogenous) PGC-1␣ is localized within the cell nucleus, but outside the nucleolus; (ii) PGC-1␣ is a very short lived protein with a t1 ⁄ 2 of ϳ 0.5 h; (iii) inhibition of the proteasome with MG132 markedly slowed PGC-1␣ degradation; and (iv) an N-terminal variant of 270 amino acids (NT-PGC-1␣) which is predominantly cytoplasmic, is stable (t1 ⁄ 2 Ͼ 7 h). To determine the subpathway of the ubiquitin proteasome system involved in PGC-1␣ degradation, we characterized the degradation of a C-terminally truncated PGC-1␣, PGC182, and its lysine 3 arginine mutants. PGC182 and lysine-less PGC182 are predominantly nuclear and rapidly degraded by the ubiquitin proteasome system. Thus, the nuclear N terminus-dependent pathway appears to play a dominant role in the dynamic balance of cellular PGC-1␣ levels.
PGC-1␣ functions as a master transcriptional regulator of cellular energy homeostasis via control of mitochondrial biogenic and energy metabolic networks, especially in brown fat and cardiac and skeletal muscle where it governs many aspects of fuel metabolism and ATP production. PGC-1␣ coactivates multiple transcription factors involved in mitochondrial biogenesis, oxidative phosphorylation, fatty acid oxidation, and glycogen and glucose flux (29). Given the potency of this coactivator, its activity must be tightly controlled in accordance with energy needs and substrate availability. Indeed, chronic overexpression of PGC-1␣ in heart results in toxicity leading to cardiomyopathy due to exuberant mitochondrial proliferation (24,30). It is well established that the expression of PGC-1␣ is highly regulated at the gene transcriptional level accounting, at least in part, for its inducibility and tissue-specific expression (5,7). Little attention has been directed, however, to the mechanism(s) underlying or the regulation of PGC-1␣ degradation.
Puigserver et al. (9) suggested that PGC-1␣ phosphorylation between residues 262 and 298 served as p38 MAPK targets that slow PGC-1␣ degradation. More recently, Sano et al. (11) demonstrated nuclear localization of N-terminal FLAG-tagged PGC-1␣ and suggested that it was degraded via the ubiquitin proteasome system although no half-life data were reported. Further, they noted two PEST-like regions (amino acids 80 -144 and 255-270) and suggested that the ubiquitination and degradation of PGC-1␣ were dependent upon the integrity of the C terminus containing arginine-serine-rich domains and an RNA-recognition motif (amino acids 565-798). C-terminal deletion mutants showed that PGC-1␣ (amino acids 1-565) was both nuclear and cytoplasmic, whereas PGC-1␣ (1-292) was excluded from the nucleus. In addition, Zhang et al. (4) have also suggested that the C-terminal region is essential for targeting PGC-1␣ for ubiquitination and degradation based on their findings that an alternative 3Ј splice variant (NT-PGC-1␣) of amino acids 1-270 appears to be more stable than PGC-1␣.
Our data herein differ somewhat from the previously published observations. First, we found that both endogenous and exogenously expressed PGC-1␣ is rapidly degraded (t1 ⁄ 2 ϳ 0.5 h) in all cell types examined, including HL1, HeLa, and C2C12. This degradation rate is very rapid and typical of many transcription factors and oncoproteins (31). Second, we found that PGC182 is equally rapidly degraded (t1 ⁄ 2 ϳ 0.5 h) (Figs. 2 and 4) in all cells examined. Sano et al. (11) reported that N-terminal constructs of PGC-1␣ (e.g. amino acids 1-292) required the C-terminal region for ubiquitination and proteasomal degradation, and Zhang et al. (4) reported that the N-terminal (1-270) NT-PGC-1␣ was stable. Our data demonstrate C-terminal truncated PGC182 ubiquitination (Figs. 5 and 6) and rapid  (t1 ⁄ 2 ϳ 0.5 h) proteasome-dependent degradation (Figs. 2, 4, and  7). One potential and significant difference between our studies and those of Sano et al. (11) is the use of N-terminal FLAG tags on all of Sano et al. (11) constructs. As we show here using lysine-less mutants of PGC182, PGC182 can be rapidly ubiquitinated and degraded via its N terminus. It is possible that in the Sano et al. studies (11) the FLAG tag interfered with the physiological ubiquitination and degradation pathway. Third, we demonstrate that the C-terminal region(s) of PGC-1␣ are not required for the N-terminal region to be ubiquitinated and degraded. In addition, although PGC-1␣ appears to contain two PEST-rich regions, one of which is contained within PGC182 (i.e. amino acids 80 -144), there are at present no data that support the role of this region in PGC-1␣ degradation. Furthermore, although our data demonstrate similar half-lives of PGC-1␣ and PGC182, the 270 residue-N-terminal variant (NT-PGC-1␣) is both remarkably stable (t1 ⁄ 2 Ͼ 7 h) and localized to the cytoplasm. Finally, although the N-terminal constructs of PGC-1␣ of Sano et al. (11) and Zhang et al. (4) are relatively stable, they have not determined whether the limitation to degradation is the ubiquitination of the target protein or the inability of the ubiquitinated substrate to be degraded by the proteasome. PGC182, a nuclear protein, is however both ubiquitinated and degraded by the proteasome.
It has been well recognized that single ubiquitin moieties are coupled to the ubiquitin-activating enzyme (E1) and ubiquitin-   DECEMBER 17, 2010 • VOLUME 285 • NUMBER 51 conjugating enzymes (E2s) via a thiol-ester bond (31). The recent reports that two proteins (Herpesvirus MIR1 and ubiquitin-conjugating enzyme-7) are conjugated to a multiubiquitin chain via thiol-ester linkage (27,28) deserve comment. This multiubiquitin chain thiol-ester linkage to a target protein is sensitive to cleavage upon reduction (e.g. ␤-mercaptoethanol). Our results in Fig. 6, which demonstrate the identical pattern and abundance of ubiquitin conjugates of PGC182 and lysineless PGC182 mutants under nonreducing and reducing conditions, and results in Fig. 7 of unaltered ubiquitin-mediated degradation of lysine-less/cysteine-less PGC182 eliminate the possibility of thiol-ester-linked ubiquitin conjugates.

Ubiquitin-mediated Degradation of PGC-1␣
Our observation that PGC182 and its lysine-less derivatives are both nuclear and rapidly ubiquitinated and degraded via the proteasome and that NT-PGC-1␣ is both very stable and cytoplasmic also deserves comment. In terms of localization, fulllength PGC-1␣ is predominantly nuclear, as has been observed by others (3,11). A putative nuclear localization sequence has been suggested within the C-terminal region (amino acids 651-668) (32). Our observation that PGC182 and its lysine-less derivatives are predominantly nuclear is of interest because it contains no apparent nuclear localization sequence, either classical or bipartite, as for example is found in MyoD (21). Thus, either PGC182 has a nonclassical nuclear localization sequence or is chaperoned into the nucleus in association with one or more carriers. If the latter is the case, the chaperone abundance and/or efficiency must be remarkable because the observations with PGC182 were made following overexpression of the protein at a cellular level. Further studies will address these alternatives. Presumably, NT-PGC-1␣ is unable to associate with chaperone carriers, interact with distinct cytoplasmic proteins, or has a dominant nuclear export sequence in residues 182-270.  Fig. 1 or (B) fixed, and localization of NT-PGC-1␣ was visualized via immunofluorescence of antibody to PGC-1␣ as described in Fig. 4. The t1 ⁄2 for NT-PGC-1␣ was Ͼ7 h; two experiments are shown in filled circles in A.  Perhaps more important is the demonstration that PGC182 is degraded via the N terminus-dependent pathway. The N terminus-dependent pathway is relatively recently described. Here, the initial ubiquitin moiety is covalently attached to the free and exposed ␣-amino group of the N-terminal residue of the substrate (20). Thereafter, a polyubiquitin chain is built upon the initial conjugated ubiquitin. Substrates for this pathway include at least a dozen proteins, several of which are transcription factors or other key regulatory molecules: e.g. MyoD (15,21), HPV16 oncoprotein E7 (16), the Id1 and Id2 developmental regulators (18,19), p16 INK4a (20), p19 ARF , p14 ARF (17). The N terminus-dependent pathway is the predominant degradation pathway for several of these substrates including Id1 and Id2.
Despite the recognition of the N terminus as a site for ubiquitination (14), in only three instances (HPV58-E7, p21, and ERK3) has direct biochemical sequence evidence established ubiquitin covalently attached to the ␣NH 2 terminus of the target protein (14,33). Most often the evidence is indirect and includes polyubiquitin-mediated proteasome-dependent rapid degradation of the lysine-less target protein, either naturally occurring lysine-less (e.g. p14 ARF or HPV58-E7) or lysine-less mutants of the wild type protein (e.g. MyoD, LMP1). Many of these target proteins can be stabilized by addition of a bulky lysine-less peptide (e.g. 6ϫmyc or His 6 -4HA). However, addition of HA tags to lysine-less p19 ARF did not provide stability (17). Recently, Anderson et al. (12) suggested that PGC-1␣ is degraded by the proteasome within the nucleus. Using N-terminal GFP-tagged PGC-1␣, they demonstrated increased nuclear accumulation of PGC-1␣ upon oxidative stress. However, they did not demonstrate nuclear degradation. We, however, have found rapid degradation of N-terminal EGFP, 6-X-LL-myc, and p14 ARF tagged lysine-less PGC182 (data not shown). In addition, other N-terminally degraded proteins have not been examined following the addition of bulky N-terminal peptides (e.g. HMG-CoA reductase (34) and neurogenin (35)). Our results with PGC182 demonstrate polyubiquitination and rapid proteasome-dependent degradation for wild type PGC182 and its lysine-less mutants, strongly supporting the N terminus-dependent pathway.
The rapid ubiquitin proteasome-dependent degradation of PGC-1␣ and PGC182 raises the issue as to whether this regulatory mechanism is influenced by metabolic perturbations. Indeed, steady-state levels of PGC-1␣ and PGC182 decrease upon incubation in low serum media in HepG2 and HeLa cells. This decrease in PGC-1␣ and PGC182 levels occurred concomitantly with an increase in ubiquitin proteasome-dependent degradation rates (Figs. 9 and 10). These observations suggest that the degradation rates of PGC-1␣ are influenced by substrate availability (e.g. fatty acids) and/or serum regulatory factors. Future studies aimed at the physiological regulation of PGC-1␣ degradation are warranted.
Using nuclear localization sequence and nuclear export sequence mutants of MyoD, we demonstrated that the N terminus-dependent pathway primarily functions within the nucleus (21). Thus, the results seen in Figs. 4 and 5 strongly suggest that the N terminus-dependent pathway is the predominant route of PGC182 degradation. In addition, it is likely that PGC182 is degraded via this pathway within the nucleus as is MyoD (25); however, definitive proof awaits resolution of the nuclear-mediated uptake/export mechanism(s) as discussed above. Whether this is the case for the full-length molecule awaits further study. Consistent with this notion, NT-PGC-1␣ is both stable and cytoplasmic. It is tempting to speculate that NT-PGC-1␣ serves a specific transcriptional coregulatory role that requires longer action following the potent acute transcriptional coactivating effects of full-length PGC-1␣. Taken together, these data strongly suggest that a nuclear ubiquitinproteasome system governs the rapid degradation of PGC-1␣ and its derivatives.