Fatty acids regulate perilipin5 in muscle by activating PPARδ.

The surface of lipid droplets (LDs) in various cell types is coated with perilipin proteins encoded by the Plin genes. Perilipins regulate LD metabolism by selectively recruiting lipases and other proteins to LDs. We have studied the expression of perilipins in mouse muscle. The glycolytic fiber-enriched gastrocnemius muscle expresses predominantly Plin2-4. The oxidative fiber-enriched soleus muscle expresses Plin2-5. Expression of Plin2 and Plin4-5 is elevated in gastrocnemius and soleus muscles from mice fed a high-fat diet. This effect is preserved in peroxisome proliferator-activated receptor (PPAR)α-deficient mice. Mouse muscle derived C2C12 cells differentiated into glycolytic fibers increase transcription of these Plins when exposed to various long chain fatty acids (FAs). To understand how FAs regulate Plin genes, we used specific activators and antagonists against PPARs, Plin promoter reporter assays, chromatin immunoprecipitation, siRNA, and animal models. Our analyses demonstrate that FAs require PPARδ to induce transcription of Plin4 and Plin5. We further identify a functional PPAR binding site in the Plin5 gene and establish Plin5 as a novel direct PPARδ target in muscle. Our study reveals that muscle cells respond to elevated FAs by increasing transcription of several perilipin LD-coating proteins. This induction renders the muscle better equipped to sequester incoming FAs into cytosolic LDs.


Cloning and mutagenesis of the Plin5 reporter
The mouse Plin2 and Plin4 LUC reporters have been described elsewhere ( 19,41 ). The full-length mouse Plin5 promoter ( Ϫ 2324/ +244) was amplifi ed by a PCR strategy described previously ( 51 ), cloned into pPCR-Script (Stratagene), digested out using Hind III, and inserted into the pGL3-Basic luciferase reporter vector (Promega, Madison, WI). Site-directed mutagenesis of the DR-1 the energy status of the organism is high (postprandial), but facilitates lipolysis when stored energy needs to be released (fasted). These events are controlled by the phosphorylation status of perilipin1, which determines recruitment of lipolytic enzymes to the LD surface. This regulatory mechanism cannot be compensated for by other perilipins in the lack of perilipin1 (20)(21)(22)(23). Distinct functional roles for the remaining perilipins are less clear. Other perilipin members similarly interact with lipases and associated factors at the LD surface (24)(25)(26)(27). However, functional compensation precludes characterization of individual perilipins in cells and mice ( 28 ). It is clear that Plin2 -and Plin5 -null mice have reduced hepatic and cardiac accumulation of lipids, respectively ( 29,30 ). Perilipin5 has a unique ability to facilitate physical linkage of LDs and mitochondria when ectopically expressed ( 31,32 ), but the signifi cance of this association is unclear. Depending on cell type, ectopic expression of perilipin5 either prevents ( 7 ) or enhances lipolysis ( 15 ). An obligatory role for per-ilipin5-mediated association of LDs and mitochondria for enhanced oxidation ( 31,33 ) might explain the discrepancies in the initial reports. The functions of perilipin3 and perilipin4, other than binding to LDs, are poorly understood.
Depending on physiological conditions and the presence of other perilipin family members, perilipins may be cytosolic, bound to LDs, or rapidly degraded and thus nearly absent in the cell ( 7,34,35 ). Due to their role in lipolysis, the type of perilipins being expressed and bound to LDs are an important determinant of cellular LD metabolism. It is therefore important to identify transcription factors and pathways that control expression of the various Plin genes. Several of the Plin genes contain evolutionary conserved cis-regulatory elements occupied by peroxisome proliferator-activated receptors (PPARs). The PPARs belong to the nuclear receptor superfamily and consist of the isotypes PPAR ␣ , PPAR ␤ / ␦ (hereafter referred to as PPAR ␦ ), and PPAR ␥ . They all heterodimerize with retinoid X receptors (RXRs) onto mainly DR-1 type PPAR response elements (PPREs) in the promoter region of target genes ( 36,37 ). Adipose expression of Plin1 and Plin4 is switched on by binding of PPAR ␥ to their respective promoter regions ( 19,(38)(39)(40) . Plin2 is regulated by PPAR ␣ (41)(42)(43) and PPAR ␦ (44)(45)(46) in various cell types, whereas expression of Plin3 seems unaffected by activation of PPARs ( 19,41 ). Regulation of Plin5 by PPARs is poorly understood. Expression of Plin5 is enhanced by activation of PPAR ␣ ( 7,15,16 ), but no PPRE has been identifi ed in the Plin5 gene.
Little is known regarding transcription factors important for expression of perilipins in muscle. Given the known regulation of Plin genes by PPARs in other tissues, we analyzed FA and PPAR regulation of Plin genes. Our analyses revealed an unexpected importance for PPAR ␦ as a FA sensor regulating the expression of the Plin4 and Plin5 genes in muscle. We further identifi ed a conserved PPRE in the Plin5 gene. This PPRE is essential for FA-stimulated expression of perilipin5 and establishes Plin5 as a direct PPAR ␦ target gene.
(result not shown). These assays confi rm that both cell lines were well differentiated into contractile myotubes.
Unless otherwise indicated, C2C12 and Sol8 cells were seeded in 12-well dishes at a density of 3 × 10 4 cells/well. Two days later, differentiation was initiated by changing to Diff-medium. For transfection experiments, C2C12 and Sol8 cells were seeded at 6 × 10 4 cells/well in antibiotic-free medium. The following day, cells were given 1 ml antibiotic-free Diff-medium, prior to transfection with 2 g DNA:4 l Lipofectamine2000 complexed in 200 l OPTI-MEM (Life Technologies Corporation). After 6 h, medium was replaced with Diff-medium containing antibiotics and allowed to grow for a maximum of 4 days before being harvested.
Following binding of primary antibodies, membranes were incubated with species-specifi c horseradish peroxidase-labeled secondary antibodies (Abcam, goat to rabbit IgG-HRP, #ab6721, 1:10,000 and rabbit to mouse IgG-HRP, #524567, 1:10,000) and binding detected using ECL Plus (GE Healthcare), or alkaline effi ciency. The knockdown effi ciency was found to be comparable 3 and 4 days after siRNA transfection.

Reporter gene expression assay
C2C12 cells (5 × 10 3 cells/well) were seeded in 96-well dishes in 75 l antibiotic-free medium. The next day, 75 ng DNA (50 ng reporter, 10 ng expression vectors, and 5 ng pRL) and 0.5 l Lipofectamie2000 were mixed in 2 × 10 l OPTI-MEM I (Life Technologies Corporation) and added to cells incubated in 75 l Diffmedium. After 5 h, culture medium was replaced with Diff-medium and incubated for 2 days. Fresh medium containing FAs and PPAR antagonists was added for an additional 24 h. Cells were washed in 1× PBS and lysed in 20 l Passive Lysis buffer (#E194A, Promega). Dual luciferase activity was determined using Dual-Luciferase® Reporter Assay System (#E1910, Promega) and luciferase activity measured with a Synergy 2 Luminometer (BioTek, Winooski, VT).

ChIP experiments
Chromatin immunoprecipitation (ChIP) experiments were performed as described previously with minor modifi cations ( 53 ). Briefl y, C2C12 cells were cross-linked with 1% formaldehyde for 10 min. Cross-linking was terminated by 10 min incubation with 0.125 M glycine. Cells were washed twice in cold 1× PBS and harvested in lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.0) containing complete Protease Inhibitor Cocktail (Roche, #4693116001). Lysed cells were sonicated using a Bioruptor (Diagenode, Belgium) to fragments of 300-500 bp. Chromatin phosphatase-conjugated-labeled species-specifi c secondary antibodies using Western Breeze® Chemiluminescent kit (Life Technologies Corporation). Chemiluminescent signals were visualized with exposure to Hyperfi lm ECL (GE Healthcare). Carestream MI SE was used to quantify Western blots.

Preparation and analysis of RNA
Cells were lysed in 500 l 1× Total RNA Lysis Solution (#4305895, Life Technologies Corporation) per well (12-well plate) and frozen at Ϫ 80°C before isolation. Total RNA from cell extracts was isolated using ABI 6100 Nucleic Acid Prep-Station using the preprogrammed "RNA-Cell method" (Life Technologies Corporation).
Total RNA (cells, 12 ng/ l; muscle, 50 ng/ l; and liver, 12 ng/ l) was reverse transcribed into single-stranded cDNA using high capacity cDNA reverse transcription kit (Life Technologies Corporation, #4368814). Quantitative real-time PCR amplification (1 l cDNA reaction in 20 l reaction volume) was performed using TaqMan® Universal PCR Master Mix on an ABI 7900HT system (Life Technologies Corporation) operating with standard settings. RNA was analyzed using predesigned TaqMan® Low Density Custom Arrays (liver and soleus) or predesigned single assays (gastrocnemius and cell cultures). Assays used: Plin1, #Mm00558672_m1; Plin2, #Mm00475794_m1; Plin3,

Animal experiments
All animal use was approved and registered by the Norwegian Animal Research Authority. Mice were housed in a temperature controlled (22°C) facility with a strict 12 h light/dark cycle. Animals were euthanized by cervical dislocation; tissue samples were dissected, snap-frozen in liquid nitrogen, and stored at Ϫ 80°C until further analysis.
Male backcrossed congenic PPAR ␣ Ϫ / Ϫ mice (B6.129S4-Ppara tm1Gonz N12; Jackson Laboratory, Bar Harbor, ME) and PPAR ␣ +/+ controls (C57BL/6J; B and K Universal Ltd., Norway) 9 weeks of age were ad libitum fed a standard chow diet [64% carbohydrate, 31.5% was collected by centrifugation, concentration determined by A260, and diluted in RIPA buffer (0.1% SDS, 0.1% Na-deoxycholate, 1% Triton X-100, 1 mM EDTA, 0.5 mM EGTA, 140 mM NaCl, 10 mM Tris-HCl, pH 8.0) and immunoprecipitated with 2 g antibody against RNAPII (Santa Cruz Biotechnology, CA; #sc-899) overnight at 4°C in the presence of protein A beads (GE Healthcare). Beads were washed three times in RIPA buffer and eluted in 1% SDS with 0.1 M NaHCO 3 . Chromatin was de-cross-linked by adding 0.2 M NaCl and incubating overnight at 65°C. DNA was purifi ed by phenol-chloroform extraction, precipitated in ethanol with sodium acetate, and dissolved in water. DNA enrichment was quantifi ed by real-time PCR (ABI, 7900HT) using SYBR Green Master Mix (Life Technologies Corporation). The following primers were used:  determine if a HFD changes the expression of Plin genes in a FA-oxidative muscle type, Plin mRNA expression levels were evaluated in soleus muscle from wild-type (WT) and PPAR ␣ knock-out (KO) mice fed a chow or high-fat diet. PPAR ␣ KO mice were included in the study, because activation of PPAR ␣ is known to stimulate transcription of Plin2 and Plin5 in liver ( 7,15,16,(41)(42)(43). Plin2 mRNA expression increased 2-fold by HFD in WT but not in PPAR ␣ KO mice ( Fig. 1 ). Plin5 mRNA expression was lower in PPAR ␣ KO mice compared with WT mice, similar to what was previously reported in liver ( 7 ). Interestingly, expression of Plin5 increased by similar magnitude in both genotypes fed a HFD (2.5-fold and 2.9-fold in PPAR ␣ KO and WT , respectively), demonstrating that regulation of Plin5 can occur independent of PPAR ␣ in muscle. In accord with the changes in gene expression, perilipin2 and perilipin5 protein increased with a HFD in WT mice ( Fig. 1B ). In the absence of PPAR ␣ , both perilipin2 and perilipin5 proteins were substantially lower. Perilipin2 was not detected, whereas perilipin5 levels increased by HFD also in PPAR ␣ KO mice. The mRNA and protein levels of Plin3 and Plin4 remained relatively unchanged by treatment or genotype. protein, 2% fat (fat source soya oil)] or a high-fat diet (HFD) [36% carbohydrate, 20% protein, 35.5% fat (fat source lard); #F3282, 1/2 ″ pellet (BioServ, Frenchtown, NJ)] for a period of 13 weeks.

Statistical methods
All results are presented as means with standard error of the mean (SEM) or standard deviation (SD). One-way analysis of variance (ANOVA) followed by Tukey's multiple comparison tests or twotailed Student's t -tests were used to assess signifi cance ( P < 0.05).

HFD feeding of mice increases expression of selective perilipins in muscle
Prolonged feeding of mice with a HFD increases circulating FAs and stimulates uptake and utilization of FAs as the preferred energy source in peripheral tissues. To medium supplemented with various FAs complexed to BSA. The expression of Plin2 and Plin4-5 all increased by various FAs, but with different magnitude ( Fig. 2A ). Plin4 and Plin5 were mainly induced by unsaturated long chain FAs, whereas Plin2 mRNA was elevated by both saturated and unsaturated long chain FAs. Similar to what we have shown in liver ( 41 ), the expression of Plin3 remained unchanged.
All of the added FAs stimulated formation of small intracellular LDs (result not shown). To further verify that the added FAs were incorporated into intracellular LDs, we determined perilipin2 and perilipin3 levels ( Fig. 2B ). In contrast to perilipin3, perilipin2 is known to be posttranslationally regulated and rapidly degraded by the proteasome in the absence of intracellular LDs ( 41 ). The solid increase in perilipin2 and unchanged levels of perilipin3 confi rms that the FAs were incorporated into intracellular LDs. No clear protein signals were observed for the per-ilipin4 or perilipin5 proteins, in agreement with the low mRNA expression of these Plins.

The Plin genes are differently regulated by selective activation of PPARs
Long chain FAs have previously been shown to stimulate Plin2 mRNA expression in various tested cell types ( 41,46,55 ), but the mechanism involved has not been clarifi ed. FAs might activate genes by acting as physiological ligands for PPARs ( 36 ). We therefore determined the expression level of each PPAR isotype during myotubedifferentiation and used the Ct values at day 0 as an indication of relative expression of each PPAR isotype. PPAR ␦ was highly expressed with unaffected levels during differentiation in both cell lines. PPAR ␥ was slightly downregulated during differentiation, whereas PPAR ␣ was expressed at low levels with increased expression during differentiation (supplementary Fig. III).
To determine if there is a PPAR isotype-dependent regulation of the Plin genes in muscle cells, cells were differentiated for 6 days and stimulated with specifi c PPAR activators for 24 h ( Fig. 3A ). Activation of PPAR ␣ (WY-14643; 10 M) in C2C12 cells had little effect on the Plin mRNA levels, except for an increase in Plin5 mRNA (5-fold). Given the low basal expression of Plin5 in C2C12 cells (Ct = 35), such a minimal regulation is expected. Activation of PPAR ␦ (GW501516; 0.1 M) increased mRNA content of Plin2 (4-fold), Plin4 (5-fold), and Plin5 (46fold), whereas Plin3 mRNA remained unchanged. Activation of PPAR ␥ (Rosi, Tro, and GW1929; 1 M) increased mRNA levels of Plin4 (about 3-fold) and Plin5 (2-fold). The PPAR ␥ antagonist (GW9662; 1 M) had no effect on the mRNA levels of the Plin genes.
The profound response to activation of PPAR ␦ compared with other PPARs might be explained by the relatively higher expression level of this particular PPAR in cultured C2C12 cells. To determine the effi ciency for each PPAR to regulate Plin genes, C2C12 cells were transfected to ectopically express PPAR ␣ , PPAR ␦ , or PPAR ␥ , differentiated for 3 days, and stimulated for 24 h with PPAR isoform specifi c ligands ( Fig. 3B ). Comparable ectopic expression of Next, we analyzed Plin mRNA content in gastrocnemius, which contains a mixture of glycolytic and oxidative fi bers. In this tissue, Plin2 and Plin4 mRNAs increased by HFD in a PPAR ␣ -dependent manner ( Fig. 1C ), whereas the Plin5 mRNA increased by HFD in both WT and PPAR ␣ KO mice. Essentially the same changes in regulation were observed at the protein level (result not shown). These results implicate that a HFD alters the protein and mRNA levels of several perilipins regardless of fi ber composition, and that other factors in addition to PPAR ␣ may regulate transcription of Plin5.

FAs regulate Plin genes in cultured myotubes
The increased expression of several Plin genes by the diet rich in lipids may be directly mediated by FAs or secondarily due to complex physiological changes that occur with increased levels of circulating FAs. To be able to identify molecular mechanisms inducing Plin genes in muscle tissues, we continued our investigation using cultured muscle cells. We established culturing conditions and confi rmed differentiation of myoblasts into myotubes (see Experimental Procedures) in two different myotube cell cultures, C2C12 and Sol8 cells. Cells were seeded at day Ϫ 2, grown to confl uence, and subjected to myotube differentiation for up to 7 days. The expression of the Plin genes varied somewhat in the two cell lines, but was less affected by differentiation in C2C12 cells (see supplementary Fig. II). As only C2C12 cells expressed Plin2-5, this cell line was used in further studies.
To test if the Plin genes are regulated by FAs, C2C12 cells were differentiated for 6 days and cultured for 24 h in  the various PPARs was confi rmed by expressing 6xHis-PPAR vectors in C2C12 cells and visualizing the fusion proteins with an antibody recognizing the His epitope ( Fig. 3C ). Ectopic expression and activation of PPAR ␣ induced Plin2 and Plin5 mRNAs. Very similar results were observed with ectopic expression and activation of PPAR ␦ , which induced expression of Plin2, Plin4, and Plin5. Ectopic expression and activation of PPAR ␥ 1 or PPAR ␥ 2 foremost increased Plin4 mRNA. These results demonstrate that each PPAR differently regulates expression of Plin2-5 mRNAs in C2C12 cells.
To determine if addition of FAs and PPAR ␦ alters expression of the perilipin4 and perilipin5 proteins, C2C12 cells were differentiated in the presence of BSA-OA and/ or the PPAR ␦ agonist (GW501516). The fold induction of Plin2, Plin4, and Plin5 mRNAs ( Fig. 3D ) were somewhat higher than overnight stimulation of differentiated myotubes (compare Fig. 2 and Fig. 3B ). Stimulation with the PPAR ␦ agonist during differentiation induced the short perilipin5 protein [amino acids 16-463, see ( 7 )], whereas no clear signal was observed for OA treatment ( Fig. 3E ). A clear band indicative for perilipin4 protein was not ob-served. Given the low levels of perilipin5 protein, we tested if Plin5 behaves similarly to Plin2, which is posttranslationally stabilized by the cellular content of LDs ( 7,34 ). Vectors expressing the full-length perilipin2, the shorter alternative translated perilipin2 (plin2 a123-425), perilipin5, and perilipin5 a16-463 were transfected into C2C12 cells. As expected, expression of the two perilipin2 isoforms was highly dependent on BSA-OA supplementation of the culture media ( Fig. 3F ). In contrast, ectopic expression of the perilipin5 isoforms was unaffected by addition of BSA-OA. Together, these results show that activation of PPAR ␦ induces Plin5 mRNA and protein in C2C12 cells.

FAs are unable to induce Plin mRNAs in the presence of a PPAR ␦ antagonist
To test if the observed FA effect on Plin mRNAs depends on a particular PPAR, differentiated C2C12 cells were incubated with a combination of BSA-OA and BSA-LA (FAs; 50 M each) in the presence of selective PPAR inhibitors GW6471 for PPAR ␣ ( 56 ), GSK0660 for PPAR ␦ ( 57 ), and GW9662 for PPAR ␥ ( 58 ). Incubation with FAs increased mRNA levels of Plin2, Plin4, and Plin5. Coincubation with Fig. 8. Oleic acid or PPAR ␦ agonist treatment of mice elevates Plin5 in soleus muscle. Mice were gavaged twice (36 and 12 h before being euthanized) with vehicle (0.5% CMC), GW501516 (5 mg/kg), or 400 l glyceryl trioleate/triolein. Mice were euthanized at the onset of the light cycle. A: Relative gene expression of Plin2-5 in soleus muscle relative to TBP. Results are presented as mean ± SEM (n = 4-6 per group). Statistical differences from CMC treatment (* P < 0.05, ** P < 0.01). B: Western blot of perilipin4, perilipin5, and grp78 (control) proteins in soleus muscle. Each treatment is represented by four independent mice. C: Relative levels of perilipins/grp78 (n = 4, difference from CMC treatment, ** P < 0.01 ). assays with Plin5 WT or mutated reporter constructs. C2C12 cells were transfected with reporters and expression vectors encoding for RXR ␣ and/or PPAR ␦ and differentiated into myotubes for 3 days. Cells were then stimulated with vehicle or the PPAR ␦ activator GW501516 (0.1 M) for 24 h. Transfection with the Plin5 WT or mutated reporters had similar basal activity, demonstrating that the mutation itself did not affect basal transcriptional activity. Coexpression with RXR ␣ and PPAR ␦ , and stimulation with GW501516 gradually increased Plin5 WT reporter activity up to a maximal 70-fold increase ( Fig. 7C ). In contrast, no induction was observed with the mutated Plin5 reporter construct.
We fi nally tested if the PPAR ␦ -dependent regulation of Plin5 gene expression by FAs depends on the identifi ed response element. Plin5 WT or mutated reporters were transfected into cells together with empty or RXR ␣ and PPAR ␦ expression vectors and stimulated as shown. For the Plin5 WT reporter, stimulation with GW501516 or FAs gave an expected increase in reporter activity, whereas GSK0660 attenuated the FA effect ( Fig. 7D ). In contrast, for the Plin5 mutated reporter construct, none of the treatments had any effect on reporter activity.

Expression of Plin5 is induced by a PPAR ␦ agonist and oleic acid in vivo
Our results from cell studies support that FAs regulate expression of Plin4 and Plin5 in muscle by activating PPAR ␦ . To determine if the same occurs in muscle tissue, mice were given an oral gavage of a PPAR ␦ agonist (GW501516) or FAs in the form of triglycerides (glyceryl trioleate; triolein). Treatment with GW501516 increased Plin4 mRNA and showed a tendency for an induction of Plin5 mRNA in soleus muscle ( Fig. 8A ), whereas triolein treatment increased both Plin4 and Plin5 mRNAs. The same inductions occurred in gastrocnemius muscle (results not shown).The induction was stronger for Plin5, and only the perilipin5 protein was signifi cantly elevated by these treatments ( Fig. 8B, C ).

DISCUSSION
Activation of PPAR ␣ and PPAR ␥ are known to stimulate expression of Plin genes in liver ( 7,15,16,(41)(42)(43) and adipose tissue ( 19,(38)(39)(40), respectively. By using PPAR isoform-specifi c agonists and antagonists, we demonstrate that PPARs selectively regulate Plin genes in muscle cells. PPAR ␦ is required for FA-induced transcription of Plin5 through a conserved DR-1 element in intron 1 of the Plin5 gene. The identifi cation of this functional response element classifi es Plin5 as a novel direct PPAR target gene, similar to Plin1 , Plin2 , and Plin4 . Plin3 seems to be an exception in the family, by not being regulated by PPARs.
The different responses to PPAR activation are likely important for the distinct Plin tissue expression profi les ( 7,19 ). In muscle, enhanced PPAR ␣ and PPAR ␦ activation increases Plin5 expression, whereas the adipose-enriched Plin4 is preferentially induced by enhanced PPAR ␥ activity. In contrast, Plin2 shows no clear PPAR isoform preference, and is nonresponsive to manipulation of PPAR ␦ expression. antagonists for PPAR ␣ or PPAR ␥ did not affect the FAstimulating effect. In contrast, coincubation with the PPAR ␦ antagonist blunted the effect of FAs on Plin2, Plin4, and Plin5 mRNAs ( Fig. 4A ). The PPAR ␦ antagonist was also effective in preventing induction of the perilipin2 protein ( Fig. 4B ).
To determine if the observed increase in Plin mRNAs by FAs was due to transcriptional stimulation, Plin2 ( 41 ), Plin4 ( 19 ), and Plin5 luciferase reporters were cotransfected with PPAR expression vectors into myoblast C2C12 cells. Cells were subsequently differentiated into myotubes prior to 24 h incubation with FAs and PPAR antagonists. Ectopic expression of each PPAR stimulated basal reporter activity with different effi ciency (not shown). Addition of FAs increased reporter activity further, and in this context, PPAR ␦ had a unique effect ( Fig. 5A-C ). Coexpression of PPAR ␦ resulted in a marked boost in Plin4 and Plin5 reporter activity upon activation with FAs, whereas the PPAR ␦ antagonist reversed the effect of FAs.
To strengthen the observation that FAs increase Plin mRNAs by stimulating transcriptional activity, we analyzed Pol II recruitment to the Plin genes using ChIP with primers located 100-200 bp downstream of the transcriptional start sites. Stimulation with either FAs or the PPAR ␦ agonist (GW501516) increased Pol II recruitment to the Plin2 , Plin4 , and Plin5 genes, whereas coincubation with the PPAR ␦ antagonist (GSK0660) reversed the effect of FAs ( Fig. 5D ). Taken together, our results suggest that FAs stimulate transcription of several Plin genes in a PPAR ␦dependent manner.

Silencing of PPAR ␦ attenuated the effect of FAs on Plin5 and Plin4 mRNAs
To determine the effect of FAs in cells with reduced expression of PPAR ␦ , C2C12 cells were reverse transfected with scramble siRNA or siRNAs against PPAR ␦ and subjected for myotube differentiation. PPAR ␦ mRNA was knocked down by 70% from day 2 to day 4 post-transfection ( Fig. 6 ). Two days post-transfection, cells were incubated with the PPAR ␦ activator (GW501516; 0.1 M), FAs (100 M) alone, or in combination with the PPAR ␦ antagonist (GSK0660, 1 M). Incubation with the FAs increased Plin2, Plin4, and Plin5 mRNA levels in scramble-transfected cells, with the FA effect attenuated upon addition of the PPAR ␦ antagonist ( Fig. 6 ). Strikingly, the stimulating effect on Plin5 mRNA by FA incubation was abolished in PPAR ␦ siRNA-transfected cells, underlying the importance of PPAR ␦ in Plin5 regulation. Silencing of PPAR ␦ also blunted FA-stimulated induction of Plin4 mRNA, but had little effect on Plin2 mRNA.

The Plin5 intron 1 contains an evolutionary conserved PPRE
The Plin1 , Plin2 , and Plin4 promoters all contain functionally characterized PPREs ( 19,41 ). Our results prompted us to analyze the Plin5 gene for the presence of a functional PPRE. A promising DR-1 type element was identifi ed in intron 1, in an area conserved to the human Plin5 gene ( Fig. 7A, B ). To test the functionality of this element, we mutated the DR-1 element and performed reporter In addition to PPAR redundancy, the distinct regulation of Plin2 may also be infl uenced by other transcription factors ( 59,60 ). Several factors may infl uence recruitment of PPAR isoforms to a particular PPRE and subsequent transcriptional regulation. Among these are tissue variability in chromatin packing within each Plin locus, specifi c coregulator recruitment, the presence of activating or repressing ligands, or the nucleotide composition of the PPREs and nearby binding sites ( 36 ). The level of endogenous ligands clearly plays a role in our cell studies. PPAR ␦ is highly expressed in muscle tissues and C2C12, yet perilipin5 is only detected in C2C12 cells after prolonged culturing in the presence of a specifi c PPAR ␦ agonist. Another factor is likely the nucleotide compositions of the various PPREs found to regulate Plin genes. These sequences are highly conserved across species (human, mouse, and rat), but vary considerably among the various Plin genes ( 19,41 ). This suggests that the PPAR isoform regulating each Plin gene is evolutionarily conserved and that this PPAR-specifi c regulation is physiologically important.
Expression of Plin5 was initially described to correlate with PPAR ␣ activity ( 7,15,16 ), but confl icting observations provided evidence for alternative regulation of the gene. In the lack of PPAR ␣ , the fold-induction of Plin5 is preserved by fasting in liver ( 7 ). We now demonstrate that Plin5 expression is elevated in muscle by a HFD in the absence of PPAR ␣ . Fasting and a HFD can be viewed as contrary physiological conditions with low or excess energy, respectively, but have elevated circulating FAs in common, caused by hormone-stimulated lipolysis of adipose TAG or increased digestion and uptake of FAs. Increased expression of Plins has been observed in various cell types upon stimulation with FAs ( 41,46,61 ), but the molecular mechanisms have been unclear. Our data provide a molecular explanation for regulation of Plin5 in the absence of PPAR ␣ . Long chain unsaturated FAs may bind directly to the ligand binding pocket of PPAR ␦ and regulate its transcriptional activity ( 62 ), which in turn stimulates transcription of the Plin5 gene. The observation that PPAR ␦ may sense circulating free FAs and regulate a subset of hepatic genes independent of PPAR ␣ ( 63 ) supports such a mechanism. The discovery of a phospholipid as the endogenous PPAR ␣ ligand ( 64 ) also questions the role of PPAR ␣ as a direct FA sensor. The role of PPAR ␣ may rather involve the receptor's ability to stimulate transcription of genes facilitating muscular FA uptake ( 65 ). Our data suggest that PPAR ␦ is the physiological regulator of perilipin5 in muscle.
PPAR ␦ is nearly ubiquitously expressed and the predominate PPAR isoform expressed in rodent skeletal muscle. In contrast to PPAR ␣ ( 65 ), forced muscle-specifi c expression of PPAR ␦ improves endurance in mice and promotes a shift from glycolytic to oxidative muscle fi bers ( 66,67 ). Oxidative muscle tissues also contain higher levels of perilipin5 [see ( 7 ) and Fig. 1 ]. The direct regulation of Plin5 by PPAR ␦ may therefore explain the uneven distribution of Plin5 expressed in muscle fi bers. Other factors found to drive formation of oxidative muscles includes PGC-1 ␣ ( 68 ), PGC-1 ␤ ( 69 ), ERR ␣ ( 70 ), ERR ␥ ( 71 ), and the corepressor NCoR1 ( 72 ). Overexpression of PGC-1 ␣ in muscle may stimulate Plin5 expression ( 73 ), but the importance of the other factors have not been studied. Nevertheless, it is clear that Plin5 belongs to the pool of genes that distinguish oxidative type I from glycolytic type II fi bers.
The function of genes enriched in oxidative fi bers is primarily linked to FA oxidation, myosin fi ber types, mitochondrial biogenesis, and increased mitochondrial oxidative capacity (66)(67)(68)(69)(70)(71)(72). The role of perilipin5 in this setting is not clear. Due to low endogenous perilipin5 expression in cultured cells, much of our molecular understanding of the protein is based on cellular studies with ectopic perilipin5 expression. When overexpressed, perilipin5 recruits Abhd5 ( 24 ), ATGL ( 25,26 ), and HSL ( 27 ) to the surface of LDs, and promotes association of LDs and mitochondria ( 31 ). It remains to be elucidated if a portion of these effects are caused by high ectopic expression. Although it is unclear how interactions between perilipin5 and the above mentioned proteins regulate lipolysis, accumulating evidence from mice demonstrate that perilipin5 preserves LDs. Per-ilipin5-null mice lack LDs in the heart ( 30 ) whereas cardiacspecifi c perilipin5 transgenic mice accumulate LDs ( 74,75 ). Alteration in cardiac perilipin5 primarily affects TAG storage, which would be expected based on the preferred binding of perilipin5 to LDs fi lled with TAG ( 47 ). The role of perilipin4 is less clear. PPAR ␥ stimulates perilipin4 expression in adipocytes ( 19 ), which points to a role in energy storage. Perilipin4 may be recruited to nascent LDs formed in cultured adipose cells exposed to high concentrations of FAs ( 14 ). A more recent publication demonstrates that perilipin4 preferentially binds to LDs fi lled with CEs with an ability to enhance accumulation of such LDs when ectopically expressed ( 47 ).
Additional functional analyses are required to fully understand the function of Plin4 and Plin5 in LD metabolism and their roles in muscle physiology. The regulation of Plin5 by PPAR ␦ may provide insight into why lipid stores in muscles are benefi cial and detrimental in athletic and obese subjects, respectively. PPAR ␦ -mediated regulation of Plin4 and Plin5 may render the muscle tissue better equipped to fi ne-tune LD metabolism by having increased levels of these LD binding proteins. When present at the surface of LDs, they may help to preserve LDs and increase the cellular capacity to prevent lipotoxicity.