LSDP5 is a PAT protein specifically expressed in fatty acid oxidizing tissues

Biochimica et Biophysica Acta 1771 (2007) 210 – 227 www.elsevier.com/locate/bbalip

LSDP5 is a PAT protein specifically expressed in fatty acid oxidizing tissues Knut Tomas Dalen a,b,⁎, Tuva Dahl a , Elin Holter a , Borghild Arntsen a , Constantine Londos b , Carole Sztalryd c , Hilde I. Nebb a,⁎ a

c

Department of Nutrition, Institute of Basic Medical Sciences, University of Oslo, P.O. Box 1046 Blindern, N-0316 Oslo, Norway b Laboratory of Cellular and Developmental Biology, NIDDK, National Institutes of Health, Bethesda, MD, 20892-8028, USA Geriatric research, Education and Clinical Center, Baltimore Veterans Affairs Health Care Center, Department of Medicine, School of Medicine, University of Maryland, Baltimore, MD, 21201, USA Received 19 October 2006; received in revised form 25 November 2006; accepted 28 November 2006 Available online 8 December 2006

Abstract The PAT family (originally named for Perilipin, ADFP and TIP47) now includes four members: Perilipins, ADFP, TIP47 and S3-12. Significant primary sequence homology and the ability to associate with lipid storage droplets (LSDs) are well conserved within this family and across species. In this study, we have characterized a novel PAT protein, lipid storage droplet protein 5 (LSDP5) of 463 residues. A detailed sequence analysis of all murine PAT proteins reveals that LSDP5, TIP47 and ADFP share the highest order of sequence similarity, whereas perilipin and S3-12 have more divergent carboxyl- and amino-termini, respectively. Ectopically-expressed YFP-LSDP5 or flag-LSDP5 fusion proteins associate with LSDs. In accord with recent published data for perilipin, forced expression of LSDP5 in CHO cells inhibits lipolysis of intracellular LSDs. The LSDP5 gene is primarily transcribed in cells that actively oxidize fatty acids, such as heart, red muscle and liver. Expression of LSDP5 is stimulated by ligand activation of peroxisomal proliferator-activated receptor alpha (PPARα), and significantly reduced in liver and heart in the absence of this transcription factor. PPARα is generally required for regulation of fatty acid metabolism during fasting, but fasting induces LSDP5 mRNA in liver even in the absence of PPARα. © 2006 Elsevier B.V. All rights reserved. Keywords: LSDP5; Adipophilin; ADFP; S3-12; Perilipin; TIP47; PAT; TAG; Fatty acid; PPAR

1. Introduction Most mammalian cells are able to store triacylglycerols (TAG)1, cholesterol esters or other lipids in relatively small (< 1 μm diameter) lipid storage droplets (LSDs) which can be used as an energy source or for membrane biogenesis [1]. For

Abbreviations: ADFP, adipose differentiation-related protein; ATGL, adipocyte triaclyl glycerol lipase; CMC, carboxymethyl-cellulose; FA, fatty acid; HSL, hormone-sensitive lipase; LSD, lipid storage droplets; LSDP5, lipid storage droplet protein 5; OA, oleic acid; PPAR, peroxisomal proliferatoractivated receptor; PPRE, peroxisomal proliferator response element; TIP47, tail-interacting protein of 47 kDa; TAG, triacylglycerol; WAT, white adipose tissue ⁎ Corresponding authors. Knut Tomas Dalen is to be contacted at Laboratory of Cellular and Developmental Biology, NIDDK, National Institutes of Health, Bethesda, MD, 20892-8028, USA. Fax: +1 301 496 5239. Hilde Irene Nebb, fax: +47 22851398. E-mail addresses: [email protected] (K.T. Dalen), [email protected] (H.I. Nebb). 1388-1981/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2006.11.011

decades, these LSDs have been viewed as simple lipids reservoirs, but this view has changed mainly due to the discovery of a family of structurally related LSD binding proteins. Perilipins [2–4] were the first proteins that were experimentally demonstrated to associate with the LSD surface. Soon thereafter, the previously cloned mouse adipose differentiation-related protein (ADFP) [5] was found on LSDs in many cells [6] and tissues [7]. Two other proteins, tailinteracting protein of 47 kDa (TIP47)/placental tissue protein 17 (pp17) and S3-12 were subsequently cloned [8–10] and reported to bind to LSDs [11,12]. Structurally and functionally conserved family related proteins are also found in nonmammalian species such as Drosophila melanogaster and Dictyostelium discoideum [13]. Proteomics studies of LSDs from various cells demonstrate that PAT proteins (named after Perilipin, ADFP and TIP47) are among the most abundant proteins on the surface of LSDs [14–17]. Perilipin, ADFP and TIP47 exhibit high sequence identity within an amino-terminal PAT-1 domain and weaker homology

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

in the central and carboxyl-terminal PAT-2 domain [13,18]. The amino-terminal segment of S3-12 shares limited identity to the PAT-1 domain, but the remaining protein shares significant sequence homology to ADFP and TIP47 in the carboxylterminus [10,19] and is considered as a peripheral member [13,18]. All of the above proteins contain putative 11-mer helical repeats in the central sequence, which are also found in other lipid associated proteins such as synucleins, apolipoproteins, phosphate cytidyltransferases and dehydrins [20]. Although yet to be experimentally proven, it is likely that LSD targeting of PAT proteins is facilitated by these amphipathic helical repeats. The tissue distribution of the PAT proteins is well characterized. TIP47 and ADFP are both expressed ubiquitously. Whereas TIP47 mRNA is expressed at similar levels in most tissues examined [19,21], ADFP is transcriptionally regulated by fatty acids (FAs) [22] and fasting [21], leading to mRNA enrichment in FA metabolizing organs [6,19]. Expression of perilipin is largely confined to adipose and steroidogenic cells [2,3,23]. S3-12 protein expression is likely restricted to adipose cells [10,12], even though high levels of S3-12 mRNA has been found in skeletal muscle and heart [19]. To date, the function of only one of the PAT proteins, perilipin, has been firmly established. The generation of perilipin null mice provided a strong basis for the functional studies. Disruption of the Plin gene results in a lean mice with a 70% decreased adipose tissue mass [24,25], largely due to changes in the behavior of its adipocytes. A lack of perilipin surrounding adipose LSDs leads to a constitutively high basal lipolysis, but also to a loss in the ability to respond to lipolytic stimuli. The control of basal versus stimulated lipolysis is normally controlled by the phosphorylation state of the perilipin protein [25–30], and the defect in response upon lipolytic stimuli in Plin null mice reflects the inability of lipases, such as hormone-sensitive lipase (HSL), to bind to the LSD [27]. To date, only fragmentary functional knowledge has been reported for the other PAT proteins. The large majority of publications report only detection of expression or changes in the expression of these genes in various cell types. Some recently emerging data using cultured cells show that ADFP protects LSDs from degradation [31], implying a more general role for the PAT proteins in inhibition of lipolysis. Expression of ADFP is important in liver, being a major LSD binding protein in mice during fasting [21]. A modestly lower hepatic TAG content in Adfp null mice is so far the only phenotype discovered by disruption of the Adfp gene [32]. Little is known about the S312 and TIP47 proteins in lipid metabolism, except for their known targeting to LSDs in cells cultured in the presence of FAs [11–13]. The transcriptional regulation of the PAT genes suggests that they are tightly linked to fatty acid metabolism. Several of the PAT genes are transcriptionally regulated by members of the peroxisomal proliferator-activated receptors (PPARs). The PPAR family consists of three isotypes PPARα, PPARβ/δ and PPARγ, that belong to a subfamily of nuclear receptors that heterodimerize with retinoid X receptors (RXRs) and regulate transcription by binding to specific PPAR response elements

211

(PPREs) in the promoter region of target genes [33]. The PPARs are expressed in a tissue-specific manner: PPARγ is highly enriched in white adipose tissue (WAT) and macrophages [34], PPARα in liver and fatty acid metabolizing tissues, such as muscle, heart and kidney [35], whereas PPARβ/δ is more ubiquitously expressed [36]. Expression of the PPARs generally correlates well with the tissue expression profile of the PAT proteins. S3-12 [19] and perilipin [19,37–39] are regulated by PPARγ, ADFP by PPARα [21,40,41] and PPARβ/δ [42–44], whereas TIP47 seems not to be regulated by PPARs [19,21]. In this report we describe a fifth and novel member of the PAT family, with highest sequence similarity to TIP47 and ADFP. Like the other PAT proteins, LSDP5 binds to the surface of LSDs and protects them from lipolytic degradation. This novel PAT member is transcriptionally regulated by PPARα, and mainly expressed in fatty acid oxidizing cells such as heart, red muscle and liver. The transcriptional regulation of the Lsdp5 gene is similar to the regulation of the Adfp gene. These two PAT proteins are co-regulated in heart and liver upon physiological changes such as fasting and re-feeding. During our preparation of this manuscript, another research group published analysis of the similar protein, designated as MLDP of 448 residues [45]. In this paper we compare and contrast our analysis of the larger LSDP5 protein of 463 residues with those in the concurrent independent study. 2. Experimental procedures 2.1. Materials Restriction enzymes were obtained from Promega (Madison, Wisconsin). Cell culture reagents, 4-chloro-6-(2,3-xylidino)-2-pyrimidinylthioacetic acid (WY-14643), Oil red O, forskolin, isobutylmethylxanthine (IBMX), oligonucleotides and chemicals were purchased from Sigma (St. Louis, Missouri). All cell culture plasticware was obtained from Corning incorporated (Corning, New York).

2.2. Identification, cloning and genomic analysis of Lsdp5 and the remaining PAT genes A partial mouse LSDP5 protein sequence was identified using the highly conserved carboxyl-terminal motif found in the PAT proteins S3-12, ADFP and TIP47 [19,46] using protein-protein BLAST (Matrix: PAM30) [47]. The protein sequence was used in a further BLAST search using tblastn. These BLAST hits were used in nucleotide-nucleotide BLAST searches to obtain a full-length mouse LSDP5 cDNA sequence. The human and rat LSDP5 cDNA sequences were identified by BLAST using the mouse LSDP5 sequence. Mouse and human LSDP5 cDNAs were cloned with RT-PCR using Omniscript RT kit (Qiagen, Valencia, CA) from mouse liver total RNA (C57BL/ 6J-strain) and human heart mRNA (Clontech, Mountain View, CA, #636532), followed by PCR amplification using PfuUltra (Stratagene, La Jolla, CA) with PCR settings as described [48]. Primers used are listed in Table 1. The amplified PCR products were cloned into pPCR-Script (Stratagene), sequenced (Macrogen, Korea) and found to correspond 100% to the predicted cDNA sequences. Nucleotide–nucleotide BLAST against the genome sequence databases were used to determine the chromosomal location of LSDP5 in the human, mouse and rat genomes. The murine genomic organization of M6prbp1, Lsdp5 and S3-12 on chromosome 17 was determined by analyzing a BAC clone (bMQ217-F16 [49]) containing these PAT genes. Briefly, a successful gap-repair subcloning of individual PAT genes, as well as a joint subcloning of the Lsdp5 and S3-12 genes, confirms the presence of all PAT genes, as well as co-localization of the Lsdp5 and S3-12 genes within this single BAC clone. To determine the distance

212

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

Table 1 PCR primers used in this study Primer name pSG5 expression vectors 5 m LSDP5 CDS (aa1–463) 5-m-LSDP5-CDS (aa16–463) 3-m-LSDP5-CDS 5-h-LSDP5-CDS 3-h-LSDP5-CDS 5-m-perilipin-CDS 3-m-perilipin-CDS 5-m-S3-12-CDS 3-m-S3-12-CDS 5-h-ADFP-CDS 3-h-ADFP-CDS pEYFP-c1 fusion vectors 5-m-LSDP5-EYFP-C1 (aa1–463) 5-m-LSDP5-EYFP-C1 (aa16–463) 3-m-LSDP5-EYFP-C1 5-m-ADFP-EYFP-C1 3-m-ADFP-5-EYFP-C1 5-m-perilipin-EYFP-C1 3-m-perilipin-EYFP-C1 5-m-TIP47-EYFP-C1 3-m-TIP47-EYFP-C1 Flag expression vector 5-m-LSDP5-Flag (EcoRI) 3-m-LSDP5-Flag (EcoRI) Probes 5-m-LSDP5-probe 3-m-LSDP5-probe 5-h-LSDP5-probe 3-h-LSDP5-probe Quantitative RT-PCR Cyclophilin B (#NM_011149) forward Cyclophilin B reverse ADFP (#NM_007408) forward ADFP reverse LSDP5 forward LSDP5 reverse

Primer sequence 5′-TAAGATCTTGCACCCAGGGATCTGATTCTC-3′ 5′-TAAGATCTACACAGCAGAATGTCCGGTGA-3′ 5′-TAAGATCTAGGGGGCAGTAGAGACCTCGATA-3′ 5′-TAAGATCTAGGTGACCCTGTTTGCAGCAC-3′ 5′-TAAGATCTGAGTTGGGCCTGATTCCAAAGA-3′ 5′-TAAGATCTTGCTTTGCAGCGTGGAGAGTAAG-3′ 5′-TAAGATCTTAAAGGAAAGGCCCTTGACGAGA-3′ 5′-TAGAATTCTTACCCTGACTGGTCCAGGTGACT-3′ 5′-TAGGATCCGGACGCGTGATGCTTCTTTACTCT-3′ 5′-TAGAATTCTGCAGTCCGTCGATTTCTTTCTC-3′ 5′-TAGAATTCTGCACTAGTGATAGGGGCAGGTT-3′ 5′-TAGAGCTCAGGAAATGGACCAGAGAGGTGAAG-3′ 5′-TAGAGCTCGAATGTCCGGTGATCAGACAGCT-3′ 5′-TAGGTACCCCTCGATAGTCAGAAGTCCAGCTC-3′ 5′-TAGGTACCAAAATGGGAGCAGCAGTAGTGGAT-3′ 5′-TAGGTACCAGGAGGGGTTTACTGAGCTTTGAC-3′ 5′-TAGAGCTCGGATGTCAATGAACAAGGGCC-3′ 5′-TAGGTACCGCAGTCTGCTCAGCTCTTCTTGC-3′ 5′-TAGAGCTCCCATGTCTAGCAATGGTACAGAT-3′ 5′-TAGGTACCTCCCTACTTCCCTTCAGGGGTTT-3′ 5′-TTTGAATTCATGTCCGGTGATCAGACAGCT-3′ 5′-AAAGAATTCTCAGAAGTCCAGCTCTGGCAT-3′ 5′-TGCAGGGGACTAGACAAATTGGA-3′ 5′-CTCCAGGGTCAAACTTCGAGACA-3′ 5′-TTTGCAGCACGATGTCTGAAGAAG-3′ 5′-AAGTAGCCCTGCTGTCTCCTCTGA-3′ 5′-CAAGCTGAAGCACTACG-3′ 5′-AGGCCGTTCTAGCTTC-3′ 5′-CTACGACGACACCCAT-3′ 5′-CATTGCGGAATACGGAG-3′ 5′-AGGGGACTAGACAAATTGG-3′ 5′-GCTTCTCCGACTTGCC-3′

List of primers used to clone mouse (m) and human (h) PAT sequences into pSG5 (Stratagene), pEYFP-c1 (Clontech) and pcDNA3-Flag [51] vectors, and primers used for Quantitative RT-PCR analysis. Primers used to clone the remaining PAT proteins have been described previously [19]. between the M6prbp1 and Lsdp5 genes, two unique restriction sites were inserted in the M6prbp1 and Lsdp5 genes. Following digestion, the obtained fragments were separated with pulse shift gel electrophoresis, which estimated the distance between the M6prbp1 and Lsdp5 genes to approximately 160 kb, in agreement with the murine genome reference sequence.

2.3. Sequence comparison of the PAT proteins Sequence identity and similarity among the PAT proteins were determined by Blast2 [50] and clustalW alignment (using the emma interface in EMBOSS explorer; http://embossgui.sourceforge.net). Each PAT protein was aligned individually against LSDP5 to determine sequence homology to LSDP5. Accession numbers for the mouse PAT proteins used for comparison were: S312 (#NP_065593), perilipin (#AAN77870), ADFP (#NP_031434) and TIP47 (#BAB28291). The amino terminus of S3-12 (exons 1–4), the consensus of the 33-residue motif repeated 29 times (exon 5 and 6), and the carboxyl-terminus (exons 7–9) were aligned separately.

2.4. Generation of expression vectors and YFP and Flag fusion proteins The mouse perilipin and S3-12 cDNAs were cloned from mouse adipose tissue mRNA as described for the cloning of the LSDP5 cDNAs (primers used,

Table 1). The resultant cDNAs were subsequently excised from the pPCR-Script vector and cloned into the pSG5 expression vector (Stratagene). The cloning of the other PAT cDNAs has been described previously [19]. To generate YFP fusion vectors, PAT cDNAs were PCR amplified using PfuUltra (Stratagene) and cloned in frame with the pEYFP-C1 vector (Clontech). The pPCR-Script or pSG5 vectors containing full-length PAT cDNAs were used as templates (primers used, Table 1). Cloning of Flag-LSDP5 (aa16-463, PAT1) was performed using pSG5LSDP5 as template. The PCR-product was first cloned into pPCR-Script (Stratagene) followed by EcoRI digestion for recloning into pcDNA3-Flag [51].

2.5. Culturing of cells, fluorescence and lipolysis studies Cos-1 cells (ATCC) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum, 2 mM L-glutamine, penicillin (50 U/ml) and streptomycin (50 μg/ml). CHO cells were cultured in F-12 Nutrient Mix Ham supplemented with 10% fetal calf serum, L-glutamine and antibiotics. Cells were grown at 37 °C in 5% CO2. Cos-1 cells were seeded on glass coverslips in 6 well dishes (2 × 105 cells/ well), using antibiotic free medium the day before transfection. The cells were transfected for 4 h with 1 μg DNA complexed to 2 μl Lipofectamine2000 (Invitrogen, Carlsbad, CA) in serum-free DMEM. After transfection, cells were incubated for 24 h in regular growth medium supplemented with BSA or oleic

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227 acid (OA; 100 μM) complexed to fat-free BSA (OA:BSA; ratio 2.5:1) [21]. Cells were washed 3 times with 1× PBS and fixed with 3% paraformaldehyde and 0.025% glutaraldehyde for 10 min before permeabilization with 0.2% Triton-X-100 in PBS for 5 min at room temperature. The cells were blocked with 5% goat serum (Jackson Immuno-Research Laboratories, Inc., West Grove, PA) in PBS/BSA (10 mg/ml) before incubation with polyclonal rabbit Flag antibody (Sigma #F7425; 1:1000) for 1 h at room temperature. After removal of anti-Flag by washing three times 5 min in PBS/BSA, cells were treated with Alexa Fluor 568 goat anti-rabbit (Molecular Probe #A-11036, 1:500) for 1 h at room temperature. To visualize neutral lipid droplets, Bodipy 493/503 (Molecular Probes #D-3922) was dissolved in ethanol at 1 mg/ml and added to the secondary antibody solution to a final concentration of 20 μg/ml. Cells were washed six times for 2 min in PBS before mounting on microscope slides using Fluorosave (Calbiochem #345789, La Jolla, CA). Images were obtained using TCS SP multiband confocal imaging system (Leica, Microsystems GmbH, Wetzlar Germany) using a 100× objective. CHO cells were seeded on 35 mm glass bottom culture dishes (MatTek corporation, Ashland, MA, #P35G-1.5-14-C) and transfected for 5 h using 1.2 μg DNA complexed to 5 μl Lipofectamine2000 (Invitrogen). After transfection, cells were incubated in the presence of BSA or OA-BSA (400 μM) for 19 h, followed by 3 h incubation in the presence of BSA or OA-BSA containing Bodipy 558/568 C12 (Molecular Probe #A-11036, 20 μg/ ml). Cells were washed 3 times in 1× PBS to remove unincorporated stain, and incubated in regular growth medium during imaging. Images were obtained using LSM 510 Meta confocal imaging system (Zeiss, Germany) using a 100× objective. LSDP5-aa16–463 and Perilipin A cDNAs were cloned into the pcDNA3.1 (+)-vector (Invitrogen) and transfected into CHO-K1 cells using Lipofectamine 2000 (Invitrogen). Stable transfectants were obtained by selection with G-418 (600 μg/ml), and clones with high expression were selected for further studies. Lipolysis experiments were performed as described previously [27,28]. Briefly, cells were plated in 24 well dishes (105 cells/well) and loaded with OA-BSA (400 μM) for 24 h to promote LSD formation. [3H]OA (0.4 μCi/well) was added as a tracer. Under these loading conditions, the vast majority of OA is incorporated into TAG in intracellular LSDs, and all OA released upon lipolytic stimulation originates from this TAG [28]. Cells were washed in PBS containing 4% BSA, followed by incubation in efflux medium during the lipolysis experiment. Efflux medium contained 2.5 μM of the acyl-CoA synthethase inhibitor Triacsin C (Biomol, Plymouth Meeting, PA), to prevent reesterification of OA, and 1% fat-free BSA, as a FA acceptor. The efflux of [3H]OA into medium was determined by scintillation counting. Basal lipolysis was measured in the absence of further additions, whereas stimulated lipolysis was measured upon addition of 10 μM forskolin and 1 mM IBMX.

2.6. Generation of LSDP5 antibody, protein measurements and Western blot analysis The antibodies that recognize the LSDP5 were raised in rabbit and affinity purified against the LSDP5 peptide KVAEVQRSVDALQ (Rockland Immunochemicals for Research, Gilbertsville, PA). The two LSDP5 protein isoforms were ectopically expressed in CHO cells seeded in 12 well dishes. Cells were transfected for 6 h with 1.2 μl DNA complexed to 5 μl of Lipofectamine 2000 in OptiMem (Invitrogen, Carlsbad, CA). After transfection, cells were incubated for 24 h in regular growth medium supplemented with OA-BSA (100 μM) [21]. Cells were harvested in lysis buffer (1× PBS, 1% NP- 40, 0.1% SDS and 1× Complete Protease Inhibitor (Boehringer Mannheim)). Frozen mouse liver, heart and muscle tissues were homogenized in lysis buffer using a UltraTurrax® T8 (IKA Labortechnik, Staufen Germany) for 20–30 s. Protein concentrations were measured with the BC Assay method (Interchim, France, #FT-40840). Proteins from whole cell extracts were separated by a 10% Tris–HCl SDSPAGE (Criterion™ Precast Gel, Bio-Rad) and transferred to a PVDF membrane (Hybond-P, Amersham Biosciences) by electro transfer [21]. Membranes were incubated with rabbit anti LSDP5 (0,5 μg/ml) or rabbit anti-flag (Sigma #F7425; 1:5000) followed by incubation with secondary antibody (horseradish peroxidase conjugated goat anti-rabbit antibody (Southern Biotech, Birmingham, Alabama; #4050-05; 1:8000)). Bound antibody was detected using enhanced chemiluminescence (ECLplus, Amersham Biosciences) and visua-

213

lized with Hyperfilm MP (Amersham Biosciences). Membranes were stripped and incubated with primary antibody against β-actin (diluted 1:10,000, Sigma, #A5441). To determine the size of the LSDP5 protein, proteins from transfected CHO cells and heart were separated on a NuPAGE 4–12% Bis–Tris Gel (Invitrogen) at 150 V for 5 h. Proteins were transferred to Nitrocellulose and incubated with primary (LSDP5, 0.5 μg/ml) and secondary (goat polyclonal to rabbit IgG H and R, (HRP), 1:10,000 (abcam, Cambridge, MA) antibodies. Bound antibodies were detected using Supersignal West Pico Chemiluminescent substrate (Pierce, Rockford, IL).

2.7. Preparation and analysis of RNA 2.7.1. Northern analysis Total RNA from mouse tissues was extracted with TRIZOL®Reagent (Invitrogen). Total RNA (10 μg) was separated on a 1% agarose formaldehyde/ MOPS gel, wet blotted onto Hybond-N membrane (Amersham Biosciences) using SSPE as described [48]. Hybridization and stripping of membranes, human 12 lane MTN-blot (Clontech, #7780-1) and Mouse Message MAP RNA (Stratagene, #775900-12) were performed as recommended (Clontech, #PT1200-1). Probes were generated by radio labeling of cDNAs with [α-32P] dCTP (Perkin Elmer, Wellesley, MA) with the of Megaprime DNA labelling System (Amersham Biosciences). Northern blots were visualized by Hyperfilm MP, scanned with Personal Densitometer SI and analyzed using ImageQuant™ software (Amersham Biosciences). Templates for labeling of probes were generated with PCR (primers, Table 1). The other cDNAs used have been described previously [19]. 2.7.2. Quantitative RT-PCR analysis Total RNA from liver and muscle was isolated using TRIZOL®Reagent. Total RNA (2 μg) was reverse transcribed into cDNA using the Omiscript RT kit (Qiagen). Heart tissue (10 mg) was lysed and homogenized in Lysis Buffer using an Ultra Turrax homogenizing Instrument (IKA Labortechnik, Staufen, Germany) prior to mRNA isolation using the MagNA Pure LC mRNA Isolation Kit II and the MagNA Pure LC Instrument (Roche, Germany). Heart mRNA was reverse transcribed using the Transcriptor First Strand cDNA Synthesis Kit (Roche). Quantitative RT-PCR was performed on a LightCycler 2.0 instrument system using LightCycler FastStart DNA Master SYBR Green I (Roche). Cycling conditions were: Initial denaturation at 95 °C for 10 min, followed by 45 cycles at 95 °C for 10 second, annealing at 55 °C for 5 s and elongation at 72 °C for 12 s (Primers, Table 1). Relative expression was calculated against Cyclophilin B as a reference gene using the LightCycler Relative Quantification Software (Roche).

2.8. Animal experiments All animal use was approved and registered by the Norwegian Animal Research authority. Mice were maintained in a temperature controlled (22 °C) facility with a strict 12-h light/dark cycle and fed ad libitum (ad lib), unless otherwise specified. The synthetic PPARα activator (WY-14643) was dissolved in vehicle (1% w/vol carboxymethyl-cellulose (CMC) in H2O) and 10 μl vehicle/g body weight was given to each mouse. All mice were euthanized by cervical dislocation and tissues were rapidly dissected, quickly frozen in liquid nitrogen and stored at − 80 °C until RNA extraction. For generation of the multi tissue RNA blot, C57BL/6J mice (15 weeks) were fed ad lib (controls) or fasted for 12 h. Both groups were killed at the end of the dark cycle. The following tissues were dissected: white muscle (gastrocnemius), red muscle (soleus), heart (whole), spleen, kidney, liver and WAT (epididymal). For western analysis, mice were fed ad lib, fasted for 24 h or fasted 24 h and re-fed for 12 h. The Ppara−/− mice (Sv/129 (Jae Substrain)) were obtained from The Jackson Laboratory (Bar Harbor, Maine US) and have been described elsewhere [52]. 129/SvEvTac@Bom (Bomholtgård, Denmark) was used as wild type controls (Ppara+/+). In the 7 days feeding experiment, Ppara+/+ and Ppara−/− mice (12–18 weeks, 25–35 g) were fed daily with vehicle (CMC) or WY-14643 (10 mg/kg) with the last dose given 4 h before mice were euthanized. In the fasting re-feeding experiments, male Ppara+/+ (7–9 weeks) and Ppara−/− mice (6–8 weeks) were divided into three groups: nonfasted (control), fasted and re-

214

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

fed. The control group was fed ad lib, the fasted group was fasted 24 h, and the re-fed group was fasted for 24 h and then re-fed for 12 h. The mice were euthanized at the end of the dark cycle.

3. Results 3.1. Sequence comparison of human and rodent LSDP5 We had repeatedly detected an unknown transcript that weakly cross-hybridized with the TIP47 cDNA probe on human and mouse Northern blots (result not shown). In an attempt to identify this TIP47-related sequence, human, mouse and rat protein EST databases were searched with a carboxyl-terminally located motif highly conserved among S3–12, ADFP and TIP47 [19,46]. Several partial EST sequences with high sequence identity to the carboxyl-terminal motif (Fig. 1A) were found among mouse, human and rat sequences (not shown). These sequences were used in further searches to obtain full-length open reading frame sequences in order to clone this novel gene. The open reading frame is predicted to encode a protein of 463 residues in mouse, 463 residues in human and 475 residues in rat. The majority of these residues are conserved among these species (Fig. 1B). By being the fifth mammalian PAT protein identified, we suggest to rename this gene lipid storage droplet protein 5 (LSDP5), deposited in the gene bank with accession numbers: Mouse LSDP5: #DQ473305 and human LSDP5: #DQ839131. The mouse LSDP5 protein sequence is identical to the previously described PAT1 sequence of 448 residues ([13] #: NP_080150) and the recent published MLDP [45], except for the identification of a more amino-terminal translation start codon that predicts the protein to be larger than the initially deposited PAT1. This other translation start site is in frame with the start codon originally reported for PAT1, but adds 15 residues to the amino-terminus. These additional residues are predicted present in the rat, but not in the human LSDP5 protein. To determine which of the two translation start sites are used in vivo, a new antibody against mouse LSDP5 was developed (see Experimental procedures). The antibody's specificity was first confirmed by using protein extracted from Cos-1 cells transfected with pSG5-expression vectors containing full-length cDNA for all of the PAT genes incubated in the presence of BSA or OA complexed to BSA (OA-BSA; 100 μM). When performing western blotting using the LSDP5 antibody, a clear distinct signal with the expected size (∼ 54 kDa) was observed only from the cells transfected with pSG5-LSDP5 vector (Supplemental Fig. 1). As a positive control, Flag-LSDP5 protein was additionally applied on the gel. Re-probing the western blot with an antibody against the flag epitope, gave an identical protein band as the LSDP5 antibody (data not shown). The antibody was used to determine the size of the LSDP5 protein expressed in vivo. Expression vectors containing the long form of LSDP5 (LSDP5 aa1–463) and the short form lacking the 15 amino-terminal residues (LSDP5 aa16–463; PAT1 and MDLP) were transiently transfected into CHO cells incubated with OA-BSA (100 μM). Protein extracts from

transfected cells together with endogenous protein extracted from heart were subsequently separated with SDS-PAGE and the expressed LSDP5 proteins detected by Western analysis using the LSDP5 antibody. LSDP5 aa16–463 migrated at a lower molecular weight than both LSDP5 aa1–463 and endogenous LSDP5 expressed in heart (Fig. 1C). This suggests that the alternative start codon indeed is used, and that heart transcribe a 463 residues LSDP5 protein of ∼ 54 kDa. 3.2. The LSDP5 gene is encoded by nine exons and is located adjacent to S3–12 on chromosome 17 Mouse Lsdp5 is encoded by nine exons (Fig. 2A), with the typical gt-ag splicing acceptor/donor sites. The originally reported translation start site for the 448 residues PAT1 (and MLDP) protein starts with the second nucleotide in exon 3. Our newly identified translation start site is in exon 2, a more universal localization of a translational start site. The PAT-1 domain and the region containing putative α-helical 11-mer repeats are encoded by exons 4–5 and exon 6, respectively. The Plin and Adfp genes have been mapped to mouse chromosomes 7 [18] and 4 [53], respectively (Table 2). No genetic information has been reported for the remaining PAT genes, but the assembly of the murine genome suggests that the Lsdp5 gene is located on chromosome 17, adjacent to the S3-12 gene (Fig. 2B), whereas M6prbp1 (encodes for TIP47) is located 160 kb upstream of Lsdp5, separated from this PAT gene by eight unrelated genes. We have experimentally confirmed the murine genomic organization by analyzing a BAC clone containing these PAT proteins (see Experimental procedures). The same genomic organization is also found in the human and rat genomes, suggesting a linked evolution of these PAT genes. The transcribed LSDP5 mRNA is closely related to the S3–12 and TIP47 transcripts (up to 65% similarity between LSDP5 and TIP47 is found in one region, result not shown). The adjacent S3-12 and Lsdp5 genes are therefore likely to have evolved from gene duplication of an earlier version of the M6prbp1 gene. A non-transcribed Adfp pseudogene has also been found close to the Adfp gene [53]. 3.3. LSDP5 shows protein sequence identity to the PAT family A comparison of the mouse LSDP5 protein sequence with the other PAT proteins reveals highest sequence similarity to TIP47 and ADFP (55% and 51% over the entire protein sequence, respectively) (Fig. 3A and B). The homology to perilipin is high in the amino-terminal PAT-1 domain region [1,18] and in the 11-mer repeated region, but weakens rapidly in the remaining internal and amino-terminal regions (46% sequence similarity is found from residues 33–182 in LSDP5). The homology to S3–12 is restricted to the region after the PAT-1 domain (42% from residue 130 in LSDP5 corresponding to the sequence after residue 1146 in S3–12). S3–12 contains a 33-residues motif repeated 29 times [10], which is likely to fold into putative 11-mer repeats (underlined). Structurally similarly 11-mer repeats are identified in all PAT proteins, with a variable degree of residue similarity to the

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

215

Fig. 1. Identification of LSDP5 and protein sequence similarity among species. (A) The carboxyl-terminal motif used in BLAST search to identify LSDP5. The carboxyl-terminal motif is conserved in all PAT proteins, except for perilipin. Residues that are found to constitute a hydrophobic cleft in TIP47 [46] are marked with asterisks. (B) Sequence comparison of mouse, rat and human LSDP5 proteins. Similarity in each position is symbolized in the lower line. Identical residues (x), similar residues (:) and different residues (·). The mouse and human LSDP5 sequences have been verified by cloning and sequencing. The rat sequence is predicted from analyses of currently available EST sequences and gene assembly (accession number: #XM_576698). The cloned human LSDP5 sequence contained a T at position 964 generating a Trp at residue 306. A nearly equal amount of T and C (C generates Arg at this position) among available DNA sequences in the data base, suggests a single nucleotide polymorphism at this position (marked with S). (C) Protein extract from CHO cells transiently transfected with pSG5, pSG5-LSDP5-aa1–463, pSG5LSDP5-aa16–463 or from whole heart were separated by SDS-PAGE and LSDP5 detected by Western analysis using a LSDP5 specific antibody. The LSDP5 protein expressed in heart migrate with a molecular weight identical to protein translated from the pSG5-LSDP5-aa1–463 vector. The experiment was repeated three times with different dilutions of the proteins, and identical migration pattern was observed for all protein dilutions.

repeats found in S3–12. Although yet to be experimentally proven for the PAT proteins, similar amphipathic 11-mer helical repeats are suggested to facilitate binding of other types of lipid

binding proteins to lipid surfaces [20]. Except for the similarity in the 11-mer repeated region, residues from 1 to 1180 in S3-12 align poorly to the other PAT proteins, and residues that

216

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

Fig. 2. The mouse LSDP5 gene is located on chromosome 17 and consists of 9 exons. (A) Schematic structure of the Lsdp5 gene. The predicted translated sequence is shown in black. The PAT-1 domain is encoded by exon 4 to exon 5. The putative α-helical 11-mer repeats is encoded by exon 6, and the highly conserved hydrophobic pocked by exon 9. (B) The PAT genes M6prbp1 (encodes for TIP47), Lsdp5 and S3-12 are all located on mouse chromosome 17. Lsdp5 and S3-12 are adjacent and separated by only 1.8 kb. A 160 kb region, with 8 identified unrelated genes, separates the M6prbp1 and Lsdp5 genes. This genomic organization was verified by analyses of a BAC clone, bMQ-217-F16, containing all three PAT genes [49]. It is not determined if the region containing the tandem repeated 33-residues motifs in the S3-12 gene is encoded by one or two exons (marked with ?).

constitute the PAT-1 domain in much lesser conserved. The recently-identified crystal structure of the carboxyl-terminus of TIP47 revealed a clustering of residues that constitute a hydrophobic cleft [46]. Residues that constitute and surround this cleft are highly conserved in all PAT proteins, except for perilipin, which appears to have a unique carboxyl-terminus. A higher order of homology among the PAT genes Lsdp5, Adfp and M6prbp1 is also evident at the exon–intron boundaries. A similarity in the exon–intron boundaries has previously been shown in the amino-termini coding-regions of the Plin, Adfp and M6prbp1 genes [18]. A more extensive analysis including all PAT genes, demonstrates that Lsdp5, M6prbp1 and Adfp have identical exon–intron boundaries. By contrast, the carboxyl-terminal coding-region of Plin (from exon 7) is spliced differently than the carboxyl-terminal coding-

regions of the other four PAT genes (Fig. 3A). Similarly, the amino-terminal coding-region of S3-12 is spliced differently than the amino-termini of the other PAT genes. 3.4. LSDP5 binds to the surface of lipid storage droplets Given its primary sequence homology with other PAT proteins, we tested the ability of LSDP5 to associate with the LSD surface, a common feature of all known PAT proteins [3,6,11,12]. Flag-LSDP5 (LSDP5-aa16–463) was transiently transfected into Cos-1 cells incubated with BSA or OA-BSA (100 μM). Intracellular lipids and flag-LSDP5 were stained with Bodipy 493/503 or Alexa Fluor 568 secondary antibody, respectively, before signals were visualized using immunofluorescence microscopy. In the absence of OA-BSA, none or

Table 2 The genes in the PAT family Gene

Aliases

Chromosome

Location

Gene size (kb)

Exons

Transcript size (kb)

CDS

Residues

Ref

Plin

perilipin

Mouse 7

7 D2

11.5

9

ADRP adipophilin TIP47, pp17

Human 15 Mouse 4 Human 9 Mouse 17

15q26 4 38.9 cM 9p22.1 17 D

15.0 13.5 11.8 11.6

9 8 8 8

Human 19 Mouse 17 Human 19 Mouse 17

19p13.3 17 D 19p13.3 17 D

29.4 9.3 ND 5.9

8 9 (8)** ND 9

1551 1266 1044 735* 1569 1278 1314 1314 978 1305 4212 ND 1392

516 421 347 244 522 425 437 437 325 434 1403 ND 463

[3]

Adfp

PeriA-2.7 PeriB-3.6 PeriC-1.4 PeriD-2.0 2.9 1.9 1.9 2.2 1.9 2.2 5.4 ND 1.9

Human 19

19p13.3

12.9

8

1.9

1392

463

–

M6prbp1

S3-12 Lsdp5

KIAA1881 PAT1 MLDP

[67] [5] [7] – [8] [10] – –

Genetic information for the five characterized and identified PAT genes in mouse and human. The mouse Adfp gene has been mapped to chromosome 4 [53]. The characterization of the mouse Plin gene is obtained from [18]. Genetic information for the other PAT genes is obtained from human and mouse databases. The genomic localization of the PAT proteins on mouse chromosome 17 has been experimentally verified by restriction digestion and PCR characterization of BAC clone bMQ217F16 [49]. References for cloning of the PAT cDNAs are given to the right. ND, not determined. *, A periD protein has so far not been identified. **, The extraordinary nucleotide-invariability in the region encoding for the 33-residue repeats in S3-12 makes the assembly of this region uncertain. An alternative genomereference sequence suggests that the tandem repeats are encoded by one exon, with a predicted total number of eight exons in the S3-12 gene.

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

217

few intracellular LSDs were formed and the Flag-LSDP5 construct was visible throughout the cell (Fig. 4A). In the presence of OA-BSA, numerous LSDs were formed, and a distinct fluorescence signal surrounding the LSDs could be observed in cells expressing the Flag-LSDP5 (Fig. 4B). Fluorescence was not observed in cells expressing the Flag epitope alone, demonstrating that the LSDP5 sequence is essential to target the Flag-LSDP5 construct to the LSDs. To verify localization of LSDP5 to LSDs using another approach, YFP-LSDP5 fluorescence fusion proteins were used for direct immunofluorescence studies in living cells. YFPfusion constructs of LSDP5 and LSDP5-aa16–463 were transiently expressed in CHO cells incubated with BSA or OA-BSA (400 μM). LSDs were stained by incubating the cells with the fluorescent fatty acid analogue Bodipy 558/568 C12, which is incorporated into LSD TAG [54]. In the absence of OA-BSA, few LSDs were observed in the cells and the YFPLSDP5 (Fig. 5A), YFP-LSDP5-aa16–463, and YFP signals (not shown) were visible throughout the cell. As found for the Flag-LSDP5 construct, the LSDP5-YFP construct did not localize to a particular cellular compartment in the absence of LSDs. In cells incubated with OA-BSA, numerous LSDs surrounded by distinct ring-shaped YFP-LSDP5 or YFPLSDP5-aa16–463 fluorescence signals were observed, with little fluorescence elsewhere in the cells (Fig. 5B). No particular difference in the signals was observed for the two LSDP5 constructs, which suggests that the 15 amino-terminal residues are irrelevant for LSD association of the LSDP5 protein. YFPLSDP5 fluorescent signals were furthermore found to be similar to fluorescent signals obtained from TIP47, ADFP or perilipin YFP-fusion constructs in lipid loaded Cos-1 cells (Supplemental Fig. 2). Although less evident for the TIP47 fusion construct, all YFP-PAT constructs promoted a clustering of LSDs in one or a few larger pools. By contrast, LSDs were more randomly distributed in the cytosol in cells transfected with YFP (Fig. 5B) or untransfected cells (not shown). The confocal scanning microscope allowed us to visualize cross-sections of LSDs, demonstrating association of YFP-LSDP5 only on LSD surfaces. Altogether, these analyses demonstrate that ectopically expressed LSDP5 behaves similarly to the previouslycharacterized PAT proteins.

smaller (∼2 kb) transcript in the mouse tissues. Further analysis will be required to characterize these alternative transcriptional variants. It is, however, clear that these signals are not a result of cross-hybridization with any of the other PAT mRNAs, since the other PAT proteins have different tissue expression profiles [19]. The tissue selective expression pattern of LSDP5 was furthermore found to be highly similar to the tissue distribution of the fasting induced nuclear receptor PPARα (Fig. 6A and B).

3.5. Tissue expression of LSDP5 is highly restricted to heart, red muscle and liver

Previous characterization of the promoters of other PAT genes has established perilipin and S3-12 to be regulated by PPARγ [19,37–39] and ADFP to be regulated by PPARα [21,40,41] and PPARβ/δ [42–44]. The expression of LSDP5 correlates well with that of PPARα (Fig. 6), and a transcriptional regulation of LSDP5 by this PPAR family member might therefore be expected. To determine if LSDP5 is induced by activation of PPARα, wild type (Ppara+/+ ) and Ppara null mice (Ppara−/−) were fed with vehicle (CMC) or a potent synthetic PPARα activator (WY-14643; 10 mg/kg) daily for 1 week. Four hours after the last administration, the mice were euthanized and liver and heart tissues were subjected to quantitative RT-PCR analysis. A six-fold induction of LSDP5 mRNA was found in the liver of WY-14643 fed Ppara+/+ mice, but no regulation was observed in the Ppara−/− mice (Fig. 7A). Similar changes were

The so far characterized PAT genes are expressed in a tissue specific manner. In order to determine the tissue expression of LSDP5, commercial human and mouse multi tissue blots were probed with LSDP5 cDNA probes. LSDP5 was found to be expressed in skeletal muscle and liver, followed by weaker expression in heart and kidney among 12 different human tissues examined (Fig. 6A). Essentially the same distribution was observed in RNA samples from C57/BL mice, where liver, heart, and skeletal muscle showed significant expression level (Fig. 6B). In both species, the major transcript was about 2.5 kb. In addition, an alternative larger (∼5 kb) transcript was observed in human heart, and a

3.6. The expression of LSDP5 is induced by fasting The related PAT member, ADFP, is highly induced in liver by fasting [21], and it was therefore interesting to examine if the expression of LSDP5 is similarly regulated. Male C57/BL mice were fed ad libitum (ad lib) or fasted for 12 h before being euthanized. Various tissues were dissected and mRNA expression levels were determined by Northern analysis (Fig. 6C). The highest expression was found in heart, followed by red fiber muscle (soleus), liver and mixed fiber muscle (gastrocnemius). After 12 h of fasting, the LSDP5 mRNA content was increased in liver and, to a lesser extent, in the other tissues. A weaker signal, unchanged by fasting, was also observed in epididymal WAT. The LSDP5 protein expression was examined in liver, heart and soleus and gastrocnemius muscle. Male mice (C57/BL) were fed ad lib, fasted (24 h) or fasted (24 h) and re-fed (12 h) prior to determination of the LSDP5 protein content by Western analysis. The LSDP5 protein was highly increased in liver and weakly induced in heart after 24 h of fasting (Fig. 6D). In both tissues, the expression declined to baseline levels after 12 h of re-feeding. In soleus muscle, the LSDP5 protein was highly expressed, but seemed to be relatively unaffected by fasting or re-feeding. In agreement with the low mRNA expression level, the LSDP5 protein was weakly detected in gastrocnemius muscle. In all tissues, a non-specific protein band of smaller size (marked with an asterisk) was unaltered by the fasting or refeeding treatment. 3.7. The expression of LSDP5 in liver is induced by a PPARα activator

218

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

219

Fig. 4. LSDP5-flag associates with the lipid storage droplet surface. Cos-1 cells were transfected with pcDNA-Flag-vector or Flag-fused to cDNA coding for LSDP5aa16–463 and incubated with BSA or OA-BSA (100 μM) for 24 h. The cells were fixed with paraformaldehyde/glutaraldehyde prior to staining with anti-Flag together with Bodipy 493/503 for visualizing lipids. Cellular localization of LSDP5 was determined with immunofluorescence confocal laser microscopy. Left panels show immunofluorescent signal (red), middle panels show Bodipy staining (green), and right panels show merged images. (A) Expression of Flag-LSDP5 in cells incubated with BSA. (B) Expression of Flag and Flag-LSDP5 in cells incubated with OA-BSA.

observed for the PPARα responsive PAT member ADFP (seven-fold induction). The effect of WY-14643 feeding was much weaker in heart. A marginal (1.25-fold) induction of the LSDP5 and ADFP mRNAs was found (Fig. 7B), and it is arguable if this induction represents a biological relevant regulation. Interestingly, the basal expression level of LSDP5 in liver was strongly reduced (approximately 90% lower) in the Ppara null mice compared to the wild type mice. The same trend was observed in heart where the expression of LSDP5 mRNA was approximately 60% lower in Ppara null mice. By contrast, as reported previously [21], the expression

of ADFP was relatively unaffected by the removal of the Ppara gene. 3.8. The LSDP5 gene is induced during fasting independently of PPARα The ADFP gene is induced during fasting independently of a functional Ppara gene in liver [21]. Given the clear requirement for PPARα for basal expression of LSDP5, it was therefore interesting to determine if the same regulation applies to the LSDP5 gene. Ppara+/+ and Ppara−/− mice were fed ad lib

Fig. 3. LSDP5 sequence identity with other PAT members. (A) Protein sequence comparison of murine LSDP5, TIP47, ADFP, S3-12 and perilipin. The PAT-1 domain [marked with *] [1,18] is highly conserved in all PAT members except for S3-12. Hydrophobic residues that constitute and surrounds a hydrophobic cleft in TIP47 [46] are highly conserved for all PAT proteins, except for perilipin [positions marked with H]. All PAT proteins contain five putative 11-mer α-helical repeats with variable degree of sequence homology to the consensus of the 33-residues motif [insert is underlined] found in S3-12 [10,19]. The two less conserved residues in the 33residues repeat are shown with small characters [vx]. Residues that are identical or highly similar in more than three of the PAT proteins are shown in bold type. Exon– intron boundaries for each PAT gene are superimposed with shaded gray boxes of the first coded residue in each exon. (B) Schematic drawing of sequence identity (and similarity) of mouse LSDP5 against the other PAT members. Overall, LSDP5 is highly identical in sequence to TIP47 and ADFP (from residues 33 to 401) with 36 and 32% sequence identity (70 and 71% similarity), respectively. All PAT proteins have unique extreme amino- and carboxyl-termini of varying sizes. LSDP5 has considerably longer unique amino- and carboxyl-terminus compared to the mostly related PAT members TIP47 and ADFP.

220

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

Fig. 5. LSDP5-YFP selectively associates with the lipid storage droplet surface. CHO cells were transfected with pEYFP-C1-vector or YFP-fused to cDNA coding for LSDP5 (aa1–463) or LSDP5-aa16–463 and incubated with BSA or OA-BSA (400 μM) for 24 h. The cells were incubated in the presence of Bodipy 558/568 C12 in BSA or OA-BSA supplemented medium for 3 h before cells were imaged. Cellular localization of the YFP-LSDP5 proteins was determined with immunofluorescence confocal laser microscopy. Left panels show immunofluorescent signal (green), middle panels show Bodipy staining (red), and right panels show merged images. (A) Expression of YFP-LSDP5 in cells incubated with BSA. (B) Expression of YFP-LSDP5, YFP-LSDP5-aa16–463 and YFP in cells incubated with OA-BSA.

(control), fasted for 24 h (fasted) or fasted 24 h and re-fed for 12 h (re-fed). In liver, the LSDP5 mRNA was induced after fasting in Ppara+/+ mice (5.3-fold), and lowered to basal expression level after re-feeding for 12 h (Fig. 8A). A similar induction after fasting was also observed in Ppara−/− mice (4.9fold). However, due to the considerably lower basal expression of LSDP5 in Ppara−/− mice (90% lower), the LSDP5 mRNA expression after fasting was still lower than that found in ad lib fed Ppara+/+ mice. As reported previously [21], the expression

of ADFP was induced to similar levels during fasting in both Ppara+/+ and Ppara−/− mice (16- and 13-fold, respectively). Fasting also increased the LSDP5 protein in the Ppara+/+ mice, whereas the expression of the LSDP5 protein was below the detection limit in the Ppara−/− mice (Fig. 8B). The effect of fasting was weaker in heart, where the LSDP5 and ADFP mRNAs were induced 1.5 and 4.1-fold in Ppara+/+ mice, respectively (Fig. 8C). In Ppara−/− mice, the induction during fasting was approximately the same for LSDP5 (1.7-

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

221

Fig. 6. Tissue expression of LSDP5. (A) Expression of LSDP5 mRNA in human tissues. Expression of PPARα and RXRα are shown for comparison. (B) Expression of LSDP5 mRNA in mouse tissues (C57/BL strain). Expression of PPARα and RXRα are shown for comparison. For both blots, the most abundant expressed transcript (2.5 kb) is in agreement with the predicted mRNA transcript size. Additional uncharacterized transcripts are observed in human liver (∼ 5 kb) and in specific mice tissues (∼ 2 kb). (C) Expression of LSDP5 and PPARα mRNA in mouse tissues involved in fatty acid metabolism (C57/BL strain). Muscle gastrocnemius (Mgast.), muscle soleus (M-soleus), whole heart, spleen, kidney, liver and epididymal WAT (epi. WAT). Mice were fed ad lib or fasted for 12 h and killed at the end of the dark cycle. Each well contains equal amount of RNA (12 μg) pooled from three individual mice. The blot was probed with 36B4 as an internal control. 36B4 is expressed at lower levels in liver than the other tissues. (D) LSDP5 protein expression in mice fed ad lib, fasted for 24 h or fasted 24 h and re-fed 12 h (C57/BL strain). Each lane contains whole protein extract from liver (20 μg), heart (10 μg), muscle gastrocnemius (M-gast.; 50 μg) and muscle soleus (M-soleus; 50 μg) pooled from six mice. Flag-LSDP5 protein, isolated from transiently expressed Cos-1 cells, confirms the correct size for the detected LSDP5 protein. The experiment was repeated once, with similar results. The identity of the unspecific unregulated band (marked with an asterisk) is unknown.

fold) but lower for ADFP (2.5-fold). These results suggest that the fasting induced physiological regulation of the Lsdp5 gene occurs independent of a functional Ppara gene. These experiments show that a functional PPARα is important for basal expression of the LSDP5 gene in liver and heart, in marked contrast to its insignificant role for basal expression of the ADFP gene. 3.9. LSDP5 protects stored lipids from degradation Perilipin is known to be essential for hormonal regulation of lipolytic rate in adipose tissue. It inhibits lipolysis in its nonphosphorylated form and stimulates lipolysis when phosphorylated [25–30]. Recent experiments in cultured cells suggest that coating of LSDs with ADFP [31] or TIP47 [55] protects them from degradation, implying a more general role for PAT proteins in inhibition of lipolysis. To determine if LSDP5 prevents lipolysis of LSDs, we developed CHO cells that stably express LSDP5 or perilipin A and compared the lipolytic activity in these cell lines to wild type CHO cells. Wild type CHO cells coat LSDs with ADFP [6,28], and under basal conditions, the control cells rapidly hydrolyzed their LSD TAG and released OA to the medium (Fig. 9A). As reported previously [28], lipolysis was strongly suppressed, with a much lower efflux of OA into the medium, in CHO cells with perilipin-coated LSDs. Lipolysis was similarly suppressed in CHO cells with LSDP5coated LSDs. Upon lipolytic stimulation using isoprotenol and IBMX, lipolysis in the control cells was slightly inhibited, whereas lipolysis in cells expressing perilipin was stimulated

after a delay of ∼ 1 h (Fig. 9B). By contrast, like cells expressing PKA mutated forms of perilipin [28], cells expressing LSDP5 were unresponsive to stimulation. 4. Discussion Four genes have been characterized previously as members of the PAT family: Plin (perilipins), Adfp (adipose differentiation-related protein/adipophilin), M6prbp1 (pp17/TIP47) and the more peripherally-related S3-12 (S3-12) [1,18]. The novel PAT gene, Lsdp5, is clustered with M6prbp1 and S3-12 in the human, mouse and rat genomes. The linked evolution of these PAT genes is intriguing, and is also found for members of the closely related apolipoproteins, which are transcribed from apoA1/C3/A4/A5 and apoE/C1/C2 gene clusters [56,57]. Such clustering is found to facilitate transcriptional as well as functional interplay among the lipoproteins. With a variable degree of efficiency, activation of liver X receptors stimulates transcription of all genes in the apoE/C1/C2 gene cluster [58]. The apoA5 and apoC3 proteins have opposing effects on plasma triglyceride levels (reviewed in [59]). Additional studies will be required to clarify if clustering of the three PAT genes plays a role in the transcriptional regulation or function of these PAT genes. The functional knowledge on the PAT proteins is limited, but available experimental data suggest that all PAT proteins operate independently. The clearly different tissue distribution and transcriptional regulation of M6prbp1, S3-12 and Lsdp5 [10,19,45] and this report), does not suggest any transcriptional co-regulation of these PAT genes.

222

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

Fig. 7. Expression of LSDP5 and ADFP is induced by activation of PPARα. Male Ppara+/+ and Ppara−/− mice (Sv/129 strain) were gavaged with vehicle (CMC) or WY-14643 (10 mg/kg) for 1 week and euthanized 4 h after the last administration of ligand. Tissues were rapidly taken out and subjected to RNA isolation and quantitative RT-PCR analysis. (A) Relative mRNA expression of LSDP5 and ADFP in liver. (B) Relative mRNA expression of LSDP5 and ADFP in heart. The bargraphs show expression correlated against cyclophilinB. The results are presented as Mean ± SEM (n = 6 for each group). Statistical difference against controls (CMC) was determined using two-tailed student t-test (*P < 0.05; **P < 0.01; ***P < 0.001). Asterisks on Ppara−/− (CMC) represent difference relative to Ppara+/+ (CMC).

A shorter protein form of LSDP5 was previously discovered by its sequence homology to the PAT-1 domain and originally designated as PAT1 [13]. The recently published mouse MLDP protein [45] is identical to the deposited PAT1 protein of 448 residues ([13], #NP_080150). However, a more amino-terminal translation start site, present in rat and mouse LSDP5 mRNAs, predicts the mouse protein to consist of 463 residues. The use of this alternative translational start site is supported experimentally. Ectopically expressed LSDP5 aa1–463, but not LSDP5 aa16–463, migrates with similar molecular weight as LSDP5 protein endogenously expressed in heart (Fig. 1C) and brown adipose tissue (result not shown). The current report on MLDP presented a phylogenetic analysis based on full-length PAT protein sequences which predicted MLDP to be closer to perilipin than to TIP47 and ADFP [45]. Our detailed analysis of all PAT proteins reveals a different and more defined classification of the PAT protein family. The Lsdp5, M6prbp and Adfp genes have identical exon–intron boundaries, and encode for proteins that share a high order of sequence similarity throughout the majority of their protein sequences. By contrast, the carboxyl- and aminotermini coding-regions of the Plin and S3-12 genes, respectively, have different exon–intron boundaries which encode for more divergent residues. The lack of a PAT-1 domain in S3-12

and the carboxyl-terminal hydrophobic cleft in perilipin, make the putative 11-mer repeated regions the single motif entirely conserved among all PAT proteins. Extrapolation from studies on other lipid binding proteins [20], suggest that the 11-mer repeats facilitate binding of the PAT proteins to LSDs. Studies using truncated constructs support that the 11-mer regions are important for LSD binding of ADFP [54] and LSDP5/MLDP [45], whereas other carboxyl-terminal sequences might be more important for perilipin [60]. Thus far, no specific roles have been firmly assigned to the PAT-1 domain and the carboxyl-terminal hydrophobic cleft. It is believed that both the amino- and carboxyl-termini regions of perilipin are important for regulation of lipolysis [25–30]. The unique carboxyl- and amino-termini regions of perilipin and S3-12, respectively, suggest a more divergent evolution of these domains, which perhaps enables these domains to fold into more divergent structural units. Such a structural diversity may be a way for these PAT proteins to have a greater degree of variation in functional behavior and to execute unique biological functions. A signature feature of all previously characterized PAT proteins is their ability to bind to the surface of LSDs in cells cultured in the presence of FAs [6,11–13,61]. The same phenomenon was also found for LSDP5 when transfected into

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

223

Fig. 8. Expression of LSDP5 during fasting and re-feeding. Male Ppara+/+ and Ppara−/− mice (Sv/129 strain) were fed ad lib (control), fasted for 24 h (fasted) or fasted 24 h and re-fed for 12 h (re-fed) before being euthanized. Tissues were rapidly taken out and subjected to RNA isolation, quantitative RT-PCR analysis and western analysis. (A) Relative mRNA expression of LSDP5 and ADFP in liver. (B) Expression of the LSDP5 protein in liver. The LSDP5 protein band is marked with an arrow. The identity of the unspecific band (marked with an asterisk) is unknown. The blot was reprobed with β-actin to evaluate equal loading of protein. (C) Relative mRNA expression of LSDP5 and ADFP in heart. The bar-graphs show expression normalized to cyclophilinB. The results are presented as Mean ± SEM (Ppara+/+ mice, n = 5 for each group, Ppara−/− mice, n = 4 for each group). Statistical differences against controls were determined using two-tailed student t-test (*P < 0.05; **P < 0.01; ***P < 0.001). Asterisks on Ppara−/− (CMC) represent difference relative to Ppara+/+ (CMC).

COS-1 or CHO cells (this manuscript) or mouse Leydig cells [45]. LSDP5 functionally now joins other members of the PAT protein family, such as perilipin [28], ADFP [31] and TIP47 [55] in protecting TAG in LSDs from hydrolysis by lipases. Although our experiments were performed with the short form (LSDP5-aa16–463), it is unlikely that the full-length LSDP5 would behave differently, since no difference was observed in LSD binding among the PAT-1/MLDP and LSDP5 proteins. One lipase known to be associated with LSDs in CHO cells is adipocyte triaclyl glycerol lipase (ATGL) [62,63]. Further studies are required to address if ATGL is the operative lipase in these experiments. Nevertheless, the observation that PKA activation stimulates lipolysis in perilipin expressing CHO cells, suggests that phosphorylation of perilipin facilitates ATGL actions in these cells, similar to what has been reported

for HSL in other types of cells [27]. By contrast, the lack of any PKA sites in LSDP5 and ADFP render these proteins unresponsive to PKA-mediated lipolytic stimulation. Altogether, our structural and functional analyses of LSDP5, as well as studies performed by others [45], warrants classification of LSDP5 as a new member of the PAT family. Thus far, only two PAT proteins (LSD-1 and LSD-2) have been identified in Drosophila melanogaster [13]. By contrast, five PAT proteins have been identified in mammals, from which one may infer that coating of the LSD surface must be of considerable importance. The great variety of tissue specialization in the utilization of lipid stores in mammalian cells, likely dictate a need for the variety of LSD coating. All PAT proteins bind to LSDs, but they exhibit differences in their (1) tissue distribution, (2) apparent binding affinity to LSDs, (3) protein

224

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

Fig. 9. LSDP5 protects stored lipids from degradation. Wild type or CHO cells stably expressing LSDP5-aa16–463 or perilipin A were loaded with [3H]OA for 24 h, and the efflux of [3H]OA to the medium was tracked over 3 h. (A) Basal lipolysis. Efflux was measured in the absence of any additions. (B) Stimulated lipolysis. Efflux was measured from cells treated with 1 mM IBMX and 10 μM forskolin to promote activation of PKA. Each point is presented as Mean ± SEM (n = 8).

stability and (4) transcriptional regulation. One may assume that these differences evolved to fine tune fatty acid metabolism, and in particular formation of LSDs, to accommodate specificity in fatty acid metabolism in each type of cell. All tissues analyzed so far, are found to transcribe more than one PAT protein, but none of the PAT proteins show identical expression patterns [3,6,19]. The concurrent work found MLDP mRNA expression in heart liver and muscle, but MLDP protein expression only in heart [45]. Among the organs analyzed, we found LSDP5 mRNA and protein to be significantly expressed in heart, red muscle and liver. Examination of the Unigene and GeoProfiles databases suggests mRNA expression in additional tissues such as mammary gland and brown adipose tissue. A significant expression of the LSDP5 protein in brown adipose tissue has been confirmed experimentally (result not shown). Such data imply that LSDP5 expression is more restricted than the more ubiquitously expressed TIP47 and ADFP [6,19,21] and expressed in tissues with low expression of S3-12 and perilipin [2,23]. The concurrent work did not specify the type of muscle analyzed [45], which is essential as we show that LSDP5 protein is highly expressed in soleus (red muscle) but nearly undetectable in gastrocnemius (a mixed red and white muscle). Red muscle contains a higher amount of FA transporters and oxidizes more FAs as energy source than white muscle [64]. The heart, where LSDP5 is highly expressed, relies primarily on FAs as an energy source [65], and the liver is a major site of β-oxidation during fasting [66]. These data suggest a preferred expression of LSDP5 in cells and tissues that actively oxidize FAs, a broader expression pattern than originally suggested for MLDP [45]. All identified PAT proteins bind to LSDs [3,6,11,12], but they differ in LSD affinity and protein stability in the absence of LSDs. ADFP associates with LSDs within most cells examined [6], but is rarely observed in adipose cells that express perilipin [6,27], despite high ADFP mRNA expression [5]. However, in the absence of perilipin, ADFP is the major adipose LSD binding protein [25]. A putative hypothesis is that, perilipin prevents ADFP binding adipose LSD, leaving the highly unstable ADFP protein to be actively degraded by proteasomes [31]. This

suggests a hierarchy in binding affinity to the LSDs, where perilipin binds stronger to LSDs than ADFP. TIP47 protein stability is not dependent on the presence of LSDs, and only a fraction of the TIP47 protein is bound to LSDs ([11,13] and this work), suggesting it to be bind with low affinity to LSDs. The YFP-LSDP5 and flag-LSDP5 fusion proteins are easily observed in the cytosol in the absence of LSDs, but are primarily located on the surface of LSDs in cells incubated in the presence of FAs. In addition, our in vivo analyses demonstrate a consistent relationship between the LSDP5 mRNA expression level and LSDP5 protein content in cells. This suggests that the LSDP5 protein is stable in the absence of LSDs (similar to TIP47), but has a stronger affinity to LSDs (more like perilipin). In favor of the role of PAT proteins in lipid metabolism, most PAT genes are transcriptionally regulated by PPARs. S3-12 [19] and perilipin [19,37–39] are regulated by PPARγ, ADFP by PPARα [21,40,41] and PPARβ/δ [42–44], whereas TIP47 seems not to be regulated by PPARs [19,21]. The Lsdp5 gene is transcriptionally stimulated by PPARα activators and fasting ([45] and this manuscript), highly similar to the regulation of the Adfp gene. However, one intriguing difference is the highly reduced expression of LSDP5 in the absence of a functional Ppara gene. Although this reduction in expression might be due to differences in genetic background between the mice strains, only a marginal change in expression was observed for other well characterized PPARα target genes, such as ADFP, Malic enzyme 1 and acyl-coenzyme A oxidase-1 ([21] and current manuscript). The ADFP promoter contains a structurally and functionally conserved DR-1 element that binds PPARs [19,21,42]. Sequence comparison of the human, rat and mouse LSDP5 promoters has so far not revealed a conserved DR-1 type response element, and the murine promoter does not respond to PPARα activation in transfection studies (result not shown). It is therefore likely that the molecular mechanisms involved in the PPARα mediated regulation of the ADFP and LSDP5 genes are different. All PAT proteins, with the exception of S3-12, are now found to prevent lipolysis of LSDs. This function is most extensively studied for perilipin, which has a unique role in the regulation of

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

adipose lipolysis. The hormonal regulation of adipose lipolysis is tightly linked to the phosphorylation state of the perilipin protein [25–30], and this regulation is not restored by the elevated ADFP protein [25] nor by the highly expressed S3-12 or TIP47 proteins [10,19] in the lack of perilipin. The expression of LSDP5 is likely too low in adipose tissue to compensate for the abundantly expressed perilipin protein. A similarly unique role has yet to be found for the other PAT proteins. Unlike perilipin, the ADFP protein is highly expressed in a broad range of cell types [6,7], which suggests it plays a basic role in lipid metabolism. Still, the only phenotype detected in the ADFP null mice is a modest reduction in hepatic TAG content [32]. The lack of a more severe phenotype is thus far best explained by a functional compensation by TIP47 [55]. Based on the current functional data, and the high order of primary sequence similarity ([31,55] and this manuscript), it is likely that LSDP5 is able to functionally overlap with ADFP and TIP47 in their prevention of lipolysis of LSDs. It is therefore interesting to note that the expression of LSDP5 (determined by quantitative PCR analyses) and TIP47 [21] is considerably lower than ADFP in liver (result not shown). Hence, it might be that liver expression of these two PAT proteins is insufficient to fully compensate in the ADFP null mice, which generates a liver phenotype. By contrast, the similar transcriptional regulation of the ADFP and LSDP5 genes, suggests that in tissues with high co-expression, such as heart and muscle, LSDP5 together with TIP47 might compensate more efficiently in the absence of ADFP. A more comprehensive study on these PAT proteins using appropriate animal models is currently being performed to address this hypothesis. Acknowledgments The authors are grateful to Stine M Ulven and Sverre Holm for help with animal studies. This work was founded by grants from Institute of Medical Faculty at the University of Oslo, The Norwegian Research Council, the Novo Nordisk Foundation, Henning and Johan Throne-Holst's foundation, and partial by the Intramural Research Program of the NIDDK, National Institutes of Health. A part of this work was generously supported by a post doc travel grant from Hennig and Johan Throne-Holst's foundation to KTD. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbalip.2006.11.011. References [1] C. Londos, D.L. Brasaemle, C.J. Schultz, J.P. Segrest, A.R. Kimmel, Perilipins, ADRP, and other proteins that associate with intracellular neutral lipid droplets in animal cells, Semin. Cell Dev. Biol. 10 (1999) 51–58. [2] A.S. Greenberg, J.J. Egan, S.A. Wek, N.B. Garty, E.J. Blanchette-Mackie, C. Londos, Perilipin, a major hormonally regulated adipocyte-specific phosphoprotein associated with the periphery of lipid storage droplets, J. Biol. Chem. 266 (1991) 11341–11346.

225

[3] A.S. Greenberg, J.J. Egan, S.A. Wek, M.C. Moos Jr., C. Londos, A.R. Kimmel, Isolation of cDNAs for perilipins A and B: sequence and expression of lipid droplet-associated proteins of adipocytes, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 12035–12039. [4] E.J. Blanchette-Mackie, N.K. Dwyer, T. Barber, R.A. Coxey, T. Takeda, C.M. Rondinone, J.L. Theodorakis, A.S. Greenberg, C. Londos, Perilipin is located on the surface layer of intracellular lipid droplets in adipocytes, J. Lipid Res. 36 (1995) 1211–1226. [5] H.P. Jiang, G. Serrero, Isolation and characterization of a full-length cDNA coding for an adipose differentiation-related protein, Proc. Natl. Acad. Sci. U. S. A. 89 (1992) 7856–7860. [6] D.L. Brasaemle, T. Barber, N.E. Wolins, G. Serrero, E.J. BlanchetteMackie, C. Londos, Adipose differentiation-related protein is an ubiquitously expressed lipid storage droplet-associated protein, J. Lipid Res. 38 (1997) 2249–2263. [7] H.W. Heid, R. Moll, I. Schwetlick, H.R. Rackwitz, T.W. Keenan, Adipophilin is a specific marker of lipid accumulation in diverse cell types and diseases, Cell Tissue Res. 294 (1998) 309–321. [8] E. Diaz, S.R. Pfeffer, TIP47: a cargo selection device for mannose 6phosphate receptor trafficking, Cell 93 (1998) 433–443. [9] N.G. Than, B. Sumegi, G.N. Than, G. Kispal, H. Bohn, Cloning and sequence analysis of cDNAs encoding human placental tissue protein 17 (PP17) variants, Eur. J. Biochem. 258 (1998) 752–757. [10] P.E. Scherer, P.E. Bickel, M. Kotler, H.F. Lodish, Cloning of cell-specific secreted and surface proteins by subtractive antibody screening, Nat. Biotechnol. 16 (1998) 581–586. [11] N.E. Wolins, B. Rubin, D.L. Brasaemle, TIP47 associates with lipid droplets, J. Biol. Chem. 276 (2001) 5101–5108. [12] N.E. Wolins, J.R. Skinner, M.J. Schoenfish, A. Tzekov, K.G. Bensch, P.E. Bickel, Adipocyte protein S3-12 coats nascent lipid droplets, J. Biol. Chem. 278 (2003) 37713–37721. [13] S. Miura, J.W. Gan, J. Brzostowski, M.J. Parisi, C.J. Schultz, C. Londos, B. Oliver, A.R. Kimmel, Functional conservation for lipid storage droplet association among perilipin, ADRP, and TIP47 (PAT)-related proteins in mammals, Drosophila, and Dictyostelium, J. Biol. Chem. 277 (2002) 32253–32257. [14] D.L. Brasaemle, G. Dolios, L. Shapiro, R. Wang, Proteomic analysis of proteins associated with lipid droplets of basal and lipolytically stimulated 3T3-L1 adipocytes, J. Biol. Chem. 279 (2004) 46835–46842. [15] Y. Fujimoto, H. Itabe, J. Sakai, M. Makita, J. Noda, M. Mori, Y. Higashi, S. Kojima, T. Takano, Identification of major proteins in the lipid dropletenriched fraction isolated from the human hepatocyte cell line HuH7, Biochim. Biophys. Acta 1644 (2004) 47–59. [16] P. Liu, Y. Ying, Y. Zhao, D.I. Mundy, M. Zhu, R.G. Anderson, Chinese hamster ovary K2 cell lipid droplets appear to be metabolic organelles involved in membrane traffic, J. Biol. Chem. 279 (2004) 3787–3792. [17] E. Umlauf, E. Csaszar, M. Moertelmaier, G.J. Schuetz, R.G. Parton, R. Prohaska, Association of stomatin with lipid bodies, J. Biol. Chem. 279 (2004) 23699–23709. [18] . X. Lu, J. Gruia-Gray, N.G. Copeland, D.J. Gilbert, N.A. Jenkins, C. Londos, A.R. Kimmel, The murine perilipin gene: the lipid droplet-associated perilipins derive from tissue-specific, mRNA splice variants and define a gene family of ancient origin, Mamm. Genome 12 (2001) 741–749. [19] K.T. Dalen, K. Schoonjans, S.M. Ulven, M.S. Weedon-Fekjaer, T.G. Bentzen, H. Koutnikova, J. Auwerx, H.I. Nebb, Adipose tissue expression of the lipid droplet-associating proteins S3-12 and perilipin is controlled by peroxisome proliferator-activated receptor-gamma, Diabetes 53 (2004) 1243–1252. [20] R. Bussell Jr., D. Eliezer, A structural and functional role for 11-mer repeats in alpha-synuclein and other exchangeable lipid binding proteins, J. Mol. Biol. 329 (2003) 763–778. [21] K.T. Dalen, S.M. Ulven, B.M. Arntsen, K. Solaas, H.I. Nebb, PPAR {alpha} activators and fasting induce the expression of adipose differentiation-related protein in liver, J. Lipid Res. 47 (2006) 931–943. [22] J. Gao, H. Ye, G. Serrero, Stimulation of adipose differentiation related protein (ADRP) expression in adipocyte precursors by long-chain fatty acids, J. Cell. Physiol. 182 (2000) 297–302.

226

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227

[23] D.A. Servetnick, D.L. Brasaemle, J. Gruia-Gray, A.R. Kimmel, J. Wolff, C. Londos, Perilipins are associated with cholesteryl ester droplets in steroidogenic adrenal cortical and Leydig cells, J. Biol. Chem. 270 (1995) 16970–16973. [24] J. Martinez-Botas, J.B. Anderson, D. Tessier, A. Lapillonne, B.H. Chang, M.J. Quast, D. Gorenstein, K.H. Chen, L. Chan, Absence of perilipin results in leanness and reverses obesity in Lepr(db/db) mice, Nat. Genet. 26 (2000) 474–479. [25] J.T. Tansey, C. Sztalryd, J. Gruia-Gray, D.L. Roush, J.V. Zee, O. Gavrilova, M.L. Reitman, C.X. Deng, C. Li, A.R. Kimmel, C. Londos, Perilipin ablation results in a lean mouse with aberrant adipocyte lipolysis, enhanced leptin production, and resistance to diet-induced obesity, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 6494–6499. [26] C. Londos, D.L. Brasaemle, C.J. Schultz, D.C. Adler-Wailes, D.M. Levin, A.R. Kimmel, C.M. Rondinone, On the control of lipolysis in adipocytes, Ann. N. Y. Acad. Sci. 892 (1999) 155–168. [27] C. Sztalryd, G. Xu, H. Dorward, J.T. Tansey, J.A. Contreras, A.R. Kimmel, C. Londos, Perilipin A is essential for the translocation of hormonesensitive lipase during lipolytic activation, J. Cell Biol. 161 (2003) 1093–1103. [28] J.T. Tansey, A.M. Huml, R. Vogt, K.E. Davis, J.M. Jones, K.A. Fraser, D.L. Brasaemle, A.R. Kimmel, C. Londos, Functional studies on native and mutated forms of perilipins. A role in protein kinase A-mediated lipolysis of triacylglycerols, J. Biol. Chem. 278 (2003) 8401–8406. [29] J.T. Tansey, C. Sztalryd, E.M. Hlavin, A.R. Kimmel, C. Londos, The central role of perilipin a in lipid metabolism and adipocyte lipolysis, IUBMB Life 56 (2004) 379–385. [30] S.C. Souza, K.V. Muliro, L. Liscum, P. Lien, M.T. Yamamoto, J.E. Schaffer, G.E. Dallal, X. Wang, F.B. Kraemer, M. Obin, A.S. Greenberg, Modulation of hormone-sensitive lipase and protein kinase A-mediated lipolysis by perilipin A in an adenoviral reconstituted system, J. Biol. Chem. 277 (2002) 8267–8272. [31] G. Xu, C. Sztalryd, X. Lu, J.T. Tansey, J. Gan, H. Dorward, A.R. Kimmel, C. Londos, Post-translational regulation of adipose differentiation-related protein by the ubiquitin/proteasome pathway, J. Biol. Chem. 280 (2005) 42841–42847. [32] B.H. Chang, L. Li, A. Paul, S. Taniguchi, V. Nannegari, W.C. Heird, L. Chan, Protection against fatty liver but normal adipogenesis in mice lacking adipose differentiation-related protein, Mol. Cell. Biol. 26 (2006) 1063–1076. [33] S. Mandard, M. Muller, S. Kersten, Peroxisome proliferator-activated receptor alpha target genes, Cell. Mol. Life Sci. 61 (2004) 393–416. [34] P. Tontonoz, E. Hu, R.A. Graves, A.I. Budavari, B.M. Spiegelman, mPPAR gamma 2: tissue-specific regulator of an adipocyte enhancer, Genes Dev. 8 (1994) 1224–1234. [35] I. Issemann, S. Green, Activation of a member of the steroid hormone receptor superfamily by peroxisome proliferators, Nature 347 (1990) 645–650. [36] O. Braissant, F. Foufelle, C. Scotto, M. Dauca, W. Wahli, Differential expression of peroxisome proliferator-activated receptors (PPARs): tissue distribution of PPAR-alpha, -beta, and -gamma in the adult rat, Endocrinology 137 (1996) 354–366. [37] N. Arimura, T. Horiba, M. Imagawa, M. Shimizu, R. Sato, The peroxisome proliferator-activated receptor gamma regulates expression of the perilipin gene in adipocytes, J. Biol. Chem. 279 (2004) 10070–10076. [38] S. Nagai, C. Shimizu, M. Umetsu, S. Taniguchi, M. Endo, H. Miyoshi, N. Yoshioka, M. Kubo, T. Koike, Identification of a functional peroxisome proliferator-activated receptor responsive element within the murine perilipin gene, Endocrinology 145 (2004) 2346–2356. [39] M. Shimizu, A. Takeshita, T. Tsukamoto, F.J. Gonzalez, T. Osumi, Tissueselective, bidirectional regulation of PEX11 alpha and perilipin genes through a common peroxisome proliferator response element, Mol. Cell. Biol. 24 (2004) 1313–1323. [40] U. Edvardsson, A. Ljungberg, D. Linden, L. William-Olsson, H. PeilotSjogren, A. Ahnmark, J. Oscarsson, PPARalpha activation increases triglyceride mass and adipose differentiation-related protein in hepatocytes, J. Lipid Res. 47 (2006) 329–340. [41] P. Targett-Adams, M.J. McElwee, E. Ehrenborg, M.C. Gustafsson, C.N.

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53] [54]

[55]

[56]

[57]

[58]

Palmer, J. McLauchlan, A PPAR response element regulates transcription of the gene for human adipose differentiation-related protein, Biochim. Biophys. Acta 1728 (2005) 95–104. A. Chawla, C.H. Lee, Y. Barak, W. He, J. Rosenfeld, D. Liao, J. Han, H. Kang, R.M. Evans, PPARdelta is a very low-density lipoprotein sensor in macrophages, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 1268–1273. M. Schmuth, C.M. Haqq, W.J. Cairns, J.C. Holder, S. Dorsam, S. Chang, P. Lau, A.J. Fowler, G. Chuang, A.H. Moser, B.E. Brown, M. Mao-Qiang, Y. Uchida, K. Schoonjans, J. Auwerx, P. Chambon, T.M. Willson, P.M. Elias, K.R. Feingold, Peroxisome proliferator-activated receptor (PPAR)-beta/ delta stimulates differentiation and lipid accumulation in keratinocytes, J. Invest. Dermatol. 122 (2004) 971–983. K.A. Tobin, N.K. Harsem, K.T. Dalen, A.C. Staff, H.I. Nebb, A.K. Duttaroy, Regulation of ADRP expression by long-chain polyunsaturated fatty acids in BeWo cells, a human placental choriocarcinoma cell line, J. Lipid Res. 47 (2006) 815–823. T. Yamaguchi, S. Matsushita, K. Motojima, F. Hirose, T. Osumi, MLDP, a novel PAT family protein localized to lipid droplets and enriched in the heart, is regulated by peroxisome proliferator-activated receptor {alpha}, J. Biol. Chem. 281 (2006) 14232–14240. S.J. Hickenbottom, A.R. Kimmel, C. Londos, J.H. Hurley, Structure of a lipid droplet protein; the PAT family member TIP47, Structure (Cambridge) 12 (2004) 1199–1207. S.F. Altschul, T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller, D.J. Lipman, Gapped BLAST and PSI-BLAST: a new generation of protein database search programs, Nucleic Acids Res. 25 (1997) 3389–3402. K.T. Dalen, S.M. Ulven, K. Bamberg, J.-Å. Gustafsson, H.I. Nebb, Expression of the insulin responsive glucose transporter GLUT4 in adipocytes is dependent on liver X receptor alpha, J. Biol. Chem. 278 (2003) 48283–48291. D.J. Adams, M.A. Quail, T. Cox, W.L. van der, B.D. Gorick, Q. Su, W.I. Chan, R. Davies, J.K. Bonfield, F. Law, S. Humphray, B. Plumb, P. Liu, J. Rogers, A. Bradley, A genome-wide, end-sequenced 129Sv BAC library resource for targeting vector construction, Genomics 86 (2005) 753–758. T.A. Tatusova, T.L. Madden, BLAST 2 sequences, a new tool for comparing protein and nucleotide sequences, FEMS Microbiol. Lett. 174 (1999) 247–250. A. Bavner, J. Matthews, S. Sanyal, J.A. Gustafsson, E. Treuter, EID3 is a novel EID family member and an inhibitor of CBP-dependent coactivation, Nucleic Acids Res. 33 (2005) 3561–3569. S.S. Lee, T. Pineau, J. Drago, E.J. Lee, J.W. Owens, D.L. Kroetz, P.M. Fernandez-Salguero, H. Westphal, F.J. Gonzalez, Targeted disruption of the alpha isoform of the peroxisome proliferator-activated receptor gene in mice results in abolishment of the pleiotropic effects of peroxisome proliferators, Mol. Cell. Biol. 15 (1995) 3012–3022. D.P. Eisinger, G. Serrero, Structure of the gene encoding mouse adipose differentiation-related protein (ADRP), Genomics 16 (1993) 638–644. P. Targett-Adams, D. Chambers, S. Gledhill, R.G. Hope, J.F. Coy, A. Girod, J. McLauchlan, Live cell analysis and targeting of the lipid dropletbinding adipocyte differentiation-related protein, J. Biol. Chem. 278 (2003) 15998–16007. C. Sztalryd, M. Bell, X. Lu, P. Mertz, S. Hickenbottom, B.H. Chang, L. Chan, A.R. Kimmel, C. Londos, Functional compensation for adipose differentiation-related protein (ADFP) by TIP47 in an adfp nullembryonic cell line, J. Biol. Chem. 281 (2006) 34341–34348. M.J. Hoffer, M.H. Hofker, M.M. van Eck, L.M. Havekes, R.R. Frants, Evolutionary conservation of the mouse apolipoprotein e-c1–c2 gene cluster: structure and genetic variability in inbred mice, Genomics 15 (1993) 62–67. S.K. Karathanasis, Apolipoprotein multigene family: tandem organization of human apolipoprotein AI, CIII, and AIV genes, Proc. Natl. Acad. Sci. U. S. A. 82 (1985) 6374–6378. P.A. Mak, B.A. Laffitte, C. Desrumaux, S.B. Joseph, L.K. Curtiss, D.J. Mangelsdorf, P. Tontonoz, P.A. Edwards, Regulated expression of the apolipoprotein E/C-I/C-IV/C-II gene cluster in murine and human macrophages. A critical role for nuclear liver X receptors alpha and beta, J. Biol. Chem. 277 (2002) 31900–31908.

K.T. Dalen et al. / Biochimica et Biophysica Acta 1771 (2007) 210–227 [59] K.W. van Dijk, P.C. Rensen, P.J. Voshol, L.M. Havekes, The role and mode of action of apolipoproteins CIII and AV: synergistic actors in triglyceride metabolism? Curr. Opin. Lipidol. 15 (2004) 239–246. [60] A. Garcia, A. Sekowski, V. Subramanian, D.L. Brasaemle, The central domain is required to target and anchor perilipin A to lipid droplets, J. Biol. Chem. 278 (2003) 625–635. [61] D.L. Brasaemle, T. Barber, A.R. Kimmel, C. Londos, Post-translational regulation of perilipin expression. Stabilization by stored intracellular neutral lipids, J. Biol. Chem. 272 (1997) 9378–9387. [62] R. Zimmermann, J.G. Strauss, G. Haemmerle, G. Schoiswohl, R. BirnerGruenberger, M. Riederer, A. Lass, G. Neuberger, F. Eisenhaber, A. Hermetter, R. Zechner, Fat mobilization in adipose tissue is promoted by adipose triglyceride lipase, Science 306 (2004) 1383–1386.

227

[63] A.L. Magra, P.S. Mertz, J.S. Torday, C. Londos, Role of adipose differentiation-related protein in lung surfactant production: A reassessment, J. Lipid Res. 47 (2006) 2367–2373. [64] B.B. Rasmussen, R.R. Wolfe, Regulation of fatty acid oxidation in skeletal muscle, Annu. Rev. Nutr. 19 (1999) 463–484. [65] M.S. Jafri, S.J. Dudycha, B. O'Rourke, Cardiac energy metabolism: models of cellular respiration, Annu. Rev. Biomed. Eng. 3 (2001) 57–81. [66] S. Kersten, J. Seydoux, J.M. Peters, F.J. Gonzalez, B. Desvergne, W. Wahli, Peroxisome proliferator-activated receptor alpha mediates the adaptive response to fasting, J. Clin. Invest. 103 (1999) 1489–1498. [67] J. Nishiu, T. Tanaka, Y. Nakamura, Isolation and chromosomal mapping of the human homolog of perilipin (PLIN), a rat adipose tissue-specific gene, by differential display method, Genomics 48 (1998) 254–257.