Originally published In Press as doi:10.1194/jlr.M300279-JLR200 on October 1, 2003
Journal of Lipid Research, Vol. 45, 17-31, January 2004
Copyright © 2004 by American Society for Biochemistry and Molecular Biology
PPAR
controls the intracellular coenzyme A concentration via regulation of PANK1 gene expression1
Gayathri Ramaswamy*,
,
Mohammad A. Karim2,*,
K. Gopal Murti
,** and
Suzanne Jackowski3,*,
* Protein Science Division, St. Jude Children's Research Hospital, Memphis, TN 38105-2794
Department of Infectious Diseases, and Scientific Imaging Shared Resource, St. Jude Children's Research Hospital, Memphis, TN 38105-2794
Departments of Molecular Sciences, University of Tennessee Health Science Center, Memphis, TN 38163
** Pathology, University of Tennessee Health Science Center, Memphis, TN 38163
1 The nucleotide sequences reported in this paper have been submitted to the GenBankTM/EBI Data Bank with accession numbers AY027661, AY027662, and AY027663. 
Published, JLR Papers in Press, October 1, 2003. DOI 10.1194/jlr.M300279-JLR200
2 Present address of M. A. Karim: Division of Endocrinology, Department of Medicine, Central Arkansas Veterans Healthcare System and University of Arkansas for Medical Sciences, Little Rock, AR 72205.
3 To whom correspondence should be addressed. e-mail: suzanne.jackowski{at}stjude.org
 |
ABSTRACT
|
|---|
Pantothenate kinase (PanK) is thought to catalyze the first rate-limiting step in CoA biosynthesis. The full-length cDNA encoding the human PanK1
protein was isolated, and the complete human PANK1 gene structure was determined. Bezafibrate (BF), a hypolipidemic drug and a peroxisome proliferator activator receptor- (PPAR) agonist, specifically increased hPANK1 mRNA expression in human hepatoblastoma (HepG2) cells as a function of time and dose of the drug, compared with hPANK1ß, hPANK2, and hPANK3, which did not significantly increase. Four putative PPAR response elements were identified in the PANKI promoter, and BF stimulated hPANK1 promoter activity but did not alter the mRNA half-life. Increased hPANK1 mRNA resulted in higher hPanK1 protein, localized in the cytoplasm, and elevated PanK enzyme activity. The enhanced hPANK1 gene expression translated into increased activity of the CoA biosynthetic pathway and established a higher steady-state CoA level in HepG2 cells.
These data are consistent with a key role for PanK1 in the control of cellular CoA content and point to the PPAR transcription factor as a major factor governing hepatic CoA levels by specific modulation of PANK1 gene expression.
Abbreviations: BF, bezafibrate; HSS, Hallervorden-Spatz syndrome; P-Pan, 4'-phosphopantothenate; P-PanSH, 4'-phosphopantetheine; Pan, pantothenate; PanK, pantothenate kinase; PKAN, pantothenate kinase-associated neurodegeneration; PPAR, peroxisome proliferator activator receptor-; RACE, rapid amplification of cDNA ends
Supplementary key words pantothenate • pantothenate kinase • peroxisome proliferator activator receptor- • fatty acid ß-oxidation • liver
|
INTRODUCTION
|
|---|
Pantothenate (Pan) is a B complex vitamin (B5) and is a nutritional requirement in mammals. Pan is the precursor for the biosynthesis of CoA, the predominant acyl group carrier in cells. Acyl moieties are attached to the terminal sulfhydryl of CoA, and these compounds participate in over 100 different reactions in intermediary metabolism (1–4). Short-chain CoA thioesters, such as acetyl-CoA or succinyl-CoA, are the most abundant components of the CoA pool (5) and are important intermediates in the tricarboxylic acid cycle, which coordinates carbon utilization and oxidative energy production. CoA is also an essential cofactor in long-chain fatty acid metabolism. CoA transfers fatty acids to and from the mitochondrial carnitine transferases and carries all of the intermediates of fatty acid ß-oxidation in peroxisomes and mitochondria. CoA acyl-thioesters are also the substrates for the acyltransferases and desaturases involved in membrane phospholipid formation, triglyceride synthesis, and protein acylation. Thus, CoA participates in the major anabolic and catabolic pathways critical to cell growth and function.
Pan is phosphorylated by pantothenate kinase (PanK) to yield 4'-phosphopantothenate (P-Pan) in the first enzymatic reaction leading to CoA. PanK activity controls the cellular level of CoA in bacteria (6) and mammals (7, 8), and there are multiple Pank genes that are differentially expressed and regulated in mammalian tissues (4, 9–11). The first mammalian gene identified was the mouse Pank1 gene, which encodes two proteins that differ at the N terminus due to alternative splicing of distinct initiating exons (9). PanK1 enzyme activity is feedback inhibited by both unacylated CoA and acetyl-CoA (9), whereas the PanK1ß isoform is less sensitive to feedback regulation by acetyl-CoA and is stimulated by unacylated CoA (8). The human PANK2 gene also encodes two protein isoforms, one of which preferentially localizes to mitochondria (11). Mutations in the human PANK2 coding sequence create a metabolic imbalance leading to either HARP (hypoprebetalipoproteinemia, acanthocytosis, retinitis pigmentosa, pallidal degeneration) syndrome (12) or a progressive childhood neuropathy called PanK-associated neurodegeneration (PKAN) (10) characterized by iron accumulation in the basal ganglia. The human genetic data suggest that dysfunctional CoA biosynthesis due either to defective expression or to sequence alteration of a single PanK isoform can selectively impact the function of specific tissues.
PanK1 is expressed at the highest levels in mouse liver, kidney, and heart tissues (8, 10). Before the discovery of multiple PanK isoforms, it was noted that the levels of liver PanK activity (13, 14) decreased in conjunction with CoA levels following starvation and increased following feeding (15–19), increased in diabetes (20, 21), and increased after treatment with hypolipidemic agents (17, 22, 23), suggesting that changes in PanK expression may be involved in altering the physiological status of the tissues. The peroxisome proliferator activator receptor- (PPAR) is a transcription factor that is highly expressed in liver, regulates the expression of fatty acid oxidation genes (24), and is activated by hypolipidemic agents such as the fibrates and WY146,43 (25). Activated PPAR binds to the retinoic acid receptor RXR, and this complex binds to specific DNA sequences called peroxisome proliferator response elements (PPREs) located in the promoter regions of target genes to upregulate their expression (26–29).
This work investigates the role of PanK gene expression in regulating CoA biosynthesis in human hepatoblastoma (HepG2) cultured cells. The full-length human PANK1 transcript was identified and the human PANK1 gene structure determined. The human PanK1ß isoform (30, 31) is homologous with the mouse PanK1ß (8). Like the murine PanK1 isoform, the human PanK1 protein is significantly longer than the human PanK1ß at the N terminus. The expressed human PanK1 protein is biochemically active, its expression is stimulated by the PPAR agonist, bezafibrate (BF), and increased PanK1 activity resulted in elevated CoA levels. These data describe a mechanism for stimulation of CoA biosynthesis via increased PanK activity in human liver cells and point to regulation of PanK1 isoform expression as a component of the physiological response to hypolipidemic agents.
|
EXPERIMENTAL PROCEDURES
|
|---|
Materials
Sources of supplies were as follows: American Radiolabel Co., Inc., D-[1-14C]pantothenic acid (55 mCi/mmol) and D-[2,3-3H]pantothenic acid (25 Ci/mmol); American Type Culture Collection, human HepG2 cells and African Green Monkey kidney (COS7) cells; Amersham Pharmacia Biotech, deoxycytidine 5'-[-32P]triphosphate (3,000 Ci/mmol), ECL Western blotting detection kit, Rediprime II labeling kit; Analtech, Silica gel H TLC plates; Analytical Luminescence Laboratory/BD Pharmingen, Monolight® 2010 Luminometer; Biodesign International, Saco, ME, sheep anti-bovine catalase antibody; BioWhittaker, Dulbecco's modified Eagle's medium (DMEM) and phenol red-free DMEM; Boehringer Mannheim, RNaseA, terminal transferase; Calbiochem, protease inhibitor cocktail set III, sodium pantothenate, CoA, 4'-phosphopantetheine (P-PanSH); Clontech, Expand long template PCR system, ExpressHyb hybridization solution, Marathon-readyTM liver and kidney cDNA libraries, human brain, heart, kidney, liver, skeletal muscle and testis poly(A)+ RNA; Falcon, culture plasticware; Hyclone, fetal bovine serum (FBS) and 0.25% trypsin; Invitrogen (Gibco) Life Technologies, Inc., cDNA cycle kit, TRIzol LS reagent, TA cloning kit, pcDNA3.1(-) vector, glutamine, penicillin/streptomycin, phosphate-buffered saline, lipofectAMINETM and lipofectAMINE plusTM reagent, Pan-free DMEM; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, fluorescein (FITC)-conjugated goat anti-rabbit IgG, RNase A, Texas Red-coupled donkey anti-sheep antibody; Kodak, BioMax MR X-ray films; Media Cybergenetics, Silver Spring, MD, ImageProPlus software; Micron Separations Inc., Nitropure nitrocellulose transfer membranes; Misonix Sonicator 3000; Molecular Dynamics, Storm 860 Phosphor Imager, and Typhoon Phosphor Imager; Molecular Probes, mitotracker red, propidium iodide; Nalge Nunc International, Labtek glass chambers; Novatech, PAN-free DMEM; Promega, pGL3 vectors, restriction endonucleases, Taq polymerase, Luciferase Assay kit; Roche, 5'/3'-rapid amplification of cDNA end (RACE) kit; Schleicher and Schuell, TURBOBLOTTERTM Rapid Downward Transfer System; Sigma Aldrich Chemical Co., delipidated fetal calf serum, dimethyl sulfoxide, BF. All other supplies were reagent grade or better.
Identification of the hPanK1-specific transcript sequence
Human BACmid AQ076749 was identified by performing a BLAST search using the mouse Pank1 cDNA (AF347700) as the query. The 5'-end of the hPANK1 transcript was found with 5'-RACE using the exon 1-specific reverse primer, 5'-CGCCACCGACGAGTTTCAACAT-3', and human kidney poly(A)+ RNA as template for first-strand cDNA synthesis. The homopolymeric tailing of the purified cDNA was perfomed using terminal transferase and dATP according to the manufacturer's instructions. The dA-tailed, first-strand cDNA was used as the template for first round of PCR using exon 1-specific 5'-GGGAATCAAGTTCTCTGGGCTTGCA-3' as reverse primer and an oligo (dT) anchor primer, 5'-GACCACGCGTATCGATGTCGACTTTTTTTTTTTTTTTTT-3', as the forward primer. A second round of PCR was performed using the product from the first-round PCR as the template and using a gene-specific nested reverse primer, 5'-GGGCTTTGTATGGCGATGTGAAAACC-3', and a PCR anchor primer, 5'-GACCACGCGTATCGATGTCGAC-3', as the forward primer. The PCR product was subcloned into pCR2.1 and the sequence of the product verified. The 129 bp sequence was found to be identical to a portion of the genomic sequence of human BAC AQ076749. Another round of 5'-RACE was performed using human kidney poly(A)+ RNA as the template and using exon 1-specific reverse primer 5'-GGGCTTTGTATGGCGATGTGAAAACC-3' and the oligo(dT) anchor primer mentioned above. The second round of PCR was performed using the product from the first round as the template and using a gene-specific nested reverse primer, 5'-GAACCGTTTCAGGAGGAATACGG-3', and PCR anchor primer as the forward primer. The 184 bp product was subcloned into pCR2.1, and the sequence of the plasmid was determined. This sequence marked the end of the hPanK1 transcript.
Tissue distribution of hPANK isoforms
Expression of the hPANK genes was evaluated by RT-PCR using the cDNA cycle kit according to the manufacturer's instructions. First-strand cDNA synthesis mediated by reverse transcriptase was done using a common primer, 5'-ACTTCCGGATGCTCTTCAGG-3', and poly(A)+ RNAs (1 µg) from human brain, heart, kidney, liver, skeletal muscle, and testis (Clonetech) were used as templates. An aliquot (3 µl) of the RT reaction was used as the template for PCR. PCR detected both hPanK1 and hPanK1ß isoforms in a multiplex reaction using the above-mentioned common primer as the reverse primer together with 5'-CTGCGGAGGAGGATGGAC-3' and 5'-GCCGAAGTGCATTATTTCACA-3' as hPanK1- and hPanK1ß-specific forward primers, respectively. The thermal cycling program was denaturation at 94°C for 30 s, annealing at 57°C for 30 s, and extension at 72° for 45 s. The total number of thermal cycles was 50. The PCR products were separated on a 1.5% agarose gel. The expected sizes of the hPanK1 and hPanK1ß PCR products were 163 bp and 318 bp, respectively.
Isolation of total RNA
HepG2 cells were cultured in 150 mm dishes that were placed on ice, and the cells were scraped into 20 ml cell culture medium. The cell suspension was harvested by centrifugation at 106 g for 6 min at 4°C. The media was aspirated, and the pellet was resuspended in 6 ml of RNA lysis buffer [4 M guanidinium thiocyanate, 10 mM Tris-HCl (pH 7.4), 2% ß-mercaptoethanol, 2% Sarkosyl, and 10 mM EDTA]. RNA was extracted using the guanidine method (32). Alternatively, RNA was extracted using TRIzol LS reagent according to the manufacturer's instructions.
Detection and quantitation of hPanK isoforms
A primer common to both the hPanK1 and hPanK1ß isoforms, 5'-GTCAAATACTTCCGGATGCTC-3', present in exon 2, was used for first-strand cDNA synthesis by reverse transcription. For hPanK2 and hPanK3, the primers used for cDNA synthesis were 5'-AGCCAATGTTCACCAGAAGC-3' and 5'-TGACTCCTGAGCCAATGTTC-3', respectively. The liver fatty acid binding protein (L-FABP) and GAPDH cDNAs were synthesized using 5'-CATGGTATTGGTGATTATGTCGC-3' and 5'-GACCACCTGGTGCTCAGTGTAG-3', respectively. RT was done using 2 µg of total RNA as template and the cDNA Cycle Kit according to the manufacturer's instructions. An aliquot of the RT reaction (3 µl) was used as template to amplify the corresponding genes by PCR. An hPanK1 nested reverse primer, 5'-ACTTCCGGATGCTCTTCAGG-3', was used to amplify both hPanK1 and hPanK1ß isoforms by PCR. This primer was present in exon 2 of the hPANK1 gene. Primers 5'-CTGCTCCCTCAGCATGACTC-3' and 5'-GACTGAAGCTCATTCAACTGG-3' were used, respectively, as hPanK1- or hPanK1ß-specific forward primers for PCR. For hPanK2, the forward and reverse PCR primers used were 5'-AGCTGAAGGACCTGACTCTG-3' and 5'-AGCCAATGTTCACCAGAAGC-3'; for hPanK3, the forward primer was 5'-TATCACAGCAGAGGAAGAGC-3' and the reverse primer was 5'-TGACTCCTGAGCCAATGTTC-3'; for L-FABP, the forward primer was 5'-GCAGGTCAGTCGTGAAGAGG-3' and the reverse primer was 5'-CATGGTATTGGTGATTATGTCGC-3'; and for GAPDH, the forward and reverse primers were 5'-CAACAGCCTCAAGATCATCAGC-3' and 5'-GACCACCTGGTGCTCAGTGTAG-3', respectively. The thermal cycling program was denaturation at 94°C for 1 min, annealing at 57°C for 30 s, and extension at 72°C for 45 s. Hot start at 94°C for 5 min was done prior to addition of Taq DNA polymerase. Thirty cycles amplified the hPanK isoforms, and 22 cycles amplified for L-FABP and GAPDH, as predetermined to be in the linear range of accumulation of the individual PCR products as a function of cycle time. The sizes of the RT-PCR products were: hPanK1, 202 bp; hPanK1ß, 190 bp; hPanK2, 368 bp; hPanK3, 488 bp; L-FABP, 371 bp; and GAPDH, 424 bp. An aliquot of each PCR reaction (15 µl) was fractionated by electrophoresis on a 1% agarose gel and transferred to Hybond-N+ nylon membrane using the TURBOBLOTTERTM Rapid Downward Transfer System according to the manufacturer's instructions.
The RT-PCR products were detected by Southern hybridization. The PCR products corresponding to the genes were used as template for synthesis of radioactive probes. The probes were labeled with Redivue deoxycytidine 5'-[-32P]triphosphate using the Rediprime II labeling kit according to the manufacturer's instructions. The blots were hybridized using 2 x 106 cpm/ml of radiolabeled probe in ExpressHyb hybridization solution according to the manufacturer's instructions. Following overnight hybridization, the blots were washed and exposed to phosphor screens for 2 h, and the bands were quantitated using a Storm 860 Phosphor Imager and the ImageQuant software. Alternatively, the PCR products were run on a 1% agarose gel containing 50 µg/ml of ethidium bromide. The ethidium bromide fluorescence was quantitated using the Typhoon Phosphor Imager.
Construction of reporter gene plasmids
NheI and XhoI restriction sites were introduced at the 5' and 3' ends, respectively, of the 2,436 bp hypothetical hPANK1 promoter region by PCR using 5'-GCTAGCTAATTACCAAGTGCCTCC-3' and 5'-CTCGAGGCTTGCAGAGAGGAGATG-3' as forward and reverse primers and using the Expand long template PCR system. The PCR product contained the promoter region plus the transcriptional start site and the untranslated 5' sequence of exon 1 up to the translational start site. DMSO (5%) was included in the PCR reaction, and human Bacmid AQ076749 was used as the template. The thermal cycling conditions were denaturation at 94°C for 30 s, annealing at 58°C for 30 s, and extension at 72°C for 4.5 min for 40 cycles. The PCR product was cloned into pCR2.1 vector and transformed into Escherichia coli strain INVF'; plasmids from multiple transformants were sequenced at the Hartwell center for Biotechnology at St. Jude Children's Research Hospital. The plasmid with the correct DNA sequence was digested with XhoI and subcloned into the pGL3 vector upstream of the luciferase reporter sequence. Plasmid with the promoter in the sense orientation (5' to 3') was called P1, and plasmid with the promoter in the reverse orientation (3' to 5') was called P1-reverse.
Luciferase assays
HepG2 cells were grown to 80% confluence in 6-well plates in phenol red-free DMEM, 10% FBS, and 2 mM glutamine. Cells were transfected with 2 µg of plasmid P1, P1-reverse, or pGL3 basic vector; or 0.5 µg of simian virus 40 (SV40)-driven pGL3 promoter vector using lipofectAMINE plusTM reagents according to the manufacturer's instructions. The duration of transfection was 5 h, following which phenol red-free DMEM, 20% delipidated fetal calf serum, and 2 mM glutamine were added. Twenty-four hours post-transfection, the medium was replaced with complete medium containing phenol red-free DMEM, 10% delipidated fetal calf serum, and 2 mM glutamine. The cells were maintained in complete medium for 24 h, following which fresh complete medium was added to the cells. DMSO or BF (0.5 mM) was added to cells. After incubation for 72 h, the cells were analyzed for luciferase activity using the Luciferase Assay kit according to the manufacturer's instructions. Luciferase activity was measured in the Monolight® 2010 Luminometer. Data were normalized to protein content in the lysate as determined by the Bradford method (33).
Cloning and assembly of hPanK1 cDNA
Based on the sequence information, 5'-GCTAGCTTCACATCGCCATACAAAGCC-3' and 5'-CCATGGGAATGGCGGCCTGTTCTTTCTCCC-3' were used as hPANK1-specific primers to introduce NheI and NcoI restriction endonuclease sites at the 5'- and 3'- ends of exon 1. PCR-mediated cDNA synthesis was performed using human Bacmid AQ076749 as the template, and the product was called Fragment 1 and was cloned into pCR2.1. Human liver Marathon-ReadyTM cDNA was used as template to perform 5'-RACE using the human PANK1 gene-specific reverse primer 5'-CTCGCTGCCCATCTGAATGAACC-3' and an adaptor primer, 1, 5'-CCATCCTAATACGACTCACTATAGGGC-3', as the forward primer for the first round of cDNA synthesis. The cDNA product from the first round was used as a template for the second round of PCR-mediated cDNA synthesis using a nested hPANK1 gene-specific forward primer, 5'-GGACGTCTCGGATCCCAG-3', and a nested adaptor primer, 2, 5'-ACTCACTATAGGGCTCGAGCGGC-3'. This fragment contained exon 1ß and a portion of exon 2 of the hPANK1 gene and was called Fragment 2 and was cloned into pCR2.1. PCR-mediated cDNA synthesis was performed using human kidney Marathon-ReadyTM cDNA as template and hPANK1-specific forward and reverse primers 5'-CGTCCACCTGGAACTGAAAAACCTG-3' and 5'-CCGAGCAATGGAGCCAATG-3', respectively. A 724 bp cDNA fragment corresponding to exon 2 was obtained. AatII and BstUI restriction enzyme sites were introduced at the 5'- and 3'- ends, respectively, of this 724 bp product by PCR using 5'-GACGTCCACCTGGAACTGAA-3' and 5'-CGCGCACATCCGAGCAATGG-3' as forward and reverse primers. This fragment was called Fragment 3 and was cloned into pCR2.1. Poly(A)+ RNA from human kidney was used as a template to perform 3'-RACE using an oligo(dT)-containing adaptor primer 5'-GGCCACGCGTCGACTAGTACTTTTTTTTTTTTTTTTT-3' for reverse transcription, and the resulting cDNA was used as template for PCR-mediated cDNA synthesis using the hPANK1 gene-specific forward primer 5'-CATTGGCTCCATTGCTCGG-3' and the abridged universal amplification primer 5'-GGCCACGCGTCGACTAGTAC-3' as the reverse primer. A 1,676 bp RT-PCR product was obtained and contained exon 6 and the coding region of exon 7. A BstEII site was introduced at the 5'-end of this cDNA product by PCR-mediated cDNA synthesis using 5'-TGGTCACCATCACCAACAACATTGGCTCCATTGCTCGG-3' as the forward primer and 5'-GCGGCCCCGACTTTCACCTTCTCCAGCAG-3' as the reverse primer. This 313 bp PCR product was called Fragment 4 and was cloned into pCR2.1. Multiple independent transformants were isolated from each PCR cloning, and the DNA sequences of the plasmids were determined using M13 forward and reverse primers at the Hartwell Center for Biotechnology at St. Jude Children's Hospital.
Fragments 1, 2, 3, and 4 were assembled and cloned into pcDNA3.1(-), a mammalian expression vector. First, Fragment 2 was subcloned into pBluescriptKS(+/-) using unique SpeI and XhoI restriction sites. Next, Fragment 1 in pCR2.1 and Fragment 2 in pBluescriptKS(+/-) were digested with KpnI and NcoI. The 823 bp product obtained from digestion of Fragment 1 contained exon 1 and was ligated to the 3,045 bp fragment obtained from the digestion of Fragment 2 in pBluescriptKS(+/-), which lacked exon 1ß. The ligated product, called "1+2", was transformed into E. coli strain INVF'. Fragments 3 and 4, each cloned into pCR2.1, were digested with XbaI and BstEII. The 738 bp product from the Fragment 4 plasmid was ligated into the 4,168 bp product from the Fragment 3 plasmid. The ligated product was called "3+4" and was transformed into INVF' cells. Fragment 3+4 was then subcloned into pBluescriptKS(+/-) using SpeI and XhoI restriction sites. The complete hPANK1 cDNA was assembled by digesting Fragments 1+2 and 3+4 with KpnI and AatII. The 977 bp product obtained from the Fragment 1+2 digestion was ligated into the 3,937 bp product obtained from the digestion of Fragment 3+4. The ligated product was transformed into INVF' cells. Finally, the assembled hPANK1 cDNA in pBluescriptKS(+/-) was digested with NheI and XbaI. The 1,984 bp hPANK1 cDNA insert was ligated into the 5,304 bp pcDNA3.1(-) vector, which had been digested with NheI and XbaI. The ligated product was transformed into E. coli strain INVF'.
Cell transfection and PanK assays
COS7 cells were cultured in DMEM plus 10% FBS and grown to 80% confluence in 100 mm culture dishes. COS7 cells were transfected with 10 µg of plasmid DNA using lipofectAMINETM reagent according to the manufacturer's instructions. Cells were harvested 48 h after transfection and were resuspended in hypotonic lysis buffer [10 mM Tris (pH 7.5), 1 mM EDTA, 1x protease inhibitor cocktail set III], incubated on ice for 30 min, and lysed by sonication at maximum power for 3 min in 30 s pulses at 4°C. The cell lysate was centrifuged at 5,500 g for 10 min at 4°C, and the supernatant was used for PanK assays. The assay mixture contained 100 mM Tris (pH 7.5), 20 mM MgCl2, and freshly made 1 mM ATP and D-[1-[14C]pantothenic acid (50.5 µM; specific activity, 55 mCi/mmol) plus the indicated amount of lysate protein and was incubated at 37°C for 15 min. The assay was stopped, and the P-Pan product was quantitated as described previously (8).
PanK activity in HepG2 cells was determined as follows: cells were scraped from culture dishes and washed with PBS, and the cell pellets were resuspended in an equal volume of ice-cold hypotonic lysis buffer and incubated in a rotator at 4°C for 1 h. The cells were then lysed by sonication at maximum power for 6 min in 30 s pulses at 4°C. The cell lysate was centrifuged at 5,500 g for 10 min at 4°C. The supernatant protein was quantitated by the Bradford method (33) and was used for PanK assays. The PanK assay was performed as described above.
Immunological detection of hPanK1 protein
A peptide ECYYGENPTNPELCQ was synthesized and coupled to keyhole limpet hemocyanin by the Hartwell Center for Biotechnology, St. Jude Children's Research Hospital, and was sent to Rockland Inc., Gilbertsville, PA for raising rabbit polyclonal antiserum against PanK1. Affinity purification of the polyclonal antisera was performed as described previously (9). COS7 cells were transfected and lysed using hypotonic lysis buffer as described above. Aliquots of the crude cell lysate from transfected and untransfected cells were fractionated by SDS-PAGE on 10% gels. The separated proteins were electroblotted onto nitrocellulose membrane, and the hPanK1 protein was detected by immunoblotting using the ECL detection kit according to the manufacturer's instructions. PanK1-specific antibody was used at a dilution of 1:500 (stock 0.4 mg/ml) as the primary antibody, which was incubated with the membrane for 1 h at room temperature. Horseradish peroxidase-labeled anti-rabbit IgG was used as the secondary antibody and was diluted to 1:5,000 prior to incubation with the membrane for 1 h at room temperature. The blots were washed five times with 100 ml of TBS-T and developed using the ECL detection reagents according to the manufacturer's instructions. The chemiluminescent signal was detected using X-ray films.
Immunostaining, confocal microscopy, and quantitation of hPanK1 protein expression
HepG2 cells were seeded at a density of 5 x 104 in DMEM plus 10% FBS in 2-well Labtek glass chambers with DMSO alone or DMSO with 0.5 mM BF (0.25 M stock solution made in DMSO) for indicated times of incubation at 37°C in 5% CO2, 95% air. The cells were washed with PBS (pH 7.4) three times for 2 min each and permeabilized in 0.5% Triton X-100 in PBS for 5 min at room temperature. The cells were fixed with 4% formaldehyde in PBS for 30 min at room temperature and incubated with 1% BSA in PBS for 1 h to block nonspecific cellular binding sites. The cells were incubated with the affinity-purified rabbit polyclonal anti-PanK1 primary antibody at the indicated dilutions in PBS for 45 min at 37°C. The anti-PanK1 antibody was removed, and cells were washed five times with PBS. FITC-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories, Inc.) diluted 1:50 in PBS was added to the cells, followed by incubation at 37°C for 45 min. The cells were again washed five times with PBS. For nuclear staining, cells were first treated with 1:100 dilution of RNase A [10 mg/ml stock in 10 mM Tris (pH 7.5), Jackson ImmunoResearch Laboratories, Inc.] for 20 min at 37°C to digest RNA, followed by incubation with 5 µg/ml propidium iodide for 5 min at 37°C. The RNase digestion was performed to eliminate nonspecific staining of RNA by propidium iodide. At this point, the plastic chambers on slides were removed, the slides were washed with PBS, and a drop of p-phenylenediamine medium (34) was added to the slides to prevent bleaching of the fluorochrome. The slides were then covered with glass coverslips and sealed with nail polish. In double-staining experiments to localize hPanK protein in mitochondria or peroxisomes, the cells were first stained for the protein, followed by incubation with mitotracker red (Molecular Probes, Eugene, OR) or anti-catalase (sheep anti-bovine catalase antibody; Biodesign International) and Texas Red-coupled donkey anti-sheep antibody (Jackson ImmunoResearch Laboratories, Inc.).
The images were obtained using a Leica TCS NT SP confocal microscope. The microscope is equipped with an argon laser to detect green (FITC) fluorescence (absorption/emission = 494 nm/518 nm) and a Krypton laser to detect red (propidium iodide or Texas Red) fluorescence (absorption/emission = 568 nm/615–617 nm). Single optical sections (0.5 µm thick) were obtained from cells, and the images were sequentially scanned to minimize cross-talk between the green and red channels. Individual sections or three-dimensional images constructed from multiple optical sections were used to display the data in both green and red channels using the Leica imaging software. Overlay images from green and red channels were used to show the colocalization of the two fluorochromes. To maintain consistency and accuracy, samples were examined at the same laser power and at the same photo multiplier tube gain settings in both channels.
The quantitation of hPanK1 protein expression was done using a previously described procedure (35) and ImageProPlus software (Media Cybergenetics). Mean fluorescence intensity (FITC-green) of the hPanK1 protein signal in the cytoplasm was measured in control cells and in those treated with BF. Single-section digital images containing 20 cells were used in the analysis to determine the fluorescence intensity of the cytoplasm (single pixel intensity x number of pixels per cell); the nuclei, defined by the red fluorescence of the propidium iodide, were excluded from the analysis.
Metabolic labeling
HepG2 cells were grown to 35% confluence in 100 mm culture dishes in regular complete DMEM plus 10% serum. Cells were washed twice with PBS and radiolabeled with D-[2,3-3H]pantothenic acid (33.3 nM, specific activity 25 Ci/mmol) in PAN-free DMEM. At the times indicated, cells were washed with PBS and harvested by trypsinization using 0.25% trypsin. Following inactivation of the trypsin, the cells were separated from the medium and cell pellets were stored at -80°C. The pellets were resuspended in 50 µl of hypotonic lysis buffer and lysed by sonication as described above. The cell lysate was centrifuged at 10,620 g for 10 min at 4°C. Aliquots (10 µl) of the lysates were treated with DTT for 3 min and spotted onto silica gel H TLC plates to separate the radiolabeled metabolites. The plates were developed with butanol-acetic acid-water (5:2:4; v/v/v) as the solvent system. Plates were scraped (0.5 cm zones) into scintillation vials, and radioactivity was counted following deactivation of the silica gel and addition of scintillation fluid. The metabolites were identified and quantified by comigration with authentic standards for Pan, CoA, and P-PanSH (6).
|
RESULTS
|
|---|
Structure of the human PANK1 gene
A BLAST search of the National Center for Biotechnology Information human genomic database identified a high throughput genomic contig, AL157400, that resembled the mouse Pank1 cDNA and gene structure (9). A comparison of the mouse and human PanK1 DNA sequences suggested, however, that the human PanK1 open reading frame was longer than the mouse protein due to the absence of a translational initiating methionine preceded by upstream stop codons. Thus, the complete cDNA sequence of hPanK1 was determined using 5'-RACE, RT-PCR, and 3'-RACE techniques as described under Experimental Procedures. Compared with mPanK1, the hPanK1 transcript encoded an additional 38 amino acids at the N terminus of the protein, and in-frame stop codons were located upstream of the initiating methionine. The structure of the human PANK1 gene was determined by comparing the human PANK1 and ß cDNA sequences with the genomic sequence of AL157400 (Fig. 1
; Table 1). The hPANK1 transcript initiates at exon-1, which consists of 989 bp: 705 bp constitute the coding region, and the remaining 284 bp constitute the 5'-untranslated region. hPANK1ß initiates at exon-1ß, which consists of 195 bp: 30 bp of coding sequence, with the remaining 165 bp constituting the 5'-untranslated region. Either of these exons are spliced into exons 2 through 7, which are common to both the transcripts. The common exon 7 consists of 1,478 bp, which includes 1,421 bp of 3'-untranslated region as determined by 3'-RACE. The hPANK1 transcript is 3.5 kb, and the hPANK1ß transcript is 2.7 kb. The PANK1 gene was mapped to human chromosome 10q23.3-23.31 based on the human genome project, and this location was confirmed by fluorescence in situ hybridization using human BACmid AQ076749 as the probe (data not shown). In summary, the human PANK1 gene encodes two protein isoforms, PanK1 and PanK1ß. The hPANK1 open reading frame encodes a protein of 598 amino acids, whereas the hPANK1ß encodes a 373 amino acid protein (Fig. 2)
. These proteins differ in their amino termini but share a common carboxy terminal domain consisting of 363 amino acids.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 1. Diagram of the human PANK1 gene structure. The exon/intron arrangement of the 62 kb human PANK1 gene is diagrammatically illustrated along with the hPANK1 and hPANK1ß transcripts that encode distinct PanK1 proteins. The two hPANK1 isoforms differ at exon 1 and utilize common exons 2–7. The darkened boxes depict the coding regions within the exons, and the lighter boxes indicate the untranslated sequences. The diagram is not to scale, and the approximate intron sizes are indicated. The exact sizes of the introns and exons are listed in Table 1. The arrows indicate the translational start sites. The translational stop sites are depicted by the asterisks.
|
|
Tissue distribution of the human PANK1 isoforms
Poly(A)+ RNA isolated from various human tissues was used as template in a multiplex RT-PCR experiment to determine the relative expression levels of hPANK1 and hPANK1ß mRNAs (Fig. 3)
. hPANK1 mRNA was expressed in brain, heart, kidney, liver, skeletal muscle, and testis, whereas hPANK1ß was expressed in much lower levels in kidney, liver, brain, and testis, and was not detected in heart and skeletal muscle. The control lane illustrates the expected PCR products using hPANK1 and hPANK1ß cDNA expression plasmids as templates.
hPANK expression in HepG2 cells
Cultured HepG2 cells were chosen for investigation because the PANK1 and PANK1ß isoforms are expressed (see below) in this commonly used cell line, in which fatty acid metabolic gene expression is responsive to PPAR-activating ligands (36). Cells were incubated with BF, a PPAR agonist, at concentrations up to 0.7 mM in complete medium containing 10% serum for 24 h. BF binds to serum components (37), and the BF concentrations used in our work matched those required for a genetic response in this system (38). Incubation with BF at the concentrations indicated did not change the doubling time or confluent cell density of the HepG2 cells (data not shown). Relative expression levels of the hPANK isoforms and the L-FABP- and GAPDH-positive and negative control genes, respectively, were determined by semiquantitative RT-PCR. The basal PANK1 was expressed at a 2-fold higher level compared with PANK1ß using this assay. The hPANK1 mRNA increased significantly in response to 0.5 mM BF (Fig. 4A, B)
and increased up to 5-fold following treatment with 0.7 mM of the drug. In contrast, the levels of hPANK2 and hPANK3 did not change, and hPANK1ß decreased slightly following incubation with the drug. L-FABP mRNA expression was significantly enhanced in response to increasing concentrations of the agonist (Fig. 4A, C), as reported previously for this PPAR-responsive gene (9), whereas GAPDH remained unchanged (Fig. 4A). These data show that BF increased the expression of PPAR target genes in the HepG2 cell system and that hPANK1 was the only PANK isoform that responded to BF in a dose-dependent manner.
hPANK1 mRNA exhibited a detectable increase at 12 h following 0.5 mM BF treatment, and reached a maximum at 24 h that was maintained at 48 h (Fig. 5A, B)
. However, the levels of hPANK2 and hPANK3 expression did not change in response to the drug, and the hPANK1ß message decreased slightly (Fig. 5A, B), in agreement with the BF dose response shown in Fig. 4. The relative level of the control L-FABP mRNA increased following 6 h of BF treatment and was about 6-fold higher after 24 h (Fig. 5A, C). These results indicated that the specific expression of hPANK1 mRNA was responsive to a PPAR agonist in a time-dependent manner.
hPanK1 mRNA half-life
Increased mRNA stability in response to BF treatment may have accounted for the accumulation of hPANK1. This point was tested by treating HepG2 cells with 0.5 mM BF or DMSO (vehicle control) for 20 h, then adding actinomycin D (5 µg/ml) to inhibit further RNA synthesis. RNA samples were prepared at time intervals up to 1 h following actinomycin D treatment, and the hPANK1 mRNA levels were determined by RT-PCR (Fig. 6)
. GAPDH mRNA expression was chosen both as an RNA loading control and as a negative control for actinomycin D treatment, because the half-life of GAPDH message is greater than 20 h in HepG2 cells (39, 40). The level of hPANK1 message decreased with actinomycin D at the same rate in both control and BF-treated cultures (Fig. 6A, B) to about 23% of the initial amount after 1 h of treatment. The mRNA half-lives were determined from the slopes of the lines obtained by linear regression analysis, and there was no significant difference between the hPANK1 message stability in control (40 min) or BF-treated cells (35 min). These data ruled out decreased mRNA degradation as a mechanism for heightened hPANK1 mRNA levels following BF treatment.
Response of the hPANK1 promoter to a PPAR agonist
We next investigated the role of the hPANK1 promoter in the increased expression of hPANK1 mRNA in response to BF. Genomic DNA sequence (2.4 kb) 5' to the transcript start site, called the "PANK1 promoter" region, was evaluated for common promoter sequences and PPRE(s). Four putative PPREs were identified in the PANK1 promoter region (Fig. 7A)
, with midpoints at -771 bp, -1,210 bp, -1,370 bp, and -2,141 bp relative to the transcription start site, which was defined as 0 bp. The putative PPRE present at -1,370 bp matched 100% with the consensus for the human PPRE (41), and the remaining three PPREs deviated from the consensus by one base pair. Reporter gene constructs containing the 2.4 kb of putative promoter sequence (Fig. 7A) plus the untranslated region of exon 1 were made by subcloning the genomic DNA in the sense (P1) and the anti-sense (P1-reverse) orientations relative to the luciferase open reading frame in the pGL3- vector. The activity of the promoter region in both orientations was determined by luciferase assays in HepG2 cell lysates following transfection with the reporter constructs. P1 promoter activity was clearly evident in cells transfected with the sense construct (Fig. 7B), and this activity doubled upon treatment with 0.5 mM BF (Fig. 7B). The P1 promoter cloned in the reverse orientation was inactive. The activities of the pGL3- vector lacking promoter sequences (negative control) and the pGL3+ vector containing the SV40 promoter upstream of the luciferase open reading frame (positive control) did not change in response to the drug. These results indicated that the hPANK1 promoter activity increased in cultured HepG2 cells following treatment with BF, consistent with the BF-elevated hPANK1 mRNA. These data support the conclusion that PPAR elevates hPANK1 expression in response to BF via one or more of the PPREs located in the promoter region.
View larger version (46K):
[in this window]
[in a new window]
|
Fig. 7. Analysis of the hPANK1 promoter region. A: Putative peroxisome proliferator response elements (PPREs) in a 2.4 kb DNA sequence 5' to the transcript start site are indicated. The transcriptional start site is numbered as 0 and the positions of the upstream bases are indicated by negative numbers. The midpoints of the sequences of the PPREs are 5'-AGACTCTGACTCC (-771 bp), 5'-GGAGTCATAAACAGA (-1,210 bp), 5'-GGGAACAGATGCA (-1,370 bp), and 5'-AGGGATAGAGTGA (-2,141 bp). The human PPRE consensus sequence is (A/G, G, G/A, N, C/A/T/G, A/C/G, A/C/G/T, A/G/C, G/T/A, G/T, T/G, C/G, A/C). B: HepG2 cells were transfected with hPANK1 promoter-luciferase reporter plasmids constructed in the sense (P1) and antisense (P1-Rev) orientation and treated with 0.5 mM BF 48 h later. Luciferase activity was measured at 72 h and normalized to cell lysate protein amount in each sample. Gray bars represent the luciferase activity in DMSO-treated cells; cross-hatched bars represent luciferase activity in cells treated with 0.5 mM BF. The simian virus 40 promoter vector (pGL3+) is a positive activity control; the pGL3(-) is a vector control lacking promoter sequence. The error bars represent triplicate data points, and the P value (P < 0.001) was determined using the Sigmastat program. The results are derived from two independent experiments.
|
|
Analysis of the PANK1 promoter using the TRANSFAC database (42) revealed the presence of additional binding sites for transcription factors involved in the regulation of carbohydrate and lipid metabolism. A putative carbohydrate response element was present at -1,110 bp. A carbohydrate response element is present in the promoter of glycolytic genes such as pyruvate kinase and lipogenic genes such as S14 (43–46). A putative sterol-regulatory element binding protein (SREBP) site was present at -1,075 bp. This site is present in the promoter regions of genes involved in coordinating glucose and lipid metabolism (46–48). A putative glucose response element (GlRE) or upstream stimulatory factor binding site (E-box) was found at -337 bp, similar to the GlRE identified in a PPAR promoter (49, 50). This analysis suggests that several other regulatory factors may control hPanK1 expression. In particular, the regulation by glucose through the GlRE element and by lipids via the SREBP1 site appear to be the most important to investigate in the future in light of the central importance of CoA to lipid and carbohydrate metabolism.
hPANK1 cDNA encodes a functional PanK
The human PANK1 open reading frame is larger than the homologous mouse sequence and encodes an additional 38 amino acids at the N terminus of the protein. To confirm that the larger human PanK1 protein was, in fact, active in the cellular context, the cDNA was assembled and cloned into pcDNA3.1(-), a mammalian expression vector. The hPanK1 protein is made up of 598 amino acids (Fig. 2), with 235 amino acids arising from exon 1 and the remaining 363 amino acids derived from exons 2 through 7 (Fig. 1). The calculated molecular mass of the protein is 65.8 kDa. COS7 cells were transfected with the hPanK1 cDNA, and cell lysates were analyzed for PanK enzyme activity (Fig. 8A)
. The hPANK1-transfected cells exhibited significantly higher PanK-specific activity, compared with control cells transfected with vector plasmid alone. Control COS7 cell lysates had very low activity, detectable only at concentrations >250 µg protein (data not shown). Our affinity-purified rabbit polyclonal PanK1 anti-peptide antibody detected the hPanK1 protein in the crude lysate from hPANK1-transfected cells with an estimated size of 66 kDa (Fig. 8B), consistent with the calculated size from the cDNA sequence. The endogenous protein was not detected using the immunoblotting technique at protein loads up to 88 µg/lane (not shown). This observation is consistent with the low enzyme-specific activity in cell lysates (see below) and indicates that hPanK1 is not an abundant protein in HepG2 cells. This conclusion is consistent with a rate-controlling role for PanK1 in the CoA biosynthetic pathway.
BF elevated hPanK1 protein levels in HepG2 cells
HepG2 cells were treated with DMSO or BF for 72 h and examined for the expression and localization of hPanK1 protein using immunofluorescence and confocal microscopy (Fig. 9)
. The cells exhibited fluorescence in the cytoplasm (green fluorescence, Fig. 9, left); the nuclei were stained red with propidium iodide (Fig. 9, middle), and the overlay of the two probes is shown in Fig. 9, right. The fluorescence pattern is largely diffuse, with some punctate staining, indicating that the majority of protein is cytosolic. The punctate staining suggested that a portion of the protein may be associated with a subcellular organelle. However, we do not think that this is the case, because double staining of the cells with anti-hPanK1 antibody and mitotracker dye (specific for mitochondria) or anti-catalase antibody (specific for peroxisomes) showed that the majority of hPanK1 did not localize with the mitochondrial or peroxisomal markers (data not shown), and the punctate staining remained in the following controls. Two control experiments were performed to determine if components of the hPanK1 staining were nonspecific. First, the PanK1 antibody was preincubated with the peptide epitope prior to staining the cells. These results (Fig. 9G, H, I) show a significant reduction in total fluorescence, confirming that the majority of the signal arose from antibody binding to the hPanK1 epitope. Second, the extent of nonspecific staining by the FITC-coupled secondary antibody was determined by incubation with this secondary antibody alone. These results showed even lower levels of cytoplasmic fluorescence with secondary antibody alone (Fig. 9J, K, L). In both controls, the punctate green fluorescence was present while the diffuse cytoplasmic green fluorescence was not present, leading to the conclusion that hPanK1 protein was cytoplasmic and the punctate staining was nonspecific.
View larger version (116K):
[in this window]
[in a new window]
|
Fig. 9. Immunofluorescence/confocal images of hPanK1 protein expression in HepG2 cells. The cells were treated with DMSO (A, B, C) or 0.5 mM BF (D–L) for 72 h, and the hPanK1 protein was visualized by immunofluorescence and confocal microscopy. The panels on the left show fluorescence in the green channel (hPanK staining); panels in the middle show red fluorescence (nuclear staining); and panels on the right represent overlay images of left and middle panels. In DMSO-treated cells, cytoplasmic green fluorescence is visible (A, B, C). In BF-treated cells, there was an increase in the cytoplasmic fluorescence (D, E, F). Controls in which the primary antibody was incubated with 5-fold molar excess of the PanK antigenic epitope peptide before staining show a reduction in the cytoplasmic fluorescence intensity (G, H, I). Minimal cytoplasmic fluorescence is seen in cells stained with the secondary FITC-coupled antibody only (J, K, L). The gain settings for A–F were 605 (green) and 679 (red). For G–I, the gain settings were 594 (green) and 647 (red). The gain settings for J–L were 718 (green) and 679 (red).
|
|
The levels of hPanK1 were quantitated in control and BF-treated cells by measuring the fluorescence intensities in confocal images (Fig. 10)
. Measurements were made in cells treated with either DMSO or BF at intervals between 0 h and 72 h and the relative fluorescent intensities plotted. In cells treated with BF, the expression of hPanK1 protein increased by a factor of 1.7 at 24 h and was maintained at this elevated level up to 72 h (see also Fig. 9D, E, F). The protein levels did not significantly change between 0 h and 72 h in DMSO-treated cells. These data are consistent with an increase in cellular PanK1 protein occurring in response to BF treatment.
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 10. Quantitation of hPanK1 protein expression. The hPanK1 protein amount was quantified for 20 cells per time point in duplicate using ImageProPlus software as described under Experimental Procedures. Mean fluorescence intensity of hPanK1 in DMSO control cells (closed circle) or in BF-treated cells (open circle) was determined using the gain settings reported in Fig. 9A–F. The error bars represent duplicate data points.
|
|
BF increased cellular PanK-specific enzyme activity
HepG2 cells were treated with DMSO or 0.5 mM BF for times up to 48 h, and the total PanK enzyme specific activity was determined in crude cell lysates. In the DMSO-treated cells, there was no significant change in PanK enzyme activity between 0 h and 72 h of treatment (Fig. 11)
. BF treatment triggered an increase in PanK enzyme-specific activity as a function of time to a level that was approximately twice that in the control at 48 h. The PanK enzyme activity increase in response to BF was consistent with the magnitude of the observed increase in PanK1 mRNA (Fig. 5) and PanK1 protein expression (Fig. 10) levels.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 11. BF-elevated cellular PanK-specific enzyme activity. HepG2 cells were untreated (0 h) or treated with DMSO or 0.5 mM BF for 12 h, 24 h, and 48 h. PanK enzyme activity was determined using 350 µg soluble lysate protein using the assay described under Experimental Procedures. The results represent two independent experiments performed in duplicate. The error bars represent duplicate data points.
|
|
BF elevated cellular CoA and P-PanSH levels
Metabolic radiolabeling of HepG2 cells with [3H]Pan was performed to investigate the impact of BF-stimulated PanK expression on CoA biosynthesis. Cells were harvested at various times between 12 h and 72 h of BF treatment, and the soluble extract was fractionated by TLC to identify the cellular levels of CoA and various intermediates in the biosynthetic pathway. The major radiolabeled cellular pools were Pan, P-PanSH, and CoA, and these three intermediates constituted >95% of the total. There was no significant change in the total amount or ratio of phosphorylated CoA pathway intermediates in control cells between 12 h and 72 h, indicating that the radiolabeling had reached metabolic equilibrium by 12–24 h. In contrast, the phosphorylated PanK products increased after 12 h of BF treatment, reaching a maximum at 24 h that was 3.6x higher than in control cells (Fig. 12A)
, and this level was maintained up to 72 h. The amount of cellular CoA in BF-treated cells increased to a level 2.6-fold higher than that in DMSO-treated controls (Fig. 12B), and the P-PanSH increased 3.6-fold after 24 h of BF exposure (Fig. 12C). These data show that BF-stimulated PanK expression was reflected in a higher steady-state CoA level in HepG2 cells. These data are consistent with a key rate-limiting role for PanK in the CoA biosynthetic pathway, and the finding that P-PanSH also was elevated in response to BF points to the P-PanSH adenylyltransferase as a second control point in the CoA biosynthetic pathway.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 12. BF increased intracellular CoA and 4'-phosphopantetheine (P-PanSH). HepG2 cells were labeled with D-[2,3-3H]pantothenic acid (33.3 nM, specific activity 25 Ci/mmol) in pantothenate (Pan)-free DMEM in the presence of DMSO or 0.5 mM BF for 0 h, 12 h, 24 h, 48 h, and 72 h. The distribution of radiolabeled Pan metabolites was quantified in cell extracts using TLC as described under Experimental Procedures, and radioactivity was normalized to total protein in the cell extract. A: Total phosphorylated Pan metabolites in control (closed circle) and BF- (open circle) treated cells. B: CoA levels in cells treated with DMSO (closed circle) or BF (open circle). C: P-PanSH levels in control (closed circle) or BF-treated (open circle) cells. The results represent two independent experiments performed in duplicate. The error bars represent duplicate data points.
|
|
|
DISCUSSION
|
|---|
Our results point to the PPAR-mediated regulation of PanK1 expression as an important mechanism underlying the modulation of hepatic CoA levels in response to the physiological status of the animal (51). The metabolic state (starvation or feeding) (15–19), pathological conditions (diabetes) (20, 21), and hypolipidemic drugs (fibrates) (17, 22, 23) all significantly alter the cellular levels of CoA, and these fluctuations are reflected by concomitant changes in tissue PanK activity (13, 14). CoA participates in most of the reactions catalyzed by PPAR target genes (i.e., fatty acid ß-oxidation), and CoA and its thioesters are also regulators of key enzymes in intermediary metabolism, such as pyruvate dehydrogenase, citrate lyase, and carnitine palmitoyl transferase. Thus, the observed adjustments in the concentration of CoA/acetyl-CoA likely contribute to the redirection of carbon flux that occurs during fasting, in diabetes, and following treatment with hypolipidemic drugs. The human PANK1 gene structure (Fig. 1) reveals that the PanK1 and PanK1ß transcripts arise from the utilization of alternate transcriptional start sites, as reported in the mouse (9). The 1ß exon is removed by splicing the 1 exon to exon 2 to yield the PanK1 transcript. A survey of human tissue mRNA reveals that PanK1 expression predominates (Fig. 3), and BF stimulation of PanK expression is selective for the PanK1 isoform in the HepG2 cultured cell model (Fig. 4). The presence of four DNA sequences that correspond to PPREs in the 1 promoter region (Fig. 7A), coupled with enhanced PanK1 promoter activity in response to BF treatment of HepG2 cells (Fig. 7B), points to transcriptional regulation of PanK1 as a primary mechanism for the control of CoA levels by hypolipidemic agents. The elevation of PanK1 mRNA (Figs. 4, 5) leads to the concomitant increase in PanK1 protein, PanK enzyme activity, and CoA content (Figs. 10, 11, 12). The elevation of PanK1 expression and CoA content in the HepG2 model may be lower than that observed in liver, because HepG2 cells express limited amounts of PPAR protein (52). Thus, our model for CoA tissue regulation by PPAR-regulated PanK1 expression is consistent with the need to significantly increase the availability of CoA to support the accelerated processing of CoA thioesters through the ß-oxidation pathway.
The biochemical analysis of allosteric PanK control (5, 8, 9, 53, 54), coupled with the genetic regulatory mechanism described in this report, points to PanK as a master switch in CoA production. Pan is usually present in excess and is rapidly taken up by cells via an active transport system (55, 56). Thus, the intracellular level of CoA is controlled by the amount of cellular PanK activity, which is a function of both the stringency of feedback regulation and the total amount of PanK protein. Feedback inhibition of PanK by CoA is a homeostatic mechanism that sets the upper limit for the intracellular CoA concentration at a given level of PanK expression (5, 8, 9, 53, 54). Feedback inhibition is critical, because in the absence of regulation, the CoA levels would depend on the dietary supply of Pan and would rise until all available Pan is converted to CoA. The level of PanK expression establishes the intracellular level of CoA by changing the set point established by feedback regulation. Increased expression of PanK will cause the CoA/CoA thioester pool to expand until it reaches a new steady-state level that effectively switches off the flux through the pathway via the feedback inhibition loop. The fact that mammalian PanK isoforms have different sensitivities to feedback regulation (9) illustrates that CoA levels will change by different degrees, depending not only on the amount of PanK expressed but also on which isoform is being genetically regulated. The accumulation of P-PanSH following increased PanK expression (Fig. 12) (8) indicates that P-PanSH adenylyltranferase, CoA phosphodiesterase, and/or organelle-specific transport systems (see below) are a second site for controlling total CoA content. This secondary regulatory point has been observed before using plasmid-driven overexpression of both the bacterial PanK protein (57) and the mouse PanK1ß (8). Bacterial P-PanSH adenylyltransferase is likely feedback inhibited by CoA (58), but this property has not been explored in the related mammalian multifunctional CoA synthase.
The compartmentation of the CoA biosynthetic pathway is of considerable interest, because the highest concentrations of CoA exist in organelles. In general, cytosolic concentrations of CoA are estimated at between 0.02 and 0.06 mM, whereas mitochondrial concentrations range from 2.2 to 2.6 mM (59, 60). Peroxisomes also have high concentrations of CoA (61, 62). The requirement for high levels of CoA in peroxisomes is consistent with the increases in CoA biosynthesis and tissue CoA levels following treatment with peroxisome proliferators (17). The observation that fibrates do not elicit their hypolipidemic effects in PAN-deficient animals (63) illustrates that accelerated CoA production is essential to the biogenesis of functional peroxisomes. Our model for CoA elevation via PPAR-stimulated PanK1 expression (22, 23) provides a mechanism for the boosting of CoA biosynthesis to supply the growing CoA demand of the expanding peroxisomal compartment. Our work indicates that hPanK1 is a cytosolic protein (Fig. 9), and sequence analysis did not reveal the presence of known peroxisome targeting signals. These data suggest either that intact CoA is transported into peroxisomes or that peroxisomes contain a CoA synthase (P-PanSH adenylyltransferase/dephosphoCoA kinase) and import P-PanSH from the cytosol. Recently, a CoA hydrolase was identified in peroxisomes (64, 65), which provides a mechanism for this organelle to decrease the CoA concentration. Clearly, the regulation of organelle CoA content is complex and likely governed by the interplay among CoA biosynthesis, transport, and degradation.
The importance of CoA compartmentation is highlighted by the recent discovery that Hallervorden-Spatz syndrome (HSS), an autosomal recessive neurodegenerative disorder (66), likely arises from dysregulated mitochondrial CoA metabolism. The disorder is associated with the deposition of iron in the basal ganglia (67) and serves as a model for complex neurodegenerative diseases, such as Parkinson's disease (68), Alzheimer's disease (69), and Huntington's disease (70), which, like HSS, all exhibit pathologic accumulation of iron in the brain. Mutations that inactivate one isoform of human PanK, PANK2, are the causative agent of HSS, and the disease was renamed PanK-associated neurodegeneration (10). The similarity of the pathology of PKAN to that of the disorders listed above led Zhou et al. (10) to suggest that perturbed Pan metabolism may be a general feature of neurodegenerative diseases. Importantly, one of the isoforms encoded by the PANK2 gene localizes to mitochondria (11). This discovery suggests that mitochondria may have the entire complement of CoA biosynthetic enzymes. Previous work hypothesized that mitochondrial CoA arises either from P-PanSH transport followed by conversion to CoA by a mitochondrial form of the bifunctional CoA s
TOP
ABSTRACT
INTRODUCTION
