Fasting-Induced Transcription Factors Repress Vitamin D ...Fasting-Induced Transcription Factors Repress Vitamin D Bioactivation, a Mechanism for Vitamin D Deﬁciency in Diabetes

Fasting-Induced Transcription Factors Repress Vitamin DBioactivation, a Mechanism for Vitamin D Deficiency inDiabetesSanna-Mari Aatsinki,1,2,3 Mahmoud-Sobhy Elkhwanky,1,2 Outi Kummu,1,2 Mikko Karpale,1,2 Marcin Buler,1,2

Pirkko Viitala,1 Valtteri Rinne,3 Maija Mutikainen,4 Pasi Tavi,4 Andras Franko,5,6,7 Rudolf J. Wiesner,5

Kari T. Chambers,8 Brian N. Finck,8 and Jukka Hakkola1,2

Diabetes 2019;68:918–931 | https://doi.org/10.2337/db18-1050

Low 25-hydroxyvitamin D levels correlate with the prev-alence of diabetes; however, the mechanisms remainuncertain. Here, we show that nutritional deprivation–responsive mechanisms regulate vitamin D metabolism.Both fasting and diabetes suppressed hepatic cytochromeP450 (CYP) 2R1, the main vitamin D 25-hydroxylase respon-sible for the first bioactivation step. Overexpression ofcoactivator peroxisome proliferator–activated receptorg coactivator 1-a (PGC-1a), induced physiologically by fast-ing and pathologically in diabetes, resulted in dramaticdownregulation of CYP2R1 in mouse hepatocytes in anestrogen-related receptor a (ERRa)–dependent manner.However, PGC-1a knockout did not prevent fasting-inducedsuppression of CYP2R1 in the liver, indicating that additionalfactors contribute to the CYP2R1 repression. Furthermore,glucocorticoid receptor (GR) activation repressed the liverCYP2R1, suggesting GR involvement in the regulation ofCYP2R1. GR antagonist mifepristone partially preventedCYP2R1 repression during fasting, suggesting that gluco-corticoids and GR contribute to the CYP2R1 repressionduring fasting. Moreover, fasting upregulated the vitaminD catabolizing CYP24A1 in the kidney through the PGC-1a-ERRa pathway.Our studyuncovers amolecularmechanismfor vitamin D deficiency in diabetes and reveals a novelnegative feedback mechanism that controls crosstalk be-tween energy homeostasis and the vitamin D pathway.

Vitamin D is an endocrine regulator of calcium homeosta-sis, but emerging evidence has indicated a role in theregulation of energy homeostasis (1,2). There is strongevidence that vitamin D affects adipogenesis and adipocytelipid metabolism; however, the effects vary between speciesand cell models used (1,3). Both Vdr2/2 and Cyp27b12/2

mice display a lean phenotype with decreased fat mass,probably because of increased energy expenditure (1,4). Incontrast, in humans, low 25-hydroxyvitamin D (25-OH-D)associates with obesity (1,2).

Vitamin D deficiency is a widespread medical healthproblem worldwide (5). Ample epidemiological evidenceassociates vitamin D deficiency with the prevalence ofmetabolic diseases, and low 25-OH-D levels have beenreported to correlate with the incidence of type 1 and type2 diabetes (6–9). However, the causal relationship is un-certain, and the mechanisms involved are unclear. Fur-thermore, the intervention studies aiming at diabetesprevention with vitamin D supplementation have beenlargely disappointing or inconclusive (6).

Vitamin D is a prohormone activated in two enzymaticsteps: 25-hydroxylation in the liver and 1a-hydroxylationin the kidney to produce the main active form 1a,25-dihydroxyvitamin D (1a,25-(OH)2-D) (10). Vitamin D sta-tus is usually assessed by measuring the main circulating

1Research Unit of Biomedicine, Pharmacology and Toxicology, University of Oulu,Oulu, Finland2Medical Research Center Oulu, Oulu University Hospital and University of Oulu,Oulu, Finland3Admescope Ltd., Oulu, Finland4A.I. Virtanen Institute for Molecular Sciences, University of Eastern Finland,Kuopio, Finland5Institute of Vegetative Physiology, Medical Faculty, University of Köln, Köln,Germany6Institute for Clinical Chemistry and Pathobiochemistry, Department for DiagnosticLaboratory Medicine, University Hospital Tübingen, Tübingen, Germany7German Center for Diabetes Research, Neuherberg, Germany8Department of Medicine, Washington University School of Medicine, St. Louis, MO

Corresponding author: Jukka Hakkola, [email protected]

Received 28 September 2018 and accepted 26 February 2019

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1.

S.-M.A. and M.-S.E. contributed equally to this work.

© 2019 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, and thework is not altered. More information is available at http://www.diabetesjournals.org/content/license.

918 Diabetes Volume 68, May 2019

METABOLISM

https://doi.org/10.2337/db18-1050http://crossmark.crossref.org/dialog/?doi=10.2337/db18-1050&domain=pdf&date_stamp=2019-04-09mailto:[email protected]://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://www.diabetesjournals.org/content/licensehttp://www.diabetesjournals.org/content/license

vitamin D metabolite (i.e., 25-OH-D) (5). Cytochrome P450(CYP) 2R1 is the predominant vitamin D 25-hydroxylase inthe liver (11,12). A genetic defect in the CYP2R1 gene hasbeen shown to cause an inherited form of vitamin D de-ficiency and rickets in children (13–15). Large-scale studieshave identified CYP2R1 gene variants as one of the majorgenetic determinants of low 25-OH-D levels (16–18). Ad-ditionally, polymorphism in the CYP2R1 gene has beenassociated with susceptibility to type 1 diabetes (19,20).However, little is known about the regulation of CYP2R1,and vitamin D 25-hydroxylation has even been consideredan unregulated step (10).

CYP24A1 enzyme catalyzes 24-hydroxylation of 1a,25-(OH)2-D and 25-OH-D (10). 24-Hydroxylation is the maininactivation step limiting the effect of the main vitamin Dreceptor (VDR) ligand 1a,25-(OH)2-D but also reduces thepool of 25-OH-D available for 1a-hydroxylation. CYP24A1plays an important role in the regulation of vitamin Daction and is under stringent regulatory control (10).

In this report, we show that CYP2R1 and its catalyticactivity vitamin D 25-hydroxylation are suppressed in theliver during fasting and in both type 1 and type 2 diabetesmouse models. Mechanistically, we demonstrate involve-ment of at least two molecular pathways: the peroxisomeproliferator–activated receptor g coactivator 1-a (PGC-1a)/estrogen-related receptor a (ERRa) axis and the glucocor-ticoid receptor (GR). Furthermore, CYP24A1 is inducedduring fasting under the control of PGC-1a. Altogether,these results indicate that energy metabolism–regulatingfactors control vitamin D metabolism and establish repres-sion of vitamin D bioactivation as an important, novelmechanism inducing vitamin D deficiency in diabetes.

RESEARCH DESIGN AND METHODS

Animal ExperimentsAll animal experiments were approved by the NationalAnimal Experimental Board, Finland, or the local animalcommittees (the Animal Studies Committee of Washing-ton University School of Medicine, St. Louis, MO, and theAnimal Use and Care Committee of Semmelweis Univer-sity, Budapest, Hungary) or the local government author-ities (Bezirksregierung Köln/Landesamt für Natur, Umweltund Verbraucherschutz, North Rhine-Westphalia, Cologne,Germany). All animals were housed in standard conditionswith a 12-h dark-light cycle. If not otherwise stated, at theend of the experiments, the mice were killed with CO2inhalation and neck dislocation, and tissues were collectedand frozen in liquid nitrogen. All animals were male.

Fasting ExperimentsDBA/2 male mice, obtained from the Laboratory AnimalCenter, University of Oulu, aged 8–12 weeks, were fastedfor 12 or 24 h. The animals had free access to drinkingwater. At the end of the experiment, the mice were anes-thetized with a solution containing fentanyl-fluanisone(Hypnorm) and midazolam (Dormicum) and sacrificed. In-dependent repetition of the experiment gave similar results.

Pgc-1a2/2 mice (21), aged 3–4 months, and Pgc-1a+/+

littermates were fasted for 12 h, after which the animalswere sacrificed and the tissues collected. Liver-specificPgc-1b2/2mice (22) or Pgc-1b+/+ littermates, aged 8 weeks,were housed individually and fasted for 24 h. Animals weresacrificed and tissues harvested and snap frozen in liquidnitrogen. Both knockout (KO) lines were in the C57BL/6background. Wistar rats, aged 10 weeks, were fasted for24 h as has been described previously (23).

High-Fat Diet TreatmentC57BL/6 mice, aged 8 weeks, were fed a high-fat diet (HFD)(60% fat, TD.06414; Envigo) or regular chow (5% fat,TD.2018; Envigo) for 16 weeks. The mice were fastedfor 12 h overnight, and fasting blood glucose was deter-mined. The mice were sacrificed by CO2 inhalation, andblood was drawn into an EDTA-primed syringe from venacava and tissues collected. The chow diet and HFD contained39.39mg/kg and 52.35mg/kg vitamin D, respectively. Theestimated amount of the chow diet eaten by a mouse was4 g/day. However, the HFD-treated mice ate less (i.e.,3.1 g/day). Therefore, the daily intake of the vitamin D was;0.16 mg/day for both the chow and the HFD-treated mice.

Streptozotocin TreatmentExperimental type 1 diabetes was induced by intraperito-neal (i.p.) injections of C57BL/6 mice with streptozotocin(STZ) as has been described previously (24,25).

Treatment With Nuclear Receptor Agonists andInhibitorsFor ERRa inhibition in the kidney, the C57BL/6 mice, aged8 weeks, were treated i.p. with an ERRa inverse agonist,XCT790 (0.48mg/kg) dissolved in DMSO/corn oil or vehicleonce daily for 3 days. Twelve hours after the last XCT790injection, mice were either fed or fasted overnight for anadditional 12 h, and the tissues were collected.

For the dexamethasone treatment, C57BL/6 mice, aged8 weeks, were fasted 1 h before an i.p. injection ofdexamethasone (3 mg/kg) dissolved in DMSO/corn oilor vehicle only. The mice were fasted 6 h and the tissuescollected. In some experiments, simultaneously with thedexamethasone injections, the GR was inhibited with theGR antagonist mifepristone (i.p. 100 mg/kg) dissolved inDMSO/corn oil. In all the treatment groups, the vehicleamount was kept similar.

To inhibit GR in the liver of fasting mice, C57BL/6 mice,aged 8–9 weeks, were injected with mifepristone (i.p.100 mg/kg) or vehicle (DMSO/corn oil) two times (the firstinjection at 9 A.M. and the second injection at 9 P.M.), andsubsequently, the mice were either fed or fasted overnightfor an additional 12 h, and the tissues were collected.

Cell CultureMouse primary hepatocytes were isolated from maleDBA/2 (Figs. 2A and 4A) or C57BL/6 mice (LaboratoryAnimal Center, University of Oulu), aged 8–10 weeks, asdescribed (26) and cultured inWilliam’s Emedium containing

diabetes.diabetesjournals.org Aatsinki and Associates 919

2.0 g/L D-glucose and insulin 5 mg/L, transferrin 5 mg/L,sodium selenate 5 mg/L, and 10% FBS (Sigma-Aldrich, St.Louis, MO). The cultures were maintained for an additional12–24 h in serum-free William’s E medium before adenoviralinfections or chemical treatments. HepG2-cells (ATCC, Man-assas, VA) were maintained in basic DMEM (4.5 g/L glucose)supplied with 10% FBS and 1% penicillin-streptomycin (allGibco, Invitrogen, Carlsbad, CA).

RNA Preparation and Quantitative RT-PCRFrom fasted male DBA/2 mouse livers, RNAs wereextracted with the cesium chloride centrifugation method(27). From all the other tissues and cell samples, total RNAextraction was performed using either TRI reagent orRNAzol reagent (Sigma-Aldrich) according to the manu-facturer’s protocol. One microgram of RNA was used forcDNA synthesis using p(dN)6 random primers (RocheDiagnostics, Mannheim, Germany) and Maloney murineleukemia virus reverse transcriptase (Promega, Madison,WI) or RevertAid Reverse Transcriptase (Thermo FisherScientific, Waltham, MA). The quantitative PCR (qPCR)reactions were performed using SYBR Green chemistry orTaqMan chemistry (Applied Biosystems, Foster City, CA).The sequences for the primers and TaqMan probes arelisted in Table 1. The fluorescence values of the qPCRproducts were corrected with the fluorescence signals ofthe passive reference dye (ROX). The mRNA levels oftarget genes were normalized against 18S rRNA, GAPDH,

or TBP control levels using the comparative CT (DDCT)method.

Adenoviruses and Short Hairpin RNA Knockdown inCellsPGC-1a-2x9 and PGC-1a-L2L3M plasmids were pro-vided by Dr. Donald McDonnell (Duke University Schoolof Medicine, Durham, NC). Recombinant adenovirusesexpressing green fluorescent protein (GFP-Ad), LacZ(LacZ-Ad), and PGC-1a (PGC-1a-Ad, PGC-1a-2x9-Ad,PGC-1a-L2L3M-Ad) were prepared as described previously(26). For overexpression of PGC-1a, PGC-1a-2x9, andPGC-1a-L2L3M, multiplicity of infection (MOI) 0.5 wasused for each virus. ERRa-Ad was purchased from VectorBiolabs (Malvern, PA). Mouse primary hepatocytes wereinfected with adenoviruses in William’s E growth mediumwithout serum for the indicated time periods before RNAor protein extractions. Adenoviruses containing scram-bled short hairpin RNA (shRNA) (shScr-Ad) and ERRa-targeting shRNA (shERRa-Ad) were purchased from VectorBiolabs. For the shRNA experiments, mouse primaryhepatocytes were first infected with either shScr orshERRa at MOI 30 in William’s E medium for 24 h, afterwhich cells were infected with either PGC-1a-Ad or con-trol virus LacZ-Ad at MOI 2. After 48 h, the cells werecollected and RNA isolated. The efficiency of the knock-down was tested by measuring ERRa mRNA by qPCR.For inhibition of the ERRa by XCT790 in combination

Table 1—Sequences of the PCR primers

Gene Forward primer (59–39) Reverse primer (59–39)

mPGC-1a GCAGGTCGAACGAAACTGACa CTCAGCCTGGGAACACGTTAa

TCCTCCTCATAAAGCCAACCb GCCTTGGGTACCAGAACACTb

CTGCTCTGGTTGGTGAGGAc GCAGGCTCATTGTTGTACTGc

AGCCGTGACCACTGACAACGAGd GCTGCATGGTTCTGAGTGCTAAGd

PEPCK GGTGTTTACTGGGAAGGCATC CAATAATGGGGCACTGGCTG

mERRa ATCTGCTGGTGGTTGAACCTG AGAAGCCTGGGATGCTCTTG

mCYCS CCAAATCTCCACGGTCTGTTC ATCAGGGTATCCTCTCCCCAG

mATP5B GGTTCATCCTGCCAGAGACTA AATCCCTCATCGAACTGGACG

mVDR GAATGTGCCTCGGATCTGTGG GGTCATAGCGTTGAAGTGGAA

rCYP2R1 AAACTACAACCAATGTGCTCCG CTTCCCAAGAAGGCCTCCTGT

18S CGCCGCTAGAGGTGAAATTC CCAGTCGGCATCGTTTATGG

TBP GAATATAATCCCAAGCGATTTG CACACCATTTTTCCAGAACTG

mPGC-1b CTCCAGGCAGGTTCAACCC GGGCCAGAAGTTCCCTTAGG

mGR AGCTCCCCCTGGTAGAGAC GGTGAAGACGCAGAAACCTTG

mNR1D1 TCCCCAAGAGAGAGAAGCAA CTGAGAGAAGCCCACCAAAG

mTAT TGCTGGATGTTCGCGTCAATA CGGCTTCACCTTCATGTTGTC

mGAPDH GGTCATCATCTCCGCCCC TTCTCGTGGTTCACACCCATC

mG6PASE CATCAATCTCCTCTGGGTGG TGCTGTAGTAGTCGGTGTCC

mCYP2R1 Mm01159414_m1 (Life Technologies)

mCYP24A1 Mm00487245_m1 (Life Technologies)aUsed for measuring PGC-1amRNA in Figs. 4C and 5G. bUsed for measuring PGC-1amRNA in Fig. 3A and E and Supplementary Fig. 2L.cFor adenoviral expressed PGC-1a detection. dUsed for measuring PGC-1a mRNA in Fig. 5C and Supplementary Fig. 2I.

920 Diabetes Represses Vitamin D Activation Diabetes Volume 68, May 2019

http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1

with the PGC-1a overexpression, PGC-1a-Ad was used atMOI 2.

Western BlotCYP2R1 was detected from mouse liver microsomal frac-tions. Microsomal fractions were prepared by differentialcentrifugation (28), and protein content was measured byBradford reagent (Bio-Rad, Hercules, CA). CYP24A1 wasdetected from total protein fractions prepared as describedpreviously (29). Protein fractions were subjected to precastSDS-PAGE (10–12% in polyacrylamide) (Bio-Rad) and trans-ferred to polyvinylidene fluoride or nitrocellulose membrane(Millipore, Billerica, MA). Membranes were incubated withappropriate primary antibody in 5% skimmed milk or Amer-sham ECL Prime Blocking Reagent in Tris-buffered salinewith 0.1% Tween, usually overnight, followed by secondaryhorseradish peroxidase–conjugated antibody incubation. Theimmunoreactive bands were visualized with Chemilumines-cent Peroxidase Substrate-1 reaction (Sigma-Aldrich), andAmersham ECL start Western Blotting Detection Reagent(GE Healthcare, Little Chalfont, U.K.) and quantified byQuantity One or Image Studio software.

Vitamin D 25-Hydroxylase AssayMouse liver microsomal samples (0.5 mg/mL protein) weresubjected to incubation with 2 mmol/L cholecalciferoltogether with 0.1 mol/L PBS and preincubated for5 min. Enzymatic reactions were initiated with the addi-tion of 0.5 mmol/L NADPH and quenched using ice-coldacetonitrile at a 1:1 volume ratio after 40-min shaking at37°C. Samples were mixed well and kept at 220°C untilanalyzed using liquid chromatography–tandem mass spec-trometry (LC-MS/MS).

LC-MS/MSQuantitative analysis of 25-OH-D was performed by re-versed-phase LC (ACQUITY UPLC; Waters, Milford, MA)combined with MS detection (Xevo T-QS triple quadrupolemass spectrometer; Waters). The chromatographic separa-tion was carried out with an ACQUITY UPLC BEH ShieldRP18 column (2.13 50mm, 1.7 mm) (Waters). The columntemperature was set to 45°C, and a gradient elution withmobile phase A (0.5% formic acid) and mobile phase B(15% isopropanol and 85% acetonitrile) at a flow rate of500 mL $min21 was used. The elution gradient consisted ofraising the part of mobile phase B from 20% up to 90% in3.5 min followed by column equilibration until 4.5 min. Theinjection volume was 4 mL. The retention time of 25-OH-Dwas 2.82 min. The monitored fragmentation reaction wascharge/mass ratio 383 . charge/mass ratio 211. Positivemode of electrospray ionization was used as the ionizationsource. Data were processed with MassLynx MS version 4.1software (Waters).

Measurement of the Plasma 25-OH-DThe plasma level of the 25-OH-D was measured usinga 25-Hydroxy Vitamin Ds EIA kit (Immunodiagnostic

Systems, Tyne and Wear, U.K.) by ValiFinn (Oulu, Finland)according to the manufacturer’s protocol.

Bioinformatics Analysis of the Mouse Cyp2r1 GenePromoter and Reporter Gene AssaysThe mouse Cyp2r1 gene promoter ERRa binding sites werepredicted using MatInspector software (30) (Genomatix,Munich, Germany). The 1.2 kb (210 to21,220 bp, relative tothe transcription start site [TSS]) promoter region of themouse Cyp2r1 gene was amplified with PCR from the C57BL/6mouse DNA by using forward primer TTCTCGAG-CTTCAAGCCTTAAAATGATGTGAG (TT is extranucleotidefor efficient binding of the restriction enzyme, CTCGAG isthe XhoI restriction site) and reverse primer TTAAGCTT-CTACGAACCAGTCCGGAGC (TT is extranucleotide for effi-cient binding of the restriction enzyme, AAGCTT is theHindIII restriction site) and inserted upstream of the fireflyluciferase gene reporter in the pGL3-basic vector. The pGL3vector containing the wild-type (WT) promoter construct wassubjected to site-directed mutagenesis using a QuikChange IISite-Directed Mutagenesis Kit to generate a construct con-taining the promoter sequence with mutated ERRa bindingsite (21,117 to 21,122 bp, relative to the TSS).

The Cyp2r1 promoter constructs were transiently trans-fected into HepG2 cells using the FuGene transfectionreagent (Promega) together with the pRL-TK Renilla lucif-erase reporter to normalize for transfection efficiency.Empty pGL3-basic vector was used as a negative control.Twenty-four hours after transfection, PGC-1a and ERRawere overexpressed using the adenovirus at MOI 2. TheLacZ-Ad–infected cells were used as a negative control.Cells were incubated for a further 24 h, and the luciferaseactivities were measured using Dual-Glo Luciferase AssaySystem (Promega) and Varioskan Flash equipment (ThermoFisher Scientific). The firefly luciferase values were nor-malized with the Renilla luciferase values. The data areexpressed as relative to the LacZ-Ad–infected cells.

MicroarrayThe DNA microarray experiment to study the PGC-1a–regulated genes in mouse (C57BL/6) primary hepatocyteswas done as previously described (23). Microarray data canbe accessed at the National Center for BiotechnologyInformation Gene Expression Omnibus (GEO), with theaccession number GSE114485.

Analysis of the Published ChromatinImmunoprecipitation Sequencing DataThe published chromatin immunoprecipitation sequencing(ChIP-seq) data PPARGC1A and ESRRA ChIP-seq in HepG2(accession number GSE31477) (31) and NR3C1 (GR) ChIP-seqin mouse liver (GSE72084) (32) were retrieved and analyzedby using the Cistrome database (33) and visualized using theUniversity of California, Santa Cruz, genome browser.

Statistical AnalysisThe statistical data analysis was performed using Graph-Pad Prism software (GraphPad, La Jolla, CA). Unless


otherwise stated, the comparison of means of two groupswas done by Student two-tailed t test, whereas multiplegroups were compared by one-way ANOVA followed byTukey post hoc test. Differences were considered signifi-cant at P , 0.05.

RESULTS

Fasting and Diabetes Repress CYP2R1, the Vitamin D25-Hydroxylase, in the LiverWe observed that fasting represses CYP2R1 expression inthe mouse liver in vivo. Remarkably, the expression ofCYP2R1mRNAwas strongly repressed by 50% already after12-h fasting and further suppressed by 80% after 24 h (Fig.1A). CYP2R1 expression was regulated in negative correla-tion with the fasting-induced gluconeogenic gene phospho-enolpyruvate carboxykinase (PEPCK) (r = 20.605, P =0.0011) (Fig. 1B and Supplementary Fig. 1A). Furthermore,CYP2R1 protein was effectively decreased at both time pointsto 45% after 12 h and 33% after 24 h compared with controls(Fig. 1C). Consistent with the mRNA and protein results,fasting strongly decreased the liver microsomal vitamin D25-hydroxylase activity to 54% after 12 h and below thedetection level after 24-h fasting (Fig. 1D). However, inaccordance with the long half-life of the metabolite, theplasma level of 25-OH-D was not affected by short-term,12-h fasting compared with the fed controls (SupplementaryFig. 1B). In addition to mouse, fasting repressed CYP2R1expression in rat liver after 24-h fasting (Fig. 1E).

Hepatic CYP2R1 was repressed also in the mouse di-abetes models. In the HFD-induced mouse model of obe-sity and type 2 diabetes (Supplementary Fig. 1C and D),hepatic CYP2R1 mRNA was repressed by 45% (Fig. 1F).Consistent with the CYP2R1 mRNA, the plasma level of25-OH-D was significantly reduced in the HFD-treatedmice compared with the chow-fed controls (Fig. 1G). Alsoin the type 1 diabetic mouse model (Supplementary Fig. 1E),induced with STZ, CYP2R1 mRNA was repressed by 43%and protein by 29% (Fig. 1H and I). Analysis of publishedmicroarray data (accession number GSE39752) supports thefinding that CYP2R1 is repressed in the livers of STZ-treatedmice (Supplementary Fig. 1F). Altogether, these data showa clear modulation of vitamin D bioactivation by themetabolic state. CYP2R1 expression and the vitamin D25-hydroxylation were markedly repressed in the livers offasted as well as diabetic animals.

PGC-1a-ERRa Pathway Represses CYP2R1ExpressionWe next investigated the mechanisms of CYP2R1 repres-sion and hypothesized that nutrition-responsive coactiva-tor PGC-1awould be involved in this process since PGC-1aplays a central role in the fasting response and in un-controlled diabetes (34). PGC-1a overexpression in mouseprimary hepatocytes with PGC-1a-Ad downregulatedCYP2R1 strongly and dose dependently, resulting inonly an 11% expression at MOI 1 compared with GFP-Adcontrol (Fig. 2A). To explore the mechanism in more detail,

we transduced mutant PGC-1a into hepatocytes (Supple-mentary Fig. 2A). Gaillard et al. (35) described PGC-1amutants selective for nuclear receptor interactions; PGC-1a-L2L3M mutant is unable to bind any nuclear receptors,whereas PGC-1a-2x9 mutant interacts selectively with nu-clear receptors ERRa or HNF-4a. Interestingly, PGC-1a-2x9 downregulated CYP2R1 expression almost similarly tothe WT (Fig. 2B), whereas the L2L3M mutation abolishedthe CYP2R1 repression and even resulted in weak induction(Fig. 2B). These results indicate that an interaction witha nuclear receptor, most probably ERRa, is indispensable forPGC-1a–mediated CYP2R1 suppression. Supporting thishypothesis, several ERRa target genes (35) were upregu-lated by the WT and the PGC-1a-2x9 mutant (Supplemen-tary Fig. 2B–D). Moreover, the majority of the PGC-1a-2x9mutant–induced genes are dependent on ERRa (35).

Several approaches were used to explore the role ofERRa in the regulation of CYP2R1. First, we performedERRa knockdown by shERRa-Ad virus combined withPGC-1a-Ad treatment in mouse primary hepatocytes.PGC-1a-Ad induced the expression of ERRa about four-fold, as expected (36) (Supplementary Fig. 2E and F). ERRaknockdown abolished CYP2R1 suppression by PGC-1a-Ad(Fig. 2C). Furthermore, we confirmed the role of ERRa byusing ERRa inverse agonist XCT790. Indeed, 2 mmol/LXCT790 prevented CYP2R1 repression by PGC-1a (Fig.2D) without any effect on PGC-1a or ERRa expression(Supplementary Fig. 2G and H). Analysis of public Encyclo-pedia of DNA Elements data indicates that PGC-1a andERRa bind to two common regions within the humanCYP2R1 gene in HepG2 cells (31) (Supplementary Fig.2M). We performed a bioinformatics promoter analysisof the mouse Cyp2r1 gene with MatInspector softwareand identified a potential ERRa binding site in the proximalpromoter.When21.2 kb Cyp2r1-59-luciferase-reporter con-struct was transfected into human hepatoma HepG2 cellsand the cells were infected with PGC-1a-Ad, luciferaseactivity was repressed; however, mutation of the ERRabinding site at position 21,117 to 21,122 bp (relative tothe TSS) abolished the PGC-1a response (Fig. 2E). Inter-estingly, ERRa-Ad did not have an effect on the luciferaseactivity, indicating a crucial need for PGC-1a (Fig. 2E).Altogether, these data indicate that ERRa plays a novel,indispensable role in PGC-1a–mediated downregulation ofCYP2R1 expression in mouse hepatocytes.

Next, we investigated the CYP2R1 repression mecha-nism in vivo by using PGC-1a KO mice fasted for 12 h.Interestingly, PGC-1amRNA was not significantly inducedby fasting in the livers of the Pgc-1a+/+ animals comparedwith the fed control animals (Supplementary Fig. 2I). On theother hand, PEPCK expression was significantly induced4.9-fold by fasting in the Pgc-1a+/+ mice and 3.4-fold in thePgc-1a2/2 mice, indicating a fasting response (Supplemen-tary Fig. 2J). CYP2R1 expression was repressed in the fastedPgc-1a+/+ mouse livers down to 25% compared with the fedcontrols (Fig. 2F). Interestingly, PGC-1a KO was not suf-ficient to abolish the CYP2R1 repression by fasting (Fig. 2F).


http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1

Figure 1—Fasting and diabetes repress CYP2R1 expression and function in liver.A andB: Fasting represses the CYP2R1mRNAbut inducesPEPCK mRNA in mouse liver (controls and 12-h fasted n = 9, 24-h fasted n = 8). C and D: Fasting decreases the CYP2R1 protein (n = 4) andthe vitamin D 25-hydroxylase activity in the liver microsomes (controls and 24-h fasted n = 6, 12-h fasted n = 5). E: Twenty-four-hour fastingreduces CYP2R1mRNA in rat liver (controls n = 3, fasted n = 4). F andG: HFD-induced obesity and type 2 diabetes in mouse downregulatesCYP2R1mRNA in the liver and reduces the plasma 25-OH-D (chow n = 10, HFD n = 9).H and I: The CYP2R1mRNA and protein were reducedin the liver of the type 1 diabetic mousemodel (STZ-treatedmice) (controls n = 8, STZ n = 4). The box and whisker plots indicate theminimum,25th percentile, median, 75th percentile, andmaximum. In addition, the mean is indicated with +. In the dot plots, the mean is indicated. Datawere analyzed in panels A, B, and D with one-way ANOVA (Tukey post hoc test, 95% CI) and in panels E–H with two-tailed t test. *P, 0.05,**P , 0.01, ***P , 0.001.


Curiously, the expression of PGC-1bwas induced morepotently in the livers of fasted PGC-1a KOmice (6.1-fold)compared with the WT mice (2.0-fold) (Fig. 2G). Inaddition, ERRa induction was significant only in the

PGC-1a KO mice (Supplementary Fig. 2K). We thereforehypothesized that PGC-1b, a close relative of PGC-1a aswell as a fasting-inducible factor (37), could compensatefor the chronic loss of PGC-1a in the KO mice or could

5' 3'

Figure 2—The PGC-1a-ERRa pathway represses CYP2R1. A: PGC-1a-Ad reduces the CYP2R1 mRNA in mouse primary hepatocytes (n =3). B: PGC-1a–mediated suppression of CYP2R1 requires interaction with ERRa (n = 6). C and D: ERRa knockdown by shERRa-Ad (n = 6) orERRa inhibition by XCT790 (XCT) (DMSO n = 4, 1 mmol/L XCT n = 6, 2 mmol/L XCT n = 5) abolishes the suppression of CYP2R1 by PGC-1a.E: An ERRa binding site in the Cyp2r1 promoter mediates PGC-1a–prompted reduction of the luciferase activity (n = 12 PGC-1a-Adexperiments [left], n = 5 ERRa-Ad experiments [right]). F: PGC-1a KO does not abolish the CYP2R1 repression by fasting in mouse liver(PGC-1a+/+ n = 7, PGC-1a2/2 n = 6).G: PGC-1a KO potentiates PGC-1b induction by fasting.H: Liver-specific PGC-1b KO (LS-PGC-1b2/2)does not abolish the CYP2R1 repression by fasting in mouse liver (n = 4). The box and whisker plots indicate the minimum, 25th percentile,median, 75th percentile, and maximum. In addition, the mean is indicated with +. Data were analyzed in panels A–D with one-way ANOVA(Tukey post hoc test, 95%CI) and in panels E–Hwith two-tailed t test. *P, 0.05, **P, 0.01, ***P, 0.001, ****P, 0.0001; #P, 0.05, ###P,0.001.


http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1

even play an independent role in the suppression ofCYP2R1.

We therefore investigated the fasting effect on CYP2R1also in the liver-specific PGC-1b KO mice. In 24-h fastedWT mice, CYP2R1 was again repressed by 35% (Fig. 2H).However, PGC-1b KO did not reverse the repression of

CYP2R1 (Fig. 2H). PGC-1b KO did not affect the PGC-1ainduction by fasting (Supplementary Fig. 2L).

Activation of GR Represses CYP2R1GR activated by cortisol is another key pathway controllingthe fasting response in the liver and is also activated in

Figure 3—Activation of GR represses CYP2R1. A and B: Treatment with dexamethasone (DEXA) reduces the CYP2R1 mRNA and protein inmouse liver (n = 7).C–F: TheGR antagonist mifepristone (MIF) attenuates the repression of CYP2R1 andNR1D1 and the induction of PGC-1aand TAT by DEXA in mouse liver (n = 7). G and H: The effect of GR antagonist MIF on fasting response of CYP2R1 and ANGPTL8 in mouseliver (vehicle n = 8, MIF [fed] n = 8, MIF [fast] n = 7). The box and whisker plots indicate the minimum, 25th percentile, median, 75th percentile,and maximum. In addition, the mean is indicated with +. Data were analyzed in panel Awith two-tailed t test and in panelsC–Hwith one-wayANOVA (Tukey post hoc test, 95% CI). **P , 0.01, ***P , 0.001, ****P , 0.0001; ##P , 0.01, ###P , 0.001, ####P , 0.0001.


http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1

diabetes (38). To investigate the role of GR in the regu-lation of CYP2R1, we treated mice with a GR agonist,dexamethasone, for 6 h. The treatment decreased liverCYP2R1 mRNA levels by 49% along with changes in theexpression of several known GR target genes, includingANGPTL8, NR1D1, and TAT (38,39), and increased PGC-1a levels 3.3-fold (Fig. 3A), suggesting involvement of GRin the regulation of CYP2R1 in vivo. GR is also a knowninteraction partner for PGC-1a (34). The CYP2R1 proteinwas decreased 26% by 6-h dexamethasone treatment (Fig.3B). Furthermore, analysis of the published microarraydata (accession number GSE24256) (40) supports theeffect of dexamethasone on CYP2R1 expression in mouseliver (Supplementary Fig. 3A).

To verify the role of GR, we cotreated mice withdexamethasone and the GR antagonist mifepristone. Mi-fepristone completely prevented induction of PGC-1a andTAT by dexamethasone and inhibited repression ofCYP2R1 and NR1D1 (Fig. 3C–F). Furthermore, analysisof published ChIP-seq data indicates that GR binds toCyp2r1 gene proximal promoter in mouse liver (41) (Sup-plementary Fig. 3B). To evaluate the role of GR in thefasting-mediated repression of CYP2R1, we investigatedthe effect of pharmacological inhibition of GR by mifepristone during fasting. Mifepristone abolished therepression of the control gene ANGPTL8 (Fig. 3H).Furthermore, mifepristone partially, but significantly,prevented the effect of fasting on the CYP2R1 in theliver (repression decreased from 56 to 18%); however,the fasting effect still remained statistically significant(Fig. 3G).

Fasting Induces CYP24A1 in the Kidney Through thePGC-1a-ERRa Pathway

DNA microarray analysis of PGC-1a–responsive genes inmouse hepatocytes indicated that CYP24A1, the vitamin D24-hydroxylase and the main inactivator of active vitaminD (42,43), was among the most upregulated genes (Table2). This was further confirmed by PGC-1a-Ad dose-response experiments using qPCR (Fig. 4A).

CYP24A1 is primarily regulated by VDR (44), and VDRis an interaction partner for PGC-1a (45). We thereforeexplored the role of VDR in the PGC-1a–mediated in-duction of CYP24A1. After treating mouse primary hep-atocytes with the VDR ligand calcitriol, CYP24A1 wasinduced 120-fold (Fig. 4B–D). PGC-1a-Ad induced CYP24A1much more potently, up to 20,000-fold. However, com-bined treatment with calcitriol did not potentiate CYP24A1induction by PGC-1a (Fig. 4B). Furthermore, CYP24A1protein could be detected after PGC-1a-Ad transductionbut not after calcitriol treatment (Fig. 4E). PGC-1a-Adtreatment had no effect on VDR expression, but ERRa wasinduced 11-fold (Fig. 4F and G). Altogether, these resultssuggest that VDR does not play a major role in the CYP24A1induction by PGC-1a.

We next investigated the putative role of ERRa in theCYP24A1 induction by PGC-1a. PGC-1a-2x9 mutant in-duced CYP24A1 similarly to the WT PGC-1a, whereasPGC-1a-L2L3M mutant had no effect, suggesting involve-ment of ERRa (Fig. 4H). Indeed, induction of CYP24A1was abolished by ERRa knockdown, indicating a novelPGC-1a-ERRa pathway–mediated regulatory mechanismfor CYP24A1 (Fig. 4I).

Table 2—The top 10 up- and downregulated genes in mouse primary hepatocytes transduced with PGC-1a-Ad compared withGFP-Ad–treated control cells

Gene symbol Gene name Fold change

UpregulatedG6pc Glucose-6-phosphatase catalytic subunit 388.9Nr0b2 Nuclear receptor subfamily 0 group B member 2 290.7Cyp24a1 Cytochrome P450 family 24 subfamily A member 1 179.0Slc16a5 Solute carrier family 16 member 5 133.2Ldhb Lactate dehydrogenase B 135.9Cox7a1 Cytochrome C oxidase subunit 7A1 79.3Cidec Cell death–inducing DFFA-like effector C 108.9Upp1 Uridine phosphorylase 1 68.1Usp2 Ubiquitin-specific peptidase 2 46.0Hpdl 4-Hydroxyphenylpyruvate dioxygenase like 46.8

DownregulatedFbxo5 F-box protein 5 214.3E2f7 E2F transcription factor 7 212.6Fignl1 Fidgetin like 1 211.5Uhrf1 Ubiquitin like with PHD and ring finger domains 1 210.8Ccne2 Cyclin E2 210.14632417K18Rik RIKEN cDNA 4632417K18 gene 28.9Cdc2a Cyclin D1 27.3Ccnf Cyclin F 27.3Shcbp1 SHC binding and spindle associated 1 27.2Aurka Aurora kinase A 26.9

Full data available at GEO with the accession number GSE114485.


http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-1050/-/DC1

Figure 4—The PGC1a-ERRa pathway mediates the CYP24A1 induction in hepatocytes. A: PGC-1a induces the CYP24A1 expression inmouse primary hepatocytes (n = 3). B: PGC-1a-Ad induces the CYP24A1 expression much more potently than calcitriol treatment in mouseprimary hepatocytes (PGC-1a-Ad and LacZ-Ad MOI 1, calcitriol 200 nmol/L, n = 3). C: PGC-1a was upregulated as expected by the PGC-1a-Ad (PGC-1a-Ad and LacZ-AdMOI 1, calcitriol 200 nmol/L, n = 3).D: PEPCKwas induced as expected by PGC-1a induction (PGC-1a-Adand LacZ-Ad MOI 1, calcitriol 200 nmol/L, n = 3). E: CYP24A1 protein was detected by immunoblotting after the PGC-1a induction but notafter calcitriol treatment (PGC-1a-Ad and LacZ-Ad MOI 1, calcitriol 200 nmol/L, n = 3). F and G: PGC-1a did not affect the VDR expression,whereas ERRa was induced in mouse primary hepatocytes (PGC-1a-Ad and LacZ-Ad MOI 1, calcitriol 200 nmol/L, n = 3). H: CYP24A1induction by PGC-1a requires interaction with ERRa in the mouse primary hepatocytes (n = 6). I: The ERRa knockdown abolishes theinduction of CYP24A1by PGC-1a inmouse primary hepatocytes (n = 6). The bars indicatemean6SD. The box andwhisker plots indicate theminimum, 25th percentile, median, 75th percentile, andmaximum. In addition, themean is indicatedwith +. The data were analyzedwith one-way ANOVA (Tukey post hoc test, 95% CI). In panels A, H, and I, some control values without PGC-1a induction were below the detectionlevel and are not shown. *P , 0.05, **P , 0.01, ***P , 0.001, ****P , 0.0001; #P , 0.05, ###P , 0.001, ####P , 0.0001.


Although we detected CYP24A1 in hepatocytes afterPGC-1a overexpression, we could not detect CYP24A1 inthe mouse liver in vivo. Indeed, the main site for CYP24A1expression and function is the kidney (42). CYP24A1 wasinduced up to 2.7-fold in the kidney by fasting (Fig. 5A) inagreement with a previous study (46). However, the fast-ing effect was abolished in the PGC-1a KOmice (Fig. 5B–D),indicating that PGC-1a plays an indispensable role inthe CYP24A1 induction by fasting in the kidney. Further-more, ERRa antagonist XCT790 treatment diminished thefasting-mediated induction of CYP24A1, suggesting thatfasting induces CYP24A1 in the kidney through the PGC-1a-ERRa pathway (Fig. 5E–H).

DISCUSSION

Plasma 25-OH-D level is regularly used as the measure ofvitamin D status. The reason behind this is the rather long

half-life of 25-OH-D of ;2 weeks (47). Furthermore,25-hydroxylation has been considered not to be underefficient metabolic control, and 25-OH-D should thusreflect the intake of vitamin D (48). Several CYP enzymes,including CYP2R1, CYP27A1, CYP2D25, CYP2C11, andCYP3A4, have been reported to be capable of vitamin D25-hydroxylation in vitro (48). However, genetic studies,expression data, and catalytic properties all suggest thatCYP2R1 is the main vitamin D 25-hydroxylase enzyme inliver. Indeed, both the KO studies in mouse and the geneticevidence in humans indicate that defect in the CYP2R1 generesults in vitamin D deficiency (i.e., low plasma 25-OH-Dlevel) (12,13,49). In humans, this has been shown to resultin symptomatic rickets, vitamin D–dependent rickets type1B (15). These patients respond poorly to ordinary vita-min D supplementation but may benefit from 25-OH-Dtreatment (15).

Figure 5—The PGC1a-ERRa pathway mediates the CYP24A1 induction in the kidney by fasting. A: Twelve-hour fasting induces theCYP24A1 in themouse kidney (n = 10).B: PGC-1aKO abolishes the induction of the CYP24A1 by fasting in themouse kidney (PGC-1a+/+ n =7, PGC-1a2/2 n = 6).C: PGC-1awas not detected in the kidneys of either fed or fasted PGC-1a2/2mice.D: ERRamRNA in the kidneys ofWTand PGC-1a2/2 mice. E: Inhibition of ERRa by XCT790 (XCT) attenuates the CYP24A1 induction by fasting in the mouse kidney (vehicle n =10, XCT n = 8). F–H: PEPCK, PGC-1a, and ERRa were measured as control genes in the XCT790 experiment. The box and whisker plotsindicate theminimum, 25th percentile, median, 75th percentile, andmaximum. In addition, themean is indicatedwith +. Datawere analyzed inpanels A–D with two-tailed t test and in panels E–H with one-way ANOVA (Tukey post hoc test, 95% CI). *P, 0.5, **P, 0.01, ***P, 0.001,****P , 0.0001.


We now show that the CYP2R1 enzyme may be re-pressed also functionally at the level of gene regulation.Twelve-hour fasting suppressed liver microsomal vitaminD 25-hydroxylation;50%, and after 24-h fasting, we wereunable to detect any 25-OH-D formation. Thus, the firstvitamin D bioactivation step is under the strict control ofnutritional state. Although the acute food deprivationresulted in a strong effect on vitamin D 25-hydroxylaseactivity, this was not reflected in the plasma 25-OH-Dconcentration, presumably because of the long half-life of25-OH-D (47). Therefore, it seems unlikely that short-term fasting would have a significant effect on vitamin Dfunctions at the systemic level. This raises the question ofthe physiological purpose of the CYP2R1 repression duringfasting. A likely explanation is that fasting launches phys-iological adjustment as precaution for possible longer-termfood shortage. This may potentially be related to the role ofvitamin D in energy homeostasis (1) and could have beenevolutionarily beneficial during periods of starvation. Al-ternatively, 25-OH-D could have some unknown localfunction in liver. Furthermore, we observed induction ofCYP24A1 in the kidney during fasting. This is a mechanismthat limits the level of 1a,25-(OH)2-D and consequentlyactivation of VDR (10). The CYP24A1 induction and theCYP2R1 repression are expected to suppress vitamin Dsignaling in a synergistic manner.

The suppression of vitamin D bioactivation by fasting-activated mechanisms has important implications in thecontext of human metabolic diseases. Hepatic signaling

pathways triggered physiologically during fasting displaytypically prolonged, constant activation in diabetes. Theclassical consequence is increased activation of gluconeogen-esis, resulting in fasting hyperglycemia (50). Since in diabetes,unlike in short-term fasting, the activation of thesemolecularmechanisms is long-standing, the suppression of vitamin D25-hydroxylation will eventually lead to a lower plasma25-OH-D level. In agreement with this theory, we observedreduced plasma 25-OH-D concentrations in the HFD-treatedmice. Thus, we propose that repression of vitamin D bio-activation represents a novel mechanism that plays a role invitamin D deficiency in diabetes.

PGC-1a is one of the major molecular factors regulatinggluconeogenesis and other metabolic pathways activatedin the diabetic liver (34,50). We now show that PGC-1a,ERRa dependently, also represses vitamin D bioactivationand, thus, establishes regulation of vitamin D metabolismas a novel metabolic function under the control of PGC-1a.Furthermore, the CYP24A1 induction by fasting in thekidney was demonstrated to be under the control of PGC-1a-ERRa. Thus, the PGC-1a-ERRa pathway appears toplay a major role in the crosstalk between energy homeo-stasis and vitamin D metabolism. Interestingly, a recentstudy showed that Cyp2r1-deficiency in zebrafish affectedlipid metabolism through vitamin D–regulated function ofPGC-1a (51). PGC-1a and vitamin D metabolism couldthus form a regulatory loop.

Although, the PGC-1a-ERRa pathway was found to bean effective regulator of CYP2R1, the PGC-1a KO did not

Figure 6—A proposed model of how nutrition-sensing factors regulate vitamin D metabolism in response to fasting and diabetes. Weproposed that at least twomolecular pathways are involved in the suppression of the CYP2R1 by fasting in the liver. These samemetabolism-regulating pathways are activated in diabetes. The first one is mediated by the fasting-inducible cofactor PGC-1a through nuclear recep-tor ERRa. The second one is mediated through the cortisol/GR pathway. CYP2R1 repression results in suppression of vitamin D25-hydroxylation, the first bioactivation step in the liver. On the other hand, fasting induces CYP24A1 in the kidney through a mecha-nism involving the PGC-1a-ERRa pathway. Induction of CYP24A1 is expected to induce catabolism of vitamin D. Suppression of the25-hydroxylation in the liver and induction of the deactivation step in the kidney may lead to a lower plasma level of 25-OH-D and, inturn, vitamin D deficiency.


prevent CYP2R1 suppression during fasting. This indicatesthat additional molecular mechanisms play a role in theregulation of CYP2R1 during fasting. Activation of GR wasfound to be a second mechanism capable for CYP2R1suppression. Indeed, cortisol levels are increased duringfasting as well as in diabetes (38). By using pharmacolog-ical inhibition of GR, we could partially prevent fasting-mediated repression of CYP2R1, suggesting that GR isinvolved in CYP2R1 repression by fasting. However, wecannot exclude the possibility that additional regulatorymechanisms mediate CYP2R1 repression by fasting. Fromthe point of view of drug therapy, the observed repressionof CYP2R1 by pharmacological glucocorticoid treatmentmay explain the observed association between glucocorti-coid use and vitamin D deficiency (52).

In summary, our results reveal a novel crosstalk be-tween energy homeostasis and the vitamin D pathway,suggesting a physiological need for suppression of vitaminD signaling during nutrient deprivation (Fig. 6). This maybe related to the role of vitamin D in energy metabolism(1,3). Altogether, our study provides a mechanism thatmay explain the lower vitamin D levels in patients withdiabetes and suggests that vitamin D deficiency is a con-sequence, not the cause, of diabetes.

Acknowledgments. The authors thank Ritva Tauriainen (University ofOulu) for technical assistance. The help of Dr. Anastasia Georgiadi and SanderKersten (University of Wageningen, Wageningen, the Netherlands) with themicroarray study is acknowledged.Funding. The study was financially supported by the Scholarship Fund of theUniversity of Oulu (Tyyni Tani Fund) to M.-S.E., the Academy of Finland (grants267637, 292540) and the Sigrid Juselius Foundation to P.T., the CologneExcellence Cluster on Cellular Stress Responses in Aging-associated Diseasesand Center of Molecular Medicine Cologne of the Medical Faculty to R.J.W.,National Institutes of Health grant R01-DK-104735 to B.N.F., and the Academy ofFinland (grant 286743) and the Diabetes Research Foundation to J.H.Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. S.-M.A. and M.-S.E. performed a majority of theexperiments and measurements. S.-M.A., M.-S.E., and J.H. designed the study,analyzed the data, and wrote the manuscript. S.-M.A. and M.B. performed thePGC-1amicroarray experiment. O.K. and M.K. performed the HFD experiment andhelped with the other animal experiments. P.V. performed the 12- and 24-h fastingexperiments in mice and helped with the hepatocyte primary cultures. V.R.performedmeasurements of the 25-OH-D by LC-MS/MS. M.M. and P.T. performedthe PGC-1a KO mouse experiments. A.F. and R.J.W. performed the STZ mousestudy. K.T.C. and B.N.F. performed the liver-specific PGC-1b KOmouse study. J.H.supervised the overall conduct of the study. All authors read and approved the finalmanuscript. J.H. is the guarantor of this work and, as such, had full access to allthe data in the study and takes responsibility for the integrity of the data and theaccuracy of the data analysis.Data Availability. Data sharing: The data sets generated and/or analyzedduring the current study are available in the GEO repository (our microarray forPGC-1a overexpression in mouse primary hepatocytes, accession numberGSE114485; published microarray for dexamethasone treatment in mouse livers,GSE24256; published PPARGC1A and ESRRA ChIP-seq in HepG2, GSE31477; andpublished NR3C1 [GR] ChIP-seq in mouse liver, GSE72084). Resource sharing: Theresources generated and/or analyzed during the current study are available fromthe corresponding author on reasonable request.

Prior Presentation. Parts of this study were presented in abstract format the 10th International Meeting of the International Society for the Study ofXenobiotics, Toronto, Ontario, Canada, 29 September–3 October, 2013; the6th Sino-Finn Life Science Forum: From Systems Biology to TranslationalMedicine, Helsinki, Finland, 17–18 August 2015; Nuclear Receptors: FromMolecules to Humans, Ajaccio, France, 24–28 September 2015; the 4thHelmholtz-Nature Medicine Diabetes Conference, Munich, Germany, 18–20September 2016; and European Molecular Biology Organization/EuropeanMolecular Biology Laboratory Symposium: Metabolism in Time and Space:Emerging Links to Cellular and Developmental Programs, Heidelberg,Germany, 11–13 May 2017.

References1. Bouillon R, Carmeliet G, Lieben L, et al. Vitamin D and energy homeostasis: ofmice and men. Nat Rev Endocrinol 2014;10:79–872. Rosen CJ, Adams JS, Bikle DD, et al. The nonskeletal effects of vitamin D: anEndocrine Society scientific statement. Endocr Rev 2012;33:456–4923. Abbas MA. Physiological functions of vitamin D in adipose tissue. J SteroidBiochem Mol Biol 2017;165:369–3814. Narvaez CJ, Matthews D, Broun E, Chan M, Welsh J. Lean phenotype andresistance to diet-induced obesity in vitamin D receptor knockout mice correlateswith induction of uncoupling protein-1 in white adipose tissue. Endocrinology2009;150:651–6615. Holick MF. The vitamin D deficiency pandemic: approaches for diagnosis,treatment and prevention. Rev Endocr Metab Disord 2017;18:153–1656. Van Belle TL, Gysemans C, Mathieu C. Vitamin D and diabetes: the oddcouple. Trends Endocrinol Metab 2013;24:561–5687. Kayaniyil S, Harris SB, Retnakaran R, et al. Prospective association of 25(OH)D with metabolic syndrome. Clin Endocrinol (Oxf) 2014;80:502–5078. Harinarayan CV. Vitamin D and diabetes mellitus. Hormones (Athens) 2014;13:163–1819. Norris JM, Lee HS, Frederiksen B, et al.; TEDDY Study Group. Plasma25-hydroxyvitamin D concentration and risk of islet autoimmunity. Diabetes 2018;67:146–15410. Christakos S, Dhawan P, Verstuyf A, Verlinden L, Carmeliet G. Vitamin D:metabolism, molecular mechanism of action, and pleiotropic effects. Physiol Rev2016;96:365–40811. Zhu JG, Ochalek JT, Kaufmann M, Jones G, Deluca HF. CYP2R1 is a major,but not exclusive, contributor to 25-hydroxyvitamin D production in vivo. Proc NatlAcad Sci U S A 2013;110:15650–1565512. Cheng JB, Levine MA, Bell NH, Mangelsdorf DJ, Russell DW. Genetic evi-dence that the human CYP2R1 enzyme is a key vitamin D 25-hydroxylase. ProcNatl Acad Sci U S A 2004;101:7711–771513. Al Mutair AN, Nasrat GH, Russell DW. Mutation of the CYP2R1 vitamin D25-hydroxylase in a Saudi Arabian family with severe vitamin D deficiency. J ClinEndocrinol Metab 2012;97:E2022–E202514. Thacher TD, Fischer PR, Singh RJ, Roizen J, Levine MA. CYP2R1 mutationsimpair generation of 25-hydroxyvitamin D and cause an atypical form of vitamin Ddeficiency. J Clin Endocrinol Metab 2015;100:E1005–E101315. Molin A, Wiedemann A, Demers N, et al. Vitamin D-dependent rickets type 1B(25-hydroxylase deficiency): a rare condition or a misdiagnosed condition? J BoneMiner Res 2017;32:1893–189916. Wang TJ, Zhang F, Richards JB, et al. Common genetic determinants ofvitamin D insufficiency: a genome-wide association study. Lancet 2010;376:180–18817. Manousaki D, Dudding T, Haworth S, et al. Low-frequency synonymouscoding variation in CYP2R1 has large effects on vitamin D levels and risk ofmultiple sclerosis [published correction appears in Am J Hum Genet 2018;103:1053]. Am J Hum Genet 2017;101:227–23818. Jiang X, O’Reilly PF, Aschard H, et al. Genome-wide association study in79,366 European-ancestry individuals informs the genetic architecture of25-hydroxyvitamin D levels. Nat Commun 2018;9:260


19. Ramos-Lopez E, Brück P, Jansen T, Herwig J, Badenhoop K. CYP2R1(vitamin D 25-hydroxylase) gene is associated with susceptibility to type1 diabetes and vitamin D levels in Germans. Diabetes Metab Res Rev 2007;23:631–63620. Cooper JD, Smyth DJ, Walker NM, et al. Inherited variation in vitamin D genesis associated with predisposition to autoimmune disease type 1 diabetes. Diabetes2011;60:1624–163121. Lin J, Wu PH, Tarr PT, et al. Defects in adaptive energy metabolism with CNS-linked hyperactivity in PGC-1alpha null mice. Cell 2004;119:121–13522. Chambers KT, Chen Z, Crawford PA, et al. Liver-specific PGC-1beta de-ficiency leads to impaired mitochondrial function and lipogenic response tofasting-refeeding. PLoS One 2012;7:e5264523. Buler M, Aatsinki SM, Skoumal R, et al. Energy-sensing factors coactivatorperoxisome proliferator-activated receptor g coactivator 1-a (PGC-1a) andAMP-activated protein kinase control expression of inflammatory mediators inliver: induction of interleukin 1 receptor antagonist. J Biol Chem 2012;287:1847–186024. Franko A, von Kleist-Retzow JC, Neschen S, et al. Liver adapts mito-chondrial function to insulin resistant and diabetic states in mice. J Hepatol2014;60:816–82325. Franko A, Huypens P, Neschen S, et al. Bezafibrate improves insulin sen-sitivity and metabolic flexibility in STZ-induced diabetic mice. Diabetes 2016;65:2540–255226. Arpiainen S, Järvenpää SM, Manninen A, et al. Coactivator PGC-1alpharegulates the fasting inducible xenobiotic-metabolizing enzyme CYP2A5 in mouseprimary hepatocytes. Toxicol Appl Pharmacol 2008;232:135–14127. Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ. Isolation of biologicallyactive ribonucleic acid from sources enriched in ribonuclease. Biochemistry 1979;18:5294–529928. Raunio H, Valtonen J, Honkakoski P, et al. Immunochemical detection ofhuman liver cytochrome P450 forms related to phenobarbital-inducible forms inthe mouse. Biochem Pharmacol 1990;40:2503–250929. Lämsä V, Levonen AL, Leinonen H, Ylä-Herttuala S, Yamamoto M, Hakkola J.Cytochrome P450 2A5 constitutive expression and induction by heavy metals isdependent on redox-sensitive transcription factor Nrf2 in liver. Chem Res Toxicol2010;23:977–98530. Cartharius K, Frech K, Grote K, et al. MatInspector and beyond: promoteranalysis based on transcription factor binding sites. Bioinformatics 2005;21:2933–294231. Charos AE, Reed BD, Raha D, Szekely AM, Weissman SM, Snyder M. A highlyintegrated and complex PPARGC1A transcription factor binding network in HepG2cells. Genome Res 2012;22:1668–167932. Goldstein I, Baek S, Presman DM, Paakinaho V, Swinstead EE, Hager GL.Transcription factor assisted loading and enhancer dynamics dictate the hepaticfasting response. Genome Res 2017;27:427–43933. Mei S, Qin Q, Wu Q, et al. Cistrome Data Browser: a data portal for ChIP-Seqand chromatin accessibility data in human and mouse. Nucleic Acids Res 2017;45:D658–D66234. Yoon JC, Puigserver P, Chen G, et al. Control of hepatic gluconeo-genesis through the transcriptional coactivator PGC-1. Nature 2001;413:131–138

35. Gaillard S, Grasfeder LL, Haeffele CL, et al. Receptor-selective coactivators astools to define the biology of specific receptor-coactivator pairs. Mol Cell 2006;24:797–80336. Schreiber SN, Knutti D, Brogli K, Uhlmann T, Kralli A. The transcriptionalcoactivator PGC-1 regulates the expression and activity of the orphan nuclearreceptor estrogen-related receptor a (ERRa). J Biol Chem 2003;278:9013–901837. Lai L, Leone TC, Zechner C, et al. Transcriptional coactivators PGC-1alphaand PGC-lbeta control overlapping programs required for perinatal maturation ofthe heart. Genes Dev 2008;22:1948–196138. Opherk C, Tronche F, Kellendonk C, et al. Inactivation of the glucocorticoidreceptor in hepatocytes leads to fasting hypoglycemia and ameliorates hyper-glycemia in streptozotocin-induced diabetes mellitus. Mol Endocrinol 2004;18:1346–135339. Dang F, Wu R, Wang P, et al. Fasting and feeding signals control the os-cillatory expression of Angptl8 to modulate lipid metabolism. Sci Rep 2016;6:3692640. Duma D, Collins JB, Chou JW, Cidlowski JA. Sexually dimorphic actions ofglucocorticoids provide a link to inflammatory diseases with gender differences inprevalence. Sci Signal 2010;3:ra7441. Lim HW, Uhlenhaut NH, Rauch A, et al. Genomic redistribution of GRmonomers and dimers mediates transcriptional response to exogenous gluco-corticoid in vivo. Genome Res 2015;25:836–84442. Prosser DE, Jones G. Enzymes involved in the activation and inactivation ofvitamin D. Trends Biochem Sci 2004;29:664–67343. Jones G, Prosser DE, Kaufmann M. 25-Hydroxyvitamin D-24-hydroxylase(CYP24A1): its important role in the degradation of vitamin D. Arch BiochemBiophys 2012;523:9–1844. Schuster I. Cytochromes P450 are essential players in the vitamin D signalingsystem. Biochim Biophys Acta 2011;1814:186–19945. Savkur RS, Bramlett KS, Stayrook KR, Nagpal S, Burris TP. Coactivation of thehuman vitamin D receptor by the peroxisome proliferator-activated receptorgamma coactivator-1 alpha. Mol Pharmacol 2005;68:511–51746. Hagenfeldt-Pernow Y, Ohyama Y, Sudjana-Sugiaman E, Okuda K, Björkhem I.Short-term starvation increases calcidiol-24-hydroxylase activity and mRNA levelin rat kidney. Eur J Endocrinol 1994;130:608–61147. Jones KS, Assar S, Harnpanich D, et al. 25(OH)D2 half-life is shorter than25(OH)D3 half-life and is influenced by DBP concentration and genotype. J ClinEndocrinol Metab 2014;99:3373–338148. Zhu J, DeLuca HF. Vitamin D 25-hydroxylase - four decades of searching, arewe there yet? Arch Biochem Biophys 2012;523:30–3649. Thacher TD, Levine MA. CYP2R1 mutations causing vitamin D-deficiencyrickets. J Steroid Biochem Mol Biol 2017;173:333–33650. Wu H, Deng X, Shi Y, Su Y, Wei J, Duan H. PGC-1a, glucose metabolism andtype 2 diabetes mellitus. J Endocrinol 2016;229:R99–R11551. Peng X, Shang G, Wang W, et al. Fatty acid oxidation in zebrafish adiposetissue is promoted by 1a,25(OH)2D3. Cell Rep 2017;19:1444–145552. Skversky AL, Kumar J, Abramowitz MK, Kaskel FJ, Melamed ML. Associationof glucocorticoid use and low 25-hydroxyvitamin D levels: results from the NationalHealth and Nutrition Examination Survey (NHANES): 2001-2006. J Clin EndocrinolMetab 2011;96:3838–3845


Documents

Fasting-Induced Transcription Factors Repress Vitamin D ...Fasting-Induced Transcription Factors Repress Vitamin D Bioactivation, a Mechanism for Vitamin D Deﬁciency in Diabetes