Altered lipoprotein metabolism in P2Y13 knockout mice

Altered lipoprotein metabolism in P2Y13 knockout mice

Biochimica et Biophysica Acta 1801 (2010) 1349–1360 Contents lists available at ScienceDirect Biochimica et Biophysica Acta j o u r n a l h o m e p ...

1MB Sizes 5 Downloads 79 Views

Biochimica et Biophysica Acta 1801 (2010) 1349–1360

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b b a l i p

Altered lipoprotein metabolism in P2Y13 knockout mice Daniël Blom a,⁎, Ting-ting Yamin a, Marie-France Champy d, Mohammed Selloum d, Elodie Bedu d, Ester Carballo-Jane a, Lynn Gerckens a, Silvi Luell a, Roger Meurer a, Jayne Chin a, John Mudgett b, Oscar Puig c a

Department of Cardiovascular Diseases, Merck & Co, Inc., Rahway, NJ 07065, USA Department of Genetically Engineered Models Initiative, Merck & Co, Inc., Rahway, NJ 07065, USA Department of Molecular Profiling Research Informatics, Merck & Co, Inc., Rahway, NJ 07065, USA d Institut Clinique de la Souris/Mouse Clinical Institute BP10142, 67404 Illkirch Cedex, France b c

a r t i c l e

i n f o

Article history: Received 22 April 2010 Received in revised form 10 August 2010 Accepted 25 August 2010 Available online 15 September 2010 Keywords: GPR86 P2ry13 HDL Gene expression profiling Reverse Cholesterol Transport SREBP

a b s t r a c t The purinergic receptor P2Y13 has been shown to play a role in the uptake of holo-HDL particles in in vitro hepatocyte experiments. In order to determine the role of P2Y13 in lipoprotein metabolism in vivo, we ablated the expression of this gene in mice. Here we show that P2Y13 knockout mice have lower fecal concentrations of neutral sterols (− 27% ± 2.1% in males) as well as small decreases in plasma HDL (− 13.1% ± 3.2% in males; − 17.5% ± 4.0% in females) levels. In addition, significant decreases were detected in serum levels of fatty acids and glycerol in female P2Y13 knockout mice. Hepatic mRNA profiling analyses showed increased expression of SREBP-regulated cholesterol and fatty acid biosynthesis genes, while fatty acid β-oxidation genes were significantly decreased. Liver gene signatures also identified changes in PPARα-regulated transcript levels. With the exception of a small increase in bone area, P2Y13 knockout mice do not show any additional major abnormalities, and display normal body weight, fat mass and lean body mass. No changes in insulin sensitivity and oral glucose tolerance could be detected. Taken together, our experiments assess a role for the purinergic receptor P2Y13 in the regulation of lipoprotein metabolism and demonstrate that modulating its activity could be of benefit to the treatment of dyslipidemia in people. © 2010 Elsevier B.V. All rights reserved.

1. Introduction High density lipoprotein cholesterol (HDL-c) has been shown to play an important role in the transport of cholesterol from peripheral cells to the liver in a process called reverse cholesterol transport. Increasing the reverse transport of cholesterol via HDL, either by increasing the HDL particle pool size or by enhancing the flux of cholesterol through this pathway, is thought to have a cardioprotective effect [1–4]. The liver is capable of binding HDL-c in a scavenger receptor BI (SR-BI) dependent and independent way. SR-BI dependent endocytosis of HDL-c by the liver is thought to lead to selective uptake of cholesterol, followed by recycling of the protein moiety back into the circulation [5,6]. The SR-BI independent binding of HDL-c involves uptake by the liver, probably followed by degradation of the entire HDL particle [7–9]. Recently, a high affinity, non-SR-BI, HDL binding site on the liver was shown to be the β-chain of the ATP synthase complex [10–12]. The ATPase synthase complex is a major mitochondrial inner membrane protein involved in the synthesis of ATP, and the identification of the β-subunit of

⁎ Corresponding author. Merck & Co, Inc., 126 E Lincoln Avenue, RY80T-A100, Rahway, NJ 07065, USA. Tel.: +1 732 594 5002; fax: +1 732 594 1169. E-mail address: [email protected] (D. Blom). 1388-1981/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2010.08.013

this complex on the plasma membrane was therefore a surprising one. Martinez and colleagues [10] showed that binding of lipid-free ApoAI to the ATPase β chain induced an increase in ATP hydrolysis, indicating a reversal of the catalytic activity of this protein at the plasma membrane. This hydrolysis of ATP was accompanied by a concomitant increase in HDL endocytosis by ATPase β-chain-expressing liver cells. The hydrolysis of ATP, and thus the generation of ADP, upon binding of lipid-free ApoAI to β-chain ATPase suggested a role for a purinergic receptor in the uptake of holo-HDL particles by liver cells. Indeed, the endocytosis of triglyceride-rich HDL2 (TG-HDL2) by HepG2 liver cells could be mimicked by the administration of ADP, but not other tri- or dinucleophosphates, allowing for uncoupling of ApoAI binding to the β-chain and the holo-HDL particle endocytosis step [13]. By performing RNA knockdown experiments, Jacquet and colleagues showed the nature of this purinergic receptor to be the Gαi protein coupled receptor P2Y13 [13]. P2Y13 was simultaneously identified by three different research groups and has also been called GPR86 or GPR94 [14–16]. P2Y13 is widely expressed in human tissues (as assessed by Northern blot and RT-PCR), with highest expression in the spleen, followed by liver, placenta, leukocytes and brain [14,15]. The antithrombotic agent AR-C69931MX was developed as a P2Y12 antagonist and showed partial agonism on P2Y13, as measured by its potential to lower forskolin-induced cAMP levels in CHO cells [13]. When administered to human HepG2 cells, both ADP and AR-C69931MX were

1350

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

shown to stimulate the uptake of both TG-HDL2 as well as HDL3, a process that was dependent on the presence of functional P2Y13 since this effect was not observed in P2Y13 knockdown HepG2 cells and because HepG2 cells do not express P2Y12 [13]. Taken together, these results suggest that P2Y13 plays a role in the metabolism of HDL and the clearance of these lipoprotein particles by the liver. A better understanding of the role of P2Y13 in this process could contribute to the development of small molecules that could alter the levels of plasma HDL, thereby enabling therapeutic intervention with cardiovascular disease. In this study, we describe the phenotypic analysis of P2Y13 knockout mice and assess the role of this gene in the metabolism of HDL cholesterol. 2. Materials and methods 2.1. Animals and phenotypic analyses of P2Y13 knockout mice The P2Y13 knockout mice (Line 1952; Genbank accession AK008013; gi: 12841935) were generated at Deltagen (San Mateo, CA, USA). All animal procedures described were approved by the Merck Research Laboratories (Rahway, NJ) Institutional Animal Care and Use Committee, and were in conformity with Public Health Service policy on humane care and use of laboratory animals. Body weights were monitored once a week during 10 weeks, between 6 and 16 weeks of age. At 8 weeks of age, animals were switched from an autoclaved standard rodent chow diet (D04 from SAFE, France) to high fat high carbohydrate (HFHC) diet (D12492 from Research Diets, USA). Body composition (lean and fat mass content) was evaluated by dual X-Ray analysis on a densitometer PIXImus (GE Medical systems) at the age of 7 weeks (mice fed chow diet) or by quantitative magnetic resonance (qNMR, Minispec-Bruker) at the age of 23 weeks (mice fed HFHC diet for 15 weeks). For the qNMR analyses, mice were placed individually into the insert tube of the machine, followed by careful closure of the insert with a plunger in order to gently restrain the mouse and placement into the machine for the measurement. The total measurement took approximately 2 min per mouse and assessed the percentage of lean, fat and free body fluid content. X-ray analysis was performed on young, anaesthetized mice (intraperitoneal injection of Ketamine/Xylazine) in order to also evaluate the bone mineral density. The qNMR measurement did not give bone mineral density and was performed on conscious mice. Oral glucose tolerance test (OGTT) was performed at the age of 16 weeks, after 8 weeks of HFHC diet feeding on 16 h-fasted mice. After measurement of the fasted glucose level on a drop of blood collected from the tail (time 0), mice were dosed via oral gavage with a solution of 20% glucose in sterile water at a dose of 2 g glucose/kg body weight. Blood was collected from the tail for glycaemia measurements at 15, 30, 45, 60, 90, 120, 150 and 180 min after injection of the glucose solution. Glucose levels were measured on total blood using Accu-Chek active glucose test strips (Roche Diagnostics). Insulin sensitivity test was performed at the age of 18 weeks, after 10 weeks of HFHC diet feeding on mice on 16-h fasted mice. After measurement of the basal glucose level on a drop of blood collected from the tail (time 0), mice were injected intraperitoneally with a solution of insulin (Umuline® Rapide - Ely Lilly) in a sterile saline solution (0.9% NaCl) at a dose of 0.5 UI/kg. Blood was collected for glucose determination from the tail at 15, 30, 45, 60 and 90 min after injection of the insulin solution. Blood parameters were determined at the age of 8 weeks (upon standard chow diet) and 24 weeks (after 16 weeks of HFHC diet feeding) using an Olympus AU400 analyzer (Olympus SA, Rungis, France) with kits and controls supplied by Olympus for total cholesterol, HDL-cholesterol, and triglycerides measurement. Free fatty acid levels were measured on the Olympus AU400 using a kit from Wako (Wako Chemical Inc, Richmond, USA) and glycerol, using a

kit from Randox Laboratories (United Kingdom). Internal quality control materials (Olympus) were analyzed on a daily basis to monitor our precision throughout the experiment. Lipoprotein profiles from P2Y13 KO mice were generated by fractionation of plasma or serum using Superose-6 size exclusion chromatography (GE LifeSciences, Inc.). Total cholesterol levels in the column effluent were continuously measured via in-line mixture with a commercially available enzymatic colorimetric cholesterol detection reagent (Total Cholesterol E, Wako USA) followed by downstream spectrophotometric detection of the reaction products at 600 nm absorbance. The first peak of cholesterol eluted from the column was attributed to VLDL, the second peak to LDL and the third to HDL; the area under each peak was calculated using software provided with the FPLC. To calculate the cholesterol concentration for each lipoprotein fraction, the ratio of the corresponding peak area to total peak area was multiplied by the total cholesterol concentration measured in the sample. All data collected in the phenotypic analyses described above were subjected to statistical analysis using the Student's t-test (p b 0.05). For all analyses, the Graphpad Prism 5 program was used. Statistically significant changes or differences are indicated in each respective figure. 2.2. In vivo RCT assay AcLDL (Intracel, cat #RP-045, 2.5 mg/ml) was labeled with 3Hcholesterol (NET-139 Cholesterol, [1,2-3H(N)], Perkin Elmer (Waltham, MA, USA), 40 Ci/mmol, 1 mCi/ml (37 MBq/ml) in ethanol) as follows: 750 μl of 3H-cholesterol (1 mCi/ml) in solution in ethanol were placed in a glass vial and ethanol was evaporated under nitrogen. Once all ethanol was evaporated, 3.75 mg of AcLDL were added, the vial tightly capped and incubated at 37 °C for 18 h. After incubation, free label was eliminated by purifying AcLDL through a HiTrap desalting column (Sephadex G-25 Superfine, GE Healthcare Bio-Sciences AB, Uppsala, Sweden). On the day of the study, 3H-cholesterol AcLDL was adjusted to 40 × 106 dpm/ml with saline/1 mM EDTA. 200 μl of the resulting solution, containing 8 × 106 dpm 3H-Cholesterol AcLDL, was injected intravenously (i.v.) into each mouse (via tail vein). Clearance of 3H-cholesterol AcLDL was measured in mice anesthetized with sodium pentobarbital (60 mg/kg). Mice were i.v. injected with 200 μl of 3H-cholesterol AcLDL solution (adjusted to 40 × 106 dpm/ml), and blood samples collected by retro-orbital bleed at 1, 2, 5, 10 and 30 min post-dosing. After the last sample was collected, mice were euthanized and livers collected for measurement of 3H-cholesterol and processed as described below. Plasma samples were counted in 10 ml of scintillation cocktail. Clearance of 3Hcholesterol AcLDL was expressed as percentage of total injected dpm remaining in serum at any given time. Four mice from each genotype were used in this study. Reappearance of 3H-cholesterol was measured in conscious mice. Mice were injected i.v. with 200 μl of 3H-cholesterol AcLDL solution (adjusted to 40 × 106 dpm/ml), and mice were then individually housed in wire bottom-fitted cages in order to collect feces. Blood samples were collected at 0.5, 3, 6 (via retro-orbital bleed, under isoflurane anesthesia) and 24 h (via cardiac bleed in mice euthanized with CO2) post-dosing. After euthanasia, liver was collected for measurement of 3H-cholesterol. The whole liver was mixed with 5 ml of 2.5 M KOH, followed by incubation for 2 h at 60 °C. Subsequently, 5 ml of absolute ethanol was added to all samples, and mixed thoroughly by vortexing. Subsequently, 10 ml of hexane was added to each tube, and after vigorous vortexing, the aqueous and organic phases were allowed to separate. 2.5 ml of the organic phase was mixed with 10 ml of scintillation cocktail (Scintisafe 2, Fisher Scientific, Pittsburgh, PA, USA) and radioactivity measured using a β-counter. The remaining organic phase was removed, and samples re-extracted with 10 ml of hexane. The aqueous phase was

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

bile acids counted. 24 h feces were collected and dried overnight at room temperature. Feces were mixed with 10 ml of 2.5 M KOH, and the extraction of neutral sterols carried out as described above for liver. All 3 H values are expressed as percentage of injected dpm.

then acidified with concentrated HCl (12 M), at a 1 ml HCl: 5 ml aqueous phase ratio. After acidification, 12 ml of ethyl acetate was added, and phases were allowed to separate. 2.5 ml of the organic phase was mixed with 10 ml of Scintisafe 2 and radioactivity associated with

A

1351

BH

Neo

Targeting vector

Wildtype allele

H

B

1

H

2

Chromosome 3

BH

Neo

Mutant allele

H

B

1

Wildtype fragment Mutant fragment

H

2

BH

BH

15 6 kBp ~15.6 ~12.9 kBp

Probe

B

C 1 2 3 4 5 6 7 8 T E

1

2

3

4

5

7

6

9

11

13

15

17

19

21

T T E

E

W

H

K

W H 1 H2 H3 H4

K

D P2Y 13 140 0

125

50 25

* WT

P2Y13-/-

Male

* WT

P2Y13-/-

Female

100 80 60 40

*

120

mRNA (AU)

mRNA (AU)

mRNA (AU)

75

140

*

120

100

0

LacZ

LacZ

100 80 60 40 20

20

0

0 WT

P2Y13-/-

Male

WT

P2Y13-/-

Female

Fig. 1. Targeted disruption of the mouse P2Y13 gene. (A) Targeting strategy. Shown are the structures of the targeting construct, the wild type allele and the disrupted (mutant) P2Y13 allele. Homologous recombination of the targeting construct with the wildtype allele on chromosome 3 results in the replacement of 26 base pairs in exon 2 of P2Y13 by a NeomycinLacZ (Neo) cassette. This Neo-lacZ cassette contains BamHI (B) and HindIII (H) restriction sites, which can be used in Southern blot analysis of ES cell and genomic tail DNA. (B) Southern blot of genomic DNA isolated from embryonic stem cells, digested with HindIII (lanes 1–4) or BamHI (lanes 5–8) and hybridized with a DNA fragment located outside of the 5' arm of the targeting construct (indicated in A). Lanes 1, 2, 5 and 6 contain genomic DNA from parental ES cells; lanes 4 and 8 contain DNA isolated from ES cells (line 6800) that were used for generation of the described knockout mice. Lanes 3 and 7 contain DNA isolated from a second ES line that was positive for recombination, but not used for the generation of mice. E: Endogenous P2Y13 fragment; T: Targeted P2Y13 fragment. (C) Genomic PCR on tail DNA from wild type (W), heterozygous (H) and homozygous (K) mutant P2Y13 mice. All odd-numbered lanes show multiplexed PCR reactions with three primers that allow for the identification of both wild-type as well as mutant allele. All evennumbered lanes show PCR reactions with two primers that recognize the endogenous P2Y13 allele only. The left-hand panel (lanes 1–6) shows a schematic of expected PCR fragment patterns and sizes. The right-hand panel (lanes 7–22) shows the generated PCR fragments. DNA extracted from wild type (WT DNA) and targeted ES cells (ES DNA) were used as controls (lanes 19 + 20 and 20 + 21, respectively). H1–H4 indicate four independent heterozygous mutant mice. (D) Detection of P2Y13 and LacZ mRNA by quantitative real time PCR (TaqMan) on liver RNA isolated from male and female homozygous P2Y13 knockout (−/−) and wild type (WT) littermate mice. Shown are the relative RNA levels in arbitrary units (AU ± SEM), determined from three individual mice per group, 4 PCR reactions each. *p b 0.01 WT vs. KO.

1352

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

2.3. RNA microarray analysis

the RNA sample was performed using Promega SV Total RNA Isolation System. Total RNA concentration and purity were assessed by NanoDrop ND-1000. Total RNA (2 μg) was amplified to cRNA following standard protocols by Affymetrix. cRNA was hybridized to an Affymetrix Merck custom array containing probesets for 38,385 transcripts. Washing and scanning was performed following standard protocols. Microarray data has been deposited in the GEO

Livers from 11-week-old mice on regular chow diet were isolated, immediately snap-frozen in liquid nitrogen and stored at −80 °C until needed. RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. After the ethanol precipitation step in the TRIzol extraction procedure, a cleanup of

A

Body weight Change to HFHC diet

Weight (grams)

50

Male WT Male P2Y13-/-

40

Female WT Female P2Y13-/-

30

20

10 5

6

7

8

9

10

11

12

13

14

15

16

17

age (weeks)

B

Bone mineral content

Bone area *

0.40

8.0

bone area (cm2)

0.35

BMC (g)

0.30 0.25 0.20 0.15 0.10 0.05

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0

0 WT

P2Y13-/-

WT

Male

P2Y13-/-

WT

Female

C

P2Y13-/-

WT

Male

% Fat Mass

P2Y13-/-

Female

% Lean Mass

40

70

30

60

lean (%)

fat (%)

*

20

50

40

10

30

0 WT

P2Y13-/-

Male

WT

P2Y13-/-

Female

WT

P2Y13-/-

Male

WT

P2Y13-/-

Female

Fig. 2. Phenotypic analysis of P2Y13 knockout mice. (A) Body weight of P2Y13 knockout and wild type littermates were monitored from 6 through 16 weeks of age, for both genders. At 8 weeks of age, mice were placed on HFHC diet. Depicted are average values ± SEM derived from 8 wild type male, 7 P2Y13−/− male, 9 wild type female and 9 P2Y13−/− female mice. (B) Bone mineral content and bone area were determined in 7-week-old mice by DEXA scan. Bars indicate average values, dots individual measurements from 9 wild type male, 8 P2Y13−/− male, 9 wild type female and 9 P2Y13−/− female mice. *p b 0.05 WT vs. KO. (C) Fat mass and lean mass were determined by qNMR on 23-week-old mice, fed a high fat and high carbohydrate diet for 15 weeks. Depicted are average values± SEM derived from 8 wild type male, 8 P2Y13−/− male, 9 wild type female and 9 P2Y13−/− female mice. (D) Oral Glucose Tolerance Test was carried out on 16-week-old mice that were fed HFHC diet for 8 weeks. Depicted are average values± SEM. Numbers same as in panel A. (E) Intraperitoneal insulin sensitivity tests were carried out on 18-week-old mice that were fed HFHC diet for 10 weeks. Depicted are average values± SEM. Numbers same as in panel A. *p b 0.05 WT vs. KO.

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

D

OGTT

glucose (mg/dL)

450

1353

Male WT

400

Male P2Y13-/-

350

Female WT

300

Female P2Y13-/-

250 200 150 100 50 0 0

25

50

75

100

125

150

175

200

time (min)

E

IPIST 225

Male WT

200

Male P2Y13-/-

glucose (mg/dL)

175

Female WT

150

Female P2Y13-/-

125 100 75 50 25

*

*

0 0

10

20

30

40

50

60

70

80

90

100

time (min) Fig. 2 (continued).

database (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE20920. Raw intensity data was normalized using the RMA algorithm [17]. Subsequently, data was prefiltered to remove transcripts detected as "absent" by using the MAS5 algorithm (Affymetrix) with p N 0.05 in at least 50% of replicates in all treatment groups [18]. 14413 probesets (sequences) remained for further analysis. The Rosetta error model was used to estimate variance confidence. Probesets with Rosetta error model p N 0.05 were kept but depicted in grey in the heatmaps to indicate their lower confidence. ANOVA (p b 0.05) was used to identify differentially expressed genes. A threshold of 1.2 on the fold change of the combined replicates was imposed to select only for robust changes in gene expression (above 20% transcript abundance). Gene set enrichment was performed against Gene Ontology, KEGG, Swissprot and Panther gene families. Pathway analysis was performed by using the Ingenuity Pathway Analysis tool (Ingenuity® Systems, www.ingenuity.com). For the verification of transcript level changes, the TaqMan RT-PCR kit from Applied Biosystems (Foster City, CA, USA) was used. In a 96well plate, 5 μl (500 ng) of total RNA from each liver sample was added to 95 μl of the master mix (consisting of 1× Taqman RT buffer, 5.5 μM MgCl2, 500 μM of each dNTP, 2.5 μM random hexamer, 0.4 U/μl of RNase inhibitor, and 1.25 U/μl of Reverse Transcriptase), followed by incubations at 25 °C for 10 min, 48 °C for 30 min and 95 °C for 5 min. The generated cDNA was then used to amplify P2Y13 cDNA using ABI primer/probes Mm00546978_m1 and Mm01951265_s1, which recognize the exon1-exon2 boundary sequence and the 3'UTR of P2Y13 mRNA, respectively. The Neomycin/LacZ cassette used for ablation of the P2Y13 gene contains its own promotor, transcriptional

termination and mRNA polyadenylation signals. Therefore, insertion of this Neo/LacZ cassette into the P2Y13 gene should lead to an ablation in P2Y13 mRNA expression. mRNA amplification was carried out using the ABI PRISM 7900HT System (Applied Biosystems). Assays were performed in 384-well optical plates in triplicate, each reaction containing 4 μl of cDNA and 16 μl of TaqMan PCR reagent mixture consisting of 1× Master Mix, 10 μM P2Y13 primer/probe (or 10 μM GAPDH primer/probe) and RNase-free water. Results were expressed in fold difference (calculated as 2−ΔΔCt) between P2Y13 knockout and wild type mice. The Graphpad Prism 5 program was used to determine statistically significant differences in RNA expression and p-values are indicated in each respective figure. 3. Results 3.1. Generation of P2Y13 knockout mice In order to gain insight into the physiological role of P2Y13 and its possible involvement in the metabolism of HDL, we set out to study the effect of ablating its expression in the mouse. The P2Y13 open reading frame was disrupted by homologous recombination with a Neomycin-LacZ cassette-containing targeting vector which resulted in the deletion of 26 base pairs from exon 2 (see Fig. 1A). The resulting open reading frame encodes a truncated form of P2Y13, which consists of the first 29 amino acids and lacks all seven putative transmembrane domains, rendering it non-functional. The genomic disruption of P2Y13 was confirmed by Southern blot analysis of embryonic stem cell DNA as well as by PCR on genomic tail DNA from generated mice (Fig. 1B and C). The Neomycin-LacZ cassette contains

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

its own promoter, transcription termination and mRNA polyadenylation sequences. Insertion of this cassette into the P2Y13 gene was therefore anticipated to result in ablation of P2Y13 mRNA expression in the generated knockout mice. To verify this, we performed quantitative RT-PCR analysis on liver mRNA, which showed a total lack of P2Y13 expression in both male and female P2Y13−/− mice (Fig. 1D). The distribution of the genotypes of the F2 offspring was in accordance with Mendelian distribution, indicating no detrimental effect of P2Y13 ablation in the mouse. In addition, mice that were homozygous for the 26 base pair P2Y13 deletion were viable, fertile and showed no gross abnormalities. 3.2. Body weight and bone density analysis in P2Y13−/− mice The generated P2Y13 knockout mice were backcrossed five generations to C57Bl/6 mice (96.9% B6N, 3.1% SVJ129) and subjected to extensive phenotypic analyses, using wild type littermates as controls. To assess whether ablation of P2Y13 had any effect on the body weight of mice, we recorded their body weights once a week from the age of 6 to 16 weeks. At 8 weeks old, the mice were switched from a regular chow diet to a high fat and high carbohydrate (HFHC) diet. As shown in Fig. 2A, the body weights of wild type control mice and P2Y13 knockout mice were comparable on a regular chow diet. Also upon feeding an HFHC diet, no significant differences between P2Y13 knockout and wild type mice could be detected, for both genders. At 7 weeks of age, the bone density of P2Y13 knockout mice was determined by measurement of the bone mineral content and bone area through Dual Energy X-ray Absorption (DEXA) scan. This analysis showed significant increases in bone area for male as well as female P2Y13 knockout mice (Fig. 2B).

mice, for both genders (Fig. 3). These analyses showed significant reductions in glycerol and fatty acid levels in female P2Y13 knockout mice (Fig. 3). 3.5. Altered lipoprotein levels in P2Y13 knockout mice Since P2Y13 has been suggested to play a role in the metabolism of lipoprotein particles, we measured total cholesterol levels in P2Y13 knockout and wild type control mice on chow diet. Total cholesterol levels were found to be significantly lower in both male (−11%) and female (−26%) P2Y13 knockout mice when compared to wild type mice (Fig. 4A). Given that total cholesterol levels reveal no information on the nature of lipoprotein particles (HDL, LDL, VLDL) or changes in their plasma levels, we subjected the plasma of these

A

Free Fatty Acids 1.5 1.2

FFA (mmol/L)

1354

0.9 0.6 0.3 0.0

WT

3.3. Normal body composition of P2Y13 knockout mice

WT

P2Y13-/-

Female

B

Glycerol 0.50

Glycerol (mmol/L)

Quantitative NMR (qNMR), which allows for the simultaneous detection of body fat, lean body mass and body fluids, was used to determine differences in these physical properties between P2Y13 knockout and control mice. qNMR was carried out on 23-week-old mice, after feeding an HFHC diet for 15 weeks. This analysis did not show any statistically significant differences between P2Y13 knockout and control mice for body weight, fat mass and lean mass, for both male as well as female mice (Fig. 2C).

P2Y13-/-

Male

0.40 0.30 0.20 0.10

3.4. Glucose tolerance and insulin sensitivity tests 0.00

WT

P2Y13-/-

Male

C

WT

P2Y13-/-

Female

Triglycerides 0.8 0.7 0.6

TG (mmol/L)

The effect of P2Y13 deletion on glucose tolerance was measured by performing an oral glucose tolerance test (OGTT) in 16-week-old mice, after 8 weeks of feeding an HFHC diet. As shown in Fig. 2D, in neither male nor female P2Y13 knockout mice could we detect a difference in the clearance of glucose from the plasma compared to their wild type litter mate controls, indicating that P2Y13 does not have an effect on glucose tolerance in mice. Also the effect of deletion of P2Y13 on insulin sensitivity was measured, by monitoring the plasma glucose levels after intraperitoneal administration of a fixed dose of insulin in 18-week-old mice that were fed a high fat and high carbohydrate diet for 10 weeks. Although the female P2Y13 knockout mice had significantly lower glucose levels at the onset of the experiment than wild type mice, the clearance rate of glucose upon administration of insulin was not different in these knockout mice as compared to female wild type mice (Fig. 2E). The glucose measurements upon insulin administration in male mice showed a somewhat different picture. Although the P2Y13 knockout mice showed somewhat higher glucose level at the zero time point, the clearance of glucose upon administration of insulin was not statistically different from that in wild type mice (Fig. 2E). At 8 weeks of age, the plasma levels of free fatty acids, glycerol and triglycerides were measured in both P2Y13 knockout and wildtype

0.5 0.4 0.3 0.2 0.1 0.0

WT

P2Y13-/-

Male

WT

P2Y13-/-

Female

Fig. 3. Metabolite plasma levels of P2Y13 knockout mice. At 8 weeks of age, mice fed a regular chow diet were subjected to measurement of plasma levels of free fatty acids (A), glycerol (B) and triglycerides (C). Bars indicate average values, dots individual measurements from 7 wild type male, 9 P2Y13−/− male, 9 wild type female and 9 P2Y13−/− female mice. *p b 0.05 WT vs. KO.

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

A

B

Total Cholesterol 140

p=0.047

p=0.039

p=0.023

p=0.045

100

mg/dL L

100

mg/d dL

HDL 120

120

1355

80 60

80 60 40

40

20

20 0

0 WT

P2Y13-/-

WT

Male

C

P2Y13-/-

P2Y13-/-

ns

WT

Male

D

LDL 30

WT

Female

VLDL 25

p=0.01

25

P2Y13-/-

Female

ns

ns

20

mg/dL

mg/dL

20 15 10

15 10 5

5 0

0 WT

P2Y13-/-

Male

WT

P2Y13-/-

Female

WT

P2Y13-/-

Male

WT

P2Y13-/-

Female

Fig. 4. Lipoprotein profile of P2Y13 knockout mice. Serum was isolated from 8-week-old mice (6 per genotype and gender on regular chow diet). One aliquot was subjected to total cholesterol measurement (A), the remainder was subjected to size exclusion chromatography. Continuous cholesterol detection of the eluate allowed for the determination of cholesterol in the HDL (B), LDL (C) and VLDL (D) fractions. Indicated are average values of 6 determinations ± SEM. ns: not significant.

mice to size exclusion chromatography (Appendix A). As shown in Fig. 4B, both male and female P2Y13 knockout mice had slightly, but significantly lower plasma HDL levels (−13.1% ± 3.2% and −17.5% ± 4.0%, respectively) in comparison to wild type control mice. Plasma LDL levels were found to be unchanged in male P2Y13 knockout mice, while female knockout mice showed significantly lower LDL cholesterol levels when compared to wild type mice (Fig. 4C). No significant changes could be detected in the VLDL levels of P2Y13 knockout mice compared to wild type mice (Fig. 4D). Given the effect of P2Y13 deletion on plasma lipoprotein levels, we assessed the effect of these lipid changes on reverse cholesterol transport (RCT) in P2Y13 knockout mice. A surrogate for RCT is the flux of isotopically labeled cholesterol from macrophages to the liver. As shown in Fig. 5A (left-hand panel), the clearance of intravenously administered, acetylated 3H-cholesterol-LDL is a rapid process that is identical in P2Y13 knockout mice and wild type littermates. Previous experiments have shown this fast clearance to be most likely mediated by liver-resident macrophages in a scavenger-receptor dependent manner [19]. The reappearance of 3H-cholesterol in HDL particles in the circulation of mice can be measured after longer periods of time and was significantly lower in P2Y13 knockout mice as compared to wild type littermates (Fig. 5A, right-hand panel). The area under the curve (AUC) for wild type mice was 96 ± 3.3 (mean ± SEM), compared to 81.8 ± 3.8 in P2Y13 knockout mice (p b 0.05). In P2Y13 knockout mice, significantly less 3H-cholesterol was detected in the total plasma cholesterol pool (3.2 × 105 ± 0.05 × 105 dpm/mg cholesterol) compared to wild type mice (3.5 × 105 ± 0.08 × 105 dpm/mg cholesterol; p b 0.01). Throughout the experiment, hepatic levels (a composite of hepatocytes and liver-resident macrophages) of isotope-labeled cholesterol and bile acids remained identical

between P2Y13 knockout and wild type mice (Fig. 5B). However, the amount of isotope-labeled cholesterol in the feces of P2Y13 knockout mice detected over a 24-h period was ~25% lower than that of their wild type littermates (Fig. 5C). The amount of fecal isotope-labeled bile acids showed no statistical difference between P2Y13 knockout and wild type mice (Fig. 5C, right-hand panel). Also the amount of feces collected over 24 h was no different between wild type and knockout mice (not shown). Taken together, these results indicate that reverse cholesterol transport is decreased in P2Y13 knockout mice. 3.6. Hepatic mRNA profiling analysis Given the changes in lipoprotein metabolism in P2Y13 knockout mice and the central role of the liver in this process, we assessed gene expression changes in this organ by making use of microarray mRNA profiling. When only mRNA expression changes of more than 20% were considered, we found 3471 mRNA sequences that were significantly different (p b 0.05) between male wild type and P2Y13 knockout mice as well as female wild type and P2Y13 knockout mice. Supervised clustering analysis revealed striking differences in mRNA expression changes between male and female P2Y13 knockout mice (Fig. 6A). In both genders, approximately 50% of the changes concerned upregulation of mRNA expression. Of the 1883 and 1694 mRNA sequences that were changed in male and female P2Y13 knockout mice, respectively, only a small number was common between both genders (see Fig. 6A). As shown in Fig. 6A, the nature of the changes in hepatic mRNA expression are different between male and female P2Y13 knockout mice and suggests that ablation of P2Y13 has a different impact on metabolic processes in each gender and is in

1356

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

A

ClearanceReappearance

40 30

Plasma % injected DPM

20 10

WT

5 P2Y13-/-

0 1

5

10

30

1

3

6

minutes

24

hours

Neutral Sterols (% injected DPM in Liver)

60 WT

50

KO

40 30 20 10 0

2.0 WT KO

1.5

1.0

0.5

0.0

30 min

C

Bile Acids (% injected DPM in Liver)

B

24 hours

30 min

Neutral sterols

Bile acids 6

*

1.5

1.0

0.5

0.0 Wild Type

P2Y13-/-

Bile acids (% injected DPM in Feces)

N Neutral sterols

2.0

(% Injected DPM in Feces)

24 hours

5 4 3 2 1 0 Wild Type

P2Y13-/-

Fig. 5. Kinetics of 3H-cholesterol in P2Y13 knockout mice. (A) Kinetics of 3H-cholesterol in the P2Y13 knockout mice. 17-week-old male wild type or P2Y13 knockout mice (n = 4) were injected with 3H-cholesterol-AcLDL, and the fate of the label determined over time. Clearance was measured over a 30-min period, whereas reappearance into the circulation (as a surrogate for RCT) was measured for a 24-h period. 3H was rapidly cleared from the circulation in both groups of animals. However, reappearance of 3H was decreased in the P2Y13 knockout mice (AUCWT = 96.5 ± 3.3, vs AUCKO = 81.8 ± 3.8, mean ± SEM, p b 0.05). (B) Incorporation of 3H into liver neutral sterols and bile acids. Livers were collected at the termination of the time-courses described in panel A, and neutral sterols and bile acids extracted from whole livers. Radioactivity associated with each fraction was measured by scintillation counting. No significant differences were observed between the two genotypes in either fraction. (C) Incorporation of 3H into fecal neutral sterols and bile acids. Feces were collected continuously for 24 h, and neutral sterols and bile acids extracted. Radioactivity associated with each fraction was measured by scintillation counting. There were no differences in the amount of radioactivity associated with the bile acid fraction, but P2Y13 knockout mice had significantly lower levels of 3H associated with the neutral sterol fraction suggesting a decrease in RCT. *p b 0.01.

line with some of the observed phenotypic differences between the two genders described above. 3.7. Pathway and network analysis of hepatic mRNA signatures To gain insight into the metabolic processes that were affected in P2Y13 knockout mice, we performed gene set enrichment analyses. Significant enrichment was found for sets of genes involved in the metabolism of fatty

acids, steroids, lipids and cholesterol (p=1.49×10− 7; Fig. 6B). Ingenuity pathway analysis was used to determine molecular pathways and networks enriched in the P2Y13 knockout signature. A network around Srebp1 was found to be upregulated in female P2Y13 KO mice (Fig. 6C), with changes in genes like Ldlr, Sqle, Insig1, Hmgcs1, Dhcr7 and Sc5dl, all of which are involved in cholesterol synthesis or metabolism. This suggests an upregulation of fatty acid and cholesterol synthesis genes. Conversely, hepatic β-oxidation was one of the downregulated

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

1357

A 896

828

758

699

228 62

P2Y13 WT Male P2Y13 WT Female P2Y13 KO Male P2Y13 KO Female

B Gene Set Lipid, fatty acid and steroid metabolism Select regulatory g y molecule Steroid metabolic process Steroid biosynthetic process Basolateral plasma membrane cholesterol metabolic process Steroid metabolism

Expectation

P-Value

1.69E-04 1.87E-03 6.07E-03 3.58E-02 3.98E-02 4.16E-02 9 . 62 E-02

1.49E-07 1.57E-06 5.37E-06 3.21E-05 3.41E-05 3.75E-05 8.91E- 05

C n=1694 sequences

Fig. 6. P2Y13 gene expression signature. (A) mRNA profiling shows that ablation of P2Y13 affects gene expression in a gender-specific manner. Several distinct gene clusters are observed specifically for males (top for each genotype) and females (bottom for each genotype). Changes for most of the genes are common, but in most cases did not reach significance in one of the sexes. Shown are the results of a one-way ANOVA analysis (p b 0.05) for 3471 transcripts whose relative expression changes were more than 20%. Results were obtained from 11-week-old mice on chow diet; 6 mice per genotype and gender. (B) Gene set enrichment analysis shows lipid metabolism and cholesterol synthesis as significantly enriched biological processes in the female P2Y13 knockout signature. (C) Shown is the Ingenuity network around the transcription factor Srebp. Nodes in red are upregulated mRNA changes. Srebp and several of its gene targets are differentially expressed in female P2Y13 knockout mice. Network was generated using Ingenuity Pathway Analysis tool (Ingenuity® Systems, www.ingenuity.com).

1358

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

pathways, with acyl-CoA dehydrogenase (Acad9; EC 1.3.99.3) and acetylCoA C-acetyltransferase (Hadha; EC2.3.1.9) transcript levels decreased by more than 20% (not shown). The encoded proteins are involved in the oxidation of fatty acids, suggesting a lower level of catabolism of fatty acids in the liver of P2Y13 knockout mice. We subsequently subjected these function-enriched gene sets to an intersection with liver and adipose tissue Bayesian networks [20,21]. This intersection analysis identified the significant (N20%; p = 0.05) upregulation of genes involved in the synthesis of cholesterol and fatty acids in the liver of P2Y13 knockout mice (Fig. 7). The increase in Srebpregulated genes involved in fatty acid and cholesterol synthesis, as well as the reduction in fatty acid oxidation genes suggests that the P2Y13 knockout mice have adjusted their hepatic mRNA expression to accommodate the metabolic changes detected in these mice. 4. Discussion In this study, we describe the phenotypic analysis of mice deficient in the purinergic G protein coupled receptor P2Y13. Although the expression of P2Y13 is highest in human tissues such as the spleen, brain, placenta as well as in neutrophils and mononuclear cells, experimental studies have mainly focussed on a role for P2Y13 in the liver, where expression levels are modest [14–16]. Jacquet and colleagues [13] demonstrated a role for P2Y13 in the uptake of HDL by HepG2 hepatocytes: upon binding of Apolipoprotein AI to membraneassociated F1-ATPase, ATP is hydrolyzed to ADP, which in turn binds and activates the purinergic receptor P2Y13. The activation of P2Y13 was shown to lead to an increase in HDL uptake by HepG2 cells and to depend on the presence of functional P2Y13, since knockdown of its mRNA expression prevented the increased uptake of HDL [13]. Moreover, direct stimulation of P2Y13 by administration of ADP or the partial agonist AR-C69931MX to HepG2 cells or to isolated and perfused mouse livers also resulted in increased uptake of HDL. These results suggested that P2Y13 activation is downstream of the F1-ATPasedependent ApoAI binding step and that P2Y13 activation alone suffices for the increased HDL uptake by hepatocytes. The results presented in this study support a possible role for P2Y13 in the metabolism of lipoprotein particles. Ablation of this purinergic receptor in mice resulted in decreased fecal sterol concentrations in an in vivo RCT assay and resulted in a small reduction in circulating HDL levels in comparison to wild type mice. No changes in total body weight, fat mass or lean body mass could be detected in these knockout mice, suggesting that the observed changes in fecal sterol and plasma HDL cholesterol concentrations were not the consequence of a decreased food intake in the P2Y13 KO mice. Also no changes could be detected in insulin sensitivity and oral glucose tolerance, however, small increases in bone area were observed in both male and female knockout mice. The observed decrease in circulating HDL levels in P2Y13 knockout mice seems to be at odds with the proposed role for P2Y13 in hepatic uptake of HDL, which suggested that inactivation of this receptor would result in decreased HDL uptake by the liver and hence higher rather than lower circulating levels of HDL. Several scenarios can be envisioned that might account for this apparent discrepancy. First, there is the possibility that total ablation of P2Y13 in vivo leads to compensatory changes in the expression levels of other genes that regulate cellular cAMP concentration or uptake of lipoprotein particles, resulting in a disconnect between in vitro and in vivo observations. A second possibility is that P2Y13 does not alter HDL metabolism directly, but rather alters the turnover of metabolites in both hepatic and extra-hepatic tissues that in turn alter lipoprotein metabolism.

1359

An example of such metabolites could be fatty acids generated by lipolysis in adipocytes that can alter PPARα activity in the liver. In such a case, the ablation of P2Y13 would have different effects in isolated liver cells and in a whole body. Our in vivo reverse cholesterol transport experiments with P2Y13 knockout mice indicate a decrease in cholesterol transport to the feces, arguing for a role of P2Y13 in this process. Given the small decrease in circulating HDL cholesterol levels in the P2Y13 knockout mice, we cannot formally rule out that this decrease in RCT is simply a consequence of lower HDL levels per se, rather than a defect in the on or offloading capacity of HDL to or from cells. That notwithstanding, it does argue for a role for P2Y13 in the metabolism of HDL. The liver RNA profiling analyses show a significant enrichment for upregulated genes that are involved in the synthesis of cholesterol as well as fatty acids, indicative of an increased SREBP tone in the livers of P2Y13 knockout mice. These hepatic mRNA expression changes might not be a direct consequence of P2Y13 ablation, but rather an indirect one: the reduction in reverse cholesterol transport capacity and circulating HDL levels in the P2Y13 knockout mice could conceivably result in a lower influx of cholesterol from HDL into hepatocytes. As a consequence, the SREBP tone in the liver would increase and lead to an upregulation of genes involved in the synthesis of cholesterol. An additional interesting finding from our mRNA expression profile studies is the downregulation of some of the genes involved in fatty acid beta oxidation and the upregulation of several genes involved in the biosynthesis of fatty acids in the liver of P2Y13 knockout mice. While the latter phenomenon can also be explained by an increased SREBP tone in the liver [22], the downregulation of beta oxidation genes suggests a decrease in activity of the nuclear hormone receptor PPARα. PPARα is the main regulator of fatty acid beta oxidation and is activated by fatty acids and certain hypolipidemic drugs [23–25]. Since P2Y13 knockout mice have lower circulating levels of fatty acids, and therefore conceivably lower activation of liver PPARα, the downregulation of genes involved in fatty acid beta oxidation could function as a compensatory mechanism, perhaps to maintain physiological levels of serum fatty acids required as a source of energy for many tissues [26,27]. Although the studies described above focus on the elimination of P2Y13 activity, it is very well possible that increasing P2Y13 activity would have the opposite physiological effects, such as an increase in the flux of cholesterol from peripheral cells to the feces and a small increase in plasma HDL levels. Agonizing this G protein coupled receptor would therefore possibly be of benefit for the treatment of cardiovascular diseases. An additional advantage of agonizing P2Y13 could be the activation of hepatic PPARα. Such activation would lead to an increase in the rate of fatty acid beta oxidation and thus benefit insulin sensitivity [24,26]. In addition, high plasma HDL levels have been shown to be of benefit to the cardiovascular system [25,28,29]. We currently do not know the possible negative effects of stimulating P2Y13 in vivo, but liver RNA profiling and lipoprotein level measurements in mice either overexpressing or treated with a P2Y13-specific agonist should give disclosure on this matter. Although transgenic mice overexpressing P2Y13 have not been generated yet, the P2Y12 antagonist AR-C69931MX has been shown to act as a partial agonist on P2Y13 [14,30]. Since AR-C69931MX has a very short half-life, in vivo mouse experiments with this compound will be difficult to carry out, but modifications of it might make it possible to obtain a small molecule agonist that can be used for in vivo experiments. This should allow for disclosure on the suitability of P2Y13 as a target for therapeutic intervention and for additional insight into the physiological role of this G Protein Coupled Receptor and its mechanism of action in lipoprotein metabolism.

Fig. 7. Molecular networks in the P2Y13 knockout liver mRNA signature. The mouse liver and adipose Bayesian network is enriched for genes present in the female P2Y13 knockout signature. Nodes indicated in blue represent genes related to lipid metabolism.

1360

D. Blom et al. / Biochimica et Biophysica Acta 1801 (2010) 1349–1360

Supplementary data to this article can be found online at doi:10.1016/ j.bbalip.2010.08.013. Acknowledgments We thank Lauretta Le Voci and Elizabeth Polizzi Somers for their help with data analysis and manuscript preparation and Dr. Debraj GuhaThakurta for his help and advice.

References [1] G.A. Francis, R.J. Perry, Targeting HDL-mediated cellular cholesterol efflux for the treatment and prevention of atherosclerosis, Clin. Chim. Acta 286 (1999) 219–230. [2] P.P. Toth, M.H. Davidson, Therapeutic interventions targeted at the augmentation of reverse cholesterol transport, Curr. Opin. Cardiol. 19 (2004) 374–379. [3] G.F. Lewis, D.J. Rader, New insights into the regulation of HDL metabolism and reverse cholesterol transport, Circ. Res. 96 (2005) 1221–1232. [4] G.F. Lewis, Determinants of plasma HDL concentrations and reverse cholesterol transport, Curr. Opin. Cardiol. 21 (2006) 345–352. [5] Y. Zhang, A.M. Ahmed, T.L. Tran, J. Lin, N. McFarlane, D.R. Boreham, S.A. Igdoura, R. Truant, B.L. Trigatti, The inhibition of endocytosis affects HDL-lipid uptake mediated by the human scavenger receptor class B type I, Mol. Membr. Biol. 24 (2007) 442–454. [6] T.A. Pagler, S. Rhode, A. Neuhofer, H. Laggner, W. Strobl, C. Hinterndorfer, I. Volf, M. Pavelka, E.R. Eckhardt, D.R. van der Westhuyzen, G.J. Schutz, H. Stangl, SR-BImediated high density lipoprotein (HDL) endocytosis leads to HDL resecretion facilitating cholesterol efflux, J. Biol. Chem. 281 (2006) 11193–11204. [7] D.L. Silver, N. Wang, X. Xiao, A.R. Tall, High density lipoprotein (HDL) particle uptake mediated by scavenger receptor class B type 1 results in selective sorting of HDL cholesterol from protein and polarized cholesterol secretion, J. Biol. Chem. 276 (2001) 25287–25293. [8] F. Rinninger, M. Brundert, I. Brosch, N. Donarski, R.M. Budzinski, H. Greten, Lipoprotein lipase mediates an increase in selective uptake of HDL-associated cholesteryl esters by cells in culture independent of scavenger receptor BI, J. Lipid Res. 42 (2001) 1740–1751. [9] M. Brundert, J. Heeren, H. Greten, F. Rinninger, Hepatic lipase mediates an increase in selective uptake of HDL-associated cholesteryl esters by cells in culture independent from SR-BI, J. Lipid Res. 44 (2003) 1020–1032. [10] L.O. Martinez, S. Jacquet, J.P. Esteve, C. Rolland, E. Cabezon, E. Champagne, T. Pineau, V. Georgeaud, J.E. Walker, F. Terce, X. Collet, B. Perret, R. Barbaras, Ectopic beta-chain of ATP synthase is an apolipoprotein A-I receptor in hepatic HDL endocytosis, Nature 421 (2003) 75–79. [11] J.R. Morrison, G.A. McPherson, N.H. Fidge, Evidence for two sites on rat liver plasma membranes which interact with high density lipoprotein, J. Biol. Chem. 267 (1992) 13205–13209. [12] R. Barbaras, X. Collet, H. Chap, B. Perret, Specific binding of free apolipoprotein A-I to a high-affinity binding site on HepG2 cells: characterization of two highdensity lipoprotein sites, Biochemistry 33 (1994) 2335–2340. [13] S. Jacquet, C. Malaval, L.O. Martinez, K. Sak, C. Rolland, C. Perez, M. Nauze, E. Champagne, F. Terce, C. Gachet, B. Perret, X. Collet, J.M. Boeynaems, R. Barbaras, The nucleotide receptor P2Y13 is a key regulator of hepatic high-density lipoprotein (HDL) endocytosis, Cell. Mol. Life Sci. 62 (2005) 2508–2515.

[14] D. Communi, N.S. Gonzalez, M. Detheux, S. Brezillon, V. Lannoy, M. Parmentier, J.M. Boeynaems, Identification of a novel human ADP receptor coupled to G(i), J. Biol. Chem. 276 (2001) 41479–41485. [15] T. Wittenberger, H.C. Schaller, S. Hellebrand, An expressed sequence tag (EST) data mining strategy succeeding in the discovery of new G-protein coupled receptors, J. Mol. Biol. 307 (2001) 799–813. [16] D.K. Lee, T. Nguyen, K.R. Lynch, R. Cheng, W.B. Vanti, O. Arkhitko, T. Lewis, J.F. Evans, S.R. George, B.F. O'Dowd, Discovery and mapping of ten novel G proteincoupled receptor genes, Gene 275 (2001) 83–91. [17] R.A. Irizarry, B.M. Bolstad, F. Collin, L.M. Cope, B. Hobbs, T.P. Speed, Summaries of Affymetrix GeneChip probe level data, Nucleic Acids Res. 31 (2003) e15. [18] D. Greco, D. Leo, U. di Porzio, C. Perrone Capano, P. Auvinen, Pre-filtering improves reliability of Affymetrix GeneChips results when used to analyze gene expression in complex tissues, Mol. Cell. Probes 22 (2008) 115–121. [19] T.J. Van Berkel, A. Van Velzen, J.K. Kruijt, H. Suzuki, T. Kodama, Uptake and catabolism of modified LDL in scavenger-receptor class A type I/II knock-out mice, Biochem. J. 331 (Pt 1) (1998) 29–35. [20] E.E. Schadt, C. Molony, E. Chudin, K. Hao, X. Yang, P.Y. Lum, A. Kasarskis, B. Zhang, S. Wang, C. Suver, J. Zhu, J. Millstein, S. Sieberts, J. Lamb, D. GuhaThakurta, J. Derry, J.D. Storey, I. Avila-Campillo, M.J. Kruger, J.M. Johnson, C.A. Rohl, A. van Nas, M. Mehrabian, T.A. Drake, A.J. Lusis, R.C. Smith, F.P. Guengerich, S.C. Strom, E. Schuetz, T.H. Rushmore, R. Ulrich, Mapping the genetic architecture of gene expression in human liver, PLoS Biol. 6 (2008) e107. [21] Y. Chen, J. Zhu, P.Y. Lum, X. Yang, S. Pinto, D.J. MacNeil, C. Zhang, J. Lamb, S. Edwards, S.K. Sieberts, A. Leonardson, L.W. Castellini, S. Wang, M.F. Champy, B. Zhang, V. Emilsson, S. Doss, A. Ghazalpour, S. Horvath, T.A. Drake, A.J. Lusis, E.E. Schadt, Variations in DNA elucidate molecular networks that cause disease, Nature 452 (2008) 429–435. [22] J.D. Horton, N.A. Shah, J.A. Warrington, N.N. Anderson, S.W. Park, M.S. Brown, J.L. Goldstein, Combined analysis of oligonucleotide microarray data from transgenic and knockout mice identifies direct SREBP target genes, Proc. Natl Acad. Sci. USA 100 (2003) 12027–12032. [23] C. Dreyer, G. Krey, H. Keller, F. Givel, G. Helftenbein, W. Wahli, Control of the peroxisomal beta-oxidation pathway by a novel family of nuclear hormone receptors, Cell 68 (1992) 879–887. [24] J.P. Berger, T.E. Akiyama, P.T. Meinke, PPARs: therapeutic targets for metabolic disease, Trends Pharmacol. Sci. 26 (2005) 244–251. [25] J. Plutzky, Peroxisome proliferator-activated receptors as therapeutic targets in inflammation, J. Am. Coll. Cardiol. 42 (2003) 1764–1766. [26] L. Michalik, J. Auwerx, J.P. Berger, V.K. Chatterjee, C.K. Glass, F.J. Gonzalez, P.A. Grimaldi, T. Kadowaki, M.A. Lazar, S. O'Rahilly, C.N. Palmer, J. Plutzky, J.K. Reddy, B.M. Spiegelman, B. Staels, W. Wahli, International Union of Pharmacology. LXI. Peroxisome proliferator-activated receptors, Pharmacol. Rev. 58 (2006) 726–741. [27] S. Kersten, J. Seydoux, J.M. Peters, F.J. Gonzalez, B. Desvergne, W. Wahli, Peroxisome proliferator-activated receptor alpha mediates the adaptive response to fasting, J. Clin. Invest. 103 (1999) 1489–1498. [28] N. Vu-Dac, K. Schoonjans, V. Kosykh, J. Dallongeville, J.C. Fruchart, B. Staels, J. Auwerx, Fibrates increase human apolipoprotein A-II expression through activation of the peroxisome proliferator-activated receptor, J. Clin. Invest. 96 (1995) 741–750. [29] B. Staels, N. Vu-Dac, V.A. Kosykh, R. Saladin, J.C. Fruchart, J. Dallongeville, J. Auwerx, Fibrates downregulate apolipoprotein C-III expression independent of induction of peroxisomal acyl coenzyme A oxidase. A potential mechanism for the hypolipidemic action of fibrates, J. Clin. Invest. 95 (1995) 705–712. [30] F. Marteau, E. Le Poul, D. Communi, C. Labouret, P. Savi, J.M. Boeynaems, N.S. Gonzalez, Pharmacological characterization of the human P2Y13 receptor, Mol. Pharmacol. 64 (2003) 104–112.