The prorenin receptor and its soluble form contribute to lipid homeostasis

Abstract

Obesity is associated with alterations in hepatic lipid metabolism. We previously identified the prorenin receptor (PRR) as a potential contributor to liver steatosis. Therefore, we aimed to determine the relative contribution of PRR and its soluble form, sPRR, to lipid homeostasis. PRR-floxed male mice were treated with an adeno-associated virus with thyroxine-binding globulin promoter-driven Cre to delete PRR in the liver [liver PRR knockout (KO) mice]. Hepatic PRR deletion did not change the body weight but increased liver weights. The deletion of PRR in the liver decreased peroxisome proliferator-activated receptor gamma (PPARγ) and triglyceride levels, but liver PRR KO mice exhibited higher plasma cholesterol levels and lower hepatic low-density lipoprotein receptor (LDLR) and Sortilin 1 (SORT1) proteins than control (CTL) mice. Surprisingly, hepatic PRR deletion elevated hepatic cholesterol, and up-regulated hepatic sterol regulatory element-binding protein 2 (SREBP2) and 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG CoA-R) genes. In addition, the plasma levels of sPRR were significantly higher in liver PRR KO mice than in controls. In vitro studies in HepG2 cells demonstrated that sPRR treatment upregulated SREBP2, suggesting that sPRR could contribute to hepatic cholesterol biosynthesis. Interestingly, PRR, total cleaved and noncleaved sPRR contents, furin, and Site-1 protease (S1P) were elevated in the adipose tissue of liver PRR KO mice, suggesting that adipose tissue could contribute to the circulating pool of sPRR. Overall, this work supports previous works and opens a new area of investigation concerning the function of sPRR in lipid metabolism and adipose tissue–liver cross talk.

NEW & NOTEWORTHY Hepatic PRR and its soluble form, sPRR, contribute to triglyceride and cholesterol homeostasis and hepatic inflammation. Deletion of hepatic PRR decreased triglyceride levels through a PRR-PPARγ-dependent mechanism but increased hepatic cholesterol synthesis through sPRR-medicated upregulation of SREBP-2. Our study highlighted a new paradigm of cross talk between the liver and the adipose tissue involving cholesterol and sPRR.

INTRODUCTION

The obesity epidemic, currently affecting >35% of the population in the United States, is associated with several deleterious changes in lipid metabolism, including increased serum lipid levels, and glucose/insulin homeostasis. Hyperlipidemia is a risk factor for cardiovascular disease (CVD) and is estimated to be responsible for more than half of cardiovascular mortality (1–3).

The prorenin receptor (PRR) is a component of the renin angiotensin system involved in the regulation of blood pressure and fluid volume. PRR is a 350-amino-acid protein with a single transmembrane domain and has been identified as the receptor for (pro)renin, renin in its active form and prorenin in its inactive form. Increasing evidence suggests that the function of PRR is not restricted to blood pressure control (4–6). PRR can also interact with the membrane receptor complex of the Wnt/β-catenin pathway or with the vacuolar-type proton-pump ATPase (H-ATPase) two crucial pathways involved in tissue and organ development (4, 7–10).

Cousin et al. (11) identified a putative cleavage site in the extracellular part of PRR for furin that cleaved off the amino-terminal domain of PRR in the Golgi apparatus to generate a 28 KDa form of PRR, namely, soluble PRR (sPRR). The sPRR can be retained inside cells (12) and secreted into plasma (11) and in the media of cells and urine (13, 14). Indeed, sPRR is released into the media of different cultured cells, including renal cell lines (15), human glomerular epithelial cells (11), vascular smooth muscle cells (15), and adipocytes (16). Site-1 protease (S1P) and A Disintegrin and Metalloproteinase 19 (ADAM19) were also identified as enzymes involved in sPRR generation (15, 17). Evidence indicates that both PRR and sPRR play roles in obesity and energy and lipid metabolism (16, 18–20). During the development of obesity, the adipose tissue produced more PRR and the levels of plasma sPRR increased (16). Additionally, we previously showed that adipose PRR KO prevented the development of obesity and drastically decreased fat mass (16). Paradoxically, the deletion of PRR in adipose tissue also increased plasma sPRR levels (16). Interestingly, PRR is also expressed in the liver, and the deletion of PRR in adipose tissue induced an increase in hepatic PRR expression and hepatic cholesterol and triglycerides levels. To understand the biological function of hepatic PRR, Lu et al. (18) silenced PRR in hepatocytes and found that PRR deletion decreased cellular LDL uptake and reduced SORT1 and low-density lipoprotein receptor (LDLR) proteins levels. In mice, the knockout of PRR in liver, using antisense oligonucleotides (ASO), impaired hepatic LDL clearance (20). However, plasma sPRR levels were not quantified, and the relative contribution of sPRR to lipid homeostasis was not examined. In this study, we first aimed to determine whether the deletion of PRR specifically in liver affects lipid homeostasis by using a different approach. Instead of using ASO, an adeno-associated viral (AAV)-thyroxine binding globulin (TBG) promoter-Cre recombinase vector was administered to PRR floxed mice (liver PRR KO). We next examined plasma sPRR levels and investigated the relative contribution of PRR and sPRR to lipid metabolism.

EXPERIMENTAL PROTOCOL

Animals

All procedures involving animals were conducted with the approval of the University of Kentucky Institutional Animal Care and Use Committee (University of Kentucky IACUC protocol number: 2013-1109) and in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

AAV Experiment

An adeno-associated virus, serotype 8, was generated by the Genetic Technologies CORE of the University of Kentucky. Cre recombinase was driven by a liver-specific promoter, i.e., thyroxine-binding globulin (TBG) promoter (TBG^p-Cre). PRR floxed male mice (8- to 12-mo-old) with loxP sites flanking exon 2 of the PRR gene (PRR^flox/flox) were treated with a single intraperitoneal injection at a dosage of 1.0 × 10¹¹ genome copies (gc) virus (liver PRR KO, n = 5 mice) or with saline (CTL, n = 6 mice). Mice were provided water and normal chow diet ad libitum (18% protein, Global Diet; Teklad Harlan Madison, WI). Three weeks after injection, tissues were harvested, snap-frozen in liquid nitrogen, and stored at −80°C. Blood was collected in tubes (4°C) containing EDTA (0.2 mol/L), centrifuged at 2,500 g for 10 min, and plasma was stored at −80°C. Body composition was measured by nuclear magnetic resonance spectroscopy (Echo MRI, Houston, TX), which assessed whole body fat, lean mass, and total water mass in conscious mice. The mice were placed in cylindrical holders and inserted in the EchoMRI machine for a short scan time (1–2 min).

Quantification of Plasma and Tissues Components

Plasma concentrations of sPRR were quantified using a soluble (Pro)renin Receptor ELISA kit (Immuno-Biological Laboratories Co, Minneapolis, MN). Plasma concentrations of cholesterol and triglyceride were measured using a Wako kit (Wako Chemicals, Richmond, VA). The antibody anti-PRR of the sPRR ELISA kit recognized the same epitope for sPRR and full-length PRR and assessed cleaved sPRR and noncleaved sPRR contents in the tissue but did not recognize the truncated form. Therefore, total cleaved sPRR and noncleaved sPRR contents of epididymal fat (EF) were assessed using the soluble (Pro)renin Receptor ELISA kit (Immuno-Biological Laboratories Co, Minneapolis, MN).

Lipoprotein Cholesterol and Apolipoprotein Distribution

The cholesterol distribution among lipoprotein classes was determined after separation by gel filtration chromatography based on a method described previously (21). An aliquot of plasma was diluted in 0.9% NaCl, 0.05% EDTA/NaN3 and centrifuged at 2,000 g for 10 min to remove any particulate debris. The supernatant was transferred to a glass insert contained in a GC vial. After loading the vial into an autosampler set at 4°C (G1329A, Agilent Technologies, Santa Clara, CA), 40 µL of the sample was injected onto a Superose 6 10/300 or Superose 6 Increase 10/300 (GE Healthcare Life Sciences, Pittsburgh, PA) chromatography column. Under the control of an isocratic pump (G1310A/B, Agilent Technologies, Santa Clara, CA), the sample was separated at a flow rate of 0.4 mL/min with an eluent containing 0.9% NaCl, 0.05% EDTA/NaN3. The column effluent was mixed with a total cholesterol enzymatic reagent (C7510, Pointe Scientific, Canton, MI) running at a flow rate of 0.125 mL/min, and the mixture was passed through a knitted reaction coil (EPOCOD, Aura Industries Inc., San Diego, CA) in a 37°C H₂O jacket. The absorbance of the reaction mixture was read at 500 nm using a variable wavelength detector (G1314F, Agilent Technologies, Santa Clara, CA). The signal was subsequently integrated using an Agilent OpenLAB Software Suite (Agilent Technologies, Santa Clara, CA). For apolipoprotein distribution, plasma was separated by gel filtration chromatography as described above, and then 3-min fractions were collected between 18 and 48 min for further analysis by Western blot.

Quantification of Liver Lipids

Liver lipid content was determined based on the method described by Carr et al. (22). A piece of frozen liver was thawed and minced with a razor blade. Following transfer to a tared 16- × 100-mm glass tube, the wet weight of the tissue was measured using an analytical balance. To extract the lipids from the tissue, 3 mL of 2:1 chloroform:methanol (CHCl₃:MeOH) was added and incubated at 55°C for 2 h. After centrifuging the tube at 1,500 g for 10 min, the lipid extract was transferred to a new 16- × 100-mm glass screw-top tube. The tube containing the extracted liver was washed with 2 mL of 2:1 CHCl₃:MeOH and centrifuged at 1,500 g for 10 min. The lipid extract and wash were combined, and the solvent was evaporated under nitrogen at 55°C. The dried lipid extract was dissolved in 6 mL of 2:1 CHCl₃:MeOH. After the addition of 1.2 mL diluted H₂SO₄ (0.05%, v/v), the sample was vortexed for 20 s, and the phases were separated by centrifugation as described above. The upper aqueous phase was removed and an aliquot (typically 1 mL) of the bottom, lipid-containing organic phase was transferred to a new 16- × 100-mm glass screw-top tube using a volumetric glass pipet. After adding 2 mL of 1% Triton-X100 dissolved in CHCl₃, the organic solvent was evaporated under nitrogen at 55°C. The dried sample was dissolved in 1 mL water and heated at 60°C for 10 min. After vortexing and centrifuging as above, samples dissolved in 2% Triton-X100/water were analyzed for cholesterol and triglycerides using a Wako kit (Wako Chemicals, Richmond, VA). Standards for the triglyceride assay were created using vegetable oil and were dissolved in 2% Triton-X100/water. Organic-solvent-resistant, Teflon-lined caps were used to seal the tubes throughout the protocol.

Tissue RNA Extraction and Quantitative RT-PCR

RNA was extracted from the liver and epididymal fat using the SV Total RNA Isolation System (Promega, Madison, WI). Total RNA was quantified with a NanoDrop 2000 spectrophotometer (Wilmington, DE), and cDNA was synthesized using qscript cDNA SuperMix (Quanta Biosciences, Gaithersburg, MD). Real-time quantitative PCR was performed with PerfeCTa SYBR Green FastMix (Quanta BioSciences, Gaithersburg, MD). All the primers sequences are described in Table 1.

Table 1.

Primer sequences

Target Gene	Forward	Reverse
Human GAPDH	GGAGCGAGATCCCTCCAAAAT	GGCTGTTGTCATACTTCTCATGG
Human PRR	GTGTTTTGGGGAACGAGTTTTAGT	TCCTGGTATAGGCCAATTTCCA
Human SREBP2	AACGGTCATTCACCCAGGTC	GGCTGAAGAATAGGAGTTGCC
Mouse 18 s	GTAACCCGTTGAACCCCATT	CCATCCAATCGGTAGTAGCG
Mouse ABCG5	TGTCCTACAGCGTCAGCAAC	TGAGGATTTGCCTGTCCCAC
Mouse ApoB	ATGGTGAAAGGTCCCATATGCT	TGATGGACCTGCTGTAGCTTG
Mouse CEBPβ	CAAGATGCGCAACCTGGAGA	GACAGCTGCTCCACCTTCTT
Mouse CEBPα	AAACAACGCAACGTGGAGAC	TGTCCAGTTCACGGCTCAG
Mouse COX 2	CACCCTGACATAGACAGTGAAAG	CTGGGTCACGTTGGATGAGG
Mouse FABP4	GGAACCTGGAAGCTTGTCTC	TGATGCTCTTCACCTTCCTG
Mouse HMGCoA-R	GCGTAAGCGCAGTTCCTTC	CTCACAGTCCTTGGATCCTCC
Mouse LDLR	AGGCTGTGGGCTCCATAGG	TGCGGTCCAGGGTCATCT
Mouse LXRβ	CCGACAGAGCTTCGTCC	CCCACAGACACTGCACAG
Mouse MCP-1	TTAAAAACCTGGATCGGAACCAA	GCATTAGCTTCAGATTTACGG GT
Mouse Perilipin	GGTACACTATGTGCCGCTTCC	CTTTGCGCTCCGCCTCT
Mouse PPARγ	GAGAGGTCCACAGAGCTGATT	TCGCTGATGCACTGCCTATG
Mouse PRR	TGGTGGCGGGTGCTTTAGGA	AGCCCGTGGCCGGTGGAATA
Mouse SREBP-1c	GGAACTTTTCCTTAACGTGGGC	AGCATGTCTTCGATGTCGTTCA
Mouse SREBP2	GCGTTCTGGAGACCATGGA	ACAAAGTTGCTCTGAAAACAAATCA
Mouse TNF-α	CCCTCACACTCAGATCATCTTCT	GCTACGACGTGGGCTACAG
Mouse PREF-1	AGTGCGAAACCTGGGTGTC	GCCTCCTTGTTGAAAGTGGTCA
Mouse S1P	AGTCATCACGTCCCCTGAAAA	GCCGACTTGAATGTCTTCCTG
Mouse Cyclin D1	GCGTACCCTGACACCAATCTC	ACTTGAAGTAAGATACGGAGGGC
Mouse SREBP-1c	GGAACTTTTCCTTAACGTGGGC	AGCATGTCTTCGATGTCGTTCA
Mouse Furin	CCTGCTAGGTCGGGATGATTC	AGGGACGTGTATCAGGAGCC

Liver Immunostaining

Livers were collected, fixed in paraformaldehyde, and embedded in paraffin blocks. Sections were stained with hematoxylin and eosin and examined by light microscopy (magnification ×20 and ×63).

Western Blotting

Protein from the frozen liver and epididymal fat were extracted in ice-cold Tris buffer enriched with Roche cOmplete cocktail inhibitor with a Geno/Grinder 2010 (SPEX SamplePrep, Metuchen, NJ) and were submitted to SDS-PAGE on precast polyacrylamide gel (Mini-PROTEAN TGX, 4%–20%, Bio-Rad Laboratories, Hercules, CA). Proteins were transferred onto a polyvinylidene difluoride membrane, which was blocked in 5% nonfat dried milk in Tris-buffered saline with 0.1% Tween 20 (TBST). Membranes were then incubated with anti-PRR antibody (No. HPA003156, Sigma, St Louis, MO), anti-LDLR (No. ab30532 Abcam, Cambridge, MA), anti-peroxisome proliferator-activated receptor gamma (PPARγ) (No. 2435, Cell Signaling Technology, Inc., Danvers, MA), anti-Perilipin (No. 9349, Cell Signaling Technology, Inc.), anti-FABP4 (No. 2120, Cell Signaling Technology, Inc.), anti-CEBPα (No. 8178, Cell Signaling Technology, Inc.), SORT1 (No. 20681, Cell Signaling Technology, Inc.), or anti-GAPDH (No. 5174, Cell Signaling Technology, Inc.) in TBST 5% nonfat dried milk. Following incubation with horseradish peroxidase (HRP)-conjugated anti-rabbit secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, PA), proteins were imaged using a Syngene PXi imager (Syngene, Frederick, MD). The levels of proteins were quantified using ImageJ software (National Institutes of Health, Bethesda, MD) and normalized to GAPDH levels. For the quantification of lipoprotein in fast-performance liquid chromatography (FPLC) fractions, 10 μL of pooled fractions were separated on a 4%–20% SDS-PAGE gel, transferred to PVDF membranes, and incubated overnight with anti-ApoA1 (No. K23001R, Meridian Life Science, Inc., Memphis, TN) or ApoB48/100 (No. K23300R, Meridian Life Science, Inc.) antibodies. After incubation with HRP-conjugated anti-rabbit secondary antibody (No. A6154, Sigma, St Louis, MO), proteins were revealed on an X-ray film and quantified as described above.

In Vitro Experiments

HepG2 cells (ATCC) were cultured in the presence of Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) of a mixture of penicillin and streptomycin. PRR (siPRR) was silenced by mouse Stealth RNA interference designed for Atp6ap2 gene (PRR) using Lipofectamine 2000 (Thermo Fisher Scientific), and according to the procedure recommended by the manufacturer. HepG2 cells (ATCC) were treated with mouse recombinant sPRR-His-Tag (100 ng/mL, residues 18–276, Genscript) or with vehicle for 24 h. 3T3-L1 cells were cultured in the presence of DMEM supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) of a mixture of penicillin and streptomycin and treated with mouse recombinant sPRR-His-Tag (10–200 ng/mL) or with the vehicle for 24 h. Primary preadipocytes from adipose PRR KO mice and control mice were isolated by collagenase digestion and treated with sPRR (100 ng/mL) and vehicle.

Statistical Analysis

Data are represented as means ± SE. Statistical differences between groups were analyzed by one-way ANOVA followed by Holm–Sidak post hoc analysis for multiple comparisons. For ANOVA models, normality was assessed using the Shapiro–Wilk test, and when P < 0.05, response variables were log-transformed. Statistical differences with control were analyzed by t tests for in vitro studies. Statistical outliers were identified using the Grubbs test (GraphPad QuickCalcs).

RESULTS

Deletion of PRR in Liver Induced Hepatomegaly and Disturbed Lipid Homeostasis

Male PRR^flox/flox mice treated with AAV-TBG-Cre (liver PRR KO) exhibited significantly reduced liver PRR mRNA (Fig. 1A) and protein levels (Fig. 1B) compared with mice treated with the control vehicle (CTL). The body weight of liver PRR KO mice was not significantly different from that of CTL mice (Fig. 1C). The deletion of PRR in the liver did not significantly change the total fat and lean masses (expressed as a percent of the body weight, Fig. 1D). Moreover, the weights of the retroperitoneal fat, epididymal fat, and subcutaneous fat were not significantly different in liver PRR KO mice compared with in CTL mice (Table 2). Interestingly, liver weights increased significantly in liver PRR KO compared with in CTL mice (Table 2).

Figure 1.

Deletion of hepatic PRR did not change the body weight and the fat mass. A: hepatic mRNA expression of PRR in control (CTL) and liver PRR KO male mice (KO). B: hepatic PRR protein levels (35 KDa) of control (CTL) and liver PRR KO male mice (KO). C: body weight of PRR in control (CTL) and liver PRR KO male mice (KO). D: body composition of PRR in control (CTL) and liver PRR KO male mice (KO) assessed by EchoMRI (CTL: n = 6 mice and KO: n = 5 mice). Data are means ± SE. *P < 0.05 compared with the vehicle. CTL, control; KO, knockout; PRR, prorenin receptor.

Table 2.

Organ weights of liver PRR KO and CTL male mice fed a standard diet

Weight, g	CTL	Liver PRR KO
Retroperitoneal fat	0.436 ± 0.137	0.408 ± 0.100
Epididymal fat	0.966 ± 0.116	0.767 ± 0.098
Subcutaneous fat	0.354 ± 0.051	0.343 ± 0.076
Liver	1.822 ± 0.120	2.369 ± 0.084*

Value are means ± SE. CTL, n = 6 mice and KO, n = 5 mice. *P < 0.05 compared with CTL. CTL, control; KO, knockout.

Additionally, the deletion of hepatic PRR increased the levels of plasma total and free cholesterol (Fig. 2, A and B) and LDL cholesterol (Fig. 2, D and E) and did not change the plasma triglyceride levels (Fig. 2C). Hepatic LDLR protein levels (nonglycosylated form, 95 KDa) and SORT1 protein levels decreased significantly in liver PRR KO mice compared with in CTL mice (Fig. 3, A and B).

Figure 2.

Deletion of hepatic PRR increased plasma cholesterol. A: plasma total cholesterol concentration in control (CTL) and liver PRR KO male mice (KO). B: plasma-free cholesterol concentration PRR in control (CTL) and liver PRR KO male mice (KO). C: plasma triglyceride levels in control (CTL) and liver PRR KO male mice (KO). (CTL: n = 5 mice and KO: n = 5 mice). Data are means ± SE. *P < 0.05 compared with the vehicle. D: plasma cholesterol distribution determined by FPLC in two individual control (CTL) and two liver PRR KO male mice (KO). E: Western blot analysis of lipoprotein fractions isolated by FPLC from two pooled plasma of control (CTL) and liver PRR KO male mice (KO) (CTL: n = 2 mice and KO: n = 2 mice). FPLC, fast-performance liquid chromatography; KO, knockout; PRR, prorenin receptor.

Figure 3.

Deletion of hepatic PRR decreased hepatic triglyceride contents but promoted inflammation. A: representative Western blot of hepatic LDLR, PPARγ, and SORT1 proteins in control (CTL) and liver PRR KO male mice (KO). B: hepatic LDLR (95 kDa), PPARγ (glycosylated form, 75 kDa), SORT1 (100 kDa), and GAPDH (36 kDa) protein quantification (arbitrary units) in control (CTL) and liver PRR KO male mice (KO). C: hepatic triglyceride contents in control (CTL) and liver PRR KO male mice (KO). D: hepatic mRNA expression of genes involved in inflammation in control (CTL) and liver PRR KO male mice (KO). E: histopathologic examination of the liver in control (CTL) and liver PRR KO male mice (KO). Councilman bodies (black arrow heads), intracytoplasmic globules (open arrow heads), and inflammatory cells infiltrates (arrows) are showed on ×63 magnification. F: plasma ALT levels in control (CTL) and liver PRR KO male mice (KO) (CTL: n = 6 mice and KO: n = 5 mice). Data are means ± SE. *P < 0.05 compared with the vehicle. KO, knockout; LDRLR, low-density lipoprotein receptor; PRR, prorenin receptor; SORT1, Sortilin 1.

Because we previously showed that the deletion of PRR in adipose tissue downregulated PPARγ and consequently prevents triglyceride accumulation in the lipid droplets of adipocytes, we aimed to determine whether hepatic PRR deletion influenced PPARγ and triglyceride contents in the liver. Indeed, PPARγ protein levels (glycosylated form, 75 KDa) and hepatic triglyceride levels decreased significantly in the liver of liver PRR KO mice compared with those in the liver of CTL mice (Fig. 3, A–C). Additionally, hepatic proinflammatory effectors TNF-α and MCP-1 mRNA levels and plasma ALT levels increased significantly in liver PRR KO mice compared with in CTL mice (Fig. 3, D and F). Histopathologic examination of the liver revealed distinct hallmarks of hepatocyte injury and regeneration (including numerous councilman bodies, intracytoplasmic globules, and increased mitoses) in the liver of PRR KO mice compared with the liver of mice injected with saline (Fig. 3E). Moreover, NAS analysis confirmed increased lobular inflammation that is characterized by the presence of numerous neutrophilic microabscesses in liver PRR KO mice compared with CTL mice (CTL: score = 0.8; liver PRR KO: score = 2.5, P < 0.05). Although the liver PRR KO achieved only a 50% liver-specific PRR knockdown, the knockdown was sufficient to induce a profound phenotype. Because AAV-TBG Cre is well recognized to be a specific and efficient tool to target gene in hepatocytes (23, 24), one reason could be attributed to the fact that PRR mRNA and proteins levels were assessed in the total liver, which also contains other types of cells (endothelial cells, stellate cells, Kupffer cells, and lymphocytes).

Deletion of Hepatic PRR Elevated Hepatic Cholesterol and Upregulated SREBP2 and HMG CoA-Reductase

The deletion of PRR in the liver induced a significant increase in the total cholesterol content in the liver of liver PRR KO mice compared with in the liver of CTL mice (Fig. 4A, P < 0.05). Furthermore, genes involved in cholesterol synthesis, especially sterol regulatory element-binding protein 2 (SREBP2) and 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG CoA-R), were significantly upregulated in the liver of liver PRR KO mice compared with liver of CTL mice (Fig. 4B, P < 0.05). Together, our data suggest that hepatic PRR is involved in the regulation of endogenous cholesterol synthesis.

Figure 4.

Deletion of hepatic PRR stimulated endogenous cholesterol biosynthesis pathway in the liver. A: hepatic cholesterol contents in control (CTL) and liver PRR KO male mice (KO) (CTL: n = 6 mice and KO: n = 5 mice). B: expression of genes in the liver of control (CTL) and liver PRR KO male mice (KO) (CTL: n = 6 mice and KO: n = 5 mice). C: plasma sPRR levels of PRR in control (CTL) and liver PRR KO male mice (KO). (CTL: n = 5 mice and KO: n = 5 mice). The mRNA levels of PRR (D) and SREBP2 (E) in HepG2 cells transfected with siPRR and treated with vehicle or recombinant mouse sPRR-HisTag (100 ng/mL) for 24 h (total of 5–7 replicates from 2 experiments). Data are means ± SE. *P < 0.05 compared with the vehicle. KO, knockout; PRR, prorenin receptor.

Deletion of Hepatic PRR Increased Circulating sPRR That Could Promote the Upregulation of Hepatic SREBP2

Interestingly, liver PRR KO mice exhibited elevated plasma sPRR levels compared with CTL mice vehicle (Fig. 4C). To determine whether circulating sPRR could contribute to endogenous hepatic cholesterol synthesis, HepG2 cells were transfected with siPRR and treated with or without a mouse recombinant sPRR-His-Tag (Fig. 4, D and E). PRR expression increased significantly in HepG2 cells treated with sPRR compared with that in HepG2 cells treated with vehicle (Fig. 4D). The silencing of PRR in HepG2 cells did not change SREBP2 mRNA levels. However, sPRR treatment induced a significant increase in SREBP2 expression, suggesting that sPRR mediated SREBP2 upregulation independently of PRR expression (Fig. 4E).

PRR and sPRR Programed Adipose Tissue Fate through Upregulation of Genes Involved in Adipogenesis and Fat Mobilization

Because sPRR plasma levels doubled in liver PRR KO mice, we next investigated the origin of this increase. Interestingly, total cleaved sPRR and noncleaved sPRR content levels were significantly higher in the epididymal fat of liver PRR KO mice than in that of CTL mice (Fig. 5A). We next investigated whether the increase in the total cleaved sPRR and noncleaved sPRR contents was mediated by an upregulation of PRR expression and an increase in furin and S1P, the two enzymes involved in PRR cleavage. Interestingly, the deletion of hepatic PRR increased PRR expression (Fig. 5B) and elevated the level of PRR protein in the adipose tissue (Fig. 5C). Furin and S1P mRNA expression increased significantly in the adipose tissue of liver PRR KO compared with that in the adipose tissue of CTL mice, suggesting that the deletion of hepatic PRR induced a compensatory increase in plasma sPRR levels and that the adipose tissue could participate in the increase in plasma sPRR levels. Because we previously demonstrated that deleting PRR in the adipose tissue lowers the expression of the genes involved in adipogenesis, lipid synthesis, and trafficking (16, 25), we assessed the adipogenesis status of the adipose tissue of liver PRR KO mice. In line with the increase in adipose PRR expression, PPARγ and perilipin gene expressions increased significantly in the epididymal fat of liver PRR KO mice compared with those in the epididymal fat of CTL mice (Fig. 5B). Additionally, PRR, PPARγ, and perilipin protein levels were higher in the epididymal fat of liver PRR KO mice than in that of CTL mice (Fig. 5C). Interestingly, the genes involved in lipolysis (ATGL) and cell cycling (cyclin D1) also increased, suggesting that the deletion of PRR in the liver activated the overall molecular machinery of adipogenesis. To confirm the relative contribution of sPRR to the regulation of PPARγ expression, 3T3-L1 cells were treated with and without mouse recombinant sPRR-His-Tag. Our results showed a significant increase in PRR and PPARγ gene expression with sPRR treatment (200 ng/mL), suggesting that sPRR might influence adipocyte differentiation by stimulating PPARγ and PRR gene expression (Fig. 6, A and B). To determine whether sPRR-mediated increase in PPARγ expression is dependent on PRR, isolated primary preadipocyte from adipose PRR KO mice were treated with sPRR or vehicle (Fig. 6C). sPRR treatment increased PPARγ expression in primary preadipocytes, but sPRR treatment failed to restore PPARγ expression in PRR KO preadipocytes, suggesting that sPRR-mediated upregulation of PPARγ is PRR-dependent.

Figure 5.

Deletion of hepatic PRR elevated total cleaved sPRR and noncleaved sPRR contents in adipose tissues and stimulated adipogenesis. A: total cleaved sPRR and non-cleaved sPRR contents in the epididymal fat of control (CTL) and liver PRR KO male mice. B: mRNA expression of genes involved in adipogenesis in epididymal fat of control (CTL) and liver PRR KO male mice. C: representative Western blot and protein quantification of PRR (35 kDa), PPARγ (53 kDa), perilipin (62 kDa), FABP4 (15 kDa), CEBPα42 (42 kDa), CEBPα 28 (28 kDa), and GAPDH in the epididymal fat of control (CTL) and liver PRR KO male mice (KO). KO, knockout; PRR, prorenin receptor. *P < 0.05 compared with control.

Figure 6.

sPRR induced an increase in PRR and PPARγ gene expression. The expression of PRR (A) and PPARγ (B) genes in 3T3-L1 cells treated with vehicle or with mouse recombinant sPRR-HisTag (0–200 ng/mL) for 24 h. Data are means ± SE of a total of 12 replicates from 3 experiments. t Test was performed. *P < 0.05 compared with control. C: PPARγ gene expression of primary pre-adipocytes isolated from adipose PRR KO mice and control mice and treated with sPRR (100 ng/mL) and vehicle (3 mice/groups). Results are expressed as the mean of PPARγ gene expression from primary pre-adipocyte that reached confluence and initiated adipocyte differentiation. KO, knockout; PRR, prorenin receptor.

DISCUSSION

Our study demonstrated that sPRR and hepatic PRR contribute to triglyceride and cholesterol homeostasis. To the best of our knowledge, no prior studies examined the role of sPRR in lipid homeostasis. Although the liver PRR KO achieved only a 50% liver-specific PRR knockdown, deleting hepatic PRR induced a significant increase in hepatic cholesterol and an upregulation of SREBP2 and HMG CoA-R but also elevated circulating sPRR. Therefore, we investigated whether the stimulation of cholesterol synthesis was attributed to a lack of hepatic PRR or was due to elevated levels of plasma sPRR. In vitro studies demonstrated that sPRR treatment upregulated SREBP2 gene expression independently of PRR, suggesting that circulating sPRR could contribute to hepatic cholesterol biosynthesis. Moreover, deleting hepatic PRR elevated total cleaved and noncleaved sPRR contents, furin, and S1P protein levels in epididymal fat, indicating that the increase in circulating sPRR likely originated from adipose tissue. Finally, we demonstrated that sPRR could participate in adipocyte differentiation through a PRR/PPARγ-dependent mechanism.

The silencing of PRR in HEK293, A431, and HepG2 cells impaired LDL uptake by decreasing SORT1 protein abundance, a regulator of lipid secretion and clearance, and by reducing LDLR abundance (18). Moreover, the inhibition of hepatic PRR with a N-acetylgalactosamine (P)RR antisense oligonucleotide (ASO-PRR) reduced hepatic LDLR protein levels, inducing an elevation of plasma cholesterol in mice fed a normal diet (20). Additionally, the deletion of hepatic PRR resulted in cholesterol enrichment of LDL particles (20). Similarly, in the present study, the deletion of PRR in liver using AAV-TBG induced hepatomegaly and hypercholesterolemia and decreased LDLR protein levels (18, 20). In contrast, we found that hepatic cholesterol levels increased in liver PRR KO mice, whereas hepatic cholesterol levels were unchanged in ASO-PRR mice fed a normal diet. The discrepancy between the results could be attributed to the difference between circulating sPRR levels in the two mouse models used. Our data indicated that plasma sPRR levels increased in liver PRR KO and could have participated in the elevation of hepatic cholesterol levels via stimulation of endogenous cholesterol synthesis. However, circulating sPRR was not quantified in ASO-PRR mice, and therefore, the relative contribution of sPRR to the phenotype was not investigated. In fact, the use of antisense oligonucleotides could have affected nontargeted RNAs or influenced the expression of PRR and circulating sPRR in ASO-PRR mice (26, 27). We previously showed that the deletion of adipocyte PRR reduced the adipose tissue mass through a downregulation of PPARγ gene expression and reduced the genes involved in lipid transport and synthesis (16, 25). In ASO-PRR mice, PRR protein tended to decrease in the adipose tissue, and the fat mass was reduced in high fat (HF)-fed ASO-PRR mice, suggesting that adipogenesis and adipose sPRR levels might have been altered (20).

SREBP-2 is an essential transcription factor for de novo cholesterol synthesis, notably through its positive regulation of HMG-CoA reductase gene expression (28–30). In our study, elevated hepatic cholesterol is associated with an upregulation of SREBP-2 and HMG-CoA reductase gene expression, suggesting a stimulation of cholesterol synthesis. Because circulating sPRR was elevated in liver PRR KO, we investigated the relative contribution of PRR and sPRR to SREBP-2 regulation. sPRR treatment elevated SREBP-2 gene expression, whereas PRR silencing had no effect on SREBP-2 transcripts, suggesting that sPRR participated in SREBP-2 regulation independently of PRR. Therefore, sPRR might be a new potential contributor to hepatic lipid metabolism by initiating endogenous cholesterol synthesis via SREBP-2.

The deletion of PRR in liver increased plasma cholesterol, hepatic cholesterol synthesis, and liver weights. A similar phenotype was characterized in humans with the Niemann–Pick type C (NPC) disease and in NPC KO mice. NPC is a protein involved in the movement of unesterified cholesterol from the lysosome to the cytosol (31). In NPC KO mice, Beltroy et al. (31) suggested that the increase in hepatic cholesterol synthesis was due to the inability of the sterol trapped in the lysosomal compartment to become a part of the cholesterol metabolism in the hepatocyte. Consequently, the hepatocyte sensed a shortage in cholesterol and, to compensate, increased cholesterol synthesis. Because we showed that sPRR increased hepatic cholesterol synthesis and that hepatic SORT1 protein levels decreased in liver PRR KO, the increase in plasma cholesterol could have resulted from an increase in hepatic cholesterol synthesis (that involved sPRR) combined with a decrease in cellular uptake (mediated by a PRR/LDLR/SORT1-dependent mechanism). In addition, the liver size of NPC KO mice correlated positively with an increase in hepatic cholesterol levels, suggesting that the increase in the liver weights of liver PRR KO resulted from the increase in cholesterol contents.

We previously demonstrated that the increased plasma sPRR in standard diet (SD)- or HF-fed adipose-PRR KO mice originated from the liver (16, 25). In the present study, we demonstrated that the deletion of PRR in liver increased PRR, furin, S1P, and total cleaved and noncleaved sPRR levels in the adipose tissue, indicating that the increased plasma sPRR levels likely originated from the adipose tissue in liver PRR KO mice. Our studies suggest that a compensatory mechanism occurs between the liver and the adipose tissue to counteract the lack of a functional tissue PRR. Previous studies suggested that cholesterol stimulated SREBP-1c to upregulate PPRAγ expression in the adipose tissue and consequently increased free cholesterol storage in lipid droplets, lipogenesis, and adipocyte differentiation (32). Because Wang et al. (33) showed that PRR is a direct target gene of PPARγ and because the deletion of PRR in the liver increased circulating cholesterol and adipose SREBP-1c levels, the elevated circulating cholesterol levels in liver PRR KO could have stimulated adipose SREBP1c and PPARγ and consequently PRR expression. In turn, because we demonstrated previously that PRR controlled PPARγ expression as well, the adipose PRR could have upregulated PPARγ. Hence, a liver-adipose tissue cross talk is essential for PRR regulation. In the present study, we showed that sPRR upregulated PRR gene expression in HepG2 cells. This finding is consistent with our recent report showing an increase in hepatic PRR gene expression in female mice infused with sPRR (25). Interestingly, sPRR treatment cannot rescue the expression of its own gene in siPRR-treated cells, suggesting that sPRR needs the full-length PRR to exert its action.

PRR is expressed abundantly in the adipose tissue, and accumulating evidence demonstrates that PRR is a master regulator of adipogenesis. In addition, the development of obesity induces an increase in PRR gene expression in the adipose tissue of mice, rats, and humans (16, 34, 35). Treatment with the handle region peptide, a PRR blockade, reduced fat mass and adipocyte size, as well as leptin and inflammatory cytokines levels (36). Furthermore, the silencing of PRR in 3T3-L1 cells reduced PPARγ and FABP4 gene expression, indicating that PRR is an important regulator of PPARγ (16). In line with previous works, the present study indicated that the upregulation of PRR in adipose tissue was associated with an upregulation of PPARγ, FABP4, and perilipin. Additionally, sPRR promotes PRR and PPARγ gene expression in 3T3-L1 cells. Thus, sPRR could stimulate adipogenesis through a PRR-induced PPARγ-dependent mechanism. We acknowledge that this stimulation was insufficient to change the fat mass in the mice fed a standard diet but instead activated the molecular machinery of adipogenesis, and a second hit (such as a high-fat diet) may be required to trigger this effect.

Similar to the study of Ren et al. (20), hepatic triglyceride levels and PPARγ expression were reduced in liver PRR KO mice. A previous report (37) demonstrated that PPARγ knockdown in the hepatocyte was associated with a decrease in hepatic triglyceride levels. Additionally, in AZIP liver PPARγ KO mice, the reduction in liver triglyceride levels was attributed to an increase in triglyceride clearance (38). Because we showed that plasma triglyceride levels remained unchanged in liver PRR KO mice and because Ren et.al. (20) showed reduced hepatic triglyceride levels in high fat-fed ASO PRR KO, one could speculate that the downregulation of hepatic PPARγ influenced not only the liver triglyceride clearance but also the liver triglyceride synthesis (39).

In conclusion, the remarkable phenotype of the liver PRR KO mouse model demonstrated the importance of the liver PRR in cholesterol and triglyceride homeostasis, and in liver inflammation and injury. In addition, PRR and sPRR contributed to lipid homeostasis by stimulating regulatory pathways involved in LDLR clearance, cholesterol synthesis, and adipocyte differentiation. Further investigation is needed to determine the mechanism by which sPRR participated in lipid homeostasis.

GRANTS

This work was supported by NIH Grant R01-HL-142969, the American Heart Association Grant 13SDG17230008, the National Institute of General Medical Sciences Grant P30 GM127211, and the University of Kentucky, Center for Clinical and Translational Sciences Grant UL1TR001998.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

E.G. and F.B.Y. conceived and designed research; E.G., G.A., N.R.S., G.A.G., and R.E.T. performed experiments; E.G., G.A.G., and R.E.T. analyzed data; E.G., B.T.S., N.R.S., G.A.G., R.E.T., and F.B.Y. interpreted results of experiments; E.G. and F.B.Y. prepared figures; E.G. and F.B.Y. drafted manuscript; E.G., G.A., A.P., K.N., B.T.S., N.R.S., G.A.G., R.E.T., and F.B.Y. edited and revised manuscript; E.G., N.R.S., R.E.T., and F.B.Y. approved final version of manuscript.