Hypercholesterolemia in rats with hepatomas: Increased oxysterols accelerate efflux but do not inhibit biosynthesis of cholesterol

Authors

  • Takeshi Hirayama,

    1. Division of Gastroenterology and Hepatology, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki, Japan
    Search for more papers by this author
    • T. Hirayama and A. Honda contributed equally to this work.

  • Akira Honda,

    1. Division of Gastroenterology and Hepatology, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki, Japan
    2. Ibaraki Prefectural Institute of Public Health, Mito, Ibaraki, Japan
    Search for more papers by this author
    • T. Hirayama and A. Honda contributed equally to this work.

  • Yasushi Matsuzaki,

    Corresponding author
    1. Division of Gastroenterology and Hepatology, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki, Japan
    2. Division of Gastroenterology and Hepatology, Tokyo Medical University, Kasumigaura Hospital, Ami, Ibaraki, Japan
    • Division of Gastroenterology and Hepatology, Tokyo Medical University, Kasumigaura Hospital, 3-20-1 Ami-Machi-Chuoh, Inashikigun, Ibaraki 300-0395, Japan
    Search for more papers by this author
    • fax: (81) 29-888-3463

  • Teruo Miyazaki,

    1. Ibaraki Prefectural Institute of Public Health, Mito, Ibaraki, Japan
    Search for more papers by this author
  • Tadashi Ikegami,

    1. Division of Gastroenterology and Hepatology, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki, Japan
    2. Division of Gastroenterology and Hepatology, Tokyo Medical University, Kasumigaura Hospital, Ami, Ibaraki, Japan
    Search for more papers by this author
  • Mikio Doy,

    1. Ibaraki Prefectural Institute of Public Health, Mito, Ibaraki, Japan
    Search for more papers by this author
  • Guorong Xu,

    1. Department of Medicine, University of Medicine and Dentistry of New Jersey—New Jersey Medical School, Newark, NJ
    2. Veterans Affairs Medical Center, East Orange, NJ
    Search for more papers by this author
  • Michael Lea,

    1. Department of Medicine, University of Medicine and Dentistry of New Jersey—New Jersey Medical School, Newark, NJ
    Search for more papers by this author
  • Gerald Salen

    1. Department of Medicine, University of Medicine and Dentistry of New Jersey—New Jersey Medical School, Newark, NJ
    2. Veterans Affairs Medical Center, East Orange, NJ
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Abstract

Hypercholesterolemia is an important paraneoplastic syndrome in patients with hepatoma, but the nature of this defect has not yet been identified. We investigated the molecular mechanisms of hypercholesterolemia in a hepatoma-bearing rat model. Buffalo rats were implanted in both flanks with Morris hepatoma 7777 (McA-RH7777) cells. After 4 weeks, tumor weight was 5.5 ± 1.7 g, and serum cholesterol level increased from 60 ± 2 to 90 ± 2 mg/dL. Protein and mRNA expression of the ATP-binding cassette transporters A1 and G1 (ABCA1 and ABCG1) was markedly higher in tumors than in livers. These increases were associated with activation of liver X receptor α (LXRα) as a result of the increased tissue oxysterol concentrations. The accumulation of oxysterols in the hepatomas appeared to be caused mainly by the upregulation of cholesterol biosynthesis, despite the increased tissue sterol concentrations. Overexpression of the sterol regulatory element-binding protein (SREBP) processing system relative to sterol concentration contributed to the resistance to sterols in this tumor. In addition, bile acid biosynthesis was inhibited despite the reduced expression of the small heterodimer partner (SHP) and activated LXRα, which also appeared to contribute to the accumulation of oxysterols followed by the acceleration of cholesterol efflux. In conclusion, hypercholesterolemia in McA-RH7777 hepatoma-bearing rats was caused by increased cholesterol efflux from tumors as a result of activation of LXRα. Overexpression of the SREBP processing system contributed to the activation of LXRα by maintaining high oxysterol levels in tissue. (HEPATOLOGY 2006.)

Hypercholesterolemia is one of the well-known paraneoplastic manifestations that may occur in patients with hepatomas.1, 2 It presents in 10%–25% of patients with hepatomas2, 3 and is an unfavorable prognostic factor for these patients.4 Hypercholesterolemia associated with hepatoma has also been reported in rats implanted with some types of Morris hepatomas.5–7 Although the exact mechanism remains to be identified, hypercholesterolemia in rats has been reported to be a result of the increased release of cholesterol from the tumor into the blood.6, 8

Many years ago, Siperstein et al. reported that cholesterol feedback control, which is characteristic of normal liver function, was consistently lost in all hepatomas regardless of animal species, degree of tumor differentiation, or method of induction of the hepatoma.9 However, further studies demonstrated that dietary cholesterol slightly but significantly inhibited the activity of HMG-CoA reductase, the rate-limiting enzyme in the cholesterol biosynthetic pathway, in rat hepatomas.10 In addition, the injection of mevalonolactone, a precursor of cholesterol, into rats bearing hepatomas suppressed cholesterol biosynthesis in the tumors.11 Therefore, defective cholesterol feedback regulation in hepatomas was considered to result from the secondary effects of altered cholesterol metabolism rather than from an inherent defect in the cholesterol biosynthetic pathway itself. In fact, studies using other murine hepatomas have suggested that reduced lipoprotein receptor number12, 13 and limited lipoprotein access to the hepatoma tissue14 contribute to the desensitized feedback regulation. However, it has not been determined whether the acceleration of cholesterol efflux from the hepatoma is the cause of the desensitized cholesterol feedback control or a consequence of the defective feedback regulation.

The recent explosion of research into nuclear factors including nuclear receptors has provided new insights into the regulation of cholesterol metabolism. De novo cholesterol synthesis is regulated by sterol regulatory element-binding proteins (SREBPs) that are synthesized in the endoplasmic reticulum and are released to the nucleus by sterol-sensitive proteolysis.15 Efflux and transport of cholesterol from cells to extracellular acceptors or matrices are controlled by the liver X receptor α (LXRα, NR1H3), which is an oxysterol receptor,16 and cholesterol catabolism (bile acid synthesis) is mainly regulated by the farnesoid X receptor (FXR, NR1H4), a bile acid–activated receptor.16

The current study was undertaken to explore the molecular mechanisms that elicit hypercholesterolemia in hepatoma-bearing animals. The Morris hepatoma 7777 (McA-RH7777) is an extensively studied rat liver tumor that is poorly differentiated and grows rapidly when implanted in the flanks of Buffalo rats and is one of the hepatomas that can cause paraneoplastic hypercholesterolemia in host animals.5, 8 Using this model, we demonstrated that cholesterol efflux from the tumor was accelerated by the upregulation of the ATP-binding cassette transporters A1 and G1 (ABCA1 and ABCG1). In addition, this upregulation was a consequence of overexpression of the SREBP processing system in McA-RH7777 hepatomas.

Abbreviations

SREBP, sterol regulatory element binding protein; LXRα, liver X receptor α; ABC, ATP-binding cassette transporter; CDCA, chenodeoxycholic acid; CA, cholic acid; HMGCR, HMG-CoA reductase; SCAP, SREBP cleavage-activating protein; INSIG, insulin-induced gene; SHP, small heterodimer partner; HNF, hepatocyte nuclear factor.

Materials and Methods

Cell Culture.

The rat hepatoma cell line, McA-RH7777 was obtained from American Type Culture Collection (Rockville, MD). Stock cultures were grown and maintained in DMEM (low glucose) supplemented with 10% (v/v) fetal bovine serum. The cultures were incubated at 37°C in a humidified incubator containing 5% CO2 and 95% air.

Animal Experiment.

Eleven male Sprague-Dawley rats of the Buffalo strain (mean ± SEM [range] of 144 ± 7 g [103–169 g]) were purchased from Charles River Japan Inc. (Yokohama, Japan), and were kept under regular 12-hour light-dark cycles (7 AM–7 PM) with free access to standard chow and water. Six rats were injected with 2 × 106 of McA-RH7777 cells in each flank site, and the other five rats served as controls. After 2 weeks two hepatoma-implanted rats and after 4 weeks the remaining four hepatoma-implanted rats and five control rats were sacrificed between 1 PMand 2 PM under diethyl ether anesthesia. The tumors were carefully dissected and weighed, and blood and liver were collected. Tumor and liver specimens were immediately frozen in liquid N2 and stored at −70°C until later use. The animal protocol was approved by the Institutional Animal Care and Use Committee at Tsukuba University, Tsukuba, Japan.

Serum Lipid Concentrations.

Serum total cholesterol and triacylglycerol concentrations were measured by enzymatic methods using the Cholesterol E-Test Wako and Triglyceride E-Test Wako, respectively (Wako Pure Chemical Industries, Osaka, Japan)

Tissue Lipid Concentrations.

Concentrations of cholesterol17 and oxysterols18 in the tumors and livers were determined using our previously described methods of gas chromatography–mass spectrometry with selected-ion monitoring (GC-SIM). Triacylglycerol concentrations were measured according to the method of Repa et al.19 except that the Triglyceride E-Test Wako was used for the colorimetric assay.

Tissue Bile Acid Concentrations.

Bile acid concentrations and profiles were determined in tumors and livers according to our previously described method20 with some modifications. In brief, deuterium-labeled bile acids were added as internal recovery standards to the whole homogenate and incubated in 5% KOH at 80°C for 20 minutes. Bile acids were extracted by a Bond Elut C18 cartridge and subjected to enzymatic hydrolysis with choloylglycine hydrolase. The resulting deconjugated bile acids were converted into the ethyl ester dimethylethylsilyl (DMES) ether derivatives and quantified by high-resolution GC-SIM. The column oven was programmed to change from 100°C to 280°C at 30°C/min after a 1-minute delay from the start time, and the multiple ion detector was focused on m/z 459.3294 for chenodeoxycholic acid (CDCA), m/z 461.3451 for lithocholoc acid (LCA), m/z 486.3529 for α-muricholic acid (α-MCA), m/z 561.3795 for β-muricholic acid (β-MCA), m/z 563.3952 for deoxycholic acid (DCA) and ursodeoxycholic acid (UDCA), m/z 665.4453 for cholic acid (CA), m/z 463.3546 for [2H4]CDCA, m/z 465.3702 for [2H4]LCA, m/z 567.4203 for [2H4]DCA and [2H4]UDCA, and m/z 669.4705 for [2H4] CA. [2H4]CA was used as an internal recovery standard for α-MCA and β-MCA.

Enzyme Assays.

Microsomes and mitochondria were prepared from tumors and livers by differential ultracentrifugation,17 and the protein concentrations were determined by the method of Bradford.21 Microsomal HMG-CoA reductase (HMGCR) activity was measured using a previously described method22 with some modifications. Briefly, 0.1 mg of microsomal protein was incubated for 30 min at 37°C in 100 mmol/L potassium phosphate buffer (pH 7.4) containing 0.1 mmol/L EDTA, 50 mmol/L KCl, 10 mmol/L DTT, a NADPH generating system, and 30 nmol [14C]HMG-CoA (final volume 150 μL). The reaction was terminated with the addition of 20 μL of 6N HCl and [3H]mevalonolactone (40,000 dpm). After lactonization, the products were separated by thin-layer chromatography, and mevalonolactone was measured by dual-label liquid scintillation counting. The activities of microsomal cholesterol 7α-hydroxylase (CYP7A1) and mitochondrial cholesterol 27-hydroxylase (CYP27A1) were measured using our previously described methods.23 Microsomal oxysterol 7α-hydroxylase (CYP7B1) activity was determined by stable-isotope dilution mass spectrometry using [2H7]7α,27-dihydroxycholesterol as an internal recovery standard. The reaction mixture (final volume 250 μL) consisted of 100 mmol/L potassium phosphate buffer (pH 7.4) containing 0.1 mmol/L EDTA, 5 mmol/L DTT, 30 μmol/L 27-hydroxycholesterol (dissolved in 6 μL of ethanol), and 0.1 mg of microsomal protein. Incubation was initiated by the addition of NADPH (final concentration 1.2 mmol/L) and continued for 15 min at 37°C. The reaction was terminated with 2.5 mL of 2:1 (v/v) chloroform–methanol and 100 ng of [2H7]7α,27-dihydroxycholesterol. Extraction, purification, and quantification of 7α,27-dihydroxycholesterol were performed as described previously.17

RNA Measurements.

Total RNA was extracted from frozen tissue by a MagNA Pure LC RNA Isolation Kit III (Roche Diagnostics, Mannheim, Germany). Reverse transcription was performed on 1 μg of total RNA using a 1st Strand cDNA Synthesis Kit for RT-PCR (Roche). The real-time quantitative PCR assay was performed in triplicate using aliquots of the cDNA (originating from 20 ng each of total RNA) with the FastStart DNA Master SYBR Green I and a LightCycler (Roche). The sequences of the oligonucleotide primer pairs used to amplify the RNA are listed in Table 1. PCR amplification began with a 10-minute preincubation step at 95°C, followed by 40 cycles of denaturation at 95°C for 10 seconds, annealing at 62°C for 10 seconds, and elongation at 72°C for 10 seconds. The relative concentration of the PCR product derived from the target gene was calculated using LightCycler System software. A standard curve for each run was constructed by plotting the crossover point against the log concentration. The concentration of target molecules in each sample was then calculated automatically by reference to this curve (r = −1.00), and the results were standardized to total RNA concentration. Amplification products were checked by electrophoresis on 3% agarose gels, and the specificity of each PCR product was assessed by melting curve analysis.

Table 1. Primer Sequences Used in mRNA Quantification by Real-Time Reverse-Transcription PCR
mRNAGenbank Accession No.ForwardReverseAmplicon length (bp)
  1. Real-time quantitative PCR was performed as described in Materials and Methods section.

LXRαNM_0316275′-TGCTCTCCGAGATCTGGG-3′5′-TCCTCAGTCTGCTCCACC-3′110
ABCA1NM_1780955′-AGAGCTAGGTCTCCCTT-3′5′-CACTGCCCCTGTAATGG-3′151
ABCG1NM_0535025′-CAACACAATGCAGACAGT-3′5′-CAGGGTGGATAGAGGATTT-3′157
ABCG5NM_0537545′-TGTGACCCTGGCATCTAT-3′5′-ATCATTGGACCAGTTCAGT-3′116
ABCG8NM_1304145′-GCTTCCTGTCCCTGTATTA-3′5′-GGGATCATTGTCCTCAGT-3′147
SR-BINM_0315415′-CCCAGAAGACACCACGA-3′5′-GTGTGGACAGTGTGACATCT-3′144
SREBP1XM_2133295′-CCATTGACAAGGCCATGC-3′5′-GGTCATGTTGGAAACCACGC-3′162
FASNM_0173325′-TGGTGGCTGCTACAAACA-3′5′-GGAGAAACAACCAGTGTGACAA-3′167
SREBP2XM_2169895′-CATCTTCCCCTCTCCTTCCTAT-3′5′-CCCAGCTTGACAATCATCTGC-3′131
SCAPXM_2172795′-GGAGCTGTGGAGGAAATTGT-3′5′-CCCGTGGATTCAGGTGTAGT-3′128
INSIG-1NM_0223925′-GTTCCAGAAAAGCCTCACAG-3′5′-ACGTGAACGCATCTTTAGGA-3′159
INSIG-2NM_1780915′-TTTCTCTATGTCCGCTCTTG-3′5′-GACTTCTGCCCTCTTCATTC-3′140
HMGCRNM_0131345′-ACGAATTGTGTGTGGCACT-3′5′-CCTTCTTGGTGCATGTTCCT-3′143
LDLRNM_1757625′-TCTTCCTCCCCATTGCAC-3′5′-CGAGGGGTAGGTATAGCC-3′173
FXRNM_0217455′-GAATGACCACAAGTTCACC-3′5′-AAGAAGGGAAGTCCAATACC-3′120
SHPD865805′-TGAAAGGCCGCACTCAGC-3′5′-CCCAGGTAAGGGAAGGCAT-3′155
FTFNM_0217425′-GCAAACAGGCAGAAGACT-3′5′-GTCTAGGGGTGGGAACTT-3′114
HNF4αNM_0221805′-CCTCACCTGATGCAAGAAC-3′5′-TGATGGCTGTGGAGTCTC-3′138
HNF6NM_0226715′-AGGTGGCTCAGCGTATCA-3′5′-CCCGTGTTCTTGCTCTTTC-3′248
CYP7A1NM_0129425′-TCTACCCAGACCCTTTGACT-3′5′-CTTGATTTCTTGGACGGCA-3′167
CYP7B1XM_3422185′-TAGGACTAAACCACAGTCGC-3′5′-TGCAGCCTTATTCCGCTA-3′128
CYP8B1NM_0312415′-GGTACGCTTTCGCTACCG-3′5′-GGTAGGGGTGCTTTGGAG-3′111
CYP27A1NM_1788475′-TGAAGTCTGTGTCCCGCAT-3′5′-TGTGCCAGGGCTTCTGACAA-3′121
BSEPNM_0317605′-CCCAAGGCTACTGGTAT-3′5′-GCCTTCGGTACAATCAAC-3′191
NTCPNM_0170475′-CAGAAACTCAAGATCAACCT-3′5′-CTGCTTACTGGAAAGTTTGGAA-3′153
β-ActinNM_0311445′-CCTGTATGCCTCTGGTCGTA-3′5′-CCATCTCTTGCTCGAAGTCT-3′260

SDS-PAGE and Immunoblot Analysis.

The microsomal fraction (including the membrane fraction) was resolved by SDS-PAGE on a 5%–20% gradient gel (e-PAGEL, ATTO Corporation, Tokyo, Japan) and transferred to a poly(vinylidene difluoride) (PVDF) membrane (Immobilon-P, Millipore, Bedford, MA). Immunoblot analyses of rat ABCA1, ABCG1, SREBP cleavage-activating protein (SCAP), insulin-induced genes 1 and 2 (INSIG-1 and INSIG-2), and β-actin were conducted with rabbit polyclonal antibody against human ABCA1, goat polyclonal antibodies against human ABCG1, human SCAP, human INSIG-1 and rat INSIG-2 (Santa Cruz Biotechnology, Santa Cruz, CA), and monoclonal anti-β-actin antibody (Sigma, St. Louis, MO), respectively. The membrane was blocked for 1 hour in 5% fat-free milk in TBS-T (Tris-buffered saline/0.1% Tween-20) and incubated with the primary antibody against either ABCA1 (1:100 dilution), ABCG1 (1:200 dilution), SCAP (1:200 dilution), INSIG-1 (1:200 dilution), INSIG-2 (1:200 dilution), or β-actin (1:1000 dilution) in 5% fat-free milk in TBS-T overnight at 4°C. The blot was washed three times for 10 min in TBS-T and incubated with an HRP-conjugated donkey antigoat IgG antibody (Santa Cruz) for ABCG1, SCAP, INSIG-1, INSIG-2, an HRP-conjugated donkey anti-rabbit IgG antibody (Amersham, Buckinghamshire, UK) for ABCA1, or with an HRP-conjugated sheep antimouse IgG antibody (Amersham) for β-actin. After washing, the bands were visualized by exposure to film (Hyperfilm ECL, Amersham) with an ECL Western Blotting Analysis System (Amersham) according to the manufacturer's instructions. The gradient gel was calibrated with prestained molecular-weight markers (Bio-Rad Japan, Tokyo, Japan).

Statistics.

Data are reported as means ± SEMs. The statistical significance of differences in the results of the groups was evaluated by the Student's two-tailed t test, and a difference of P < .05 was considered significant.

Results

Four weeks after the hepatomas were implanted in the flanks, tumor weight was 5.5 ± 1.7 g (n = 4), and serum cholesterol concentration had increased by 50% compared to the control baseline (Table 2). As shown in Fig. 1, expression of mRNA of ABCA1 and ABCG1, essential transporters for the efflux of cholesterol from tissue to serum, was elevated in tumors 10.3-fold and 7.2-fold, respectively, compared with that in the control livers. These increases in mRNA were associated with increased expression of the corresponding proteins.

Table 2. Differences between Control and Tumor-Bearing Rats
 Control rats at 4 wk (n = 5)Tumor-bearing rats at 2 wk (n = 2)Tumor-bearing rats at 4 wk (n = 4)
  • Means ± SEMs are given.

  • *

    P < .01, significantly different from control rats at 4 wk.

  • P < .0001, significantly different from control rats at 4 wk.

  • P < .01, significantly different from tumor-bearing rats at 2 wk.

Body weight (g)278 ± 8214 ± 26232 ± 9*
Weight gain (g/week)32 ± 336 ± 325 ± 1
Liver weight (g)11.3 ± 0.49.6 ± 0.99.4 ± 0.2*
Tumor weight (g)0.9 ± 0.25.5 ± 1.7
Serum total cholesterol (mg/dL)60 ± 266 ± 790 ± 2
Serum triacylglycerol (mg/dL)79 ± 686 ± 1369 ± 6
Figure 1.

(A) Relative expression of ABCA1, ABCG1 and β-actin mRNA in control livers, host livers, and tumors. Quantitative real-time PCR was performed with the primers described in Table 1. All data were standardized for total RNA. Mean expression in control livers was set to 1.0. Data are presented as means ± SEMs. (B) ABCA1, ABCG1 and β-actin protein in control livers and tumor tissues. Microsomal fractions (50 μg, 20 μg and 10 μg protein per lane for ABCA1, ABCG1 and β-actin, respectively) were subjected to SDS-PAGE analysis.

Figure 2 shows expression of LXRα and its target genes, except for ABCA1 and ABCG1. Although the expression of LXRα mRNA did not change in tumors, other LXRα target genes, that is, ABCG5, ABCG8, scavenger receptor class B type I (SR-BI), and SREBP1, were all markedly upregulated in tumors. In addition, mRNA expression of fatty acid synthase (FAS), a target gene of SREBP1, was also increased in tumors compared with that in control livers.

Figure 2.

Relative expression of mRNA of LXRα and its target genes in control livers, host livers, and tumors. Quantitative real-time PCR was performed with the primers described in Table 1. All data were standardized for total RNA. Mean expression in control livers was set to 1.0. Data are presented as the means ± SEMs.

To demonstrate the abundance of ligands for LXRα in tumors, the concentrations of cholesterol (Table 3) and oxysterols (Fig. 3) in tumor and liver specimens were compared. Total cholesterol concentration was increased 1.9-fold in tumors compared with that in control livers. The increase was derived from the accumulation of unesterified cholesterol rather than from the esterified form. The concentrations of 24S-hydroxycholesterol, 24S,25-epoxycholesterol, 25-hydroxycholesterol, and 27-hydroxycholesterol in tumors were significantly elevated compared with those in livers. In contrast, the concentration of 7α-hydroxycholesterol was significantly lower in tumors than in livers.

Table 3. Hepatic Cholesterol and Triacylglycerol Concentrations in Livers and Tumors
 Control livers (n = 5)Host livers (n = 4)Tumors (n = 4)
  • Means ± SEMs are given.

  • *

    P < .0001, significantly different from control livers.

  • P < .005, significantly different from host livers.

  • P < .005, significantly different from control livers.

  • P < .05, significantly different from host livers.

Cholesterol free (μmol/g liver)4.9 ± 0.25.3 ± 0.38.1 ± 0.4*
 Esterified (μmol/g liver)0.7 ± 0.31.2 ± 0.32.2 ± 0.7
 Total (μmol/g liver)5.5 ± 0.36.5 ± 0.510.3 ± 1.1
 Ester/total (%)11 ± 518 ± 420 ± 6
Triacylglycerol (μmol/g liver)8.2 ± 1.412.2 ± 2.013.3 ± 4.2
Figure 3.

Concentrations of oxysterol in control livers, host livers, and tumors. Data are presented as means ± SEMs (7α-OH, 7α-hydroxycholesterol; 22R-OH, 22R- hydroxycholesterol; 24S-OH, 24S-hydroxycholesterol; 24S, 25-epoxy, 24S,25-epoxycholesterol; 25-OH, 25-hydroxycholesterol; 27-OH, 27-hydroxycholesterol).

Activity and gene expression of key enzymes involved in the biosynthesis and catabolism of cholesterol and oxysterols are summarized in Fig. 4. In tumors, HMGCR activity was markedly upregulated and CYP7A1 activity markedly downregulated compared with those in control livers, which was associated with increased expression of HMGCR mRNA and decreased expression of CYP7A1 mRNA, respectively. In contrast, neither activity nor mRNA expression of CYP27A1 was significantly different between tumors and livers. Interestingly, CYP7B1 activity was not significantly different between tumors and livers, but expression of CYP7B1 mRNA was markedly reduced in tumors compared with in livers.

Figure 4.

Activity and relative expression of mRNA of enzymes involved in cholesterol and bile acid biosynthesis in control livers, host livers, and tumors. Quantitative real-time PCR was performed with the primers described in Table 1. All data were standardized for total RNA. Expression in controls was set to 1.0. Data are presented as means ± SEMs.

Figure 5A compares expression of mRNA of SREBP2 and related proteins between tumors and livers. SREBP2, SCAP, INSIG-1, and INSIG-2 were markedly upregulated in tumors. Expression of the LDL receptor, another target gene for SREBP2, was also upregulated in tumors. As shown in Fig. 5B, increases in the expression of SCAP, INSIG-1, and INSIG-2 mRNA were associated with increased expression of the corresponding proteins.

Figure 5.

(A) Relative expression of mRNA of SREBP2 and related proteins in control livers, host livers, and tumors. Quantitative real-time PCR was performed with the primers described in Table 1. All data were standardized for total RNA. Mean expression in control livers was set to 1.0. Data are presented as means ± SEMs. (B) SCAP, INSIG-1, and INSIG-2 protein in control liver and tumor tissues. Microsomal fraction (20 μg, 50 μg and 50 μg protein per lane for SCAP, INSIG-1 and INSIG-2, respectively) were subjected to SDS-PAGE analysis.

Bile acid composition and concentration in tumors were compared with those in livers (Fig. 6). In tumors, all bile acids except LCA were markedly reduced (total 7.9% of control livers), but the CA:CDCA+MCAs ratio did not change significantly between tumors and control livers (2.0 ± 0.3 vs. 1.4 ± 0.2, NS). Because CDCA is readily converted to α- or β-MCA in rat livers,24 the CA:CDCA+MCAs ratio in rats can relate to the CA:CDCA ratio in humans. However, this tumor does not appear to easily convert CDCA into MCAs because the MCAs:CDCA ratio in tumors was markedly low compared with that in control livers (0.4 ± 0.1 vs. 7.4 ± 1.1, P < .005).

Figure 6.

Bile acid concentrations in control livers, host livers, and tumors. Data are presented as means ± SEMs.

Expression of other genes involved in bile acid metabolism is depicted in Fig. 7. FXR and its small heterodimer partner (SHP, NR0B2) were significantly downregulated in tumors. In contrast, α-fetoprotein transcription factor (FTF, NR5A2) and hepatocyte nuclear factor 4α (HNF4α, NR2A1) were significantly upregulated in tumors. CYP8B1 and CYP7A1 were markedly downregulated, and expression of the bile salt export pump (BSEP) and the Na+/taurocholate cotransporting polypeptide (NTCP) were also very low in tumors.

Figure 7.

Relative expression of mRNA for FXR and related proteins in control livers, host livers, and tumors. Quantitative real-time PCR was performed with the primers described in Table 1. All data were standardized for total RNA. Mean expression in control livers was set to 1.0. Data are presented as means ± SEMs.

Discussion

Our results clearly demonstrate that expression of both ABCA1 and ABCG1 mRNA and protein was markedly increased in McA-RH7777 hepatomas implanted in the flanks of Buffalo rats. It should be mentioned that in the present investigation mRNA expression was normalized to total RNA concentration instead of to housekeeping gene expression. Previous reports have demonstrated that the use of housekeeping genes as internal standards is inappropriate for comparing mRNA expression in normal and malignant tissues25 and that normalization to total RNA concentration is an acceptable alternative.26 In fact, expression of β-actin mRNA and protein was markedly higher in the hepatoma tissues than in the normal livers (Fig. 1).

Upregulation of the ABCA1 and ABCG1 genes in McA-RH7777 hepatomas appears to explain the mechanism of hypercholesterolemia in the tumor-bearing rats. It is well known that HDL is the major lipoprotein increased in this model.5, 7 ABCA1 mediates the efflux of cellular cholesterol to lipid-poor apolipoprotein AI, which forms a nascent HDL.27 ABCG1 does not mediate cholesterol efflux to lipid-poor apolipoprotein AI but stimulates efflux to both smaller (HDL-3) and larger (HDL-2) subclasses of HDL.28 Thus, upregulation of these transporters accelerates cholesterol efflux and leads to elevation of HDL concentration. SR-BI is another mediator that controls the bidirectional flux of cholesterol between tissues and HDL.27 The direction of net flux changes depending on the cholesterol gradient, so it is not clear whether upregulated SR-BI in this tumor also contributes to cholesterol efflux.

Why are ABCA1 and ABCG1 markedly upregulated in this hepatoma? It is known that these transporter genes are the targets of a nuclear receptor, LXRα.29, 30 Because our results demonstrated that all the other LXRα target genes we measured (ABCG5, ABCG8, SREBP1c,16 and SR-BI31) were coordinatedly upregulated in this tumor, the stimulation of both ABCA1 and ABCG1 can be explained by activated LXRα. It should be noted that although we quantified total SREBP1 instead of SREBP1c, the 1c transcript predominates over the 1a transcript in McA-RH7777 cells.32 The LXRα is activated by several oxysterols including 24S-hydroxycholesterol, 24S,25-epoxycholesterol, 25-hydroxycholesterol, and 27-hydroxycholesterol,33, 34 and their concentrations were all significantly elevated in the tumors. Therefore, the accumulation of tissue oxysterols appears to be an important factor in the acceleration of cholesterol efflux from McA-RH7777 hepatomas.

Oxysterols are formed by hydroxylation of the side chain of cholesterol, and their tissue concentrations change depending on tissue cholesterol levels.35, 36 Under normal conditions, high concentrations of cholesterol in tissue stimulate oxysterol production, and the oxysterols, as well as cholesterol, downregulate cholesterol biosynthesis37 by inhibiting HMGCR, the rate-limiting enzyme in the cholesterol biosynthetic pathway. In McA-RH7777 hepatomas, however, HMGCR was markedly upregulated despite the increased concentrations of both oxysterols and cholesterol.

Oxysterols and cholesterol both inhibit SREBP cleavage by inducing the escort protein, SCAP, to bind to INSIGs.37 If the tumor has any SCAP mutations, binding of SCAP to INSIGs is prevented, so SREBP processing is resistant to inhibition by sterols.38 However, a previous report demonstrated that exogenously added 25-hydroxycholesterol markedly inhibited SREBP processing in McA-RH7777 cells,32 which suggests the functions of SCAP and INSIGs were intact in this tumor. When SCAP and INSIGs are intact, the relative amounts of SCAP and INSIGs determine sensitivity to endogenous sterol levels.37 For example, in cells that express excess SCAP relative to INSIGs, neither cholesterol nor 25-hydroxycholesterol inhibits SREBP processing. In contrast, overexpression of INSIGs inhibits SREBP processing.39 In our results, the expression of SCAP and INSIGs proteins, and SREBP2, SCAP, and INSIGs, mRNA were all equally overexpressed in the tumors. Therefore, the resistance to sterols is not explained by a single defect in the SREBP processing system but appears to be caused by an excess of the SREBP processing system relative to sterols in the tumors (Fig. 8). When more of the SREBP processing system was expressed, more sterols were needed to suppress SREBP cleavage, so that steady-state cholesterol concentration under feedback regulation had to be shifted higher. In addition, because higher oxysterol levels accelerate cholesterol efflux through LXRα activation, cholesterol concentration in the tumors did not easily reach the steady-state level.

Figure 8.

Cholesterol feedback regulation in (A) liver and (B) McA-RH7777 tumor. In the tumor, an excess of the SREBP processing system relative to tissue sterol concentrations accelerated SREBP cleavage and stimulated de novo sterol synthesis. Increased oxysterols activated LXRα and stimulated sterol efflux, so that tissue sterol concentrations did not reach levels necessary for the suppression of SREBP cleavage. Inhibited bile acid synthesis also contributed to the elevation of sterol concentrations followed by the acceleration of sterol efflux in the tumor. The inhibition may be explained, at least in part, by the activation of SREBPs. However, the existence of unknown factor(s) that inhibited bile acid metabolism and upregulated the SREBP processing system is suggested. ER, endoplasmic reticulum.

Decreased catabolism may also have contributed to maintaining the high tissue cholesterol and oxysterol concentrations. In McA-RH7777 hepatomas, the activity and mRNA expression of the rate-limiting enzyme in the classic bile acid biosynthetic pathway, CYP7A1, were markedly downregulated. CYP7A1 metabolizes cholesterol and oxysterols, including 24S-, 25-, and 27-hydroxycholesterols.40, 41 Therefore, although CYP7A1 is not the only enzyme that metabolizes oxysterols, the total capacity for oxysterol catabolism may be reduced. It is noteworthy here that the activity of CYP7B1, that is, 27-hydroxycholesterol 7α-hydroxylase, was retained in tumor tissues despite markedly reduced expression of CYP7A1 and CYP7B1 mRNA, suggesting 7α-hydroxylation of 27-hydroxycholesterol is catalyzed by an enzyme other than CYP7A1 and CYP7B1 in McA-RH7777 hepatomas.

The reason for the markedly suppressed classic bile acid biosynthetic pathway in this hepatoma remains unknown, but there is a striking abnormality in the regulation of CYP7A1. The expression of SHP was extremely low in accordance with markedly reduced tissue bile acid concentrations. Both downregulation of SHP and activation of LXRα42 should stimulate CYP7A1 expression, at least in rats, but CYP7A1 did not respond to these changes. We also determined the mRNA levels of HNF4α43 and HNF6,44 liver-specific transcription factors crucial for basal level transcription of CYP7A1, but no downregulation of these factors was observed. These results may suggest the tumors possess mutations of CYP7A1. However, the attenuated expression of the mRNA of several essential genes involved in bile acid metabolism, that is, CYP7A1, CYP7B1, CYP8B1, BSEP, and NTCP, suggests they are coordinatedly downregulated by a common factor. SREBPs may be such candidate factors. It has been reported that SREBP1 interferes with recruitment of the peroxisome proliferator-activated receptor-γ coactivator-1 (PGC-1) to HNF4α in the transcriptional regulation of hepatic gluconeogenic enzymes.45 Because the interaction between HNF4α and PGC-1 is also crucial for the transactivation of CYP7A1,46 SREBP1 may inhibit CYP7A1 by the same mechanism. In fact, transgenic mice expressing active SREBP1a exhibit reduced expression of CYP7A1.47 In addition, the transcription of CYP7B1 was reported to be suppressed by SREBP1 and SREBP2 via interaction with the general transcriptional activator Sp1.48 On the other hand, the effects of SREBPs on BSEP or NTCP have yet to be studied, and the regulation of CYP8B1 by SREBPs remains controversial. SREBP2 has been reported to suppress CYP8B1 promoter activity by interacting with FTF.49 In contrast, in another study SREBP2 did not exhibit any effect on the CYP8B1 promoter, whereas SREBP1 stimulated CYP8B1 promoter activity.50 Therefore, further investigation is needed to determine a putative common factor that coordinatedly downregulates bile acid metabolism in McA-RH7777 hepatomas.

In summary, hypercholesterolemia in rats bearing McA-RH7777 hepatomas was caused by the upregulation of ABCA1 and ABCG1 as a result of the activation of LXRα. High concentrations of oxysterols in tissue maintained by overexpression of the SREBP processing system were responsible for the active LXRα in this tumor. In addition, inhibited bile acid biosynthesis by an unknown factor or factors also contributed to keeping tissue sterol concentrations elevated. Hypercholesterolemia is not a common biochemical feature of patients with hepatomas. Our results suggest the concomitant upregulation of cholesterol biosynthesis with downregulation of bile acid biosynthetic pathways may be a characteristic feature of the hepatomas that can cause paraneoplastic hypercholesterolemia in the host animals.

Ancillary