Retinoids and their target genes in liver functionsand diseasesGoshi Shiota and Keita Kanki
Division of Molecular and Genetic Medicine, Department of Genetic Medicine and Regenerative Therapeutics, Graduate School of Medicine,Tottori University, Yonago, Japan
Goshi Shiota, Division of Molecular andGenetic Medicine, Department of GeneticMedicine and Regenerative Therapeutics,Graduate School of Medicine, TottoriUniversity, Yonago 683-8503, Japan. Email:[email protected]
AbstractRetinoids have been reported to prevent several kinds of cancers, including hepatocellularcarcinoma (HCC). Retinoic acid (RA) coupled with retinoic acid receptor/retinoid Xreceptor heterodimer exerts its functions by regulating its target genes. We previouslyreported that transgenic mice, in which RA signaling is suppressed in a hepatocyte-specificmanner, developed liver cancer at a high rate, and that disruption of RA functions led to theincreased oxidative stress via aberrant metabolisms of lipid and iron, indicating thatretinoids play an important role in liver pathophysiology. These data suggest that exploringthe metabolism of retinoids in liver diseases and their target genes provides us with usefulinformation to understand the liver functions and diseases. Consequently, the alteredmetabolism of retinoids was observed in liver diseases, including non-alcoholic fatty liverdisease. In this review, we summarize the metabolism of retinoids in the liver, highlight thefunctions of retinoids in HCC, non-alcoholic fatty liver disease, and alcoholic liver disease,and discuss the target genes of RA. Investigation of retinoids in the liver will likely help usidentify novel therapies and diagnostic modalities for HCC.
Financial support: This work was supported in part by a grant-in-aid from the Ministry ofEducation, Culture, Sports, Science, and Technology of Japan.
IntroductionHepatocellular carcinoma (HCC) is the third most common cancerand is reportedly increasing worldwide.1 Prognosis of HCCpatients has improved due to the progress of local therapies ofHCC. However, biological features of HCC result in high rates ofsecondary occurrence of HCC after the successful treatment ofprimary tumor. Therefore, novel therapies are required to suppressmalignant potential of HCC.
The major cause of HCC is hepatitis C virus (HCV) in Westerncountries. In Japan, 80% of HCC patients have HCV as a cause ofcancer. The occurrence of HCC is caused by direct or indirecteffects of HCV core protein. The direct effects of HCV coreprotein on the development and progression of HCC include
production of reactive oxygen species (ROS), and altered signaltransduction of mitogen-activated protein kinases such as JNK,p38, and ERK1/2.2 The indirect effects of HCV core proteininclude abnormal turnover rate of hepatocyte death and regenera-tion, which is due to the oxidative stress generated by HCV,leading to DNA damage and mutation. In addition, activation ofhepatic stellate cells (HSCs), which migrate along the space ofDisse between hepatocytes and endothelial cells, indirectly pro-motes HCC occurrence.3 As demonstrated in experimental animalmodels, activation of HSCs induces overproduction of transform-ing growth factor-b and platelet-derived growth factor, whichpromote HCC.4,5 In addition, HSC activation and the progressionof liver diseases are associated with the loss of lipid dropletscontaining vitamin A.6,7 Since epidemiological data suggest that
doi:10.1111/jgh.12031
bs_bs_banner
33Journal of Gastroenterology and Hepatology 2013; 28 (Suppl. 1): 33–37
vitamin A acts as an inhibitor of carcinogenesis in several organs,including stomach, breast, lung, prostate, and liver,8–12 we proposethe hypothesis that the loss of vitamin A in HSCs is a contributingfactor in the progression of HCC. Importantly, hepatitis B surfaceantigen-positive individuals with lower serum retinol concentra-tions have sevenfold higher risk of HCC compared with those withhigher serum retinol concentrations,12 suggesting that reducedretinol content is a high risk factor of HCC.
Mode of action and functions ofretinoids in the liverHSCs are the largest cellular sites of retinoid storage in the body,storing as much as 80% of the entire body’s retinoid supply.13
Retinol is obtained from the diet as retinyl esters and fromb-carotene from plants.14 In the blood, retinol complexed withchylomicron and retinol-binding protein 4 is taken up by hepato-cytes, and then retinol is converted to either retinyl esters or toretinal, and subsequently to retinoic acid (RA). RAs act as agonistswhen bound to the retinoic acid receptor a, b, or g (RARs), and theretinoid X receptor a, b, or g (RXRs) (Fig. 1). The heterodimer ofRAR/RXR activates the transcription of many target genes, exert-ing many potent biological functions with respect to the regulationof cell proliferation and differentiation.13
Decreased vitamin A, an inhibitor of carcinogenesis, in chronicliver disease, such as liver cirrhosis, led us to examine the hypoth-esis that the loss of vitamin A in HSCs is a cause of HCC. Toexplore the role of RAs in the liver, we developed transgenic miceexpressing the dominant negative form of RARa in a hepatocyte-specific manner.15 These mice developed microvesicular steatosisand spotty focal necrosis at 4 months of age, and developedhepatic adenoma and HCC after 12 months of age.15 Mitochondrialb-oxidation of fatty acids was downregulated, whereas peroxiso-mal b-oxidation of fatty acids and microsomal b-oxidation of fattyacids were upregulated (Fig 2). In addition, formation of H2O2
and 8-hydroxy-2’-deoxyguanosine was increased. Expression
of b-catenin and cyclin D1 was enhanced, and T-cell factor-4(TCF-4)/b-catenin complex was also increased. In addition tothese phenomena, accelerated formation of ROS caused death andproliferation of hepatocytes, and hepatocarcinogenesis. Further-more, iron overload was observed in the liver of these mice,suggesting that loss of RA signal leads to iron deposition.16 Takentogether, these data suggest that RAs play an important role inpreventing the occurrence of HCC in association with fatty acidmetabolism, iron metabolism, and Wnt signaling.15,16
Alcohol, retinoids, and HCCIt is well known that the contents of retinoids in human livertissues are decreased in fatty liver, alcoholic hepatitis, and livercirrhosis.6 Although alcohol is known to enhance hepatocarcino-genesis, the mechanism of this action remains to be solved. Adachiet al. reported that the retinoid contents in HCC specimens andtheir surrounding tissues in patients with a high intake of alcoholwere inversely correlated with the estimated cumulative lifetimeethanol consumption, suggesting that alcohol abuse promoteshepatocarcinogenesis by depleting retinoids.7 Additionally, excessethanol intake reduces the liver’s uptake of retinyl esters as part oflipoproteins17 and induces the malabsorption of retinoids by dam-aging the intestinal epithelium.18 In conclusion, there is a closerelationship among alcohol, retinoids, and HCC.
Retinoid metabolism in non-alcoholicfatty liver disease (NAFLD)The increasing prevalence of metabolic syndrome reflects a sig-nificant increase in patients with NAFLD. A subset of NAFLDpatients develop a more aggressive form of fatty liver as non-alcoholic steatohepatitis (NASH), in which patients are at thegreater risk for progression to liver cirrhosis, liver failure, andHCC.19,20 A prospective cohort study has shown that 5-year cumu-lative incidence of HCC in NASH was 7.6%, and that older ageand advanced fibrosis were important risk factors for HCC.21
These data led us to investigate the mechanisms underlying theoccurrence of HCC in NASH patients.
Inside the cell, retinol is metabolized by various enzymes.22 Invitamin A-sufficient states in the liver, retinol taken up by hepato-cytes is transferred to perisinusoidal HSCs for storage. A largeportion of vitamin A is stored in lipid droplets of HSCs. Cellularretinol binding protein 1 (CRBP1) and retinol-esterifying enzyme(LRAT) are important for esterification of retinol, and acyl-CoA:diacylglycerol acyltransferase 1 (DGAT1) and acyl-CoA:diacylglycerol acyltransferase 2 (DGAT2) play importantroles in the formation of retinyl esters as well as triglyceridesynthesis. Hepatocytes predominantly express DGAT2 mRNA,while DGAT1 is expressed in both hepatocytes and HSCs.23
Conversely, hydrolysis of retinyl esters is catalyzed by carboxy-lesterase 1 (CES1).24 Cytosolic medium-chain alcohol dehydroge-nase enzymes, such as aldehyde dehydrogenase 1 (ADH1),aldehyde dehydrogenase 2 (ADH2), aldehyde dehydrogenase 3(ADH3), retinol dehydrogenase 5, retinol dehydrogenase 10(RDH10), and retinol dehydrogenase 11 (RDH11), are involved inthe oxidation of retinol to retinal. Oxidation of all-trans retinal toall-trans retinoic acid (ATRA) is catalyzed by retinal dehydroge-nase 1, retinal dehydrogenase 2, and retinal dehydrogenase 3. The
RAR (α, β, γ)
RXR (α, β, γ)mRNA
NucleusCytoplasma
ROH
RCHO
ATRA
9cRA
RE
RE
ROH
RBP
Chy
Transcription of target genes(AGGTCA)
RAR·RXR
Figure 1 Action mode of retinoic acid in the cells. Once retinol is takenup by the cell, retinol is oxidized to retinal, and subsequently to retinoicacid. The retinoic acid coupled with the heterodimer of RXR/RAR bindsto the retinoic acid-responsive element of target genes. ATRA, all-transretinal to all-trans retinoic acid; RAR, retinoic acid receptor; RBP, retinobinding protein; RCHO, retinal; RE, retinyl ester; ROH, retinol; RXR,retinoid X receptor.
Retinoids and their target genes G Shiota and K Kanki
34 Journal of Gastroenterology and Hepatology 2013; 28 (Suppl. 1): 33–37
heterodimer of RARs with RXRs functions as transcription factorto regulate the target genes of RA, binding to DNA sequencescalled RA-responsive element localized within the promoter oftarget genes.22 Partitioning of RAs between the two receptors isregulated by cellular retinol binding protein 2 and fatty acidbinding protein 5. These proteins specially deliver RAs from thecytosol to nuclear RAR and RXR, followed by activation of avariety of RAR/RXR downstream target genes. These includeRARa2, RARb2, CRBP1, cellular retinoic acid binding protein 1(CRABP1), ADH3, cytochrome P45026A1 (CYP26A), B-celltranslocation gene 2 (Btg2), tissue transglutaminase 2 (TGase2),and phosphoenolpyruvate carboxykinase (PEPCK). The catabo-lism of ATRA is an important mechanism regulating RA levels incells and tissues. CYP26A1 is capable of metabolizing ATRA topolar metabolites, including 4-hydroxy retinoic acid, 4-oxo ret-inoic acid, 18-hydroxy retinoic acid, and 5,6-epoxy retinoic acid.
CYP26A1 can sense the concentration of RA and regulate theoxidative metabolism of ATRA. CRABP1 is also involved in regu-lating RA degradation.25
In the liver tissues with NASH, RA-metabolism-related geneswere examined by real-time reverse transcription–polymerasechain reaction (Fig. 3).22 High expression levels of LRAT, DGAT1,and DGAT2, as well as CES1, imply that mutual conversionbetween retinyl esters and retinol is active in the liver tissues ofNASH. Additionally, upregulation of CRBP1, ADH1, ADH2,ADH3, RDH10, RDH11, DHRS3, and DHRS4 is also observed,suggesting that oxidation of retinol to retinal is actively performed.ALDH1 and ALDH3 were highly expressed. Taken together, con-version of retinol to retinal, and subsequently to RA, is enhancedin the NASH liver tissues. While we found high expression of allthe RA-metabolism-related genes analyzed in this study, expres-sion of the target genes was variable; expression of CRBP1,
PPARα
PeroxisomeMicrosome
Cyp 4a10↑Cyp 4a12↑Cyp 4a14↑
dicarboxylic acids
NucleusWnt
Signal
Cyclin D1
Mitochondria
Fatty acidβ-oxidation
Fatty acidβ-oxidation
PPARβ TCF-4/β-catenin
Fatty acidω-oxidation
H2O2
H2O2
FA
steatosiscell death
hepatocyte proliferation
hepatocarcinogenesis
8-OHdG
8-OHdG
Figure 2 Abnormal status of the cells byloss of retinoid signals. In mitochondria,b-oxidation of fatty acids is suppressed,leading to the formation of steatosis. In per-oxisome and microsome, b-oxidation andv-oxidation of fatty acids is enhanced, respec-tively. These induce reactive oxygen speciesproduction, leading to death and proliferationof hepatocytes and hepatocarcinogenesis.
Figure 3 Altered metabolism of retinoic acidin non-alcoholic steatohepatitis (NASH). Con-version of retinol to retinal, subsequently toretinoic acid is actively enhanced. Degradationof retinoic acid catalyzed by CYP26A1 isextremely enhanced. If this status continuesin the liver with NASH, the retinoids will con-sequently be lost, which contributes to thedevelopment of liver cirrhosis and hepatocel-lular carcinoma. ATRA, all-trans retinal to all-trans retinoic acid; RA, retinoic acid; RAR,retinoic acid receptor.
G Shiota and K Kanki Retinoids and their target genes
35Journal of Gastroenterology and Hepatology 2013; 28 (Suppl. 1): 33–37
CYP26A1, and PEPCK was increased, but expression of RARa2and TGase2 was decreased, while expression of RARb2, ADH3,and Btg2 remained unchanged. On the contrary, expression ofCYP26A1 was extremely high, suggesting that degradation ofATRA is very active. These data suggest that metabolism of RA isvery active in NASH, and continuous active state of RA metabo-lism causes subsequent loss of RA in the liver tissues with NASH,which may contribute to the progression of steatohepatitis to livercirrhosis and HCC.
Target genes of RAIt has been reported that more than 532 genes serve as regulatorytargets of RA.26 These include 27 genes that are the direct targetsof RA, which are regulated via the RXR/RAR heterodimer boundto a DNA response element of these genes, 105 candidate genes,and 267 genes influenced by RA, although the regulatory mecha-nisms are unclear. The indirect regulation includes the actions ofintermediate transcription factors and non-specific associationswith other proteins.26 The direct regulation is involved in retinoidresponse elements. The classical retinoid response element of atarget gene is a direct repeat of the motif 5’-PuG(G/T)TCA-3’spaced by 1,2, or 5 base pairs (DR1, DR2, and DR5, respec-tively).27 The DR2 and DR5 elements preferentially bind RXR/RAR heterodimer with RXR monomer binding the 5’ motif.RARb2, CYP26, Hoxa-1, Hoxd-4, and HNF3a have DR-5 in thepromoter region of each gene, and are the target genes of RA(Fig. 4). Exploring target genes of RA is essential for identifyingthe favorable effects of RA in the liver, and will potentially lead tothe application of these genes in the clinical setting as biomarkersand therapeutic tools. These efforts will hopefully result inimproving the prognosis of the patients with liver diseases in thenear future.
Conflict of interestThe authors declare no conflict of interest.
References1 Umemura T, Ichijo T, Yoshizawa K et al. Epidemiology of
hepatocellular carcinoma in Japan. J. Gastroenterol. 2009; 44:102–7.
2 Koike K. Hepatitis C virus contributes to hepatocarcinogenesis bymodulating metabolic and intracellular signaling pathways.J. Gastroenterol. Hepatol. 2007; 22: S108–111.
3 Amann T, Bataille F, Spruss T et al. Activated hepatic stellate cellspromotes tumorigenicity of hepatocellular carcinoma. Cancer Sci.2009; 100: 646–53.
4 Factor VM, Kao CY, Santoni-Rugiu E et al. Constitutive expressionof mature transforming growth factor beta1 in the liver accelerateshepatocarcinogenesis in transgenic mice. Cancer Res. 1997; 7:89–95.
5 Campbell JS, Hughes SD, Gilbertson DG et al. Platelet-derivedgrowth factor C induces liver fibrosis, steatosis, and hepatocellularcarcinoma. Proc. Natl Acad. Sci. U.S.A. 2005; 102: 3389–94.
6 Leo MA, Lieber CS. Hepatic vitamin A depletion in alcoholic liverinjury. N. Engl. J. Med. 1982; 307: 597–601.
7 Adachi S, Moriwaki H, Muto Y et al. Reduced retinoid content inhepatocellular carcinoma with special reference to alcoholconsumption. Hepatology 1991; 14: 776–80.
8 Larsson SC, Bergkvist L, Naslund I et al. Vitamin A, retinol, andcarotenoids and the risk of gastric cancer: a prospective cohort study.Am. J. Clin. Nutr. 2007; 85: 497–503.
9 Toniolo P, van Kappel AL, Akhmedkhanov A et al. Serumcarotenoids and breast cancer. Am. J. Epidemiol. 2001; 153:1142–7.
10 Michaud DS, Feskanich D, Rimm EB et al. Intake of specificcarotenoids and risk of lung cancer in 2 prospective US cohorts.Am. J. Clin. Nutr. 2000; 72: 990–7.
11 Schenk JM, Riboli E, Chatterjee N et al. Serum retinol and prostatecancer risk: a nested case-control study in the prostate, lung,colorectal, and ovarian cancer screening trial. Cancer Epidemiol.Biomarkers Prev. 2009; 18: 1227–31.
12 Yuan J-M, Gao Y-T, Ong C-N et al. Prediagnostic level of serumretinol in relation to reduced risk of hepatocellular carcinoma.J. Natl. Cancer Inst. 2006; 98: 482–90.
13 Lee YS, Jeong W. Retinoic acid and hepatic stellate cells in liverdiseases. J. Gastroenterol. Hepatol. 2012; 27 (Suppl. 2):76–9.
14 Mongan NP, Gudas LJ. Diverse action pf retinoid receptors in cancerprevention and treatment. Differentiation 2007; 75: 853–70.
15 Yanagitani A, Yamada S, Yasui S et al. Retinoic acid receptor adominant negative from causes steatohepatitis and liver tumors intransgenic mice. Hepatology 2004; 40: 366–75.
16 Tsuchiya H, Akechi Y, Ikeda R et al. Suppressive effects of retinoidson iron-induced oxidative stress in the liver. Gastroenterology 2009;136: 341–50.
17 Sato M, Lieber CS. Changes in vitamin A status after acute ethanoladministration in the rat. J. Nutr. 1982; 112: 1188–96.
18 Underwood BA. Vitamin A in animal and human nutrition. In:Sporn MB, Roberts AB, Goodman DS, eds. The Retinoids, Vol. 1.New York: Academic Press, 1984; 281–392.
19 Ong JP, Younossi ZM. Epidemiology and natural history of NAFLDand NASH. Clin. Liver Dis. 2007; 11: 1–16.
20 Safaty L, Lemoine M. Definition and history of metabolic steatosis:clinical aspects of NAFLD, NASH and cirrhosis. Diabetes Metab.2008; 34: 634–7.
Figure 4 Target genes of retinoic acid. The classical retinoid responseelement of a target gene of retinoic acid is a direct repeat of the motif5’-PuG(G/T)TCA-3’ spaced by 5 base pairs (DR5). RARb2, CYP26A1,Hoxa-1, Hoxd-4, and HNF3a have DR-5 in their promoter regions. ATRA,all-trans retinal to all-trans retinoic acid; RA, retinoic acid; RAR, retinoicacid receptor; RXR, retinoid X receptor.
Retinoids and their target genes G Shiota and K Kanki
36 Journal of Gastroenterology and Hepatology 2013; 28 (Suppl. 1): 33–37
24 Sanghani SP, Davis WI, Dumaual NG et al. Identification ofmicrosomal rat liver carboxylesterases and their activity with retinylpalmitate. Eur. J. Biochem. 2002; 269: 4387–98.
25 Boylan JF, Gudas LJ. The level of CRABP-I expression influencesthe amounts and types of all-trans-retinoic acid metabolites in F9tetracarcinoma stem cells. J. Biol. Chem. 1992; 67: 21486–91.
26 Balmer JE, Blomhoff R. Gene expression regulation by retinoic acid.J. Lipid Res. 2002; 43: 1773–908.
27 McGrane MM. Vitamin A regulation of gene expression: molecularmechanism of a prototype gene. J. Nutr. Biochem. 2007; 18:497–508.
G Shiota and K Kanki Retinoids and their target genes
37Journal of Gastroenterology and Hepatology 2013; 28 (Suppl. 1): 33–37
