A retinoic acid receptor β2 agonist attenuates transcriptome and metabolome changes underlying nonalcohol-associated fatty liver disease
2021; Elsevier BV; Volume: 297; Issue: 6 Linguagem: Inglês
10.1016/j.jbc.2021.101331
ISSN1083-351X
AutoresXiao‐Han Tang, Marta Melis, Changyuan Lu, Andrew Rappa, Tuo Zhang, José Jessurun, Steven S. Gross, Lorraine J. Gudas,
Tópico(s)Peroxisome Proliferator-Activated Receptors
ResumoNonalcohol-associated fatty liver disease (NAFLD) is characterized by excessive hepatic accumulation of fat that can progress to steatohepatitis, and currently, therapeutic options are limited. Using a high-fat diet (HFD) mouse model of NAFLD, we determined the effects of the synthetic retinoid, AC261066, a selective retinoic acid receptor β2 (RARβ2) agonist, on the global liver transcriptomes and metabolomes of mice with dietary-induced obesity (DIO) using genome-wide RNA-seq and untargeted metabolomics. We found that AC261066 limits mRNA increases in several presumptive NAFLD driver genes, including Pklr, Fasn, Thrsp, and Chchd6. Importantly, AC261066 limits the increases in the transcript and protein levels of KHK, a key enzyme for fructose metabolism, and causes multiple changes in liver metabolites involved in fructose metabolism. In addition, in cultured murine hepatocytes, where exposure to fructose and palmitate results in a profound increase in lipid accumulation, AC261066 limits this lipid accumulation. Importantly, we demonstrate that in a human hepatocyte cell line, RARβ is required for the inhibitory effects of AC261066 on palmitate-induced lipid accumulation. Finally, our data indicate that AC261066 inhibits molecular events underpinning fibrosis and exhibits anti-inflammatory effects. In conclusion, changes in the transcriptome and metabolome indicate that AC261066 affects molecular changes underlying multiple aspects of NAFLD, including steatosis and fibrosis. Therefore, we suggest that AC261066 may have potential as an effective therapy for NAFLD. Nonalcohol-associated fatty liver disease (NAFLD) is characterized by excessive hepatic accumulation of fat that can progress to steatohepatitis, and currently, therapeutic options are limited. Using a high-fat diet (HFD) mouse model of NAFLD, we determined the effects of the synthetic retinoid, AC261066, a selective retinoic acid receptor β2 (RARβ2) agonist, on the global liver transcriptomes and metabolomes of mice with dietary-induced obesity (DIO) using genome-wide RNA-seq and untargeted metabolomics. We found that AC261066 limits mRNA increases in several presumptive NAFLD driver genes, including Pklr, Fasn, Thrsp, and Chchd6. Importantly, AC261066 limits the increases in the transcript and protein levels of KHK, a key enzyme for fructose metabolism, and causes multiple changes in liver metabolites involved in fructose metabolism. In addition, in cultured murine hepatocytes, where exposure to fructose and palmitate results in a profound increase in lipid accumulation, AC261066 limits this lipid accumulation. Importantly, we demonstrate that in a human hepatocyte cell line, RARβ is required for the inhibitory effects of AC261066 on palmitate-induced lipid accumulation. Finally, our data indicate that AC261066 inhibits molecular events underpinning fibrosis and exhibits anti-inflammatory effects. In conclusion, changes in the transcriptome and metabolome indicate that AC261066 affects molecular changes underlying multiple aspects of NAFLD, including steatosis and fibrosis. Therefore, we suggest that AC261066 may have potential as an effective therapy for NAFLD. Nonalcohol-associated fatty liver disease (NAFLD), which is defined as the accumulation of intrahepatic triglycerides without excessive alcohol intake and is usually associated with obesity, has become a primary cause of chronic liver disease. NAFLD can progress through histologically and clinically defined stages to nonalcohol steatohepatitis (NASH) or liver cirrhosis (1Yeh M.M. Brunt E.M. Pathological features of fatty liver disease.Gastroenterology. 2014; 147: 754-764Abstract Full Text Full Text PDF PubMed Scopus (194) Google Scholar). In the United States, the number of NAFLD cases is projected to reach over 100 million in 2030, and 27% of adult NAFLD cases progress to NASH (2Estes C. Razavi H. Loomba R. Younossi Z. Sanyal A.J. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease.Hepatology. 2018; 67: 123-133Crossref PubMed Scopus (656) Google Scholar). Excessive accumulation of lipids in the liver induces liver stress and injury, resulting in the fibrogenesis and inflammation often observed in NASH (3Friedman S.L. Neuschwander-Tetri B.A. Rinella M. Sanyal A.J. Mechanisms of NAFLD development and therapeutic strategies.Nat. Med. 2018; 24: 908-922Crossref PubMed Scopus (903) Google Scholar). Despite an emerging NAFLD health crisis worldwide, to date there is no single FDA-approved therapy for preventing and/or treating NAFLD other than dietary intervention, weight loss, and medications for insulin resistance and hyperlipidemia (4Sumida Y. Yoneda M. Current and future pharmacological therapies for NAFLD/NASH.J. Gastroenterol. 2018; 53: 362-376Crossref PubMed Scopus (250) Google Scholar). Therefore, it is crucial to identify and target the underlying molecular mechanisms that cause NAFLD to find a novel therapy. Here we explore a vitamin A agonist that defines a potential therapeutic approach. Carotenoids and rerinoids, including vitamin A (retinol) and its metabolites, such as all-trans retinoic acid (RA), exert regulatory functions on multiple physiological processes (5Chen Y. Clarke O.B. Kim J. Stowe S. Kim Y.K. Assur Z. Cavalier M. Godoy-Ruiz R. von Alpen D.C. Manzini C. Blaner W.S. Frank J. Quadro L. Weber D.J. Shapiro L. et al.Structure of the STRA6 receptor for retinol uptake.Science. 2016; 26: 353-381Google Scholar, 6Brun P.J. Grijalva A. Rausch R. Watson E. Yuen J.J. Das B.C. Shudo K. Kagechika H. Leibel R.L. Blaner W.S. Retinoic acid receptor signaling is required to maintain glucose-stimulated insulin secretion and β-cell mass.FASEB J. 2015; 29: 671-683Crossref PubMed Scopus (30) Google Scholar), including lipid metabolism and hyperglycemia control (6Brun P.J. Grijalva A. Rausch R. Watson E. Yuen J.J. Das B.C. Shudo K. Kagechika H. Leibel R.L. Blaner W.S. Retinoic acid receptor signaling is required to maintain glucose-stimulated insulin secretion and β-cell mass.FASEB J. 2015; 29: 671-683Crossref PubMed Scopus (30) Google Scholar, 7Berry D.C. Noy N. All-trans-retinoic acid represses obesity and insulin resistance by activating both peroxisome proliferation-activated receptor beta/delta and retinoic acid receptor.Mol. Cell. Biol. 2009; 29: 3286-3296Crossref PubMed Scopus (234) Google Scholar, 8Trasino S.E. Benoit Y.D. Gudas L.J. Vitamin A deficiency causes hyperglycemia and loss of pancreatic β-cell mass.J. Biol. Chem. 2015; 290: 1456-1473Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar). Retinoic acid receptors (RARs α, β, and γ subtypes) are transcription factors that heterodimerize with retinoid X receptors (RXRs, α, β, and γ subtypes) and bind the endogenous agonist, RA, to regulate gene expression (9Tang X.H. Gudas L.J. Retinoids, retinoic acid receptors, and cancer.Annu. Rev. Pathol. 2011; 6: 345-364Crossref PubMed Scopus (389) Google Scholar). In humans, histological stages of NAFLD, including mild and severe steatosis, NASH, and hepatocyte necrosis, show a strong inverse correlation with hepatic retinol levels (10Chaves G.V. Pereira S.E. Saboya C.J. Spitz D. Rodrigues C.S. Ramalho A. Association between liver vitamin A reserves and severity of nonalcoholic fatty liver disease in the class III obese following bariatric surgery.Obes. Surg. 2014; 24: 219-224Crossref PubMed Scopus (26) Google Scholar). Hepatic retinoic acid levels are significantly lower in human NAFLD samples than in normal liver samples (11Zhong G. Kirkwood J. Won K.J. Tjota N. Jeong H. Isoherranen N. Characterization of vitamin A metabolome in human livers with and without nonalcoholic fatty liver disease.J. Pharmacol. Exp. Ther. 2019; 370: 92-103Crossref PubMed Scopus (23) Google Scholar), and we also showed an inverse correlation between steatosis and hepatic retinol and retinyl palmitate levels (12Trasino S.E. Tang X.H. Jessurun J. Gudas L.J. Obesity leads to tissue, but not serum vitamin A deficiency.Sci. Rep. 2015; 5: 15893Crossref PubMed Scopus (51) Google Scholar). RA, a vitamin A metabolite and the endogenous agonist for all three RARs α, β, and γ, attenuated diet-induced liver steatosis in mice, indicating that activation of retinoic acid signaling could be novel therapy for NAFLD (6Brun P.J. Grijalva A. Rausch R. Watson E. Yuen J.J. Das B.C. Shudo K. Kagechika H. Leibel R.L. Blaner W.S. Retinoic acid receptor signaling is required to maintain glucose-stimulated insulin secretion and β-cell mass.FASEB J. 2015; 29: 671-683Crossref PubMed Scopus (30) Google Scholar). We showed that the synthetic retinoid AC261066 (13Lund B.W. Piu F. Gauthier N.K. Eeg A. Currier E. Sherbukhin V. Brann M.R. Hacksell U. Olsson R. Discovery of a potent, orally available, and isoform-selective retinoic acid beta2 receptor agonist.J. Med. Chem. 2005; 48: 7517-7519Crossref PubMed Scopus (53) Google Scholar) corrected hyperglycemia in type 2 diabetes mouse models, limited hepatic lipid accumulation, and prevented early hepatic fibrogenic events in a high-fat diet (HFD)-induced NAFLD mouse model (14Trasino S.E. Tang X.H. Jessurun J. Gudas L.J. A retinoic acid receptor β2 agonist reduces hepatic stellate cell activation in nonalcoholic fatty liver disease.J. Mol. Med. (Berl.). 2016; 94: 1143-1151Crossref PubMed Scopus (23) Google Scholar, 15Trasino S.E. Tang X.H. Jessurun J. Gudas L.J. Retinoic acid receptor β2 agonists restore glycaemic control in diabetes and reduce steatosis.Diabetes Obes. Metab. 2016; 18: 142-151Crossref PubMed Scopus (24) Google Scholar). However, we did not explore the molecular mechanisms involved in AC261066's actions in depth in our prior work. Here, in addition to delineating the effects of AC261066 on the physiology, transcriptome, and metabolome in a related HFD-driven NAFLD mouse model, we establish a causal role for RARβ in regulating lipid metabolism and demonstrate that AC261066 acts through RARβ. These novel and important data suggest that AC261066 could be a useful drug for the treatment of NAFLD. Here we found that AC261066 effectively limited liver steatosis and glucose excursion induced by a HFD with 60% kcal from fat, supporting our previous findings in a similar DIO (diet-induced obesity) model (14Trasino S.E. Tang X.H. Jessurun J. Gudas L.J. A retinoic acid receptor β2 agonist reduces hepatic stellate cell activation in nonalcoholic fatty liver disease.J. Mol. Med. (Berl.). 2016; 94: 1143-1151Crossref PubMed Scopus (23) Google Scholar, 15Trasino S.E. Tang X.H. Jessurun J. Gudas L.J. Retinoic acid receptor β2 agonists restore glycaemic control in diabetes and reduce steatosis.Diabetes Obes. Metab. 2016; 18: 142-151Crossref PubMed Scopus (24) Google Scholar, 16Melis M. Tang X.H. Trasino S.E. Patel V.M. Stummer D.J. Jessurun J. Gudas L.J. Effects of AM80 compared to AC261066 in a high fat diet mouse model of liver disease.PLoS One. 2019; 14e0211071Crossref PubMed Scopus (6) Google Scholar) (Fig. S1). Thus, we used this 60% HFD model to explore further the molecular mechanisms by which AC261066 exerts these potentially beneficial effects in the livers of these HFD-fed mice. Since RARβ is a transcription factor in the nuclear receptor superfamily (9Tang X.H. Gudas L.J. Retinoids, retinoic acid receptors, and cancer.Annu. Rev. Pathol. 2011; 6: 345-364Crossref PubMed Scopus (389) Google Scholar), we next explored the mechanism(s) of action of AC261066 by using RNA-seq to assess global changes in liver transcripts. A total of 4069 transcripts (differentially expressed genes (DEGs)) differed significantly between the livers from HFD-fed and chow-fed mice, including increases in 1942 genes and decreases in 2127 transcripts, respectively (q < 0.1) (q = p value adjusted for the false discovery rate) (Fig. 1A and Table S1). Additionally, we found that compared with the HFD group, the HFD+AC261066 group showed 746 significantly changed DEGs, including increases in 397 transcripts and decreases in 349 transcripts (q < 0.1) (Fig. 1B and Table S1). Hierarchical clustering of DEGs in the different groups is shown in the heatmaps in Fig. S2 (A, HFD/chow; B, HFD+AC261066 (HFD+AC261)/HFD). Importantly, 225 transcripts increased in the HFD/chow were reduced in the HFD+AC261066/HFD, while 286 transcripts decreased in the HFD/chow were increased in the HFD+AC261066/HFD (Table S2). These gene overlaps between HFD/chow and HFD+AC261066/HFD were more significant than expected random chances, with a p-value < 2.2e-16 using Fisher's exact tests. These data indicate that AC261066 limits the HFD-induced transcript changes in the liver. To probe the relevance of our HFD model to human NAFLD, we used disease signature pathway analysis by DisGeNET (17Piñero J. Queralt-Rosinach N. Bravo À. Deu-Pons J. Bauer-Mehren A. Baron M. Sanz F. Furlong L.I. DisGeNET: A discovery platform for the dynamical exploration of human diseases and their genes.Database (Oxford). 2015; 2015bav028Crossref PubMed Scopus (491) Google Scholar) (Fig. 1, C and D and Table S3), as a discovery platform designed to probe a variety of gene and disease associations. The disease signature analysis revealed that the increased transcripts in the HFD/chow were associated with signature/pathways related to NAFLD, including "Liver cirrhosis experimental," "Fibrosis, Liver," "Fatty Liver," "Cholestasis," and "Nonalcoholic Fatty Disease" (Fig. 1C). The decreased transcripts in the HFD+AC261066/HFD were also associated with signature/pathways related to NAFLD, including "Liver cirrhosis experimental," "Fibrosis, Liver," "Fatty Liver," "Cholestasis," and "Nonalcoholic Fatty Disease" (Fig. 1D). These data indicate that our HFD mouse model is clinically relevant and mimics the gene signature of advanced human NAFLD, including liver fibrosis and liver cirrhosis, and that AC261066 impacted some pathways involved in NAFLD development. Using the Gene Set Enrichment Analysis (GSEA) with Gene Ontology (GO) Biological Process Database (18Subramanian A. Tamayo P. Mootha V.K. Mukherjee S. Ebert B.L. Gillette M.A. Paulovich A. Pomeroy S.L. Golub T.R. Lander E.S. Mesirov J.P. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles.Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 15545-15550Crossref PubMed Scopus (21539) Google Scholar), we discovered that compared with the chow group, the pathways with transcripts increased in the HFD group included "Arachidonic acid metabolic process" and "Long-chain fatty acid metabolic process," while pathways with transcripts decreased in the HFD group included "Cytoplasmic translation" and "Cellular response to glucocorticoid stimulus." Compared with the HFD group, the transcripts that decreased in the HFD+AC261066 group were enriched in "Extracellular matrix organization," "Fatty acid metabolic process," "Long-chain fatty acid metabolic process," "Unsaturated fatty acid metabolic process," and "Tricarboxylic acid metabolic process." This suggests that these transcript changes were associated with HFD-induced NAFLD and AC261066's protective actions on NAFLD (Figs. 1 and S1). Of note, we found that 282 transcripts changed in the HFD/chow are also altered by HFD in a previous study (19Soltis A.R. Kennedy N.J. Xin X. Zhou F. Ficarro S.B. Yap Y.S. Matthews B.J. Lauffenburger D.A. White F.M. Marto J.A. Davis R.J. Fraenkel E. Hepatic dysfunction caused by consumption of a high-fat diet.Cell Rep. 2017; 21: 3317-3328Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar) (Table S4). Using multiomics data, Krishnan et al. (20Chella Krishnan K. Kurt Z. Barrere-Cain R. Sabir S. Das A. Floyd R. Vergnes L. Zhao Y. Che N. Charugundla S. Qi H. Zhou Z. Meng Y. Pan C. Seldin M.M. et al.Integration of multi-omics data from mouse diversity panel highlights mitochondrial dysfunction in non-alcoholic fatty liver disease.Cell Syst. 2018; 6: 103-115.e7Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar) reported key driver genes underlying NAFLD progression: Pklr (Pyruvate Kinase L/R), Fasn, Thrsp, and Chchd6. We also observed significantly greater hepatic mRNA levels of these four genes in the HFD group compared with the chow group (Fig. 2). Interestingly, these transcript levels were lower in the HFD+AC261066 (Abbreviation AC261) compared with the HFD group (Fig. 2, A and B). Further, the protein levels of these driver genes, PKLR, THRSP, and FASN, were greater in the HFD group by 51.8, 1.5, and 5.1-fold, respectively, than in chow. These proteins, especially PKLR and FASN, were lower by 2.7 and 3.9-fold, respectively, in the HFD+AC261066 compared with the HFD group (Fig. 2, C and D). Moreover, in the HFD group, FASN protein was detected in the livers enriched for lipid droplets (Fig. 2E). Taken together, these data suggest that AC261066 mitigates HFD-induced NAFLD, at least in part, by regulating NAFLD driver gene expression. This HFD-driven NAFLD model produced severe liver steatosis (i.e., >66% of fat within the hepatocytes (21Petta S. Maida M. Macaluso F.S. Di Marco V. Cammà C. Cabibi D. Craxì A. The severity of steatosis influences liver stiffness measurement in patients with nonalcoholic fatty liver disease.Hepatology. 2015; 62: 1101-1110Crossref PubMed Scopus (142) Google Scholar), and AC261066 limited lipid accumulation in the liver (Fig. S1, B and C)). Therefore, we assessed the key genes involved in de novo lipogenesis, fatty acid import, and disposal in the liver. In line with our previous report (15Trasino S.E. Tang X.H. Jessurun J. Gudas L.J. Retinoic acid receptor β2 agonists restore glycaemic control in diabetes and reduce steatosis.Diabetes Obes. Metab. 2016; 18: 142-151Crossref PubMed Scopus (24) Google Scholar), the heatmap and the quantitative comparison of the transcript levels derived from the RNA-seq data (Fig. 3, A and B) show that AC261066 mitigated the HFD-induced changes in key transcripts involved in these processes, including Pparg and Srebf1. AC261066 also greatly reduced the HFD-induced increase in PPARγ protein (Fig. 3, C and D). Transcript levels of Scd1, Acaca (ACC), and Mlxipl (ChREBP), genes involved in de novo lipogenesis, did not differ among all groups (Fig. S3). CD36, a fatty acid translocase protein regulated by PPARγ, mediates uptake of circulating fatty acids by the liver and contributes to the increased uptake of lipids in NAFLD and NASH (22Miquilena-Colina M.E. Lima-Cabello E. Sánchez-Campos S. García-Mediavilla M.V. Fernández-Bermejo M. Lozano-Rodríguez T. Vargas-Castrillón J. Buqué X. Ochoa B. Aspichueta P. González-Gallego J. García-Monzón C. Hepatic fatty acid translocase CD36 upregulation is associated with insulin resistance, hyperinsulinaemia and increased steatosis in non-alcoholic steatohepatitis and chronic hepatitis C.Gut. 2011; 60: 1394-1402Crossref PubMed Scopus (250) Google Scholar). Cd36 mRNA levels in the HFD and HFD+AC261066 groups were 2.8-fold and 1.6-fold higher, respectively, than in chow (Fig. 3, A and B). Strikingly, the CD36 protein levels in the HFD and the HFD+AC261066 groups were 13.6- and 1.9-fold higher, respectively, than in the chow group (Fig. 3, C and D). In the HFD group, we detected the CD36 protein in liver parenchyma with the highest levels of lipid droplets, and CD36 was enriched on the plasma membrane (Fig. 3E). In contrast, we detected lower levels of CD36 in both the chow and the HFD+AC261066 groups, consistent with previous reports (22Miquilena-Colina M.E. Lima-Cabello E. Sánchez-Campos S. García-Mediavilla M.V. Fernández-Bermejo M. Lozano-Rodríguez T. Vargas-Castrillón J. Buqué X. Ochoa B. Aspichueta P. González-Gallego J. García-Monzón C. Hepatic fatty acid translocase CD36 upregulation is associated with insulin resistance, hyperinsulinaemia and increased steatosis in non-alcoholic steatohepatitis and chronic hepatitis C.Gut. 2011; 60: 1394-1402Crossref PubMed Scopus (250) Google Scholar) (Fig. 3E). Thus, part of AC261066's actions may be explained by reduced fatty acid uptake in the liver. Dgat2 and Mogat1 play crucial roles in the synthesis of triglycerides, the primary storage form of intracellular lipid (23Shi Y. Cheng D. Beyond triglyceride synthesis: The dynamic functional roles of MGAT and DGAT enzymes in energy metabolism.Am. J. Physiol. Endocrinol. Metab. 2009; 297: E10-E18Crossref PubMed Scopus (142) Google Scholar). Cidea mRNA, important for lipid droplet formation, is expressed at a low level in healthy liver and is robustly increased in steatosis (24Hall A.M. Brunt E.M. Chen Z. Viswakarma N. Reddy J.K. Wolins N.E. Finck B.N. Dynamic and differential regulation of proteins that coat lipid droplets in fatty liver dystrophic mice.J. Lipid Res. 2010; 51: 554-563Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar). AC261066 attenuated HFD-induced increases in hepatic Dgat2, Mogat1, and Cidea mRNAs (Fig. 3B), suggesting that this may be another effector pathway for AC261066. MTTP (microsomal triglyceride transfer protein) and VLDL (very low density lipoprotein) play primary roles in the export of triglycerides from the liver (25Donnelly K.L. Smith C.I. Schwarzenberg S.J. Jessurun J. Boldt M.D. Parks E.J. Sources of fatty acids stored in liver and secreted via lipoproteins in patients with nonalcoholic fatty liver disease.J. Clin. Invest. 2005; 115: 1343-1351Crossref PubMed Scopus (2144) Google Scholar), and VLDLR (VLDL receptor) hinders the function of VLDL by binding to them. Compared with the chow group, HFD decreased Mttp and Vldlr mRNAs, respectively, and in the HFD+AC261066 group these changes were attenuated (Fig. 3, A and B). Collectively, these transcript data suggest that AC261066's effects extend to multiple aspects of lipid metabolism, including endogenous lipogenesis and fatty acid transport. To test whether these changes in gene expression are reflected at the metabolic level, we used an untargeted metabolomics approach (see Experimental procedures). We discovered that the liver levels of 343 metabolites were increased and 303 were decreased by HFD (p < 0.1). AC261066 increased 164 and decreased 172 metabolites (Table S5). Strikingly, the changes in some metabolites elicited by AC261066 in the HFD group were in opposite directions, as might be expected, suggesting that HFD alters hepatic metabolism and that AC261066 ameliorates some HFD-induced metabolite changes related to changes in the transcriptome and in the protein levels of key genes. In addition to the changes in triglyceride levels in the liver, shown in Fig. S1, metabolomics studies show that both HFD and AC261066 affect the levels of some fatty acids and other lipid species, including phosphatidylcholines (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), and lysophospholipids in the liver (Fig. S4). These changes in the phospholipids and lysophospholipids are quite interesting. Pathway analysis shows that among the top pathways increased in the HFD/chow are "Warburg Effect," "Fructose and Mannose Degradation," "Ketone Body Metabolism," "Glycolysis," and "De Novo Triacylglycerol Biosynthesis"; among top pathways decreased in HFD/chow were "Lysine Degradation," "Arginine/Proline Metabolism," and "Tryptophan Metabolism" (Fig. S5A). In contrast, the top pathways decreased in the HFD+AC261066/HFD included "Warburg Effect" and "Fructose and Mannose Degradation," whereas the top pathways increased included "Methionine Metabolism" and "Glycine/Serine Metabolism" (Fig. S5B). These data suggest that in our HFD-driven NAFLD model, some carbohydrate, lipid, and amino acid metabolism pathways are markedly altered as compared with the chow group, and that AC261066 treatment reversed/prevented these metabolite alterations. Fructose metabolism promotes de novo lipid biosynthesis in the liver and is hypothesized to be a key contributor to NAFLD progression, including NASH (26Jensen T. Abdelmalek M.F. Sullivan S. Nadeau K.J. Green M. Roncal C. Nakagawa T. Kuwabara M. Sato Y. Kang D.H. Tolan D.R. Sanchez-Lozada L.G. Rosen H.R. Lanaspa M.A. Diehl A.M. et al.Fructose and sugar: A major mediator of non-alcoholic fatty liver disease.J. Hepatol. 2018; 68: 1063-1075Abstract Full Text Full Text PDF PubMed Scopus (274) Google Scholar), and our data support this idea. Our RNA-seq data show that transcripts involved in fructose metabolism were increased in the HFD compared with the chow group, and these increases were attenuated in the HFD+AC261066 group, including (a) Slc2a2 (Glucose transporter 2, GLUT2), a fructose transporter; (b) Khk (Ketohexokinase), the first fructose metabolizing enzyme that rapidly phosphorylates fructose to generate fructose-1-P (fructose-1-phosphate); and (c) Xdh (xanthine dehydrogenase), an enzyme that produces uric acid and contributes to oxidative stress (26Jensen T. Abdelmalek M.F. Sullivan S. Nadeau K.J. Green M. Roncal C. Nakagawa T. Kuwabara M. Sato Y. Kang D.H. Tolan D.R. Sanchez-Lozada L.G. Rosen H.R. Lanaspa M.A. Diehl A.M. et al.Fructose and sugar: A major mediator of non-alcoholic fatty liver disease.J. Hepatol. 2018; 68: 1063-1075Abstract Full Text Full Text PDF PubMed Scopus (274) Google Scholar) (Fig. 4A). The KHK protein levels also correlated with KHK mRNA levels: the HFD/chow ratio was 2.73 and the HFD+AC261066/chow ratio was 1.8 (Fig. 4, A–C). Additionally, Khk transcript levels in all treatment groups positively correlated with those of all NAFLD driver genes (Fig. S9A), the lipogenesis promoting transcription factors Srebf1 and Pparg, and Xdh (Fig. S9B). These data suggest that fructose metabolism plays an important role in HFD-induced NAFLD development, and that AC261066 limits HFD-induced increases in fructose metabolism in the liver. We also found that the ratios of fructose, fructose-1-P (fructose-1-phosphate), IMP (inosine monophosphate), and uric acid in HFD/chow were 2.4, 2.1, 2.6, and 2.7, respectively, indicating that fructose metabolism was increased by HFD. In the HFD+AC261066/chow, the ratios of fructose, fructose-1-P, IMP, and uric acid were 1.8, 0.9, 1.8, and 1.7, indicating that AC261066 limited these HFD-induced increases in fructose metabolites (Figs. S5C and 4D). Moreover, these metabolite levels were positively correlated with each other (Fig. S9C). Since the dietary fructose in the chow and the HFD was comparable in our study (datasheets, Pico Diet, and BioServ Diet), we explored potential mechanisms related to the increase in the hepatic fructose level we observed in the HFD group. Endogenous fructose is produced from glucose through the sorbitol (polyol) pathway in which glucose is first converted to sorbitol by aldose reductase; then, sorbitol is further metabolized to fructose by sorbitol dehydrogenase. Hyperglycemia activates the polyol pathway (27Hannou S.A. Haslam D.E. McKeown N.M. Herman M.A. Fructose metabolism and metabolic disease.J. Clin. Invest. 2018; 128: 545-555Crossref PubMed Scopus (182) Google Scholar). Although we did not detect changes in the mRNA levels of the enzymes aldose reductase (Akr1b1) and sorbitol dehydrogenase (Sord) in our RNA-seq data (Fig. S6), the hepatic sorbitol level in the HFD group was 1.6-fold greater than in chow, and the HFD+AC261066 group exhibited sorbitol levels comparable to those in the chow group (Fig. 4D). Fig. S9C shows a positive correlation between the levels of fructose and sorbitol. Therefore, the elevation of fructose in the HFD group likely resulted from both increased transport of fructose via GLUT2 and from the activated sorbitol pathway. Our data suggest that AC261066 negatively regulates both hepatic fructose transport and the hyperglycemia-activated polyol pathway. To determine if these effects of AC261066 were direct effects on hepatocytes, we first employed the AML12 hepatocyte cell line. Neither control (vehicle-treated) or fructose-treated cells had detectable lipid droplets, while palmitate alone caused a low level of lipid droplet formation (Fig. 4F). Palmitate+fructose-treated cells exhibited a dramatic increase (>10-fold) in the lipid droplet level compared with other groups (Fig. 4F), indicating that fructose promotes endogenous lipogenesis. Importantly, we also found that AC261066 effectively limited the accumulation of lipid droplets induced by palmitate alone and the combination of palmitate and fructose, demonstrating that AC261066 can directly inhibit lipid accumulation in hepatocytes in this model system. To elucidate whether AC261066, a selective RARβ2 agonist (13Lund B.W. Piu F. Gauthier N.K. Eeg A. Currier E. Sherbukhin V. Brann M.R. Hacksell U. Olsson R. Discovery of a potent, orally available, and isoform-selective retinoic acid beta2 receptor agonist.J. Med. Chem. 2005; 48: 7517-7519Crossref PubMed Scopus (53) Google Scholar), limits liver steatosis via RARβ in hepatocytes, we examined AC261066's effects on lipid accumulation in parental and RARβ knockout (RARβ KO) HepG2 cells generated using CRISPR/Cas9 technology (Fig. 5A). After determining the success of the Crispr/Cas9 editing by Sanger sequencing (Fig. 5B) and Next Generation Sequencing (Fig. 5C), we evaluated the effects of AC261066 on the mRNA levels of RARβ and RARβ2, the latter being the most abundant RARβ isoform. We found that in the HepG2 parental line after a 72 h treatment with 2 μM AC261066, the RARβ and RARβ2 mRNA levels were increased by 3.3-fold (±0.0004; p = 0.0007) and 7.3-fold (±0.001; p = 0.03), respectively (Fig. 5D). By contrast, we observed no changes in the HepG2 RARβ KO cells treated in parallel. Next, we treated parental and RARβ KO HepG2 cells with oleate and palmitate for 48 h ± 2 μM AC261066 and found that (i) RARβ KO HepG2 cells accumulated much greater levels of lipids (10.76-fold (±5.96; p < 0.0001)) than parental cells; (ii) treatment with AC261066 mitigated the accumulation of lipids in the parental cells by 3.39-fold (±5.5; p = 0.04) but did not show any effects in the oleate and palmitate+AC261066-treated RARβ
Referência(s)