Hepatic glucose sensing: does flux matter?

Abstract

In this issue of the JCI, Denechaud et al. report studies investigating the role of the liver X receptors (LXRs) LXRα and LXRβ in carbohydrate sensing by the liver (see the related article beginning on page 956). The results of this study, which utilized LXRα/β double-KO mice, strongly contradict a recent Nature report that proposed that LXRα/β sense glucose independent of metabolic flux. The reported findings further support a key role for the carbohydrate-responsive element–binding protein (ChREBP) in the regulation of lipogenic genes by glucose and dietary carbohydrates.

The liver plays a central and vitally important role in glucose homeostasis by continuously adapting its own metabolism to the availability of nutrients and overall energetic needs of the body. The conversion of carbohydrates and amino acids into fatty acids, a process referred to as lipogenesis, is important for the preservation of excess energy. This metabolic conversion involves the induction of a set of lipogenic genes that have long been known to be regulated by dietary intake, particularly carbohydrates. However, because many lipogenic genes are also regulated by insulin and the concentration of insulin itself is acutely sensitive to changes in blood glucose concentration, only in recent years have we learned how carbohydrate, principally glucose, is able to modulate gene expression in a cell-autonomous manner.

The ability of mammalian cells to sense changes in glucose has long been thought to depend on changes in metabolic flux, as determined by the specific hexose transporters and hexokinases expressed in a given cell. Hepatocytes express both glucose transporter 2 (GLUT2) and glucokinase (GK; also known as hexokinase 4), both of which exhibit affinities for glucose (K^m and glucose S^0.5, respectively) in the low millimolar range. This enables metabolic flux in hepatocytes to essentially parallel any change in blood glucose concentration (1). In this regard, the liver mimics the pancreatic β cell, the prototypical glucose-sensitive cell type. However, functional parallels between these 2 cell types begin to fade beyond this point, due to the very distinct physiological functions of the liver and β cell in maintaining glucose homeostasis.

While mammals are thought to sense glucose in a manner that directly reflects glucose flux, other mechanisms that do not rely on catalysis exist in both plants and yeast. For instance, the plant hexokinase HXK1 remains capable of signaling after being mutated in a manner that eliminates all catalytic activity (2). A recent study in Nature by Mitro et al. reported that glucose is able to directly bind to and increase the transcriptional activity of the liver X receptors (LXRs) (3), thereby suggesting that these nuclear receptors also function as glucose sensors in the liver. If true, this would establish a new paradigm for glucose sensing in mammalian cells. The study by Denechaud et al. in this issue of the JCI (4) was directed at exploring this proposed paradigm by assessing the effects of a high-carbohydrate diet on hepatic lipogenesis in mice lacking both LXRα and LXRβ.

SREBP-1c and ChREBP: key players in hepatic lipogenesis

Over recent years, information has emerged indicating that SREBP-1c and carbohydrate-responsive element–binding protein (ChREBP) both contribute to the regulation of lipogenesis in liver, albeit in very distinct ways. SREBP-1c, the SREBP isoform expressed at high levels in the liver (5), is a member of the basic helix-loop-helix (bHLH) leucine zipper family of transcription factors. This family of transcription factors was originally identified through its role in regulating cholesterol and fatty acid biosynthesis by sterols (6, 7). However, SREBP-1c was later found to be regulated by insulin signaling (8). SREBP-1c affects the expression of target genes by binding to sterol response elements, usually located near the promoter region of a target gene. In addition to the regulation of SREBP-1c gene expression by insulin, insulin appears to regulate the cleavage of SREBP-1c, which also directly affects the amount of SREBP-1c in the nucleus (9). SREBP-1 KO mice exhibit profound abnormalities in lipid metabolism (10).

ChREBP, another bHLH transcription factor, also plays a central role in the dietary regulation of hepatic gene expression. This factor was identified by its binding to carbohydrate response elements (ChOREs) in a variety of glucose-responsive, lipogenic genes, including those encoding acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS), and liver–pyruvate kinase (L-PK) (11, 12). Since the initial discovery of ChREBP, this protein has been found to play a central role in the regulation of hepatic genes by glucose. Indeed, mice without this transcription factor exhibit a marked shift in substrate utilization and ATP hydrolysis by the liver, thereby leading to a markedly hypoenergetic state (13).

Glucose metabolism, initiated by GK, leads to an increase in xyulose-5-phosphate, an intermediate in the hexose monophosphate shunt pathway. At low glucose concentrations, ChREBP is an inactive protein found mostly in the cytoplasm due to its being phosphorylated at Ser196 by protein kinase A. However, at high glucose concentrations, ChREBP becomes dephosphorylated, probably by protein phosphatase 2A (PP2A), thereby stimulating its translocation to the nucleus (14, 15) as shown in Figure 1.

Figure 1

Transcriptional regulation of hepatic lipogenesis by glucose and insulin. In this issue, Denechaud et al. (4) describe a series of studies that strongly challenge a recent report by Mitro et al. (3) suggesting that LXRα/β has a direct glucose-sensing function in liver. Denechaud et al. provide strong evidence for ChREBP, not the LXRs, in mediating the effect of glucose on genes in the liver (4). The regulation of hepatic lipogenesis involves several transcriptional regulatory networks. Three transcription factors, ChREBP, LXRα/β, and SREBP-1c, contribute to this regulation, but in an often incestuous manner. For instance, LXRα/β has been found to regulate the transcription of both ChREBP and SREBP-1c. ChREBP, LXRα/β, and SREBP-1c bind to ChOREs, LXR response elements (LXREs), and sterol response elements (SREs), respectively. Moreover, DNA binding by ChREBP and LXRα/β requires heterodimerization with Max-like protein (Mlx) and retinoid X receptor (RXR), respectively. Genes that regulate hepatic lipogenesis have been found to contain 1 or more ChOREs, LXREs, and/or SREs. Some genes, such as for FAS, are regulated by all 3 transcription factors. The authors demonstrate that glucose flux is necessary for regulation of hepatic genes by glucose and that phosphorylation of ChREBP at Ser196 is reciprocally regulated by glucose flux. ChREBP is located in the cytoplasm under low plasma glucose concentration due to the phosphorylation of Ser196. As the plasma glucose level rises, glucose enters hepatocytes and is rapidly equilibrated due to the low-K_m glucose transporter GLUT2. Glucose phosphorylation is then initiated by GK, leading to the formation of xyulose-5-phosphate (X-5-P) by the hexose monophosphate shunt pathway. Increased xyulose-5-phosphate in turn appears to cause dephosphorylation of Ser196 on ChREBP (probably through activation of PP2A), thereby allowing ChREBP to enter the nucleus and affect gene expression. Insulin appears to affect hepatic lipogenesis principally by increasing transcription of both ChREBP and SREBP-1c.

Do the LXRs, upstream modulators of ChREBP and SREBP1c, function as glucose sensors?

LXRα and LXRβ, closely related members of the nuclear hormone receptor superfamily, have also been found to contribute to the regulation of hepatic lipogenesis (16). While first classified as orphan nuclear receptors, both were discovered to bind oxysterols, metabolites of cholesterol, and to regulate genes critical to cholesterol efflux and the conversion of cholesterol to bile acids (17). Moreover, LXRα and LXRβ have recently been found to bind to regulatory elements in the genes encoding SREBP-1c and ChREBP (18), thereby positioning each of these transcription factors to regulate a wide number of genes, some of which encode enzymes involved in fatty acid metabolism (19). All of these responses are abolished in LXRα/β double-KO mice (19).

Recently, Mitro et al. reported that glucose is able to bind and stimulate the transcriptional activity of LXRs (3). This group reported that both d- and l-glucose, as well as glucose-6-phosphate (G-6-P), the initial metabolite of d-glucose, bind to and activate the LXRs, thereby leading to the activation of LXR target genes, including ChREBP, in HepG2 cells (3). On account of these data, Mitro et al. proposed that the LXRs function directly as glucose sensors in the liver. Because the LXRs lie upstream of both SREBP-1c and ChREBP in the glucose and insulin signaling pathways in liver, they are certainly positioned to be able to regulate many downstream genes (Figure 1). However, questions were raised immediately about this study (20), including the fact that the experiments were performed in HepG2 cells, a human hepatocellular carcinoma cell line that responds poorly, if at all, to glucose, and that there is no stereo-specific discrimination between d- and l-glucose. In the model proposed by Mitro et al., glucose would activate the expression of lipogenic genes by the activation of the LXRs. This in turn would increase the expression of both ChREBP and SREBP-1c. However, if this model were true, then the ability of glucose to induce the expression of lipogenic genes would be lost in LXRα/β double-KO mice.

New evidence against a glucose-sensing function by the LXRs

The current report by Denechaud et al. (4) provides new insights into the respective roles of ChREBP, SREBP-1c, and the LXRs in the signal transduction pathways involved in regulating lipid synthesis in the liver. The authors performed a series of experiments directed at clarifying the separate roles of ChREBP, SREBP-1c, and LXRα/β in the nutritional regulation of hepatic genes.

First, using C57BL/6J mice that were either fed or fasted, as well as primary hepatocytes from these mice, Denechaud et al. show that LXR activation by the synthetic agonist T0-901317 is not able to overcome the posttranslational modifications of ChREBP mediated by glucose metabolism and that glucose flux is necessary for the transcriptional activation and nuclear localization of ChREBP (4). These studies made use of a phosphospecific antibody developed by the authors that enabled the phosphorylation of ChREBP at Ser196 to be directly monitored. T0-901317, the LXRα/β-activating agonist, did not affect either the dephosphorylation or nuclear translocation of ChREBP nor the expression of the gene encoding L-PK, which is known to be directly regulated by ChREBP. Finally, the authors demonstrated that refeeding fasted LXRα/β double-KO mice with a high-carbohydrate diet increases ChREBP gene expression and the translocation of this protein into the nucleus.

Second, the authors compared the extent to which LXRs contribute to the regulation of 8 hepatic genes that play key roles in glycolysis and lipogenesis by comparing the effects of fasting and refeeding a high-carbohydrate diet in wild-type and LXRα/β double-KO mice (4). This experiment revealed that in LXRα/β double-KO mice there was a lack of induction of genes regulated principally by SREBP-1c and LXRα/β and that the high-carbohydrate diet–mediated induction of ChREBP expression and its translocation to the nuclei of hepatocytes was unaffected. In interpreting this experiment, the authors dutifully note that there are different regulational patterns. Specifically, L-PK and ACC are principally regulated by ChREBP and stearoyl-CoA desaturase 1 (SCD1) and glyceraldehyde 3-phosphate acyltransferase (GPAT) are under the transcriptional control of SREBP-1c and/or the LXRs, whereas FAS is synergistically regulated by all 3 factors.

Third, the authors examined the regulation by glucose of L-PK and ACC gene expression in hepatocytes from LXRα/β double-KO mice (4). These genes, both of which contain well-defined ChOREs, were adequately induced by the addition of glucose, a finding that is directly contrary to what would be expected if LXRα and LXRβ were acting as hepatic glucose sensors. However, the induction by glucose of both L-PK and ACC did not occur after the cells were treated with an siRNA-expressing recombinant adenovirus that knocked down ChREBP.

Fourth, Denechaud et al. used a fluorescence resonance energy transfer assay to test the ability of glucose and several glucose metabolites to induce the recruitment of several coactivator peptides by the LXR ligand-binding domain (4). This experiment provided another direct test of the model proposed by Mitro et al. (3), namely, that LXRα and LXRβ function as hepatic glucose sensors by recruiting transcriptional coactivators in response to increasing concentrations of glucose or its metabolites. While the LXR agonist T0-901317 induced the recruitment of cofactor peptides to both LXRα and LXRβ, as would be expected, neither glucose nor G-6-P, when added directly to this cell-free assay, had any effect. Moreover, by performing cotransfection experiments using chimeric LXRα/β-Gal4 DNA-binding proteins and a Ga14-responsive luciferase reporter in HepG2 cells (the same cell line used by Mitro et al.), Denechaud et al. showed that glucose had no potentiating effect (4).

When considered together, existing knowledge about the function of ChREBP in the liver and the results of the current study by Denechaud et al. strongly challenge the findings of Mitro et al. and simultaneously greatly strengthen the case for ChREBP being the principal determinant for the regulation of hepatic genes by glucose and dietary carbohydrate. Indeed, these new findings provide very convincing evidence that LXRα and LXRβ do not have a direct glucose-sensing function in the liver.

So, does metabolic flux really matter for the regulation of hepatic genes by glucose? The answer, as has been thought to be the case for decades, remains yes.

Where to from here?

While ChREBP, SREBP-1c, and LXRs all play integral roles in the regulation of lipogenesis by the liver, much remains to be learned about how they are regulated, especially by insulin and glucose. In particular, a great deal more remains to be determined about the regulation of ChREBP. Specifically, how does phosphorylation regulate the activity and nuclear transport of this protein; what protein kinases and phosphatases are involved; and do specific metabolites play a role? New and emerging technologies may make such determinations possible. Indeed, only then will our understanding of this intriguing aspect of liver function be complete.

Acknowledgments

The authors are supported by NIH grant DK60667 (to M. Shiota) and DK42502 and DK72473 (to M.A. Magnuson).

Address correspondence to: Mark A. Magnuson, Center for Stem Cell Biology, Vanderbilt University School of Medicine, 9475 MRB-IV, 2213 Garland Avenue, Nashville, Tennessee 37232, USA. Phone: (615) 322-7006; Fax: (615) 322-7236; E-mail: mark.magnuson@vanderbilt.edu.

Footnotes

Nonstandard abbreviations used: ACC, acetyl-CoA carboxylase; ChORE, carbohydrate response element; ChREBP, carbohydrate-responsive element–binding protein; FAS, fatty acid synthase; GK, glucokinase; G-6-P, glucose-6-phosphate; L-PK, liver–pyruvate kinase; LXR, liver X receptor.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.118:841–844 (2008). doi:10.1172/JCI35137.

See the related article at ChREBP, but not LXRs, is required for the induction of glucose-regulated genes in mouse liver.

References

1. Magnuson, M., and Matschinsky, F. 2004. Glucokinase as a glucose sensor: past, present and future. InGlucokinase and glycemic disease: from basics to novel therapeutics. Karger. Basel, Switzerland. 1–17.
2. Moore, B., et al. 2003. Role of the Arabidopsis glucose sensor HXK1 in nutrient, light, and hormonal signaling. Science. 300:332-336.
   View this article via: PubMed Google Scholar
3. Mitro, N., et al. 2007. The nuclear receptor LXR is a glucose sensor. Nature. 445:219-223.
   View this article via: PubMed Google Scholar
4. Denechaud, P.-D., et al. 2008. ChREBP, but not LXRs, is required for the induction of glucose-regulated genes in mouse liver. J. Clin. Invest. 118:956-964.
   View this article via: JCI Google Scholar
5. Shimomura, I., Shimano, H., Horton, J.D., Goldstein, J.L., Brown, M.S. 1997. Differential expression of exons 1a and 1c in mRNAs for sterol regulatory element binding protein-1 in human and mouse organs and cultured cells. J. Clin. Invest. 99:838-845.
   View this article via: JCI PubMed Google Scholar
6. Brown, M.S., Goldstein, J.L. 1997. The SREBP pathway: regulation of cholesterol metabolism by proteolysis of a membrane-bound transcription factor. Cell. 89:331-340.
   View this article via: PubMed Google Scholar
7. Brown, M.S., Goldstein, J.L. 1999. A proteolytic pathway that controls the cholesterol content of membranes, cells, and blood. Proc. Natl. Acad. Sci. U. S. A. 96:11041-11048.
   View this article via: PubMed Google Scholar
8. Foretz, M., et al. 1999. ADD1/SREBP-1c is required in the activation of hepatic lipogenic gene expression by glucose. Mol. Cell. Biol. 19:3760-3768.
   View this article via: PubMed Google Scholar
9. Hegarty, B.D., et al. 2005. Distinct roles of insulin and liver X receptor in the induction and cleavage of sterol regulatory element-binding protein-1c. Proc. Natl. Acad. Sci. U. S. A. 102:791-796.
   View this article via: PubMed Google Scholar
10. Shimano, H., et al. 1999. Sterol regulatory element-binding protein-1 as a key transcription factor for nutritional induction of lipogenic enzyme genes. J. Biol. Chem. 274:35832-35839.
    View this article via: PubMed Google Scholar
11. Bergot, M.O., Diaz-Guerra, M.J., Puzenat, N., Raymondjean, M., Kahn, A. 1992. Cis-regulation of the L-type pyruvate kinase gene promoter by glucose, insulin and cyclic AMP. Nucleic Acids Res. 20:1871-1877.
    View this article via: PubMed Google Scholar
12. Shih, H.M., Liu, Z., Towle, H.C. 1995. Two CACGTG motifs with proper spacing dictate the carbohydrate regulation of hepatic gene transcription. J. Biol. Chem. 270:21991-21997.
    View this article via: PubMed Google Scholar
13. Burgess, S.C., et al. 2008. Carbohydrate-response element-binding protein deletion alters substrate utilization producing an energy-deficient liver. J. Biol. Chem. 283:1670-1678.
    View this article via: PubMed Google Scholar
14. Nishimura, M., Fedorov, S., Uyeda, K. 1994. Glucose-stimulated synthesis of fructose 2,6-bisphosphate in rat liver. Dephosphorylation of fructose 6-phosphate, 2-kinase:fructose 2,6-bisphosphatase and activation by a sugar phosphate. J. Biol. Chem. 269:26100-26106.
    View this article via: PubMed Google Scholar
15. Kabashima, T., Kawaguchi, T., Wadzinski, B.E., Uyeda, K. 2003. Xylulose 5-phosphate mediates glucose-induced lipogenesis by xylulose 5-phosphate-activated protein phosphatase in rat liver. Proc. Natl. Acad. Sci. U. S. A. 100:5107-5112.
    View this article via: PubMed Google Scholar
16. Zelcer, N., Tontonoz, P. 2006. Liver X receptors as integrators of metabolic and inflammatory signaling. J. Clin. Invest. 116:607-614.
    View this article via: JCI PubMed Google Scholar
17. Repa, J.J., Mangelsdorf, D.J. 2002. The liver X receptor gene team: potential new players in atherosclerosis. Nat. Med. 8:1243-1248.
    View this article via: PubMed Google Scholar
18. Cha, J.Y., Repa, J.J. 2007. The liver X receptor (LXR) and hepatic lipogenesis. The carbohydrate-response element-binding protein is a target gene of LXR. J. Biol. Chem. 282:743-751.
    View this article via: PubMed Google Scholar
19. Tobin, K.A., et al. 2002. Liver X receptors as insulin-mediating factors in fatty acid and cholesterol biosynthesis. J. Biol. Chem. 277:10691-10697.
    View this article via: PubMed Google Scholar
20. Lazar, M.A., Willson, T.M. 2007. Sweet dreams for LXR. Cell Metab. 5:159-161.