Vanin-1 Is a Key Activator for Hepatic Gluconeogenesis

  1. Chang Liu1
  1. 1Jiangsu Key Laboratory for Molecular and Medical Biotechnology and College of Life Sciences, Nanjing Normal University, Nanjing, Jiangsu, China
  2. 2Department of Geriatrics, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu, China
  1. Corresponding author: Chang Liu, changliu{at}njnu.edu.cn.
  1. S.C. and W.Z. contributed equally to this work.

Abstract

Vanin-1 (VNN1) is a liver-enriched oxidative stress sensor that has been implicated in the regulation of multiple metabolic pathways. Clinical investigations indicated that the levels of VNN1 were increased in the urine and blood of diabetic patients, but the physiological significance of this phenomenon remains unknown. In this study, we demonstrated that the hepatic expression of VNN1 was induced in fasted mice or mice with insulin resistance. Gain- and loss-of-function studies indicated that VNN1 increased the expression of gluconeogenic genes and hepatic glucose output, which led to hyperglycemia. These effects of VNN1 on gluconeogenesis were mediated by the regulation of the Akt signaling pathway. Mechanistically, vnn1 transcription was activated by the synergistic interaction of peroxisome proliferator–activated receptor γ coactivator 1α (PGC-1α) and hepatocyte nuclear factor-4α (HNF-4α). A chromatin immunoprecipitation analysis indicated that PGC-1α was present near the HNF-4α binding site on the proximal vnn1 promoter and activated the chromatin structure. Taken together, our results suggest an important role for VNN1 in regulating hepatic gluconeogenesis. Therefore, VNN1 may serve as a potential therapeutic target for the treatment of metabolic diseases caused by overactivated gluconeogenesis.

Introduction

Hepatic gluconeogenesis plays a crucial role in maintaining blood glucose homeostasis in mammals. Gluconeogenesis is activated by fasting, which means that the synthesis and secretion of glucagon and glucocorticoids are induced to enhance hepatic glucose production via the stimulation of pepck and g6pase expression (1,2). In contrast, insulin suppresses gluconeogenesis via the lipid kinase phosphatidylinositol 3-kinase pathway and correspondingly activates glycogen synthesis when sufficient food is available (3). Thus, gluconeogenesis is finely regulated by positive and negative regulators to achieve a balance. However, gluconeogenesis is constitutively induced due to insulin resistance in certain pathophysiological conditions, such as type 2 diabetes, and leads to excess glucose secretion and aggravated hyperglycemia (4,5).

Peroxisome proliferator–activated receptor γ coactivator 1α (PGC-1α) is one of the most important metabolic regulators and has been shown to stimulate various physiological processes by selectively activating nuclear receptors and transcriptional factors (6). Of note, PGC-1α robustly regulates gluconeogenesis. PGC-1α is elevated in the liver at a fasted or insulin-resistant state when gluconeogenesis is activated (7). In contrast, the liver-specific knockdown of PGC-1α significantly improves glucose intolerance during hyperglycemia (8). At the molecular level, PGC-1α facilitates the transcriptional activity of the glucocorticoid receptors FOXO1 and hepatocyte nuclear factor-4α (HNF-4α) on the promoter of gluconeogenic genes (9,10). These findings indicate that PGC-1α is a critical factor in the regulation of gluconeogensis and glucose homeostasis.

Vanin-1 (VNN1) is a glycosylphosphatidyl inositol–anchored pantetheinase that is highly expressed in the liver, gut, and kidney (11). Its pantetheinase activity hydrolyzes pantetheine into pantotheic acid (vitamin B5) and cysteamine (12). Pantotheic acid acts as the structural component of CoA (13), which indicates that VNN1 may be involved in fatty acid metabolism. Conversely, cysteamine inhibits the γ-glutamylcysteine synthetase (γ-GCS)–mediated regulation of the endogenous GSH pool and therefore affects the cellular redox status. VNN1-deficient mice exhibit significantly increased resistance to stress, and their intestinal inflammation is reduced (1416). Importantly, the link between VNN1 and metabolic diseases has been recently revealed. For example, VNN1 protects islet β-cells from streptozotocin-induced injury and regulates the development of type 1 diabetes (17). In the liver, the mRNA expression levels of VNN1, as well as its activity, are induced upon fasting (18,19). A clinical study showed that the concentrations of VNN1 in pooled human urine distinguished diabetic patients with macroalbuminuria from those with normal albuminuria (20). In addition, the combination of VNN1 and MMP9 has been suggested as a novel blood biomarker panel to discriminate pancreatic cancer–associated diabetes from type 2 diabetes (21). However, these findings are descriptive, and the detailed mechanism through which VNN1 regulates energy metabolism remains unknown.

Based on the findings mentioned above, we aimed to explore whether VNN1 regulates hepatic gluconeogenesis and if this process requires PGC-1α participation. To answer this question, we used the gain-of-function and loss-of-function strategy to manipulate the VNN1 expression levels both in vivo and in vitro based on adenoviral transduction and small interfering RNA (siRNA) delivery. Our studies revealed an important role for VNN1 in the regulation of gluconeogenesis and suggest that VNN1 inhibition may be a therapeutic target for the treatment of metabolic disorders associated with overactivated gluconeogenesis.

Research Design and Methods

Male C57BL/6 mice and diabetic db/db mice on a C57BKS background were purchased from the Model Animal Research Center of Nanjing University (Nanjing, Jiangsu, China). All animal procedures in this investigation conform to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH publication 85-23, revised in 1996) and the approved regulations were set by the Laboratory Animal Care Committee at Nanjing Normal University (permit 2090658, issued 20 April 2008). For fasting-induced gluconeogenesis, the mice were subjected to 2, 5, and 16 h fasting or 16 h fasting followed by 20-h refeeding. To induce insulin resistance, 4-week-old mice were fed a high-fat diet (HFD, fat content 60%; Research Diets, New Brunswick, NJ) for 2 months (22). To specifically manipulate the target gene expression in liver, we transduced adenoviruses targeting genes or siRNA oligonucleotides against the target genes into mice through tail-vein injection. For these experiments, adenoviruses were concentrated and purified at 1.5 × 109 plaque-forming units. Three days after the injection, all mice were killed and the livers were collected. Wortmannin (Sigma-Aldrich) was intraperitoneally injected into mice at a dose of 1.5 mg/kg body weight 12 h before killing when necessary (23).

Pantetheinase Activity Assay

The substrate, pantothenate-AMC, was chemically synthesized using β-alanine 7-amido-4-methylcoumarin trifluoroacetic acid (TFA) salt (H-β-Ala-AMC.TFA, 36 mg, 1 eq) and R-(-)-pantolactone. We used pantothenate-AMC as substrate at 37°C for 30 min to assay the pantetheinase activity; the hydrolysis catalyzed by VNN1 yields pantothenic acid and detectable free fluorescent AMC (excitation 340 nm; emission 460 nm) (24). The liver extracts were washed three times with PBS and lysed with potassium phosphate buffer (100 mmol/L, pH 7.5) containing 0.1% Triton X-100 and 0.6% sulfosalicylic acid. The protein concentrations were determined with a bicinchoninic acid protein assay. The enzymatic assay was performed using 10 μg of liver extract in phosphate buffer (100 mmol/L potassium phosphate buffer, pH 7.5) containing 2 mmol/L pantothenate-AMC, 0.01% BSA, 0.5 mmol/L dithiothreitol, 5% DMSO, and 0.0025% Brij-35 in a total reaction mixture volume of 100 μL. The reactions were carried out at 37°C in the presence or absence of liver extract, and the fluorescence (excitation 350 nm; emission 460 nm) was recorded every 2 min, with the change in fluorescence measured over a 30-min period. A standard curve was generated using purified recombinant VNN1 (Sino Biological, Daxing, China) at the same buffer conditions described above. The VNN1 activity was normalized for total protein content. The pantetheinase activity was calculated by determining the slope at 30 min from fitting the data to the standard curve and normalizing it for total protein content.

Statistical Analysis

The groups of data are presented as the mean ± SD. The data were analyzed using a one-way ANOVA followed by Fisher LSD post hoc test. The calculations were performed using the SPSS for Windows version 12.5S statistical package (SPSS, Chicago, IL). A value of P < 0.05 was considered statistically significant.

Results

Gluconeogenic Signals Induce Hepatic VNN1 Expression

Fasting-refeeding cycles are robust nutritional signals to regulate gluconeogensis. As expected, the hepatic mRNA expression of pgc-1α, an important metabolic regulator, was induced by fasting and peaked at 5 h (Fig. 1A). Subsequently, its expression gradually declined. Two gluconeogenic genes (pepck and g6pase) were induced after 2 h of fasting, and this expression persisted during the entire starvation period. Interestingly, the hepatic expression of vnn1 mRNA was also induced in mice subjected to fasting and decreased upon refeeding at 20 h. The protein expression levels of VNN1 in the liver showed a similar pattern (Fig. 1B and Supplementary Figure 1A). Conversely, diabetic db/db mice and HFD feeding–induced obese mice developed insulin resistance and showed overactivated hepatic gluconeogenesis. We found that the VNN1 mRNA and protein expression levels were increased in livers of these mice (Fig. 1CE and Supplementary Fig. 1B and C). Furthermore, VNN1 is a pantetheianse, and the biological functions of VNN1 are thought to depend on its pantetheianse enzymatic activity. In our settings, the pantetheianse activity of VNN1 increased in the liver homogenates of fasted and db/db mice (Fig. 1F and G), which is consistent with the mRNA accumulation of VNN1.

Figure 1

Gluconeogenic signals induce hepatic VNN1 expression. RT-qPCR (A) and Western blot (B) analysis of VNN1 and gluconeogenic gene expression in the liver from mice subjected to fasting (2, 5, and 16 h) or fasting 16 h followed by 20-h refeeding. *P < 0.05 and **P < 0.01 vs. refed group, n = 6. RT-qPCR (C) and Western blot (D) analysis in the liver from wild-type or db/db mice. **P < 0.01 vs. WT, n = 5. E: mRNA and protein analysis in the liver from mice fed normal diet (ND) or HFD for 2 months. **P < 0.01 vs. ND, n = 6. F: Pantetheinase activity analysis in the liver homogenates from mice described in A. **P < 0.01 vs. refed group, n = 6. G: Pantetheinase activity analysis in the liver from wild-type or db/db mice. **P < 0.01 vs. WT, n = 5. H: Regulation of VNN1 expression by 8-Br cAMP in vitro. AML-12 mouse hepatocytes were treated with 1 mmol/L 8-Br cAMP for an indicated time period, and VNN1 expression was quantified by RT-qPCR (top) and Western blot (bottom) analysis. **P < 0.01 vs. 0 h. I: Reporter gene analysis of transcriptional activity of mouse vnn1 promoter. mvnn1-luc was transfected into AML-12 cells. Thirty-six hours later, cells were treated with 8-Br cAMP (1 mmol/L) for another 12 h. **P < 0.01 vs. mvnn1-luc alone. J: Western blot analysis of VNN1 expression in the primary hepatocytes treated with dexamethasone (Dex, 1 μmol/L), 8-Br cAMP (1 mmol/L), or hydrocortisone (HCS, 1 μmol/L) for 6 h. All the data are represented as the mean ± SD. CTL, control; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

In vitro studies indicated that 8-Br cAMP, an analog of the natural signal molecule cAMP that activates gluconeogenesis, stimulated vnn1 expression in a time-dependent manner in AML-12 mouse hepatocytes (Fig. 1H). 8-Br cAMP also increased the transcriptional activity of the vnn1 promoter (Fig. 1I). In addition to cAMP, fasting-induced hormones, such as glucocorticoids, can robustly activate gluconeogenesis. However, we found that 8-Br cAMP, but not glucocorticoids (dexamethasone and hydrocortisone), induced VNN1 protein expression in mouse primary hepatocytes (Fig. 1J). These data suggested that cAMP is one of the major players to activate VNN1 expression, whereas glucocorticoids have a modest effect on VNN1 expression.

VNN3 is another member of the pantetheinase family. In contrast to the robust induction of VNN1 in gluconeogenic settings, the hepatic expression of vnn3 remained stable in fasted mice and was only moderately increased (2.1-fold) in db/db mice (Supplementary Fig. 2A and B). Functionally, the knockdown of VNN3 in the livers of db/db mice (Supplementary Fig. 2C) did not improve the glucose intolerance (Supplementary Fig. 2D) and insulin resistance (Supplementary Fig. 2E) of these mice. The hepatic expression levels of gluconeogenic genes also remained unchanged (Supplementary Fig. 2F and G). Taken together, these results suggested that VNN3 does not regulate gluconeogenesis.

VNN1 Activates Hepatic Gluconeogenesis

As shown in Fig. 2A and B, the forced expression of exogenous VNN1 in AML-12 cells by adenovirus infection induced the mRNA and protein expression levels of gluconeogenic genes in a dose-dependent manner. In contrast, VNN1 overexpression did not alter the mRNA expression levels of baf60a (a subunit of SWI/SNF chromatin remodeling complex), cyp7a1 (a regulator involved in bile acid metabolism), and fas (fatty acid synthase). Functionally, the overexpression of VNN1 in mouse primary hepatocytes increased the glucose production rate (Fig. 2C). In vivo studies indicated that the basal blood glucose levels during starvation were higher in mice that specifically overexpressed VNN1 in the liver (Supplementary Table 1). These mice also showed impaired glucose tolerance and insulin sensitivity (Fig. 2D and E). At the molecular level, the hepatic expression levels of gluconeogenic genes were increased in response to VNN1 overexpression (Fig. 2F and G), whereas baf60a, cyp7a1, and fas remained unchanged. These data suggest that the impact of VNN1 is specific to the regulation of gluconeogenesis and is not a general disturbance of the hepatic metabolism. Finally, the pantetheinase activity of VNN1 in the liver homogenates of these mice significantly increased (Fig. 2H). Taken together, our results strongly suggested that VNN1 activates gluconeogenesis.

Figure 2

VNN1 promotes hepatic gluconeogenesis. A and B: Regulation of gluconeogenic gene expression by VNN1 in vitro. RT-qPCR (A) and Western blot analysis (B) of VNN1 and gluconeogenic gene expression in AML-12 cells transduced with adenoviruses encoding GFP (Ad-GFP, as control) or VNN1 (Ad-VNN1) for 48 h. **P < 0.01 vs. Ad-GFP. C: Glucose output assay in mouse primary hepatocytes. Cells were infected by adenoviruses for 48 h, and glucose production rate was assessed. 8-Br cAMP (1 mmol/L, treatment time 12 h) was used as a positive control. **P < 0.01 vs. Ad-GFP. D: Glucose tolerance test. Mice were transduced with adenoviruses encoding GFP or VNN1 through tail-vein injection (0.1 absorbance units per mouse). Three days later, mice were fasted for 16 h and the glucose tolerance test assay was performed. **P < 0.01 vs. Ad-GFP, n = 5. E: Insulin tolerance test. Mice were treated as in D. The fasting time was shortened to 6 h and the insulin tolerance test assay was performed. *P < 0.05 and **P < 0.01 vs. Ad-GFP group, n = 5. F and G: VNN1 and gluconeogenic gene expression in the liver. RT-qPCR (F) and Western blot (G) analysis in liver of mice treated as in D. *P < 0.05 and **P < 0.01 vs. Ad-GFP group, n = 5. H: Pantetheinase activity analysis in the liver homogenates from mice treated as in D. **P < 0.01 vs. Ad-GFP group, n = 5. All the data are represented as the mean ± SD. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Knockdown of VNN1 Ameliorates Hyperglycemia

We next used a loss-of-function strategy to confirm the above findings. The liver-specific knockdown of VNN1 led to improved glucose tolerance in normal fasted mice (Fig. 3A). Consistently, the expression of hepatic gluconeogenic genes was suppressed upon the depletion of VNN1 (Fig. 3B and C). In a pathophysiological setting, hepatic VNN1 knockdown reduced the food and water intake of db/db mice (Fig. 3D). This treatment also relieved glucose intolerance and insulin resistance in animals (Fig. 3E and F). Accordingly, the mRNA and protein expression levels of gluconeogenic genes were repressed (Fig. 3G and H). Finally, the pantetheinase activity of VNN1 in the liver homogenates of these mice decreased (Fig. 3I). In agreement with these results, VNN1 knockdown in AML-12 cells repressed the 8-Br cAMP–induced increase of the transcriptional activity of gluconeogenic gene promoters (Supplementary Fig. 3A). VNN1 knockdown also inhibited gluconeogenic gene expression in AML-12 cells (Supplementary Fig. 3B and C) and decreased the glucose production rate in mouse primary hepatocytes (Supplementary Fig. 3D) upon 8-Br cAMP challenge.

Figure 3

Liver-specific knockdown of VNN1 relieves hyperglycemia. A: Glucose tolerance test analysis. Normal mice were transduced with scrambled siRNA oligonucleotides (as control) or siRNA oligonucleotides against VNN1 through tail-vein injection (2.5 mg/kg). Three days later, mice were fasted for 16 h and the glucose tolerance test assay was performed. scra, scrambled siRNA; siVNN1, VNN1 siRNA. B: RT-qPCR analysis of vnn1 and gluconeogenic gene expression in the liver from mice treated as in A. C: Western blot analysis. D: Analysis of food intake (g/mouse/day) and water drinking (mL/mouse/day) in db/db mice transduced with scrambled or VNN1 siRNA oligonucleotides for 3 days. E: Glucose tolerance test analysis. F: Insulin tolerance test analysis. G: RT-qPCR analysis of vnn1 and gluconeogenic gene expression in the liver from db/db mice. H: Western blot analysis. I: Pantetheinase activity analysis. For all the panels, *P < 0.05 and **P < 0.01 vs. scrambled siRNA. Each group included at least six mice. All the data are represented as the mean ± SD. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Akt Signal Pathway Mediates VNN1 Regulation of Gluconeogenesis

The insulin-Akt signaling axis plays an important role in the regulation of gluconeogenesis. Therefore, we investigated whether this signaling cascade is involved in mediating VNN1’s effects on glucose control. As shown in Fig. 4A and B, the overexpression of VNN1 reduced the expression levels of phosphorylated Akt both in vivo and in vitro. Conversely, the liver-specific knockdown of VNN1 in db/db mice resulted in a marked increase in phospho-Akt levels (Fig. 4C). Similarly, the transfection of VNN1 siRNA into AML-12 cells increased the basal level and showed synergistic effects with insulin on Akt phosphorylation (Fig. 4D). To further prove the causal relationship between the Akt signaling pathway and VNN1 regulation, we used wortmannin (a specific phosphatidylinositol 3-kinase/Akt inhibitor) to treat db/db mice with liver-specific VNN1 knockdown. Our data demonstrated that wortmannin alone further increased the basal glucose level and worsened impaired glucose tolerance in db/db mice. Additionally, the beneficial effects of VNN1 knockdown on glucose intolerance and insulin resistance were attenuated when mice were subsequently treated with wortmannin (Fig. 4E and F). These results strongly suggest that the Akt signaling pathway is required to regulate VNN1 in hepatic gluconeogensis. Because VNN1 is a glycosylphosphatidyl inositol–anchored enzyme, identifying the molecule that mediates the signal cascade from VNN1 to the insulin-Akt signaling pathway is of interest. Previous studies demonstrated that VNN1 prevents peroxisome proliferator–activated receptor γ (PPARγ) nuclear translocation and antagonizes its transcriptional activity in epithelial cells (25). We also observed similar phenomena in our system. Confocal imaging and the Western blotting of nuclear extracts indicated that VNN1 knockdown led to an increase in PPARγ translocation from the cytoplasm to the nuclei in AML-12 cells (Supplementary Fig. 4A and B). Given that PPARγ activates the insulin-Akt signaling pathway (26) and inhibits gluconeogenesis (27), PPARγ may act as a mediator of signal transduction from VNN1 to the Akt pathway.

Figure 4

Akt signaling pathway mediates VNN1’s regulation of gluconeogenesis. A: Hepatic Akt phosphorylation levels in mice with liver-specific overexpression of VNN1. Top: Western blot analysis of phospho-Akt levels. Bottom: Quantitative results. **P < 0.01 vs. Ad-GFP, n = 5. B: Akt phosphorylation levels in AML-12 cells. Cells were infected by adenoviruses encoding GFP or VNN1 for 48 h and then treated with 100 nmol/L insulin for 0, 5, 10, and 15 min. **P < 0.01 vs. 0 min; #P < 0.05 and ##P < 0.01 vs. Ad-GFP at each time point. C: Hepatic Akt phosphorylation levels in db/db mice with liver-specific knockdown of VNN1. *P < 0.05 vs. scrambled siRNA, n = 5. D: Akt phosphorylation levels in AML-12 cells. Cells were transfected with scrambled or VNN1 siRNA oligonucleotides for 48 h and then treated with 100 nmol/L insulin for 0, 5, 10, and 15 min. **P < 0.01 vs. 0 min; #P < 0.05 and ##P < 0.01 vs. scrambled siRNA at each time point. E: Glucose tolerance test analysis. Scrambled or VNN1 siRNA oligonucleotides were transduced into db/db mice as described above, and then wortmannin (1.5 mg/kg) was intraperitoneally injected. Twelve hours later, mice were fasted for 16 h and glucose tolerance test analysis was performed. F: RT-qPCR (top) and Western blot (bottom) analysis in the liver samples from db/db mice treated as above. For E and F, +P < 0.05 and ++P < 0.01, *P < 0.05 and **P < 0.01 vs. db/db plus scrambled siRNA group; #P < 0.05 and ##P < 0.01 vs. db/db plus VNN1 siRNA treatment. n = 5. All the data are represented as the mean ± SD. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; INS, insulin; scra, scrambled siRNA.

PGC-1α and HNF-4α Synergistically Activate VNN1 Transcription

PGC-1α is a key nuclear coactivator of gluconeogensis and may serve as an upstream regulator for vnn1 transcription. Indeed, PGC-1α overexpression increased the VNN1 expression levels both in vivo (Fig. 5A) and in vitro (Supplementary Fig. 5A and B). In contrast, PGC-1α knockdown inhibited VNN1 expression in the liver (Fig. 5B) and in cultured cells (Supplementary Fig. 5C and D). Epigenetically, histone hyperacetylation is associated with transcriptional activation. Conversely, H3K9me2 is typically found in heterochromatin and silenced genes. Remarkably, we found that PGC-1α overexpression led to a robust increase in AcH3 levels with a corresponding decrease in H3K9me2 levels on the vnn1 promoter (Fig. 5C). The knockdown of PGC-1α caused opposite results (Fig. 5D). These results indicate that PGC-1α stimulates vnn1 transcription by altering the local chromatin environment from a repressive to an active state. We next attempted to identify the transcriptional factors that mediate the activation of vnn1 transcription by PGC-1α. A bioinformatics analysis indicated that the vnn1 promoter (−267 to +1) possessed two HNF-4α binding sites (nrmotif.ucr.edu/NRBSScan/H4SBM.htm) (28). Reporter gene assays demonstrated that PGC-1α augmented the transcriptional activity of HNF-4α on a vnn1 promoter reporter (Fig. 5E). However, when the HNF-4α binding sites were mutated, the synergistic effects of PGC-1α and HNF-4α on the transcription of the vnn1 promoter disappeared (Fig. 5E and Supplementary Fig. 6). To confirm that HNF-4α is required by PGC-1α to activate VNN1, we knocked down HNF-4α in AML-12 cells and found that HNF-4α knockdown blocked the PGC-1α–induced epigenetic changes of the vnn1 promoter toward activation (Fig. 5F). Consistently, the activation of vnn1 transcription and translation by PGC-1α was repressed by HNF-4α siRNA (Fig. 5G). Lastly, we explored whether VNN1 acts as a downstream effector for PGC-1α–controlled gluconeogenesis. We knocked down VNN1 in PGC-1α–overexpressed AML-12 cells and found that VNN1 knockdown inhibited the positive effects of PGC-1α on gluconeogenic gene expression (Fig. 5H). Moreover, the glucose output of mouse primary hepatocytes was decreased by VNN1 knockdown (Fig. 5I).

Figure 5

PGC-1α and HNF-4α activate vnn1 transcription. A: Gene expression levels in mice with liver-specific overexpression of PGC-1α. Mice were transduced with adenoviruses encoding GFP or PGC-1α through tail-vein injection (0.1 absorbance units per mouse). Three days later, the hepatic expression levels of VNN1 and gluconeogenic genes were assessed by RT-qPCR (top) and Western blot (bottom) analysis. *P < 0.05 and **P < 0.01 vs. Ad-GFP, n = 5. B: Gene expression levels in mice with liver-specific knockdown of PGC-1α. Mice were transduced with adenoviruses encoding scrambled siRNA (as control) or siRNA against PGC-1α through tail-vein injection. Three days later, mice were subjected to 16-h fasting followed by RT-qPCR (top) and Western blot (bottom) analysis. **P < 0.01 vs. scrambled siRNA, n = 5. C: Chromatin immunoprecipitation assays with indicated antibodies using AML-12 cells infected with adenoviruses encoding GFP or PGC-1α. The enrichment was quantified by qPCR analysis. *P < 0.05 and **P < 0.01 vs. Ad-GFP. D: Chromatin immunoprecipitation assays in AML-12 cells. Cells were infected with adenoviruses encoding scrambled siRNA or siRNA against PGC-1α for 36 h and then treated with 8-Br cAMP for another 12 h. *P < 0.05 and **P < 0.01 vs. 8-Br cAMP group. E: Reporter gene assays in AML-12 cells transfected with indicated plasmids. **P < 0.01 vs. the basal levels; ##P < 0.01 vs. cotransfection of PGC-1α and HNF-4α. F: Chromatin immunoprecipitation assays in AML-12 cells. The cells were transfected with scrambled or HNF-4α siRNA oligonucleotides for 12 h and then infected by adenoviruses encoding GFP or PGC-1α for another 36 h. **P < 0.01 vs. scrambled siRNA plus Ad-PGC-1α. G: RT-qPCR (top) and Western blot (bottom) analysis of PGC-1α, HNF-4α, and VNN1 expression in AML-12 cells treated as in F. **P < 0.01 vs. the basal levels; ##P < 0.01 vs. scrambled siRNA plus Ad-PGC-1α. H: RT-qPCR (top) and Western blot (bottom) analysis of gluconeogenic gene expression. AML-12 cells were transfected with scrambled or VNN1 siRNA oligonucleotides for 12 h and then infected by adenoviruses encoding GFP or PGC-1α for another 36 h. I: Glucose output assay. For H and I, *P < 0.05 and **P < 0.01 vs. the basal levels; #P < 0.05 and ##P < 0.01 vs. scrambled siRNA plus Ad-PGC-1α. All the data are represented as the mean ± SD. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; scra, scrambled siRNA.

DISCUSSION

VNN1 is assumed to primarily function as a pantetheinase, which hydrolyzes pantetheine into pantothenic acid (also called vitamin B5 or pantothenate) and cysteamine (12). Recent studies have revealed additional functions of VNN1. For example, VNN1-deficient mice are resistant to intestinal inflammation, oxidative stress, and experimental colitis (15,16). VNN1 also plays a critical role in malaria susceptibility, psoriasis, carcinogenesis, and cardiovascular disease (2932). Importantly, the vnn1 gene is one of the major targets of PPARα in the mouse liver, which strongly implies that VNN1 is involved in the regulation of energy metabolism (18). In the current study, we characterized VNN1 as a novel factor that activates hepatic gluconeogenesis. This characterization was initially based on the induction of VNN1 at the transcriptional level during fasting. To unmask the impact of VNN1 during fasting glucose homeostasis in vivo, we adopted VNN1 gain- or loss-of-function manipulation methods in mouse models. Our data suggested that VNN1 activates gluconeogenesis and increases glucose output under the control of the PGC-1α/HNF-4α complex, which is mediated by the Akt signaling pathway (Fig. 6).

Figure 6

A model illustrating the physiological functions of VNN1 in various tissues, highlighting the role of VNN1 in the regulation of hepatic gluconeogenesis in a PGC-1α–dependent manner.

In addition to its role in glucose homeostasis, VNN1 also affects lipid metabolism. Hepatic VNN1 expression is induced by the agonists of PPARα, a key transcriptional factor involved in fatty acid oxidation (18,33). The oral administration of triglycerides or PPARα ligands, such as WY14643 and fenofibrate, leads to a robust increase of VNN1 expression in the mouse liver (18). In agreement with previous findings, our study showed that hepatic VNN1 expression was induced in db/db and diet-induced obese mice, which exhibited severe hepatic steatosis. More importantly, the liver-specific knockdown of VNN1 improved the hepatic steatosis in these animal models (S.C., W.Z., J. Qian, and C.L., unpublished data). These findings suggested that VNN1 may be a culprit in the pathogenesis of lipid dysregulation and accumulation. Future studies are required to fully elucidate the role of VNN1 in lipid metabolism, including fatty acid oxidation, lipid synthesis, and the mobilization and/or conversion of triglycerides, fatty acids, and cholesterol.

The recognition of the physiological functions of VNN1 has been significantly extended in recent years. Due to its pantetheinase activity, VNN1 is actively involved in the progression of inflammatory reactions by generating cysteamine. Therefore, it plays important roles in the pathogenesis of diseases related to inflammation. For example, VNN1 protects pancreatic islets from streptozotocin-induced cell death and retards the development of type 1 diabetes (17). In the intestine, VNN1 promotes an inflammatory reaction and intestinal injury by decreasing the activity of γ-glutamylcysteine synthetase and reducing the stores of reduced glutathione (15,16). Finally, because cysteamine is a transglutaminase inhibitor (34,35), VNN1 is implicated in the development of Huntington disease, which is characterized by the accumulation of highly cross-linked insoluble proteins. These proteins may be associated with an upregulation of transglutaminase (summarized in Fig. 6). Conversely, recent studies indicate that glucose homeostasis is modulated by the chronic inflammation associated with metabolic stress (36,37). Genetic models in mice have demonstrated that the deletion of key inflammatory mediators improves glucose tolerance in obesity-induced insulin resistance (38). The detrimental effects of proinflammatory pathways on glucose homeostasis are partly achieved via the inhibitory serine phosphorylation of insulin receptor substrate 1 by Jun NH2-terminal kinase. In turn, uncontrolled hyperglycemia could further contribute to chronic inflammation. For example, the activity of nuclear factor-κB is increased via the modification of O-GlcNAc under high-glucose conditions (39). In our study, we found that the pantetheinase activity of VNN1 was increased in liver homogenates from db/db mice, which have chronic inflammation. Therefore, exploring the role of VNN1 pantetheinase activity in the control of gluconeogenesis is of particular interest to future studies to integrate inflammation and glucose homeostasis via VNN1.

VNN1 knockout (KO) mice have been generated for >10 years and serve as a useful tool to elucidate the physiological functions of VNN1. However, Roisin-Bouffay et al. (17) reported that the basal glucose levels and glucose tolerance do not change in VNN1 KO mice, which differs from our results. This inconsistency may be due to the microenvironment difference between whole-body KO and tissue-specific knockdown mice, which has been observed in many other studies. For example, Bmal1 is a core component of the mammalian circadian clock machinery. Investigators reported that global bmal1-null mice exhibit impaired glucose tolerance. In contrast, liver-specific bmal1 KO mice show increased glucose tolerance (40). Because we mainly focused on the physiological function of VNN1 in the liver, we believe that our animal models are more straightforward and specific to address our question.

PGC-1α is intensively involved in the regulation of gluconeogenesis. In adults, starvation induces PGC-1α expression in the liver via glucagon and glucocorticoid signaling. Once activated, PGC-1α initiates hepatic gluconeogenesis by coactivating key transcription factors, such as HNF-4α, glucocorticoid receptors, and FOXO1 (9,10). In accordance with these observations, PGC-1α KO mice and RNA interference–mediated liver-specific PGC-1α knockdown mice display impaired gluconeogenic gene expression and hepatic glucose production (8,41). These mice develop hypoglycemia upon fasting. The strategy to target PGC-1α to regulate glucose homeostasis has been proposed; however, its specificity and safety cannot be guaranteed because PGC-1α is a versatile molecule. In contrast, the biological functions of VNN1 are relatively focused. In addition, it serves as a downstream effector of PGC-1α and is required for the action of PGC-1α in gluconeogenesis. Indeed, PGC-1α expression peaked after 5 h of fasting in our system. The expression of VNN1 increased later and peaked after 16 h of fasting. In addition, the fact that the expression levels of PGC-1α and VNN1 peak at different time points suggests that these two factors may regulate gluconeogenesis at different stages of fasting. In the early stage of fasting when PGC-1α is robustly induced, gluconeogenesis is mainly orchestrated by PGC-1α. When fasting persists and PGC-1α expression levels decline, VNN1 will relay and perpetuate gluconeogenesis. Conversely because it is a membrane protein, VNN1 is more easily manipulated than PGC-1α, which localizes in the nuclei. Collectively, aiming at VNN1, the downstream target of PGC-1α, without disturbing PGC-1α itself, can lead to a more specific regulation of body glucose levels.

Inflammation, as well as oxidative stress, makes important contributions to the pathogenesis of metabolic diseases (37). Obese patients suffer from chronic low-grade inflammation, as indicated by increased plasma levels of C-reactive protein, inflammatory cytokines, and multifunctional proteins, such as leptin and osteopontin (42). Inflammation is also prominent in diabetic elderly patients (43). Inflammatory mediators are known to affect liver function. For instance, tumor necrosis factor-α stimulates hepatic lipogenesis and promotes hyperlipidemia (44). Conversely, reactive oxygen species are pro-oxidant factors that are generated during diabetes development, and lipid peroxide-mediated tissue damage has been observed in both type 1 and type 2 diabetes (45). In detail, diabetes-associated free radical injury, i.e., the accumulation of lipid peroxidation products, depletion of GSH, decreases in the GSH/GSSG ratio, and downregulation of key antioxidant enzymes, has been detected in various tissues, including the liver, with deleterious consequences on hepatic functions and cognitive impairments observed in different diabetes animal models. VNN1 plays a critical role in the regulation of inflammation and oxidative stress. VNN1-deficient mice display downregulated intestinal inflammation and a tissue resistance to oxidative stress (15,16). This protection is a result of the enhanced γ-glutamylcysteine synthetase activity in liver, which is due to the absence of cysteamine and leads to elevated stores of glutathione, the most potent cellular antioxidant (14). Based on these findings, VNN1 could serve as one of the nodes to interconnect inflammation, oxidative stress, and metabolic dysregulation. If so, the induction of PGC-1α, an upstream regulator of VNN1, would provide cell-autonomous protection to antagonize the deleterious effects of VNN1 because PGC-1α is a master player to evoke antioxidant defense.

Akt is a crucial regulator of glucose transportation, glycolysis, glycogen synthesis, and gluconeogenesis. Impaired Akt activation is associated with abnormal gluconeogenesis and glucose intolerance (46). In contrast, the restoration of insulin-induced Akt phosphorylation improves insulin sensitivity and glucose tolerance in mice that are fed an HFD (47). In the current study, we found that VNN1 overexpression blocked insulin-induced Akt phosphorylation in AML-12 cells, whereas VNN1 knockdown restored Akt phosphorylation in db/db mice. More importantly, the Akt inhibitor wortmannin attenuated VNN1 knockdown–induced inhibition in gluconeogenic gene expression. This finding provided correlative evidence that Akt inactivation may be involved in VNN1 signaling and its effects on glucose homeostasis and insulin sensitivity. The role of VNN1 in the regulation of other components involved in the insulin signaling pathway is interesting to study, such as insulin receptor substrate 1/2.

Histone undergoes extensive posttranslational modifications in response to external signals, including acetylation, methylation, phosphorylation, and SUMOylation (48,49). These modifications allow the fine tuning of the chromatin structure and gene expression in a context-dependent manner. To date, little is known about the epigenetic regulation of the vnn1 promoter. Our findings showed that PGC-1α altered the local chromatin environment of the HNF-4α binding site on the vnn1 promoter from a repressive to an active state. Consistently, PGC-1α positively regulates vnn1 gene expression. Our findings extend the current recognition of the epigenetic regulation of gluconeogenesis via VNN1 and highlight the importance of PGC-1α in this process.

In summary, our results strongly implicate VNN1 as a central player that orchestrates gluconeogenic programs. The identification of the specific role of VNN1 in hepatic gluconeogenesis would expand our knowledge to understand the important metabolic coordination necessary for proper blood glucose control. Conversely, VNN1 may serve as a pharmaceutical target to treat metabolic diseases with overactivated gluconeogenesis. In fact, a recent report demonstrated the design, synthesis, and characterization of a novel pantetheine analog, RR6, which acts as a selective, reversible, and competitive VNN1 inhibitor at nanomolar concentrations. The oral administration of RR6 in rats completely inhibits plasma VNN1 activity and alters the plasma lipid profile (50). More specific inhibitors of VNN1 will most likely be synthesized to attenuate obesity-related hyperglycemia.

Article Information

Acknowledgments. The authors thank Dr. Jiandie Lin (Life Sciences Institute, University of Michigan, Ann Arbor, MI) for providing adenoviruses (Ad-GFP, Ad-PGC-1α, Ad-Scrambled, and Ad-PGC-1α RNAi) and plasmids encoding PGC-1α, HNF-4α, mpepck-luc, and mg6pase-luc promoters.

Funding. This work was supported by grants from the National Basic Research Program of China (973 Program) (2012CB947600 and 2013CB911600), the National Natural Science Foundation of China (31171137 and 31271261), the Program for New Century Excellent Talents in University by the Chinese Ministry of Education (NCET-11-0990), the Program for the Top Young Talents by the Organization Department of the CPC Central Committee, the Collaborative Innovation Center for Cardiovascular Disease Translational Medicine (Nanjing Medical University), and the Priority Academic Program Development of Jiangsu Higher Education Institutions.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. S.C. designed and performed research and analyzed data. W.Z. performed research and analyzed data. C.T. and X.T. performed research. L.L. provided the substrate, pantothenate-AMC, for vanin-1 activity analysis. C.L. designed research, analyzed data, and wrote the manuscript. C.L. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

  • Received May 17, 2013.
  • Accepted February 13, 2014.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.

References

| Table of Contents

This Article

  1. Diabetes vol. 63 no. 6 2073-2085
  1. Supplementary Data
  2. All Versions of this Article:
    1. db13-0788v1
    2. 63/6/2073 most recent