Switch Of Glycolysis To Gluconeogenesis Against Hepatocarcinoma

Objective:Gluconeogenesis by which glucose is biosynthesized leading to a continuous glucose supply to vital organs is a fundamental feature of normal hepatocytes. Whether this gluconeogenic activity is also present in malignant hepatocytes remains largely unexplored, despite that an answer may give rise to novel therapeutic strategies to target glycolysis, one hallmark of malignant cells.Methods:(1) The expressions of PEPCK, G6Pase and11β-HSD1,11β-HSD2in human hepatocarcinoma were analyzed by real-time PCR and immunohistochemisty. To further clarify the exact situation of11β-HSD1/11β-HSD2in hepatocarcinoma patients,58human hepatocarcinoma specimens were immunohistochemically analyzed and the relative expressions of11β-HSDl and11β-HSD2were quantified. Then, the ratios of11β-HSDl/11β-HSD2were calculated and the relationship between11β-HSD1/11β-HSD2and patients’ survival time and tumor recurrence was analyzed by the Kaplan-Meier survival method.(2) The expressions of11β-HSD1,11β-HSD2and PEPCK, G6Pase in murine hepatocarcinoma were analyzed by RT-PCR, Western blot and immunohistochemisty. Then,11β-HSDl-overexpression or11β-HSD2-knockdown H22was constructed and these engineered H22tumor cell lines were inoculated into the mice, the growth of tumor was monitored. To further investigate the effect of gluconeogenesis on hepatocarcinoma, the PEPCK and G6Pase expressions in11β-HSD1-overexpression or11β-HSD2-knockdown H22tumor were analyzed by RT-PCR.(3) In vitro, H22cells were treated with0,0.1,1or10μM dexamethasone for7days. Then, the expressions of PEPCK and G6Pase were determined by RT-PCR and Western blot and cellular glucose was measured with glucose assay kit. In vivo, The BALB/c mice were subcutaneously injected with3x105H22cells for4days, and then treated with the intraperitoneal injections of different concentrations of dexamethasone (1.25,2.5and5μg/g) or saline once per day for16days. The growth of tumor was monitored.16days later, the mice were sacrificed and hepatocarcinoma was separated for PEPCK and G6Pase expression analysis and tissue glucose was measured with glucose assay kit. To exclude the influence of different inoculation sites on dexamethasone treatment, BALB/c mice liver were inoculated with H22cells and treated with0μg/g,1.25μg/g,2.5μg/g or5μg/g dexamethasone.16days later, the mice were sacrificed and livers were separated and analyzed. To further investigated the influence of dexamethasone on non-liver cancers,3×105non-liver derived melanoma B16tumor cells were subcutaneously injected into C57BL/6mice for4days, and then treating them with different concentrations of dexamethasone, the growth of tumor was monitored. To confirm that the antitumor effect of dexamethasone is via the gluconeogenetic pathway,3-MPA, the PEPCK-selective inhibitor was used and growth of tumor was monitored. To further strengthen dexamethasone as a potential agent in the treatment of hepatocarcinoma, prednisone, a dehydrogenated inactive form of glucocorticoids, was additionally tested in H22tumor cells in vitro and H22tumor-bearing mice in vivo.(4) The molecular basis of dexamethasone affecting the glucose metabolism of hepatocarcinoma was further investigated. H22tumor-bearing mice were treated with different concentrations of dexamethasone for seven days. A panel of metabolism-related genes (HK2, PFK1, PKM2, LDHA, PDHA1, CS, SDHA, ACL, ACC, HMGCR, G6PD, GPD1) in tumor tissues was analyzed by RT-PCR. Then, the expressions of LDH and GPD1in tumor tissues were further analyzed by real-time PCR and Western blot, and lactate in tumor tissues was measured.Results:(1) Both the transcripts and proteins of PEPCK and G6Pase were strikingly lower in the tumor tissues but much higher in the peritumoral liver tissues, as shown by real time PCR and immunohistochemical staining. In line with these results, it was clear that11β-HSD1was downregulated but11β-HSD2was upregulated in hepatocarcinoma tissues. Moreover,58human hepatocarcinoma specimens were immunohistochemically analyzed and the relative expressions of11β-HSD1and11β-HSD2were quantified, which revealed the inversely expressed11β-HSD1and11β-HSD2in hepatocarcinoma relative to normal liver tissue. The ratios of11β-HSD1/11β-HSD2were calculated and the patients’samples were split into2classes (high and low) according to the median value in the whole set of58samples. The Kaplan-Meier survival analysis showed that patients with high ratio had a significant longer survival time and lesser recurrence than those with low ratio.(2) When we used murine hepatocarcinoma tumor cell line H22to generate hepatocarcinoma, the expressions of PEPCK and G6Pase were found to be strikingly decreased and were not affected by fasting, as shown by RT-PCR, Western blot and immunohistochemical staining. Surprisingly,11β-HSD1was markedly downregulated and11β-HSD2was markedly upregulated in hepatocarcinoma tissues, compared to normal liver tissues. In addition, the primary hepatocarcinoma cells and normal hepatocytes, isolated from tumor-bearing mice, also showed downregulation of11β-HSD1and upregulation of11β-HSD2. Not therefore unexpectedly,11β-HSD1-overexpression or11β-HSD2-knockdown both restored gluconeogenesis with upregulations of PEPCK and G6Pase in H22tumor tissues. As a result, the inoculation of these engineered H22tumor cell lines to the mice resulted in the inhibition of tumor growth and the corresponding prolonged survival of the mice.(3) In vitro, dexamethasone-treated H22tumor cells showed the upregulation of PEPCK and G6Pase expressions under the concentrations of1and10μM. Consistently, the levels of intracellular glucose were also found to be increased. In line with these in vitro results the in vivo dexamethasone treatments resulted in increases in PEPCK and G6Pase expressions and tissue glucose in all the mice, compared to the saline control. Moreover, dexamethasone treatment showed significant inhibition of the ectopic H22tumor growths and orthotopic H22tumor growths. But dexamethasone did not significantly suppress tumor growth and only produced marginal effects, suggesting a relative selectivity of dexamethasone for hepatocarcinoma but not non-liver cancers. It was found that the intragastric administration of3-MPA the PEPCK-selective inhibitor effectively counteracted the inhibitory effects of dexamethasone on H22tumor. Unlike dexamethasone and its efficacy, prednisone a dehydrogenated inactive form of glucocorticoids did not show the upregulation of PEPCK and G6Pase or any antitumor effect.(4) LDHA and GPD1were found to be downregulated after dexamethasone treatment and the lactate level was also downregulated.Conclusions:We show here that gluconeogenesis was not present in human or mouse malignant hepatocytes. Two critical enzymes11β-HSDl and11β-HSD2that regulate glucocorticoid activities were found to be expressed inversely in malignant hepatocytes, resulting in the inactivation of endogenous glucocorticoids and the loss of gluconeogenesis. In patients’hepatocarcinoma, the expressions of11β-HSD1and11β-HSD2are strikingly linked to patients’ prognosis and survival. Dexamethasone, the active form of synthesized glucocorticoids, is capable of restoring gluconeogenesis in malignant cells by bypassing the abnormal regulation of11β-HSD enzymes, leading to therapeutic efficacy against hepatocarcinoma. These findings reveal that the reversed expressions of11β-HSD1and11β-HSD2may play an important role in the switch gluconeogenesis to glycolysis in hepatocarcinoma, begging it to be a new hepatocarcinoma treatment target.