Progesterone increases blood glucose via hepatic progesterone receptor membrane component 1 under limited or impaired action of insulin

Hepatic gluconeogenesis is the main pathway for blood glucose maintenance activated during fasting. Retardation of insulin action, such as in diabetes mellitus, activates gluconeogenesis during the fed state. While the role of progesterone (P4) in diabetes is controversial, the P4 receptor, progesterone receptor membrane component 1 (PGRMC1), is known to stimulate pancreatic insulin secretion. We investigated the role of P4, via hepatic PGRMC1, during gluconeogenesis. The PGRMC1 binding chemical, AG-205, induced PGRMC1 monomer (25 kDa) abundance, and increased PEPCK expression and glucose production in parallel with cyclic AMP (cAMP) induction in Hep3B cells. PGRMC1-mediated cyclic AMP was inhibited by an adenylate cyclase inhibitor (MDL-12,330A). PEPCK suppression in Pgrmc1 KO hepatocyte was not observed after treatment of MDL-12,330A. PGRMC1 knockdown or overexpression systems in Hep3B cells confirmed that PGRMC1 mediates PEPCK expression via phosphorylation of cAMP-response element binding protein (CREB). CREB phosphorylation and PEPCK expression in primary hepatocytes were greater than that in PGRMC1 knock-out hepatocytes. Progesterone increased PGRMC1 expression, which induced cAMP and PEPCK induction and glucose production. In vivo, P4 suppressed gluconeogenesis following plasma insulin induction under normal conditions in a mouse model. However, P4 increased blood glucose via gluconeogenesis in parallel with increases in PGRMC1 and PEPCK expression in mice in both insulin-deficient and insulin-resistant conditions. We conclude that P4 increases hepatic glucose production via PGRMC1, which may exacerbate hyperglycaemia in diabetes where insulin action is limited.

To gain an insight into the mechanism of regulation of PGRMC1 and PEPCK expression, an adenylate cyclase inhibitor (MDL) was used to determine whether PGRMC1-mediated cAMP accumulation is due to adenylate cyclase activity. As a result, PGRMC1 expression was increased (p < 0.05, 1.52-vs. control) in the AG-205 group and remained high (p < 0.05, 1.56-fold vs. control) in the AG-205 + MDL group ( Fig. 2A). PEPCK expression was also increased (p < 0.05, 1.24-fold vs. control) in the AG-205 group, but the increase was suppressed (p < 0.05, 72% vs. AG-205 group) in the AG-205 + MDL group ( Fig. 2A). While cAMP level was suppressed in the MDL group (p < 0.05, 56.5% vs. control), the induction of cAMP was observed in the AG-205 group (p < 0.05, 1.24-fold vs. control), and the induction through AG-205 was blocked by MDL treatment (p < 0.05, 68.6% vs. AG-205 group), as shown in Fig. 2B. We therefore confirmed the involvement of adenylate cyclase in the PGRMC1-cAMP-PEPCK expression pattern; an illustration of the suggested pathway is provided in Fig. 2C. To confirm PGRMC1mediated PEPCK induction by AG-205, we additionally monitored the expression of PGRMC1 monomer and PEPCK in primary WT and Pgrmc1 KO hepatocytes with AG-205. Following induction (p < 0.05, 1.56-fold vs. control) of PGRMC1 by AG-205, the PEPCK protein was also increased (p < 0.05, 1.48-fold vs. control) in primary hepatocytes (Fig. 3A). Importantly, the expression of PEPCK was decreased (p < 0.05, 36.1% vs. WT hepatocytes) in Pgrmc1 KO primary hepatocytes, and AG-205 was not sufficient to increase PEPCK expression (Fig. 3B). Through these results, we confirmed that AG-205 regulates PEPCK via the PGRMC1 monomer abundance. To confirm the involvement of adenylate cyclase, we treated MDL-12,330A in primary hepatocyte. As expected, the www.nature.com/scientificreports/ expression of PEPCK was suppressed (p < 0.05, 43.6%) in Pgrmc1 KO hepatocyte compared to WT hepatocytes (Fig. 3C). The expression of PGRMC1 was also inhibited in WT hepatocyte + MDL-12,330A (p < 0.05, 39.4% vs. WT control hepatocytes), therefore, PGRMC1 might be related to adenylate cyclase (Fig. 3C). However, after MDL-12,330A treatment, a clear reduction of PEPCK expression was not observed in the Pgrmc1 KO hepatocyte, which was seen in the WT hepatocyte (Fig. 3C).
PEPCK regulation by PGRMC1 in hepatocytes. As PGRMC1 regulates PEPCK according to adenylate cyclase activity, we next focused on CREB phosphorylation which is crucial initiator for PEPCK induction. To evaluate the effect of PGRMC1 on phosphor-CREB and PEPCK expression, we performed PGRMC1 knockdown in Hep3B cells in low-glucose medium. When the expression of PGRMC1 was decreased (47.6% vs. control) by PGRMC1 siRNA, the expression of PEPCK was also decreased (p < 0.05, 60.2% vs. control) in the PGRMC1 siRNA group (Fig. 4A). The ratio of phosphorylated/total CREB was concomitantly decreased (p < 0.05, 60.6% vs. control) in the PGRMC1 siRNA group (Fig. 4A). Consistently, the expression of phosphorylated/total CREB and PEPCK and was decreased (p < 0.05, 62% and 70.6%, respectively, vs. control) in PGRMC1 siRNA group even after P4 treatment (Fig. S2). When the expression of PGRMC1 was increased (p < 0.05, 1.56-fold vs. control) using a PGRMC1 overexpression vector, the ratio of phosphorylated/total CREB and expression of PEPCK was also increased (p < 0.05, 1.53-and 1.48-fold, respectively, vs. control) in the PGRMC1 overexpression group (Fig. 4B). Based on these results, we confirmed that hepatic PGRMC1 can directly control the expression levels of PEPCK protein via phosphorylation of CREB without the influence of hormones or growth factors. Using primary hepatocytes from WT or Pgrmc1 KO mice, we confirmed the relationship between PGRMC1 and PEPCK expression in a more physiologically relevant condition. Hepatocytes were incubated in low glucose medium (50 mg/dl) for 24 h. Firstly, the mRNA expression of murine Pepck was suppressed (p < 0.05, 21.7% vs. WT hepatocytes) in Pgrmc1 KO primary hepatocytes (Fig. S3). Moreover, the ratio of phosphorylated/total CREB and expression of PEPCK were suppressed (p < 0.05, 81.8%, and 34.4%, respectively, vs. WT hepatocytes) in Pgrmc1 KO hepatocytes (Fig. 4C).
Progesterone induces gluconeogenesis in mice in which insulin effects are retarded. We investigated whether P4 increases gluconeogenesis in vivo via PGRMC1 induction. According to previous studies, while P4 is known to induce gluconeogenic genes, it also promotes insulin secretion in the pancreas. We compared the effects of P4 on gluconeogenesis in normal, insulin-deficient, and insulin-resistant states. Under typical physiological conditions, P4 suppressed the levels of blood glucose in a pyruvate tolerance test. Blood glucose levels were decreased (p < 0.05, 83.1, 81.1, and 83.6%, respectively, vs. control) at 45, 60, and 90 min in the P4 group after sodium pyruvate injection (Fig. 6A). The area under the curve (AUC) was also decreased in the P4-treated group (Fig. 6A). Blood glucose levels were decreased after 45 min, which implies that P4 promoted insulin action because blood glucose levels in the P4 group were similar to those of the control at 15 min. Consistently, the level of plasma insulin (μIU/min) was increased (p < 0.05, 2.16-fold vs. control) in the P4 group (Fig. 6A). The expression of PGRMC1 was increased (p < 0.05, 1.5-fold vs. control), but the ratio of phosphorylated/total CREB and PEPCK expression were not changed (Fig. 6A). www.nature.com/scientificreports/ Conversely, when we triggered insulin deficiency by high-dose of STZ (100 mg/kg), the levels of blood glucose were increased (1.07-, 1.16-, 1.24-, 1.13-, and 1.23-fold, respectively, vs. control) at 15, 30, 45, 60, and 90 min in the P4 group after pyruvate injection (Fig. 6B). The AUC was also increased in the P4 group (Fig. 6B). Plasma insulin levels were similarly low between the control and P4 groups (Fig. 6B). The expression of PGRMC1, the ratio of phosphorylated/total CREB, and the expression of PEPCK were increased (p < 0.05, 1.54-, 1.83-, 1.31fold, respectively, vs. control) in the P4 group (Fig. 6B).
Similarly, when we triggered insulin resistance by feeding mice a high-fat diet and injecting them with a low dose of STZ (30 mg/kg), the levels of blood glucose were increased (1.24-, 1.16-, 1.38-, and 1.51-fold, respectively, vs. control) at 15, 30, 60, and 90 min in the P4 group after pyruvate injection (Fig. 6C). The AUC was also increased in the P4 group (Fig. 6C). Plasma insulin levels were similar between the control and P4 groups (Fig. 6C). The expression of PGRMC1, the ratio of phosphorylated/total CREB, and the expression of PEPCK were increased (p < 0.05, 4.61-, 2.93-, and 1.97-fold, respectively, vs. control) in the P4 group (Fig. 6C). Collectively, these results are consistent with in vitro data in the present study, showing the induction of hepatic gluconeogenesis by P4 when insulin action was limited.

Discussion
Previous research has reported that a high dose of P4 disturbs glucose homeostasis linked with diabetes mellitus 15 . Moreover, P4 receptor-knockout mice were less prone to hyperglycaemia 16 . While P4 seems to be the reason for insulin resistance that triggers diabetes mellitus during gestation 15 , the role of P4 in gluconeogenesis remains unclear. In our study, we identified the relationship between PGRMC1 and PEPCK and further examined the effect of P4 on gluconeogenesis via PEPCK induction. AG-205 increased PEPCK expression via PGRMC1 and adenylate cyclase, and the PGRMC1-mediated PEPCK expression involved with phosphorylation of CREB. Consistent with a previous study reporting that P4 interacts with PGRMC1 17 , P4 increased PGRMC1 and PEPCK expression, cAMP level, and glucose production. These observations are important because the role of PGRMC1 could be highlighted with direct evidence by modulating PGRMC1 in vitro. Following PTT analysis in a mouse model, P4 increased gluconeogenesis via induction of PGRMC1 and PEPCK when insulin action is limited. www.nature.com/scientificreports/ Our study suggests that PGRMC1 may be an emerging target for controlling gluconeogenesis and diabetes in an insulin-resistant state. Using siRNA or overexpression plasmids regulating PGRMC1 levels, we confirmed that hepatic PGRMC1 regulates PEPCK, a key enzyme in gluconeogenesis. Moreover, the PGRMC1 ligand AG-205 accumulates PGRMC1 monomer, and this induction of PGRMC1 monomer enhanced PEPCK expression through cAMP accumulation. Although glucagon and epinephrine are also known to induce intracellular cAMP production and PKA activation 18 , intracellular cAMP accumulation occurred upon an increase in PGRMC1 levels without glucagon or epinephrine in our study. In addition, MDL-12,330A completely inhibited expression of PEPCK in both WT and Pgrmc1 KO hepatocytes (Fig. 3C), suggesting PGRMC1 regulates PEPCK through cAMP and adenylate cyclase dependent manner. After the increase in cAMP, activated PKA then enters the nucleus, phosphorylates the cAMP regulatory element-binding protein (CREB), and induces Pepck transcription 19 . Knockdown of hepatic CREB substantially reduced blood glucose production in a ZDF T2DM rat model 20 . PGRMC1 regulates PEPCK expression via CREB phosphorylation, considering the results of PGRMC1 knockdown and overexpression in Hep3B cells. It was also confirmed that the PGRMC1 regulates the ratio of phosphor/total CREB and expression of PEPCK in primary hepatocytes. Therefore, increased cAMP production through the induced increase in PGRMC1 levels will lead to an increased rate of gluconeogenesis.
AG-205 is known as a PGRMC1 inhibitor based on many previous studies but has also been shown to increase PGRMC1 monomer levels 21 . In addition, a previous study showed that AG-205 treatment decreased dimeric and oligomeric forms of PGRMC1 but increased the monomeric form of PGRMC1 22 . Importantly, the PGRMC1 monomer could also be functional, in addition to the multimeric forms of PGRMC1 23 . In the present study, we observed induction of PGRMC1 monomers (25 kDa) after AG-205 treatment, similar to previous reports. We confirmed that increased levels of PEPCK after AG-205 treatment were consistent with results representing PGRMC1 knockdown and overexpression, respectively. Furthermore, it was clear that AG-205 induced PEPCK through induction of PGRMC1 monomer because PEPCK levels in PGRMC1 KO hepatocytes did not increase upon AG-205 treatment.
A derivative of P4, 17α-hydroxyprogesterone, can lower the occurrence rate of preterm delivery 24 , which is prevalent in moderately overweight and obese pregnant women 25 . However, in a clinical study, women treated with 17α-hydroxyprogesterone caproate during pregnancy were predisposed to gestational diabetes compared to the control group 26 . While P4 contributes to insulin secretion 8,9 , it can also result in insulin resistance and the concurrent induction of gluconeogenic genes 7 . To exclude the effect of P4-mediated insulin, we used in vitro conditions to investigate the role of P4 in hepatocytes. In vitro P4 increased PGRMC1, PEPCK, cAMP, and glucose production. However, P4 seems to suppress gluconeogenic activity in vivo as P4 promotes insulin secretion in normal mice. Interestingly, P4 also induces phosphorylation of CREB, PEPCK, and gluconeogenesis in mice under conditions of insulin-deficient and insulin-resistant conditions. Accordingly, P4 induces hepatic gluconeogenesis and maintains high glucose levels in a state that insulin has almost no effect due to insulin resistance. This is important as normal pregnancy usually includes insulin resistance 27 , and P4 might induce gestational diabetes by increasing gluconeogenesis when the effects of insulin are not considered.
Type II diabetes is characterized by hyperglycaemia during a fasting state and by a prolonged rise in blood glucose levels after the fed state. This type of diabetes is associated with insulin resistance and promotes gluconeogenesis despite high insulin levels 28 . Furthermore, increased alpha-cell function and consequent hyperglucagonemia have been recognized as contributors to hyperglycaemia in type II diabetic patients because of gluconeogenesis activation 29 . In an early study, gestational P4 had relative diabetogenic properties, decreased glucose uptake, and increased hepatic glucose production 15 . It is very interesting that treatment of pregnant women with P4 for the prevention of recurrent preterm birth increased the risk of gestational diabetes mellitus 26 . In this study, we demonstrated that P4-mediated PGRMC1 induction increases gluconeogenesis under insulin-resistant conditions such as type II diabetes. In conclusion, our study suggests that the regulation of gluconeogenesis by PGRMC1 modulation holds great potential as a new therapeutic approach to manage and treat type II diabetes.
Methods chemicals. MDL-12,330A hydrochloride was purchased from Tocris Bioscience. Progesterone was purchased from Tokyo Chemical Industry. Streptozotocin was purchased from Santa Cruz Biotechnology. Sodium pyruvate, glucagon, and dexamethasone were purchased from Sigma-Aldrich.
Animals. C57BL/6J WT and Pgrmc1 KO male mice were housed in a pathogen-free facility at Chungnam National University under a standard 12 h light:12 h dark cycle and fed standard chow with water provided ad libitum. All mouse experiments were approved and performed in accordance with the Chungnam Facility Animal Care Committee (CNU-00606).
For the pyruvate tolerance test (PTT), mice were injected with P4 (1 mg/kg) for 3 days and fasted for 24 h as of the last injection. Mice were injected with sodium pyruvate (2 g/kg), and blood glucose was monitored by tail snipping. For insulin-deficiency, mice were pre-treated with streptozotocin (STZ) (100 mg/kg) before 1 week and were assessed by PTT. For the insulin-resistant state, mice were fed a high-fat diet for 8 weeks. On the first day of week 6 of the high-fat diet, mice were injected with a low dose of STZ (30 mg/kg). After 8 weeks of high-fat diet, mice were assessed by PTT. Streptozotocin was dissolved in citrate buffer (pH 4.4) and injected into the intraperitoneal cavity (IP). Progesterone was injected into the subcutaneous inguinal area. A high-fat diet (Research Diets, New Brunswick, NJ) was composed of carbohydrate (20% kcal), protein (20% kcal), and fat (60% kcal).
Blood glucose level. Blood  Glucose production assay. Hep3B cells were maintained in glucose-free medium containing 2 mM sodium pyruvate and 1 nM dexamethasone. The medium was prepared according to the manufacturer's protocol (Sigma-Aldrich).
RnA isolation, Reverse transcription, and qRt-pcR. RNA  Western blotting. Liver and Hep3B cells were lysed and sonicated for protein extraction. Protein samples were subjected to 8-12% SDS-PAGE after 5 min of boiling at 100 °C. Gels were blotted by wet transfer with Bio-Rad Power Pac at 350 V for 1 to 2 h. PVDF membranes were blocked for 1 h in bovine serum albumin (BSA) and incubated with primary antibodies overnight at 4 °C. Membranes were then incubated with secondary antibodies diluted with 1:10,000 in BSA overnight at 4 °C. The results were detected with enhanced chemiluminescence (ECL) solution (Cyanagen) and ChemiDoc (Fusion Solo, Vilber Lourmat). Primary polyclonal antibodies used were: rabbit anti-β-actin (sc-130656, Santa Cruz Biotechnology, USA) and rabbit anti-PEPCK (10004943, Cayman Chemical, USA). Rabbit monoclonal antibody to PGRMC1 (#13856) was purchased from Cell Signaling Technology, Inc. Rabbit monoclonal antibody to phosphor CREB (ab32096, Abcam), CREB (ab32515, Abcam) were used. The secondary antibody used was mouse anti-rabbit IgG (211-032-171 anti-rabbit, Jackson laboratory).
For PGRMC1 overexpression, Hep3B cells were transfected with 2.5 µg of human PGRMC1 expression plasmid and Lipofectamine 2000 (Invitrogen) in Opti-MEM (Gibco) medium according to the manufacturer's protocol. For PGRMC1 knockdown, Hep3B cells were transfected with control siRNA or PGRMC1 siRNA #1 (5′-CAG UAC AGU CGC UAG UCA A-3′) and #2 (5′-CAG UUC ACU UUC AAG UAU CA-U-3′) purchased from Bioneer (Daejeon, Korea). primary hepatocyte isolation. Primary cell culture was performed as described previously 30 . The hepatocytes of mice were isolated by collagenase digestion and the Percoll Gradient method. In detail, mice were anesthetized and perfused with Ca 2+ and Mg 2+ free-Hanks' Balanced Salt Solution (HBSS) containing EDTA (1 mM), and then digested with a collagenase solution containing liberase (0540119001, Sigma). Livers were removed and rinsed twice with HBSS, digested with a collagenase solution containing liberase (0540119001, Sigma-Aldrich) and then gently teased with forceps until they were in solution.
The cell suspensions were filtered through a sterile 40 μm nylon cell strainer (93040, SPL) to remove undigested tissue and connective tissue. The cells were centrifuged for 5 min at 1000 rpm and resuspended in medium. The resuspended cells were centrifuged using 40% Percoll for 15 min at 2000 rpm with the brake option off. After centrifugation, the healthy hepatocytes were washed twice with DMEM supplemented with 5% FBS, and then seeded into well tissue culture plates.
Statistical analysis. Data are reported as the mean ± standard deviation (SD). Differences between means were obtained by Student's t-test and one-way ANOVA followed by a Dunnett post analysis was performed using Graph Pad Software (GraphPad Inc., San Diego, CA). www.nature.com/scientificreports/