TCF7L2 involvement in estradiol- and progesterone-modulated islet and hepatic glucose homeostasis

To evaluate the role of TCF7L2, a key regulator of glucose homeostasis, in estradiol (E2) and progesterone (P4)-modulated glucose metabolism, mouse insulinoma cells (MIN6) and human liver cancer cells (hepG2 and HUH7) were treated with physiological concentrations of E2 or P4 in the up- and down-regulation of TCF7L2. Insulin/proinsulin secretion was measured in MIN6 cells, while glucose uptake and production were evaluated in liver cancer cells. E2 increased insulin/proinsulin secretion under both basal and stimulated conditions, whereas P4 increased insulin/proinsulin secretion only under glucose-stimulated conditions. An antagonistic effect, possibly concentration-dependent, of E2 and P4 on the regulation of islet glucose metabolism was observed. After E2 or P4 treatment, secretion of insulin/proinsulin was positively correlated with TCF7L2 protein expression. When TCF7L2 was silenced, E2- or P4-promoted insulin/proinsulin secretion was significantly weakened. Under glucotoxicity conditions, overexpression of TCF7L2 increased insulin secretion and processing. In liver cancer cells, E2 or P4 exposure elevated TCF7L2 expression, enhanced the activity of insulin signaling (pAKT/pGSK), reduced PEPCK expression, subsequently increased insulin-stimulated glucose uptake, and decreased glucose production. Silencing TCF7L2 eliminated effects of E2 or P4. In conclusion, TCF7L2 regulates E2- or P4-modulated islet and hepatic glucose metabolism. The results have implications for glucose homeostasis in pregnancy.

Proinsulin and insulin release tests. MIN6 cells (5 × 10 5 cells/well) were seeded in six-well plates and incubated overnight. Cells were treated with the vehicle control or sex hormones for 24 h. Proinsulin or insulin was released into the medium and measured using the proinsulin or insulin ELISA kit. The proinsulin or insulin secretion was normalized to viable cell numbers.
Glucose uptake and production assays. Cells (2.5 × 10 5 cells/well) seeded in six-well plates were treated with sex hormones for 24 h. The medium was replaced with serum-and glucose-free DMEM for another 4 h. The cells were then washed with PBS and incubated in serum-and glucose-free DMEM containing the fluorescent glucose analog 2-deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4-yl)amino]-d-glucose (2-NBDG, 40 μ M) and insulin (100 nM) for 30 min. After incubation, cells were washed with PBS, and fluorescence retained by the cells was measured by flow cytometer (BD Bioscience, San Jose, California, USA). Data from 10,000 cells were collected and analyzed.
Cells (5 × 10 4 cells/well) seeded in 24-well plates were treated with sex hormones for 24 h. The medium was then replaced with serum-and glucose-free DMEM containing 20 mmol/l d-lactate and 2 mmol/l sodium pyruvate for 20 h. The medium was subsequently assayed for glucose with a commercial kit (Invitrogen). The glucose production was normalized to viable cell numbers.
Qualitative RT-PCR. Total RNA was extracted using a commercial kit, and real-time PCR was performed with the 9700 Real-time PCR System (Applied Biosystems, Carlsbad, California, USA) and the SDS 2.1 software (Applied Biosystems). The PCR primers are shown in Supplemental Table 2. Each PCR run started with the incubation of samples at 50 °C for 2 min, followed by incubation at 95 °C for 10 min, and 40 cycles of incubation at 95 °C for 15 s and at 59 °C for 30 s. Data were normalized by quantifying the amount of amplified cDNA products by calculating the ratio of the amount of cDNA relative to the amount of GAPDH cDNA.

Results
Effects of estradiol and progesterone on β-cell survival and function. To evaluate the effects of E 2 and P 4 on β -cell proliferation, we treated MIN6 cells with 100 nM E 2 , 1 μ M P 4 , or 100 nM E 2 plus 1 μ M P 4 . We found that the sole use of E 2 or P 4 increased the number of viable cells (Fig. 1A), but E 2 and P 4 together significantly decreased the number of viable cells at 24, 48, and 72 h (Fig. 1A, P < 0.05). E 2 can stimulate progesterone receptor (PR) activity 18 . Overactivation of P 4 /PR signaling exerted an unfavorable effect on β -cell survival and function [19][20][21] . Thus, we further exposed the cells to a low concentration of E 2 (10 nM) plus P 4 and found a significant increase in the number of viable cells compared with those treated with high concentration of E 2 (100 nM) plus P 4 (Fig. 1A, P < 0.05). A similar result was observed after PR was blocked with antagonist RU486 (40 μ M) but not after ER was blocked with ICI182780 (10 μ M) (Fig. 1A).
Insulin and proinsulin secretion were evaluated after sex hormone treatment in MIN6 cells. When the cells were cultured in E 2 alone for 24 h, they exhibited significantly increased insulin/proinsulin secretion under basal and glucose-stimulated conditions (Fig. 1B,C, P < 0.05) compared with the control cells. Cells treated with P 4 alone showed significantly increased insulin/proinsulin secretion only under glucose-stimulated conditions (Fig. 1B,C, P < 0.05). However, co-treatment of E 2 and P 4 demonstrated a much weaker effect on insulin/proinsulin secretion than the effect caused by E 2 or P 4 treatment alone. After reducing the concentration of E 2 or blocking PR, cells treated with E 2 plus P 4 showed significantly increased insulin/proinsulin secretion (Fig. 1B,C, P < 0.05). The insulin/proinsulin stimulatory index and insulin processing capability (reflected by the proinsulin-to-insulin ratio) followed the same tendency as the insulin/proinsulin secretory dynamic ( Fig. 1D-F).
Silencing of TCF7L2 weakens estradiol-and progesterone-promoted β-cell insulin secretion. To assess the role of TCF7L2 in E 2 -or P 4 -modulated β -cell function, we analyzed the TCF7L2 protein content after sex hormone exposure. E 2 markedly increased TCF7L2 protein content under both basal and stimulated conditions, whereas P 4 sharply increased TCF7L2 protein content under stimulated conditions ( Fig. 2A). Similar effect was found in TCF7L2 mRNA levels after sex hormone exposure (Supplemental Fig. 1). The TCF7L2 protein content showed a significantly positive correlation with the insulin/proinsulin secretion after E 2 or P 4 treatment (Supplemental Table 3; Figs 1 and 2).
We knocked down TCF7L2 in MIN6 cells and found that the number of viable cells significantly decreased at 24, 48, and 72 h ( Fig. 2B, P < 0.05). Decreased TCF7L2 expression led to significantly lower insulin/proinsulin secretion (Fig. 2C,D, P < 0.05). More importantly, the downregulation of TCF7L2 markedly weakened the E 2 -or P 4 -promoted insulin/proinsulin secretion (Fig. 2C,D, P < 0.05). E 2 did not significantly change the insulin stimulatory index, whereas P 4 remarkably increased insulin and proinsulin stimulatory indexes (Fig. 2E,F). The downregulation of TCF7L2 weaken the P 4 -elevated insulin/proinsulin stimulatory index (Fig. 2E,F). Furthermore, P 4 -promoted insulin processing persisted after the downregulation of TCF7L2, as reflected by the significantly decreased proinsulin-to-insulin ratio compared with that of the control (Fig. 2G, P < 0.05).

Overexpression of TCF7L2 increases insulin secretion and processing under glucotoxicity status.
To mimic diabetic glucotoxicity, we exposed MIN6 cells to a high glucose concentration (33.3 mM) for 72 h. Although an increase in basal secretion of insulin was observed, we found significantly reduced insulin secretion and increased proinsulin-to-insulin ratio at stimulated conditions in high glucose-exposed MIN6 cells compared with 5.5 mM glucose-cultured cells (P < 0.05), thereby indicating impaired glucose-stimulated insulin secretion and processing capability in the setting of glucotoxicity.
Under glucotoxicity conditions, we found that P 4 cannot stimulate TCF7L2 protein content (Fig. 3A) and cannot reduce the proinsulin-to-insulin ratio under stimulated condition (Fig. 3G). By contrast, E 2 treatment was associated with similar effects on insulin and proinsulin secretion as under normal conditions (Figs 2 and 3) Overexpression of TCF7L2, led to increased number of viable cells (Fig. 3B), significantly increased basal and stimulated insulin secretions (Fig. 3C, P < 0.05) and enhanced insulin processing capability, as reflected by a significantly decreased proinsulin-to-insulin ratio under glucotoxicity conditions (Fig. 3G, P < 0.05). However, the stimulatory indexes remained unchanged (Fig. 3E,F, P > 0.05).
TCF7L2 participates in estradiol-and progesterone-regulated hepatic glucose metabolism. Since the liver plays a central role in controlling glucose uptake and production, we assessed the effect of sex hormones on hepatic glucose metabolism. We treated hepG2 and HUH7 cells with 100 nM E 2 , 1 μ M P 4 , or 100 nM E 2 plus 1 μ M P 4 . The numbers of viable cells were similar in each sex hormone-treated group compared with the controls at 24, 48, and 72 h (P > 0.05). We also found a significantly increased insulin-stimulated glucose uptake and decreased glucose production after E 2 , P 4 , or combined treatment (Fig. 4A,B). Moreover, TCF7L2 protein content markedly increased after E 2 , P 4 , or combined treatment (Fig. 4C).
After TCF7L2 downregulation in hepG2 cells, the effect caused by E 2 or P 4 disappeared (Fig. 5A,C). Overexpression of TCF7L2 increased insulin-stimulated glucose uptake and decreased glucose production in cells with or without sex hormone treatment (Fig. 5B,D).
To further investigate the mechanism of sex hormone-modulated hepatic glucose metabolism, we examined the expression of several key molecules after E 2 or P 4 treatment. We found that both E 2 and P 4 reduced the expression of PEPCK but increased the expression of pAKT, pGSK, and pERK1/2 proteins (Fig. 5E). By silencing TCF7L2, the effect of E 2 -and P 4 -stimulated phosphorylation of AKT, GSK, and ERK1/2 was significantly weakened. However, P 4 induced the elevated IRS2 expression in the presence and absence of TCF7L2. Additionally, silencing of TCF7L2 increased the PEPCK expression but decreased the expression of GLUT2 and IRS2 in each group (Fig. 5E), whereas overexpression of TCF7L2 produced the opposite effect (Fig. 5F). The quantification of western blots is presented in Supplemental Fig. 2. We also found the same effects of E 2 and P 4 in mRNA levels of PEPCK, GLUT2, and IRS2 (Supplemental Fig. 3).

Discussion
The increased levels of estrogenic and progestational hormones can generally lead to insulin resistance and elevated glucose levels. However, pregnant women were found to exhibit lower blood glucose levels than those of non-pregnant ones 22 . Updated guidelines recommend lower blood glucose values to diagnose GDM 22 . E 2 and P 4 induce opposing physiological effects on the female reproductive system and other organ systems, such as the cardiovascular system 22 . The present study demonstrates that the two sex hormones may exert an antagonistic effect on regulating islet glucose homeostasis. Our results reveal that acute treatment of E 2 or P 4 leads to elevated β -cell proliferation, increased insulin/proinsulin production. This response seems to be an adaptation of islets to maintain maternal glucose homeostasis, even under conditions of glucotoxicity. However, combined E 2 and P 4 exposure resulted in an opposite effect, namely decreased cell proliferation and reduced glucose-stimulated responsivity and insulin processing. A previous report implied that the antagonistic effect permitted P 4 to stimulate islet cell proliferation, but the effect was suppressed by E 2 supplementation 6 . However, the sole use of E 2 did not affect β -cell proliferation 23 but prevented apoptosis under oxidative stress injury 24 .
To explain this series of events, it is necessary to focus first on the paradoxical effect of P 4 on β cells. Previous studies have shown that the physiological concentration of P 4 stimulates β -cell proliferation 6 and increases insulin secretion under conditions of oxidative stress 25 . By contrast, pharmacological or supra-physiological concentrations of P 4 will exert an unfavorable effect on β -cell survival and function 21,26 . Overstimulation of P 4 /PR signaling will lead to reduced β -cell viability 19,20 . E 2 can markedly stimulate PR activity in β cells 18 ; thus, we assumed that E 2 may reinforce P 4 /PR signaling and cause an amplification effect. As expected, we found that the antagonistic effect can be significantly attenuated by reducing the concentration of E 2 during cotreatment or by inhibiting the PR. By contrast, the ER antagonist cannot achieve such an effect because it inhibits the E 2 signal but, moreover, cannot influence overstimulated P 4 /PR signaling. Overall, our results suggest a role for the interplay between sex hormones and the development of GDM. However, further studies, are with regard to this observation.
TCF7L2 is the most powerful, recognized susceptibility gene for diabetes. The association of TCF7L2 gene polymorphisms with almost all subtypes of diabetes (e.g., type 1 and 2 diabetes, latent autoimmune diabetes in adults, hepatogenous diabetes, and post-transplant diabetes) has been demonstrated 27-30 . Zhou et al. proposed a master role for TCF7L2 in regulating insulin biosynthesis, secretion, and processing 10 . But evidence from Boj et al. was less convincing 14 . Our results revealed that TCF7L2 was essential for glucose-stimulated insulin/proinsulin secretion and processing, as well as β -cell proliferation. Furthermore, under conditions of glucotoxicity, the overexpression of TCF7L2 will markedly increase insulin processing and secretion. These findings attest to the significant role of TCF7L2 in islet function and indicate a potential clinical role of TCF7L2 in treating diabetes.
We further found a positive correlation between E 2 -or P 4 -increased TCF7L2 protein content and E 2 -or P 4 -promoted insulin/proinsulin secretion. Downregulation of TCF7L2 remarkably weakened the effect induced by E 2 or P 4 , thereby indicating that E 2 -or P 4 -promoted insulin/proinsulin secretion can, at least in part, be TCF7L2 dependent. Moreover, the E 2 -promoted stimulated insulin processing is TCF7L2 required; whereas the P 4 -promoted stimulated insulin processing may be independent of TCF7L2 because the effect persisted in the absence of TCF7L2 expression. In addition, TCF7L2 may not be essential for the E 2 -or P 4 -increased stimulatory index since downregulation of TCF7L2 did not alter the effects in stimulatory index. Under conditions of glucotoxicity, E 2 increased TCF7L2 protein content under basal and stimulated conditions as it did under the normal, non-glucotoxic conditions to achieve similar effects on β -cell secretion. In contrast, P 4 did not further increase TCF7L2 protein content under stimulated conditions. Thus, insulin processing was unaffected. However, P 4 still induced enhanced insulin/proinsulin secretion under stimulated conditions. The results suggest that TCF7L2 is essential for insulin processing. However, TCF7L2 was not required for P 4 -promoted stimulated insulin/proinsulin secretion in β cells that were glucotoxic.
Given the central role of the liver in the regulation of glucose homeostasis, we also assessed the effects of E 2 and P 4 on hepatic glucose metabolism. Although the opposing effects of E 2 and P 4 on oxidative stress processes have been reported in vitro [31][32][33] , a recent in vivo study using microarray analysis showed that E 2 , P 4 , and their co-treatment actually displayed similar patterns and networks of hepatic gene expression 34 . Consistent with the previous study, we found that E 2 and P 4 presented similar phenotypes and protein content profiles in regulating hepatic glucose metabolism. Both E 2 and P 4 can improve insulin-stimulated glucose uptake, possibly through the enhanced activation of hepatic insulin signaling (pAKT/pGSK). E 2 and P 4 can also reduce gluconeogenesis by repressing PEPCK expression. Both E 2 and P 4 significantly increased hepatic ERK activation, which can increase glycogen synthesis and attenuate glucose output 35 . In addition, the opposing effects of P 4 and E 2 were detected in the expression of IRS2 and GLUT2, which may further explain the stronger effect of P 4 than E 2 in regulating hepatic glucose uptake.
Combined with previous studies, our study has demonstrated that TCF7L2 represses hepatic glucose production by regulating gluconeogenic genes [13][14][15][16] . In addition, we revealed that TCF7L2 mediated hepatic insulin signaling (IRS2-AKT-GSK), glucose sensing (GLUT2), and energy metabolism (ERK). Norton et al. have revealed that TCF7L2 binds directly to multiple genes that are important in hepatic glucose metabolism (e.g., PCK1, IRS2, AKT2 and PDK4) using CHIP-Seq in vitro 36 . Therefore, TCF7L2 may modulate hepatic glucose homeostasis in various ways far more than gluconeogenesis 37 . Moreover, this study demonstrates that TCF7L2 is involved in E2-or P 4 -modulated liver glucose metabolism. E 2 or P 4 -stimulated activation of AKT/GSK and ERK1/2 signaling seems to be TCF7L2-dependent because silencing of TCF7L2 clearly reverses sex hormone-induced phosphorylation of AKT/GSK and ERK1/2 proteins. In contrast, the P 4 -induced overexpression of IRS2 may be TCF7L2-independent because the effect did not diminish after the silencing of TCF7L2.
This study has several limitations. First, the interplay between different concentrations of E 2 and P 4 is complex. We did not assess the correlation between TCF7L2 and E 2 plus P 4 . The genomic and non-genomic mechanisms of the antagonistic effect need further research in future well-designed studies. Second, this work is an in vitro study with short-term sex hormone treatment. Thus, in vivo studies are needed to further document our findings.
In conclusion, both E 2 and P 4 at physiological concentrations modulate glucose homeostasis to meet pregnancy demands, including increased insulin secretion and processing, enhanced hepatic insulin signaling, and reduced gluconeogenesis. The reported insulin resistance caused by E 2 or P 4 may be attributable to other mechanisms, such as peripheral insulin resistance and insulin binding 38,39 . The antagonistic effect of E 2 and P 4 on the regulation of islet glucose metabolism, possibly concentration-dependent, was also observed. Therefore, pregnant women with different combinatorial amounts of sex hormones may be at differential risk for developing GDM. Finally, TCF7L2 is an important regulator of both islet and hepatic glucose metabolism. TCF7L2 participates in E 2 -or P 4 -modulated glucose homeostasis, particularly in hepatic glucose production. Compared with P 4 , E 2 seems to be more dependent upon expression of TCF7L2 to regulate glucose metabolism.