High glucose stimulates expression of aldosterone synthase (CYP11B2) and secretion of aldosterone in human adrenal cells

Aldosterone synthase is the key rate‐limiting enzyme in adrenal aldosterone production, and induction of its gene (CYP11B2) results in the progression of hypertension. As hypertension is a frequent complication among patients with diabetes, we set out to elucidate the link between diabetes mellitus and hypertension. We examined the effects of high glucose on CYP11B2 expression and aldosterone production using human adrenal H295R cells and a stable H295R cell line expressing a CYP11B2 5′‐flanking region/luciferase cDNA chimeric construct. d‐glucose (d‐glu), but not its enantiomer l‐glucose, dose dependently induced CYP11B2 transcription and mRNA expression. A high concentration (450 mg·dL−1) of d‐glu time dependently induced CYP11B2 transcription and mRNA expression. Moreover, high glucose stimulated secretion of aldosterone into the media. Transient transfection studies using deletion mutants/nerve growth factor‐induced clone B (NGFIB) response element 1 (NBRE‐1) point mutant of CYP11B2 5′‐flanking region revealed that the NBRE‐1 element, known to be activated by transcription factors NGFIB and NURR1, was responsible for the high glucose‐mediated effect. High glucose also induced the mRNA expression of these transcription factors, especially that of NURR1, but NURR1 knockdown using its siRNA did not affect high glucose‐induced CYP11B2 mRNA expression. Taken together, it is speculated that high glucose may induce CYP11B2 transcription via the NBRE‐1 element in its 5′‐flanking region, resulting in the increase in aldosterone production although high glucose‐induced NURR1 is not directly involved in the effect. Additionally, glucose metabolism and calcium channels were found to be involved in the high glucose effect. Our observations suggest one possible explanation for the high incidence of hypertension in patients with diabetes.

Aldosterone synthase is the key rate-limiting enzyme in adrenal aldosterone production, and induction of its gene (CYP11B2) results in the progression of hypertension. As hypertension is a frequent complication among patients with diabetes, we set out to elucidate the link between diabetes mellitus and hypertension. We examined the effects of high glucose on CYP11B2 expression and aldosterone production using human adrenal H295R cells and a stable H295R cell line expressing a CYP11B2 5 0 -flanking region/luciferase cDNA chimeric construct. D-glucose (D-glu), but not its enantiomer L-glucose, dose dependently induced CYP11B2 transcription and mRNA expression. A high concentration (450 mgÁdL À1 ) of D-glu time dependently induced CYP11B2 transcription and mRNA expression. Moreover, high glucose stimulated secretion of aldosterone into the media. Transient transfection studies using deletion mutants/nerve growth factor-induced clone B (NGFIB) response element 1 (NBRE-1) point mutant of CYP11B2 5 0 -flanking region revealed that the NBRE-1 element, known to be activated by transcription factors NGFIB and NURR1, was responsible for the high glucose-mediated effect. High glucose also induced the mRNA expression of these transcription factors, especially that of NURR1, but NURR1 knockdown using its siRNA did not affect high glucose-induced CYP11B2 mRNA expression. Taken together, it is speculated that high glucose may induce CYP11B2 transcription via the NBRE-1 element in its 5 0 -flanking region, resulting in the increase in aldosterone production although high glucoseinduced NURR1 is not directly involved in the effect. Additionally, glucose metabolism and calcium channels were found to be involved in the high glucose effect. Our observations suggest one possible explanation for the high incidence of hypertension in patients with diabetes.
The number of patients with diabetes mellitus is increasing every year, and 382 million people in the world were estimated to be affected in 2013 [1]. Among patients with diabetes, hypertension is one of the most frequently observed complications. In Japan, the incidence of hypertension in patients with diabetes is approximately 60%, which is twice that in nondiabetic people [2]. The etiology of hypertension in patients with diabetes is partially explained by the effect of hyperinsulinemia on renal proximal tubules due to insulin resistance [3]. Additionally, endothelial dysfunction and atherosclerosis induced by diabetes mellitus may also contribute to the progression of hypertension [4]. However, the direct involvement of high glucose on the etiology of hypertension in patients with diabetes still remains uncertain.
The renin-angiotensin-aldosterone system (RAAS) is known as the main humoral pathway involved in the etiology of hypertension, and aldosterone, the final product of the pathway, plays an important role in the progression of hypertension and vascular damages in combination with sodium [5]. Aldosterone is synthesized in the zona glomerulosa of the adrenal cortex from cholesterol catalyzed via side chain cleavage enzyme (CYP11A1), 3b-hydroxysteroid dehydrogenase (3b-HSD), steroid 21-hydroxylase (CYP21), and aldosterone synthase (CYP11B2), which is the key rate-limiting enzyme in aldosterone production [6]. Aldosterone synthase gene (CYP11B2) expression is mainly regulated by angiotensin II (AII) and potassium via transcription factors including Nur-related factor 1 (NURR1) [7]. Recently, genetic analyses of KCNJ5, ATP1A1, ATP2B3, and CACNA1D have revealed that chronic overexpression of CYP11B2 induces not only aldosterone hypersecretion but also the formation of primary aldosteronism [8], resulting in the progression of severe hypertension. Moreover, aberrant WNT signaling caused by mutations in CTNNB1 has also been recognized to be involved in the formation of primary aldosteronism [9]. In order to investigate the direct link between hypertension and diabetes mellitus, we here examined the effects of high glucose on CYP11B2 expression and aldosterone secretion using human adrenal H295R cells.

Cell culture
H295R cells or CYP11B2-H295R cells were grown with a 1 : 1 mixture of Dulbecco's modified Eagle's medium and Ham's F12 medium supplemented with 10% fetal bovine serum (FBS), Insulin-Transferrin-Selenium-G Supplements (Invitrogen, Carlsbad, CA, USA), 1.25 mgÁmL À1 BSA (Sigma), 5.35 lgÁmL À1 linoleic acid (Sigma), 100 UÁmL À1 penicillin, and 100 lgÁmL À1 streptomycin. Cells were cultured in a humidified incubator at 37°C with 5% CO 2 . As the D-glu concentration in the media was~100 mgÁdL À1 , we added either concentrated D-glu or L-glu solution to adjust the final concentration. For example, the 450 mgÁdL À1 D-glu concentration was composed of 100 mgÁdL À1 D-glu from the media and 350 mgÁdL À1 D-glu from the concentrated D-glu solution, and its osmolality-adjusted control was composed of 100 mgÁdL À1 D-glu from the media and 350 mgÁdL À1 L-glu from the concentrated L-glu solution. In some experiments, 2-deoxy-D-glu, 3-O-methyl-D-glu, D-sorbitol, or D-fructose was used instead of L-glu. Moreover, CYP11B2-H295R cells were incubated either with angiotensin II receptor blockers (ARBs) or with calcium channel blockers (CCBs) in the presence of 450 mgÁdL À1 D-glu.

RNA Preparation and quantitative real-time PCR
When H295R cells were grown to 60% confluence in 24multiwell plates, they were exposed to several concentrations of D-glu or D-glu plus L-glu for the indicated times, and their total RNA was extracted using SepasolÒ-RNA I Super G (Nacalai Tesque, Kyoto, Japan) according to the manufacturer's instructions. Total RNA were subjected to reverse transcription (RT) reaction using Prime-Script Reverse Transcriptase (Takara Bio, Ohtsu, Japan) with random 6-mer and oligo dT primers according to the manufacturer's instructions. Thereafter, the obtained templates were used for quantitative real-time PCR (95°C, 3 min for one cycle; 95°C, 15 s; 60°C, 10 s; 72°C, 20 s for 40 cycles) either with iQ Supermix (Bio-Rad, Hercules, CA, USA; for CYP11B2, CYP11B1, HSD3B2, and CYP17) or with THUNDERBIRDÒ SYBRÒ qPCR Mix (TOYOBO, Osaka, Japan; for others) by DNA Engine thermal cycler attached to Chromo4 detector (Bio-Rad). The sequences of the primers and TaqMan probes are shown in Table 1.

Transient transfection and luciferase assay
H295R cells were plated to 60% confluence in 24-multiwell plates. Thereafter, they were transiently transfected with 200 ng luciferase reporter plasmids and 100 ng pCMV-bgal using LipofectamineÒ 2000 Transfection Reagent (Life Technologies, Carlsbad, CA, USA) for 24 h. The cells were then exposed to D-glu or L-glu for the indicated times and concentrations. They were thereafter washed with PBS, and the cell extracts were prepared using Glo Lysis Buffer (Promega). Luciferase activity was measured using Bright-Glo reagents (Promega), and b-galactosidase activity was simultaneously measured. Data were normalized by the b-galactosidase activities. When the stable CYP11B2-H295R cells [7] were used, only the luciferase activity was measured.

Measurement of aldosterone/cortisol concentration
H295R cells were plated to 60% confluence in 24-multiwell plates. Thereafter, they were exposed to either 100 mgÁdL À1 D-glu, 450 mgÁdL À1 D-glu, 100 mgÁdL À1 D-glu plus 100 nmolÁL À1 AII (for aldosterone), or 450 mgÁdL À1 D-glu plus 100 nmolÁL À1 AII (for aldosterone) for 72 h. The aldosterone and cortisol concentrations of the media were thereafter measured by Aldosterone EIA Kit and Cortisol EIA Kit (Cayman Chemical), respectively, after their extraction with dichloromethane according to the manufacturer's instructions. The obtained data were normalized by the protein concentrations measured by Protein Assay Kit (Bio-Rad).

Statistical analyses
All data are presented as mean AE SEM. For the statistical analyses, ANOVA followed by post hoc Tukey's test was performed. P < 0.05 was considered statistically significant.

Effects of high glucose on CYP11B2 expression and aldosterone secretion
We first examined the effects of high glucose on CYP11B2 mRNA expression using H295R cells. As shown in Fig. 1A, D-glu levels above 270 mgÁdL À1 significantly induced CYP11B2 mRNA expression. Timecourse experiments in the presence of 450 mgÁdL À1 D-glu demonstrated that high glucose induced CYP11B2 mRNA expression after 48 h (Fig. 1B). We next examined the effect of high glucose on CYP11B2 transcription using stable CYP11B2-H295R cells [7], and also observed similar stimulatory effects in both the dose-response (Fig. 1C) and time-course (Fig. 1D) experiments. We also examined the effect of high glucose (450 mgÁdL À1 D-glu) on the mRNA expression of other enzymes/protein involved in adrenal steroidogenesis. As shown in Fig. 2, high glucose treatment significantly induced the mRNA expression of 11bhydroxylase gene (CYP11B1) (A) and steroidogenic acute regulatory protein gene (StAR) (E) after 48-h incubation, while high glucose treatment significantly decreased that of CYP11A1 (D) after 24-h incubation. High glucose treatment did not affect the mRNA expression of 3b-HSD gene (HSD3B2) (B) and CYP21 (C), while it tended to decrease, although not significantly, the expression of 17a-hydroxylase/17,20 lyase gene (CYP17) (F). We then examined the effect of high glucose on aldosterone secretion from H295R cells. As shown in Fig. 3A, incubation of the cells in the presence of 450 mgÁdL À1 D-glu for 72 h significantly induced aldosterone secretion into the media, which was comparable to the AII-induced aldosterone secretion. Incubation with 450 mgÁdL À1 D-glu plus AII did not further increase the aldosterone secretion (Fig. 3A). It can be concluded that high glucose induces CYP11B2 transcription and mRNA expression resulting in the increase in aldosterone secretion. In contrast, although high glucose induced the mRNA expression of CYP11B1 ( Fig. 2A), it did not induce cortisol secretion into the media (Fig. 3B) probably due to the decreasing trend of CYP17 mRNA expression (Fig. 2F).

Identification of the element(s) responsible for the high glucose-induced CYP11B2 transactivation
In order to identify the element(s) responsible for the high glucose-induced CYP11B2 transactivation, we examined the effects of high glucose on the promoter activity of CYP11B2 5 0 -flanking region deletion mutants by comparing the effects between D-glu (450 mgÁdL À1 ) and L-glu (100 mgÁdL À1 D-glu and 350 mgÁdL À1 L-glu) using H295R cells. As shown in Fig. 4A, although high glucose-induced CYP11B2 transactivation was significantly observed in À1521/+2luc, it was not observed in À747/+2-luc, À135/+2-luc, À106/+2-luc, or À65/+2-luc. These data indicate that the region between À1521 and À747 may be responsible for the high glucose effect. As NBRE-1 element [11], which is known to be transactivated by NURR1 and NGFIB [11,14], is located within the region (À766/À759), we next examined the effect of high glucose on the element. As shown in Fig. 4B, point mutation of NBRE-1 element (NBRE-1 mut) completely suppressed the high glucose effect. These data indicate that the element responsible for the high glucoseinduced CYP11B2 transactivation may possibly be the NBRE-1 element.
Effects of NURR1 siRNA on the high glucoseinduced CYP11B2 mRNA expression In order to examine the involvement of NURR1 in the high glucose-induced CYP11B2 mRNA expression, we next transfected either the control or NURR1 siRNA into H295R cells, and thereafter treated the cells with 450 mgÁdL À1 D-glu for 48 h. As shown in Fig. 6A, NURR1 mRNA expression was significantly decreased by the transfection of NURR1 siRNA as compared to that of control siRNA in the presence of either 100 mgÁdL À1 D-glu or 450 mgÁdL À1 D-glu, suggesting the efficient knockdown of endogenous NURR1 mRNA. However, NURR1 knockdown by its siRNA transfection did not affect the high glucose-induced CYP11B2 mRNA expression in comparison with control siRNA transfection (Fig. 6B). These data indicate that other NR4A family members or other transcription factors may be involved in the high glucoseinduced CYP11B2 mRNA expression via the NBRE-1 element.

Effects of 2-deoxy-D-glucose, 3-O-methyl-Dglucose, D-sorbitol, and D-fructose on CYP11B2 mRNA expression
We next examined the involvement of glucose metabolism on the high glucose-induced CYP11B2 expression. When we treated H295R cells with either 2-deoxy-D-glu, which could be phosphorylated but could not be metabolized further [15,16], or 3-Omethyl-D-glu, which could not be phosphorylated [16], the induction of CYP11B2 mRNA expression was not observed (Fig. 7). These data suggest that Dglu metabolization may be more necessary for the induction than glucose 6-phosphate. Moreover, incubation with D-sorbitol or D-fructose, both of which are D-glu metabolites via the polyol pathway [17], did not affect CYP11B2 mRNA expression (Fig. 7), indicating that the pathway may not be involved in the induction.

Effects of ARBs and CCBs on the high glucoseinduced CYP11B2 transcription
We next examined the effects of ARBs and CCBs on the high glucose-induced CYP11B2 transcription. As shown in Fig. 8A, each ARB at 1 lmolÁL À1 did not affect the high glucose effect. In contrast, each CBB dose dependently inhibited the high glucose-induced CYP11B2 transactivation (Fig. 8B). It is therefore suggested that calcium channels, but not AII type 1 receptor, are involved in the high glucose effect.

Discussion
In this study, we first demonstrated the stimulatory effect of high glucose on CYP11B2 transcription and mRNA expression as well as aldosterone secretion in human adrenal cells (Figs 1 and 3A). The high glucose-induced CYP11B2 mRNA expression was not observed when we used glucose analogs, 2-deoxy-D-glu and 3-O-methyl-D-glu [15,16] (Fig. 7), suggesting that it is necessary for D-glu to be metabolized within the cells for the stimulatory effect. As we also observed the high glucose-induced StAR mRNA expression (Fig. 2E), increased StAR and CYP11B2 may coordinately induce aldosterone production. Aldosterone not only induces hypertension and vascular damage in combination with sodium [5], but is also known to inhibit glucose-induced insulin secretion in pancreatic ß-cells [18] as well as insulin signaling in peripheral tissues [19,20]. Therefore, the high glucose-induced aldosterone may induce 'a vicious cycle' in terms of the exacerbation of glucose intolerance/diabetes mellitus. Although the plasma aldosterone concentration in patients with diabetes has long been controversial, it has recently been confirmed to be significantly higher than that of normal subjects by fixing sodium/potassium intake and the time for drawing blood samples [21]. Therefore, based on our present observation, high glucose-induced aldosterone production may possibly contribute to the increased plasma aldosterone level in patients with diabetes.
Interestingly, high glucose was demonstrated to suppress SF-1 mRNA expression, although not significantly (Fig. 5C). As SF-1 is known to suppress CYP11B2 transcription [23,24], the high glucosemediated SF-1 decrease may also contribute to the induction of CYP11B2 transactivation. The mechanisms by which high glucose induces NURR1 mRNA expression remain uncertain. AII and potassium are two major factors that regulate CYP11B2 transcription [14,25]. AII is known to bind to AII type 1 receptor and activate phospholipase C to increase inositol 1,4,5trisphosphate (IP 3 ), and IP 3 induces the release of intracellular calcium from the endoplasmic reticulum, while potassium causes depolarization of the membrane allowing extracellular cytoplasmic calcium influx through the T-and L-type calcium channels [14,25]. In both cases, increased calcium leads to the activation of calcium/calmodulin-dependent kinase, resulting in the induction of NURR1 mRNA expression [14,25]. When we treated the stable CYP11B2-H295R cells with several ARBs, the high glucose-induced CYP11B2 transactivation was not affected (Fig. 8A). In contrast, when we treated the cells with several CCBs, they, especially benidipine that blocks both T-and L-type calcium channels [26], dose dependently inhibited the high glucose-induced CYP11B2 transactivation (Fig. 8B). These data indicate that high glucose may affect at least pathway(s) mediated via T-and/or Ltype calcium channels, but not pathway(s) mediated via AII type 1 receptor. Interestingly, we have recently observed the high glucose-induced mRNA expression of T-type calcium channel subunits (CaV3.1, CaV3.2, and CaV3.3) [27], which may also be involved in the high glucose effect. Further studies are needed to clarify the precise molecular mechanisms of the high glucose-induced CYP11B2 transactivation.
In summary, we here demonstrated high glucoseinduced CYP11B2 transcription and mRNA expression as well as aldosterone secretion via NURR1 induction. As our observation provides a novel insight into the etiology of hypertension in patients with diabetes, it may also lead to novel therapeutics, such as an inhibitor of CYP11B2 transcription, for patients with diabetes complicated with hypertension.