Blockade of multiple monoamines receptors reduce insulin secretion from pancreatic β-cells

Clinical use of olanzapine frequently causes severe hyperglycemia as an adverse effect. In this study, we elucidated mechanisms by which olanzapine reduced insulin secretion using the hamster pancreatic β-cell line HIT-T15. Reverse transcriptional-PCR analysis revealed expression of dopamine (D2, D3 and D4), serotonin (5-HT2A, 5-HT2B, 5-HT2C, and 5-HT6), and histamine (H1 and H2) receptors in HIT-T15 cells. Olanzapine decreased insulin secretion from HIT-T15 cells at clinically relevant concentrations (64–160 nM). A dopamine D2 agonist, D3 antagonist, and D4 antagonist suppressed insulin secretion, whereas a D2 antagonist and D3 agonist increased it. A serotonin 5-HT2B agonist slightly increased insulin secretion, while a 5-HT2C antagonist slightly decreased it. Other agonists and antagonists for serotonin receptors did not affect insulin secretion. A histamine H1 agonist increased insulin secretion, whereas an H1 antagonist and H2 agonist suppressed it. Our results suggest that dopamine (D2, D3 and D4), serotonin (5-HT2B and 5-HT2C), and histamine (H1 and H2) receptors, which are expressed on pancreatic β-cells, directly modulate insulin secretion from pancreatic β-cells. Thus, olanzapine may induce hyperglycemia in clinical settings by suppressing insulin secretion from pancreatic β-cells through inhibition of dopamine D3, serotonin 5-HT2B and 5-HT2C, and histamine H1 receptors.


Effects of dopamine receptor agonists and antagonists on insulin secretion from HIT-T15 cells.
The involvement of dopamine receptors in insulin secretion was evaluated using HIT-T15 cells ( Fig. 3). Dopamine decreased insulin secretion in a concentration-dependent manner (Fig. 3A), consistent with a previous report 13 . Bromocriptine, a dopamine D 2 receptor agonist, decreased insulin secretion in a concentration-dependent manner and this decrease remained constant (~40% of control) at dosages over 100 nM (Fig. 3B). In contrast, haloperidol, a dopamine D 2 receptor antagonist, increased insulin secretion (Fig. 3C). For the dopamine D 3 receptor, the agonist 7-hydorxy PIPAT significantly enhanced insulin secretion (Fig. 3D), whereas the antagonist NGB2904 suppressed it (Fig. 3E). Furthermore, both the dopamine D 4 receptor agonist ABT724 and antagonist sonepirazole significantly suppressed insulin secretion to ~80% of control (Fig. 3F,G). These findings suggest that stimulation of dopamine D 2 and D 4 receptors decreased insulin secretion from pancreatic β-cells, whereas stimulation of D 3 receptors increased it.

Effects of serotonin receptor agonists and antagonists on insulin secretion from HIT-T15 cells.
The involvement of serotonin receptors in insulin secretion was evaluated using HIT-T15 cells (Fig. 4).
Insulin secretion was decreased to ~80% of control by serotonin in a concentration-dependent manner (Fig. 4A). Neither the serotonin 5-HT 2A receptor agonist TCB2 nor antagonist MDL11939 had an effect on insulin secretion (Fig. 4B,C). The serotonin 5-HT 2B receptor agonist BW723C86 slightly increased insulin secretion (Fig. 4D), while the antagonist SB204741 exerted no significant effects on secretion (Fig. 4E). Stimulation of serotonin 5-HT 2C receptors by agonist Ro60-0175 did not affect insulin secretion, although blockade of serotonin 5-HT 2C receptors by antagonist SB242084 slightly decreased secretion (Fig. 4F). Similar to the serotonin 5-HT 2A receptor, neither the serotonin 5-HT 6 receptor agonist WAY181187 or antagonist SB399885 had an effect on insulin secretion (Fig. 4G,H). These results suggest that stimulation of serotonin 5-HT 2B and 5-HT 2C receptors can increase insulin secretion from pancreatic β-cells, although their contributions are much lower than those of dopamine receptors.

Effects of histamine receptor agonists or antagonists on insulin secretion from HIT-T15 cells.
We evaluated the involvement of histamine receptors in insulin secretion from HIT-T15 cells (Fig. 5). Histamine decreased insulin secretion at 1-100 nM (Fig. 5A). The histamine H 1 receptor agonist 2-pyridylethylamine (2-PEA) increased insulin secretion in a concentration-dependent manner, whereas trans-triprolidine, a histamine H 1 receptor antagonist, decreased secretion to ~40% of control (Fig. 5B,C). In contrast, the histamine H 2 receptor agonist amthamine decreased insulin secretion to ~75% of control in a concentration-dependent manner. In addition, the histamine H 2 receptor antagonist tiotidine slightly decreased  6 ), and histamine (H 1 and H 2 ) receptors in HIT-T15 cells. Total RNA extracted from HIT-T15 cells was reversetranscribed, and first-strand cDNA was synthesized. Target genes were amplified with a set of specific primers (shown in Table 2). For expression analysis of dopamine D 3 and D 4 , all serotonin, and histamine H 2 receptors, two-step PCR was performed using nested primers (shown in Table 2). PCR products were separated by electrophoresis using a 2% agarose gel and stained with ethidium bromide. M, 100-bp ladder size marker.

Olanzapine nor receptor agonists or antagonists affected the viability of HIT-T15 cells.
To evaluate whether cytotoxic effects of olanzapine and agonists or antagonists for each receptor occurred, an XTT assay was performed. As shown in Table 1, 1-h exposure of each agent tested did not affect the viability of HIT-T15 cells, indicating that alterations in insulin secretion from HIT-T15 cells induced by these compounds were not the result of cytotoxicity.

Discussion
Several reports have shown that administration of olanzapine could induce hyperglycemia 14,15 . As olanzapine exerts an antipsychotic effect by inhibiting multiple receptors for dopamine, serotonin, histamine, adrenaline, and acetylcholine, we hypothesized that olanzapine could increase hyperglycemia by suppressing insulin secretion from pancreatic β-cells through blockade of multiple receptors. In this study, we investigated the involvement of dopamine, serotonin, and histamine receptors in insulin secretion using HIT-T15 cells.
The effects of olanzapine on plasma insulin levels are incompletely understood because doses of olanzapine used in previous animal experiments were higher than used in clinical settings. Nagata et al. 10 reported that serum concentrations of insulin increased following a single intravenous infusion of olanzapine at a dose of 2.5-10 mg/ HIT-T15 cells were incubated with medium containing 1% dimethylsulfoxide (control) or olanzapine for 1 h at 37 °C. Concentrations of insulin released into the medium were determined using a rat Insulin ELISA kit. Amounts of insulin secretion were normalized to the total protein content of each well. Each value represents mean ± SD of eight trials. ***P < 0.001 with respect to control.  13 reported that olanzapine increased insulin secretion from human islets at concentrations of 1-5 µM. In clinical settings, olanzapine is administered orally at a dose of 5-20 mg daily, yielding reportedly therapeutic serum concentrations of 20-50 ng/mL (64-160 nM) 12 . In this study, olanzapine decreased insulin secretion from HIT-T15 cells by ~20% compared with controls at concentrations of 1-1000 nM (Fig. 2). Our preliminary study showed that olanzapine increased insulin secretion from HIT-T15 Concentrations of insulin released into the medium were determined using a rat Insulin ELISA kit. Amounts of insulin secretion were normalized to the total protein content of each well. Each value represents mean ± SD of four to eight trials. ***P < 0.001 with respect to control. (2019) 9:16438 | https://doi.org/10.1038/s41598-019-52590-y www.nature.com/scientificreports www.nature.com/scientificreports/ cells at concentrations of 10-30 µM (data not shown). These findings suggest that olanzapine directly suppresses insulin secretion from pancreatic β-cells at clinical concentrations. Blood glucose levels were reportedly altered by 20% when insulin secretion was altered by 25% after feeding 16 , suggesting that olanzapine can induce hyperglycemia by suppressing insulin secretion from pancreatic β-cells at clinical concentrations. Concentrations of insulin released into the medium were determined using a rat Insulin ELISA kit. Amounts of insulin secretion were normalized to the total protein content of each well. Each value represents mean ± SD of eight trials. *P < 0.05, **P < 0.01, ***P < 0.001 with respect to control. (2019) 9:16438 | https://doi.org/10.1038/s41598-019-52590-y www.nature.com/scientificreports www.nature.com/scientificreports/ We demonstrated that dopamine D 2 , D 3 , and D 4 ; serotonin 5-HT 2A , 5-HT 2B , 5-HT 2C , and 5-HT 6 ; and histamine H 1 and H 2 receptors are expressed by HIT-T15 cells (Fig. 1). A few previous studies reported expression of these receptors in human pancreas [17][18][19] . Rubí et al. 17 reported the detection of dopamine D 2 and D 4 receptor mRNAs in human islets. Bonhaus et al. 18 observed mRNAs for serotonin 5-HT 2A , 5-HT 2B , and 5-HT 2C receptors in human pancreas. Although no previous studies reported expression of histamine H 1 or H 2 receptors in human pancreas, Szukiewicz et al. 19 reported protein expression of these receptors in pancreatic β-like cells differentiated from human amniotic epithelial cells by nicotinamide treatment. Thus, the involvement of these receptors in insulin secretion may be observed even in human pancreas.
To evaluate the involvement of dopamine, serotonin, and histamine receptor subtypes in insulin secretion, we examined the effects of agonists and antagonists specific for each receptor subtype on insulin secretion from HIT-T15 cells. With regard to dopamine receptors, stimulation of the dopamine D 2 receptor suppressed insulin secretion, whereas its blockade enhanced it (Fig. 3B,C). This result is consistent with a previous report that blockade of dopamine D 2 receptor enhanced insulin secretion from human islets 13 . In contrast, stimulation of the dopamine D 3 receptor enhanced insulin secretion, whereas its blockade enhanced secretion (Fig. 3D,E). Thus, olanzapine can suppress insulin secretion via blockade of the dopamine D 3 receptor. Insulin secretion was increased by either stimulation or blockade of the dopamine D 4 receptor (Fig. 3F,G) using specific dopamine D 4 agonist ABT724 (EC 50 value for rat dopamine D 4 receptor is 12.4 nM and that for dopamine D 2 is >10 µM) or antagonist sonepiprazole (K i value for human dopamine D 4 receptor is 10 nM and those for other monoamine receptors are >2 µM), respectively, which seems contradictory 20,21 . However, these findings can be explained by differences in expression levels between dopamine D 2 and D 3 receptors. Dopamine can stimulate both dopamine HIT-T15 cells were incubated with medium containing 1% dimethylsulfoxide (control), histamine (A), 2-pyridylethylamine (2-PEA) (B, H 1 agonist), trans-triprolidine (C, H 1 antagonist), amthamine (D, H 2 agonist), or tiotidine (E, H 2 antagonist) for 1 h at 37 °C. Concentrations of insulin released into the medium were determined using a rat Insulin ELISA kit. Amounts of insulin secretion were normalized to the total protein content of each well. Each value represents mean ± SD of eight trials. ***P < 0.001 with respect to control. D 2 and D 3 receptors when the dopamine D 4 receptor is blocked. Expression levels of dopamine D 2 receptor were higher than those of dopamine D 3 receptor, as expression of mRNA for dopamine D 3 could be detected by two-step PCR with nested primers (Fig. 1). We considered decreased insulin secretion via blockade of the dopamine D 4 receptor to arise from stimulation of the dopamine D 2 receptor.
For serotonin receptors, stimulation of the 5-HT 2B receptor slightly increased insulin secretion from HIT-T15 cells, whereas blockade of the 5-HT 2C receptor decreased secretion (Fig. 4C,F). In contrast, a 5-HT 2B antagonist, 5-HT 2C agonist, and both an agonist and antagonist of 5-HT 2A and 5-HT 6 did not affect insulin secretion. Bennet et al. 22 demonstrated that stimulation of 5-HT 2B receptor increased the glucose-stimulated insulin secretion from mouse and human pancreatic β-cells by triggering downstream changes in cellular Ca 2+ flux that enhance mitochondrial metabolism. These findings suggest that serotonin 5-HT 2B and 5-HT 2C receptors can modulate insulin secretion from β-cells. Thus, inhibition of serotonin 5-HT 2B and 5-HT 2C receptors may be involved in olanzapine-reduced insulin secretion, although their contributions may be less than those of dopamine receptors. Interestingly, there is one previous reports showing that 5-HT 3 receptor-mediated insulin secretion was further enhanced in pregnant mice compared to that in normal mice 23 . Thus, it is possible that the contributions of serotonin receptors subtypes to insulin secretion are altered under diseased states.
We also evaluated the involvement of histamine receptors in insulin secretion (Fig. 5). Histamine decreased insulin secretion at 1-100 nM (Fig. 5A). Stimulation of the histamine H 1 receptor increased insulin secretion, whereas stimulation of the histamine H 2 receptor decreased it (Fig. 5B,D). Thus, insulin secretion from pancreatic β-cells can be modulated by both histamine receptor subtypes, and olanzapine can suppress insulin secretion via blockade of the histamine H 1 receptor.
The roles of endogenous monoamines in the insulin secretion have not been understood completely. Ustione and Piston 24 reported that dopamine was secreted from pancreatic β-cells simultaneously with insulin and caused negative feedback inhibition on insulin secretion. In agreement with their report, we also showed that physiological levels of dopamine suppressed the insulin secretion from pancreatic β-cells (Fig. 3). In contrast, there have been no reports regarding the roles of endogenous serotonin and histamine on insulin secretion. In this study, we used dopamine, serotonin and histamine at concentrations of 0.1-100 µM, 0.1-100 µM and 1-100 nM, respectively. These concentrations of dopamine and serotonin are higher than those in physiological concentrations which in human plasma are reportedly ~6.5 nM 25 and ~0.6 pM 26 , respectively. Thus, further studies are necessary to understand the role of endogenous monoamines in the insulin secretion at physiological conditions.
In this study, we did not confirm the expression and involvement of muscarinic acetylcholine receptors on insulin secretion from HIT-T15 cells. Iismaa et al. 27 reported that muscarinic acetylcholine receptors were expressed on rat pancreatic β-cells. Furthermore, Henquin et al. 28 reported that insulin secretion was enhanced via stimulation of muscarinic receptors. Because olanzapine inhibits the muscarinic acetylcholine receptors, it is speculated that olanzapine can suppress the insulin secretion from pancreatic β-cells via blockade of muscarinic acetylcholine receptors as well as monoamine receptors. Further studies are necessary to clarify the involvement of muscarinic acetylcholine receptors in insulin secretion from pancreatic β-cells.
In conclusion, we demonstrated that olanzapine suppressed insulin secretion from pancreatic β-cells via blockade of dopamine D 3 , serotonin 5-HT 2B and 5-HT 2C , and histamine H 1 receptors at clinical concentrations in vitro. Although further studies are necessary using human pancreatic β-cells for in vitro and in vivo animal studies, these findings shed new light on the mechanisms underlying olanzapine-induced hyperglycemia.
Cell culture. HIT-T15 cells were obtained from Sumitomo Dainippon Pharma (Osaka, Japan). Cells were cultured in Ham's F12K medium (Sigma-Aldrich) containing 10% fetal bovine serum, 100 units/mL penicillin G, 100 µg/mL streptomycin, and 10 mM glucose which corresponds to the physiological blood concentrations in human in an atmosphere of 5% CO 2 /95% air at 37 °C. Cells were subcultured once a week using 0.25% EDTA and 0.038% trypsin. Fresh medium was replaced every 2 days. Cells were used between passages 80 and 100.
RT-PCR analysis. Total RNA was extracted from HIT-T15 cells using an RNeasy Plus Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Next, total RNA was used for reverse transcription to synthesize cDNA using a ReverTra Ace qPCR RT kit (Qiagen). PCR was performed with an iCycler (Bio-Rad Laboratories, Hercules, CA, USA) using KOD-Plus-DNA polymerase (Toyobo, Osaka, Japan). Conditions for PCR were as follows: initial denaturation at 94 °C for 2 min; denaturation at 94 °C for 30 sec; annealing at optimal temperatures for dopamine, serotonin, and histamine receptors for 30 sec; and extension at 68 °C for 1 min (35 cycles). Primers, annealing temperatures, and product sizes for each receptor are summarized in Table 2. To examine expression of mRNA for dopamine D 3 and D 4 receptors, and all serotonin receptors, we performed two-step PCR with nested primers due to their lower expression in HIT-T15 cells. Nested primers for each receptor are summarized in Table 2. Conditions for the second round of PCR were the same as those for the first round. PCR products were electrophoresed with a 2% agarose gel and visualized under ultraviolet light with ethidium bromide. www.nature.com/scientificreports www.nature.com/scientificreports/ Insulin secretion assay. Insulin secretion assays were performed according to previous reports 29,30 . Briefly, HIT-T15 cells were seeded at a density of 1.0 × 10 5 cells/well in 24-well plates and cultured for 72 h after seeding. Next, cells were pre-incubated with fresh medium containing 1% dimethylsulfoxide (DMSO) for 30 min at 37 °C. After pre-incubation, cells were incubated with fresh medium for 1 h at 37 °C. To examine the effects of olanzapine or agonists/antagonists for each receptor on insulin secretion, each compound was added to the medium at various concentrations during incubation. Compounds tested are shown in Table 3. After incubation, the concentration of insulin released into the medium was determined using a rat Insulin ELISA kit (Morinaga Institute of Biological Science, Yokohama, Japan) according to our previously reported method 31,32 . Next, residual cells were   Table 3. Agonists and antagonists specific for dopamine, serotonin, or histamine receptors used in this study.