Dopamine D2 receptor upregulates leptin and IL-6 in adipocytes[S]

Leptin is a pro-inflammatory cytokine secreted by the adipose tissue. Dopamine D2 receptors (D2Rs) have anti-inflammatory effects in the brain and kidney tissues. Mouse and human adipocytes express D2R; D2R protein was 10-fold greater in adipocytes from human visceral tissue than subcutaneous tissue. However, the function of D2R in adipocytes is not well understood. 3T3-L1 cells were treated with D2-like receptor agonist quinpirole, and immunoblot and quantitative PCR were performed. Quinpirole increased the protein and mRNA expression of leptin and IL-6, but not adiponectin and visfatin (24 h). It also increased the mRNA expression of TNF-α , MCP1, and NFkB-p50. An acute increase in the protein expression of leptin and TNF-α was also found in the cells treated with quinpirole. The leptin concentration in the culture media was increased by quinpirole-bathing the 3T3-L1 adipocytes. These quinpirole effects on leptin and IL-6 expression were prevented by the D2R antagonist L741,626. Similarly, siRNA-mediated silencing of Drd2 decreased the leptin, IL-6, mRNA, and protein expressions. The D2R-mediated increase in leptin expression was prevented by the phosphoinositide 3-kinase inhibitor LY294002. Acute quinpirole treatment in C57Bl/6J mice increased serum leptin concentration and leptin mRNA in visceral adipocyte tissue but not in subcutaneous adipocytes, confirming the stimulatory effect of D2R on leptin in vivo. Our results suggest that the stimulation of D2R increases leptin production and may have a tissue-specific pro-inflammatory effect in adipocytes.

D 2 R regulates the expression of leptin and IL-6 in a mouse adipocyte cell line.

Human subcutaneous and visceral adipocyte cells
Preadipocytes were isolated from subcutaneous and visceral fat of nonoverweight, nondiabetic donors undergoing elective abdominal surgery. Preadipocytes were cultured and then differentiated into adipocytes, using differentiation medium as described previously (14.

Mice
C57Bl/6J male mice at 20 weeks of age (Jackson Laboratory, Bar Harbor, Me) were housed in the Animal Care Facility of the University of Maryland Baltimore and were studied for the role of D 2 R on the expression and release of leptin in adipocytes in vivo. The mice were treated with quinpirole (1 mg/kg; i.v.) or vehicle; the serum samples and adipocyte tissue (subcutaneous and visceral) were harvested 2 h after injection. All studies were approved by the Animal Care and Use Committee of the University of Maryland Baltimore.

Quantitative real-time PCR
Total RNA was purified using the RNeasy RNA extraction Mini kit (Qiagen, Valencia, CA). RNA samples were converted into first strand cDNA using an RT2 First Strand kit (Qiagen). Gene expression was quantified by real-time PCR performed on an ABI Prism 7900 HT (Applied Biosystems, Foster City, CA). The assay used gene-specific primers (Qiagen) and the SYBR Green realtime PCR detection method (Qiagen). GAPDH was used as a housekeeping gene. The different genes studied were those for leptin, IL-6, TNF, MCP-1, NFB p50, and GAPDH. Data were analyzed using the Ct method (18).

Leptin concentration in the medium
Leptin concentration in cell culture medium was quantified using a commercial Kit (Cell Biolabs, INC). All assays were performed in duplicate and normalized by protein concentration.

Statistical analysis
Data are expressed as mean ± SeM. Comparisons between two groups used the Student's t-test. One-way ANOVA followed by posthoc analysis using the Holm-Sidak multiple comparison test were used to assess significant differences among three or more groups. P < 0.05 was considered statistically significant.

D 2 R and D 3 R are expressed higher in human visceral than subcutaneous adipocytes
The presence of D2R and D3R in primary cultures of human subcutaneous adipocytes and in mouse 3T3-L1 cells was confirmed by immunoblotting (Fig. 1A, B). To compare the relative abundance of D 2 R in visceral and subcutaneous adipocytes, primary cultures of human adipocytes from subcutaneous and visceral fat were studied. qRt-PCR analyses showed that D 2 R mRNA was 10-fold greater in visceral than subcutaneous adipocytes (Fig. 1C).

D 2 R regulates leptin expression via AKT in 3T3-L1 cells
Stimulation of D 2 R by quinpirole (1 M; 24 h) increased phospho AKT and leptin protein expression (+131 ± 36%, P < 0.05, n = 4) (Fig. 4). This is in agreement with the data shown in Fig. 2 where the effects were completely prevented by cotreatment with the reversible PI 3 K inhibitor LY294002 (10 M; 24 h), suggesting that AKT may be involved in the positive regulation of leptin by D 2 R. By contrast, the D 2 R-stimulated increase in IL-6 protein expression (+120 ± 24%, n = 4, P < 0.05) was not prevented completely by LY294002 (+71 ± 23%, n = 4, P < 0.05) (Fig. 4A). We speculate that PI 3 K regulates the de novo synthesis of leptin but not its release. Therefore, the presence of the residual leptin in the medium may have kept the IL-6 expression high even in the presence of PI 3 K inhibitor. To determine whether the presumed residual effect of leptin on IL-6 expression is correct, differentiated 3T3-L1 cells were treated with L39A/D40A/F41 (25 nM, 24 h), a leptin antagonist (17), and quinpirole (1 M, 24 h). Immunoblot analyses show that IL-6 protein expression was decreased in the presence of L39A/D40A/F41, and quinpirole did not increase IL-6 protein expression in the presence of the leptin antagonist (31 ± 6%, n = 5, P < 0.05) (Fig. 4C), suggesting that the presence of leptin in the medium may have been responsible for the inability of the PI 3 K inhibitor (LY294002) to block the D 2 R stimulatory effect on IL-6 expression (Fig. 4A). The presence of leptin receptor (OB-R) in 3T3-L1 cells was confirmed by immunoblotting. Its expression was not altered by D 2 R stimulation (Fig. 4B).

Dopamine decreases leptin expression via the adrenergic receptors in 3T3-L1 cells
It has been reported that dopamine suppresses leptin release in 3T3-L1 cells and human subcutaneous adipocyte tissue (10)(11)(12). However, our data show that D 2 R increases leptin expression (Figs. 2, 4, and supplementary Fig. S1). Therefore, to address the discrepancies between our results and those of others, we treated 3T3-1L cells with dopamine in the presence of D 1 R antagonist, D 2 R antagonist, -adrenergic receptor antagonist, or -adrenergic receptor antagonist, with or without insulin (Fig. 5). We found that dopamine (1 uM, 24 h) decreased leptin expression (33 ± 3% vs. control) but the D 1 R antagonist SCH 23390 (10 µM) did not prevent the inhibitory effect of dopamine on leptin expression. However, the presence of the D 2 R antagonist L741,626 enhanced the inhibitory effect of dopamine on leptin expression (84 ± 4% vs. dopamine alone), suggesting that the D 2 R increases leptin expression. The -adrenergic antagonist propranolol (10 µM) did not clearly affect the inhibitory action of dopamine on leptin expression. By contrast, the -adrenergic antagonist phentolamine (10 µM) prevented the inhibitory effect of dopamine on leptin expression. This finding, however, is somewhat different from that reported by Than et al., who showed that the inhibitory effect of dopamine (1 M) on leptin secretion were blocked by propranolol but not by phentolamine or the D 1 -like (SCH23390) or D 2 -like (haloperidol) receptor antagonists (11). We also found that dopamine increased leptin secretion in 3T3-L1 cells in the presence of both and -adrenergic receptor antagonists (dopamine 1 M + propranolol 10 µM + phentolamine 10 µM. 24 h, +37 ± 13% vs. vehicle), indicating that both adrenergic receptors may be involved in the effect of dopamine on leptin expression (23, 24,). However, dopamine did not show any effect on leptin expression in insulin-free medium (vehicle, insulin-free medium, 24h, 72 ± 8% vs. control; dopamine, medium without insulin; 24 h, 78 ± 4% vs. control), indicating that the effect of dopamine on leptin is insulin-dependent.

D 2 R stimulation increases leptin expression and release in vivo
To determine whether D 2 R affects leptin expression in vivo, C57Bl/6 mice were treated by quinpirole or vehicle; the serum leptin concentration was greater in the quipirole group (vehicle 1,094 ± 132 pg/ml vs. quinpirole 2,579 ± 599 pg/ml) than in the vehicle group 2 h after treatment (Fig. 6A). Subcutaneous and visceral tissues were collected before and after treatment and no differences were found between groups before treatment on leptin mRNA expression (data not shown). However, quinpirole increased leptin mRNA expression (Fig. 6B) in visceral adipocyte tissue (1,200 ± 366, % of control), but not in subcutaneous adipocyte tissue. These differences may be due to low D 2 R expression in subcutaneous tissue (Fig. 1) (10). IL-6 mRNA expression was also determined in the subcutaneous and visceral adipocyte tissue of mice treated with quinpirole. The expression of IL-6 did not differ between groups (Fig.  6C). Therefore, quinpirole did not increase IL-6 expression in vivo, contrary to that found in the 3T3-L1 cell line. This difference may be because leptin in the medium increases IL-6 expression via OB receptors stimulation in 3T3-L1 cells as has been demonstrated in Fig. 4.

DISCUSSION
We and others have reported that the D 2 R has antiinflammatory properties in the kidney (15,16,25) and , and -adrenergic receptor antagonist (phentolamine, 10 µM) in medium with or without insulin. Protein expression of leptin in the medium was determined by eLISA and normalized by protein concentration, one-way ANOVA; *P < 0.05, control vs. others; # P < 0.05, vs. others; Student's t-test; †P < 0.05 vs. control, n = 6/11. brain (26). However, the function of the D 2 R in adipocytes is not clear. We now report that the D 2 R and D 3 R are expressed in mouse 3T3-Ll adipocytes, where the D 2 R upregulates the mRNA and protein expression of leptin and IL-6 via the PI 3 K/AKT pathway. The stimulatory effect of the D 2 R on leptin expression is dependent upon the presence of insulin.
It has been reported that D 2 -like receptors decreased the expression of adipokines such as leptin, Il-6, TNF, and adiponectin (27). These observations were consistent with the concept of the existence of unique leptin-dopamine interactions in the hypothalamus and the hyposensitivity of the dopamine system in obesity, and may provide indirect evidence for an inhibitory effect of dopaminergic neurotransmission on leptin secretion via autonomic nerves. It has been described that the central part of the autonomic nervous system and intra-abdominal and subcutaneous fat pads are innervated by separate sympathetic and parasympathetic motor neurons (28). Moreover, some studies have shown that bromocriptine, a D 2 -like receptor agonist, decreases the leptin concentration in serum, likely via hypothalamic action (27,29); therefore, systemic stimulation of D 2 R could decrease leptin expression. In contrast, Drd2 / mice have lower serum leptin concentration compared with their wild-type littermates (30), indicating that additional mechanisms may be involved in the regulation of leptin by D 2 R. By contrast, female Drd3 / mice have increased serum leptin concentration; a high-fat diet increases serum leptin concentration in both male and female Drd3 / mice (31). Whether or not this dynamic regulation between dopamine receptors and leptin is organ-specific remains to be determined (32,33).
Recently, other authors have proposed the hypothesis that the interaction between adipokines and dopamine also occurs in the adipocyte tissue. Brown, but not white, adipocytes (23) express dopamine receptors in some studies (34). One main question is, if dopamine receptors exert an effect on adipocyte regulation, what is the source of the dopamine in adipocyte tissue? It has been shown that human adipocytes possess arylsulfatase A, and circulating dopamine sulfate may serve as the source of dopamine in adipocytes (10).
Dopamine has been shown to decrease leptin expression in human subcutaneous adipocyte tissue (10,12). 3T3-L1cells may also express D 2 R (35) (Fig. 1). Dopamine and some dopamine receptor agonists are also able to stimulate and -adrenergic receptors in the cardiovascular system and in adipocytes (11,23,24,35,36) and this effect in vivo is observed at a higher dopamine dose in humans (37). Dopamine also suppresses leptin release in 3T3-L1 cells but the effect is blocked, not by the dopamine receptor antagonists, but by propranolol, indicating the involvement of -adrenergic receptors (11). White and brown adipose tissue of Sprague-Dawley rats can synthesize catecholamines (38).
Our data show that the D 2 R agonist increases leptin and IL-6 expression in 3T3-L1 cells. However, consistent with previous studies, our data show that dopamine decreases leptin expression mainly via -adrenergic receptors in  3T3-L1 cells, while it increases leptin expression when  and adrenergic receptors are blocked by their antagonist (Fig. 5). Thus, the ability of dopamine to decrease leptin in subcutaneous adipocytes may be by the stimulation of adrenergic receptors, independent of dopamine receptors. Human D 2 R is expressed to a greater extent in visceral adipocytes than subcutaneous adipocyte tissue (Fig. 1), consistent with previous reports (10).
Dopamine and the D 1 -like receptor agonist SKF38393 decrease leptin release but increase adiponectin and IL-6 release in human adipocyte explants and differentiated primary adipocytes (10). Our data show that the D 1 R antagonist failed to block the dopamine effect on leptin expression (Fig. 5), suggesting that this action is independent of D 1 R, and that D 5 R may be involved in leptin regulation. SKF38393 can also stimulate adrenergic receptors (36), indicating that the capacity of SKF38393 to downregulate leptin expression may be independent of the dopamine receptors.
By contrast, Nakano et al. (13) showed that dopamine causes inflammation, inducing the expression of IL-6 and IL-17. The effect is blocked by the D 1 R antagonist in human primary cultures from synovial tissues, indicating that the possible pro-inflammatory effects of dopamine receptors may be an alternative target in inflammatory disease (39). In addition, a study of human subcutaneous adipocyte tissue have shown that dopamine receptors increase the expression of IL-6 (10).
Leptin increases pro-inflammatory cytokines such as IL-6 (2). Our data show that IL-6 is upregulated by D 2 R stimulation, which increases leptin release into the medium that is blocked by an OB-R receptor antagonist. Hence, leptin in the medium may increase the expression of pro-inflammatory adipokines via paracrine and autocrine action in 3T3-L1 cells via OB-R stimulation (40) (Fig. 7). Our in vivo data confirm that stimulation of D 2 R with quinpirole increases leptin synthesis from visceral adipocyte tissue because there is a remarkable increase in its mRNA expression. In addition, stimulation of D 2 R may also increase leptin release into the blood because the increase in serum leptin concentration occurred within 2 h of the treatment. Therefore, the stimulatory effect of D 2 R on leptin is confirmed in vitro and in vivo, which may have pro-inflammatory effects in adipocytes.
Taking all this into account, the physiological role of D 2 R in adipocytes in the regulation of leptin in vivo is complicated; hypothalamic D 2 R could be also involved in the leptin regulation (28). Because leptin could contribute to the inflammatory effect of visceral adipocyte tissue, we speculate that the stimulatory effect of D 2 R on leptin may have special relevance in the study of insulin-resistant diabetic patients or in patients under D 2 R agonist treatment. D 2 R-altering drugs are prescribed to patients with nervous and psychiatric disorders, hyperprolactinemia, or in the early phase of Parkinson's disease. This is the first report that shows that D 2 R in adipocytes could have pro-inflammatory effects. Visceral adipocytes have also been reported to increase the expression of pro-inflammatory adipokines that could alter liver triglyceride metabolism and increase the development of arteriosclerosis (5). However, recent reports have shown that pro-inflammatory responses in adipocyte tissue are essential for proper extracellular matrix remodeling and angiogenesis (41). Therefore, whether this mechanism has significant physiological benefits or deleterious consequences must be determined. The understanding of the action of these drugs and undesirable side-effects, especially in viscerally obese patients, may be improved by considering their ability to directly affect adipocyte function.