1α,25(OH)2D3 attenuates IL-6 and IL-1β-mediated inflammatory responses in macrophage conditioned medium-stimulated human white preadipocytes by modulating p44/42 MAPK and NF-κB signaling pathways

Background Metabolic syndrome is characterized by macrophage infiltration and inflammatory responses—metaflammation in adipose tissue. IL-6 and IL-1β could mediate the inflammatory responses in macrophage stimulated-preadipocytes by modulating MAPK and NF-κB pathways. To test this hypothesis we used antibodies to block IL-6 and IL-1β action in macrophage conditioned medium (MacCM)-stimulated human white preadipocytes. Moreover, as interventions that prevent this could potentially be used to treat or prevent metabolic syndrome, and 1α,25(OH)2D3 has previously been reported to exert an anti-inflammatory action on macrophage-stimulated adipocytes, in this study we also investigated whether 1α,25(OH)2D3 could attenuate inflammatory responses in MacCM-stimulated preadipocytes, and explored the potential anti-inflammatory mechanisms. Methods Human white preadipocytes were cultured with 25% MacCM for 24 h to elicit inflammatory responses. This was confirmed by measuring the concentrations and mRNA levels of major pro-inflammatory factors [IL-1β, IL-6, IL-8, monocyte chemoattractant protein (MCP)-1 and regulated on activation, normal T cell expressed and secreted (RANTES)] by ELISA and qPCR, respectively. IL-6 and IL-1β actions were blocked using IL-6 antibody (300 ng/ml) and IL-1β antibody (15 μg/ml), respectively. Potential anti-inflammatory effects of 1α,25(OH)2D3 were investigated by pre-treatment and treatment of 1α,25(OH)2D3 (0.01 to 10 nM) for 48 h in MacCM-stimulated preadipocytes. In parallel, western blotting was used to determine inflammatory signaling molecules including relA of the NF-κB pathway and p44/42 MAPK modified during these processes. Results MacCM enhanced the secretion and gene expression of IL-1β, IL-6, IL-8, MCP-1 and RANTES by increasing the phosphorylation levels of relA and p44/42 MAPK in preadipocytes, whereas blocking IL-6 and IL-1β action inhibited the inflammatory responses by decreasing p44/42 MAPK and relA phosphorylation, respectively. Furthermore, 10 nM of 1α,25(OH)2D3 generally inhibited the IL-6 and IL-1β-mediated inflammatory responses, and reduced both p44/42 MAPK and relA phosphorylation in MacCM-stimulated preadipocytes. Conclusions 1α,25(OH)2D3 attenuates IL-6 and IL-1β-mediated inflammatory responses, probably by inhibiting p44/42 MAPK and relA phosphorylation in MacCM-stimulated human white preadipocytes. Electronic supplementary material The online version of this article (10.1186/s13098-019-0405-2) contains supplementary material, which is available to authorized users.


Background
Metabolic syndrome is a clustering of clinical findings consisting of abdominal obesity, high glucose, high triglycerides, low high-density lipoprotein cholesterol levels, and hypertension [1]. Metaflammation in adipose tissue, characterized by infiltration of macrophages, local gene expression and secretion of pro-inflammatory factors, especially IL-6 and IL-1β [2], is considered a potential factor contributing to development of the metabolic syndrome [1]. Moreover, it is notable that inflammatory signaling molecules including relA (NF-κB p65) and MAPK family members [i.e. p44/42 (ERK1/2)] activated in adipose tissue during metabolic syndrome, could trigger metaflammation and insulin resistance [3,4]. It is therefore important to determine whether IL-6 or IL-1β mediates adipose tissue metaflammation during macrophage infiltration via modulating MAPK or NF-κB pathway, to help further elucidate the pathophysiology of metabolic syndrome.
Attenuating adipose tissue metaflammation promotes insulin sensitivity [5], and might be a potential strategy to treat or prevent metabolic syndrome. Accumulated evidence suggests that macrophages play a key role in influencing the proliferation, survival and differentiation of preadipocytes [6][7][8][9][10], which are essential in maintaining adipose tissue homeostasis [11]. Most importantly they might even induce or aggravate metaflammation by stimulating preadipocytes to express and secrete a variety of cytokines, most of which are pro-inflammatory [12][13][14]. In vitro studies have shown that 1α,25(OH) 2 D 3 exerts anti-inflammatory actions on lipopolysaccharide (LPS) and macrophage-stimulated adipocytes, by reducing the release and gene expression of major pro-inflammatory factors including IL-6, IL-8 and MCP-1 [15,16]. However, the anti-inflammatory effects of 1α,25(OH) 2 D 3 on macrophage-stimulated preadipocytes, remains to be explored in detail.
In this study, we aimed firstly to test whether IL-6 or IL-1β is critical in mediating inflammatory responses in MacCM-stimulated preadipocytes, in terms of proinflammatory gene expression and secretion; and secondly to determine the inflammatory signaling pathways activated by examining the phosphorylation of signaling molecules including relA of the NF-κB pathway and p44/42 MAPK. Finally, we investigated whether 1α,25(OH) 2 D 3 , could attenuate the inflammatory responses in MacCM-stimulated preadipocytes, and explored the potential anti-inflammatory mechanisms.

Preparation of MacCM
The human THP-1 monocytic cell line was kindly provided by Professor Helen R Griffiths (Aston University, UK). Monocytes were cultured in T75 flasks in RPMI-1640 medium (Sigma-Aldrich, UK) supplemented with 10% fetal bovine serum and 100 U/ml penicillin, 100 μg/ml streptomycin, and incubated at 37 °C in 95% air and 5% CO 2 . When the monocyte density reached 1 × 10 6 cells/ml, the differentiation to pro-inflammatory M1 dominant macrophages was induced by 100 nM phorbol 12-myristate 13-acetate (Sigma-Aldrich, UK) for 48 h, and then activated by 1 μg/ml LPS [17] and 1 mM ATP (Sigma-Aldrich, UK) for a further 24 h in RPMI-1640 before medium collection. The MacCM was filtered through 0.22 μm filters and stored at − 80 °C.

Culture of human white preadipocytes
Human white preadipocytes derived from subcutaneous adipose tissue of a female Caucasian subject (BMI 21; age 44 years) were obtained from PromoCell (Germany). The preadipocytes were cultured in T25 flasks then sub-cultured into 12-well plates (seeding density: 5000 cells per cm 2 ) in preadipocyte growth medium (PromoCell, Germany) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.25 μg/ml amphotericin B, and incubated at 37 °C in 95% air and 5% CO 2 .

MacCM stimulation and blockade of IL-6 and IL-1β action
When confluence was reached, preadipocytes (n = 6 wells per group) were cultured for 24 h in preadipocyte growth medium, which was replaced with 25% MacCM (diluted in RPMI-1640) to elicit inflammatory responses, together with IL-6 antibody (300, 350 and 450 ng/ml) or IL-1β antibody (15 μg/ml as previously established [18]) (R&D Systems, UK) to block IL-6 or IL-1β action by neutralization for a further 24 h before cell and supernatant collection. Isotype control mouse IgG at the same concentrations confirmed that non-specific binding did not block inflammatory responses (data not shown). The control received a mock treatment of 25% preadipocyte growth medium (diluted in RPMI-1640).

Measurement of inflammatory responses
Cytokine release of the pooled cell supernatants (n = 6) were screened by Human Cytokine Array Panel A following the manufacturer's instructions (R&D Systems, UK) and Molecular Imager ChemiDoc XRS+ System (Bio-Rad, UK). The results were presented as pixel density relative to the references of the arrays. The secretion levels of major pro-inflammatory factors including IL-1β, IL-6, IL-8, MCP-1 and RANTES were measured in duplicate using human ELISA kits following the manufacturer's instructions (R&D Systems, UK) and SPECTROstar Nano Microplate Reader (BMG LABTECH, Germany). The results were also normalized to total cell protein content (measured by Thermo Scientific Pierce BCA Protein Assay Kit, UK) and presented as μg(cytokine)/mg(cell protein). The mRNA levels of IL-1β, IL-6, IL-8, MCP-1 and RANTES in preadipocytes were measured in duplicate using TaqMan gene expression assays (Applied Biosystems, UK), qPCR core kit (Eurogentec, Belgium) and Stratagene Mx3005P instrument system. The results were normalized to the values of reference gene PPIA [19] and presented as fold changes of Ct value relative to controls using the 2 −ΔΔct formula [20]. The intracellular levels (densities) of inflammatory signaling molecules including relA, phosphorylated relA, p44/42 MAPK, phosphorylated p44/42 MAPK, p38 MAPK and phosphorylated p38 MAPK (New England BioLabs, UK) were measured using western blotting and Molecular Imager ChemiDoc XRS+ System (Bio-Rad, UK). The results were calculated as ratios of phosphorylated relA to relA, phosphorylated p44/42 MAPK to p44/42 MAPK and phosphorylated p38 MAPK to p38 MAPK, normalized to loading control vinculin (Abcam, UK), and presented as fold changes of density relative to controls. All the antibodies used were diluted according to the manufacturer's instructions. Methods of qPCR and western blotting were as previously described [21].

Statistical analysis
Results are presented as mean ± SEM. For statistical analysis, unpaired Student's t test was used for individual group comparisons. One-way ANOVA for independent samples was used, followed by Tukey's test for individual group comparisons. A value of p < 0.05 was regarded as statistically significant. The analysis and presentation were performed using Prism 5 (GraphPad, USA).

IL-6 and IL-1β mediate inflammatory responses in MacCM-stimulated preadipocytes
Infiltration of macrophages into adipose tissue induces metaflammation [2], which may contribute to metabolic syndrome [1]. In the initial experiments, we treated human white preadipocytes with MacCM (25%) to elicit inflammatory responses. Cytokine release was screened using proteome array and the results revealed that (Fig. 1a, b) pro-inflammatory factors including IL-6, IL-8, MCP-1 and RANTES were secreted from MacCM-stimulated preadipocytes, compared to controls. Surprisingly, IL-1β, which is vital in initiating and sustaining metaflammation in adipose tissue [22,23], was not detected. However, (Figs. 1c-g and 2) further ELISA results (unnormalized and normalized) show that besides IL-6, IL-8, MCP-1 and RANTES, the secretion levels of IL-1β were also markedly increased, but relatively lower, possibly too low to be detected by the array. In contrast, the secretion level of serpin was mildly decreased from MacCM-stimulated preadipocytes. Similarly, (Fig. 3) the qPCR results show that the mRNA levels of these pro-inflammatory factors were generally enhanced in MacCM-stimulated preadipocytes, especially IL-1β and IL-6, which were approximately 767 and 404-fold higher, compared to the control.
The plasma level of IL-6 is almost exclusively determined by whole-body adiposity and thus closely associated with metabolic syndrome [14]. Moreover, IL-6 has been found to stimulate the synthesis of C-reactive protein, which is one of the strongest indicators of metabolic risk [24]. Hence, we speculated that IL-6 might mediate inflammatory responses in MacCMstimulated preadipocytes. To test this, we blocked IL-6 action from MacCM-stimulated preadipocytes using IL-6 antibody. In preliminary work, a range of IL-6 antibody doses were tested based on calculations using the manufacturer's instructions, and (Additional file 1: Fig. S1PA and S1PB) the results show that 300 ng/ml was the optimal dose to inhibit cytokine secretion and gene expression. In accordance with this, (Fig. 2) the ELISA results (normalized) show that the secretion of IL-8 was considerably decreased by IL-6 antibody, which was 0.6-fold lower compared with MacCM-stimulated preadipocytes. Likewise, the levels of MCP-1 and RANTES were moderately decreased by 0.3-and 0.2-fold, respectively. Although the concentrations of IL-1β were numerically lower, this did not reach statistical significance. The secretion of IL-6 (and IL-1β later) is not presented, since additional IL-6 (IL-1β) antibodies in the supernatant will affect the accuracy of the ELISA results. Moreover, blocking IL-6 action exerted potent inhibitory effects on cytokine gene expression, (Fig. 3) as the mRNA levels of IL-1β, IL-6, IL-8, MCP-1 and RANTES were all significantly lowered by 0.4-to 0.7-fold. Therefore, IL-6 propagates and maintains inflammatory responses during adipose tissue metaflammation, since blocking IL-6 action inhibits the secretion and gene expression of IL-1β, IL-6, IL-8, MCP-1 and RANTES in MacCM-stimulated preadipocytes. A recent study from our group demonstrated that IL-1β could target preadipocytes to induce adipose tissue metaflammation [25]. Likewise, the current results show that IL-1β is crucial in mediating inflammatory responses in macrophage-stimulated preadipocytes, (Fig. 2) since by blocking IL-1β action from MacCMstimulated preadipocytes with IL-1β antibody (15 μg/ ml), the secretion levels of IL-6 and IL-8 were moderately reduced by 0.5-fold compared with MacCMstimulated preadipocytes. However, the release of MCP-1 and RANTES were not statistically different in the presence of MacCM compared with the control. In parallel, (Fig. 3) blocking IL-1β action markedly decreased the mRNA levels of IL-1β, IL-6, IL-8, MCP-1 and RANTES by 0.4-to 0.7-fold in MacCM-stimulated preadipocytes.

IL-6 and IL-1β increase the phosphorylation of inflammatory signaling molecules in MacCM-stimulated preadipocytes
A possible mechanism for induction of adipose tissue metaflammation is activation of the NF-κB signaling pathway via phosphorylation of transcription factor relA [5]. In our study, (Fig. 6b) MacCM considerably increased the phosphorylation levels of relA by 180%, (Fig. 6a) though the levels of NF-κB were similar compared with the control. In accordance with that, the NF-κB signaling pathway could be activated by IL-1β [5], (Fig. 6b) since the decreased phosphorylation levels of relA were significantly associated with blocking IL-1β action.
The conventional MAPKs including p44/42 and p38, are activated through phosphorylation, and are established as playing an important role in various biological processes, especially inflammation [26]. Firstly, (Fig. 6a and Additional file 2: Fig. S2A) compared to the control, MacCM had no effect on the levels of p44/42 MAPK and p38 MAPK. Secondly, (Fig. 6c) the phosphorylation of p44/42 MAPK were moderately increased by 60% in MacCM-stimulated preadipocytes. In parallel, (Additional file 2: Fig. S2B) MacCM remarkably increased the phosphorylation of p38 MAPK, which was 250% higher than observed in the control. Furthermore, it is interesting to note that IL-6 triggers inflammatory responses by activating the p44/42 MAPK signaling pathway [27], as indicated by our observation that the phosphorylation levels of p44/42 MAPK were significantly decreased by blocking IL-6 action in MacCM-stimulated preadipocytes (Fig. 6c). In addition, (Fig. 6c and Additional file 2: Fig. S2B) show that after antibody blockade of IL-1β, the phosphorylation of p44/42 MAPK and p38 MAPK were slightly lower in MacCM-stimulated preadipocytes, but the difference from control was not statistically significant. Likewise, (Fig. 6b and Additional file 2: Fig. S2B) blocking IL-6 action did not affect the phosphorylation levels of relA or p38 MAPK.

Discussion
The initial results show that in human white preadipocytes, MacCM massively enhanced the secretion and gene expression of major pro-inflammatory factors including IL-1β, IL-6, IL-8, MCP-1 and RANTES, which have been measured as indicators of adipose tissue metaflammation in keeping with published studies [14,24,[28][29][30][31]. Although downstream markers of adipose tissue inflammation such as leptin and adiponectin are of interest [32], these adipokines are usually secreted by mature adipocytes; preadipocytes express and secrete extremely low levels of adiponectin and leptin, which we did test in this study (data not shown). Interestingly, we also found that TNF-α, one of the major pro-inflammatory factors released from adipocytes in metaflammation [4], was not secreted at detectable levels from MacCMstimulated preadipocytes. Hence, it appears that preadipocytes have a unique secretome compared to mature adipocytes during metaflammation. Considered as an adipokine [14], IL-6 is also a multifaceted, pleiotropic cytokine and suggested to play a central role in the development of metabolic syndrome by inducing metaflammation and insulin resistance [33]. Hence, rather than neutralizing IL-1β in MacCM as formally established [18], we blocked IL-6 and IL-1β action during the stimulation, to determine whether either of them could mediate the inflammatory responses in MacCMstimulated preadipocytes. The present study showed that both IL-6 and IL-1β mediate MacCM-stimulated inflammatory responses in preadipocytes by increasing the phosphorylation of relA of the NF-κB pathway and p44/42 MAPK, respectively. Moreover, it might be suggested that IL-6 exerts broader inflammatory effects on preadipocytes, for most of the pro-inflammatory factors (though not IL-1β) were inhibited when IL-6 action was blocked, whilst the chemoattractants MCP-1 and RANTES were not significantly reduced by IL-1β blockade. However, it is clear that IL-1β is also important as its blockade did significantly attenuate the inflammatory responses.
Intriguingly, our results suggest that 1α,25(OH) 2 D 3 could attenuate inflammatory responses in macrophage-stimulated preadipocytes during adipose tissue metaflammation, for 1α,25(OH) 2 D 3 significantly The mRNA levels of pro-inflammatory factors a IL-1β, b IL-6, c IL-8, d MCP-1 and e RANTES were measured from cells by qPCR. Data are shown as mean ± SEM for groups of 6. The results were analyzed using one-way ANOVA with Tukey's post hoc test and confirmed by three independent experiments. A significant difference to control was indicated by ***(p < 0.001); to THP-1-MacCM by # (p < 0.05), ## (p < 0.01) and ### (p < 0.001) inhibited pro-inflammatory secretion from MacCMstimulated preadipocytes. Furthermore, 1α,25(OH) 2 D 3 inhibits the mRNA of the major pro-inflammatory factors in MacCM-stimulated preadipocytes probably by decreasing phosphorylation of relA, which is consistent with other relevant studies that phosphorylated The results were analyzed using one-way ANOVA with Tukey's post hoc test and confirmed by three independent experiments. A significant difference to control was indicated by ***(p < 0.001); to THP-1-MacCM by # (p < 0.05), ## (p < 0.01) and ### (p < 0.001) Fig. 7 1α,25(OH) 2 D 3 attenuates Il-6 and IL-1β-mediated inflammatory responses in MacCM-stimulated human white preadipocytes. Accumulated macrophages infiltrating into white adipose tissue induce metaflammation, which is embodied by inflammatory responses including increased local gene expression and secretion of pro-inflammatory factors, particularly IL-1β, IL-6, IL-8, MCP-1 and RANTES. Among these pro-inflammatory factors, by binding to its receptor, IL-1β initiates and sustains the inflammatory responses in macrophage-stimulated preadipocytes by enhancing the phosphorylation of relA of the NF-κB signaling pathway, while in the same ligand-receptor manner IL-6 mediates similar responses by enhancing the phosphorylation of p44/42 MAPK. Besides directly inhibiting the pro-inflammatory gene expression (not tested in our study), 1α,25(OH) 2 D 3 attenuates IL-1β-mediated inflammatory responses by decreasing the phosphorylation of relA in the nucleus. Moreover, 1α,25(OH) 2 D 3 can exert a non-genomic action by decreasing the phosphorylation of p44/42 MAPK in the cytoplasm, to attenuate IL-6-mediated inflammatory responses