Basic Fibroblast Growth Factor is Involved in Bisphenol S Induced Proliferation of Hemangioma Cells

The hyperproliferation of mesoblastic vascular tissues can lead to the incidence of hemangioma (IHA), which is the most common benign tumor in infants. Estrogenic signals can trigger the progression of HA via activation of gene transcription. We found that bisphenol S (BPS), one widely spread endocrine disrupting compound (EDC), can induce the proliferation of HA cells and trigger the G1 to S transition of cell cycle. Among the tested cytokines, BPS can up regulate of basic broblast growth factor (bFGF). Targeted inhibition of bFGF via its neutralization antibody can reverse BPS induced cell proliferation. Mechanistically, BPS can increase the mRNA expression of bFGF via increasing the transcription and mRNA stability. The activation of p65 and down regulation of miR-155-5p were responsible for BPS induced transcription and mRNA stability of bFGF, respectively. Conclusions: BPS can increase the expression of bFGF via activation of p65 and down regulation of miR-155-5p, which resulted in the proliferation of HA cells.


Background
Infantile hemangioma (IHA), caused by hyperproliferation of mesoblastic vascular tissues, is the most common benign tumor in infants at present (Mabeta and Pepper, 2011). HA can cause a heavy physical and mental burden to patients, particularly for patients with dis gured skin lesions (Chibbaro et al., 2018).
Pharmacotherapy and surgery are major approaches for HA therapy (Chen et al., 2018). However, there are still 25% HA patients undergoing resection had persistence of symptoms (Haggstrom et al., 2007). The understanding about initiation and risk factors for HA is warranted for discovery of novel treatment approaches and prevent of HA.
It has been reported that about 75-80% of HA patients are females (Chibbaro et al., 2018). The detailed cause of the female preponderance is not yet understood. It has been reported that level of estradiol (E2) in healthy children was signi cantly lower than that in HA patients (Sasaki et al., 1984). The serum levels of E2 in proliferating HA tissues are greater than that in involuting phase (Yu et al., 2006). Laboratory studies indicated that estrogen can promote the in vitro proliferation of HA cells, which may depend on certain growth factors (GFs) and be inhibited by tamoxifen (Xiao et al., 1999). Further, estrogen and VEGF had synergistic effects on proliferation of HA cells (Xiao et al., 2004). All these data suggested the positive roles of estrogenic signals on HA progression.
Endocrine disrupting compounds (EDCs) are environmental compounds which have similar characteristics with E2 (Rubin, 2011). They can in uence multiple endocrine related pathways and then disrupt hormone functions (Henley and Korach, 2006). Bisphenol A (BPA) is a typical EDC banned from many human consumer products due to the negative effects on human health (Lu et al., 2013). Its analog bisphenol S (BPS) are widely used as substitutes for industrial application particularly in many commercial products labeled "BPA-free" (Rochester and Bolden, 2015). Recently, numerous studies indicated that BPS can also accumulate in human body and is an urgent issue for public health (Rochester and Bolden, 2015). For example, BPS can reduce the steroid hormone synthesis and down regulate steroidogenic gene transcripts (Feng et al., 2016). BPS also has a comparable estrogenic activity as BPA (Kuruto-Niwa et al., 2005). It can induce epithelial-mesenchymal transition (EMT) in HA cells via induction of Snail (Zhai et al., 2016). While the potential effects of BPS on the progression of HA are not investigated.
The cytokines such as vascular endothelial growth factor (VEGF) and basic broblast growth factor (bFGF) are critical for HA progression (Fu et al., 2017;Przewratil et al., 2010). For example, VEGFA can facilitate the growth of HA-derived endothelial cells (HemECs) (Jinnin et al., 2008). While targeted inhibition of VEGF can inhibit the proliferation of HA cells (Pan et al., 2015). As to bFGF, its expression was associated with incidence of HA (Bielenberg et al., 1999) and proliferation of HA cells (Przewratil et al., 2010). However, the regulators about bFGF and other cytokines in HA were not well illustrated.
Our present study found that nanomolar BPS can trigger the proliferation of HA cells via increasing the expression of bFGF. While the neutralization antibody of bFGF can abolish BPS-induced proliferation of HA cells. Mechanistically, BPS increased the transcription of bFGF via activation of p65 and stabilized the mRNA of bFGF via decreasing miR-155-5p.

BPS triggers the proliferation and cell cycle transition of HA cells
To evaluate the potential functions of BPS on progression of HA, we treated cells with increasing concentration of BPS for 48 h. CCK-8 analysis showed that 10 nM BPS can signi cantly trigger the proliferation of HDEC cells (Fig. 1A), however, BPS with the concentrations greater than 1 mM can suppress proliferation of HDEC cells. Similarly, nanomolar concentrations of BPS can also trigger the proliferation of CRL-2586 cells (Fig. 1B). Considering that nanomolar was more closely to concentration of BPS observed in human body, we only focused on the potential roles of nanomolar BPS in next studies. Colony formation analysis showed that 10 nM BPS can also signi cantly trigger the colonization of both HDEC and CRL-2586 cells (Fig. 1C). Cell cycle analysis showed that population of the cells in G0/G1 was decreased in the groups treated with BPS. The distribution of S phase was increased by treatment with BPS . It indicated that BPS can trigger the proliferation of HA cells via inducing cell cycle transition.

BPS increases the expression of bFGF in HA cells
Growth factors (GFs) such as EGF and hepatocyte growth factor (HGF) are suggested to play important roles in epithelial cell proliferation (Han et al., 2011;Zelenka and Arpitha, 2008). We then tested the potential effects of BPS on the expression of various GFs including acidic broblast growth factor (aFGF), bFGF, FGF3, Insulin-like growth factor-1 (IGF-1), HGF, VEGF, and transforming growth factor beta (TGF-β). Our data showed that BPS can increase the expression of bFGF and VEGF in HDEC cells ( Fig. 2A). However, BPS can only increase the expression of bFGF in CRL-2586 cells (Fig. 2B).
Consistently, BPS can increase the expression of bFGF in HDEC cells via both time (Fig. 2C) and concentration (Fig. 2D) manners. ELISA con rmed that BPS increased expression of bFGF in both HDEC and CRL-2586 cells (Fig. 2E). All these results suggested that BPS increased the expression of bFGF in HA cells.

bFGF is involved in BPS induced proliferation of HA cells
The potential roles of bFGF were measured in BPS induced proliferation of HA cells. We found that neutralization antibody of bFGF can attenuate BPS induced proliferation of HDEC cells (Fig. 3A). However, the neutralization antibody for VEGF, which was also upregulated by BPS, had no signi cant effect on BPS induced proliferation of HDEC cells (Fig. 3B). Consistently, anti-bFGF also reversed BPS induced proliferation of CRL-2586 cells (Fig. 3C). In addition, anti-bFGF also partially attenuated BPS decreased proportions of G0/G1 phase (Fig. 3D) while increased proportions of S phase (Fig. 3E) in HDEC cells. All these data suggested that bFGF was involved in BPS induced proliferation of HA cells.

BPS can increase the transcription and mRNA stability of bFGF
We then further investigated the mechanisms responsible for upregulation of bFGF. Our data showed that BPS treatment can increase the mRNA of bFGF since treatment for 4 h. We then analyzed the promoter activity of bFGF in BPS treated HA cells by use of promoter luciferase assay. Our data showed that BPS can increase the promoter activity of bFGF in both HDEC and CRL-2586 cells (Fig. 4A). Further, BPS can increase the half-life of bFGF mRNA in HDEC cells (Fig. 4B). However, BPS had no effect on mRNA distribution in cellular fractions such as cytoplasm and nucleus in HDEC cells (Fig. 4C). Further, BPS had no effect on protein stability of bFGF in HDEC cells (Fig. 4D). These data suggested that bFGF can increase the transcription and mRNA stability of bFGF.

NF-κB is involved in BPS induced expression of bFGF in HA cells
Previous studies indicated that AP-1 and NF-κB can regulate the transcription of bFGF .
We therefore investigated whether AP-1 or NF-κB was involved in BPS induced transcription of bFGF. We found that BPS can increase the phosphorylation of p65, while not c-Fos or c-Jun, in HDEC cells (Fig. 5A).
Consistently, BPS also increased the phosphorylation of p65 in CRL-2586 cells (Fig. 5B). Further, BAY, the inhibitor of NF-κB, can reverse BPS induced upregulation of bFGF in HDEC cells (Fig. 5C). Consistently, BAY also partially attenuated BPS induced proliferation of HDEC cells (Fig. 5D). These data suggested that NF-κB was involved in BPS induced expression of bFGF in HA cells. , miR-16-5p, miR-140-5p, miR-503-5p, and miR-155-5p can directly target the 3'UTR of bFGF to regulate its expression. Our data showed that BPS can signi cantly inhibit the expression of miR-155-5p, while not others, in HDEC cells (Fig. 6A). Consistently, our data showed that BPS can also decrease the expression of miR-155-5p in CRL-2586 cells (Fig. 6B). The mimic of miR-155-5p can attenuate BPS induced upregulation of bFGF in HDEC cells (Fig. 6C). Consistently, mimic of miR-155-5p can also partially attenuate BPS induced proliferation of HDEC cells (Fig. 6D). All these data indicated that miR-155-5p is involved in BPS increased mRNA stability of bFGF.  (Vinas and Watson, 2013) and HA cells can be exposed to BPS via blood circulation, the potential effects of BPS on HA progression need further study.

Discussion
We found the upregulation of bFGF was involved in BPS induced proliferation of HA cells. The expression of bFGF and its receptor are closely associated with proliferation of infantile cutaneous hemangioma (Przewratil et al., 2010). In situ hybridization and immunohistochemical analysis con rmed that the expression of bFGF is closely correlated with incidence of hemangioma (Bielenberg et al., 1999). The bFGF is an effective stimulator of breast epithelial cells proliferation and differentiation (Korah et al., 2000). Our present study found that the neutralization antibody against bFGF can suppress the proliferation of HA cells and block promotion effect of BPS on cells proliferation, which further con rmed the essential roles of bFGF in HA progression.
We found that activation of p65 and down regulation of miR-155-5p were responsible for BPS induced transcription and upregulation mRNA stability of bFGF, respectively, in HA cells. In bovine mammary epithelial cells (BMEC), the expression of bFGF is dependent on the NF-κB and AP-1 signaling pathways Cells (2 × 10 2 cells/well) seeded in six-well plates were treated with or without BPS as the indicated conditions. Then cells were allowed to grow for 5-14 days with media changed every 5 days. The formation of colony was stained and counted according to the previous study (Kalailingam et
7 Enzyme-linked immunosorbent assay (ELISA) The levels of bFGF in medium of cells treated with or without BPS were measured by use of ELISA kit according to the manufacturer's protocol (USCN Business Co. Ltd., Wuhan, China).

Promoter activity assay
The promoter (1 kb upstream of TSS) of bFGF was cloned into the pGL-Basic plasmid to generate pGL-bFGF-Basic plasmid. Effects of BPS on the promoter activity of bFGF were tested by luciferase assay according to previously described protocol (Jiang et al., 2013). Brie y, cells were transfected with pGL-bFGF-basic and pRL-TK for 12 h and then further treated with or without BPS for the indicated time periods. The luciferase was measured by Dual-Glo Luciferase Assay system (Promega). Renilla Luciferase (R-luc) was used to normalize re y luciferase (F-luc) activity to evaluate reporter translation e ciency.

Statistical analysis
All values are expressed as the mean ± standard deviation (SD). Data were analyzed by use of SPSS 14.0 software (SPSS, Inc., Chicago, IL, USA). The student t test was used to assess the difference between two groups. P ≤ 0.05 was considered as statistically signi cant.

Consent for publication
All authors give the consent for the publish of this study.

Availability of data and material
All data and material are available.
Disclosure of potential con icts of interest Page 9/18 The authors declare no con ict of interest.    (B) HDEC cells pretreated with or without 100 nM BPS for 24 h were further treated with or without transcriptional inhibitor Act-D for the indicated times, the mRNA of bFGF was checked by qRT-PCR; (C) After treated with or without 100 nM BPS for 24 h, the mRNA of bFGF in cytoplasm and nucleus of HDEC cells was checked by qRT-PCR; (D) HDEC cells were pre-treated with or without 100 nM BPS for 24 h and further treated with CHX (100 μg/ml) for increasing time periods, the expression of bFGF was recorded (left) and quantitatively analyzed (right). Data are shown as means ± SD. **p < 0.01 compared to control.  miR-155-5p is involved in BPS increased mRNA stability of bFGF.HDEC (A) or CRL-2586 (B) cells were treated with or without 100 nM BPS for 15 min, the expression of miRNAs was checked by qRT-PCR; HDEC cells pretreated with scramble control or miR-155 mimic and then further treated with or without 100 nM BPS for 24 h, the mRNA expression of bFGF was checked by qRT-PCR (C), the cell proliferation was detected by CCK-8 kit (D). Data are shown as means ± SD. **p < 0.01 compared to control.