Introduction

Endocrine Disruptor Chemicals (EDCs) are substances that cause an alteration of the endocrine system mimicking the hormone activity, leading to functional alterations of the nervous and immune systems, embryonic development, and metabolic diseases1. Bisphenol A (BPA) is a chemical substance used in association with other chemical products to produce polycarbonate plastics, epoxy resins, and other polymeric materials, like water dispensers, storage containers, reusable beverage bottles, dental sealants, and medical devices2, and to manufacture paper products like receipts and tickets.

Heating and acidification reactions like pasteurization process, storage, sterilization, microwave heating, among others can promote the dissolution of the “ester” radicals that connect the molecules of BPA and release them in small quantities into the products that will be consumed, like food and drinks, representing a potential risk to human health. Preclinical research demonstrated that BPA can induce the development of prostate and breast cancers, precocious puberty, morphological changes of sperm cells, interruption of estral cyclicality, obesity, and alteration of the mammary gland3. BPA mimics the action of natural hormones, which compromises numerous physiological processes like reproduction. Therefore, an unusual presence of xenoestrogens can cause an alteration of endocrine homeostasis4. This EDC interacts with estrogen receptors, mimicking the role of the female hormone, affecting in a negative way the reproduction process. The mechanism of action of the estrogenic activity of BPA is the activation of the nuclear estrogen receptors alpha (ERα) and beta (ERβ)5. BPA is a lipophilic compound that can cross the placenta. Different concentrations of BPA were found in the blood of pregnant women6, and in the amniotic fluid7, but for ethical reasons, “in vitro” is preferred to "in vivo” to study the BPA effects in humans.

Mesenchymal stem cells (MSCs) were classified by Friedenstein and colleagues as non-hematopoietic cells, they are considered multipotent adult stem cells capable of replicating and transforming into other cell types to differentiate into bone, cartilage, and fat tissues8. MSCs may also be able to differentiate into cell types that do not belong to mesodermal tissue, such as nerve cells, heart muscle cells, liver cells, and endothelial cells that form the innermost layer of blood vessels. They possess immunomodulatory properties and can regulate the alterations of the immune system responses, present anti-inflammatory properties, and, once grafted into the body, promote local healing processes, migrating towards the damaged tissue. MSCs can be isolated from different tissues, such as adipose tissue, synovial membranes, trabecular and compact bone, bone marrow, placenta, and oral tissues7,9,10.

The oral cavity, due to its easy accessibility, is a good alternative source to isolate MSCs populations like dental pulp stem cells (DPSCs), stem cells from human exfoliated deciduous teeth (SHED), periodontal ligament stem cells (PDLSCs); apical papilla stem cells (APSCs), dental follicle stem cells (DFSCs) and gingival stem cells (GMSCs)11. MSCs derived from the oral cavity showed the same morphological features, functional ability, and proteomic profile as MSCs isolated from bone marrow12. Oral stem cells have been clinically used in the treatment of a wide variety of disorders comprising bone defects, graft versus host disease, autoimmune diseases, diabetes, neurological, and periodontal diseases13,14, in the treatment of pathologies and malfunctions of blood vessels, and minimizing tissue damage after a post-myocardial infarction15. The positive effect of MSCs in tissue repair and regeneration is associated with the release of soluble factors to induce an immunomodulatory environment16. The hPDLSCs are easily accessible cells. They express the MSC-associated markers CD13, CD29, CD44, CD59, CD90, CD105, and the pericyte marker STRO-1, and present self-renewal, and multi-lineage differentiation capabilities, including neural cells, chondrocytes, adipocytes, cardiomyocytes, endothelial cells, and odontoblasts10, they can regenerate the cementum/periodontal ligament-like complex to take part in the alveolar bone regeneration, and have also shown a vascular potential, initiating in vitro angiogenesis of vascular endothelial cells17. Periodontal ligament stem cells (hPDLSCs) can be differentiated into endothelial cells (e-hPDLSCs) to serve as an “in vitro” model to analyze the effects of treatment in the expression of miRNAs contained in their extracellular vesicles.

The goal of the current work was to investigate, from an epigenetic point of view, the molecular mechanisms related to angiogenesis mediated by BPA in an “in vitro” model of human periodontal ligament stem cells (hPDLSCs) differentiated into endothelial cells (e-hPDLSCs) through the analysis of the expression of miRNAs contained in the extracellular vesicles of BPA-treated and untreated e-hPDLSCs.

Methods

Chemical

The organic compound Bisphenol A [2,2-Bis(4-hydroxyphenyl) propane, 4,4’- Isopropylidenediphenol] (molecular weight: 228.9 g/mol, product number 133027) was purchased from Sigma-Aldrich (Milan, Italy).

Cell culture establishment

Tissue biopsies were scraped from healthy patients from the horizontal fibers of the periodontal ligament of human premolar teeth according a consent document signed by the patients registered in the current work. The Medical Ethics Committee formally accepted the study design at the Medical School, “G. d’Annunzio” University, Chieti, Italy (N° 266/17 April 2014). The Department of Innova-tive Technologies in Medicine and Dentistry and the Laboratory of Stem Cells and Re-generative Medicine are licensed according to the quality standard ISO 9001:2008 (N° 32,031/15/S).

Tissues were washed in phosphate-buffered saline (PBS, Lonza, Basel, Switzerland) solution with 5% of Gentamicin (Lonza, G1272), placed in a culture dish and frammented. The biopsy fragments obtained were cultured in Thera PEAK™MSCGM-CD™ Bullet Kit- chemically defined (MSCGM-CD, Lonza)18 and incubated at 5% CO2 and 37 °C. . The cell culture medium was replaced with the fresh medium every two or three days. After two weeks, cells spontaneously migrated from the tissue explants to the plate. The cells were sub-cultured until passage 2. The adherence to on a plastic substrate and the ability to differentiate in adipogenic and osteogenic lineage is previously demonstrated by Caputi et al.18.Effect of short peptides on neuronal differentiation of stem cells).

hPDLSCs endothelial differentiation

The endothelial differentation was performed through the use of specific medium kit. In detail hPDLSCs were cultured with endothelial growth medium (EGM-2, Lonza) containing basal endothelial medium-2, R3-insulin-like growth factor-1 (R3-IGF-1), growth supplements containing hydrocortisone, human fibroblast growth factor (FGF-b), human epithelial growth factor (EGF), GA-1000, heparin, ascorbic acid, 5% fetal bovine serum (FBS), and vascular endothelial growth factor 165 (VEGF-165) (EGM-2 Bullet Kit; Lonza). The cells were kept at 37 °C with 5% CO2 by replacing the medium every 3 days. After 12 days of induction for endothelial differentiation, the cells were detached from the monolayer with 0.25% trypsin–EDTA (Lonza) and evaluated. Differentiation of hPDLSCs into endothelial cells, e-hPDLSCs, was evidenced at 50–60% of confluence, control samples were hPDLSCs cultured in basal medium MSCGM-CD. To study the capillary-like tube structure formation ability, differentiated cells were cultured in 12-well culture plates, treated with 300 μl/ well of Cultrex® Basement Membrane Extract (Trevigen, Inc., Gaithersburg, MD, USA) and observed at an inverted light microscope after 1 h of culture.

Cell-viability assay

The hPDLSCs were plated at the cell density of 1.8 × 103 cells per well in a 96-well culture plate. The cell viability was analyzed through the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfo-phenyl)-2H-tetrazolium (MTT) assay (CellTiter 96® Aqueous One Solution Cell Proliferation Assay, Promega, Madison, WI, USA). After 24 h of stimulation with BPA at 1, 5, 10, and 100 µM concentration, 20 μL/well of MTT dye solution were added to the culture medium, and cells were incubated for 3 h at 37 °C. The amount of formazan crystals was directly proportional to the number of viable cells in culture, and identified by absorbance at 490 nm wavelength in the Synergy™ HT spectrophotometer, multi-detection microplate reader. (Biotech, Winooski, VT, USA), experiments were executed in triplicate.

Immunofluorescence analysis and confocal laser scanning microscopy (CLSM)

The e-hPDLSCs were stained for immunofluorescence followed by CLSM to analyze the endothelial commitment. Briefly, e-hPDLSCs and 10 μM BPA-e-hPDLSCs treated were cultured at a density of 5 × 103/well on 8-well culture glass slides (Corning, Glendale, Arizona, USA). Samples were fixed for 1 h with 4% paraformaldehyde in 0.1 M of PBS (pH 7.4) (Lonza, Basel, Switzerland) at room temperature. After several washes, the immunofluorescence assay was executed. The permeabilization was done with 0.1% Triton X-100 in PBS buffer (Lonza) for 6 min and blocked with 5% skimmed milk in PBS for 1 h. The primary antibodies used were: anti-PECAM1 (1:200) (sc-376764 Santa Cruz Biotechnology, Dallas, Texas, USA), anti-VEGF (1:200) (sc-57496 Santa Cruz Biotechnology), anti-VEGFR (1:200) (sc-271789 Santa Cruz Biotechnology), and anti-vWF (1:200) (sc-14014 Santa Cruz Biotechnology). The e-hPDLSCs and BPA-e-hPDLSCs treated were incubated with primary antibodies at room temperature for two hours, samples were incubated with secondary antibody Alexa Fluor 488 green fluorescence-conjugated goat anti-rabbit (A-11008, Invitrogen, Eurogene, Oregon, USA) at a 1:200 dilution, and Alexa Fluor 488 green fluorescence-conjugated goat anti-mouse (A-11001, hermofisher Scientific ) at a 1:200 dilution for 1 h at 37 °C. To dye the cytoskeleton actin, hPDLSCs, and e-hPDLSCs cells were incubated with Alexa Fluor 568 phalloidin red fluorescent conjugate (1:200) (A12380, Thermofisher Scientific) for 1 h, and to dye the nuclei, cells were stained with TOPRO (1:200) (T3605, Thermofisher Scientific) for 1 h. The Zeiss LSM800 confocal system (Carl Zeiss, Jena, Germany) was used to acquire images.

Western blotting analysis

The protein lysates of e-hPDLSCs and BPA-e-hPDLSCs treated (50 µg) were electrophoresed and transferred into a polyvinylidenfluoride (PVDF) membrane followed by blocking the membranes in 5% non-fat milk in PBS and 0.1% Tween-20. Then, the membranes were incubated overnight at 4 °C with the following primary antibodies: anti-PECAM1 (1:500) (sc-376764 Santa Cruz Biotechnology), anti-VEGF (1:500) (sc-57496 Santa Cruz Biotechnology), anti-VEGFR (sc-271789 Santa Cruz Biotechnology), anti-IL-vWF (1:500) (sc-14014 Santa Cruz Biotechnology), and β-actin as a loading control (1:750) (sc-69879 Santa Cruz Biotechnology). After five washings with 0.1% Tween-20 in PBS, the membranes were incubated overnight at 4 °C with peroxidase-conjugated anti-mouse (A90-116P Goat anti-mouse Bethyl Laboratories Inc Montgomery, Texas, USA) and rabbit (A 120-101P Goat anti-rabbit Bethyl Laboratories Inc) secondary antibodies at 1:5000 dilution in 1X PBS, 2.5% milk, and Tween-20 at 0.1%19. The levels of protein expression were detected using the Immobilon Crescendo (Merck Millipore, Milan, Italy) with an image documenter Alliance 2.7 (Uvitec, Cambridge, UK). Signals were measured through UVIband-1D gel analysis (Uvitec), data were normalized with values evaluated by densitometric analysis of the β-actin protein.

Extracellular vesicles extraction

EVs were isolated from the conditioned medium of e-hPDLSCs and treated BPA e-hPDLSCs. The ExoQuick TC commercial agglutinant (System Biosciences, Euroclone SpA, Milan, Italy) was utilized for EV extraction: 2 mL of ExoQuick TC solution was added to 10 mL of conditioned medium and incubated overnight at 4 °C without rotation. The following day one centrifugation step was executed at 1500 × g for 30 min to sediment the Evs20.

Extracellular vesicles microRNA array profiling

The EVs RNA was obtained using Total Exosome RNA and Protein Isolation kit (4,478,545, Thermo Fisher Scientific). MicroRNA profiling was performed by the Affymetrix MicroArray GeneChip System (Applied Biosystems, Foster City, CA, US). e-hPDLSCs and e-hPDLSCs + BPA samples were initially labeled with the FlashTag Biotin HSR labeling kit (Applied Biosystems, Foster City, CA, US) as follows: RNA concentration was adjusted for each sample and RNA Spike Control Oligos were added. The tailing reactions were performed by diluting the ATP mix in 1 mM of Tris, as indicated for total RNA, and subsequently mixing it with 10X reaction buffer, 25 mM MnCl 2, and the PAP enzyme to prepare the Poly A Tailing Master Mix in a nuclease-free tube. The Poly A Tailing master, oligo mixes, and corresponding RNA samples were incubated for 15 min at 37 °C. At this point, the 5X FlashTag Biotin HSR Ligation Mix and T4 DNA Ligase was added to each sample and the ligation mixtures were incubated at room temperature for 30 min. The reactions were then stopped by adding the HSR Stop Solution and hybridized with the Affymetrix GeneChip miRNA 4.0 arrays by using the GeneChip Hybridization, Wash and Stain Kit (Applied Biosystems, Foster City, CA, US). The hybridization cocktail was added to each biotin-labeled ligation mixture and the resulting mix was applied to every single array, which was then incubated for 42 h in the GeneChip Hybridization Oven 645 (Applied Biosystems, Foster City, CA, US)21. Arrays were then washed and stained with the same kit in the Fluidics Station 450 according to the protocol FS450_0002 and scanned by the GeneChip Scanner 3000 7G and Affymetrix Command Console (AGCC) Software (Applied Biosystems, Foster City, CA, US). Raw data was analyzed using the Transcriptome Analysis Console (TAC) 4.0 Software (Applied Biosystems, Foster City, CA, US). The GeneChip miRNA 4.0 arrays contain 100% miRBase v20 coverage of 203 organisms and specifically probes for 2578 human mature miRNAs, as well as 1996 human snoRNAs and scaRNAs. Feature intensities were extracted by using the miRNA 4.0 array library files.

Statistics

Statistical significance was acquired with GraphPad 5 software (GraphPad, San Diego, CA, USA) one-way ANOVA followed by post hoc Tukey’s multiple comparisons analysis. Values of p < 0.05 were considered statistically significant.

Ethical approval declarations

All experimental protocols were approved by a Medical Ethics Committee at the Medical School, “G. d’Annunzio” University, Chieti, Italy (N 266/17,042,014). The consent form has been signed and obtained from all patients enrolled in the present study. The biological samples were de-identifyng before the further experimental procedures. The Department of Innovative Technologies in Medicine & Dentistry and the Laboratory of Stem Cells and Regenerative Medicine are certified according to the quality standard ISO 9001:2008 (32,031/15/S). The present research have been performed in accordance with the Declaration of Helsinki.

Results

Effects of BPA on hPDLSCs cell viability

The cell viability of hPDLSCs was evaluated after BPA treatment using an MTT assay. Human PDLSCs showed a fibroblast like-morphology other than the capacity to differentiate into adipogenic and osteogenic lineages, as demonstrated by the positivity to oilred and alizarin red staining (Supplementary figure). Analyzed concentrations of BPA ranged from 1 to 100 µM. The graphics evidenced a dose-dependent reduction of cell viability as concentrations of BPA increased. Viability values remained unaltered after 24 h and after one week. Cells treated using 1 µM of BPA showed 92% cell viability, and cells BPA-treated showed 80% cell viability (Fig. 1A, B). This experiment used 10 μM of BPA for further investigations.

Fig. 1
Fig. 1
Full size image

(A) Cell viability of hPDLSCs cells treated with BPA at 1, 5, 10 and 100 μM for 24 h. Cell viability was assessed using MTT assay and normalized to control cells treated with DMSO (0.2% as final concentration). (B) Cell viability of hPDLSCs cells treated with BPA at 1, 5, 10, and 100 μM for 1 week.

Development of vascular endothelial morphology of endothelial-committed hPDLSCs.

Untreated e-hPDLSCs cells and 10 μM BPA-e-hPDLSCs treated cells were cultured on dishes covered with Cultrex gel solution. Cells were observed at light microscopy after 1, 2, 4, and 6 h. After 2 h in culture, the untreated e-hPDLSCs started to aggregate with each other by cytoplasmic process connections, and after 6 h in culture, a mesh-like structure changed into a tubular structure (Fig. 2 A1-A4). After 2 h in culture, the e-hPDLSCs 10 μM BPA-treated formed the first tubular structures, and after 6 h they were completely formed (Fig. 2 B1-B4). Furthermore, the tubular structures found in the untreated cells were comparable to those of the 10 μM BPA-treated e-hPDLSCs.

Fig. 2
Fig. 2
Full size image

Hpdlscs endothelial differentiation of untreated (A1-A4) and 10 μM BPA-treated cells (B1-B4) for 1, 2, 4 and 6 h.

Expression levels of PECAM1, VEGF, VEGFR, and vWF in BPA-treated cells.

Immunofluorescence results evidenced that PECAM1, VEGF, VEGFR, and vWF proteins were significantly more expressed in BPA-treated e-hPDLSCs, than in e-hPDLSCs untreated cells.

On the other hand, hPDLSCs untreated and BPA-treated hPDLSCs presented a noticeable reduction in these protein expressions compared to the e-hPDLSCs untreated and treated cells respectively (Fig. 3, Fig. 4, Fig. 5 and Fig. 6).

Fig. 3
Fig. 3
Full size image

PECAM-1 signaling pathway and differences in the protein expression of untreated, and 10 μm BPA-treated e- hPDLSCs cell line. Expression of PECAM-1 analyzed by confocal microscopy. Green fluorescence: PECAM-1. Red fluorescence: cytoskeleton actin. Blue fluorescence: cell nuclei. Scale bar: 20 μm.

Fig. 4
Fig. 4
Full size image

VEGF signaling pathway and differences in the protein expression of untreated, and 10 μm BPA-treated e-hPDLSCs cell line. Expression of VEGF analyzed by confocal microscopy. Green fluorescence: VEGF. Red fluorescence: cytoskeleton actin. Blue fluorescence: cell nuclei. Scale bar: 20 μm.

Fig. 5
Fig. 5
Full size image

VEGFR signaling pathway and differences in the protein expression of untreated, and 10 μm BPA-treated e-hPDLSCs cell line. Expression of VEGFR analyzed by confocal microscopy. Green fluorescence: VEGFR. Red fluorescence: cytoskeleton actin. Blue fluorescence: cell nuclei. Scale bar: 20 μm.

Fig. 6
Fig. 6
Full size image

vWF signaling pathway and diferences in the protein expression of untreated, and 10 μm BPA-treated e-hPDLSCs cell line. Expression of vWF analyzed by confocal microscopy. Green fluorescence: vWF. Red fluorescence: cytoskeleton actin. Blue fluorescence: cell nuclei. Scale bar: 20 μm.

The Western blot analysis showed expression of PECAM1, VEGF, VEGFR, and vWF significantly increased in BPA e-hPDLSCs treated cells compared to the e-hPDLSCs untreated cells (Fig. 7).

Fig. 7
Fig. 7
Full size image

Western blotting analysis. (A) PECAM-1, VEGF, VEGFR, and vWF of untreated, and 10 μm BPA-treated e-hPDLSCs cell line. Each membrane was probed with β- actin antibody to verify the loading consistency. Western blot image was representative of three different experiments. (B) Histograms represent the mean value of three separate experiments of densitometric measurements of protein bands expressed as the integrated optical intensity (IOI). The error bars showed the standard deviation (± SD). Data was analyzed by ANOVA, with a post-hoc application of t-student test. *** p < 0.001.

MicroRNA expression profiles in BPA-treated and untreated e-hPDLSCs.

Ingenuity Pathways Analysis (IPA) and its microRNA Target Filter were analyzed to disclose a possible microRNA—gene target relation between our differentially expressed microRNAs and angiogenesis markers, PECAM1, VEGF, VEGFR, and vWF, based on reported data or prediction tools (Ingenuity Expert Findings, TarBase, TargetScan, miRecords).

Comparing BPA-treated e-hPDLSCs to the untreated e-hPDLSCs, it was evidenced that a series of the initial 395 modulated miRNAs, which were significantly down-regulated (control cells), presented up-regulation expression of PECAM1, VEGF, VEGFR, and vWF.

The 10 following microRNAs, hsa-miR-1233-5p, hsa-miR-193b-5p, hsa-miR-26a-5p, hsa-miR-6084, hsa-miR-6124, hsa-miR-6165, hsa-miR-619-5p, hsa-miR-6778-5p, hsa-miR-6880-5p and hsa-miR-8075, target the gene expression of PECAM1.

The 14 following microRNAs, hsa-miR-1343-5p, hsa-miR-17-5p, hsa-miR-4270, hsa-miR-4298, hsa-miR-4441, hsa-miR-4673, hsa-miR-5196-5p, hsa-miR-6127, hsa-miR-6133, hsa-miR-6748-5p, hsa-miR-6754-5p, hsa-miR-6756-5p, hsa-miR-6875-5p and hsa-miR-939-5p, target the expression of VEGF.

The 11 following microRNAs, hsa-miR-221-3p, hsa-miR-222-3p, hsa-miR-551b-5p, hsa-miR-6511b-5p, hsa-miR-664b-5p, hsa-miR-665, hsa-miR-6762-5p, hsa-miR-6798-5p, hsa-miR-6803-5p, hsa-miR-6827-5p, and hsa-miR-6845-5p, have been shown to target VEGFR, while 3 of them, hsa-miR-2467-3p, hsa-miR-4322 and hsa-miR-6796-5p, target the expression of VWF. (Table 1). In microarray analysis, the expression levels of 395 microRNAs in BPA-treated e-hPDLSCs were significantly modulated compared to untreated e-hPDLSCs (p < 0,05) (Fig. 8).

Table 1 List of down-regulated microRNAs, that target the up-regulated angiogenic markers, PECAM1, VEGF, VEGFR, and VWF in BPA-exposed e-hPDLSCs vs e-hPDLSCs.
Full size table
Fig. 8
Fig. 8
Full size image

Volcano Plot of the total number of microRNAs expressed in BPA-treated e-hPDLSCs compared to untreated e-hPDLSCs. In red are the up-regulated ones, in green are the down-regulated ones, and in grey are the non-modulated miRNAs.

Discussion

BPA can cause epigenetic transformation, increasing angiogenic markers inside e-hPDLSCs vesicles. Due to the simple accessibility of e-hPDLSCs, the human oral cavity is an outstanding font of MSCs localized in specialized, well-characterized tissues.

Endocrine disruptor chemicals (EDCs) are exogenous substances that can interfere with the endocrine system and may lead to developmental, reproductive, immune, neurological, or metabolic diseases in humans and animals22. One of the most studied EDCs is BPA, a synthetic organic compound used to produce plastics, polyvinyl chloride (PVC), food containers, epoxy resins, and dental sealants. BPA is an EDC that can have a significant impact on the pathogenesis of hormone-dependent cancer, such as breast and prostate cancer. Indeed there are several reports on the potential role of BPA in cancer pathogenesis, demonstrating that BPA exposure induces the proliferation of human cancer cells and causes increased oxidative stress23. BPA can negatively affect the angiogenesis of the fetus and trigger malignant formations in the unborn child24.

The current study evaluated the potential of hPDLSCs to differentiate into endothelial cells (e-hPDLSCs) through the tube test formation assay during the exposure of BPA. Although there were no morphology differences compared to the untreated control, the increase in the vascularization signaling pathways ofthe angiogenic markers, platelet endothelial cell adhesion molecule (PECAM-1), vascular endothelial growth factor (VEGF), vascular endothelial growth factor receptor (VEGFR) and von Willebrand factor (vWF) evidenced their vascularization commitment. These results indicate that BPA could determine uncontrolled angiogenesis, an alteration typical of inflammation and necessary for the neovascularization process that characterizes tumor growth and metastatic spread, confirming the potential role of BPA in cancer pathogenesis25 Furthermore, a functional angiogenic process is essential for fetal development, as well as in the female reproductive cycle and in tissue repair processes26. Since BPA appears to alter the angiogenic process, it may also be involved in altered fetal development resulting from abnormal angiogenesis also in placental development27. Neovascularization, including tumor angiogenesis, is a four-step process. The basement membrane in tissues is injured locally, immediate destruction and hypoxia occur, angiogenic factors activate endothelial cells that proliferate and stabilize, and angiogenic factors continue to influence the angiogenic process. PECAM-1 is an early and sensitive marker for tumor-induced angiogenesis that is involved in the process of angiogenesis in the developing vertebrate embryo as well as during metastases formation28, VEGF and its receptor VEGFR plays a vital role not only in physiological but also in most pathological angiogenesis, such as cancer29, vWF regulates angiogenesis through multiple pathways, both intracellular and extracellular, knocking down vWF can inhibit breast cancer cell metastasis, and its overexpression significantly stimulated VEGF-A-dependent vascular proliferation by triggering the PI3K/Akt signaling pathway30. The vWF plays a role not only in homeostasis but also in tumor metastasis and angiogenesis31. Our results showing that BPA increased PECAM-1, VEGF, VEGFR, and vWF expression, highlighted the possible involvement of BPA in tumor as well as embryonic angiogenesis and confirmed the results evidenced already in other studies32,33.

Our data demonstrated that BPA determines an alteration of angiogenic pathways also from an epigenetic point of view. The IPA analysis of the microRNA vesicle content produced by the cells revealed that BPA-treated e-hPDLSCs significantly modulated 395 microRNAs compared to e-hPDLSCs. The miRNAs that target the expression of PECAM1 (microRNAs, hsa-miR-1233-5p, hsa-miR-193b-5p, hsa-miR-26a-5p, hsa-miR-6084, hsa-miR-6124, hsa-miR-6165, hsa-miR-619-5p, hsa-miR-6778-5p, hsa-miR-6880-5p and hsa-miR-8075), VEGF (hsa-miR-1343-5p, hsa-miR-17-5p, hsa-miR-4270, hsa-miR-4298, hsa-miR-4441, hsa-miR-4673, hsa-miR-5196-5p, hsa-miR-6127, hsa-miR-6133, hsa-miR-6748-5p, hsa-miR-6754-5p, hsa-miR-6756-5p, hsa-miR-6875-5p and hsa-miR-939-5p), VEGFR (called hsa-miR-221-3p, hsa-miR-222-3p, hsa-miR-551b-5p, hsa-miR-6511b-5p, hsa-miR-664b-5p, hsa-miR-665, hsa-miR-6762-5p, hsa-miR-6798-5p, hsa-miR-6803-5p, hsa-miR-6827-5p, and hsa-miR-6845-5p), and vWF (hsa-miR-2467-3p, hsa-miR-4322, and hsa-miR-6796-5p), were upregulated in our BPA-treated e-hPDLSCs compared to e-hPDLSCs, suggesting an epigenetic alteration inside the cells.

From the epigenetic point of view, the increase in the expression of the angiogenic markers and the miRNAs in the extracellular vesicles produced by BPA-treated e-hPDLSCs compared to the untreated control confirmed BPA effects on the vascularization commitment of e-hPDLSCs.

Despite the data provided are not directly translatable to human health for the limitations related to the experimentation on an in vitro model, they reinforce the results about the effect of BPA on the angiogenic process already described in literature but on other cell line. Moreover, since in the present work, the effects of BPA is evaluated during the differentiation process towards an endothelial lineage, our results provide the basis for further studies regarding the involvement of this endocrine disruptor also on the differentiation process, important not only for tumor and metastatic processes, but also for possible alterations of embryonic and fetal development related for example to maternal BPA intake during pregnancy. Furthermore, our data about different expression of miRNAs in vesicles, provide important results on how BPA interferes with the angiogenic molecular pathways also through an epigenetic manner.

Conclusion

BPA reduced cell viability of hPDLSCs in an inverse dose-dependent relationship. Although BPA-treated cells significantly express more PECAM1, VEGF, VEGFR, and vWF proteins and 395 microRNAs related to angiogenesis than untreated controls, evidencing the increase in the angiogenesis mechanism inside these cells, the vascular endothelial morphology of endothelial-committed hPDLSCs did not present differences compared to the untreated control.