Repeated Porphyromonas gingivalis W83 exposure leads to release pro-inflammatory cytokynes and angiotensin II in coronary artery endothelial cells

The role of Porphyromonas gingivalis (P. gingivalis) or its virulence factors, including lipopolysaccharide (LPS) not only has been related with periodontitis but also with endothelial dysfunction, a key mechanism involved in the genesis of atherosclerosis and hypertension that involving systemic inflammatory markers as angiotensin II (Ang II) and cytokines. This study compares the effect of repeated and unique exposures of P. gingivalis W83 LPS and live bacteria on the production and expression of inflammatory mediators and vasoconstrictor molecules with Ang II. Human coronary artery endothelial cells (HCAEC) were stimulated with purified LPS of P. gingivalis (1.0, 3.5 or 7.0 μg/mL) or serial dilutions of live bacteria (MOI 1: 100 - 1:0,1) at a single or repeated exposure for a time of 24 h. mRNA expression levels of AGTR1, AGTR2, IL-8, IL-1β and MCP-1 were determined by RT-qPCR, and IL-6, MCP-1, IL-8, IL-1β and GM-CSF levels were measured by flow cytometry, ELISA determined Ang II levels. Live bacteria in a single dose increased mRNA levels of AGTR1, and repeated doses increased mRNA levels of IL-8 and IL-1β (p < 0.05). Repeated exposure of live-P. gingivalis induced significant production IL-6, MCP-1 and GM-CSF (p < 0.05). Moreover, these MCP-1, IL-6 and GM-CSF levels were greater than in cells treated with single exposure (p < 0.05), The expression of AGTR1 and production of Ang II induced by live-P. gingivalis W83 showed a vasomotor effect of whole bacteria in HCAEC more than LPS. In conclusion, the findings of this study suggest that repeated exposure of P. gingivalis in HCAEC induces the activation of proinflammatory and vasoconstrictor molecules that lead to endothelial dysfunction being a key mechanism of the onset and progression of arterial hypertension and atherosclerosis.

. Viability of HCAECs after repeated treatments with live-P. gingivalis (A) and P. gingivalis-LPS (B). The HCAECs were stimulated to repeated live-P. gingivalis (MOI 1:100 -1:0,1) and P. gingivalis-LPS (1.0, 3.5 and 7.0 µg/mL) exposures, during 24 h. Cell viability was determined according to the fluorometric detection after reduction of resazurin in the resorufin product using AlamarBlue. 1% was considered our positive control of cell death. Percentage of cell viability with respect to the control. *Represents the statistical difference with respect to the control or without stimulus. (p < 0.05). Three independent experiments were performed; the results are presented as the means ± SEM (n = 3).
In this study, we evaluated the expression of proinflammatory cytokine mRNA and AGTR genes in HCAEC stimulated with P. gingivalis-LPS and live-P. gingivalis W83. The cells stimulated at single and repeated exposure with the different concentrations of P. gingivalis-LPS did not show significant changes for any of the pro-inflammatory markers even in comparison with the control group (Fig. 2), while repeated exposure of MOI 1:100 live-P. gingivalis compared to single exposure to the same MOI 1:100, significantly affects the expression of IL-8 and IL-1β (p < 0.05). In contrast, the expression of MCP-1 was not significantly affected by the treatments; however, an apparent reduction in its mRNA was observed after single exposure to LPS and live bacteria. ( Fig. 2A-E). On the other hand, a single exposure to live-P. gingivalis increased the AGTR1 expression ( Fig. 2A) compared to unchallenged HCAEC and challenged with live bacteria to a repeated exposure (p < 0.05), whereas AGTR2 (Fig. 2B) It was not because of the results of the treatments evaluated. P. gingivalis induces IL-1β, IL-8, IL-6, MCP-1, and GM-CSF production in HCAEC. In a pilot test, HCAEC cells were stimulated at day 1, day 3, and day 5; the supernatant was removed each 48 h to reaching a total of 7 days under repeated exposure. In general, we found similar results in a longer time frame and shorter time frame to induce the production of IL8 in the majority of LPS concentrations assessed ( Supplementary Fig. 1). However, in a short period of time an overproduction of IL8 was achieved, suggesting greater endothelial activation than with a long period of time in which endothelial activation appears to decrease. This model could be used to study endotoxin tolerance of periodontal pathogens, an aspect that requires further research. Similarly, when comparing IL6 and IL1, there were also no modifications between the 24 h and 7 day in repeated exposure models ( Supplementary Fig. 2).
The HCAECs were stimulated with P. gingivalis-LPS or live-P. gingivalis for 24 h by repeated or single exposures to evaluate the levels of IL-8, MCP-1, IL-6, IL-1β and GM-CSF in culture supernatants. Repeated exposure to live-P. gingivalis induced a significant increase in the production of IL-6, MCP-1 and GM-CSF, compared to the control group (p < 0.05), while the P. gingivalis-LPS is a single exposure only induced change in IL-1β (Fig. 3). The chemokine IL-8 was increased mainly in HCAEC challenged with live-P. gingivalis and P. gingivalis-LPS at 7 μg/mL at repeated expositions, but not significantly compared to the control (Fig. 3A). After considering the results so far, endothelial responses differed between approaches. HCAEC to repeated exhibitions with live-P. gingivalis achieved a significantly higher production of MCP-1, IL-6 and GM-CSF than treated endothelial cells at a single exposure of bacteria ( Fig. 3B-D), while producing IL-1β (Fig. 3E) was higher in HCAEC challenged with LPS at a single exposure of 1μg/mL (p < 0.05).

P. gingivalis increases angiotensin II levels in HCAEC.
Ang II concentration was evaluated in HCAEC supernatants after P. gingivalis-LPS stimulation and live-P. gingivalis in single and repeated expositions. Only live-P. gingivalis (MOI 1:100) at repeated exposure significantly increased the concentration of Ang II, compared to the live bacteria single exposure (1.9 times) and the control cells (2.1 times) (Fig. 4).

Discussion
Oral pathogens associated with periodontitis, such as P. gingivalis, have been of particular interest due to the high levels of bacteraemia and endotoxemia after routine dental procedures and everyday oral activities, such as tooth brushing 7 . The biological mechanisms underlying the potential link between periodontitis and atherosclerosis or hypertension remain unclear, mainly in terms of inflammatory and vasoactive endothelial responses to P. gingivalis with repeated exposures. In this study, we evaluated the effects of single versus repeated administration of P. gingivalis-LPS and whole bacteria on the pro-inflammatory mediators and vasoactive markers involved in HCAEC atherosclerotic response.
In relation to cell viability, there are no reports in the literature that demonstrate the in vitro effect of periodontopathogens such as P. gingivalis on HCAEC cells at repeated exposures and at concentrations greater than 2 mg/mL with LPS 33,34 . Our results demonstrated that of P. gingivalis-LPS and the bacteria did not affect the cellular viability of HCAEC at the concentrations evaluated; data similar to those reported by Chou et al. 35 where even 100 µg/mL of P. gingivalis-LPS W83 on leukocytes as PMN did not affect its viability 36 . However, depending on the type of cell line differences may occur, in this way have been described in esophageal cell lines as OE19 (adenocarcinoma), OE21 (squamous cell carcinoma), a sensitivity different from that reported in oral tumor cells (HN30), where the esophageal cancer cells were only sensitive to LPS P. gingivalis W83 concentrations of 20, 50 and 100 μg/mL (24,48 and 72 h), unlike oral cells (HN30) where the LPS increases the viability in relation to the control after 72 h of stimulus 37 .
P. gingivalis W83 strains have been known to adhere, invade and persist in bacteria-infected HCAEC 33 ; however, its ability to activate endothelial cells by chemokine and cytokine production seems to exhibit slight activation to IL-8, IL-6, and MCP-1 compared to cells infected with other strains, such as 381 or 33277 33 . On the other hand, this study demonstrated inflammatory effect of W83 strain on endothelial cells, since HCAEC exposed to repeated exposure of live-P. gingivalis induced significant increases in chemokine levels, such as IL-8 and MCP-1, and in cytokines, such as IL-6, compared with the control group, while HCAEC cells challenged with a single exposure of P. gingivalis W83 showed similar production of IL-8, IL-6 and MCP-1, compared with uninfected cells.
Increases in IL-8 are involved in the firm adhesion of rolling monocytes in the early stages of atherogenesis 38 . Similarly, IL-6 and MCP-1 increases have been implicated in the adhesion of leukocytes (mainly monocytes) to activated endothelium, which contributes to cellular migration 39 . Experimental findings suggest differential efficacy of P. gingivalis to activate HCAEC cells. P. gingivalis 381 also enhanced IL-6, IL-8 and MCP-1 production and even the adhesion of immune cells with bacteria-infected HCAEC, while other P. gingivalis strains, such as W83, induced slight activation 33,40,41 .  42 . In contrast, several authors have shown that endotoxin of P. gingivalis strain 33277 induces an important pro-inflammatory effect, increasing the production of IL-8, soluble E-selectin and MCP-1 in HCAEC 43 , while 381 strains only exerted a weak stimulatory effect on HCAEC 41 . These results have been previously described and can be attributed to the presence of the K1 capsule or the characteristics of thechemotype of P. gingivalis-LPS W83 that may present differences in the O antigen region [44][45][46] , which leads to the structural variations of the LPS that can be related with attenuation of the immune system of the host. Therefore, the low inflammatory response in HCAEC cells challenged with LPS could not be generalized to all strains of P. gingivalis 16,33 .
It has been described that P. gingivalis-LPS is an inducer of tolerance in macrophages, mainly by the suppression of the endothelial recognition receptor (TLR-4) 47,48 ; however, our results showed that LPS (7.0 µg/mL) induced an increase in the release of IL-8 at repeated exposures, which may be related to the type of endothelial recognition receptor to which P. gingivalis-LPS W83 binds (TLR-2) or other factors that may be involved by positively regulating the release of IL-8 49,50 . However, new studies are required to elucidate the mechanism of action.
Regarding the relationship between mRNA and proteins levels, cytokines as MCP-1 and IL-8 which were significantly detected in supernatants of HCAEC exposed to live-P. gingivalis, while MCP-1 mRNA levels were not upregulated. These discrepancies between mRNA and protein levels may involve the degradation rate of mRNA, which falls within a much tighter range (2-7 h for mammalian mRNAs vs 48 h for proteins) 51 . In fact, a previous report in aortic smooth muscle cells has determined that in vitro half-life of MCP-1 mRNA is approximately 45 min 52 , however, further verification using transcriptome analysis or RT-qPCR are required.
On the other hand, Ang II has been implied in atherogenesis promoting the oxidative stress in the vasculature, endothelial dysfunction and induction of an inflammatory response in the vessel wall 53 . Regardless endothelial cells are not considered a dominant source of AngII and AGTR 54 ; previous studies have shown that classic LPS from enterobacterias stimulate local and circulatory Ang II levels [55][56][57] . However, knowledge regarding the in vitro vasoactive effects of periodontopathogens as P. gingivalis-LPS on HCAECs, the cell type that has typically been used in studies of atherosclerotic diseases, is unclear yet. Contrariwise, our data showed P. gingivalis-LPS as a poor gene inducer to Ang II, AGTR1, AGTR2 and all chemokine/cytokine measured. It also suggests a weak vasoactive and inflammatory effect of endotoxin isolated from P. gingivalis W83.
Interestingly, repeated exposure of live bacteria P. gingivalis induces a greatest effect in HCAEC, suggesting that Ang II can modulate signaling cascades associated with the release of pro-inflammatory cytokines through calcium mobilization, arachidonic acid production, kinases activation (MAPKs, PKC, JAK, PI3K) or the activation of transcription factors as cAMP and NF-Kβ 17,58-60 , which may explain the possible effects presented by the stimulation with live-P. gingivalis on endothelial cells 21,55 . However, further research is required to clarify possible mechanism.
Regarding the association between Ang II and the release of proinflammatory molecules such as IL8, IL6 and MCP1, there is no evidence with periodontopathogens, while some evidence have been reported Escheriria coli-LPS (E. coli-LPS) 48,58,59 , suggesting a synergistic effect between live-P. gingivalis stimulus and Ang II.
In contrast to the high concentration of Ang II at repeated dose of live bacteria, we found a down regulation of AGTR1 and AGTR2 at the mRNA level. Similar results with human saphenous vein cells (VSMC) and rat aortic smooth muscle (HASMC) suggest that high concentrations of Ang II can induce in vitro gen downregulation or desensitization/internationalization of AGTR1 48,58,59 or AGTR2 57 . However, additional studies are required to identify the role of these receptors in endothelial dysfunction at repeated dose of periodontopathogens. www.nature.com/scientificreports www.nature.com/scientificreports/ This work represents an alternative to the traditional in-vitro approach to evaluate P. gingivalis effects on endothelial cells, since transient and frequent bacteremia or endotoxemia episodes have been clearly described in patients with periodontitis. However, the classic stimulation model performed at a single dose up to 24 h did not expose the endothelium to these immunological challenges. Therefore, in-vitro exposure to more than one (repeated) exposure of P. gingivalis on the endothelium could lead to a better understanding for the study of endothelial dysfunction and pro-inflammatory activation.

Materials and Methods
Bacterial culture and inoculum standardization. P. gingivalis (BAA-308/W83) strain was obtained from the American Type Culture Collection (ATCC) and cultured using standard methods. This strain was originally isolated from humans with oral infections (i.e., periodontitis) and has been shown to be highly virulent compared with other P. gingivalis strains 60 . Bacteria were grown in supplemented Brucella agar (0.3% Bacto agar, 0.2% yeast extract, 5% defibrinated sheep blood, 0.2% haemolyzed blood, 0.0005% hemin, and 0.00005% menadione) and incubated at 37 °C for 4 days in anaerobic conditions (Anaerogen, Oxoid, Hampshire, UK) 61 . Bacterial inoculums were prepared and standardized for P. gingivalis in RPMI-1640 (Thermo Scientific, Waltham, MA, USA) and were quantified by spectrophotometry (Thermo Scientific, Waltham, MA, USA) at specific optical densities (OD) of 0.900-0.908 at a wavelength of 620 nm, corresponding to 2,6 × 10 9 bacteria/mL. The count of colony forming units (CFU) was confirmed in triplicate under incubation conditions. Viable bacteria experiments were performed in a maximum time of two hours after having counted them, in order to avoid bacterial mortality.
LPS extraction and purification. LPS extraction was performed using hot phenol-water, as previously reported 32 , with 1.1 g of wetted P. gingivalis W83; the purification was accomplished using an enzymatic treatment with nucleases and protease, followed by size-exclusion chromatography (Sephacryl S-200 HR) with sodium deoxycholate as the mobile phase 62 . The characterization of the LPS was determined by SDS-PAGE electrophoresis, purpald assay and chromogenic LAL test, compared to the commercial LPS of P. gingivalis ATCC 33277 (InvivoGen) 32 .

Stimulation of HCAEC with LPS and viable bacteria in a single and repeated exposure model.
HCAEC cells (LONZA, Walkersville USA) were cultured in supplemented EGM2 MV medium (LONZA, Walkersville, USA). The cells were used at passage 7 in growth medium (2 × 10 5 cells/well) using 12-well culture plates (CytoOne, USA Scientific, Orlando FL, USA) and pre-incubated at 37 °C in a water-saturated atmosphere of 95% air and 5% CO 2 until reaching confluence at 20 h. Subsequently, the cells were exposed to two treatment models at different concentrations of purified LPS (1.0, 3.5 and 7.0 µg/mL) and serial dilutions of live bacteria (MOI 1:100 -1:0,1). A pilot study was conducted using longer exposure times where HCAEC cells were stimulated with P. gingivalis LPS on days 1, day 3 and day 5; the supernatant is removed every 48 hours to reach a total of 7 days under repeated exposure. The supernatant was stored at −80 C, for subsequent cytokine analysis.
In the first model the cells were stimulated for 6 h, after that, the stimulus was removed between each stimulus and replaced with a next exposure for another 6 h and for the last stimulation HCAEC the cells were exposed to another 12 h, for a total exposure time from 24 h. The supernatant was collected and storage for soluble factors measuring.
For the second treatment model, the conventional stimulus was referenced by literature in which HCAECs were stimulated with LPS at 1 µg/mL and serial dilutions of live bacteria MOI (1:100 -1:0,1) at a single exposure for 24 h 43 .
Angiotensin II concentration. Ang II levels were determined from the supernatants of the cultured HCAEC stimulated with LPS or live bacteria, using the EIA ELISA Kit (Cayman Chemical, Ann Arbor, MI, USA). To detect Ang II in the supernatants, sample extraction and ELISA were performed according to the manufacturer's instructions 65 . Statistical analysis. All experiments were repeated at least 3 times for qPCR and flow cytometry. All data were expressed as the mean ± SEM. ELISA results were performed at least 3 times in duplicate. One-way variance analysis (ANOVA) and Tukey's post hoc tests were performed for all analyses. A p-value < 0.05 was considered statistically significant.

Conclusions
Repeated exposure live-P. gingivalis induces a greater pro-inflammatory response than single exposure, described by IL-8, IL-6, MCP-1 and GM-CS in HCAEC. The expression of AGTR1 and production of Ang II induced by live-P. gingivalis W83 showed the vasomotor effect of whole bacteria in HCAEC more than LPS. The findings of this study suggest that repeated exposure of P. gingivalis in HCAEC induces the activation of proinflammatory and vasoconstrictor molecules that lead to endothelial dysfunction as a key mechanism of the onset and progression of arterial hypertension (HT) and atherosclerosis, which requires more research.