Crosstalk between Akt and NF‐κB pathway mediates inhibitory effect of gas6 on monocytes‐endothelial cells interactions stimulated by P. gingivalis‐LPS

Abstract Correlation between periodontitis and atherosclerosis is well established, and the inherent mechanisms responsible for this relationship remain unclear. The biological function of growth arrest‐specific 6 (gas6) has been discovered in both atherosclerosis and inflammation. Inhibitory effects of gas6 on the expression of inflammatory factors in human umbilical vein endothelial cells (HUVECs) stimulated by Porphyromonas gingivalis lipopolysaccharide (P. gingivalis‐LPS) were reported in our previous research. Herein, the effects of gas6 on monocytes‐endothelial cells interactions in vitro and their probable mechanisms were further investigated. Gas6 protein in HUVECs was knocked down with siRNA or overexpressed with plasmids. Transwell inserts and co‐culturing system were introduced to observe chemotaxis and adhering affinity between monocytes and endothelial cells in vitro. Expression of gas6 was decreased in inflammatory periodontal tissues and HUVECs challenged with P. gingivalis‐LPS. The inhibitory effect of gas6 on chemotaxis and adhesion affinity between monocytes and endothelial cells was observed, and gas6 promoted Akt phosphorylation and inhibited NF‐κB phosphorylation. To our best knowledge, we are first to report that gas6 inhibit monocytes‐endothelial cells interactions in vitro induced by P. gingivalis‐LPS via Akt/NF‐κB pathway. Additionally, inflammation‐mediated inhibition of gas6 expression is through LncRNA GAS6‐AS2, rather than GAS6‐AS1, which is also newly reported.

most recently. 3,4 In the aforementioned study of periodontitis, the bacterial plaque destroyed the epithelium of the periodontal pocket, allowing the entry of harmful elements (such as endotoxins, exotoxins and bacteria) into the bloodstream, causing a low-grade, systemic inflammatory condition. In addition to the direct invasion of the vessel wall, the presence of oral pathogens such as P. gingivalis (one of the key periodontal pathogens) was observed. 5 In our previous study, DNA from P. gingivalis was also found in human atheromatous plaques. 6 A separate study showed that pathogenic substances leaking into blood vessels triggered an inflammatory response that lead to endothelial dysfunction, 7 an important contributor to the initiation and progression of atherosclerotic lesions. 8 Such inflammatory responses are often due to the presence of lipopolysaccharides.
LPS-also known as lipoglycans and endotoxins-are found in the outer membranes of Gram-negative bacteria and are one of their key virulence factors. In previous studies, elevated LPS level circulating in the blood of periodontitis patients have been related to an increased risk of atherosclerosis. 9 Endothelial dysfunction has also been observed in blood vessels stimulated by LPS from periodontal pathogens. 10,11 In our previous report, P. gingivalis-LPS stimulation not only negatively affected the viability, proliferation and migration of HUVECs, but also positively promoted the secretion of adhesion molecules (ICAM-1, E-selectin) and chemokines (MCP-1, IL-8) in vitro, 12 which promoted local leucocyte infiltration and prompted the occurrence of atherosclerotic lesions.
Growth arrest-specific 6 (gas6)-a 75KD-secreted protein first found in serum-starved NIH 3T3 cells by Manfioletti in 1993-belongs to the vitamin K-dependent protein family. The gas6 protein shares a 43% amino acid identity with protein S-a negative coregulatory molecule involved in blood coagulation pathways. Gas6 contains an N-terminal carboxy glutamic acid (Gla) domain-a region rich in glutamic acid residues that is γ-carboxylated in a vitamin K-dependent reaction.
In addition, gas6 contains four epidermal growth factor-like domains and two laminin globular-like domains that contain the interaction sites for its TAM (Tyro3, Axl and Mer) receptor tyrosine kinases. 13 Gas6 is ubiquitously expressed in many cells including endothelial cells. The gas6/TAM system participates in several pathophysiological processes including thrombosis, the phagocytosis of apoptotic cells, inflammation inhibition and vascular calcification. 14 TAM-dependent pathways act as a negative feedback mechanism that suppresses inflammation. 15 Gas6 is a key homeostatic, immunological regulator of host-commensal interactions in the oral mucosa. The absence of gas6 has been shown to increase the anaerobic bacterial load and, consequently, the level of gingival inflammation in vivo. 16 In the context of atherosclerosis, Axl and Tyro3 are down-regulated in advanced human carotid plaques, 17 while Mer mutations promoted the necrosis of atherosclerotic plaques in ApoE -/mice. 18 Additionally, gas6 has been independently associated with reduced plaque height and total plaque area. 19 Protective effects of Gas6 on endothelial tight junction and permeability were also recently demonstrated in vivo. 20 Together, these data illustrate the critical role of gas6 in inflammation and atherosclerosis, and show that gas6 is likely the base molecule of the mechanisms underlying the association between periodontitis and atherosclerosis.
The earliest pathological changes of atherosclerosis involve the activation of endothelial cells, which recruit monocytes and then tether them to the intima. We observed that gas6 exerted an inhibitory effect on the mRNA expression of adhesion molecules and chemokines in HUVECs stimulated with 1μg/mL P. gingivalis-LPS. 21 However, the influence and mechanisms of gas6 on the recruiting and adhering functions of the HUVECs remained unclear. Therefore, the aims of this study were to: (a) observe the in vitro effect of gas6 on chemotaxis and adhesion of monocytes to HUVECs stimulated by P. gingivalis-LPS and (b) explore the possible mechanisms of gas6 involved in this process.
Ultra-pure P. gingivalis-LPS was purchased from InvivoGen and dissolved in endotoxin-free water at a concentration of 1 mg/mL; the resulting solution was stored at −20°C. LPS preparations were free from lipoproteins as reported by other study. 22

| Cell transfection
HUVEC cultures reaching 50%-70% confluence were transfected with gas6 siRNA (si-Gas6) with a scrambled siRNA (si-CTR) as a negative control to knock-down gas6 expression-or with pcDNA3.1(+) plasmids to overexpress gas6. To knock-down the expression level of GAS6-AS2, plasmids containing Gas6-AS2 short hairpin RNA (sh-Gas6-AS2) were used. Delivery of siRNAs, shRNAs or plasmids in this study was performed with a Lipofectamine 3000 Transfection Kit (Invitrogen). Transfection efficiency was established by determining the expression level of either gas6 or GAS6-AS2 by real-time qPCR and Western blot assays.

| Real-time PCR
Total RNA was isolated using TRizol reagent (Thermo Fisher Scientific) and reverse transcribed to cDNA according to the manufacturer's instructions. This mix (containing total cDNA, forward and reverse primer, Milli-Q water and SyberGreen reagent (Roche)) was subjected to thermal cycling performed in a 7500 Fast Time Real-Time PCR system (Applied Biosystems). PCR results were analysed using the 2 -ΔΔCT method and presented as the relevant expression level, as normalized to the level of housekeeping gene GAPDH.
All samples were amplified in duplicate, and all experiments were repeated three times. The primers used in this study were summarized on Chart 1 in Supporting Information.

| Western blot analysis
Total cellular or tissue protein was homogenized in highly efficient RIPA buffer (Solarbio) supplemented with a 1% complete protease inhibitor cocktail (Sigma-Aldrich) and, when necessary, phosphatase inhibitors. After sonication and centrifugation of the cell lysates, proteins in the supernatant were determined via BCA assay (Solarbio) and resolved on an 8% SDS-PAGE gel at 20-30 µg per lane as appropriate. These gels were electro-transferred onto a PVDF membrane.
Transfer was followed by antibody blocking of the membrane with 5% skim milk for 1 hour, incubation of the first antibody overnight at 4°C and subsequent HRP-conjugated second antibody incubation were expressed as a relative expression normalized to GAPDH level.

| Monocyte chemotaxis assay
HUVECs were seeded on basal side of 8.0-µm cell culture transwells (Corning) at a density of 3 × 10 4 cell/well. After these cultures reached full confluence, the following concentrations of P. gingivalis-LPS were used to stimulate the plated HUVEC for 24 hours: 0 μg/mL, 0.1 μg/ mL, 1 µg/mL and 10 µg/mL. When the effects of gas6 were observed, 1 μg/mL P. gingivalis-LPS were used to stimulate conditioned HUVECs for 24 hours. Then, THP-1 cells, pre-labelled with 20 µM Calcein AM for 30 minutes, were seeded onto apical side of the chamber (conc: 1 × 10 5 cell/well). Monocytes were observed transmigrating to the basal chamber after 3 hours using a Zeiss inverted microscope. Three of these images were randomly selected for analysis.

| Monocyte adhesion assay
HUVECs (1 × 10 5 cells/well) were seeded onto 12-well cell culture plates and allowed to form a cell monolayer. The cells were then stimulated by varying concentrations (0 µg/mL, 0.1 µg/mL, 1 µg/mL and 10 µg/mL) of P. gingivalis-LPS for 24 hours. In culture wells where the gas6 siRNA and overexpression plasmids were used, 1 μg/mL P. gingivalis-LPS was used to stimulate conditioned HUVECs for 24 hours. The culturing medium was replaced with fresh endothelial medium to eliminate the influence of LPS on monocytes added later.
THP-1 cells (5 × 10 5 cell/well) pre-labelled with 20 μM Calcein AM for 30 minutes were co-cultured with HUVECs for 4 hours. PBS was used to gently wash non-adherent THP-1 cells thrice; THP-1 cells that adhered to the surface of HUVECs were photographed using a Zeiss inverted microscope. Three of these images were randomly selected for analysis. A Western blotting assay was used to analyse the total protein extracted from these tissues. Immunodetection and qualification of gas6 was performed with antibodies against GAPDH and gas6 in 1:1000 dilution.

| Statistical analysis
All experiments were performed in triplicate. Results were expressed as means ± SE. An unpaired two tailed t test was performed to analyse data from two groups, and one-way analysis of variance (ANOVA) was performed to analyse data involving more than two groups. Significance analysis of data from healthy and inflammatory periodontal tissues was also performed with a t test. Values of P ≤ .05 were considered statistically significant. All data analysis was performed with SPSS software 21.0 (SPSS Inc). Corresponding symbols in figures are * for P < .05, ** for P < .01 and *** for P < .001.

| Chemotaxis and adhesion of monocytes to HUVECs was promoted by P. gingivalis-LPS stimulation
The effect of P. gingivalis-LPS infection on the expression of chemokines and adhesion molecules in HUVECs is shown in Figure 1.
The protein levels of ICAM-1 and E-selectin were significantly elevated compared to the negative control (P < .05) in HUVECs following LPS stimulation-in all three experimental concentrations. In addition, no difference in the protein level of MCP-1 or IL-8 was observed between the 0.1 μg/mL P. gingivalis-LPS experimental group and the control group (P > .05). Figure 1F shows images of THP-1 cells recruited by HUVECs in the transwell system after stimulation by the various experimental concentrations of P. gingivalis-LPS.
Monocytic chemotaxis towards HUVECs was observed to be facilitated by P. gingivalis-LPS in a dose-dependent manner. Similar effects of P. gingivalis-LPS were noted on the number of THP-1 cells that adhered to the surfaces of HUVECs, which is also demonstrated in Figure 1G.

| Gas6 inhibited P. gingivalis-LPS induced chemotaxis of monocytes towards HUVECs in vitro
As shown in Figure 2A Endothelial cells were cultured in 6-well plates and stimulated with different concentration of P. gingivalis-LPS for 24 hours, THP-1 cells were co-cultured with endothelial cells for 3 hours, images were captured after non-adherent monocytes were rinsed out gently with PBS for 3 times. Scale bars, 100 μm expression in P. gingivalis-LPS stimulated HUVECs was shown in Figure 2C-D. HUVECs that underwent gas6 knock-down also displayed increased levels of MCP-1 and IL-8 (compared to HUVECs that underwent P. gingivalis-LPS stimulation alone (P < .05), while the levels of these chemokines were conversely decreased (P < .05) in HUVECs that experienced gas6 overexpression. The effect of gas6 on chemotaxis within HUVECs (in vitro) was shown in Figure 2E.
After gas6 was knocked down and these cells underwent P. gingivalis-LPS stimulation, the number of THP-1 monocytes that migrated towards endothelial cells was considerably increased. Conversely, an inhibitory effect on chemotaxis was observed after gas6 was overexpressed in HUVECs.

| Both Axl and Mer receptors participated into the inhibitory effect of gas6
As shown in Figure 3A, Tyro3 expression within the HUVECs was not detected by Western blotting assays, and further analysis on Tyro3 receptor was therefore precluded. Axl and Mer receptors were blocked with selective small molecular inhibitors, R428 (10 μg/mL) and UNC2025 (10 nM), respectively. ICAM-1 and E-selectin expression in HUVECs were significantly elevated compared to P. gingivalis-LPS stimulation alone ( Figure 3B, P < .05).

| Akt/NF-κB pathway mediated gas6 inhibitory effect
Gas6 knock-down, followed by stimulation with P. gingivalis-LPS, significantly inhibited the expression of phosphorylated Akt

| Expression of gas6 was decreased in P. gingivalis-LPS stimulated HUVECs
The expression of gas6 and receptors, Axl and Mer, in HUVECs when  Figure 5F, P > .05). Gas6 expression was likewise inhibited when GAS6-AS2 was knocked down using GAS6-AS2 shRNAs ( Figure 5G); no difference in GAS6-AS2 expression level was observed after gas6 was knocked down or overexpressed ( Figure 5H, P > .05), from which we can conclude that GAS6-AS2 was an upstream regulatory factor for gas6 expression.

| D ISCUSS I ON
In this study, we found that gas6 protein within HUVECs inhibited the chemotaxis and adhesion of monocytes to endothelial cells stimulated by P. gingivalis-LPS. LncRNA GAS6-AS2, rather than GAS6-AS1, mediated the inhibitory effect of NF-κB on gas6 expression as was first reported.
The initial pathological process of atherosclerosis is characterized by circulating monocytes being recruited to dysfunctional gas6 is also expressed in platelets and interacts with endothelial cells, monocytes, and neutrophils. Cytokines secreted by platelets are stored in α-granules, facilitate leucocyte recruitment and participate in thrombosis. 35 Therefore, the involvement of gas6 from platelets in thrombosis cannot be ruled out. Considering the role of gas6 in immune and vascular system development 36  whether Tyro3 is also expressed in HUVECs remains to be determined. Tyro3 expression has not been detected in HUVECs via flow cytometry, 38 but was observed at the mRNA level in Tjwa's study. 33 A Western blotting assay was adopted in our studies. The monocytes group was used as a positive control, 39,40 and results indicated that LncRNA GAS6-AS2 mediated the inhibitory effect of NF-κB activation on gas6 expression ( Figure 6). Further studies regarding effect of gas6 on periodontitis and atherosclerosis in vivo may endow us with novel insights into the connection between these two diseases.

F I G U R E 6
Schematic representation for mechanisms of bi-directional regulation between gas6 and P. gingivalis-LPS in HUVECs. Expression of MCP-1, IL-8, ICAM-1 and E-selectin induced by P. gingivalis-LPS was inhibited by gas6 via Akt/NF-κB pathway; Gas6 expression in HUVECs was inhibited by P. gingivalis-LPS through NF-κB/GAS6-AS2 pathway

ACK N OWLED G EM ENTS
This work was supported by timely grants from the National Natural

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that supports the findings of this study are available in the supplementary material of this article.