Cholesterol crystals enhance TLR2- and TLR4-mediated pro-inflammatory cytokine responses of monocytes to the proatherogenic oral bacterium Porphyromonas gingivalis

Cholesterol deposits and pro-inflammatory cytokines play an essential role in the pathogenesis of atherosclerosis, a predominant cause of cardiovascular disease (CVD). Epidemiological evidence has linked periodontal disease (PD) with atherosclerotic CVD. Accordingly, viable periodontal pathogens, including Porphyromonas gingivalis, have been found in atherosclerotic plaques in humans and mice. We aimed to determine whether cholesterol crystals (CHCs) and oral bacteria synergize in the stimulation of human monocytes. Incubation of human monocytes with CHCs induced secretion of interleukin (IL)-1β, tumor necrosis factor (TNF)-α, IL-6, and IL-8. Moreover, CHCs markedly enhanced secretion of IL-1β by monocytes stimulated with the toll-like receptor (TLR) 4 agonist Escherichia coli lipopolysaccharide (LPS), and the TLR2 agonist Staphylococcus aureus lipoteichoic acid. Notably, CHCs also enhanced IL-1β secretion induced by P. gingivalis LPS and IL-1β secretion induced by whole P. gingivalis bacteria. This enhancement was abrogated by the NLRP3 inflammasome inhibitors Z-YVAD-FMK and glibenclamide. CHCs had no effect on cytokine production induced by P. gingivalis gingipains. Taken together, our findings support that CHCs, via stimulation of NLRP3 inflammasomes, act in synergy with the periodontal pathogen P. gingivalis to promote monocyte secretion of pro-atherogenic cytokines.


Introduction
Growing evidence suggests that periodontal disease (PD) is a risk factor for atherosclerotic cardiovascular disease (CVD) [1,2]. Atherosclerosis is generally accepted to be a chronic a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 inflammatory IL-10, and the chemokine IL-8. We also investigated the potential synergy between CHCs on the one hand, and Pg-LPS, Arg-gingipain and whole P. gingivalis bacteria on the other, in stimulation of these responses. In addition, we examined the role of inflammasomes in mediating P. gingivalis-and CHC-induced pro-inflammatory cytokine production.

Materials and methods Cells
Blood from anonymous, healthy blood donors attending the Blood Bank at Copenhagen University Hospital, Rigshospitalet, Denmark was used. Peripheral blood mononuclear cells (PBMCs) were isolated using Lymphoprep (Axis-shield, Oslo, Norway) gradient centrifugation. PBMCs were resuspended in RPMI 1640 with HEPES (Biological Industries, Haemek, Israel), L-glutamine and gentamicin (Fischer Scientific, Slangerup, Denmark), and monocytes were isolated using EasySep TM human CD14+ positive selection kit (STEMCELL technologies, Grenoble, France) according to the manufacturer's instructions. Purity was shown to be >90% when tested by flow cytometry.

Cholesterol crystal preparation
Cholesterol !99% (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in 95% ethanol (12.5 g/ L), heated to 60˚C, filtered, and left at room temperature to allow crystallization in 15 mL polypropylene tubes (Nunc TM , Fishcer Scientific, Slangerup, Denmark). CHCs were collected by filtering and grinding using a sterile mortar. CHCs were stored in plastic tubes at -20˚C until use. Any LPS contamination of CHCs was found to be below the detection limit of the Limulus amebocyte lysate assay QCL-1000 kit (Lonza, Walkersville, MD, USA).
Stimulation of monocytes with whole bacteria P. gingivalis (ATCC 33277) was cultured on trypticase soy blood agar plates containing 5 mg/ mL hemin and 50 μg/L Vitamin K for 4 days at 37˚C/10% CO 2 /10% H 2 O/80% N 2 . Subsequently, the bacteria were frozen in phosphate-buffered saline, pH 7.4, and kept at -80˚C. Isolated CD14+ monocytes were stimulated overnight (3x10 5 cells/well) with 4x10 5 thawed whole bacteria in RPMI 1640 with no serum added. After 20 hours at 37˚C/5% CO 2 , supernatants were harvested and analyzed for the presence of cytokines.

Analyses of cell viability using flow cytometry
Purified CD14+ cells were stained with anti-CD14-antibody (BD Bioscience, San Jose, CA) and 7-Amino Actinomycin D (7-AAD) and incubated on ice for 30 minutes. Cells were analyzed using a FACS canto II flow cytometer (BD Bioscience, San Jose, CA) and data were analyzed using FlowJo v.X, (TreeStar, Inc. Ashland, OR).

Statistical analyses
Statistical analyses were performed using GraphPad Prism 6.0 (GraphPad Software, CA, USA). Data are presented as means±SDs. Cytokine secretion upon stimulation with and without CHCs was compared using repeated measures ANOVA. Differences between paired data were assessed using paired sample t-tests. The Bonferroni correction was used where independent comparisons were made. Data were log10-transformed to obtain normal distribution when indicated. P<0.05 was considered significant.

Monocyte cytokine responses to CHCs
While examining the ability of CHCs to stimulate production of pro-inflammatory cytokines by unprimed human monocytes, we observed significant secretion of IL-1β, TNF-α, IL-6 ( Fig  1A-1C), and a statistical trend towards IL-8 secretion (Fig 1E). The CHC concentration used (2 mg/mL) was based on preliminary titration studies (S1 Fig).

Influence of CHCs on TLR2-mediated cytokine responses
Further we examined the effect of CHCs on TLR2-induced cytokine responses by stimulating monocytes with the TLR2 agonist Sa-LTA. Dose-dependent secretion of IL-1β, IL-6, IL-10, and IL-8 was observed (Fig 2). Again, IL-1β secretion was enhanced by co-incubation with CHCs at 0.1 and 1.0 μg/mL of Sa-LTA (Fig 2A), while CHCs had no effect on secretion of the other cytokines (Fig 2B-2E).

Influence of CHCs on cytokine responses to P. gingivalis gingipains
The cytokine production induced by another constituent of P. gingivalis, the cysteine protease Arg-gingipain activated by cysteine in vitro and added to monocyte cultures showed that Arggingipain induced secretion of IL-1β and TNF-α (Fig 3F and 3G), but not of IL-6, IL-10 ( Fig  3H and 3I), and IL-8 (data not shown). Furthermore, CHCs did not influence IL-1β or TNF-α secretion (Fig 3F and 3G). To further investigate the role of the NLRP3 inflammasome, we included the caspase-1/ NLRP3 inhibitor, Z-YVAD-FMK, and the ATP-sensitive potassium channel/NLRP3 inhibitor, glibenclamide in the experiments. Both Z-YVAD-FMK and glibenclamide caused a dosedependent decrease in IL-1β secretion by monocytes stimulated with a combination of Pg-LPS and CHCs, confirming that NLRP3 was critically involved in the induction of IL-1β secretion (Fig 4A and 4F). In addition, secretion of TNF-α was lowered by both Z-YVAD-FMK and glibenclamide (Fig 4B and 4G), while the two NLRP3 inhibitors had no effect on the production of IL-6, or IL-8 (Fig 4C, 4D, 4E, 4H, 4I and 4J). The data on IL-10 were ambiguous ( Fig  4D and 4I).

Role of inflammasomes in cytokine responses to Pg-LPS and CHCs
Addition of each inhibitor separately did not completely abolish secretion of IL-1β. However, addition of the inhibitors in combination at the highest concentration resulted in complete inhibition of IL-1β production (data not shown).
To examine the role of IL-1β for induction of TNF-α, IL-1 receptor antagonist (IL-1RA) was included in the experiments. IL-1RA did not affect secretion of IL-1β significantly ( Fig  5A), but inhibited the production of TNF-α (Fig 5B), indicating a role for IL-1β in induction of TNF-α, as also observed by others [34]. In addition, IL-6 production was nonsignificantly lowered after addition of IL-1RA (Fig 5C). Secretion of IL-10 and IL-8 was not affected by addition of IL-1RA (Fig 5D and 5E).
The notion that secretion of TNF-α was a consequence of stimulation with IL-1β was supported by the finding that IL-1β secretion production preceded TNF-α secretion by approximately 3 hours (Fig 5F).

Influence of CHCs on cytokine responses to whole bacteria
Incubation of monocyte cultures with P. gingivalis as whole bacteria led to production of IL-1β, TNF-α, IL-6, IL-10, and IL-8 (Fig 6A-6E). Addition of anti-TLR2 and Rs-LPS, separately or in combination, lowered secretion of all cytokines (Fig 6A-6E), demonstrating the involvement of both TLR2 and TLR4 in their induction. Anti-TLR2 and Rs-LPS inhibited the production of IL-1β by 65±18% and 44±17%, respectively, with no additive effect of their combined use (data not shown). The pan-caspase inhibitor Z-VAD-FMK completely abolished secretion of IL-1β, TNF-α, IL-6, IL-10, and IL-8 (Fig 6A-6E), demonstrating a critical role for inflammasomes in induction of all these cytokines by whole bacteria.

Discussion
Accumulation of CHCs plays a crucial role in atherogenesis, possibly via induction of proinflammatory cytokine responses through inflammasomes [5,6,11,12,25]. Infection with P. gingivalis which is thought to be a keystone pathogen in PD [20], has also been linked to the pathogenesis of atherosclerotic CVD [14,15,[19][20][21]. We therefore assessed the ability of CHCs to stimulate human monocytes for secretion of pro-inflammatory cytokines and to enhance proinflammatory cytokine responses to P. gingivalis, and we examined the role of inflammasomes and TLRs in these processes. CHCs alone were capable of inducing monocyte secretion of IL-1β, an important inflammatory mediator in CVD and PD [3,35,36]. Similar observations were made by Rajamäki et al., who used human adherent mononuclear cells and monocyte-derived macrophages [37]. Two other studies, however, found that CHCs alone did not induce IL-1β secretion by human PBMCs [12] or monocyte-derived macrophages [25]. These discrepancies may be explained by differences in incubation time (20 hours in our study versus 6 hours [12] and 6 days [25]), and the fact that we did not include GM-CSF or M-CSF [25]. We further observed that CHCs (2 mg/mL) alone induced secretion of TNF-α, IL-6, and IL-8 by monocytes. We recently found that matrix-bound cholesterol failed to stimulate monocyte-derived spindle-shaped cells for cytokine production [13], and this discrepancy may be due to differences in cholesterol formulation and differentiation stage of the cell cultures.
We show here that CHCs markedly enhance secretion of IL-1β by monocytes stimulated with the TLR4-ligand LPS from E. coli, supporting similar findings by Rajamäki et al. [37]. The Ec-LPS-induced production of TNF-α was not affected, however. The reverse pattern was found in our previous study using matrix-bound cholesterol [13]. CHCs also enhanced the secretion of IL-1β by monocytes stimulated with the TLR2-ligand lipoteichoic acid from S. aureus.
A main focus of this study was to examine whether P. gingivalis and CHCs acted synergistically in stimulation of pro-inflammatory cytokine production by monocytes, a molecular mechanism that could contribute to the association between PD and atherosclerosis. Indeed, CHCs markedly enhanced secretion of IL-1β induced by LPS isolated from P. gingivalis, in accordance with results of a recent study of human macrophages [25]. However, CHCs did not enhance the Pg-LPS-induced production of TNF-α or IL-6. Another component of P. gingivalis, Arg-gingipain, which is considered a key virulence factor, induced secretion of IL-1β and TNF-α, but not IL-6, IL-8 and IL-10. Similarly, previous studies showed that gingipains induced IL-8 and TNF-α production by human macrophages [32] and IL-1β and IL-10 production by the human monocytic cell line THP-1 [38]. Notably, CHCs did not enhance Arggingipain-mediated secretion of IL-1β by monocytes.
To our knowledge, this is the first study to assess the combined effect of CHCs and whole P. gingivalis bacteria on cytokine production by monocytes. P. gingivalis alone was a potent inducer of IL-1β, TNF-α-, IL-6-, and IL-10 secretion, as also described previously [39,40], and of IL-8 secretion, respectively, and TLR2 and TLR4 both contributed to P. gingivalis-mediated induction of cytokine responses. Notably, CHCs enhanced P. gingivalis-induced secretion of IL-1β by monocytes, but lowered the concomitant secretion of TNF-α, IL-6, IL-8, and IL-10. The underlying mechanisms and the consequences of these responses with respect to atherosclerotic plaque development and stability remain to be studied.
The pan-caspase inhibitor Z-VAD-FMK abolished the ability of CHCs to induce IL-1β secretion by monocytes and neutralized the CHC-mediated enhancement of IL-1β production induced by Pg-LPS and whole P. gingivalis. This implied that CHCs exerted their effects on IL-1β production via caspase activation, and thereby through inflammasomes. It has previously been demonstrated that CHCs activate the NLRP3 inflammasome in Ec-LPS stimulated macrophages [37] and human PBMC [12]. Here, we investigated the role of the NLRP3 inflammasome in the enhancement of IL-1β secretion induced by Pg-LPS and CHCs by use of the NLRP3 inhibitors, Z-YVAD-FMK and glibenclamide. We demonstrated that the NLRP3 inflammasome is involved in the CHC-mediated enhancement of IL-1β production by monocytes stimulated with LPS from the periodontal bacterium P. gingivalis.
In line with the findings of others [34], our experiments including IL-1RA showed that IL-1β enhanced the secretion of TNF-α, IL-6 and IL-10, but not IL-8. Also, IL-1β secretion began approximately 3 hours before TNF-α secretion.
Among the limitations of the study, it should be noted that we isolated monocytes based on their expression of CD14, which is also expressed by macrophages. The cell cultures may therefore have contained macrophages at low frequencies.
In summary, we demonstrated an inflammasome-dependent role for CHCs in enhancement of the pro-inflammatory cytokine responses of human monocytes exposed to the proatherogenic periodontal bacterium P. gingivalis. This interaction may contribute to the association between PD and atherosclerotic CVD.