Hyperlipidemia Impaired Innate Immune Response to Periodontal Pathogen Porphyromonas gingivalis in Apolipoprotein E Knockout Mice

A finely-tuned innate immune response plays a pivotal role in protecting host against bacterial invasion during periodontal disease progression. Hyperlipidemia has been suggested to exacerbate periodontal health condition. However, the underlying mechanism has not been addressed. In the present study, we investigated the effect of hyperlipidemia on innate immune responses to periodontal pathogen Porphyromonas gingivalis infection. Apolipoprotein E-deficient and wild-type mice at the age of 20 weeks were used for the study. Peritoneal macrophages were isolated and subsequently used for the study of viable P. gingivalis infection. ApoE−/− mice demonstrated inhibited iNOS production and impaired clearance of P. gingivalis in vitro and in vivo; furthermore, ApoE−/− mice displayed disrupted cytokine production pattern in response to P. gingivalis, with a decreased production of tumor necrosis factor-α, interleukin-6 (IL-6), IL-1β and monocyte chemotactic protein-1. Microarray data demonstrated that Toll-like receptor (TLR) and NOD-like receptor (NLR) pathway were altered in ApoE−/− mice macrophages; further analysis of pattern recognition receptors (PRRs) demonstrated that expression of triggering receptors on myeloid cells-1 (TREM-1), an amplifier of the TLR and NLR pathway, was decreased in ApoE−/− mice macrophages, leading to decreased recruitment of NF-κB onto the promoters of the TNF-α and IL-6. Our data suggest that in ApoE−/− mice hyperlipidemia disrupts the expression of PRRs, and cripples the host’s capability to generate sufficient innate immune response to P. gingivalis, which may facilitate immune evasion, subgingival colonization and establishment of P. gingivalis in the periodontal niche.


Introduction
Periodontal disease,a common infection-driven chronic inflammatory disease, is characterized by destruction of the supporting tissues around the tooth. Porphyromonas gingivalis, a gram-negative oral anaerobe, is implicated as a predominant contributor to human periodontitis. Within the privileged anatomical niche of periodontal pocket, P. gingivalis can replicate and destroy periodontal tissue [1]. The innate immune system uses sentinel cells (monocytes, neutrophils, macrophages, dendritic cells), bearing pattern recognition receptors (PRRs), to recognize common molecular patterns on periodontal bacterial invaders, generating the immune and inflammatory responses, which leads to clearance of the bacteria [2].
Hyperlipidemia, resulting from impaired lipid metabolism by undue intake of cholesterol or genetic deficiency, has been suggested to exacerbate periodontal parameters in periodontitis patients in epidemiologic studies [3,4,5]. High-cholesterol food or fatty acid itself may depress immune function and bactericidal effect on P. gingivalis in the host [6,7], and feeding animals a high-cholesterol diet impairs lipid metabolism and enhances periodontal destruction in lipopolysacchride (LPS) induced periodontitis in rats [8]. However, the underlying mechanism by which hyperlipidemia influences periodontal disease progress, especially the interaction between the innate immune system (including the PRRs) and the periodontal pathogen remains to be established.
Nearly half of the population above 20 years of age in the developed countries like USA and one third of the population above 35 years of age in the developing countries like China have hyperlipidemia [9,10], whereas nearly half of the Chinese adults are inflicted with moderate to severe form of periodontal diseases [11]. Therefore, it is of great importance to understand the relationship between hyperlipidemia and periodontal disease. Although it has been demonstrated that periodontal diseases could influence lipid metabolism in the serum and the blood vessel [12,13], it is not clear whether and how lipid metabolism anomaly in the blood may also affect periodontal disease progress.
Apolipoprotein E deficient (ApoE 2/2 ) mice develop severe hyperlipidemia under normal diet condition [14]. In the present study, we tested our hypothesis that long term hyperlipidemia itself, rather than diet lipid, impaired the host immune response to periodontal infection in ApoE 2/2 mice compared to healthy C57BL/6 mice. We found that long term hyperlipidemia impaired the immune response to P. gingivalis challenge by altering PRRs expression pattern in macrophages, leading an inhibited cytokine network response and decreased bacterial clearance; therefore, hyperlipidemia may lead to more severe periodontal bone loss.

Materials and Methods
Bacteria P. gingivalis 33277 was obtained from the American Type Culture Collection, and cultured in brain heart infusion (BHI) broth (Oxoid,UK) supplemented with 5 mg/mL of hemin and 0.5 mg/mL of menadione at 37uC in an anaerobic environment consisting of 90% N 2 , 5% CO 2 and 5% H 2 . Bacterial suspensions were prepared from cultures at their log phase of growth. Bacterial concentration was evaluated by measuring optical density at 600 nm (OD = 1, corresponding to 10 9 bacteria/mL), and adjusted to the desired treatment concentration by dilution with phosphate buffered solution (PBS).

Animals
One hundred six-week-old male C57BL/6 wild type (WT) mice and ApoE 2/2 mice (Jackson Laboratories, BarHarbor, Me, US) were fed a regular mouse chow diet ad libitum until about 20 wks of age to experience long term hyperlipidemia. All animal studies were performed in accordance with the policies of the Institutional Animal Care and the study has been approved by the Animal Ethics Committee at Fujian Medical University.

Serum Lipid Level
The serum total cholesterol (TC), triglyceride (TG), low density lipoprotein cholesterol (LDL-c) and high density lipoprotein cholesterol (HDL-c) concentration were determined by an enzymatic method using the protocols provided by the company (Biosino, Beijing, China). Detection sensitivity for TC, LDL-c, HDL-c and TG is 0.65 mmol/L, 0.50 mmol/L, 0.13 mmol/L and 0.15 mmol/L respectively.

Peritoneal Infection
WT and ApoE 2/2 mice were infected i.p. with P. gingivalis 33277 (5610 7 CFU). Peritoneal lavage was performed at indicated time post-infection. Serial 10-fold dilutions of peritoneal fluid were plated onto blood agar plates supplemented with hemin/menadione and cultured anaerobically for enumerating recovered peritoneal CFU. Furthermore, the periondontal lavage was centrifuged at 3006g at 4uC for 5 min. Then the cellular components were utilized in Western blot analysis for detecting iNOS protein, whereas the supernatants were used in both cytokine antibody array and ELISA to measure cytokine production.

Cytokine Profile Analysis
Cytokine production in peritoneal fluid supernatants was analyzed for levels of 62 cytokines by Raybiotech Mouse Cytokine Antibody Microarray (AAM-CYT-G3, RayBiotech, Narcross, GA, US) according to the manufacturer's instructions and as described earlier [15]. This antibody array detects simultaneously 62 mouse cytokines. Briefly, cytokine array membranes were blocked in 2 mL of 16blocking buffer for 30 min, and then incubated with 1 mL of samples at room temperature for 2 h. Samples were then decanted from each container, and the membranes were washed three times with 2 mL of l6wash buffer I, followed by two washes with 2 mL of l6wash buffer II at room temperature with shaking. Membranes were then incubated in 1:250 diluted biotin-conjugated primary antibodies at room temperature for 2 h and washed as described above before incubation in 1:1000 diluted HRP-conjugated streptavidin. After incubation in HRP-conjugated streptavidin for 60 min, membranes were washed thoroughly and exposed to a peroxide substrate (detection buffers C and D) for 5 min in the dark before imaging. Membranes then were imaged using Axon GenePix laser scanner using cy3 channel. The signal intensities were imported into RayBio H Analysis Tool. The densities were exported into Microsoft Excel, and the background intensity was subtracted prior to analysis. After subtracting background signals and normalization to positive controls, comparison of signal intensities for antigen-specific antibody spots between groups were utilized to determine relative differences in expression levels of each analyte (i.e., protein detected). Any $2-fold increase or #0.5-fold decrease in signal intensity for a single analyte between groups were considered a significant difference in expression, provided that both sets of signals were well above background (mean background62 standard deviations).

Macrophage Intracellular Bacteria Killing Assay
The intracellular P. gingivalis clearance capability of macrophages was determined by an antibiotic protection-based survival assay, as Wang et al described [17]. Briefly, following incubation of P. gingivalis with macrophages (at a multiplicity of infection [MOI] of 25:1) for 1.5 h, extracellular nonadherent bacteria were removed by washing with PBS, whereas extracellular adherent bacteria were killed by addition of gentamicin (300 mg/mL) and metronidazole (200 mg/mL) for 1 h. After washing, macrophages remained to be cultured overnight. Internalized bacteria were released by lysis of macrophages in sterile distilled water for 20 min. Serial dilutions of the lysates were plated onto blood agar dishes supplemented with hemin and menadione, and cultured anaerobically for CFU enumeration. Phagocytosis P. gingivalis was labeled with FITC as Hazenbos et al described [18], then uptake of FITC-labeled P. gingivali was detected by flow cytometry as Wang et al described [17]. Briefly, primary mouse macrophages were incubated at 37uC with FITC-labeled P. gingivalis at a MOI of 25:1 for 30 min. Phagocytosis was stopped by cooling the incubation plates on ice. Cells were washed to remove nonadherent bacteria, then extracellular fluorescence (representing attached but not internalized bacteria) was quenched with 0.2% trypan blue in some groups. Both quenched and unquenched cells were analyzed by flow cytometry (% positive cells for FITC-P. gingivalis and mean fluorescence intensity [MFI]) using the FACSCalibur (Becton-Dickinson). Association (representing both adherence and phagocytosis) or phagocytic indices were calculated using the formula (% FITC-positive cells6MFI)/ 100. When macrophages were pretreated with cytochalasin D to block phagocytosis, macrophages that were stimulated with FITClabeled P. gingivalis did not show significant fluorescence, which confirmed that trypan blue effectively quenched extracellular fluorescence.

Microarray Analysis
Two hours after infection with P. gingivalis (MOI = 25:1), total RNA of mouse macrophages was extracted using Trizol reagent (Invitrogen, Carlsbad, California, US) and was further purified by RNeasy micro kit and RNase-Free DNase Set (QIAGEN, Germany). Three independent experiments were performed for each condition. Total RNA was amplified, labeled and purified by using Affymetrix GeneChip 39IVT Express Kit to obtain biotin labeled cRNA. Following array hybridization, slides (Affymetrix, Santa Clara, CA, US) were scanned by GeneChip Scanner 3000 and Command Console Software 3.1 (Affymetrix, US). Differentially expressed genes showing a statistical significance (p,0.05, one way analysis of variance) and $2-fold change between ApoE 2/2 and the corresponding WT mice were further analyzed using SBC analysis system (Biochip, Shanghai,China): gene ontology (GO) analysis was applied to organize genes into hierarchical categories on the basis of cellular component, biological process and molecular function; enrichment pathway analysis (Kegg database) was implemented to uncover the major perturbed pathways. Complete microarray data have been deposited in the Gene Expression Omnibus public database (accession number GSE42061; http://www.ncbi.nlm.nih.gov/ geo/query/acc.cgi?acc = GSE42061).

Quantitative Real Time PCR (qPCR)
Total RNA was isolated from the cells using Trizol reagent. One microgram of total RNA was reverse-transcribed using Superscript II (Invitrogen, USA). Real-time quantitative PCR reactions were set up in triplicate using the SYBRH Premix Ex Taq TM II (Takara, Japan) and run on a 7900 HT Sequence Detection System (ABI, USA). All of the gene expression levels were normalized to the internal standard gene, GAPDH.

Western Blot
The cellular components in peritoneal fluid were collected after centrifugation, then washed in Dulbecco's phosphate-buffered saline (D-PBS) twice; for in vitro study, macrophages were harvested after trypsin-EDTA treatment, and rinsed with D-PBS for three times. Cytoplasmic protein was extracted using NE-PER nuclear and cytoplasmic extraction reagents and Halt protease inhibitor cocktails (ThermoScientific, Rockford, IL, US) according to the manufacturer's protocol. Protein concentration was determined by a Bradford protein assay using BSA for standardization, and the cytoplasmic extracts were fractionated on 10% SDS-PAGE gels, transferred to a nitrocellulose membrane and incubated overnight with rabbit polyclonal iNOS primary antibodies (dilutions 1:500) (Abcam, Cambridge, MA, US), followed by incubation with horseradish peroxidase (HRP)conjugated goat anti-rabbit secondary antibody (1:5, 000) (Gene-script, Piscataway, NJ, US). A 1:3,000 dilution of GAPDH rabbit polyclonal antibody (Genescript, USA) was used as a loading control. Bands were visualized by enhanced chemiluminescence (ECL) method. X-ray films were scanned and saved as grayscale JPEG files. The level of expression of proteins was analyzed by using the Image J software (National Institutes of Health, US), which can calculate area and pixel value statistics of user-defined selections. Each band was normalized to GAPDH. Each sample was tested in duplicate, and western blots were repeated three times.

Chromatin Immunoprecipitation Assays
Mouse macrophages (1610 7 cells per sample) were stimulated, washed with PBS, and fixed with 1% formaldehyde for 10 min at room temperature. Formaldehyde fixation was stopped with the addition of 1.25 M glycine. Sonicated chromatins were then immunoprecipitated (IP) with anti-p65 (Cell signaling Tech, US) or with control IgG Abs coupled to microbeads. DNAs from each sample (input or IP) were isolated by elusion, reverse cross-linking and proteinase K treatment. The DNA was then used as template to perform PCR amplification of the regions containing the proximal nuclear factor-kB (NF-kB) site in the promoters of TNFa, IL-6 and IL-10.

Statistical Analysis
Data were represented as the mean 6 standard deviation (SD). Mean values between the groups were compared using analysis of variance (ANOVA). Differences were considered significant when P,0.05.

Metabolic Characteristics of Experimental Animals
ApoE 2/2 mice weighed less than WT mice above 12 wks of age, and blood glucose level in ApoE 2/2 mice was not significantly different from WT mice. ApoE 2/2 mice showed significantly increased LDL-c levels in blood at 10 and 20 weeks, whereas serum triglyceride was not significantly different ( Figure 1).

Impaired Clearance of P. gingivalis in Hyperlipidemic Mice was Parallel with the Reduced iNOS Production
Based on antibiotic protection-based intracellular killing assay, we found that macrophages from hyperlipidemic ApoE 2/2 mice showed a decreased tendency to internalize P. gingivalis after 1.5 h incubation (P.0.05), whereas intracellular capability to remove P. gingivalis was impaired in macrophages from ApoE 2/ 2 mice in vitro (figure 2A and 2B). Such impaired clearance in ApoE 2/2 mice was confirmed in the peritonitis model in vivo (figure 2C).
In order to explore the changes in phagocytosis capability in macrophage, we utilized flow cytometry to measure the phago-cytosis of FITC-labled P. gingivalis, and observed that macrophages from ApoE 2/2 mice displayed similar binding and internalizing capability (figure 2D).
One of the fastest and most effective defense mechanisms for macrophages to clear invading bacteria is production of the free radical NO, which is mediated by the expression of inducible NO synthase (iNOS) [19]. Mice lacking iNOS demonstrated impaired killing of P. gingivalis and increased periodontal tissue damage upon P. gingivalis challenge [20,21]. To investigate the cause of impaired clearance of P. gingivalis in ApoE 2/2 mice, we analyzed the expression of iNOS. iNOS production was significantly decreased in the peritonitis model in ApoE 2/2 mice when compared to WT mice (figure 2E). Macrophages from ApoE 2/2 mice displayed severe disruption of iNOS expression at both the protein and mRNA levels after P. gingivalis infection (figure 2F-i, ii).

Depressed Cytokine Production during P. gingivalis Infection in ApoE 2/2 Mice
To further investigate the cause of impaired clearance of P. gingivalis in ApoE 2/2 mice, we explored the cytokine profile during P. gingivalis infection. Six hours after peritoneal infection, we observed that 23 cytokines, including TNF-a, IL-2, IL-4 and interferon c, were decreased in peritoneal fluid, whereas only 5 cytokines such as leptin, macrophage inflammatory protein-1a (MIP-1a) and platelet factor 4 (PF4) were increased in ApoE 2/2 mice when compared to WT mice using mice cytokine arrays (figure 3A-C).
Cytokine production pattern was also determined by ELISA at 0 h to 24 h after bacterial periotoneal infection. Significantly less induction of pro-inflammatory cytokines, TNF-a, IL-6, IL-1b and MCP-1, was found at various time post-infection in ApoE 2/2 mice; whereas anti-inflammatory cytokine IL-10 was not affected, and IL-12 production was not detectable during infection ( figure 3D-I). Similarly significantly blunt cytokine production was also found in P. gingivalis infected mice macrophages in vitro ( figure 4A-F).
To further clarify whether the decreased cytokine production in ApoE 2/2 mice was due to long term hyperlipidemia or the disruption of ApoE gene expression itself, we further analyzed cytokine production in ApoE 2/2 and WT mice at the age of 6 wks in peritoneal infection model. In contrast to an impaired cytokine production pattern in ApoE 2/2 mice at 20 wks of age, a modestly but significantly increased cytokine production pattern was found in peritoneal stimulation test in ApoE 2/2 mice at 6 wks of age using ELISA (not depicted).
Disrupted Gene Expression Profile after P. gingivalis Infection in ApoE 2/2 Mice Macrophages A full genome microarray was utilized to investigate global gene transcription following P. gingivalis infection in macrophages, showing that mRNA transcription was inhibited in ApoE 2/2 mouse macrophages. A hierarchical clustering algorithm was used to group qualifiers with similar patterns of expression. A heat map of genes of fold change of .5.0 was shown in figure 5A. 353 genes were up-regulated and 465 genes were down-regulated in ApoE 2/2 control when compared to WT control; whereas 148 genes were up-regulated and 236 genes were down-regulated in P. gingivalis infected ApoE 2/2 macrophages (figure 5B). We then further investigated the logical relations between the two comparisons, 130 genes were differentially expressed in both untreated and stimulated macrophages (figure 5C); furthermore, among such 130 genes, 101 genes were down-regulated and 19 were up-regulated in both untreated and stimulated macrophages ( figure 5D).
To gain insights into the mechanism of disrupted macrophage response to live P. gingivalis, we focused on differential expression genes in P. gingivalis infected macrophages utilizing gene ontology and pathway analysis. We categorized these genes as down-regulated genes (236 genes) and up-regulated genes (148 genes). In Biological Process analysis, the up-regulated genes are involved in biological adhesion; whereas down-regulated genes take part in seven biological processes majorly involved in bacteria infection such as cell killing, immune system process and response to stimulus (Table 1). In enrichment pathway analysis (Kegg), the up-regulated genes are involved in extracellular matrix-receptor interaction and p53 signaling pathway; the down-regulated genes are involved in pathways such as cytokine and cytokine receptor interaction, NOD-like receptor (NLR) and Toll-like receptor (TLR) pathway (Table 2).

Different Pattern Recognition Receptors (PRRs) Expression Pattern in ApoE 2/2 Mice
Following the clue that TLR and NLR pathway were affected in ApoE 2/2 mice, we further analyzed the PRRs expression. As shown in figure 6A, transcription of TLR-2, -4, triggering receptors expressed on myelod cells -1 (TREM-1), TREM-3 and CD14 was decreased in untreated macrophages of ApoE 2/2 mice, whereas only TREM-1 mRNA was significantly decreased in P. gingivalis infected macrophages in ApoE 2/2 mice. TLR-2 protein expression was not affected in both the untreated and stimulated macrophages, while TREM-1 expression was significantly decreased in ApoE 2/2 mice after bacteria stimulation ( figure 6B). To test the effect of TREM-1 inhibition on immune response, we blocked TREM-1 with LP-17 peptide 2 h before stimulation with P. gingivalis, and TREM-1 blockage significantly inhibited TNF-a and IL-6 production upon P. gingivalis infection at 6 h in WT mice macrophages (figure 6C). Furthermore, macrophages from WT mice treated by LP-17 peptide displayed  The proportion of total count versus the number of genes in the biological process. 3 Only the biological process with p-value ,0.05 were shown. Only biological processes that are related to innate immune response to bacterial infection were shown. doi:10.1371/journal.pone.0071849.t001 less production of cytokine when compared to macrophages from ApoE 2/2 mice.

Less p65 Binding to Gene Promoters after Decreased TREM-1 Expression
As TREM-1 is an amplifier of both the TLR and NLR pathway, down-regulation of TREM-1 may lead to inadequate signal transduction upon P. gingivalis infection. To further characterize its effect on the innate immune response to bacteria invasion, we focused on the transcriptional factor NF-kB, which plays a key role in regulating the immune response to infection. We further evaluated the capability of NF-kB p65 subunit's binding ability to DNA at TNF-a, IL-6 promoter (classic proinflammatory locus) and IL-10 promoter (anti-inflammatory locus). Recruitment of p65 to all 3 promoters in macrophages from ApoE 2/2 mice at 15 min was decreased in comparison with WT mice (Figure 7). Although no difference was observed in the recruitment of NF-kB to gene promoters at 30 min between ApoE 2/2 mice and WT mice macrophages, binding to kB sites in ApoE 2/2 mice macrophages at 30 min was still less than that in WT mice macrophages at 15 min.

Discussion
In normal mice, a well-orchestrated cytokine network maintains a finely-tuned response to bacterial challenge. Cytokines such as TNF-a and IL-1b that are considered harmful to periodontal tissue in chronic inflammation play important roles in the control of pathogen infection in the acute inflammation [22]. The principal finding of the present study is that hyperlipidemia, a kind of metabolic anomaly accompanying many obesity subjects, disrupts host capability to clear periodontal pathogen P. gingivalis, as a result of impaired capability to mount sufficient immune response to P. gingivalis. To our knowledge, this is the first report that systematically investigated the potential mechanisms by which hyperlipidemia interferes with host's normal innate response to periodontal bacterial infection.
Hyperlipidemia may directly influence disease induction and indirectly modulate immune responses [23]. Lipoproteins can bind LPS and decrease LPS-stimulated cytokine production [24]; therefore, the inhibited inflammatory response in our study may partly be attributed to the elevated low density lipoprotein in the circulation system. In the present study, we found that the ApoE 2/ 2 mice failed to generate a robust immune response. Congruent with our results, similar observation was also made during Klebsiella pneumoniae and Leishmania major infection [23,25]. As ApoE itself has an immunoinhibitory effect on inflammation [26], animals deficient in ApoE at an early age exhibit increased cytokine production in response to Listeria monocytogenes [27], Mycobacterium tuberculosis [28],Klebsiella pneumoniae [29], and display inhibited type I inflammatory response to E. coli LPS, a classical TLR4 agonist, in vivo [30].
In contrast to the impaired cytokine production in ApoE 2/2 mice at 20 wks of age, a slightly increased cytokine release to P. gingivalis was observed in vivo in ApoE 2/2 mice at 6 wks of age. Therefore, the inhibited immune response in ApoE 2/2 mice at 20 wks of age can be attributed to long term hyperlipidemia, which overcomes the effects of ApoE deficiency, leading to decreased cytokine release and iNOS production. Furthermore, iNOS deficiency may impair host's ability to kill P. gingivalis and aggravate soft tissue damage and alveolar bone loss [20,21,31]. Our study is congruent with the study of Shamshiev et al [32],  showing that dyslipidemia inhibits TLR induced activation of CD8adendritic cells in ApoE 2/2 mice in aged mice (20-30 wksage), rather than young mice (5 wk).
In agreement with the inhibited cytokine production in ApoE 2/ 2 mice, microarray data showed that overall gene response in ApoE 2/2 mice was inhibited, and biological processes related to innate immune response such as cell killing, immune system process, response to stimulus, became impaired. Such gene expression profile would alter the clearance of P. gingivalis in macrophages in vitro and in vivo. To further discover the mechanism of hyperlipidemia interfering with the innate response, pathway analysis was conducted, showing that TLR and NLR pathway was influenced by hyperlipidemia.
Recognition of invasive pathogens by immune cells relies on their capacity to detect microbial-or pathogen-associated molecular patterns (PAMPs) [19]. This process involves the coordinated actions of different membrane-associated or intracellular patternrecognition molecules, including TLRs and NLRs [33]. Ligandactivated receptors trigger the mitogen-activated protein kinases (MAPKs), NF-kB, and interferon-related factor (IRF) signal transduction pathways that induce the transcription and production of immune genes, including cytokines that are critical for the activation of innate immunity [34].
Following the clue in microarray results, we further demonstrated that although TLR-1, -2, -4 and -6, NOD-1 and -2 expressions upon P. gingivalis challenge was not affected, whereas TREM-1 became significantly inhibited in ApoE 2/2 mice. In addition, in the present study, blockage of TREM-1 by LP-17 peptide significantly inhibited IL-6 and TNF-a production in response to P. gingivalis in macrophages. As TREM-1 is an amplifier of both the TLR and NLR signal transduction pathway [35,36], an inhibited TREM-1 expression may be a potential cause leading to decreased inflammatory signal transduction and cytokine release upon P. gingivalis infection. Such effect was confirmed by the decreased recruitment of NF-kB onto the promoters of the pro-inflammatory TNF-a and IL-6, and the antiinflammatory IL-10. Although we observed an inhibited TREM-1 expression in ApoE 2/2 mice macrophages, caution is still required for the result explanation. Over-expression of TREM-1 in ApoE 2/2 macrophages would further shed insight into the contributory role of TREM-1 in the disrupted cytokine production in ApoE 2/2 mice.
Understanding kinetic properties of NF-kB interaction with its target genes, i.e. the binding and unbinding, may help us understand the inflammatory response to P. gingivalis. After degradation of their cytoplasmic inhibitors (IkBs), NF-kB dimers translocate to the nucleus and bind to thousands of 9-10 nt kB sites dispersed in the genome [37,38], generating complexes with a half-life of 45 min [39]. As promoter-bound NF-kB is in dynamic equilibrium with nucleoplasmic dimers, promoter occupancy and transcriptional activity oscillate synchronously with nucleoplasmic NF-kB [40]. In addition to its ability to bind to NF-kB in cytosol, IkBa can also dissociate NF-kB from DNA and export it to the cytosol [41]. In our experiment, we observed that binding of NF- kB in ApoE 2/2 mice macrophages at 30 min was similar to that in WT mice at 30 min, but it was still less than that in WT mice at 15 min. Therefore, NF-kB in ApoE 2/2 mice macrophages demonstrated a later and less competent binding to its promoters, leading to the decreased production of IL-1b, IL-6, TNF-a, MCP-1 both in vivo and in vitro.
Activation of monocytes by TLRs stimulants, such as LPS, results in the rapid production of proinflammatory cytokines, including TNF-a, IL-1b, and IL-6, followed later by the secretion of the anti-inflammatory IL-10. However, in our study, IL-10 production in ApoE 2/2 mice was not altered in vivo and in vitro, which was inconsistent with the decreased recruitment of NF-kB to IL-10 gene promoters. As both NF-kB and MAPK signal pathways, including ERK, JNK, and p38, are involved in the classical LPS-induced IL-10 expression [42,43], further exploration on MAPK pathway might unravel the possible role of NF-kB binding to the IL-10 promoter.
Although we observed an inhibited innate immune response and TREM-1 expression in hyperlipidemic ApoE 2/2 mice, the possible mechanism by which hyperlipidemia influences PRRs has not been fully understood. Macrophages express a host of receptors, such as CD36 and scavenger receptor A, which bind and internalize modified forms of LDL. Thus serum lipid may be incorporated into cell membranes and cytoplasm, altering immune response to PAMPs. Indeed, patients with hyperlipidemia abnormities displayed altered lipid composition of the cell membrane and an increased cholesterol contents in cytoplasm in erythrocytes and lymphocytes [44,45,46,47]. Furthermore, macrophages from ApoE 2/2 mice showed 3-fold higher membrane cholesterol compared with wild-type macrophages [48].
Subtle changes in the lipid composition of the plasma membrane, particularly the cholesterol content, may result in the modification on lipid rafts, which provide spatial organization for transmembrane proteins and act as signaling platforms responsible for coordinating outside-in and inside-out signal transduction [49]. Toll-like receptors (TLR) and interleukin receptors are found within the lipid rafts of cells treated with pro-inflammatory LPS [50,51,52]. Therefore, alteration in lipid contents in circulation system may influence cholesterol compo- Figure 8. Proposed mechanism for impaired host innate immune response to viable P. gingivalis infection by hyperlipidemia in ApoE 2/2 mice. Increased blood lipid would be incorporated into cell membrane and intracellular components of macrophages, leading to the decreased TREM-1 expression upon P. gingivalis infection. Reduced TREM-1 activation resulted in inadequate signal transduction, less p65 nuclear translocation and binding to gene promoters region at DNA; therefore, innate immune responses including iNOS production and release of cytokines, such as TNF-a, IL-6 and MCP-1, were disrupted in the hyperlipidemic host. LDL, low density lipoprotein; TREM-1, triggering receptors expressed on myeloid cells-1; PRRs, pattern recognition receptors, TLR, Toll-like receptors; NLR, NOD-like receptors; NF-kB, nuclear factor kB; IkB, inhibitory kB; IKK, IkB kinase complex; iNOS, inducible nitric oxide synthase; IL-6, interleukin 6; MCP-1, monocyte chemoattractant protein 1; TNF-a, tumor necrosis factor a. doi:10.1371/journal.pone.0071849.g008 sition in cell membrane and affect lipid rafts dynamics, leading to altered inflammatory response to bacteria infection. P. gingivilis may exploit various receptors in cell membrane to facilitate its invasion, such as TLR2, CD14, CXC-chemokine receptor 4 (CXCR4), complement receptor 3 (CD11b/CD18) [17,53,54,55]. In our study, macrophages from ApoE 2/2 mice showed inhibited transcription of TLR2, TREM-1, TREM-3, and CD14 in quiescent state, whereas only TREM-1 expression was influenced by hyperlipidemia in bacteria stimulated macrophages.
In addition to membrane receptors, P. gingivalis utilizes lipid rafts as signaling and entry platforms [56]. P. gingivalis colocalizes with lipid rafts in a cholesterol-dependent way, depletion of cellular cholesterol using methyl-beta-cyclodextrin resulted in about 50% inhibition of P. gingivalis uptake, and this effect was reversed by cholesterol reconstitution [56]. Macrophages from ApoE 2/2 mice show higher membrane cholesterol than wild-type macrophages [48]. Therefore, an increased internalization might be expected in macrophages from ApoE 2/2 mice. In contrast, we failed to observe changes in phagocytosis in macrophages. Possible roles of TREM-1 in the inside-out mechanism of P. gingivalis need to be explored further.
Consistent with Bostanci et al [57], we observed that TREM-1 takes part in the immune response to P. gingivalis in mice macrophages. Following stimulation by LPS in vitro and in vivo, monocytes secrete a soluble form of TREM-1 (sTREM-1) [58]. sTREM-1 has been found to correlate with inflammatory disease, such as sepsis [59,60], relapsing polychondritis [61] and inflammatory bowel diseases [62]. Elevated oral and systemic levels of soluble TREM-1 has been found in periodontitis [63,64]. Therefore, monitoring changes in sTREM-1 level would help understand inflammation and immune response in the periodontal niche.
The possible mechanism by which hyperlipidemia impairs innate immune responses to viable P. gingivalis infection has been illustrated and summarized in figure 8. Briefly, increased serum lipid would alter the cholesterol component in the cell membrane and cytoplasm of macrophages, which might further affect PRR expression on the cell membrane upon P. gingivalis challenge. Disrupted activation of TREM-1, a kind of PRRs amplifying the TLR transduction pathway, could lead to inhibited p65 nuclear translocation and binding to gene promoters region. Consequently, innate immune responses including iNOS production and cytokines release, such as TNF-a, IL-6 and MCP-1, might be suppressed in the hyperlipidemic host.
Given that chronic periodontitis is characterized by alternate periods of exacerbation and quiescence [65], the impaired TREM-1 expression and the inhibited innate immune response would have a significant influence on the periodontal disease progress. This inability to mount a sufficient inflammatory response may lead to inadequate elimination of the bacteria, which would in turn grow and proliferate in the periodontal niche, and cause more severe periodontal damage.
In conclusion, we have shown that hyperlipidemia inhibits inflammatory response in macrophages, which in turn cripples host's capability to remove the bacterial infection. These findings suggest that hyperlipidemia can predispose the host to oral infection, by impairing proper immune response to bacteria challenge.