Experimental Periodontitis Deteriorated Atherosclerosis Associated With Trimethylamine N-Oxide Metabolism in Mice

Ten 8-week-old male ApoE^−/− mice, and ten 8-week-old male C57BL/6J mice were obtained from the Model Animal Research Center of Nanjing University (Nanjing, China). The mice were housed in a controlled pathogen-free environment with free access to food and water for acclimatization (Cani et al., 2008), and then randomized to two groups: the control group and the experimental periodontitis group. They were fed Western Diet (WD) consisting of 21% (w/w) fat, 0.2% cholesterol, and 0% choline (TD88137–Harlan) for an additional 8 weeks. All experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Nanjing University (IACUC-D2102033).

P. gingivalis strain 33277 was cultured in a Brain Heart Infusion Broth (Beyotime Biotechnology, China) in an anaerobic environment for 48 h at 37 °C. The anaerobic state was kept by AnaeroPack (Mitsubishi Gas Chemical Company, Inc, Japan) in an anaerobic jar (Oxoid, England) (Arimatsu et al., 2014). The concentration of bacteria was determined with a spectrophotometer (SpectraMax M3, Molecular Devices, USA) at an optical density of 600 nm (OD = 10⁹ P. gingivalis per ml) (Zhang et al., 2014). A total of 10⁹ CFU of live P. gingivalis was collected by centrifugation, and then resuspended in 100 μl Phosphate Buffered Saline (PBS) with 2% carboxymethyl cellulose (Sigma-Aldrich, America). The bacterial suspension was given to each of the five mice gingival margin of the molars 5 times a week for 8 weeks. The other five mice in the control group were sham-administrated without the P. gingivalis. Twenty-four hours after the final intervention, the feces were collected from the live mice, and then the mice were anesthetized for tissues collection (Arimatsu et al., 2014).

The hemi-mandibles were scanned using a high-resolution Micro-CT (SkyScan1176, Bruker, Germany) for alveolar bone loss evaluation. After scanning, a set of slices were used for three-dimensional reconstructions. All images were reoriented such that the cement-enamel junction (CEJ) and the alveolar bone crest (ABC) appeared in the Micro-CT slice that was to be analyzed. The alveolar bone loss was measured from the CEJ to the ABC of the mesial surface of the first molar on the sagittal plane (Park et al., 2007; Madeira et al., 2013).

The collected mouse hearts were embedded in tissue freezing medium (Sakura, America), and then sliced and stained with oil red O. The stained plaque area of the aortic sinus was analyzed using Image J 1.37c (National Institutes of Health, USA) (Tomoki et al., 2011). Hematoxylin–eosin-stained (H&E) sections of the aortic arch and liver were used for morphometric analysis (Kesavalu et al., 2012).

Quantification of TMAO, TMA, Choline, Creatinine, Betaine, and L-Carnitine was performed using stable isotope dilution ultra-high-performance liquid chromatography–tandem mass spectrometry (UHPLC–MS/MS). The analysis used a Waters ACQUITY UPLC HSS T3 column (2.1 ∗ 100 mm, 1.8 um). Mobile phases A and B, respectively, comprised 0.1% formic acid and 0.1% acetonitrile in water. MS quadrupole and ion source temperatures were separately taken as 100 and 650°C. The ion transitions were m/z 76.1 → 58.1 for TMAO, m/z 60 → 40.5 for TMA, m/z 118.1 → 58.1 for Betaine, m/z 104.1 → 45 for Choline, m/z 114.1 → 44 for Creatinine and m/z 162.1 → 85 for L-Carnitine.

The mouse serum samples were collected by extracting the eyeballs, and assayed for total cholesterol (TC), triglyceride (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL), oxidized low-density lipoprotein (ox-LDL), lipopolysaccharide (LPS), interleukin (IL)-6, interleukin (IL)-1β, and tumor necrosis factor (TNF)-α. The levels of ox-LDL, LPS, IL-6, IL-1β, and TNF-α were determined by ELISA (Dakewe, China). Enzymatic colorimetry (Biosino Bio-Technology & Science Inc, China) was used for serum TC and TG analysis. The LDL and HDL concentrations were detected by the LDL/HDL-Cholesterol Kit (Biosino Bio-Technology & Science Inc, China) (Lalla et al., 2003).

The HepG2 cell line was obtained from ATCC. HepG2 cells were cultured in DMEM high glucose medium (10% fetal bovine serum, 1% penicillin and streptomycin) at 37°C and 5% CO₂ saturated humidity. When the cell adherence rate reached 80 to 90%, the cells were digested with trypsin, and the cell viability was above 90% for subpassage. One day before the experiment, the cells were digested with 0.25% trypsin and inoculated with DMEM high glucose medium containing 10% fetal bovine serum into a 6-well culture plate. The agents were added after the cells were completely adhered to the wall. There were 5 wells in each group.

The total RNA was extracted from the mouse liver, abdominal adipose, large intestine, small intestine samples, and HepG2 cells by RNAsimple Total RNA Kit (Tiangen, China). Reverse transcription PCR was carried out in the PCR Thermal Cyclers (Applied Biosystems, Thermo Fisher, America) for cDNA synthesis with PrimeScript™ RT reagent Kit with gDNA Eraser (Takara, Japan). Primers for real-time PCR were purchased from GenScript, and the primer sequences were listed in Supplemental Table 1. The gene expression analysis was realized finally in a final volume of 20 μl consisted by 0.5–5 ng cDNA, 900 nM each of the forward and reverse primers, and 10 μl iTaq Universal SYBR Green Supermix (Bio-Rad, America) in QuantStudio 7 Flex Real-Time PCR System (Applied Biosystems, Thermo Fisher, America). The PCR conditions were 2 min at 50°C, 10 min at 95°C followed by 40 cycles of two-step PCR denaturation at 95°C for 15 s and annealing extension at 60°C for 60 s. The relative amount of each studied mRNA was normalized to the relative quantity of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA, and the data were analyzed according to the 2^−△△CT method (Cani et al., 2008; Nakajima et al., 2015).

The mouse feces were collected after P. gingivalis administration for two months, and the bacterial DNA was extracted from feces by TIANamp Stool DNA Kit (Tiangen, China). The bacterial DNA extracted from feces was collected for gut microbiota analysis. The V3–V4 region of the 16S rRNA genes was amplified by PCR using a primer set (F: CCTAYGGGRBGCASCAG, R: GGACTACNNGGGTATCTAAT). The qualified DNA was used for sequencing analysis in Illumina HiSeq platform at the Novogene Bioinformatics Institute (Beijing, China) (Ma et al., 2018).

The statistical analysis of phenotypic traits was performed using SPSS22. For statistical analysis, either a student t-test or one-way ANOVA was performed, and the significance level was set at P = 0.05 which was shown in the figure legends.

Eight week old male ApoE^−/− mice were randomly divided into control and experimental periodontitis groups upon CMC and P. gingivalis administration. After 2 months of administration, the mandible specimens of the two groups were collected and scanned by Micro-CT. As shown in Figure 1A, the alveolar bone loss refers to the distance from the cementum-enamel junction (CEJ) to the alveolar crest (ABC) of the first molar. The results showed that the alveolar bone height of the periodontitis group was significantly lower than that of the control, and the difference was statistically significant (P <0.05) (Figure 1B). The experimental periodontitis group exhibited more plaque deposition under the intima than the control group (Figure 1C). Analysis of the atherosclerotic plaque area revealed a significant increase in the periodontitis group versus the control group (Figure 1D). The H&E-stained aortic sections were used for vessel wall assessment and invading mononuclear cell observation. As shown in Figures 1E, F, the experimental periodontitis group exhibited the lager intimal thickness, more plaque deposition under intima and more invading monocytes throughout arterial layers than those of the control group. As shown in Supplemental Figure 1, experimental periodontitis had significant effects on the serum lipid profile. The cholesterol levels shifted toward atherogenic levels.

The effect of experimental periodontitis on plasma TMAO, the important indicator of atherosclerosis, and its precursors were analyzed. The plasma TMAO level increased significantly in the experimental periodontitis group, compared to the control group. However, no significant difference was observed in the level of TMA and its precursors between the two groups, except that the plasma choline level decreased in the periodontitis group (P <0.01) (Figures 1G, H).

TMA, the precursor of TMAO, is a unique metabolite produced by the gut bacteria. The composition and abundance of intestinal microflora in the samples were detected by genome sequencing. There was no significant difference in the number of OTUs. The Chao1 index, ACE index, Shannon index, and Simpson index were lower in the periodontitis group than in the control group after a 2-month treatment (Figure 2A).

At the phylum level, the major 3 phyla of periodontitis and control mice were Bacteroidetes, Firmicutes, and Proteobacteria, and with no significant difference between the two groups (Figure 2B). Furthermore, the PCA analysis showed differences in microbiome composition between periodontitis and control mice (Figure 2C). The LEfSe analysis identified the characteristic bacteria at 2-month point. Lachnospiraceae_NK4A136_group and Acetatifactor were abundant in the periodontitis group, whereas Rikenellaceae_RC9_gut_group and Mycoplasmataceae were abundant in the control group (Figures 2D, E).

The analysis of different bacterial operational taxonomic units (OTUs) showed that 43 OTUs, including Lachnospiraceae_bacterium_A4 (OTU402), Lactobacillus_animalis (OTU79), and Parabacteroides_goldsteinii (OTU147) at species level, and also Lachnospiraceae_NK4A136_group (OTU15, OTU471), Mucispirillum (OTU37), and Ruminococcaceae_UCG-014 (OTU136), etc. at genus level, were significantly different between periodontitis model and control mice (Figure 2F). The full list of the OTU taxonomy is in Supplemental Table 2. As shown in Figure 2G, 37 KO ORTHOLOGY were at different levels between periodontitis and control mice, namely, K00108↗: choline dehydrogenase [EC:1.1.99.1], K00218↗: protochlorophyllide reductase [EC:1.3.1.33], K00228↗: coproporphyrinogen III oxidase [EC:1.3.3.3], etc.

FMO3 is another important factor for the formation of TMAO. After the 2-month P. gingivalis administration, the mRNA expression of FMO3 was found to be significantly increased in the mouse livers (Figure 3A). Meanwhile, the expression of IL-6 (Figure 3B) in the liver of the periodontitis group was also enhanced. Pathological observation showed an increased lymphocyte content in the liver of the experimental periodontitis mice, compared with the control group. Experimental periodontitis also induced more steatosis, and the loss of cellular boundaries in livers (Figure 3C). Further blood analysis revealed that the experimental periodontitis mice had elevated LPS levels (P <0.05), while changes in inflammatory factors IL-6 and TNF-α were consistent with LPS (Figures 3D, E).

The mRNA expression of IL-6, TNF-α, and IL-1β in the small and large intestines did not differ significantly between the two groups (Figures 3F, G). The expressions of tight junction proteins ZO-1, Claudin-1, and Occludin were analyzed for gut barrier function. In the small intestine, experimental periodontitis significantly downregulated ZO-1 expression (P <0.05), and Claudin-1 and Occludin also showed less expression in experimental periodontitis mice (Figure 3H). These three tight junction proteins had similar decreased mRNA expressions in the large intestine at 8 weeks in the experimental periodontitis group (Figure 3I). The abnormal functions of intestinal immunity may create conditions for gut microbiota and its toxins like LPS to get into the bloodstream in mice.

To understand the effect of LPS incitation on the FMO3 in the liver and the oxidation of TMA, 8-week-old male C57BL/6J mice were injected with LPS intraperitoneally at a dose of 5 mg/kg every 3 days for 2 weeks. The level of the inflammatory factors IL-6, IL-1β, and TNF-α in the liver increased significantly, compared to the control (Figure 4A). The mRNA expression of FMO3 in the liver was also increased in the LPS group (Figure 4B). Meanwhile, the plasma TMAO significantly increased in the LPS group compared to that in the control group, while there was no obvious effect on the level of plasma TMA and its precursors (Figures 4C, D).

The HepG2 cells were used to explore the direct effect of LPS and the immune effect of IL-6 on the expression of FMO3. The HepG2 cells were treated with 10, 50, and 100 μg/ml LPS, respectively, for 24 h. The mRNA expression of FMO3 increased significantly in the three LPS treatment groups, especially in the 100 μg/ml LPS group (P <0.05) (Figure 4E). Furthermore, a much more obvious increase of FMO3 expression was found after the treatment of HepG2 cells with 25 or 50 ng/ml IL-6 (Figure 4F).