Involvement of a periodontal pathogen, Porphyromonas gingivalis on the pathogenesis of non-alcoholic fatty liver disease

This study was conducted with the approval of the Ethics Committees of Yokohama City University, Osaka University, and Nara City Hospital. A total of 150 Japanese NAFLD patients (102 with NASH and 48 with NAFL) and 60 control subjects (non-NAFLD) from similar socioeconomic strata were recruited. After provision of adequate explanation, written informed consent for participation in this study was obtained from all the subjects prior to the sample collections. All control subjects (non-NAFLD) were confirmed to have normal liver function (aspartate aminotransferase (AST), alanine aminotransaminase (ALT) levels within normal range), no evidence of viral hepatitis and no evidence of the presence of criteria of the metabolic syndrome, such as diabetes mellitus and high serum cholesterol (see Table 1). They were also confirmed to not have a history of habitual alcohol drinking-in this context, all subjects with a history of excessive alcohol consumption (> 20 g/day) were excluded. Other exclusion criteria for NAFLD patients were a history of use of drugs associated with fatty liver diseases and a history of other hepatic diseases, such as chronic hepatitis C, hepatitis B (serum positive for hepatitis B surface antigen), autoimmune hepatitis, primary biliary cirrhosis (PBC), sclerosing cholangitis, hemochromatosis, alpha-antitrypsin deficiency, Wilson's disease, hepatic injury caused by substance abuse [15‐18]. None of the patients had any clinical evidence of hepatic decompensation, e.g. hepatic encephalopathy, ascites, variceal bleeding, or elevation of the serum bilirubin level to more than twofold the upper limit of normal [15‐17].

Liver specimens were obtained from NAFLD patients for the purpose of diagnosis and staging for NASH, with an 18-gauge needle biopsy apparatus (Pro-Mag, Medical Device Technologies, Gainsville, FL, USA). In our present study, 102 were NASH and 48 were NAFL (non-alcoholic fatty liver). On NASH patients, their ALT/AST levels were much increased and other many abnormal symptoms were observed. So, all of NASH patients were applied the biopsy. Among the 48 NAFL patients, 36 cases showed the abnormal ALT/AST levels, and other abnormal symptoms were observed. Therefore, they were also applied the biopsy. Other 12 NAFL patients, their ALT levels were almost normal range. It was reported that the patients of NAFLD with normal ALT level may also have histological features at risk for disease progression [19‐21]. Therefore, we performed liver biopsy for the purpose of diagnosing and staging for NASH, although the ALT level was in normal range. However, the case of application for the biopsy was strictly defined by the criteria follows; The NAFLD patients were suspected that 1) the existence of severe steatosis (such as persist evidence of severe steatosis at ultrasonography and/or CT) or 2) the fibrosis was progressed (such as platelet count was low and/or serum albmin was low and/or the hyaluronic acid was high and/or the type IV collagen 7s domain was high and/or the liver stiffness was high by transient elastography). In fact, the 12 NAFL patients were fit the cases described above. We excluded unclear borderline cases in this study, because they were not performed the biopsy. Therefore, total 150 NADLF patients (102 NASH + 36 abnormal ALT/AST-NAFL + 12 normal ALT/AST-NAFL) were applied the biopsy.

We did not perform any biopsies on control subjects, because their ALT/AST levels are normal and they have no abnormal symptoms.

The biopsy specimens were obtained from the same lobe to enable obtainment of a sufficient sample size for analysis, and stained with hematoxylin-eosin, reticulin, and Masson trichrome stains. The histopathological findings were evaluated and confirmed as NAFLD/NASH according to the criteria of Matteoni et al. [15]. The criterion for the diagnosis of NAFLD was; macrovesicular steatosis affecting at least 5% of the hepatocytes, with the cases being further classified as steatosis or steatohepatitis. In addition to the presence of steatosis, the minimum criteria for the diagnosis of steatohepatitis included the presence of lobular inflammation and either ballooning of cells or perisinusoidal/pericellular fibrosis in zone 3 of the hepatic acini [15]. The degree of steatosis was assessed as follows, based on the percentage of hepatocytes containing macrovesicular fat droplets: grade 0, no steatosis; grade 1, < 33% hepatocytes containing macrovesicular fat droplets; grade 2, 33%-66% hepatocytes containing macrovesicular fat droplets; grade 3, > 66% hepatocytes containing macrovesicular fat droplets. The severity of the necroinflammatory activity was assessed as follows: grade 1, occasional ballooned hepatocytes (mainly zone 3), scattered intraacinar neutrophils ± lymphocytes, no or mild portal inflammation; grade 2, obvious ballooning degeneration in zone 3, intra-acinar neutrophils, mild to moderate portal and intra-acinar inflammation; grade 3, widespread ballooning, intra-acinar inflammation, accumulation of neurtophils near the ballooned hepatocytes, mild to moderate portal inflammation [15‐18].

Blood samples were obtained from the patients after overnight fasting (12 h) for measurement of the serum levels of AST, ALT, gamma glutamyltranspeptidase (GGT), choline esterase, albumin, triglyceride, high density lipoprotein (HDL) cholesterol, low density lipoprotein (LDL) cholesterol, iron, ferritin, insulin, type IV collagen 7s and hyaluronic acid (HA) concentration, fasting blood glucose, and hematological parameters, including the red blood cell count, hematocrit, and hemoglobin. Serum insulin levels were measured by radioimmunoassay.

Other laboratory biochemical parameters were measured with a conventional automated analyzer. IR was calculated by the modified homeostasis model assessment of insulin resistance (HOMA-IR) using the formula: HOMA-IR = fasting insulin (lU/ml) × plasma glucose (mg/dl)/405 [22].

The specimens were processed for PCR assay using a previously reported method [23]. Briefly, bacterial cells were collected in micro-centrifuge tubes after centrifugation of 500 μl of saliva at 16,000 × g for 5 min. Next, 80 μl of a nuclear lysis buffer (Promega, Madison, WI, USA) was added, and the mixture was incubated at 80°C for 5 min, followed by the addition of 60 μl of protein precipitation solution (Promega). The proteins were removed by centrifugation at 16,000 × g for 3 min, and the DNA was purified by phenol-chloroform extraction and ethanol precipitation. The extracted DNA was then dissolved in 50 μl of TE buffer (10 mM Tris-HCl, 1 mM EDTA [pH 8.0]).

First, we confirmed the appropriate extraction of DNA by PCR using a ubiquitous primer set (5'-AGA GTT TGA TCC TGG CTC AG-3' and 5'-GGC TAC CTT GTT ACG ACT T-3') [23]. Then, P. gingivalis was detected using a PCR method with a P. gingivalis-specific primer set (5'-TGT AGA TGA CTG ATG GTG AAA ACC-3' and 5'-ACG TCA TCC CCA CCT TCC TC-3'), as described previously [24]. P. gingivalis-positive specimens were further analyzed to differentiate among their fimA genotypes using a PCR method with specific primer sets for each fimA type as previously reported [23, 25, 26]. Genomic DNA samples extracted from P. gingivalis strains ATCC33277 (type I), OMZ314 (II), 6/26 (III), HG564 (IV), HNA99 (V), and HG1691 (Ib) were used as positive controls.

Detection of other periodontal bacteria, such as Treponema denticola, Prevotella intermedia, Tannerella forsythia, Aggregatibactor actinomycetemcomitans, Campyrobacter rectus, was also performed according to a previously described method [27].

All animal experiments in the present study conformed to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health, and were approved by the institutional animal care and use committees of Osaka University Graduate School of Dentistry.

Mouse NAFLD model was produced according to the method described previously [28]. Briefly, C57BL/6J mice (6 weeks old, Charles River Japan, Tokyo) allowed free access to water and food throughout experimental period. Mice were randomly divided three groups, High-fat diet control (HFD control); High-fat diet plus P. gingivalis (HFD + P.g); and Basal diet plus P. gingivalis (Basal diet + P.g), respectively. HFD groups were administered High-fat diet 32 (Japan CLEA, Tokyo); this feed contains 506.8 kcal/100 g (57.5% from fat, 19.7% from protein, and 22.8% from carbohydrate). Basal diet group was administered basal diet (ORIENTAL YEAST Co., Ltd., Tokyo); this feed contains 360 kcal/100 g (13.3% from fat, 26.2% from protein, and 60.5% from carbohydrate). Four weeks after the start of HF or basal diet, mice were administered P. gingivalis (OMZ314 strain, 10⁷ cells/body) or vehicle via jugular vein. Because of the high-frequency of type II fimA detection in NAFLD patients, we selected the OMZ314 strain (Type II fimA) for the administration to mice. Twelve weeks after the start of HF or basal diet, body weight, liver weight, and pathological observations were performed.

The patients did not receive any medication for the treatment of impaired liver function, metabolic syndrome or NAFLD during the 3-month periodontal treatment period.

Statistical analyses were performed using the SPSS for Windows, ver. 12. Results are expressed as mean ± SEM. Statistical comparisons were made using the Student's t-test or Scheffe's method after analysis of variance (ANOVA). Differences were considered to be significant at < 0.05. In addition, multiple regression analysis was performed in the NAFLD patients and control subjects with P. gingivalis infection to identify the predictive value of P. gingivalis infection for the development/progression of NFLD, using demographic factors such as the age, history of DM and BMI.

To clarify the relationship between NAFLD and P. gingivalis infection, multiple regression analysis was performed for the NAFLD patients and the control subjects using demographic factors such as the age, history of DM and BMI. The result revealed a statistically significantly higher prevalence of P. gingivalis infection in NAFLD patients as compared with that in the control subjects, even after adjusting for age, history of DM and BMI (Table 2). This result suggests that the presence of P. gingivalis infection might be an independent predictor of the development of NAFLD.

As shown in Table 3, comparative analysis of various parameters was conducted between P. gingivalis-positive and P. gingivalis-negative NAFLD patients. No significant differences in the serum ALT, AST, γGTP, TG, HDL, LDL, Fe, FPG, IRI, HA or type-IV collagen 7S were observed between the two groups of patients. However, a tendency towards decrease of the serum levels of γGTP, TG, cholinesterase, Fe and ferritin was observed in the P. gingivalis-positive patients. In addition, significant decrease of the serum albumin was also observed in this group. Similar results were observed in the P. gingivalis-positive and -negative patients with NASH. On the other hand, a significant difference in the prevalence of DM was observed between the P. gingivalis-positive and -negative groups among patients with NASH, while no such difference was observed between the two groups among NAFLD patients.

P. gingivalis fimbriae, filamentous appendages on the bacterial surface, are classified into 6 genotypes based on the diversity of the fimA genes encoding each fimbrial subunit. It has been demonstrated that bacterial clones with types II, IV or Ib fimA are invasive, whereas those with types I, III or V fimbriae are non-invasive [23, 25, 26]. Interestingly, most of the fimA genotypes detected in the P. gingivalis-positive specimens were of the invasive types; II (50.0%), IV (14.3%), Ib (30.0%); total, 94.3% (Figure 2). These results suggest that invasive-type P. gingivalis may be involved in the progression of NAFLD.

In our HFD-induced mouse NAFLD model, accumulation of lipid in liver and marked increase in body weight are usually observed at 24 weeks, but slight at 12 weeks, after the start of HFD [28]. However, in this study, we compared body weight, liver weight, and pathological findings between P. gingivalis- and vehicle-administered group at 12 weeks after the start of HFD or basal diet. As shown in Figure 3, the mice of HFD control group showed the slight increase in body and liver weight (Figure 3A and 3B, right panels). In contrast, the mice of HFD + P.g group showed the marked increase in body and liver weight. Obvious differences of whole body and liver between HFD control and HFD + P.g group were observed (Figure 3A and 3B, left panels). In contrast, administration of Streptococcus mutans, major caries pathogen but non-periodontitis-related oral bacteria, did not show any increases in body (33.50 + 2.51 g on bacteria vs. 36.33 + 3.21 g on control, P = 0.9325) and liver weight (1.27 + 0.15 g on bacteria vs. 1.38 + 0.13 g on control, P = 0.3811). In addition, other non-virulent oral bacteria, such as Streptococcus sanguinis and Streptococcus salivalius, did not show the increases in body and liver weights (data not shown). These results indicate that the effect of marked increase in body and liver weight might be P. gingivalis-specific phenomena. On the pathological observations, marked accumulation of lipid was observed in the mice infected with P. gingivalis (Figure 3C, middle panel). Furthermore, the marked increases in ALT level and liver triglyceride level were also observed (Figure 3D). These results indicate that the infection of P. gingivalis on HFD condition accelerates the progression of NAFLD.

To clarify the possibility that periodontal treatments in NAFLD patients with periodontitis may improve the condition of NAFLD, we applied the periodontal treatments to randomly selected NAFLD patients. A tendency towards decrease of the serum AST and ALT levels with periodontal treatments was observed by 1 month after the start of the treatment procedures (Figure 4). At the period of 2 months of treatments, significant decreases of both serum AST and ALT levels were observed as compared with the levels at the baseline (P = 0.0165, and P = 0.0031, respectively). Further improvements of the serum AST and ALT levels were observed at the end of 3 months after the start of the treatment procedures (P = 0.0069, and P = 0.0009, respectively). There were no changes in body weights during the treatment period.