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

Abstract

Background: Non-alcoholic fatty liver disease (NAFLD) is a hepatic manifestation of metabolic syndrome that is closely associated with multiple factors such as obesity, hyperlipidemia and type 2 diabetes mellitus. However, other risk factors for the development of NAFLD are unclear. With the association between periodontal disease and the development of systemic diseases receiving increasing attention recently, we conducted this study to investigate the relationship between NAFLD and infection with Porphyromonas gingivalis (P. gingivalis), a major causative agent of periodontitis.

Methods: The detection frequencies of periodontal bacteria in oral samples collected from 150 biopsy-proven NAFLD patients (102 with non-alcoholic steatohepatitis (NASH) and 48 with non-alcoholic fatty liver (NAFL) patients) and 60 non-NAFLD control subjects were determined. Detection of P. gingivalis and other periodontopathic bacteria were detected by PCR assay. In addition, effect of P. gingivalis-infection on mouse NAFLD model was investigated. To clarify the exact contribution of P. gingivalis-induced periodontitis, non-surgical periodontal treatments were also undertaken for 3 months in 10 NAFLD patients with periodontitis.

Results: The detection frequency of P. gingivalis in NAFLD patients was significantly higher than that in the non-NAFLD control subjects (46.7% vs. 21.7%, odds ratio: 3.16). In addition, the detection frequency of P. gingivalis in NASH patients was markedly higher than that in the non-NAFLD subjects (52.0%, odds ratio: 3.91). Most of the P. gingivalis fimbria detected in the NAFLD patients was of invasive genotypes, especially type II (50.0%). Infection of type II P. gingivalis on NAFLD model of mice accelerated the NAFLD progression. The non-surgical periodontal treatments on NAFLD patients carried out for 3 months ameliorated the liver function parameters, such as the serum levels of AST and ALT.

Conclusions: Infection with high-virulence P. gingivalis might be an additional risk factor for the development/progression of NAFLD/NASH.

Figure 1

Detection frequency of P. gingivalis in NAFLD/NASH patients and control (non-NAFLD) subjects. A: Comparison of the detection frequency of P. gingivalis between NAFLD patients and non-NAFLD (control) subjects. B: Comparison of the detection frequencies of P. gingivalis among NASH patients, non-alcoholic fatty liver (NAFL) patients and in non-NAFLD (control) subjects. OR: odds ratio, CI: confidence interval.

Figure 2

Percentage of detection frequency of various fimA types on NAFLD patients. 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%. Type III (5.7%) is non-invasive type.

Figure 3

Effect of administration of P. gingivalis on mouse NAFLD model. A: Typical pictures of whole body (left panel) and body weight (right panel) of mice administered vehicle (HFD control) or P. gingivalis (HFD + P.g) under the high fat diet (HFD) condition. Each column represents mean + SEM from 12 to 13 independent animals. **; P < 0.01. B: Typical pictures of liver (left panel) and liver weight (right panel) of mice administered vehicle (HFD control) or P. gingivalis (HFD + P.g) under the high fat diet (HFD) condition. Each column represents mean + SEM from 12 to 13 independent animals. **; P < 0.01. C: Typical pathological pictures of liver of mice administered vehicle (HFD control) or P. gingivalis (HFD + P.g) under the high fat diet (HFD) condition, or P. gingivalis under basal diet condition (Basal diet + P.g). Each column represents mean + SEM from 12 to 13 independent animals. **; P < 0.01. D: Alterations of ALT/AST levels and liver triglyceride level. Each column represents mean + SEM from 6 to 9 independent animals. *; P < 0.05 and **; P < 0.01, respectively.

Figure 4

Improvement of the serum AST and ALT levels with periodontal treatments in NAFLD patients with periodontitis. Box plots represent the inter-quartile range (boxes), median (central horizontal lines), range (thin lines) and outliers (circles) of the serum AST or ALT levels from 10 cases. Stepwise decreases of the serum AST (P < 0.0001, Kruskal-Wallis test) and ALT (P < 0.0001, Kruskal-Wallis test) were observed during the course of the periodontal treatments.