Validación de instrumentos de recolección de datos:

We detected intravenously inoculated A. actinomycetemcomitans by QPCR from the liver tissue of 30% of infected mice; the median level was 10 GE/mg liver tissue, with a range from 3 to 1654 GE/mg. The median level was relatively low, which might be due to the non-refined bacterial DNA isolation method. However, the positive detection results support the hypothesis that A. actinomycetemcomitans may disseminate into tissues via the circulation. The paraffin-embedded liver tissue sections were stained with hematoxylin/eosin and the morphology was evaluated and graded according to the observed level of infiltration of inflammatory cells and microvesicular appearance. The effect of A. actinomycetemcomitans infection on liver morphology was clearly proinflammatory. When compared to the controls, we observed an enhanced infiltration of inflammatory cells in both the Aa group (p = 0.002) and the combined infection model (Cpn+Aa group, p = 0.023). The phenomenon was particularly pronounced in the recurrent

A. actinomycetemcomitans infection model (Figure 4).

Table 5 presents a summary of the studied parameters. The QPCR analyses on the relative expression of inflammation-related genes supported the observed increase in the inflammation status. The relative expression levels of Il-1β (p <0.001), Mcp-1 (p = 0.008), and CD68 (p = 0.002) were higher in the Aa group, and the levels of Il-1β (p <0.001) in the Cpn+Aa group, than in control mice. The protein expression of Il-1β and CD68 was defined from immunohistochemical staining of liver paraffin sections.

53 Figure 4. Infiltration of inflammatory cells into liver tissue induced by recurrent A.

actinomycetemcomitans infection. The arrows point to examples of the sites affected.

The hepatic lipid concentrations were determined from chloroform- methanol/methanol extracts of the frozen tissue samples. The phospholipid levels of liver tissue were significantly decreased among all infected mice. The reduction was especially substantial in the Aa group (mean ± standard error of mean [SEM]; 19.5 ± 2.2 mmol/mg tissue, p <0.001) when compared to control mice (88.9 ± 4.3 mmol/mg tissue). This may result from endotoxin activation of phospholipases 1A and 2A, and cause marked changes in membrane structure and function (Liu et al. 1988). Apart from the phospholipid reduction, the other measured parameters involved in lipid homeostasis were scarcely altered (Table 5). There were no significant changes in hepatic cholesterol and TG levels; however, the TG/cholesterol ratio was increased 4-fold in the Aa group (p <0.001). Lipid and lipoprotein metabolism-related gene expression was similar to that of control mice. Furthermore, the hepatic proportions of SFAs, monounsaturated fatty acids (MUFAs), and polyunsaturated fatty acids (PUFAs) remained unaltered. Moreover, the infection did not modify the serum lipid and lipoprotein metabolism-related factors. However, the TG levels were decreased (p = 0.041) in the Aa group.

The serum TNF-α concentration (p = 0.027) and LPS activity were higher (p = 0.014) in the Aa group compared to the controls. Additionally, the level of serum immunoglobulin (Ig) G class antibodies against A. actinomycetemcomitans was significantly increased among all Aa-infected mice, and exhibited a positive correlation with the degree of hepatic inflammation. The observed changes in serum parameters were related to the augmented systemic inflammation reaction and resembled the results from other A. actinomycetemcomitans mouse studies (Tuomainen et al. 2008; Zhang et al. 2010). The detection of A. actinomycetemcomitans from the liver tissue, combined with the infiltration and distribution of PBMC and neutrophils, up-regulated expression of

54

inflammation-related genes, and decreased levels of hepatic phospholipids, also indicated local infection-derived inflammation in the liver. It is reasonable to hypothesize that prolonged inflammation induces the production of ROS and proinflammatory cytokines, and may lead to ER stress, an unfolded protein response and hepatic injury. The administration of proteases and E. coli LPS in the gingival sulcus have been shown to induce hepatic inflammation and steatosis in rats (Tomofuji et al. 2007). The undetected hepatic steatosis in our study may be explained by the differences in bacterial stimuli and dosage. The recent A. actinomycetemcomitans mouse models demonstrating increased atherosclerotic lesions and an enhanced oxidation potential have used high bacterial doses and inoculations three times per week (Zhang et al. 2010; Jia et al. 2013). Our A.

actinomycetemcomitans infection model more closely resembles repeated recurrent

infections than an aggressive infection with continuous bacterial distribution.

Contrary to the pathogen burden hypothesis (Zhu et al. 2000; Georges et al. 2003), the combined infection model (Cpn+Aa group) did not result in cumulative or augmented infection-derived alterations. The systemic inflammation reaction and responses in the liver were milder than in the Aa group, and it seems that the C. pneumoniae infection may have primed the immune system.

55 Table 5. Studied parameters among A. actinomycetemcomitans-infected mice

Parameter

Changes when compared to the control group, p <0.05 Aa group,

n = 10

Cpn+Aa group, n = 10

Liver

Infiltration of inflammatory cells ↑ ↑

Il-1β (mRNA FC) ↑ ↑ Mcp-1 (mRNA FC) ↑ - CD68 (mRNA FC) ↑ - Srebp-1c (mRNA FC) - - Fasn (mRNA FC) - - Acc2 (mRNA FC) - - LDLr (mRNA FC) - -

Phospholipid (mmol/mg tissue) ↓ ↓ Cholesterol (mmol/mg tissue) - -

TG (mmol/mg tissue) - -

TG/cholesterol ↑ -

SFA (%/total FAs) - -

MUFA (%/total FAs) - -

PUFA (%/total FAs) - -

Serum TG (µmol/ml) ↓ - Cholesterol(µmol/ml) - - FFA (mg/dl) - - apoA-I (mg/ml) - - PLTP activity (µmol/ml/h) - - Adiponectin (ng/ml) - - LPS activity (EU/ml) ↑ - LBP (µg/ml) - - SAA (µg/ml) - - TNF-α (pg/ml) ↑ - IgG against A. actinomycetemcomitans (AU) ↑ ↑

↑, increased/up-regulated; ↓, decreased/down-regulated; -, no change; Aa group, recurrent A. actinomycetemcomitans infection; Acc2, acetyl-CoA carboxylase 2; AT, Cpn+Aa group, combined recurrent A. actinomycetemcomitans and chronic C. pneumoniae infections; FA, fatty acid; Fasn, fatty acid synthase; FC, fold change; FFA, free fatty acid; Il-1β; interleukin 1β; LBP, LPS binding protein; LDLr, LDL receptor; LPS, lipopolysaccharide; Mcp-1, monocyte chemoattractant protein 1; MUFA, monounsaturated fatty acid; PLTP, phospholipid transfer protein; PUFA, polyunsaturated fatty acid; SAA, serum amyloid A; SFA, saturated fatty acid; Srebp-1c, sterol responsive element binding protein 1c; TG, triglyceride; TNF-α, tumor necrosis factor α.

56

4.1.2. A. actinomycetemcomitans-derived changes in transcriptomes and fatty acid