**3. Alzheimer's Disease**

AD is currently the best-known neuropsychiatric disease that is associated with periodontitis based on clinical and experimental evidence that is accumulating rapidly. Recent epidemiological studies have pointed out that periodontitis significantly elevates the risk for AD. A prospective pilot study using qPCR identified *P. gingivalis* DNA in the CSF in seven of the 10 clinically diagnosed AD patients who had mild to moderate cognitive impairment [24]. A cross-sectional study reported that plasma TNFα and antibodies against periodontal bacteria were elevated in AD patients relative to normal controls and were independently associated with AD [32]. Another cross-sectional study demonstrated that the increased serum levels of TNFα and IL-6 in patients with AD were significantly associated with periodontitis [33]. A case-control study established a significant association between AD and the increased number of deep periodontal pockets [34]. A retrospective matched cohort study on 9291 patients with periodontitis showed that chronic periodontitis exposure for 10 years was associated with a 1.707-fold increase in the risk of developing AD [35]. Another historical cohort study with the larger sample size of 262,349 participants who su ffered from chronic periodontitis supports this finding [36]. Prospective cohort studies with relatively small sample sizes demonstrated that serum IgG antibody levels to periodontitis bacteria, such as *P. gingivalis*, *Tannerella forsythia* and *Treponema denticola* (the so-called "red complex"), were significantly increased in baseline serum drawn from subjects who were diagnosed with AD in later years compared to controls [37,38]. Even after the onset of AD, periodontitis may exacerbate cognitive impairment. A six-month observational cohort study tested cognitive function and serum pro-inflammatory markers in 52 patients with mild to moderate AD. The study showed that the presence of periodontitis at baseline was associated with a six-fold increase in the rate of cognitive decline in participants over the six-month follow-up period, and was also associated with a relative increase in the pro-inflammatory state over that period [39]. A meta-analysis based on one cross-sectional study [40] and two case-control studies [41,42], which have not been previously mentioned, and assessed as at a low risk of bias, came to the conclusion that periodontitis is significantly associated with AD (OR 1.69, 95% CI 1.21–2.35) [43]. Furthermore, severe forms of periodontitis show a more intense association with AD (OR 2.98, 95% CI 1.58–5.62) [43]. An interesting study comparing periodontal conditions between countries reported that periodontitis was more prevalent in Germany and that elderly German subjects had significantly more severe periodontal conditions and fewer remaining teeth compared to those in Japan, even after adjustment of the comprehensive risk factor [44]. Accordingly, it would be tempting to examine whether the AD prevalence in Germany is significantly higher than that of Japan by a direct comparison with the adjustment of confounding factors.

Growing evidence based on animal studies also strengthens a possible causal link of periodontitis to AD. Chronic intraperitoneal injection of *P. gingivalis* LPS (1 mg/kg/day, daily for 5 weeks) has been demonstrated to cause learning and memory deficits accompanied with intracellular amyloid β (Aβ) accumulation in neurons and microglial activation (i.e., increased expression of IL-1β and TLR2 restricted to microglia) in the hippocampus in middle-aged mice (12 months old), but not in young mice (2 months old) or in middle-aged cathepsin-B knockout mice [29]. Therefore, cathepsin B, known as an amyloid precursor protein (APP) secretase, may play a critical role in the periodontitis-exacerbated AD and could be a therapeutic target. Even a single intraperitoneal injection of *P. gingivalis* LPS (5 mg/kg) into 8-week old mice has been shown to impair spatial learning and memory with neuroinflammation (i.e., microglial activation, astrocytic activation, and increased expression of TNFα, IL-1β, and IL-6 in the cortex and/or hippocampus) and activation of the TLR4/nuclear factor-kappa B (NF-κB) signaling pathway (i.e., up-regulation of TLR4, CD14, IL-1 receptor-associated kinase 1, and phospho-p65 in the cortex) [45]. These behavioral and immuno-biochemical findings were considerably abolished by the TLR4 inhibitor TAK242, suggesting that the *P. gingivalis* LPS-induced cognitive dysfunction and neuroinflammation are mediated by the TLR4/NF-κB signaling pathway [45]. Interestingly, this study also tested *E. coli* LPS and reported that either *P. gingivalis* LPS or *E. coli* LPS caused both cognitive impairment and neuroinflammation and there was no significant di fference between the e ffects of the two LPS species [45]. Experimental chronic periodontitis, caused when live *P. gingivalis* (ATCC33277) was given by oral gavage every 48 h over 6 weeks, has been shown to impair learning and memory and elicit neuroinflammation (i.e., increased expression of TNFα, IL-1β and IL-6 in the brain) in middle-aged mice (12 months old), although not in young individuals (4 weeks old) [46]. Experimental chronic periodontitis, induced by repeated oral application of another live *P. gingivalis* W83 every 48 h over 22 weeks, has been demonstrated to increase extracellular Aβ42 amyloid plaques, ser396 residue of tau protein phosphorylation, neurofibrillary tangle formation, and neuroinflammation (i.e., microglial activation, astrocytic activation, and increased expression of TNFα, IL-1β and IL-6) in the hippocampus

of 6-week old mice [47]. Currently, there is only one study that employed APP-transgenic (Tg) mice and the study implies that periodontitis exacerbates the hallmark pathology and symptoms of AD [48]. Specifically, APP-Tg mice with periodontitis induced by oral infection with *P. gingivalis* ATCC33277, showed greater deposition of Aβ40 and Aβ42 amyloid plaques in both the hippocampus and cortex and increased brain expression of TNFα and IL-1β, compared with sham-infected APP-Tg mice. Furthermore, cognitive function was significantly impaired in the periodontitis-induced APP-Tg mice relative to sham-infected APP-Tg mice [48].

Postmortem studies have indicated the potentially causal presence of periodontopathic virulence products in AD brains. LPS from *P. gingivalis* was detected in the brain of four out of 10 AD cases by immunofluorescence and Western blot (WB) analyses, whereas *P. gingivalis* LPS was not detected in 10 age-matched non-AD controls. Dominiy et al. (2019) have performed a seminal study, identifying *P. gingivalis* DNA and gingipains, toxic proteases secreted from *P. gingivalis* in AD brains. Immunohistochemical analyses using tissue microarrays showed that gingipain immunoreactivity in AD brains was significantly greater than that in sex- and age-matched non-AD brains, and that gingipain immunoreactivity significantly correlates with tau and ubiquitin loads and AD diagnosis [24]. Using qPCR, the authors identified *P. gingivalis* DNA in the AD brains which were lysine gingipain-positive in WB and immunoprecipitated analyses [24]. In addition to postmortem brain studies, they carried out in vivo studies using wild-type mice and gingipain-knockout mice that were orally infected with *P. gingivalis* W83 every other day for 42 days. Colonization of *P. gingivalis* and Aβ42 levels were increased in the brains of the infected wild-type mice, while the colonization and Aβ42 levels were decreased in the brains of either the infected wild-type mice treated with the gingipain inhibitor COR119 or of the gingipain-knockout mice [24]. Because these findings sugges<sup>t</sup> that gingipain inhibition reduces the *P. gingivali* load and Aβ42 production in the brain, gingipain inhibitors could have therapeutic potential for patients with both AD and periodontitis.
