**1. Introduction**

Periodontitis is a chronic, oral multi-bacterial infection affecting nearly 50% of the population worldwide and is the most prevalent inflammatory disease in adults [1,2]. Periodontitis is not only an oral localized inflammatory disease, but also elicits low-grade systemic inflammation via both the release of pro-inflammatory cytokines and the invasion of periodontitis bacteria (e.g., *Porphyromonas gingivalis* (*P. gingivalis*)) along with their components (e.g., lipopolysaccharide (LPS) and flagellin) into systemic circulation [3]. Periodontitis thus causes or hastens other chronic systemic inflammatory diseases, including atherosclerosis, cardiovascular diseases, diabetes, and rheumatoid arthritis [4]. In particular, the causal link between periodontitis and infective endocarditis has been known for many decades. Besides inducing systemic inflammation, increasing evidence implies that periodontitis provokes chronic inflammation associated with activation of microglia, the immune cells in the brain, which is referred to as neuroinflammation (reviewed in [5–7]).

The concept of neuroinflammation was originally proposed for neurodegenerative disorders in the 1980s, based on two historical discoveries. The first was the immunohistochemical identification of activated microglia in association with the lesions in Alzheimer's disease (AD) brains [8]. The second was the epidemiologic finding that rheumatoid arthritics, who regularly consume anti-inflammatory agents, were relatively spared from AD [9]. In the ensuing years, activated microglia have also been found in the lesions of Parkinson's disease (PD), multiple sclerosis, and amyotrophic lateral sclerosis [10]. Neuroinflammation has thus become considered as a common prominent feature among a variety of neurodegenerative disorders. Attention on the pathogenetic role of neuroinflammation has, over the past two decades, been expanded to psychiatric disorders. Immunohistochemistry and

positron emission tomography studies have revealed microglial activation in the brain of patients with schizophrenia [11–13] and major depression (MD) [14,15], the representative endogenous psychoses. Neuroinflammation could thus be important in many pathological conditions of the brain, including both neurodegenerative disorders and psychiatric disorders (hereinafter referred to as neuropsychiatric disorders in this article).

Based on the aforementioned findings, chronic inflammation can be regarded as a common denominator of periodontitis and neuropsychiatric disorders. Specifically, neuroinflammation may causally link periodontitis to the clinical onset and development of neuropsychiatric disorders. Furthermore, through the biological mechanism of chronic inflammation, periodontitis could causally affect neuropsychiatric disorders, especially MD, because of its psychosocial e ffects, such as shame, loneliness, impaired quality of life (QOL), and impaired social status [16]. Nonetheless, this review article focuses on the biological and epidemiological evidence for possible causal links of periodontitis to the selected neuropsychiatric disorders, namely AD, MD, PD, and schizophrenia. This article also discusses an association between periodontitis and the neurological event of ischemic stroke.

### **2. How Does Periodontitis Cause Neuroinflammation?**

Neuroinflammation is a key pathogenetic connector between periodontitis and neuropsychiatric disorders. The biological mechanisms by which periodontitis causes neuroinflammation can be presumed to consist of three possibilities, as follows (Figure 1).

**Figure 1.** Scheme for presumed mechanisms by which periodontitis causes neuroinflammation. These consist of three possibilities as follows: (**1**) Peripheral pro-inflammatory cytokines associated with periodontitis communicate with the brain via the neural pathway, humoral pathway, and cellular pathway. (**2**) Periodontal bacteria/bacterial molecules can directly invade the brain either through the blood stream or via cranial nerves. (**3**) Communication between periodontal bacteria/bacterial molecules and brain-resident microglia could occur through the leptomeninges.

(**1**) Neuroinflammation can be caused by peripheral pro-inflammatory cytokines generated in systemic inflammation induced by periodontitis without the contact of periodontal bacteria/bacterial molecules with the brain tissue via three pathways, i.e., the neural pathway, the humoral pathway, and the cellular pathway. Through the neural pathway, systemic cytokines directly activate primary afferent nerves, such as the vagus nerve. The signal reaches the primary and secondary projection of the neural pathway, reaching first the nucleus tractus solitaries and subsequently, various hypothalamic

brain nuclei [17]. It has been shown that subdiaphragmatic vagotomy blocks the LPS-induced sickness behavior in rats [18], while it does not a ffect the LPS-induced synthesis of pro-inflammatory cytokines at the periphery. The humoral pathway involves the choroid plexus and circumventricular organs, which lack an intact blood–brain barrier (BBB). These leaky regions may be access points for circulating pro-inflammatory cytokines to enter into the cerebral parenchyma by volume di ffusion and elicit downstream signaling events, which are important in altering brain function [19]. The cellular pathway implicates systemic inflammation in association with both activation of endothelial cells (CECs) and an increase in circulating monocytes [19]. Systemic pro-inflammatory cytokines activate CECs, expressing receptors for TNFα and IL-1β, which in turn, signal to the perivascular macrophages located immediately adjacent to CECs [20]. These perivascular macrophages subsequently communicate with microglia and thus lead to microglial activation. Activated microglia secrete not only pro-inflammatory cytokines but also proteases and chemokines, including monocyte chemoattractant protein (MCP)-1. MCP-1 is supposed to be responsible for the recruitment of monocytes into the motor cortex, hippocampus, and basal ganglia regions, areas of the brain known to be involved in the control of behavior [21].

(**2**) Periodontal bacteria/bacterial molecules can directly invade the brain either through the blood stream or via cranial nerves. In periodontitis, a periodontal pocket is filled with periodontal bacteria/bacterial molecules that form biofilms. Since periodontal bacteria are capable of invading an intact pocket epithelium, periodontal bacteria/bacterial molecules can gain access to the circulation [22]. It has been shown that LPS deteriorates the BBB and increases its permeability through abnormal activation of matrix metalloproteinase [23]. Circulating periodontal bacteria/bacterial molecules could then penetrate into the brain through the compromised BBB. In fact, *P. gingivalis* DNA has been identified by quantitative polymerase chain reaction (qPCR) in the brain of mice orally infected with *P. gingivalis* [24], and *P. gingivalis*-derived LPS has been detected in the brains of AD patients [25]. Circulating periodontal bacteria/bacterial molecules can also enter the brain through the circumventricular organs and choroid plexuses that lack the contiguous BBB. The cranial nerve may be another entry route for periodontal pathogens into the brain. The olfactory and trigeminal nerves are known to be used by periodontal bacteria to bypass the BBB [26]. The detection of oral *Treponemas* in the trigeminal ganglia supports the idea of neural pathways [27]. Via any of these pathways, infiltration of periodontal bacteria/bacterial molecules into the brain could result in inflammatory activation of microglia, since it has been demonstrated that either *P. gingivalis* infection or LPS of *P. gingivalis* activates microglia in vitro to produce pro-inflammatory cytokines, such as tumor necrosis factor (TNF)- α, interleukin (IL)-1β and IL-6 [28,29].

(**3**) The leptomeninges could be a site of communication between periodontal bacteria and brain-resident microglia. The leptomeninges covers the brain parenchyma surface and provides a physical boundary at the cerebrospinal fluid (CSF)-blood barrier. Leptomeningeal cells express Toll-like receptors (TLRs) 2 and 4 that are the receptors for *P. gingivalis* LPS. Leptomeningeal cells can be activated by circulating *P. gingivalis* LPS and subsequently produce pro-inflammatory cytokines for the brain [30,31]. The pro-inflammatory cytokines released from leptomeningeal cells activate microglia to evoke neuroinflammation. Accordingly, the leptomeninges could be harmful by transducing peripheral inflammation, including periodontitis, into neuroinflammation.
