9.1 Introduction

Many neurodegenerative diseases cause progressive deterioration in specific areas of the nervous system. This affects both the structure and function of neural tissues. Two of the most common neurodegenerative diseases are Alzheimer disease (AD) and Parkinson disease (PD). AD initially affects cognitive function, whereas PD commonly begins by affecting motor function, although later in the PD trajectory cognitive impairment often ensues. Both conditions are chronic, debilitating and progressively limit the individual’s ability to function independently. As the likelihood of developing neurodegenerative diseases increases with age, the global burden of both AD and PD is set to rise steeply. Between 1980 and 2015, average age at death increased from 73.1 to 81.6 years in the UK and 73.7 to 78.7 years in the USA, a trend seen in most countries with an inverted population age pyramid1. The 2016 worldwide estimates of those living with AD or PD are 44 million and 6 million, respectively2. In the UK alone, the number of people with PD is approximately 16,000. For AD, this is 500,000, representing 62% of those with dementia. In 95% of cases the age of diagnosis is over 65 years; prevalence of dementia rises from 7% over the age of 65, to 13% over the age of 75, and approximately 60% of those affected overall are female3.

The annual cost of dementia in the UK is approximately £26.3 billion3, and the annual financial impact of living with PD in the UK has been estimated to be over £16,500 for each individual household affected, costing the UK economy between £450 million and £3 billion per year4. There is also a high social cost, as these diseases often affect the quality of life of carers and loved ones as well as the individuals affected. Periodontitis has for some time been associated with neurodegenerative diseases, largely because of the obvious related decline in ability to provide self-care for the teeth and surrounding tissues. However, it is now being recognised that the relationship might be bi-directional5. There is a growing body of evidence suggesting a potentially proactive role for periodontal pathogens and the periodontal chronic low grade inflammatory response in the development of some types of neurodegenerative diseases. Like many chronic inflammatory diseases, periodontitis is similarly age-related, with onset commonly occurring in mid-life.

Currently there is no cure for AD or PD; treatment of neurodegenerative conditions is limited to, at best, delaying the worsening symptoms. This is in some ways similar to the goal of conventional periodontal treatment, which is to slow the progression of periodontal destruction, ideally reaching a state of periodontal stability. The key to developing improved treatment modalities is likely to be found in discoveries relating to the pathogenesis of these conditions. Whilst the causes of the neuropathology of AD and PD remain somewhat elusive, they are likely to be multifactorial. It has been known for some time that peripheral infection is a strong risk factor for the development and progression of AD in elderly populations. Peripheral infection is a common cause of delirium in the elderly, and delirium has been shown to increase the risk of developing dementia eight-fold over the following 10 years6. Cognitive decline has also been shown to occur at a more rapid rate following a period of delirium in those already diagnosed with AD7. With increased age, and in AD and PD, the blood–brain barrier (BBB) can have increased permeability8^,9. However, it is likely that the initiating factors associated with AD or PD occur in mid-life and may be low level, chronic infections. It has recently been suggested that periodontal pathogens and/or the peripheral inflammatory state seen in periodontitis could be implicated in neurodegeneration10^,11. These pathogens may slowly increase the permeability of the BBB and produce pro-inflammatory proteins resulting in peripheral and neural inflammation, ultimately leading to neurodegeneration. Thus both AD and PD may be driven by neuroinflammation; this chapter seeks to examine its origin and to discuss its association with periodontitis and its possible role in these two neurodegenerative diseases.

9.2 Cellular and molecular mechanisms

9.2.1 Alzheimer disease (AD)

AD is the most common form of dementia; indeed, 60% to 80% of those with dementia have AD12. The likelihood of developing AD increases with age, although it is important to note that the features of AD are not simply features of the aging process. AD is a neurodegenerative condition that progresses inexorably until death, which tends to occur 3 to 12 years after diagnosis. One quarter of hospital beds in the UK are occupied by people who are over 65 years old and living with dementia13. Overall, the cost to the UK economy including loss of earnings and other factors is estimated to be over £26 billion per year3.

AD has been defined by Nelson et al14 as the presence of three factors:

● clinical dementia (cognitive impairment with a memory component that impacts, to a greater or lesser degree, on daily living skills)

● neurofibrillary tangles within the neurons of the neocortex

● amyloid plaques with amyloid beta-containing cores inside swollen and degenerating neuronal cell processes.

The clinical symptoms of progressive memory loss and cognitive impairment are the result of the toxic action of the misfolded proteins hyperphosphorylated tau (ptau) and amyloid β (Aβ) in the brain, which progressively sequester their native counterparts, leading to a breakdown in neuronal structure and function. This is combined with their activation of innate immune function, and feeds a cycle of inflammatory processes. The underlying reasons for the initiation of these events are gradually being uncovered.

9.2.1.1 Symptoms of AD

AD was named after Alois Alzheimer, a German psychiatrist and pathologist working in Frankfurt, who provided the published description of a 51-year-old woman who had previously been in his care. Alzheimer described her symptoms of dementia, including profound confusion, memory loss, auditory hallucinations, aggression and paranoia. Five years later he carried out a neuropathological examination, and described the pathology in a paper in 190715. Symptoms begin with mild forgetfulness, sometimes called mild cognitive impairment, and lead to other neuropsychological deficits such as more profound memory deficit, reduced cognition, often with visuospatial disturbances, and impaired decision making. The symptoms can manifest as mis-orientation in time and space, aphasia (difficulties in speaking or understanding communication), apraxia (an inability to carry out tasks despite intact sensory and motor systems) and agnosia (loss of the ability to recognise what objects are and their use). Additionally, there is often anosmia (loss of the sense of smell), which is often one of the first symptoms and can be predictive of conversion from mild cognitive impairment to AD16. All of these symptoms relate to the specific areas of the brain that are gradually being affected by the progression of the disease. Neuropsychiatric (behavioural) changes will also be manifested, often early in the disease process. Depression commonly affects two thirds of patients at some point, paranoia affects about a third, and aggression is seen in about a fifth of patients. Wandering is also fairly common; hallucinations are mainly auditory, occurring less frequently. Personality changes can include apathy, disinhibition and irritability. AD is a slowly progressive condition, which over the decade following diagnosis results in ever increasing severity of symptoms. This takes away the individual’s sense of self and gradually becomes more and more debilitating, ultimately causing death. The progression of symptoms correlated with the neuropathology is described in Table 9-1.

Table 9-1 The stages of AD. Symptomatic changes are taken from Reisberg et al17, neuropathological data is from recent PET scan data18. There is a general correlation between AD pathology and clinical symptoms, although this is subject to individual variability. This can be complicated by overlapping interactions between the pathology of AD and other dementias. The progression of AD is predictable even though the rate and order of clinical symptoms varies from person to person. (For further information regarding NFT and Aβ see Section 9.2.1.1 ‘Symptoms of AD’ and Section 9.2.1.4 ‘Neuropathology of AD’)

Aβ = amyloid β; AD = Alzheimer disease; NFT = neurofibrillary tangle; ptau = hyperphosphorylated tau.

9.2.1.2 Diagnosis of AD

A definitive diagnosis of AD is usually only possible after death, as it requires correlation of the clinical symptoms to dementia during life with histopathological changes in the brain. More recently, it has been made possible to visualise the presence of Aβ deposits and neurofibrillary tangles (NFTs) by PET (positron emission tomography) scans, but this is not yet generally available for diagnosis. The neuropathological changes include the formation of NFTs in neurons and extracellular amyloid plaques in the brain parenchyma. The former comprises a hyperphosphorylated form of the protein tau (ptau), which accumulates within neurons and spreads from the brainstem to the cortex. Additionally there are amyloid plaques, comprised of extracellular aggregations of a protein called Aβ. The definitive diagnosis post-mortem requires presence of these plaques and NFT, in addition to a diagnosis of dementia in life. The clinical diagnosis of AD is derived from the combination of information from the patient’s medical history, and clinical, neurological and cognitive assessments. Blood tests (e.g. vitamin B12 and levels of circulating thyroid hormones) and neuroimaging (e.g. computed tomography and magnetic resonance imaging) play an important role in determining alternative causes of dementia that need to be excluded, for instance brain tumours, cerebral infarcts and other causes of dementia, such as vascular or Lewy body disease, prior to reaching a diagnosis.

The diagnosis of AD, particularly in the earlier stages of the disease process, can be complex and is therefore best made by a specialist in the field. In 2011 the clinical diagnostic guidance for AD was updated by the US National Institute on Aging-Alzheimer’s Association Workgroups19. Dementia caused by AD is classified as either probable or possible depending on patient symptoms, clinical findings, results of special investigations and disease progression. Further criteria are used for the diagnosis of AD for those participating in clinical research studies where additional tests and imaging techniques are available. Clinical diagnostic accuracy for AD using current methods is limited, being estimated in 2001 to have at best 80% sensitivity and 70% specificity20. This is further complicated, as many of those with AD have co-existent pathologies that can affect their symptoms and progression of the condition. A significant amount of research has gone into investigating potential biomarkers that may, in the future, prove more accurate than the combined clinical assessment methods currently used. Since 2012, numerous Cochrane reviews have evaluated the potential role of biomarkers and imaging techniques for the diagnosis of AD, although to date no single biomarker has been found capable of detecting the presence or absence of AD in a reliable and accurate manner. The majority of potential biomarkers have better sensitivity than specificity and thus are better suited to excluding AD than determining its presence21.

9.2.1.3 Treatment of AD

Treatment for AD is currently symptomatic. NICE (National Institute for Health and Care Excellence, UK) approve two types of symptomatic treatments that are aimed at acetylcholine- or glutamate-producing neurons. Early in the AD disease process there is a loss of the neurotransmitter acetylcholine due to the dysfunction of cholinergic neurons based in the basal forebrain. These neurons project to the hippocampus and are important in memory formation. Acetylcholinesterase inhibitors (CEIs) (donepezil, galantamine and rivastigmine) are commonly used as treatments for mild as well as moderate AD. These CEIs potentiate the signal produced, by preventing the breakdown of acetylcholine at the synapse. NICE has also approved the drug memantine for use in the symptomatic management of moderate or severe AD, and for those who cannot take CEIs. Glutamate is an excitatory neurotransmitter that binds to the N-methyl-D-aspartate (NMDA) receptor. Over-excitation of this receptor causes an increase in intracellular calcium ions, which leads to excitotoxicity. The drug memantine binds to the NMDA receptor and inhibits the prolonged influx of calcium ions. Depression may result from coming to terms with the diagnosis and symptoms of AD, or may actually be a part of the disease process related to the slowing down of cognitive function. AD-associated depression can be treated with antidepressants in the same manner as regular depression. If there is a problem with sleeping, clonazepam, a benzodiazepine tranquiliser, may be prescribed. None of these medications slow the progression of AD symptoms. Notably, the realisation that prevention of AD by means of a healthy lifestyle, nutrition and exercise is currently the best option, may well be the turning point in slowing the escalating prevalence of this disease.

9.2.1.4 Neuropathology of AD

In AD there is a general increase in the amount of Aβ peptide (major forms are 40 and 42 amino acids in length, usually produced in a 90:10 ratio), which is cleaved from the amyloid precursor protein (APP). Due to its propensity to fibrillise, Aβ₄₂ rapidly oligomerises and aggregates in the parenchyma to form extracellular amyloid plaques, whereas the more soluble Aβ₄₀ form is cleared into blood vessels. This may, however, eventually result in amyloid angiopathy, where smooth muscle cells in the walls of cerebral arterioles are replaced by amyloid. It has been proposed that an imbalance between the formation and clearance of Aβ results in its excessive accumulation in the brain, and that this is an early and critical event in the pathogenesis of AD. This concept, later named the ‘amyloid cascade hypothesis’, was first postulated over 25 years ago22. These changes, together with altered and reduced synaptic neuronal communication, eventually lead to neuronal and cortical atrophy14. It is now understood that changes in brain microanatomy and cellular and neuronal signalling pathways can be seen 20 years prior to the diagnosis of AD. Furthermore, NFTs and Aβ are not specific for AD, and can be seen, although less abundantly, in the brains of many elderly people as a response to injury or other stimuli; however, the presence of NFTs in the neocortex is pathognomonic for AD. Similarly, amyloid plaques that are surrounded by degenerating neuronal processes often containing ptau aggregates are a hallmark of AD, unlike the diffuse amyloid plaques seen in the brains of many elderly people.

Aβ is produced in all brains and is believed to have a protective role in healthy young brains that enhances learning and memory development. Aβ has a number of physiological roles within the healthy brain, playing an important role in the innate immune activation of phagocytic cells, such as microglia and macrophages within the central nervous system (CNS)23. Aβ may function as an antimicrobial peptide in the brain24 and has been shown capable of inhibiting fungal growth and viral replication in vitro and in animal models25^,26. It is the fibrillisation of Aβ, with subsequent effects, rather than simply its presence that is pathogenic27. Under healthy physiological conditions, Aβ is cleared by three mechanisms: phagocytosis by cells, enzymatic degradation and efflux across the BBB. These processes are essential in avoiding Aβ accumulation and maintaining brain health. Efflux and subsequent degradation in the peripheral tissues is responsible for up to two thirds of Aβ clearance from the brain. Peripheral Aβ results from synthesis in the adrenal glands, organs, muscles, blood and endothelial cells as well as efflux across the BBB. Plasma concentrations of Aβ are 5 to 15 times lower than in cerebrospinal fluid (CSF), and the form of Aβ found in the periphery has less propensity to aggregate than that found in the brain of AD patients28. Comorbidity of AD with several systemic diseases, primarily diabetes mellitus and cardiovascular disease, is known to adversely affect the trajectory of AD. Aβ may be implicated in many of them; for example, accumulation of peripheral Aβ has been shown to be capable of affecting heart function and cardiac systolic dysfunction, which affects cerebral blood flow, and in turn affects Aβ synthesis and clearance in the brain29.

AD has been referred to as ‘type 3 diabetes’, due to the insulin resistance and overlapping signalling pathways seen in diabetes type 2 and AD30. Although Aβ is considered to be a necessary component in the pathogenesis of AD, it is also seen in the brains of aged individuals without the signs and symptoms of AD31. NFTs correlate better with the cognitive decline seen in Alzheimer patients32; however, amyloid may facilitate the spread of tau. Positive feedback loops between Aβ and inflammation cause a destructive and self-perpetuating cycle that can occur with or without the direct need for pathogens, either in the initiation or perpetuation of this process33. This is in contrast to periodontitis, in which the inflammation is dependent on the presence of bacteria.

There are several factors that contribute to neuroinflammation in AD. These include the misfolded Aβ and tau proteins, neuronal injury or death, and peripheral influences both from the systemic immune response to infection and, more controversially, from pathogenic virulence factors that have been able to reach the CNS. Aβ deposition in the brain is frequently initiated during the fifth decade of life, more than 15 to 20 years before AD is usually diagnosed34. The clinical diagnosis often correlates with the point at which the rising cerebral Aβ reaches a plateau. Challenges to the peripheral immune system, such as periodontitis beginning two decades earlier, by triggering the innate immune response may lead to increased Aβ production and further enhance the innate immune response. This would create a positive feedback loop, i.e. a vicious cycle. This is illustrated in Fig 9-1. Periodontitis is one such peripheral infection that commonly presents within 5 years either side of a patient’s 40th birthday35. This coincides with the age shortly before Aβ levels commonly begin to increase in the brain.

9.2.1.5 The immune system, neuroinflammation and neurodegeneration in AD

Encapsulated by the BBB, the CNS was previously believed to have a unique immune system. This theory has been discredited due to emerging evidence that immune dysfunction may be a key factor in the progression of neurodegenerative diseases such as AD36. In aged individuals, there is a gradual deterioration of the immune system that occurs as part of the natural aging process. This is termed immunosenescence, and can result in an overall immune dysregulation37 and a background low-grade chronic inflammatory state in the elderly, known as inflammaging, which is associated with many age-related conditions, including AD38. Similarly, periodontitis has been shown to cause a low-grade systemic pro-inflammatory state in patients that is likely to have a significant contributory effect on the cellular aging process39. Such immune dysfunction is likely to occur in tandem with the increased permeability of the BBB. This has been shown in animal models at an early stage in AD pathogenesis and in humans using neuroimaging techniques40. This means that there is potentially more cross-talk between the peripheral nervous system and the CNS than in younger healthy individuals, resulting in a greater influence of the peripheral immuno-inflammatory system on the CNS. The immune response can be both protective and destructive. This is true in both periodontitis and the neurodegeneration seen in AD. In the early stages of the disease process the inflammatory response is generally protective, whereas in the later stages the inflammatory response becomes upregulated, dysregulated, independent of the initial stimuli, and therefore destructive. The traditional conservative hypothesis is that neuroinflammation exacerbates the pre-existent neurodegenerative process of AD pathology41. The more controversial view held by some researchers is that neuroinflammation could be the driving force behind the pathogenesis of AD33.

Microglia and astrocytes

The resident macrophages, the microglia, together with astrocytes, support a healthy and stable environment. In addition, these cells, in conjunction with dendritic cells, provide protection and homeostasis by recognising and reacting to potentially harmful endogenous and exogenous molecules. These cells express receptors on their surface (e.g. Toll-like receptors) capable of recognising pathogen-associated molecular patterns (PAMPs). Inflammatory mediators are then released, instigating an inflammatory response. A common example of a PAMP is the bacterial endotoxin lipopolysaccharide (LPS), which is present on Gram-negative bacterial cell membranes. Microglia constitute 5% to 10% of total brain cells42, and have a long-established role in phagocytosis of pathogens and dead or damaged cells. In young, healthy brains, microglial activation is protective. Microglia communicate with almost all brain cells and control the expression of several aspects of the complement system within the CNS. They maintain homeostasis and are involved in tissue repair, but conversely they can also contribute to tissue destruction. Importantly, they also eliminate waste that has accumulated in the brain, including Aβ deposits, and contribute to the protection and remodelling of neuronal synapses. Microglia are present in increased numbers in AD brain compared with controls, often found around amyloid plaques42. Where there has been a breakdown of the BBB, macrophages from the peripheral tissues and blood may infiltrate into the brain and differentiate into microglia-like cells, which may influence the progression of neurodegenerative diseases42.

Microglia can react to components of pathogen cell walls, or misfolded proteins including ptau and Aβ, and can phagocytose or degrade Aβ using extracellular proteases. They also have receptors for advanced glycation end products (AGEs), known as RAGEs which may lead them to become hyper-responsive to neuron damage43. Microglia may be primed or pre-activated as part of the aging process, or due to contact with misfolded Aβ or tau proteins associated with AD pathology, or as the result of peripheral inflammatory signals44. Microglial activation mediates Aβ oligomer-induced synapse loss and neuronal death in AD via a number of pathways including complement induction42. Microglia are capable of changing from a benign supporting cell (the so-called ‘M2 state’) to an aggressive phagocytic state (‘M1 state’) able to engulf bacteria and other microbes and initiate inflammatory states in the parenchymal tissue45. Post-mortem analyses of AD brain tissue have demonstrated this change from a benign to an aggressive state in both microglia and astrocytes, and this is associated with a loss of neurotrophic support46. Such a pro-inflammatory response may ultimately result in cognitive decline47. Microglial priming has been shown to result in increased production of cytokines and reactive oxygen species (ROS), which, in conjunction with over-active phagocytosis, contributes to the neurodegenerative processes seen in AD48. Periodontitis is an example of peripheral inflammation that could cause microglia to be primed into a pro-inflammatory phenotype and cause damage to brain tissue. The pro-inflammatory M1 phenotype of peripheral macrophages has already been associated with tissue destruction in periodontitis and levels have been shown to rise further with increasing age49. Figure 9-2 provides an overview of the components of the brain immune response in health and AD.

LPS from Porphyromonas gingivalis has been shown to induce Aβ production in neuronal cell cultures and induce overproduction of interleukin 1 beta (IL-1β) and tumour necrosis factor alpha (TNF-α) from cell cultures of microglia primed with Aβ₄₂41. Similar observations were made recently in primary cultures of microglia and mixed hippocampal cells exposed to Aggregatibacter actinomycetemcomitans LPS 50.

Microglia become less efficient at clearing Aβ when chronically stimulated; thus chronic microglial activation (‘reactive microgliosis’) in transgenic mice following repeated LPS treatment results in increased Aβ deposition51. Gram-negative periodontal pathogens contain LPS within their cell walls and are likely to contribute to chronic stimulation of microglia once within the CNS. Until recently, less emphasis has been placed on the role of astrocytes in AD. Astrocytes are the most numerous and heterogenous of all neuroglial cells in the CNS; they control and maintain the brain environment. This includes regulation of extracellular ions, pH, neurotransmitter levels, cerebral blood flow and energy storage/distribution. Astrocytes also have roles in the immune response of the CNS; they release cytokines and chemokines and modulate the BBB52. Astrocytes contribute to the clearance of Aβ by their role in the direct phagocytosis of Aβ; they produce Aβ-degrading proteases and control the paravenous clearance of Aβ, which is dependent on the astrocytic water channel, aquaporin 433.

Astrocytes are also important for synaptic function and cognition33, and atrophy of astrocytes has been shown in mouse models of AD53. In AD brain, astrocyte atrophy begins in the entorhinal cortex, where memory formation and consolidation occurs. They are likely to be early responders to the cellular effects of AD, and may pre-date Aβ deposition34. Hypertrophic reactive astrocytes are also seen around amyloid plaques in AD post-mortem brain tissue54. Like microglia, astrocytes have been shown to be influenced by the peripheral immune system. Peripheral immune activation by LPS in mice results in both microglial and astrocyte activation in the brain55. Reactive microglia secrete IL-1α and TNF-α, which, together with a complement protein, C1q, can transform surrounding astrocytes from a responsive (A2) to a reactive (A1) phenotype. These reactive astrocytes then become cytotoxic to neurons44.

Leucocytes in the periphery and the brain

Leucocytes are believed to enter the CNS via the choroid plexus or the vascular network at the subventricular zones52. They may have a protective role in the removal of Aβ plaques, although this requires further investigation33. These innate immune cells have been shown to have a pro-inflammatory response in cognitively impaired individuals56. Neutrophils are the leucocytes that are recruited first to a site of infection or tissue damage. They are the most prevalent leucocyte present in inflamed periodontal tissues. They help to maintain periodontal health and the homeostasis between the host and microbial oral biofilm. In periodontitis the pathogens are not controlled, or eliminated sufficiently by neutrophils, to prevent chronic inflammation and degeneration of the periodontal tissues57. Neutrophils are capable of entering the human brain tissue58. Interestingly, it has also been shown that neutrophils that have migrated into the CNS acquire a toxic phenotype, which causes harm to brain tissue59. The first time neutrophils were associated with AD was when the neutrophil specific protease, cathepsin G, was found within Aβ plaques60. Subsequently, the association of neutrophils to AD has been shown by many studies in mice61 and also in clinical studies on humans62.

Neutrophils have been show to surround Aβ plaques with neutrophil extracellular traps (NETs)63, fibres that can bind pathogens. NETs can potentially cause breakdown of the BBB, and once within the brain, neutrophils and their products could cause enhanced microglial and astrocyte activation, thereby damaging neuronal cells58. Reducing the entry of neutrophils into the CNS by an antibody-mediated blockage of the entry portal (i.e. preventing LFA-1 integrin attachment), reduced amyloid and cognitive deficits in mice. Furthermore, peripheral intravenous injection of a neutrophil targeted antibody had a similar effect63. Apoptotic neutrophils are engulfed by macrophages, thus neutrophil apoptosis is a programmed cell death that avoids the release of harmful cell contents into the surroundings. In the absence of apoptosis, neutrophil necrosis results in the harmful contents of neutrophils being released into surrounding tissues and is therefore harmful to the host tissues. Some bacteria, for example P. gingivalis, can halt apoptosis64. Disruption of the normal life cycle and functions of neutrophils causing necrosis can result in increased host damage65 by triggering a host inflammatory response66. P. gingivalis has been found in the brain67 and could thus cause destruction at the BBB and within the CNS by eliciting neutrophil activation and necrotic cell death.

T cells are a heterogenous group of leucocytes that mature in the thymus and tonsils and which play a central role in the cell-mediated adaptive immune response. Overall, T cell numbers are reduced in AD and there is a higher proportion of memory T cells compared to age-matched controls68. The phenotype of circulatory T cells is altered in AD, becoming more reactive to Aβ69, with a reduced proportion of regulatory T cells responsible for controlling the immune response70. Furthermore, it has been speculated that T cells could affect the permeability of the BBB by influencing astrocytes52. Memory T cells have been found in higher numbers in the brains of those with AD, where they were found in close association with microglia71. Respiratory infection was shown to both increase Aβ deposition in the brain and increase T cell infiltration in rodent models of AD as the animals aged72. In particular, CD4+ and CD8+ regulatory cells have important roles in both periodontitis and AD, leading to an altered immune response. Thus alteration in T cell number and populations as the result of periodontitis could also influence T cell response in AD; further research is needed to elucidate the precise mechanisms involved73.

9.2.1.6 Protein mediators and modulators of neuroinflammation in AD: peripheral and central communication

There is mounting evidence of bidirectional crosstalk between the peripheral and central immune systems44, with peripheral immune activation contributing synergistically to neurodegeneration74. The heightened and chronic activation of the peripheral immune response seen in periodontitis is an example of such peripheral immune activation.

Cytokines

Cytokines are signalling proteins that are released by immune cells. The magnitude and duration of the host immune response to infection is regulated by cytokine and chemokine networks. They are produced by, and help to regulate the cellular inflammatory mediators. Microglia and astrocytes are the principle source of cytokines within the brain, though in moderate to late stage AD it is likely that a significant amount of them are released by peripheral immune cells that have infiltrated into the brain across a leaky BBB. Pro-inflammatory cytokines are created by multi-protein complexes called inflammasomes within immune cells. TNF-α levels in CSF from AD patients have been reported to be 25-fold higher than in controls75; and elevated CSF TNF-α increases the risk of patients progressing from mild cognitive impairment to AD76. In AD mouse models the cytokine load also correlated with Aβ levels77. IL-1β can favour Aβ deposition by altering the expression and proteolysis of APP78. It is thought that elevated cytokines promote the likelihood of a pro-inflammatory phenotype of microglia and macrophages within the brain before structural and cognitive changes occur33. The relationship between these cytokines is not straightforward, as in some animal studies IL-6 and TNF-α have been shown to protect against advancing AD pathology33. These cytokines have been shown in healthy rodents to be capable of crossing the BBB. Indeed they are believed to be selectively transported across it, into the brain79.

During normal healthy conditions there are few lymphocytes in the CNS; however, in the inflammatory environment cytokines such as IL-1β, IL-6 and TNF-α facilitate the entry of lymphocytes into the CNS. Once inside the brain they contribute to the inflammatory processes capable of causing cell death. These cytokines are also capable of affecting efflux transporters across the BBB. As efflux transporters are responsible for the expulsion of potentially harmful endogenous and exogenous molecules from the CNS, reduced function could lead to accumulation of Aβ and cytokines that drive the neurodegenerative process in the brain. In a mouse model of AD, peripheral LPS has been shown to cause increased levels of IL-1β, IL-6 and TNF-α, not only in the periphery, but also in the brain80. Furthermore, mice treated with intraperitoneal LPS showed increased serum levels of cytokines, which altered BBB transport of Aβ, where influx was increased and efflux decreased, resulting in an overall increase of cerebral Aβ81. These effects are seen to a greater extent in aged mice and those with pre-existent neurodegenerative conditions81. In a landmark study, oral inoculation of P. gingivalis in an AD mouse model led to significant alveolar bone loss after 5 weeks, notably resulting in increased levels of IL-1β, TNF-α and endotoxin in the brain41. Importantly, it was shown that the cognitive function was significantly impaired in these mice. Post-mortem analysis showed significantly increased Aβ deposition in the hippocampus, with an increase of Aβ₄₀ and Aβ₄₂ both in the hippocampus and cortex41. Similarly, in a recent study, repeated oral administration of P. gingivalis was also shown to lead to accumulation of this bacterium in the hippocampi of mice, with a localisation intra- and peri-nuclearly, in microglia, astrocytes, neurons and extracellularly82. Inflammatory markers along with phosphorylated tau and Aβ₄₂ were also increased, and neurodegeneration was evident in these animals compared to control animals82.

The chance of developing AD has been said to be doubled for elderly people exposed to systemic infection83 and is associated with more extreme ‘sickness’ behaviours such as apathy and depression in those patients. Greater cognitive decline has been observed in people with AD who have systemic infections associated with increased plasma IL-1β for at least 2 months after resolution of the infection47. Notably, periodontitis results in a local increase of cytokines in the affected tissues, with IL-1β, IL-6 and TNF-α produced as part of the host response84. Further work needs to be carried out on this in terms of any peripheral influence of these cytokines on the BBB.

Neuropeptides

Neuropeptides are small signalling molecules that are released by neurons, although other peripheral cell types are also capable of secreting them in the periphery in response to inflammation85. They have been identified in gingival crevicular fluid and there is increasing86 evidence that the peripheral tissue destruction seen in periodontitis can be modulated by neuropeptides including substance P, calcitonin gene-related peptide (CGRP), and neuropeptide Y, which are upregulated in periodontitis. The nervous system has been identified as a critical regulator of periodontal inflammation87 and as neuropeptides are synthesised by immune cells as well as neurons it is likely that there is two-way communication between the immune system and the nervous system to control periodontal inflammation85. This also poses a further mechanism by which the peripheral inflammatory response seen in periodontitis could affect the neuroinflammation seen in AD and vice versa. Substance P has several neuroprotective roles, one of which is believed to be its ability to stimulate non-amyloidogenic APP processing (that is, the cleavage of APP via a pathway that does not produce Aβ), thus reducing the production of toxic Aβ peptides in the brain88. CGRP potentially inhibits the infiltration of macrophages into the brain and reduces the expression of many inflammatory mediators that are believed to be upregulated in AD89.

Complement

The complement system is a cascade of antimicrobial proteins that bridges the innate and adaptive immune response and is important in maintaining a controlled immune response to pathogens in the periphery90. Complement factors have long been known to be associated with Aβ deposits in AD91. Aβ is able of activating the complement system, via the alternative pathway in vitro. Moreover, several complement proteins have also been measured at higher levels in those with AD than controls92. Polymorphisms of the CR1 gene, a critical regulator of C3 convertase in humans, have also been implicated in Aβ clearance from the brain93. Cleavage of C5 is also likely to have a detrimental effect on AD progression, as mice treated with a C5a receptor antagonist showed decreased signs of AD pathology94. Complement also has an important role in periodontal inflammation and pathogenesis though not all complement pathways and functions are known95. P. gingivalis is capable of subverting complement receptor 3 and C5a to promote a pro-inflammatory destructive response, leading to the suggestion that these could be a future therapeutic target for the management of periodontitis90. C3 is believed to be critical in the host-bacterial perpetuation of dysbiosis seen in periodontitis39, and the expression of the C3 gene is downregulated following successful periodontal therapy96. In summary, the roles of the complement system in periodontitis and AD are complex. Further research is required to elicit whether there could be interplay of complement factors between the two diseases.

9.2.1.7 Reactive oxygen and nitrogen species

ROS are oxygen free radicals and other oxygen derivatives involved in oxygen radical production97. They are involved in physiological cellular functions, where they are effectively controlled by antioxidants to prevent tissue damage. ROS production is rapidly increased in inflamed tissues, as part of the ‘respiratory burst’ from neutrophils, macrophages and other cells98. They act as mediators of inflammation and when they are not balanced by antioxidants cause oxidative stress and tissue damage99. Reactive nitrogen species (RNS) are derived from nitric oxide. Like ROS, RNS have physiological roles in normal cellular function, although when out of balance with antioxidants, they behave in a similar way to ROS and are capable of causing nitrosative stress to host cells. Peripheral neutrophils from those with periodontitis have consistently been shown to produce more ROS than those from healthy individuals100, even when they are not directly stimulated101, suggesting they have an inflammatory phenotype. The increased ROS release seen in periodontitis is likely to be not only due to the initial stimulation from periodontal pathogens, but also genetically determined102. ROS and RNS are increased in AD brains103. Oxidative stress within the brain is believed to support the formation of Aβ species that more rapidly lead to amyloid plaque formation, and high concentrations of ROS and RNS can also be directly toxic to neurons33. As these are volatile molecules with a very short life, it is unlikely that ROS or RNS molecules from the periodontal tissues cross the BBB. However, it is possible that pre-primed pro-inflammatory leucocytes cross a leaky BBB, where they could release increased amounts of ROS/RNS, potentially causing more damage to the other cells of the CNS than would otherwise be the case had the leucocytes not been pre-primed by periodontitis.

9.2.1.8 The role of microorganisms in AD

Many different types of microbes live on the human body, including bacteria, archaea, fungi, protozoa, viruses and microscopic animals. This includes both commensal organisms that live in synergy with the host and pathogens, capable of causing disease in the host. Pathobionts are capable of causing disease, yet under normal healthy circumstances live in a symbiotic relationship with the host. However, given circumstances of microbial dysbiosis these pathobionts can cause and exacerbate the chronic disease process. Periodontal pathobionts could unfavourably affect the inflammatory environment in the brain during AD pathogenesis, either indirectly through their effect on the peripheral immune system and inflammation, or centrally if they can enter the tissues of the brain. The periodontal tissues provide a unique environment to allow the ingress of periodontal pathogens into tissues, nerves and blood vessels that are anatomically close to the brain. This has been highlighted in a recent editorial written by leading experts in AD research, who report evidence that the hallmarks of AD are indicators of an infection and call for action to further investigate the correlation between microbial infection and AD pathogenesis104. Sceptics of this research believe that microbes in the brain are the result of post-mortem contamination. Though this may account for some of the species present, it cannot explain the presence of the variety of bacteria found; nor can it explain the differences in the numbers and proportion of bacterial species present between diseased and non-diseased brain, or the association of certain species to amyloid plaques105.

Potential microbial entry portals to the brain

There are several potential modes of entry for periodontal pathogens to reach the brain. The virulence and tissue invasive nature of many periodontal pathogens, which are anatomically close to the brain, has been well documented. By far the most likely, and with the most evidence, is that bacteria enter the peripheral blood stream and are then transported through the BBB into the CNS. Bacteraemia of oral origin has been recognised since the 1970s. It has been speculated that the proliferation and dilation of the periodontal vasculature and ulceration of the periodontal epithelium, seen in periodontitis, provide a larger surface area for the entry of microorganisms into the blood stream compared to healthy gingival tissue106. A systematic review and meta-analysis has determined that plaque accumulation and gingival inflammation significantly increase the prevalence of bacteraemia following tooth brushing107. Gingival inflammation has also been significantly associated with the incidence of bacteraemia following scaling and root planing108. The viability of a number of periodontal pathogens such as P. gingivalis and Treponema denticola in atherosclerotic plaques has been recognised109 and adds credence to the possibility that oral bacteria can reach the brain via the systemic circulation.

Such pathogens have the potential to linger in the systemic circulation, prolonging what was previously thought to be a transient bacteraemia110. It has also been speculated that P. gingivalis is able to promote the disruption of barrier function, potentially affecting both the junctional epithelial attachment of gingivae to the root surface and the BBB111. The other potential modes of entry for pathogens to reach the brainstem include the cranial nerve afferents and penetration into the brain structures not encapsulated by the BBB, such as the circumventricular organs, and these would thus act as an entry portal for bacteria to reach other parts of the brain. The identification of oral Treponema in the trigeminal ganglia also supports this neural route of pathogen transport112; this is also a potential route for the transfer of viruses, particularly herpes viruses. Interestingly, many of the first changes associated with the pathology of AD occur in the brainstem very close to the cranial nerve nuclei. The locus coeruleus and its noradrenergic projection neurons containing neuromelanin were the first to show the presence of abnormal hyperphosphorylated tau forms in the axon113. The serotonergic nucleus of the dorsal raphe also showed early lesions, which progressed to the central and linear raphe nuclei, and later the caudal raphe nuclei. Pathology was most severe in the dorsal raphe nucleus114. Evidence suggests that the tau pathology occurs also at an early stage in this brain region115.

Bacteria

Bacterial meningitis caused by Neisseria meningitidis is the result of a severe and acute reaction to bacteria within the brain. Some bacteria are capable of entering brain tissue without causing such an immediate and pronounced effect. An example of this is the spirochaete Treponema pallidum, responsible for the syphilitic dementia seen in tertiary syphilis, which shares the same amyloid plaque histopathology as AD116. In syphilis the initial infection with T. pallidum is known to occur many years before the symptoms of syphilitic dementia. An emerging body of evidence indicates that microbes can be present within the brain without triggering an acute response. Furthermore, the microbes present in the healthy brain have been shown to be different to those found in brains affected by AD117. Various techniques have been used to identify bacteria in human post-mortem brain tissue; earlier studies used microscopy and later studies have identified bacteria through DNA amplification and detection using polymerase chain reaction and the unbiased technology of next generation sequencing (NGS).

Numerous microbes have been isolated to a greater extent from the brains of AD patients than from age-matched non-AD patients post-mortem. In the 1990s, Miklossy118, using microscopy, described the presence of obligate oral anaerobes such as spirochaetes, in blood and CSF from Alzheimer patients but not in age-matched controls. One literature review cites evidence of oral spirochaetes in over 90% of AD brains119, and Borrelia burgdorferi and various Treponema species have been detected at much greater levels in AD brain than in controls112. Recently, Aβ was shown to be a major component of spirochaetal biofilms within amyloid plaques in AD brain tissue and it is suggested that bacterial amyloid is a constituent of AD plaques and may contribute to uncontrolled inflammatory drive105. Elevated plasma levels of antibodies to the oral bacteria A. actinomycetemcomitans, Tannerella forsythia and P. gingivalis were found in AD patients compared to controls120. Furthermore, a longitudinal study of cognitively unimpaired individuals demonstrated higher baseline serum antibody levels, specific for the oral anaerobes Prevotella intermedia and Fusobacterium nucleatum, which were shown to correlate with cognitive deficits a decade later121. Cutibacterium acnes117, Escherichia coli122 and Chlamydophila pneumoniae123 have also been identified as more commonly present in AD brains compared to controls.

Although these are not known components of oral microbial communities, it is possible that periodontal pathogens could interact with other bacteria differently in the brain than they do in periodontal tissues. The significance of this is yet to be discovered. It has not yet been determined whether bacteria in the brain of those without acute signs of infection are present in a viable form, and this will be a goal of future research in this field. Bacterial infiltration into the brain is likely to result in an immune response instigated by the native microglia and any peripheral immune cells that have infiltrated into brain tissue, including the production of cytokines and initiation of aspects of the complement cascades. The major component of the outer membrane of Gram-negative bacteria is LPS, which is capable of evoking a strong immune response. LPS from the oral anaerobe P. gingivalis has been detected in the brains of AD patients, but not in control brain tissue124. In a recent review article by Olsen et al125, it was also suggested that P. gingivalis and its secreted peptidyl arginine deiminase enzyme, through enzymatic conversion of arginine to citrulline, may contribute to AD. Citrullination of host proteins has been associated with the pathogenesis of multifactorial diseases such as rheumatoid arthritis (see Chapter 5 ‘Periodontitis and rheumatoid arthritis’) and also with AD. Citrullinated proteins have been detected in brain tissue in several studies. Overexpression of protein arginine deiminases and protein citrullination are abnormal features of neurodegeneration and inflammatory diseases and have been proposed as a possible cause of AD126. Therefore, it is possible that P. gingivalis directly contributes to neurodegeneration.

Viruses

Herpes simplex virus (HSV) and cytomegalovirus (CMV) have also been detected in the brain tissue of older adults with and without AD127. More recently, circulatory HSV antibodies have been associated with the severity of AD128. There are around 100 individual studies that indicate HSV 1 is a major risk factor for AD104. Whilst latent viruses may prove inconsequential for many people they may be reactivated by immunosuppression, stress or inflammation in the brain129. Those who have the APOE-ε4 (apolipoprotein E) allele, responsible for increasing the risk of developing sporadic AD (see Section 9.2.1.9 ‘Genetics of AD’), are likely to be more susceptible to reactivation and/or degree of damage caused by viruses within the brain130. In vitro and in mice, cerebral infection with HSV has led to the Aβ deposition and tau abnormalities typical of AD131^,132. Circulating antibodies to CMV have also been associated with AD progression, and faster cognitive decline in those infected with CMV was seen in a 5-year follow-up study of AD patients133. HSV, CMV and Epstein-Barr virus type 1 (EBV) have been isolated from periodontal pockets with much greater frequency than from healthy gingival crevices134 and have been hypothesised to be implicated in the pathogenesis of periodontitis135, potentially having a synergistic relationship with other periodontal pathogens. The periodontal tissues could therefore act as a reservoir for the dissemination of viruses to other body sites such as the brain.

Fungi

Candida albicans is known to be an opportunistic pathogen that is found in many sites within the oral cavity, including the periodontium, being recovered from the subgingival biofilm in 7% to 20% of those with periodontitis, with higher prevalence among immunocompromised hosts136. Candida species have been shown to form aggregates with bacteria in the dental biofilm, adhere to epithelium and invade into connective tissue where the collagenases and proteases they produce can degrade host proteins within the extracellular matrix136. Once within the tissues the host will mount an immune response, which will include components of the innate and adaptive arms of the immune system, with a significant proportion of the immune response controlled by binding to Toll-like receptors TLR2 and TLR4136. CSF levels of C. albicans proteins were also found to be higher in those with AD137. DNA of C. albicans and other fungi have also been found in AD patients but not controls138. Moreover, fungal matter had been identified within neurons in human post-mortem AD brain samples139. Further to this, polymicrobial infections have also been reported as being present in the entorhinal cortex and hippocampus of post-mortem AD brain including fungi and bacteria such as Burkholderia species140. Little is known regarding the significance of fungi at periodontal sites and within the brain in relation to the pathogenesis of periodontitis and AD, respectively, although further research is ongoing in these areas141.

9.2.1.9 Genetics of AD

Genetic mutations are permanent changes within the genome that can lead to disease. There are three gene mutations that lead to the inheritance of AD and each of them causes early onset familial AD (EOAD) (Fig 9-3). They include the genes APP, PSEN1 and PSEN2 which encode respectively APP, presenilin 1 and presenilin 2142. Each of these mutations affects the breakdown of APP into forms that predispose the individual to amyloid plaque formation. Polymorphisms in the APOE gene affect an individual’s risk of developing AD, but do not directly cause it142. The APOE gene has three common alleles. Of these, the APOE4 allele is the most well-known as it increases the risk of developing AD and lowers the age of onset, compared with the most prevalent allele APOE3. Genome-wide association studies (GWAS) search the genome for small genetic differences called single nucleotide polymorphisms (SNPs) that are associated with a particular condition or trait. This database143 lists several genes that are risk factors for AD or periodontitis. There are 705 AD-associated and 155 periodontitis-associated SNPs. Twenty-five SNPs that are associated with periodontitis are co-located in the same gene area as those associated with AD; there are two specific genes, ITGA8 and DLG2 for which SNPs have been identified as risk factors for AD, PD and periodontitis. These code, respectively, for integrin subunit alpha 8, known to be involved in neurite outgrowth, and discs large MAGUK (membrane-associated guanylate kinase) scaffold protein 2, which takes part in a number of interactions including the binding to NMDA channels.

Investigating these SNPs further will deepen understanding of aspects of shared pathogenesis between these neurodegenerative conditions and periodontitis. There have been a number of studies of the common vitamin D receptor (VDR) polymorphisms, and a meta-analysis demonstrated robust associations with periodontitis. Interestingly, the t allele of the Taq-1 locus appeared to be a protective factor for chronic, but not aggressive (Grade C) periodontitis in Asians, yet this was not true for white individuals144. Other notable polymorphic associations with increased risk of AD are APA1Aa, Apa-1t and Taq-Ig, particularly in those over 75≈years145. A meta-analysis showed strong associations of the TaqI and ApaI polymorphisms, particularly in Asian populations, with AD145. Vitamin D elicits neuronal protection by a number of mechanisms including reduction of inflammation, by suppressing pro-inflammatory cytokines such as TNF-α, and increased removal of Aβ₄₂146. Periodontitis may therefore have an effect on the progression and age of initiation, particularly for those individuals on the right of the spectrum (Fig 9-3), as they are influenced by other risk factors to a greater extent. Figure 9-4 gives an overview of the different factors that may contribute to the initiation and/or progression of AD.

Stay updated, free dental videos. Join our Telegram channel

Join