Parkinson's Disease: A Multisystem Disorder

Abstract

The way sporadic Parkinson's disease (PD) is perceived has undergone drastic changes in recent decades. For a long time, PD was considered a brain disease characterized by motor disturbances; however, the identification of several risk factors and the hypothesis that PD has a gastrointestinal onset have shed additional light. Today, after recognition of prodromal non-motor symptoms and the pathological processes driving their evolution, there is a greater understanding of the involvement of other organ systems. For this reason, PD is increasingly seen as a multiorgan and multisystemic pathology that arises from the interaction of susceptible genetic factors with a challenging environment during aging-related decline.

Introduction

Parkinson's disease (PD) is currently described as a multisystem neurodegenerative disease, since it concomitantly involves the central nervous system (CNS), enteric nervous system (ENS), autonomic nervous system, adaptive immune system, and gastrointestinal (GI) tract [1–5]. Clinically, it is characterized by the occurrence of both motor and non-motor symptoms (NMS) [6].

While bradykinesia associated with resting tremor and rigidity is still considered a cardinal manifestation of the disease, it is insufficient for clinical diagnosis. Thus, an established clinical diagnosis requires, besides the cardinal manifestation, the presence of at least two supporting criteria, in the absence of exclusion criteria or "red flags". The support criteria include a response to dopaminergic therapy, dyskinesia secondary to levodopa treatment, and the presence of NMS [1–6].

NMS such as hyposmia/anosmia, REM sleep behavior disorder, depression, anxiety, cardiac sympathetic denervation, and constipation, may precede dopaminergic neuronal loss and the consequent motor symptoms by several years [6, 7]. Although NMS are present during disease progression and are an important cause of reduced quality of life, their presentation and timing vary. Sleep and autonomic nervous system disturbances are present from early to later stages but dementia and psychosis are more likely to occur in later stages of the disease [5, 7]. The presence of these manifestations is a reminder of the inherent pathophysiology of PD that extends beyond the dopaminergic neuronal loss in the pars compacta [1, 6–9].

Before The Race: Risk Factors

It is thought that there is a close relationship between genetic predisposition, a favorable environment, and ageing in the onset of the pathology. Therefore, although hereditary forms of the disease have been described, most cases are sporadic with no known etiology. As a result, several possible risk factors have been suggested [1, 10–12].

Exposure to External Agents and Lifestyle

Rural living, farming and untreated well-water consumption have been correlated with an increased risk of developing PD [13, 14]. Although this epidemiological relationship is not completely understood, it is known that there is an increased risk of PD in individuals with occupational exposure to pesticides, as well as, a worse prognosis [13, 15]. This risk is clearer with pesticides that affect complex I of the mitochondrial respiratory chain and induce oxidative stress, such as Rotenone and Paraquat [1, 13–18]. Although the neurotoxic action of heavy metals has been described in animal models, a definite link between exposure and increased risk for PD remains controversial [1, 11]. Iron in its free form is capable of inducing oxidative stress and its accumulation in the substantia nigra of patients is also known [19, 20]. However, and despite the hypothesis that alterations in iron metabolism confer susceptibility to PD, the comparison of serum iron levels between PD patients and controls has yielded contradictory results [1, 19–26]. Similarly, the relationship of PD with polymorphisms in C282Y or H63D in the HFE gene, which encodes a protein responsible for regulating iron levels in liver cells and hepcidin production, also remains controversial, although some data point to a protective effect of the HFE-C282Y polymorphism [25]. Low serum zinc levels have been also described as a potential risk factor for PD [24].

Exercise is a known protective factor for PD. One of the possible reasons is the reported higher serum urate levels that have been correlated with a lower risk of PD and improved prognosis in men. However, the same relationship remains controversial in women [11, 16, 26–29]. Urate is a product of purine metabolism, which may have an antioxidant effect through activation of the Nrf2/ARE (antioxidant responsive element) pathway, protecting against the loss of dopaminergic neurons [16, 26]. The Nrf2/ARE pathway has already been indicated as a possible therapeutic target in neurodegenerative pathology, since it is responsible for regulating, both in a constitutive and induced manner, the expression of proteins essential to cellular protection against oxidative stress. This pathway is also capable of improving mitochondrial biogenesis [27, 30]. Thus, physical exercise, and thereby the rise of urate serum levels, may have dual beneficial effects on PD by preventing mitochondrial dysfunction and oxidative stress [11, 16, 26–28].

Coffee consumption has also been linked to a lower PD risk, and this relationship is most evident in men [16, 18, 26, 31]. The protective effects of coffee are thought to be mostly due to caffeine, an hypothesis strengthened by the fact that other sources of caffeine, such as teas and soft drinks, have the same effect, unlike decaffeinated coffee, which has not been shown to reduce the risk of PD [26, 32]. Nevertheless, we cannot exclude other components that may be removed during the aqueous extraction of caffeine. Among women, the decreased risk is controversial, probably because estrogens and caffeine are metabolized by the same enzyme, CYP1A2 [16, 31, 32].

Caffeine acts as a competitive adenosine receptors antagonist, having a higher affinity for the A1 and A2A adenosine receptors (A1R and A2AR). Chronic A1R blockade by caffeine is known to lead to upregulation of these receptors, resulting in decreased pro-inflammatory cytokines, while A2AR inhibition has anti-inflammatory and anti-apoptotic effects. By this route, caffeine prevents adenosine-mediated neuroinflammatory processes by decreasing microglial reactivity, glutamate release, and the release of pro-inflammatory cytokines [16, 32–34].

Smoking has been associated with a lower PD risk, but the hypothesis of a mortality bias associated with smoking has been refuted. It has been recognized that the protective effect is greater in men and among long-term smokers [16, 18, 35]. A PD premorbid personality was described; indeed, PD patients who score high on neuroticism have greater risk aversion and lower sensation-seeking scores [18, 36, 37], therefore they are thought to be less prone to smoking. However, this behavior does not fully explain the protective effect; the mechanisms inherent in this relationship are still under scrutiny [11, 13, 16, 18, 35, 38].

The protective effect of smoking is believed to be associated with nicotine, since it has been shown to inhibit the formation of α-synuclein fibrils and to reduce the production of free radicals by monoamine oxidase B (MAO B). On the other hand, MAO B metabolizes dopamine and is involved in the production of inflammatory cytokines, as well as in apoptosis [11, 13, 16, 18, 35, 38].

Patient Susceptibility

Advanced age is the greatest known risk factor for PD. However, other individual intrinsic characteristics are potential risk factors. Being a male is a risk factor, as well as a worse prognostic factor [7, 16, 26]. Males have a higher incidence of PD and a slightly higher mortality than females [39]. The presentation and age at disease onset also differ between genders. Women have a later motor symptoms onset and more benign progression [7]. Nevertheless, symptoms such as fatigue, depression, anxiety, constipation, pain, hypo/anosmia, hyperhidrosis, and propensity for severe dysphagia are more common and severe in women. Nonetheless, the higher incidence of depression, anxiety, and pain in women is not specific to PD. Women are also more likely to develop postural instability, motor complications, and hallucinations due to iatrogenic symptomatic therapy [7, 29]. Men have greater cognitive impairment, increased sexual dysfunction, severe REM sleep disturbances, association with impulse control disorder, and severe sialorrhea [7, 29].

Sex dissimilarities may depend on structural differences, such as a higher ratio of D1:D2 dopaminergic receptors in females in the dorsal and ventral striatum, but also on the contrasting hormonal background [29]. In an animal model of PD, sex differences are blurred after an oophorectomy [40]. Estradiol seems to have a neuroprotective effect by a pleiotropic action on the CNS. Its interaction with insulin-like growth factor-I has been demonstrated [41], as well as its ability to modulate cell death by inducing the expression of Bcl-2 and Bcl-w (anti-apoptotic proteins) and decreasing Bad and Bim (apoptotic proteins). Estradiol is also thought to increase the synthesis, release, and reuptake of dopamine, as well as to decrease the production of reactive oxygen species, improving mitochondrial function [42, 43].

Having a family history of PD is also a known risk factor, although the monogenic form is a minority of diagnoses [18]. However, to describe PD as monogenic or sporadic is a simplification of the genetic contribution to disease development. In reality, there is a spectrum of genetic mutations with varying degrees of penetrance, which again emphasizes the interaction between individual genetic susceptibility factors and the external environment [44, 45]. On this spectrum, we have high-penetrance mutations responsible for familial forms of PD and low-penetrance mutations have been found in sporadic forms [45, 46]. An example is that individuals with the LRRK2 locus mutation, who do not develop PD, have higher serum urate values [29]. Note that, as previously discussed, the relationship of serum urate levels with PD is not equal in both genders; it is also known that the phenotypic expression of these mutations may be dependent on the sex of the patient [16, 26, 47].

DNA methylation levels may also influence a patient's vulnerability to pathology. In blood and saliva samples from patients, co-methylation in genes associated with mitochondrial protein coding or involved in fighting oxidative stress, such as LARS2, MIR1977, and DDAH2, have been associated with PD as well [48].

The Starting Point: The Gut of a Parkinson's Patient

The relationship between the gut and the human brain has opened ample new opportunities to understand and perhaps explain neurological pathology, this being especially true in PD [49, 50].

The human gut is the home of microorganisms from the three domains of life, namely bacteria, archaea, and microeukaryotes such as fungi, as well as viruses, which together form the gut microbiome. This microbiome is now known to be essential for the development and maturation of the host nervous, immune and endocrine systems [49]. In the last decade it has been possible to assess the extent to which the microbiome composition is vulnerable to the complexity and interaction of the host with the external environment, in a way that is not limited to diet or physical exercise [51]. Moreover, mounting evidence supports intimate bidirectional crosstalk between the gut microbiome and the CNS mediated by neurotransmitters (serotonin, dopamine, and GABA), their precursors or derivatives (tryptophan and histamine), and neuromodulators (short-chain fatty acids) or other substances [52].

The irrefutable association between PD and the intestine arose from careful investigations over the natural history and evolution of the disease from very early on, showing that GI symptoms precede motor symptoms and that Lewy bodies are detectable in the gut before they appear in the CNS, leading to the hypothesis that the origin of PD lies in the gut [2, 7, 28, 52–56]. This relationship is further corroborated by the fact that patients with inflammatory bowel disease (IBD) have a higher risk of developing PD [53].

In 2007, Braak proposed the dual-hit hypothesis, which advocates that the olfactory bulb and the vagus nerve are key points to enable an unknown infectious agent in the gut to trigger PD [57]. This hypothesis is supported, in some patients, by the characteristic spreading pattern of α-synuclein aggregates. In these PD patients with a ‘body first’ phenotype, the prion-like spread of α-synuclein aggregates starts in the enteric system and reaches the CNS via the vagus nerve due to retrograde neuronal transmission. Thus, it has been documented that spreading passes through the dorsal motor nucleus of the vagus and locus coeruleus at the level of the medulla oblongata and, subsequently, ascends through the brainstem and, only then, reaches the substantia nigra [58–60].

Other hypotheses emphasize intestinal dysbiosis, that is, disruption of the balance of the microbial ecosystem, as the starting point for the progressive evolution of, at least, a subset of sporadic PD cases (Fig. 1) [1, 2, 28, 61, 62].

Fig. 1.

Schematic of the Cardoso & Empadinhas and Johnson et al. hypotheses proposing gut dysbiosis as a trigger of PD. A The Cardoso & Empadinhas hypothesis proposes that exogenous microbial toxins or those produced by the gut microbiome, in the context of gut dysbiosis, lead to mitochondrial damage. Consequently, the release of mitochondrial DAMPs leads to neuronal sterile inflammation via activation of Toll-like receptors (TLRs) and Nod-like receptors (NLRs). Ultimately, this inflammation leads to neurodegeneration of the ENS and CNS [28]; B Johnson and colleagues describes 3 phases that include triggers, facilitators, and aggravators. Triggers act transiently, e.g. trauma and exposure to toxins or pathogens, and although they are insufficient to generate the disease, they allow it to develop in the presence of facilitators. In turn, facilitators, such as mitochondrial dysfunction or gene mutations associated with PD, are conditions that precede or are concomitant to triggers, allowing CNS involvement and leading to a state of chronic systemic inflammation. Finally, aggravators include changes in autophagy and neuroinflammation that are associated with the progression of the pathology [2].

Several cohort studies have investigated the gut microbiome composition in PD patients (Table 1) and, although there are some common points in the different studies, results are also conflicting (Fig. 2) the significance of which is still uncertain given the disparities in sample size, clinical presentation, and stage of pathological progression in the patients selected [54–56, 61–65]. Still, the changes found in the microbiome of PD patients tend to corroborate the hypotheses sustaining gut dysbiosis as a trigger of PD [53, 54, 61, 63, 64]. However, the etiology beyond PD gut dysbiosis is still unclear. Since PD is intimately correlated with advanced age, one can point to its association with age-related dysbiosis, a progressive imbalance of microbial populations that translates into a decrease in the production of beneficial short-chain fatty acids as a result of microbiome functional loss [52, 66], and also to the age-related chronic pro-inflammatory state designated ‘inflammaging’ [67].

Table 1.

Comparison of eight microbiome profiling studies of Parkinson’s patients

Author (year)	Sample	Sequencing	Increased in PD Family (genus)	Decreased in PD Family (genus)
Scheperjans, et al (2015) (54)	Fecal samples (PD n=72; Controls n=72)	V1-V3 regions 16S rRNA gene.	• Enterobacteriaceae	• Prevotellaceae
Unger, et al (2016) (55)	Fecal samples (PD n=34; Controls n=34)	16S rRNA gene	• Enterobacteriaceae	• Prevotellaceae
Qian, et al (2018)(56)	Fecal samples (PD n=45; Controls n=45)	V3-V4 region of 16S rRNA gene	• Rikenellaceae (Alistipes), • Prevotellaceae (Paraprevotella) • Enterobacteriaceae (Klebsiella) • Sphingomonadaceae (Sphingomonas) • Moraxellaceae (Acinetobacter) • Comamonadaceae (Aquabacterium) • Desulfovibionaceae (Desulfovibrio) • Clostridiaceae (Clostridium ) • Lachnospiraceae • Oscillospiraceae (Butyricicoccus) • Nitrososphaeraceae (Nitrososphaera)	• Lactobacillaceae (Lactobacillus) • Chitinophagaceae (Sediminibacterium)
Petrovet al. (2017)(61)	Fecal samples (PD n=89; Controls n=66)	16S rRNA gene	Christensenellaceae (Christensenell) • Catabacteraceae (Catabacter) • Lactobacillaceae (Lactobacillu) • Oscillospiraceae (Oscillospira, Ruminococcus, Papillibacter) • Bifidobacteriaceae (Bifidobacterium)	• Lachnospiraceae (Dorea Blautia glucerasea, Coprococcus Eutactus) • Bacteroidaceae (Bacteroides) • Prevotellaceae (Prevotella) • Oscillospiraceae (Faecalibacterium, Ruminococcus callidus)
Baldini et al. (2020)(62)	Fecal samples (PD n=147; Controls n=162)	16S rRNA gene	• Verrucomicrobiaceae (Akkermansia.) • Lactobacillaceae • (Lactobacillus)	• Turicibacteraceae (Turicibacter) • Prevotellaceae (Paraprevotella - female PD)
Chiang et al. (2019)(63)	Fecal samples (PD n=80; Controls n=77)	V3-V4 region of the 16S rRNA gene	• Deferribacteraceae (Mucispirillum), • Porphyromonadaceae (Porphyromonas) • Lactobacillaceae (Lactobacillus), • Tannerellaceae (Parabacteroides)	• Prevotellaceae (Prevotella)
Li, et al (2017)(64)	Fecal samples (PD n=24; Controls n=14)	16S rRNA gene	• Enterobacteriaceae (Escherichia, Shigella, Enterococcus) • Streptococcaceae (Streptococcus Proteus)	• Lachnospiraceae (Blautia) • Oscillospiraceae (Faecalibacterium, Ruminococcus)
Hill-Burns E. et al (2017)(65)	Fecal samples (PD n=197; Controls n=130)	16S rRNA gene	• Verrucomicrobiaceae (Akkermansia) • Lactobacillaceae (Lactobacillus) • Bifidobacteriaceae (Bifidobacterium)	• Lachnospiraceae

Fig. 2.

Summary of PD microbiome profiles obtained in eight studies. Increased relative abundance of Enterobacteriaceae in four studies and decreased Prevotellaceae in PD patients reported in five studies are noteworthy. Four studies detected increased abundance of Lactobacillaceae, but one study found decreased levels in PD. The relative abundance of Lachnospiraceae was decreased in PD patients in three studies, with one study coming to opposite conclusions.

Exposure to antibiotics is another possible cause of gut dysbiosis, although its direct association with neurodegenerative diseases is still ambiguous. These drugs impact microbiome biodiversity, leading to dysbiosis that may range in extent and duration. Curiously, some antibiotics have shown neuroprotective and anti-inflammatory effects [52]. In this sense, although some studies have found no association between antibiotic intake and the incidence of PD [68], others have associated the consumption of broad-spectrum penicillin with a higher prevalence of PD [69], as well as a higher risk of PD associated with exposure to certain macrolides and lincosamides [70]. Rifampicin, a macrocyclic antibiotic, has shown neuroprotective properties in a cell model by reducing mitochondria-mediated oxidative stress, decreasing microglia activation, and suppressing the expression of α-synuclein aggregates [71]. The tetracyclines doxycycline and minocycline are also thought to possess neuroprotective properties that may have utility in neurodegenerative diseases such as PD, by way of their anti-inflammatory, anti-apoptotic, and free-radical-scavenging actions that contribute to a decrease in mitochondrial dysfunction and a reduction in microglial activation [72, 73].

It has been theorized that certain microbial toxins contribute to the development of neurodegenerative diseases. Indeed, the bacterial endotoxin lipopolysaccharide (LPS) promotes mitochondrial dysfunction and neuroinflammation, both processes widely associated with PD [2, 28, 74, 75]. Another example is the neurotoxin β-methylamino-L-alanine (BMAA), a non-proteinogenic amino-acid produced by some cyanobacteria and other microbes [76–79]. BMAA was initially associated with the amyotrophic lateral sclerosis/parkinsonism-dementia complex of Guam [80]. Since then, BMAA has been hypothesized to contribute to the development of neurodegenerative diseases in susceptible individuals [79, 81]. When present at high concentrations, this toxin was first shown to trigger excitotoxicity via N-methyl-D-aspartate receptors but given the suspicion that its contribution to the neurodegenerative process may reflect a chronic exposure at low concentrations, other mechanisms of toxicity have been investigated [75, 81, 82]. Recently, we showed that in wild-type (wt) mice, BMAA erodes a protective bacterium in ileum-associated mucosa responsible for the homeostatic immune response, leading to intestinal barrier disruption and systemic low-grade inflammation, which impacted the substantia nigra blood-brain barrier (BBB). We also found that BMAA induces mitochondrial dysfunction and fragmentation, contributing to the activation of neuronal innate immunity, which in turn leads to the activation of microglia and to neurodegeneration [83]. Interestingly, BMAA triggered α-synuclein aggregation that progressed in a caudo-rostral manner from the gut through the dorsal motor nucleus of the vagus into the substantia nigra [83]. Thus, it was hypothesized that chronic exposure of the gut to BMAA may lead to a pro-inflammatory intestinal state and to enteric mitochondrial dysfunction that, in susceptible individuals, may initially trigger ENS neurodegeneration that is possibly later translated to the CNS [81]. In addition, BMAA can progress between neurons and it seems to have a tropism for motor neurons or NADPH diaphorase-positive neurons, in which its transport is carried out retrogradely [82]. In light of this mounting evidence, and since BMAA is apparently ubiquitous in water, but not included in the routine protocols for assessment of municipal water quality [76–78], efforts should be directed toward its detection in freshwater and in seawater but also in fish and shellfish that are part of the human diet and in which this neurotoxin is known to accumulate [28].

The First Obstacle: From the Gut to the Rest of the Body

The GI boundary of the “self” is conferred by a single layer of intestinal epithelial cells (IECs) interconnected by tight junctions (TJs) and covered by mucus produced by goblet cells, into which Paneth cells secrete immunoglobulin A (IgA) and antimicrobial peptides. By itself, the intestinal barrier provides mechanical protection and selective permeability to water and ions [84, 85].

It is up to the immune system to discern which stimuli pose a survival threat to the “self”, without triggering unnecessary sterile colitis, in the face of the gut microbial population, as well as the constant passage of food [84, 86]. The shaping of this immune response remains to be understood in a way, but communication between IECs and resident macrophages is thought to be a highlight of this mystery. IEC stimulation by the presence of a healthy microbiome, via Toll-Like 4 receptors, induces interleukin-10 (IL-10) expression. In sequence, IL-10 is known to have anti-inflammatory properties and is essential for maintaining the integrity of the epithelium. On the other hand, IL-10 expression in IECs is enhanced by resident macrophage signaling [87]. In turn, the gut inflammatory status itself is responsible for shaping the phenotype acquired by resident macrophages [88].

CX3CR1+ resident macrophages are a heterogeneous group with high adaptive capacity, which concomitantly express IL-10 and TNF-α, while maintaining an alert resting state [84]. Through transepithelial extensions, these macrophages are sensitive to the intraluminal environment and respond accordingly in a phenotypic manner. In case of threat, they acquire a pro-inflammatory phenotype and communicate with dendritic cells, so that they recruit the action of adaptive immunity, or switch to a phenotype similar to that of a dendritic cell, assuming similar roles [88].

The degree of permeability of the intestinal mucosa depends on different stimuli and components. The microbiome is able to shape this barrier, for example by increasing its permeability through induction of miR-21-5p expression in epithelial cells [89]. This is a curious finding given that inhibition of miR-21-5p in an animal model of ulcerative colitis decreased inflammation as well as apoptosis through the IL-6/STAT3 pathway, demonstrating that it is a possible therapeutic target to be considered against intestinal inflammation [90]. On the other hand, excessive proteolysis results in increased permeability of the intestinal epithelium [91]. Finally, when the integrity of TJs is compromised, an increase in intestinal permeability occurs, inducing local and systemic inflammation [92].

Intestinal epithelial barrier disruption and increased permeability have been demonstrated in several pathologies, notably in the context of IBD [93], but also in PD [94, 95]. In addition, it has been recently shown that IBD patients are at significantly higher risk of developing PD [96]. Moreover, upon administration of Proteus mirabilis and its purified LPS in wt mice, a loss of dopaminergic neurons in the substantia nigra occurs, as well as neuroinflammation and aggregation of α-synuclein in the colon and the brain [97]. These findings are supported by an increase in calprotectin, α1-antitrypsin, and zonulin (the first is a marker of intestinal inflammation, and the latter two are markers of increased intestinal permeability) in fecal material from PD patients [98]. Thus, an increase in intestinal permeability in PD may be associated with a systemic pro-inflammatory state, which ultimately translates into CNS changes compatible with the motor disturbances documented in PD [28, 99, 100]. The serological increase of pro-inflammatory cytokines such as NT-proCNP, IL-6, and TNF-α in PD patients supports this hypothesis [101]. A serological increase in anti-inflammatory cytokines has also been demonstrated, and is thought to be due to compensatory mechanisms [101, 102]. Interestingly, in the plasma of patients, an increase in IL-1β and α-synuclein correlates positively with the severity of motor symptoms [103], indicating a synergistic effect of systemic inflammation, alteration in BBB permeability, and α-synuclein accumulation in PD progression

The multi-organ dysfunction in PD becomes even more evident if we focus on changes in the blood, particularly among leukocytes. Structural alterations of platelets, resulting in increased platelet activation and hypercoagulation, have been found in PD patients [104]. Deficits in the energy function of blood cells have also been found, with cells showing increased vulnerability to oxidative stress, increased glycolysis, mitochondrial dysfunction, and lower glutathione peroxidase activity in early stages of the disease [105, 106]. This demonstrates that the mitochondrial dysfunction described in PD is not limited to neurons [107]. The blood cells of PD patients express α-synuclein, and it is possible to differentiate between controls and patients based on that [108]. NLRP3 inflammasome activation has also been described in the mononuclear blood cells of patients [103].

Several studies have demonstrated changes in the lymphoproliferative response in PD patients at very early stages. Thus, a decrease in CD4+ T cells and a Th1 pro-inflammatory phenotype bias have been described [3, 4, 109]. Indeed, we have recently shown that wt mice treated with a bacterial neurotoxin (BMAA) also have a decrease in blood CD4+ T cells and an increase in plasma pro-inflammatory cytokine levels [83]. Dysfunction in Treg cells, which contributes to a pro-inflammatory microenvironment, may also be attributed to the disease [109]. α-Synuclein-specific T cells have also been identified at early stages of the disease, demonstrating an autoimmune component associated with PD [102].

Finally, the presence of respiratory symptoms in PD has been known since the first description of the disease, however the causes are not well defined [110]. Recent data addressing the gut-lung axis may shed light to this respiratory manifestation in PD. It is well described that gut dysbiosis disrupts tissue and immune homeostasis and is associated the disruption of intestinal–pulmonary cross-talk, which is linked to increased susceptibility to airway diseases and infections. Alterations of the gut–lung axis induced by gut dysbiosis in IBD patients put them at higher risk of developing pulmonary diseases [111].

The Vagus Nerve: The Highway for Alpha-synuclein Propagation

The topography of propagation of the α-synuclein aggregates is highly suggestive of the pathway taken by pathological processes [112]. Increased permeability of the intestinal mucosa is known to correlate positively with the localization of α-synuclein aggregates [113]. And, although elevated α-synuclein levels in the mucosa muscularis or submucosa is not a PD-specific outcome, this is considered a quite suggestive PD histological finding, taking into account that Lewy pathology is detected more often in colon biopsies than in rectal biopsies [114].

There is a hidden bridge between the gut and the CNS that has been unraveling in the last decade. Although the agent that unleashes the α-synuclein aggregation at the gut is still under discussion, the route that these aggregates can take to the CNS is increasingly clear. The vagus nerve has emerged as the main bridge between the gut and the CNS. This perspective is supported by the α-synuclein localization in the dorsal motor nucleus of the vagus in early stages [99] and by the protective effect that full truncal vagotomy confers for PD [115]. Another clue lies in the fact that α-synuclein aggregates can be found not only in the GI-tract but also in other organs dependent on innervation by the vagus nerve. Indeed, α-synuclein aggregates occur in the olfactory bulb, spinal cord, peripheral autonomic ganglia, submandibular gland, cardiac nerves, and ENS in PD patients before it appears in the substantia nigra pars compacta, and before neuronal loss occurs [reviewed in [116]].

Although the physiological role of α-synuclein is partly understood, the same cannot be said of its pathological characteristics, about which much remains to be discovered. Increased expression of α-synuclein may lead to inhibition of exocytosis and thus neurotransmitter release. It also seems to contribute to dopamine synthesis as well as to normal mitochondrial function. Thus, aggregates of α-synuclein also seem to induce mitochondrial dysfunction [83, 117–120].

Studies have shown that gut dysbiosis, particularly in response to microbial products, leads to the aggregation of α-synuclein, which then spreads from cell to cell via exosomes, leading to the appearance of aggregates in the recipient cell. It has been demonstrated, for example, that α-synuclein progresses, in the caudo-rostral direction, in response to exposure to BMAA and that this progression probably occurs through the vagus nerve [59, 83, 117].

The Final Hurdle: Blood-Brain Barrier and Parkinson’s Disease

The CNS is immunologically privileged due to the existence of the BBB and a disruption of this barrier translates into a greater vulnerability of the CNS to the external environment [121]. It is known that there is a positive correlation between the appearance of white matter lesions and systemic inflammation in PD [109, 122]. On the other hand, advanced age is already a risk factor for increased vulnerability of the BBB whose permeability increases naturally with aging [123]. The elderly have fewer endothelial cells available and a decreased capillary density. Also, IgG, IgA, and IgM leakage has been demonstrated in elderly brains. In addition, the involvement of the BBB in PD has been likewise demonstrated [124]. Dysfunction of TJs and transporters such as P-glycoprotein has been shown to be associated with aging and neurodegenerative diseases [124]. LPS administration, in the MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) model of PD, leads to the expression of A1 astrocytes, which are considered neurotoxic, the death of substantia nigra dopaminergic neurons, and increased permeability of the BBB and galectin-3 microglia expression [121]. This microglia galectin-3 positive phenotype is a surveillance state previously identified in the context of cerebral ischemia [125].

Microglia play a dual role in peripheral inflammation. During the acute peripheral inflammation phase, microglia contribute to BBB integrity by migrating towards the vasculature and producing claudin-5, which helps maintain TJs between endothelial cells. But if the inflammation is sustained and becomes chronic and systemic, microglia adopt a phagocytic phenotype that contributes to an increased permeability of this barrier [126].

Following LPS administration, high α-synuclein expression leads to an increase in BBB permeability [127]. Therefore, the translocation of intestinal pathological processes to the CNS is intrinsically linked to the intercellular and interregional migratory capacity of α-synuclein, which is consequently believed to have prion-like behavior in PD [128]. Its relationship with the BBB is also quite peculiar, since its passage occurs bidirectionally and is mediated by the LRP-1 (low-density lipoprotein receptor-related protein 1) receptor. This shows that α-synuclein, formed in the gut in response to an initial local inflammation, is able to reach the CNS by routes other than the vagus nerve [59, 129].

The BBB seems to be especially vulnerable in the striatum. A significant increase in BBB permeability at the posterior commissural putamen in the striatum of postmortem samples from PD patients has been reported [130]. Moreover, upon microbial neurotoxin (BMAA) treatment in wt mice we also found a decrease in BBB permeability [83]. We also consider that the BBB permeability increase may allow blood immune cells to transport α-synuclein into the brain parenchyma, as an alternative route for α-synuclein propagation. Although BBB disruption possibly plays a major role in the progression of PD, there are laboratories that see this vulnerability as an opportunity to optimize the efficacy of currently available therapies [131].

Conclusions

Currently, the PD patient is understood as a genetically susceptible individual, living in a pathology-favorable environment, subjected to various triggers, facilitators, and aggravators of the disease [2]. As a result, there is a wider involvement of the organism than previously speculated. This perspective brings new challenges, but also new opportunities for the future management and treatment of the disease [109]. In view of the absence of a disease-modifying therapy and an ineffective therapy for symptomatic non-motor manifestations, the full contemplation of PD [132] and the PD patient may indicate new paths to follow [6, 131]. This review concludes that there is strong evidence of systemic involvement in PD and a predominant role of the disruption of biological mechanical barriers, such as the intestinal barrier and the BBB, in the progression and development of the pathology [101, 106, 121, 127].

Conflict of interest

All authors claim that there are no conflicts of interest.