Parkinson’s disease (PD) is a multifactorial disorder involving both motor and non-motor symptoms caused by the progressive death of distinct neuronal populations, including dopaminergic neurons in the substantia nigra. The deposition of aggregated α-synuclein protein into Lewy body inclusions is a hallmark of the disorder, and α-synuclein pathology has been found in the enteric nervous system (ENS) of PD patients up to two decades prior to diagnosis. In combination with the high occurrence of gastrointestinal dysfunction in early stages of PD, evidence strongly suggests that some forms of PD may originate in the gut.
1. Introduction
With a continually increasing disease burden of nearly 10 million patients worldwide
[1], Parkinson’s disease (PD) is a devastating neurodegenerative disorder that is characterized by the loss of multiple neuronal populations and the aggregation of α-synuclein protein into intracellular inclusions known as Lewy bodies (LBs)
[2][3][2,3]. The progressive death of dopaminergic neurons in the substantia nigra leads to a depletion of dopamine signaling in the striatum that manifests as classic Parkinsonian symptoms, such as bradykinesia, rigidity, and resting tremors which worsen over time
[4]. In addition to impaired movement, PD patients also often experience non-motor symptoms, including depression, hyposmia, difficulty sleeping
[5], dementia
[6], and gastrointestinal issues
[7]. Autonomic dysfunction can appear decades before the onset of motor signs
[5][7][5,7], highlighting the complexity of the disorder and offering potential early points of intervention. While there is currently no cure for PD, administration of medications such as the dopamine precursor, levodopa, or surgical therapies such as deep brain stimulation can provide symptomatic relief, along with available treatments for non-motor indications.
The high occurrence of gastrointestinal dysfunction in the early stages of PD
[7], coupled with increased recognition of PD gut microbiome dysbiosis
[8] and repeated observations of α-synuclein pathology in the enteric nervous system (ENS) of PD patients
[9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][9,10,11,12,13,14,15,16,17,18,19,20,21,22,23], have collectively inspired great interest in the possibility that the disorder may originate in the gut. The gut-brain axis is a complex bidirectional signaling network by which the brain communicates with the gastrointestinal tract via the ENS that relays both sensory and motor information, bacteria-derived neuroactive molecules, and microbiome-induced cytokine release
[8]. Disruption of the gut-brain axis can lead to a range of disorders from irritable bowel syndrome to functional gastrointestinal disorders, as well as potentially mood disorders and chronic pain
[24]. In PD, patients experience a multitude of clinical symptoms that span the entirety of the gastrointestinal tract, including drooling, swallowing difficulties, delayed gastric emptying, small intestinal bacterial overgrowth, and constipation
[7]. In addition, PD gut microbiota display an enrichment of species in the
Christensenella [25],
Akkermansia [26], and
Lactobacillus [25][27][25,27] genera; depletion of species in
Bacteroides [25][27][25,27],
Clostridium [26][27][26,27], and
Faecalibacterium [28][29][28,29] genera; and decreased levels of short-chain fatty acids
[28]. Given that PD is a neurological disorder, these findings are consistent with gut-brain axis disturbances that may play a role in PD pathogenesis.
α-Synuclein aggregation is a hallmark of PD, yet the relationship of protein aggregation and neurodegeneration is still unclear despite extensive research efforts. Rare mutations in the α-synuclein gene,
SNCA, cause early-onset forms of familial PD
[30][31][32][33][34][35][30,31,32,33,34,35], and in sporadic PD (which accounts for at least 85% of cases), wild-type α-synuclein protein accumulates into LB inclusions and Lewy neurite (LN) axonal deposits
[3][36][3,36]. In 2003, Braak et al.
[37] hypothesized that PD could be staged by a topographical progression of α-synuclein lesions, the first of which appear in the dorsal motor nucleus (DMN) of the vagal nerve in the brainstem of pre-Parkinsonian patients. α-Synuclein pathology then spreads until it reaches the substantia nigra, coinciding with motor symptoms, and ultimately invades the neocortex, when patients may present with cognitive decline
[36][37][36,37].
2. The Human Enteric Nervous System
The ENS is an intricate network of neuronal cell bodies and fibers that perform a wide array of digestive functions, including moving food through the gastrointestinal tract, facilitating nutrient uptake, regulating local blood flow, and supporting the immune system
[38]. While the ENS is able to function independently of the CNS, bidirectional communication between the ENS and CNS serves to relay important information that ultimately affects organismal behavior and gastrointestinal functioning. The gut-brain axis involves the delivery of sensory information to the brain via spinal and vagal afferent pathways, and efferent motor signals to the gut by way of sympathetic and parasympathetic divisions of the autonomic nervous system
[39].
Parasympathetic innervation via the vagal nerve originates in preganglionic neurons of the DMN, which synapse onto postganglionic neurons of the ENS. The nucleus ambiguus in the brainstem also supplies vagal motor efferents specifically to the pharynx and esophagus
[40]. Vagal innervation of the ENS is densest in the upper gastrointestinal tract at the level of the esophagus and stomach and decreases more distally. With little to no vagal input to the distal colon and rectum, these regions are primarily regulated by the sacral parasympathetic nucleus of the spinal cord. The activity of the DMN and nucleus ambiguus can be modulated by sensory information that is relayed from the ENS through vagal afferent pathways to the nucleus of the solitary tract in the brainstem
[40]. While the influence of parasympathetic pathways can result in both enhancement and suppression of gut motility, sympathetic innervation of the digestive tract mainly acts to inhibit motility as a pro-survival reflex that is mediated by prevertebral sympathetic ganglia
[41].
Within the ENS, two main neuronal networks perform the complex integration of all local (intrinsic) neuronal activity, input from extrinsic sympathetic and parasympathetic neurons, as well as cues from the gastrointestinal environment. These networks, known as the myenteric (or Auerbach’s) plexus, and the submucosal (or Meissner’s) plexus, use the integrated input to determine their own sensory, motor, and secretory output. The ENS is located in the gastrointestinal wall, which is comprised of four main layers: the mucosa, submucosa, muscular layer, and adventitia or serosa (
Figure 1). The layer closest to the lumen of the gut, the mucosa, can be further divided into the epithelium, which forms the lining of the mucosa; the lamina propria, which is made up of connective tissue; and the muscularis mucosae, a layer of smooth muscle.
Figure 1. Schematic of the enteric nervous system. Shown are the major layers of the gastrointestinal wall with a simplified representation of neuronal circuitry. The myenteric (Auerbach’s) plexus contains motor circuits that control contraction and relaxation of the muscle layer whereas secretomotor and vasodilator neurons are primarily in the submucosal (Meissner’s) plexus and control local blood flow and secretion. Both plexuses receive extrinsic parasympathetic innervation from the dorsal motor nucleus of the vagus to help regulate gut motility. Sympathetic innervation and vagal sensory afferents are not shown. Figure created using Biorender.com.
Within Auerbach’s and Meissner’s plexuses, there are a multitude of neuronal subtypes (
Figure 1). In Auerbach’s plexus of the muscular layer, local motor circuits function to control gastrointestinal tract motility. These circuits are comprised of excitatory and inhibitory motor neurons that cause contraction and relaxation, respectively, of both the circular and longitudinal muscles. The excitatory motor neurons primarily use acetylcholine as their neurotransmitter, but also use tachykinins and other signaling molecules, while the inhibitory motor neurons use vasoactive intestinal polypeptide (VIP) and nitric oxide in addition to other neurotransmitters
[42]. The excitatory and inhibitory motor neurons receive local input from myenteric interneurons and sensory information from intrinsic primary afferent neurons (IPANs).
In addition to this circuitry, VIP- or acetylcholine-producing secretomotor and vasodilator neurons can be found primarily in Meissner’s plexus, with innervation from IPANs, interneurons, and extrinsic signals from vagal efferent pathways. Unlike the parallel excitatory and inhibitory pathways from the DMN of the vagus to the motor circuits of the ENS, vagal innervation of secretory ENS neurons is primarily excitatory
[40]. Secretomotor neurons control gastrointestinal secretions, while vasodilator neurons synapse onto local arterioles and regulate blood flow. Other ENS neuron types include intestinofugal neurons that synapse onto sympathetic ganglia, and motor neurons that innervate the muscularis mucosae
[42].
3. Lewy Pathology in the Enteric Nervous System in PD
The earliest report of LB pathology in the ENS of PD patients was published in 1984 by Qualman et al.
[9], who documented LBs in Auerbach’s plexus of the colon from one PD patient and the esophagus from another PD patient. The esophageal LBs were associated with ganglion cell degeneration. In 1987, Kupsky et al.
[10] found LBs in the ganglion cells of both Auerbach’s and Meissner’s plexuses of the colon and rectum in a PD patient with megacolon. A series of studies by Wakabayashi and colleagues found that LB pathology in PD is widely distributed throughout the ENS from the upper esophagus to the rectum in both Auerbach’s and Meissner’s plexuses
[11][12][13][11,12,13]. The greatest LB burden was found in the Auerbach’s plexus of the lower esophagus
[11][12][13][11,12,13], and LBs were primarily found in VIP-producing neurons although there was also rare colocalization of pathology with tyrosine hydroxylase-positive processes
[12], which may correspond to noradrenergic sympathetic fibers. No loss of enteric neurons was noted
[11][12][11,12].
Since the brain regions affected in the Braak PD staging scheme are synaptically interconnected, and one of the earliest afflicted regions is the DMN of the vagus connecting the brain to the ENS, Braak and colleagues further investigated α-synuclein in the ENS of autopsy cases staged for Lewy pathology in the CNS
[14]. In total, ten cases were examined, three of whom had been diagnosed with sporadic PD and harbored Stage 4 or 5 brain pathology, two of whom were considered incidental or presymptomatic cases having Stage 2 or 3 PD pathology, and the rest having no evidence of Lewy pathology. α-Synuclein inclusions were found in the gastric wall and the DMN of all confirmed PD and incidental cases but none of the controls, and lesions were observed in both Auerbach’s and Meissner’s plexuses among the PD cases
[14]. The presence of Lewy pathology in the ENS of incidental cases, which harbored early-stage PD brain pathology, is consistent with ENS involvement early in PD.
To date, the largest postmortem study of the ENS in PD patients and controls was conducted using the Brain Bank for Aging Research (BBAR) in Japan
[23]. The lower esophagus was examined from 46 PD patients, which included 38 who were diagnosed with PD dementia and the related disorder, Dementia with Lewy Bodies. A total of 340 controls who did not have Parkinsonism, dementia, or evidence of degeneration in the locus coeruleus or substantia nigra were used for comparison. Remarkably, LBs and LNs were observed in 41/46 PD subjects (89%) compared with 0% of the controls. Since the upper alimentary canal was previously found to have greater LB burden in PD than lower regions
[11][12][13][16][11,12,13,16], the BBAR study of the esophagus can be regarded as a highly relevant sampling of tissue. Across postmortem studies of the ENS, PD patients had more Lewy pathology in the esophagus, stomach, small intestine, large intestine, and rectum, than controls (
Figure 2A), and the total number of PD patients harboring LBs in any segment of the gastrointestinal tract (165/212; 78%) far exceeds that in controls (17/618; 3%) (
Figure 2B). These findings, therefore, support the notion that α-synuclein aggregation in the ENS is a characteristic feature of PD.
Figure 2. ENS Lewy pathology detected in postmortem studies of confirmed PD cases. (
A,
B) Each row represents one study, with the citation indicated. PD cases are to the right of the y-axis, while control cases are to the left of the y-axis. For each patient group, the number of cases with Lewy pathology (red) out of the total number of cases examined in that group is given next to the corresponding bar, with % positive for Lewy pathology given in parentheses. Only cases unique to each study are included. In (
A), graphs (from top to bottom) represent studies of the esophagus, stomach, small intestine, large intestine, and rectum. In (
B), for each study, the total number of PD or control cases with Lewy pathology detected in any gastrointestinal segment is shown. * Indicates an unknown number of subjects from these studies are overlapping. N.I., Not included in the study. N.S., Not shown in the study. Figure created using Biorender.com and GraphPad Prism 9. Refs.
[9][11][12][13][14][15][16][17][18][19][20][22][23][43][9,11,12,13,14,15,16,17,18,19,20,22,23,43].
4. Does α-Synuclein Aggregation Begin in the Gut and Spread to the Brain?
Despite ample evidence of ENS pathology in PD and the hypothesis that it may precede affliction of the CNS
[14], it remains unknown if α-synuclein aggregation follows a prion-like spreading cascade from enteric neurons, through the vagal nerve, and into the brain. α-Synuclein exhibits many prion-like properties, including the ability of aggregated conformations to self-replicate by inducing, or ‘seeding’, the aggregation of physiological α-synuclein
[44][45] [44,45]. In this way, the aggregate load within a cell can become amplified. In addition, similar to prions, α-synuclein can escape from one cell and infect a neighboring cell, as has been shown in culture systems and rodent transplantation studies in which host α-synuclein was detected in grafted cells
[46][47][46,47]. Strikingly, postmortem studies of PD patients who received fetal mesencephalic grafts into the striatum showed that 11–16 years following transplantation, grafted neurons contained LB-like α-synuclein inclusions
[48][49][48,49]. Given the young age of the grafted cells, it is unlikely that pathology developed independently of exposure to α-synuclein and/or other factors from the diseased host tissue, thereby suggesting that cell–cell transmission of α-synuclein in humans is possible.
4.1. Evidence from Incidental LB Disease and Prodromal PD
Incidental LB disease (ILBD) is thought to be a precursor to the development of PD, with subclinical LB pathology in the DMN of the vagus or other brain regions consistent with early Braak PD stages. If ILBD is in fact an incipient form of PD, then evaluation of the ENS for α-synuclein inclusions may shed light on whether protein aggregation in PD is of central or peripheral origin. Indeed, this was the motivation for Bloch et al.
[15], who examined 17 cases of ILBD postmortem and found 14 of them (82%) had α-synuclein pathology in the ENS of the esophagus. Lesions were found mainly in Auerbach’s plexus but also in Meissner’s plexus
[15].
4.2. Evidence from Human Vagotomy Studies
Besides the presence of Lewy pathology in the ENS, another key aspect of the gut-to-brain hypothesis of PD is the transmission of α-synuclein through the vagal nerve. The higher density of ENS pathology in the esophagus and stomach compared to lower regions in PD
[11][12][13][16][11,12,13,16] is consistent with the high concentration of vagal inputs to these areas. PD patient vagotomy studies may, therefore, shed light on the dependence of the disease on intact vagal innervation of the gut. A former treatment for peptic ulcer, vagotomy, is the severing of the vagal nerve either fully (as in truncal vagotomy), solely to the stomach (selective vagotomy), or most selectively to the fundus and body of the stomach only (superselective vagotomy). While only a few studies
[50][51][52][53] [50,51,52,53] have examined the potential relationship of vagal denervation with PD, the findings generally support a protective effect, potentially by reducing the ability of α-synuclein to invade the CNS.
4.3. Prion-Like Transmission of α-Synuclein in Rodents and Monkeys
There is now extensive evidence from animal models supporting the theory of prion-like transmission of α-synuclein in PD and other synucleinopathies. In mice and rats, intracerebral inoculation of recombinant α-synuclein pre-formed fibrils (PFFs), brain tissue from symptomatic α-synuclein transgenic mice, or brain tissue from human synucleinopathy patients, results in widespread deposition of LB-like inclusions that are often associated with neurodegeneration and motor dysfunction
[46][54][55][56][57][58][59][60][54,55,56,57,58,59,46,60]. Moreover, it appears that, regardless of the site of injection, aggregation propagates along synaptic connections and requires the presence of endogenous α-synuclein, similar to prions
[54][55][57][58][59][60][54,55,57,58,59,60]. In macaque monkeys, injection of PD brain tissue containing insoluble LBs into the substantia nigra or striatum caused a loss of striatal terminals followed by dopamine neuron death and diffuse α-synuclein deposits in the remaining nigral cells
[55].
4.4. C. elegans as a Powerful Model System to Study α-Synuclein Pathogenicity in PD
While rodent and non-human primate models provide essential information with regards to how α-synuclein can behave in a mammalian system, complementary animal models that offer a rapidly aging nervous system and high genetic tractability are necessary to accelerate the discovery of disease mechanisms and potential treatments. The small nematode worm,
C. elegans, provides such a platform, having a well-defined nervous system that gives rise to a complex set of behaviors
[61], orthologs for 60–80% of human genes
[62], conserved neurotransmitter signaling
[63], and suitability to rapid large-scale behavioral and phenotypic screening approaches
[61].
C. elegans is a premier model system to study aging and age-related disease, due to its short lifespan (2-4 weeks) and stereotyped age-dependent decline at the tissue, cellular, and molecular levels
[64]. In addition, transgenic expression of human α-synuclein in worms has recapitulated progressive age-dependent neuron death, protein aggregation, and behavioral deficits
[65][66][67][65,66,67], and proven useful for the study of cell autonomous disease mechanisms in dopaminergic neurons
[68][69][70][68,69,70].
In an effort to generate prion-like α-synuclein transmission models initiated in the gut of C. elegans, aour group recently published the neurotoxic effects of feeding worms human α-synuclein PFFs [71]. To our knowledge, this is the first report of α-synuclein PFF exposure in C. elegans. Similar to mouse models, scholarswe found that PFF ingestion in C. elegans promotes dopaminergic neurodegeneration, accelerates the aggregation of host α-synuclein in muscle, and induces an age-dependent motor decline. The development of these new models may serve to complement existing rodent model systems by acting as platforms for high-throughput discovery.
4.5. Alternative Hypotheses of α-Synuclein Spreading in PD
Despite mounting evidence in humans and animal models supporting the gut-to-brain hypothesis of α-synuclein transmission in PD, alternative possibilities have been proposed that fuel ongoing debate. A major criticism of the gut-origin hypothesis of PD is the lack of individuals found to have α-synuclein pathology in the ENS in the absence of pathology in the CNS. It would be expected that if α-synuclein pathology begins in the ENS and spreads to the CNS via the vagal nerve, there should be normal subjects with undiagnosed, prodromal PD that harbor ENS and/or vagal nerve pathology without evidence of lesions in the CNS. To address this issue, Beach and colleagues
[43] conducted an autopsy study of stomach and/or vagal nerve tissue from 111 normal elderly controls that had no CNS pathology, 33 ILBD cases with some CNS pathology, and 53 confirmed PD cases. None of the normal subjects were found to have α-synuclein lesions in the stomach or vagal tissue, whereas 17% and 81% of ILBD and PD cases, respectively, had stomach pathology, and 46% and 89% of ILBD and PD cases, respectively, had vagal pathology. However, the low rate of prodromal PD estimated to exist in the aged population drastically reduces the probability of detection. Rather than arguing specifically against the gut-to-brain hypothesis of PD, the discovery of pathological α-synuclein in the vagus nerve of the majority of PD patients and almost half of ILBD patients can alternatively be interpreted as supporting the vagus nerve acting as a conduit for α-synuclein transmission between the gut and the brain, potentially in either direction.