Downregulation of neuronal vasoactive intestinal polypeptide in Parkinson’s disease and chronic constipation

1. INTRODUCTION

Parkinson’s disease (PD) is a chronic neurodegenerative multi-system condition affecting about 1% of the elderly population, though a subset (~ 10%) of patients develops symptoms before 50 years of age.¹ The pathological hallmarks of PD are a progressive degeneration of the dopamine-containing neurons in the substantia nigra pars compacta along with intraneuronal aggregates of eosinophilic inclusions (mainly phosphorylated α-synuclein), i.e. Lewy bodies (LBs) and Lewy neurites (LNs).² Although PD is regarded as a prototypical movement disorder, non-motor manifestations, such as autonomic dysfunctions, particularly those involving the gastrointestinal (GI) tract, are increasingly recognised as being a major aspect of the PD clinical picture.³

Virtually all parkinsonian patients experience GI dysfunctions such as dysphagia, delayed gastric emptying (up to gastroparesis), and severe chronic constipation (CC).^4–5 Specifically, CC is a dominant manifestation in up to 80% of PD patients⁴ and it occurs 2–4 times more frequently in PD compared to age-sex matched non-PD subjects.⁶ Furthermore, the administration of L-Dopa, the most used drug to manage motor symptoms of PD patients, induces a prominent inhibitory action on GI motility, thus worsening the severity of CC.⁷ However, a delayed colonic transit is observed in PD patients also independently of drug treatment.⁸ The severe CC in parkinsonian patients is often unresponsive to first-line treatment (e.g. osmotic laxatives) and it may evolve to severe complications such as megacolon and intestinal pseudo-obstruction.^9–11

The enteric nervous system (ENS) is a prime target of the pathological process of PD as indicated by autopsies of parkinsonian patients revealing the widespread presence of LBs and/or LNs in myenteric and submucosal neurons throughout the GI tract.^12–13 Furthermore, full-thickness biopsies of patients with idiopathic CC undergoing colectomy for treatment-resistant, severe slow transit CC show changes in myenteric and submucosal neurons.¹⁴ Taken together, these features support strongly the role of altered ENS function in patients with PD and CC. The recent demonstration that colonic mucosal biopsies including submucosal tissue with its ganglionated plexus can be obtained during colonoscopy provided an important tool to investigate changes in the ENS at the cellular and molecular level occurring in patients with PD and GI dysfunction.¹⁵

Thus, the present study was designed to investigate and correlate functional and neuronal features in a cohort of PD/CC patients in comparison to non-parkinsonian CC and healthy controls. Our morphological analysis focused on enteric neurons of submucosal plexuses, which play a critical role in controlling secretomotor functions but also participate to motility regulation in the GI tract. Understanding the mechanisms underlying bowel dysfunction in PD patients would ultimately allow for a better knowledge of the management and treatment of CC in these patients.

2. MATERIALS AND METHODS

3. RESULTS

Among PD/CC patients, n= 17 (58%) completed the study; n= 7 (24%) refused colonoscopy (although they underwent TT and AM testing), n= 1 (3.5%) refused both TT and AM assessment (but that patient underwent colonoscopy), n= 3 (10%) refused TT (they had colonoscopy and AM) and n= 1 (3.5%) refused AM (but had colonoscopy and TT). In the group of CC patients, n= 9 (90%) underwent TT and AM, while n= 1 performed only AM.

4. DISCUSSION

The present study demonstrated: a) a downregulation of VIP submucosal neurons and of VIP mRNA as well as of both VIPRs, VIPR1 and VIPR2, in PD/CC and CC patients compared to controls, with the downregulation being more pronounced in PD/CC patients vs. CC patients; and b) an impairment of colonic motor and rectal sensory function in both PD/CC and CC patients. Because of the demanding nature of the study protocol that required many visits and several tests, several patients, particularly in the PD/CC group, did not complete both functional GI assessment and colonoscopy with related histopathological analysis resulting in a smaller number of subjects than the initial enrolled number. The vast majority of PD/CC patients in our study (91%) showed abnormalities of either TT or AM, indicative of sensory-motor abnormalities; 65% had a delayed TT combined with anorectal sensory-motor impairment. Taken together our data indicate a severe functional motor impairment in the colon and rectum of PD/CC patients. In contrast, in the non-parkinsonian CC group, only 33% of patients showed combined slow TT and altered AM patterns, but a high proportion (67%) displayed slow transit. Overall, the two groups, PD/CC and CC, were comparable in terms of slow transit assessed by objective measurements. Our functional data on PD/CC extend previous studies evaluating constipated parkinsonian patients.^23–24 In fact, Sakakibara et al. demonstrated that slow transit constipation was prevalent in PD/CC patients, while anorectal abnormalities were identified only in a small proportion of patients.²⁴ In contrast, Edwards et al. found anorectal manometric abnormalities in 77% of the PD/CC cases and slow transit only in 31%. In that study, the acute administration of apomorphine has been shown to improve dyssynergic defecation symptoms in a subset of patients with PD/CC, thus implying that a dopaminergic dysfunction is also responsible for dyssynergia.²³ The differences between our findings and those reported by Sakakibara et al. and Edwards et al. may be explained by the severity of PD per se and sample size, i.e. the number of patients enrolled in the two mentioned studies (approximately 50% less than the current study). In addition, the heterogeneity of pathogenetic factors contributing to CC in PD (e.g., delayed transit vs. sensory-motor abnormalities) also needs to be taken into consideration, thus different groups of patients might show differences in their GI abnormalities and symptoms. Taken together, both this current study as well as previous data indicate that both a delayed TT and abnormal anorectal manometric patterns contribute to CC in PD patients. The origin of these abnormalities is quite complex and, to date, two mechanisms have been proposed: i) ENS changes mainly in the myenteric plexus associated with LBs and LNs with or without neuronal loss;^{11, 25} and ii) an altered extrinsic nerve input to the lower gut, thereby affecting the colonic and anorectal sensory-motor function.²⁶ In the human colonic submucosa, using specific antibodies to phosphorylated α-synuclein, Lebouvier et al. showed LNs (but not LBs) in about 72% of PD/CC cases. The number of LNs positively correlated with the severity of Rome III defined PD/CC patients, although such correlation was lost when data were adjusted by age.¹⁵ The impact of LNs on altered gut physiology (transit time and sensory-motor function) remains plausible, but not yet proven. Thus, Lewy pathology may be a marker of PD involvement in the ENS, although its actual role in terms of gut dysfunction remains to be assessed. Finally, it should be stressed that the identification of Lewy pathology does not represent a reliable bio-marker of PD as strong evidence points to its demonstration in the nervous system (either central or peripheral) in patients with a variety of neurological disorders,²⁷ as well as ageing people⁴ and even in healthy subjects.²⁸

In our study, we used a recently established methodology of obtaining deep biopsies including the mucosa and submucosa during colonoscopy to obtain a submucosal preparation with the submucosal ganglia.¹⁵ The submucosa contains secretomotor / vasomotor neurons innervating mucosa and vasculature.^29–30 We have found that both PD/CC and CC patients have a significantly reduced expression of VIP, as well its receptors, at both the protein and mRNA level, whereas there were no significant changes in the expression of ChAT, a marker of cholinergic neurons. Both VIP and acetylcholine are critical signaling molecules for the regulation of colonic secretion and fluid movements^30–31 and are largely co-stored in enteric, specifically myenteric, neurons of the human GI tract. In the submucosal plexus of the human colon, however, a considerable number of VIP containing neurons are distinct from those containing acetylcholine.³² Upon release from submucosal neurons, VIP activates the specific constitutive receptor, VIPR1, expressed by the enteric epithelial lining. This results in a cAMP-related HCO3⁻ excretion and movement of Na⁺ and H₂O into the gut lumen, thereby enhancing fluid secretion.³³ The proportion of neurons immunoreactive for VIP was decreased, although not significantly in PD/CC patients vs. CC patients; however, the decrease in VIP mRNA and related receptors, VIPR1 and VIPR2, was significantly higher in PD/CC than in CC patients and controls. The decrease of VIP immunoreactive neurons in the absence of changes in other neuronal populations such as ChAT and in the total number of enteric neurons suggests a downregulation of VIP rather than neuronal loss. This is further evidenced by the concomitant decrease in VIP mRNA levels. The difference in the expression of VIP protein and mRNA may be explained by the difficulty in detecting peptides at the level of cell bodies, due to axonal transport and peptide release.³⁴ The reason(s) for the VIP downregulation remain(s) unknown. Given the crucial regulatory role exerted by VIP in gut physiology, it is reasonable to speculate that this peptide can be selectively targeted by PD.

When comparing the PD/CC and CC patients with slow transit, without taking into consideration the anorectal alteration, our data indicated a trend toward reduced VIP mRNA expression and a significant downregulation of VIPR1 and VIPR2 in slow transit PD/CC vs. CC. Overall, these findings suggest that downregulation of VIP and VIPR expression are more prominent in slow transit PD/CC than slow transit CC. The observation that human submucosal neurons project to the muscle layers in addition to the mucosa³⁵ raises the possibility that a downregulation of VIP, which inhibits GI smooth muscle, alters the normal reflex motility pattern, thus contributing to transit and anorectal abnormalities. Also, VIP containing neurons bear some important neurobiological implications. For example, whether VIP containing enteric neurons are a selective target of α-synuclein aggregation (and thereby degeneration) in PD patients remains to be determined. The identification of LNs in VIPergic neurons of the gut of parkinsonian patients supports this possibility.^36–37

The downregulation of VIP and its related receptors does not appear to be a unique feature underlying PD/CC. In fact, CC patients had a reduced expression of VIP, VIPR1 and VIPR2 compared to controls, albeit to a lower extent than PD/CC patients. It is likely that the VIP reduction shown in the investigated PD/CC patients reflects the long-standing nature of their condition which, in turn, contributes to their VIPR1 and VIPR2 downregulation. In addition, the evidence that both VIP receptors are downregulated suggests a generalized impairment of the VIP signaling. Furthermore, inflammatory changes, which are known to occur in the colonic mucosa of both PD/CC³⁸ and CC patients,³⁹ can affect the expression of VIPRs, in particular the inducible VIPR2 isoform. The involvement of other factors in addition to inflammatory stimuli, in the reduction of VIP, VIPR1 and VIPR2 mRNA expression cannot, however, be excluded. Taken together, our data suggest that a downregulation of VIP and VIPRs may contribute to the pathophysiology of CC in both parkinsonian and non-parkinsonian patients by impairing secretion in addition to inducing colonic and anorectal sensory-motor dysfunction. Secretory abnormalities can affect the composition of the fecal water content thereby leading to hard stools and resultant delayed colonic TT.⁴⁰

In conclusion, our study provides evidence for a downregulation of the VIP system, in addition to sensory-motor colonic and rectal abnormalities in patients with PD/CC. VIP containing neurons represent a major subset of submucosal, mainly secretomotor, neurons thus a decrease in VIP neurons, as well as VIPR1 and VIPR2, might lead to dysfunctional secretory processes. VIP is a major stimulant of fluid secretion, thus a reduced production of VIP is likely to result in reduced fluid secretion, which might account for the development or aggravation of CC in PD patients. Sensory-motor and secretory abnormalities are likely to represent two prominent mechanisms contributing to the pathogenesis of CC particularly in PD patients.

Supplementary Material