Correlated morphological and chemical phenotyping in myenteric type V neurons of porcine ileum

Constipation Caused by Anti-calcitonin Gene-Related Peptide Migraine Therapeutics Explained by Antagonism of Calcitonin Gene-Related Peptide's Motor-Stimulating and Prosecretory Function in the Intestine

The development of small-molecule calcitonin gene-related peptide (CGRP) receptor antagonists (gepants) and of monoclonal antibodies targeting the CGRP system has been a major advance in the management of migraine. In the randomized controlled trials before regulatory approval, the safety of these anti-CGRP migraine therapeutics was considered favorable and to stay within the expected profile. Post-approval real-world surveys reveal, however, constipation to be a major adverse event which may affect more than 50% of patients treated with erenumab (an antibody targeting the CGRP receptor), fremanezumab or galcanezumab (antibodies targeting CGRP). In this review article we address the question whether constipation caused by inhibition of CGRP signaling can be mechanistically deduced from the known pharmacological actions and pathophysiological implications of CGRP in the digestive tract. CGRP in the gut is expressed by two distinct neuronal populations: extrinsic primary afferent nerve fibers and distinct neurons of the intrinsic enteric nervous system. In particular, CGRP is a major messenger of enteric sensory neurons which in response to mucosal stimulation activate both ascending excitatory and descending inhibitory neuronal pathways that enable propulsive (peristaltic) motor activity to take place. In addition, CGRP is able to stimulate ion and water secretion into the intestinal lumen. The motor-stimulating and prosecretory actions of CGRP combine in accelerating intestinal transit, an activity profile that has been confirmed by the ability of CGRP to induce diarrhea in mice, dogs and humans. We therefore conclude that the constipation elicited by antibodies targeting CGRP or its receptor results from interference with the physiological function of CGRP in the small and large intestine in which it contributes to the maintenance of peristaltic motor activity, ion and water secretion and intestinal transit.

Classification of human enteric neurons

Major advances in our understanding of the functional heterogeneity of enteric neurons are driven by the application of newly developed, innovative methods. In contrast to this progress, both animal and human enteric neurons are usually divided into only two morphological subpopulations, “Dogiel type II” neurons (with several long processes) and “Dogiel type I” neurons (with several short processes). This implies no more than the distinction of intrinsic primary afferent from all other enteric neurons. The well-known chemical and functional diversity of enteric neurons is not reflected by this restrictive dichotomy of morphological data. Recent structural investigations of human enteric neurons were performed by different groups which mainly used two methodical approaches, namely detecting the architecture of their processes and target-specific tracing of their axonal courses. Both methods were combined with multiple immunohistochemistry in order to decipher neurochemical codes. This review integrates these morphological and immunohistological data and presents a classification of human enteric neurons which we believe is not yet complete but provides an essential foundation for the further development of human gastrointestinal neuropathology.

Somatostatin as an Active Substance in the Mammalian Enteric Nervous System

Somatostatin (SOM) is an active substance which most commonly occurs in endocrine cells, as well as in the central and peripheral nervous system. One of the parts of the nervous system where the presence of SOM has been confirmed is the enteric nervous system (ENS), located in the wall of the gastrointestinal (GI) tract. It regulates most of the functions of the stomach and intestine and it is characterized by complex organization and a high degree of independence from the central nervous system. SOM has been described in the ENS of numerous mammal species and its main functions in the GI tract are connected with the inhibition of the intestinal motility and secretory activity. Moreover, SOM participates in sensory and pain stimuli conduction, modulation of the release of other neuronal factors, and regulation of blood flow in the intestinal vessels. This peptide is also involved in the pathological processes in the GI tract and is known as an anti-inflammatory agent. This paper, which focuses primarily on the distribution of SOM in the ENS and extrinsic intestinal innervation in various mammalian species, is a review of studies concerning this issue published from 1973 to the present.

Immunohistochemical panel for the diagnosis of Hirschsprung's disease using antibodies to MAP2, calretinin, GLUT1 and S100

AIMS

The diagnosis of Hirschsprung's disease is currently based on the identification of aganglionosis and the presence of an increase in acetylcholinesterase-positive hypertrophic nerve fibres in the large bowel submucosa. However, acetylcholinesterase staining is laborious and requires a skilled technician. The aim of this study was to identify a method for diagnosing Hirschsprung's disease reliably using an immunohistochemical panel of recently proposed markers.

METHODS AND RESULTS

Sixty-nine specimens from 37 patients were evaluated. MAP2 and calretinin antibodies were shown to stain ganglia reliably in the submucosal and myenteric plexuses of normal tissue. By contrast, reduced staining of ganglia was observed in patients with Hirschsprung's disease. Staining for GLUT1 and S100 was used to evaluate the number and thickness of nerve fibres. Gain of GLUT1 and S100 expression was in contrast to the loss of calretinin and MAP2. Hypertrophic submucosal nerve fibres in Hirschsprung's disease develop a perineurium with a ring-like GLUT1 staining pattern similar in size and intensity to that observed in deeper subserosal tissue.

CONCLUSIONS

The diagnosis of Hirschsprung's disease using immunohistochemical panels could be as accurate as with conventional frozen section techniques. In particular, the use of a combination of markers for ganglia and hypertrophic nerve fibres highlighting a prominent perineurium in Hirschsprung's disease could be an alternative method.

Selective survival of calretinin- and vasoactive-intestinal-peptide-containing nerve elements in human chagasic submucosa and mucosa

Chronic Chagas' disease is frequently characterized by massive myenteric neuron loss resulting in megacolon with severely and irreversibly disturbed motility. Here, we focused on two submucosal neuron populations, immunoreactive for calretinin (CALR) or somatostatin (SOM), and their respective mucosal nerve fibres in chagasic megacolon. Surgically removed megacolonic segments of seven chagasic patients were compared with seven age- and region-matched non-chagasic control segments. Evaluation included immunohistochemical triple-staining of cryosections for CALR, SOM and peripherin or for CALR and vasoactive intestinal peptide (VIP) and of submucosal whole-mounts for CALR, SOM and the pan-neuronal marker anti-HuC/D. Submucosal neuron counts in chagasic tissue revealed neuron numbers reduced to 51.2 % of control values. In cryosections, nerve fibre area measurements revealed 8.6 % nerve fibre per mucosal area in control segments, but this value decreased to 1.5 % in megacolonic segments. In both evaluations, a disproportionate decrease of SOM-reactive nerve elements was observed. The proportions of SOM-positive neurons related to the total neuron number declined to 2 % (control 10 %) and the proportion of SOM-reactive mucosal nerve fibres related to the whole mucosal area to 0.014 % (control 1.8 %)in chagasic tissue. The second set of cryosections revealed extensive colocalization of CALR with VIP in both surviving submucosal perikarya and mucosal nerve fibres. We suggest that VIP, a neuroprotective and neuroeffectory peptide typically contained in submucosal neurons, allows both the VIP-containing neurons to endure and the patients to survive by maintaining their mucosal barrier, despite the almost complete loss of colonic motility for decades.

Chemical coding of myenteric neurons with different axonal projection patterns in the porcine ileum

The aim of this study was to perform an immunohistochemical characterization of two different myenteric neuron types of the pig displaying opposite axonal projections. These were type I neurons equipped with lamellar dendrites that projected mainly orally, and type VI neurons that displayed typical axonal dendrites and projected anally. Double immunostainings of longitudinal muscle/myenteric plexus wholemounts from ileal segments of four pigs were performed to visualize neurofilaments (NF) in combination with calcitonin gene-related peptide (CGRP), leu-enkephalin (ENK) and substance P (SP), respectively. Triple immunostainings of wholemounts, using antibodies against neuronal nitric oxide synthase (nNOS) and vasoactive intestinal peptide (VIP) as well as against VIP and galanin (GAL), were performed. We found that 78% of type I neurons immunoreacted to ENK, 21% to CGRP and 24% to SP. The NF-positive type I neurons co-reactive for one of the three above markers displayed mostly frayed outlines of both their somal contours and their broadened dendritic endings. By contrast, most of the non-coreactive type I neurons displayed rather sharply outlined somata and dendrites. No type I neuron immunoreacted to nNOS, VIP or GAL and none of the type VI NF-reactive neurons reacted to CGRP, ENK or SP. All type VI neurons investigated displayed immunoreactivity for nNOS, 92% of which were co-reactive for VIP. Co-reactivity for VIP and GAL was found in 69% of type VI neurons, 21% were positive for VIP but negative for GAL, 9% were negative for both GAL and VIP, and 1% were positive for GAL but negative for VIP. We conclude that there are two subpopulations of morphological type I neurons. One of these displays mainly oral projections and could not be further characterized in this study. The other, which may correspond to neurons innervating the longitudinal and circular muscle layers, were partly immunoreactive for ENK, CGRP and/or SP. Type VI neurons are immunoreactive for nNOS frequently co-localized with VIP and, partly, also GAL. These may be inhibitory motor neurons and are different from VIP/GAL-coreactive minineurons described earlier.

Chemical coding of submucosal type V neurons in porcine ileum

In this study, we attempted to determine the proportion of type V neurons relative to the putative whole neuron population in the two submucosal plexuses of pigs identified by their neurofilament immunoreactivity. The total neuron number was estimated in cuprolinic blue (CB)/anti-Hu protein (HU) costained wholemounts as the sum of the number of CB+/HU+, CB+/HU- and CB-/HU+ neurons. In the external submucosal plexus (ESP), HU labelled 98.6% and CB 97.3% of neurons. In the internal submucosal plexus, HU labelled 98.3%, whereas CB only marked 92.5% of neurons. Furthermore, we investigated the chemical coding of submucosal type V neurons and searched for submucosal, non-type V neurons displaying the same chemical coding as the myenteric type V neurons described earlier, i.e. the colocalization of calcitonin gene-related peptide (CGRP) and somatostatin (SOM). In order to facilitate immunohistochemical detection of neuroactive peptides, ileal segments were pretreated with colchicine prior to fixation. Type V neurons in the ESP occurred either as single cells displaying one or few prominent dendrite(s) or within aggregates displaying a dendritic tangle. In this plexus, type V neurons amounted to between 0.9 and 1.6% of all CB-stained neurons. ESP type V neurons displayed immunoreactivities for choline acetyl transferase (95.8%) and leucine-enkephalin (73.9%). All type V neurons were negative for neuronal nitric oxide synthase. Fifty-eight percent of ESP CGRP/SOM co-immunoreactive neurons displayed type V morphology, whereas 42% were non-type V neurons. Thus, the chemical coding of ESP type V neurons is in principal similar to that of the myenteric type V neurons described earlier. In the internal submucosal plexus, we found no type V neurons. In this plexus, 0.2% of all neurons counterstained with HU displayed CGRP/SOM coreactivity. As had been observed earlier concerning the myenteric type V neurons, ESP type V neurons were also closely apposed by conspicuous accumulations of boutons reactive for the same markers as the neurons themselves. Although we cannot exclude that axons of CGRP/SOM-reactive enteric, non-type V or extrinsic neurons end synaptically on type V neurons, we suggest that the main synaptic input to type V neurons originates from other type V neurons. This presents an argument for an interneuronal role of type V neurons.

Calcitonin Gene-Related Peptide: A Marker for Putative Primary Afferent Neurons in the Pig Small Intestinal Myenteric Plexus?

For years, calcitonin gene-related peptide (CGRP) has been used as a marker peptide for Dogiel type II neurons, putative intrinsic primary afferent neurons, in the pig enteric nervous system. Recently, some studies showed CGRP-positive neurons displaying distinctly different shapes. The aims of this study were to evaluate (1) the proportion of myenteric type II neurons that contain CGRP and (2) the proportion of myenteric CGRP-positive neurons that display type II vs. non-type II morphologies and to conclude if this peptide could be suited as a marker for type II neurons. For this purpose, nine myenteric whole-mounts (each one from duodenum, jejunum, and ileum, respectively, derived from three pigs) were triple-immunostained for CGRP, neurofilaments (NF), and choline acetyl transferase (ChAT). Each whole-mount was evaluated twice. First, 50 NF-stained type II neurons were selected randomly and their coreactivities for CGRP and ChAT were observed. Second, 50 CGRP-positive neurons were located randomly and their NF morphology and ChAT coreactivity were observed. Altogether, 92% of all type II neurons investigated displayed CGRP immunoreactivity, whereas 94.9% of all CGRP-reactive neurons recorded displayed type II morphology. We observed three further shapes of CGRP-positive neurons: 7 type V neurons (all were ChAT-positive; mainly in the ileal whole-mounts), 6 type I-like neurons (all were ChAT-positive), and 14 type III-like neurons (mostly ChAT-negative; mainly in duodenal and jejunal specimens). We conclude that CGRP-antibodies can be used as markers for type II neurons in the pig small intestinal myenteric plexus in quantitative studies but it should be kept in mind that up to one-tenth of CGRP-reactive neurons may be non-type II neurons. In case of single cell evaluation, CGRP-immunoreactivity alone is not suited as a marker. In such cases additional, morphological analysis is necessary.

Extrinsic and intrinsic sources of calcitonin gene-related peptide immunoreactivity in the lamb ileum: a morphometric and neurochemical investigation

To investigate extrinsic origins of calcitonin gene-related peptide immunoreactive (CGRP-IR) nerve fibres in the sheep ileum, the retrograde fluorescent tracer Fast Blue (FB) was injected into the ileum wall. Sections of thoraco-lumbar dorsal root ganglia (DRG) and distal (nodose) vagal ganglia showing FB-labelled neurons were processed for CGRP immunohistochemistry. The distribution of CGRP-IR in fibres and nerve cell bodies in the ileum was also studied. CGRP-IR enteric neurons were morphometrically analysed in myenteric (MP) and submucosal plexuses (SMP) of lambs (2-4 months). Sensory neurons retrogradely labelled with FB were scattered in T5-L4 DRG but most were located at the upper lumbar levels (L1-L3); only a minor component of the extrinsic afferent innervation of the ileum was derived from nodose ganglia. In the DRG, 57% of retrogradely labelled neurons were also CGRP-IR. In cryostat sections, a dense network of CGRP-IR fibres was observed in the lamina propria beneath the epithelium, around the lacteals and lymphatic follicles (Peyer's platches), and along and around enteric blood vessels. Rare CGRP-IR fibres were also present in both muscle layers. Dense pericellular baskets of CGRP-IR fibres were observed around CGRP-negative somata. The only CGRP-IR nerve cells were well-defined Dogiel type II neurons localised in the MP and in the external and internal components of the SMP. CGRP-IR neurons in the myenteric ganglia were significantly larger than those in the submucosal ganglia (mean profile areas: about 1,400 mum(2) for myenteric neurons, 750 mum(2) for submucosal neurons). About 6% of myenteric neurons and 25% of submucosal neurons were CGRP-IR Dogiel type II neurons. The percentages of CGRP-IR neurons that were also tachykinin-IR were about 9% (MP) and 42% (SMP), whereas no CGRP-IR neurons exhibited immunoreactivity for vasoactive intestinal peptide, nitric oxide synthase or tyrosine hydroxylase in either plexus. Thus, CGRP immunoreactivity occurs in the enteric nervous system of the sheep ileum (as in human small intestine and MP of pig ileum) in only one morphologically defined type of neuron, Dogiel type II cells. These are probably intrinsic primary afferent neurons.

Immunohistochemical characterization of putative primary afferent (sensory) myenteric neurons in human small intestine

Pseudouni- or multiaxonal Dogiel type II neurons are the intrinsic primary afferent (sensory) neurons (IPANs) in the guinea pig small intestine. Our aim was to decipher the chemical code of human myenteric type II neurons and to establish their putative vertical projections, i.e., from the myenteric plexus to the submucosa/mucosa. Additionally, we tried to distinguish them chemically from uniaxonal, dendritic type V neurons displaying, at first glance, similar shapes, i.e., smoothly contoured cell bodies with several long processes. Wholemount preparations of the myenteric plexus were immunohistochemically double or triple stained for neurofilaments (NF) and one or two of the following peptides: calbindin, calretinin (CR), calcitonin gene-related peptide (CGRP), somatostatin (SOM) and substance P (SP). In each triple stained wholemount three counts were conducted: (1) NF-positive pseudouni- or multiaxonal (type II) neurons including their reactivities for the above peptides, (2) uniaxonal or NF-negative neurons displaying coreactivities for the above peptides and (3) NF-reactive type V neurons taking into account their reactivities for the above markers. Additionally, type II neurons, which had an axon leading into (disrupted) interconnecting strands towards the submucosa were counted and somal areas of types II and V neurons were measured. The majority of myenteric type II neurons displayed coreactivities for SOM/CR (89.6%), SOM/SP (86.6%) and SP/CR (81.6%), respectively. A minority of type II neurons was positive for CGRP or calbindin. A small population with type III morphology (uniaxonal, long and slender dendrites) displayed the same coreactivities as type II neurons. In contrast, not one single type V neuron was coreactive for SOM/CR, SOM/SP or SP/CR. Out of 627 type II neurons counted in six wholemounts, 84 type II neurons displayed an axon which could be followed into disrupted interconnecting strands indicating a vertical projection pattern. Somal areas of type II neurons were twice as big as those of type V neurons (904+/-210 versus 449+/-110 microm(2)). In conclusion, most human myenteric type II neurons contain SOM, SP and CR. We suggest they are the human IPANs. Type V neurons are both morphologically and chemically distinctly different from type II neurons and may represent descending interneurons. Further studies have to decipher the type-specific chemical code of type II neurons distinguishing them also from type III neurons.

Lessons from the porcine enteric nervous system

The porcine intestinal tract possesses functional and pathological similarities to the human digestive tract and the organization of the porcine enteric nervous system, like that of the human, appears to be more complex than that of commonly investigated guinea-pig intestine. Intrinsic primary afferent neurones appear to differ in the intestines of large and small animals in terms of their chemical coding, distribution over enteric neural networks, electrophysiological behaviour and synaptic properties. Opioid receptors on afferent and motor neurones in the porcine small intestine are predominately of the delta type, whereas those in guinea-pig ileum are mu. Moreover, delta-opioid receptors associated with the myenteric and submucosal plexuses of porcine ileum that, respectively, modulate neurogenic smooth muscle contractions and mucosal ion transport appear to differ in their pharmacological characteristics. These profound interspecies and interregional differences underscore the complexity of the enteric nervous system, and the development of new drugs designed to treat human neurogastrointestinal disorders should be based on the results of investigations in homologous animal models, such as the pig.

Co-expression pattern of neuronal nitric oxide synthase and two variants of choline acetyltransferase in myenteric neurons of porcine ileum

Cholinergic enteric neurons were demonstrated immunohistochemically so far by using antibodies staining the common choline acetyltransferase (cChAT) in neurons of the central nervous system. The results of staining in the enteric nervous system of various species were, however, not satisfactory. We describe here findings obtained with a newly raised antibody against a peripheral variant of choline acetyltransferase (pChAT) in myenteric neurons of the pig small intestine. Triple labelling for pChAT/cChAT/neuronal nitric oxide synthase (nNOS) revealed 19.7% of 1664 neurons (within 40 ganglia) to be immunoreactive exclusively for pChAT whereas 29.6% were positive for cChAT alone and 18.8% were reactive only for nNOS. Colocalization of pChAT and cChAT was found in 22.4%, of pChAT and nNOS in 8.1% and of cChAT and nNOS in 1.4%. All three markers were simultaneously found in only 1 of 1664 neurons. To investigate the presence and possible colocalization of the above markers within morphologically defined neuron types, triple labelling of cChAT or nNOS with pChAT and a neurofilament (NF) antibody pool was applied and the coexpression patterns of pChAT and cChAT as well as of pChAT and nNOS in 120 neurons of each type were recorded. All type I, II, IV and V neurons displayed immunoreactivity either for one or both cholinergic markers. These neuron types were considered to be cholinergic. All type VI neurons, a descending neuron population, were negative for cChAT but positive for nNOS. However, 95% were immunoreactive for both pChAT and nNOS. The physiological significance of the possible co-existence of acetylcholine and nitric oxide within type VI neurons remains to be clarified. It is concluded that the pChAT and cChAT antibodies used here recognize partly different populations of enteric neurons in the pig. Thus, for total immunohistochemical characterization of cholinergic enteric neurons both forms of choline acetyltransferase have to be considered.

Occurrence of ACTH- and enkephalin-like peptides in the developing gut of Dicentrarchus labrax L

We carried out immunohistochemical tests in the developing gut of the sea bass Dicentrarchus labrax to follow the appearance and distribution of the immunoreactivity (IR) to antibodies against POMC-derived, adrenocorticotropic hormone (ACTH), alpha-melanocyte stimulating hormone (alpha-MSH) and beta-endorphin (beta-End), and against two enkephalins, with the aim to study a possible involvement of these molecules in the early neuro-immune-endocrine integration. Our data show that IR to antibodies against some molecules involved in the stress response, such as ACTH and enkephalins, are present in the sea bass gut from an early larval stage (4 days after hatching), before transition to the exotrophic feeding. Moreover, the present study demonstrates for the first time the presence of ACTH-like immunoreactive material in developing gut of a fish. The possible roles of tested molecules are discussed.