Distribution of urocortin in the rat's gastrointestinal tract and its colocalization with tyrosine hydroxylase

Urocortin (Ucn), a newly identified member of the corticotropin-releasing factor (CRF) family, is not only expressed in the brain, but also abundantly present in the peripheral tissues, especially in the gastrointestinal tract (GI) as determined by radioimmuoassay. In order to determine the precise localization of urocorin in the GI, we mapped the distribution of urocortin-like immunoreactivity (ir) in the GI of the rat using an immunofluorescence histochemical technique. Ucn, both in the brain and the peripheral tissues, is involved in the regulatory control of host-defense mechanism during stress. In order to study the possible involvement of the sympathetic system in the expression of GI urocortin in response to stress, we examined the effect of chemical sympathectomy on urocortin-ir and its colocalization with tyrosine hydroxylase (TH). UCn was expressed in all parietal cells of the stomach, myenteric and submucosal plexuses as well as in cells in Lieberkühn crypts of the small and large intestine. Most of the acid secreting parietal cells contained both Ucn and TH. Chemical sympathectomy did not affect Ucn immunoreactivity of parietal cells.     

Urocortin 2 expression in the rat gastrointestinal tract under basal conditions and in chemical colitis

It is becoming increasingly evident that the urocortins (Ucns) and their receptors are involved in the initiation and development of inflammation in the gastrointestinal (GI) tract. There has not been a systematic study of the basal expression of Ucns or their receptors in the GI tract. Here, we examined basal expression of Ucn 2 and its high-affinity receptor, CRF-R2 in the rat GI tract. Ucn 2 mRNA was expressed throughout the small and large intestine. Surprisingly, CRF-R2 mRNA expression was detected in only a subset of GI regions that expressed Ucn 2. Immunohistochemical study showed that both Ucn 2 immuno-reactivity (Ucn 2-IR) and CRF-R2-IR were consistently seen in the neurons of the myenteric plexus and the nerve fibers innervating the circular muscle. By and large, Ucn 2-IR was detected in all layers, including the mucosal and the submucosal layers throughout the GI regions. In contrast, CRF-R2-IR was very low or undetectable in the mucosal layers of all regions examined. The role of Ucn 2 and CRF-R2 was then examined in a rat model of chemically-induced colitis. In the early phase of colitis, Ucn 2 mRNA levels peaked, whereas, in striking contrast, CRF-R2 mRNA expression decreased approximately 2.5-fold below control levels. At the peptide level, Ucn 2-IR was specifically induced in a large population of immune cells that infiltrated the lamina propria and submucosa of the distal colon, whereas CRFR2-IR was detected in only a small fraction of infiltrated immune cells. CRF-R2-IR was dramatically reduced in the neurons of the myenteric plexus. Thus, we show, for the first time, that in the acute phase of inflammation, Ucn 2 levels are increased whereas expression levels of its only identified receptor, CRF-R2, are decreased. This suggests that Ucn 2 exerts its effects only in part via CRF-R2.     

Localization of cannabinoid receptors CB1, CB2, GPR55, and PPARα in the canine gastrointestinal tract

The endocannabinoid system (ECS) is composed of cannabinoid receptors, their endogenous ligands, and the enzymes involved in endocannabinoid turnover. Modulating the activity of the ECS may influence a variety of physiological and pathophysiological processes. A growing body of evidence indicates that activation of cannabinoid receptors by endogenous, plant-derived, or synthetic cannabinoids may exert beneficial effects on gastrointestinal inflammation and visceral pain. The present ex vivo study aimed to investigate immunohistochemically the distribution of cannabinoid receptors CB1, CB2, G protein-coupled receptor 55 (GPR55), and peroxisome proliferation activation receptor alpha (PPARα) in the canine gastrointestinal tract. CB1 receptor immunoreactivity was observed in the lamina propria and epithelial cells. CB2 receptor immunoreactivity was expressed by lamina propria mast cells and immunocytes, blood vessels, and smooth muscle cells. Faint CB2 receptor immunoreactivity was also observed in neurons and glial cells of the submucosal plexus. GPR55 receptor immunoreactivity was expressed by lamina propria macrophages and smooth muscle cells. PPARα receptor immunoreactivity was expressed by blood vessels, smooth muscle cells, and glial cells of the myenteric plexus. Cannabinoid receptors showed a wide distribution in the gastrointestinal tract of the dog. Since cannabinoid receptors have a protective role in inflammatory bowel disease, the present research provides an anatomical basis supporting the therapeutic use of cannabinoid receptor agonists in relieving motility disorders and visceral hypersensitivity in canine acute or chronic enteropathies.     

Urocortins and the regulation of gastrointestinal motor function and visceral pain

Urocortin (Ucn) 1, 2 and 3 are corticotropin-releasing factor (CRF)-related peptides recently characterized in mammals. Urocortin 1 binds with high affinity to CRF type 1 (CRF1) and type 2 (CRF2) receptors while Ucn 2 and Ucn 3 are selective CRF2 ligands. They also have a distinct pattern of distribution, both in the brain and the gastrointestinal tract, compatible with a role mediating, with CRF, the response to stress. In rats and mice, Ucn 1 injected centrally or peripherally inhibited gastric emptying and stimulated colonic propulsive motor function, mimicking the effects of stress or exogenous CRF. Centrally administered Ucn 2 inhibited gastric emptying with similar potency as CRF, while Ucn 1 and Ucn 3 were less potent. However, after peripheral administration, Ucn 1 and Ucn 2 were more potent than CRF. In mice, centrally administered Ucn 1 and 2 stimulated colonic motility with lower potency than CRF, and Ucn 3 was inactive. Studies with selective CRF1 and CRF2 antagonists demonstrated that the gastric-inhibitory and colonic-stimulatory effects of exogenously administered Ucns are mediated through CRF2 and CRF1 receptors, respectively. In addition, Ucn 2 showed visceral anti-nociceptive activity associated with the selective activation of CRF2 receptors. These observations suggest that, acting centrally and peripherally, Ucns might play a significant role in the modulation of gastrointestinal motor and pain responses during stress and stress-related pathophysiological conditions.     

Development of the submucous plexus in the large intestine of the mouse

In the small intestine of both embryonic birds and mammals, neuron precursors aggregrate first at the site of the myenteric plexus, and the submucous plexus develops later. However, in the large intestine of birds, the submucosal region is colonised by neural-crest-derived cells before the myenteric region (Burns and Le Douarin, Development 125:4335-4347, 1998). Using antisera that recognize undifferentiated neural-crest-derived cells (p75NTR) and differentiated neurons (PGP9.5), we examined the colonisation of the murine large intestine by neural-crest-derived cells and the development of the myenteric and submucosal plexuses. At E12.5, when the neural crest cells were migrating through and colonising the hindgut, the hindgut mesenchyme was largely undifferentiated, and a circular muscle layer could not be discerned. Neural-crest-derived cells migrated through, and settled in, the outer half of the mesenchyme. By E14.5, neural-crest-derived cells had colonised the entire hindgut; at this stage the circular muscle layer had started to differentiate. From E14.5 to E16.5, p75NTR- and PGP9.5-positive cells were observed on the serosal side of the circular muscle, in the myenteric region, but not in the submucosal region. Scattered, single neurons were first observed in the submucosal region around E18.5, and groups of neurons forming ganglia were not observed until after birth. The development of the enteric plexuses in the murine large intestine therefore differs from that in the avian large intestine.     

NO-Ergic Innervation of the Intestine of the Snow Sculpin Myoxocephalus brandti

Distribution, localization, and morphological peculiarities of NO-ergic nerve cells in the intestine of the snow sculpin Myoxocephalus brandti (Cottidae family) were studied using histochemical staining for NADPH-diaphorase ( NADPH-d). These cells were shown to be present in the pyloric appendages, middle and posterior parts of the intestine and in its rectal part. The NO-ergic cells are the most numerous in the myenteric plexus and circular muscle layer of all studied parts of the intestine. Single NO-ergic nerve cells are revealed in the submucosal plexus of pyloric appendages, middle and posterior parts of the intestine. No NO-ergic neural cells were found in subserosal and subepithelial plexuses, longitudinal layer of smooth muscle in all studied parts, and in the submucosal plexus of the rectal part of the intestine.     

Development of the enteric nervous system,smooth muscle and interstitial cells of Cajal in the human gastrointestinal tract

The generation of functional neuromuscular activity within the pre-natal gastrointestinal tract requires the coordinated development of enteric neurons and glial cells, concentric layers of smooth muscle and interstitial cells of Cajal (ICC). We investigated the genesis of these different cell types in human embryonic and fetal gut material ranging from weeks 4–14. Neural crest cells (NCC), labelled with antibodies against the neurotrophin receptor p75^NTR, entered the foregut at week 4, and migrated rostrocaudally to reach the terminal hindgut by week 7. Initially, these cells were loosely distributed throughout the gut mesenchyme but later coalesced to form ganglia along a rostrocaudal gradient of maturation; the myenteric plexus developed primarily in the foregut, then in the midgut, and finally in the hindgut. The submucosal plexus formed approximately 2–3 weeks after the myenteric plexus, arising from cells that migrated centripetally through the circular muscle layer from the myenteric region. Smooth muscle differentiation, as evidenced by the expression of -smooth muscle actin, followed NCC colonization of the gut within a few weeks. Gut smooth muscle also matured in a rostrocaudal direction, with a large band of -smooth muscle actin being present in the oesophagus at week 8 and in the hindgut by week 11. Circular muscle developed prior to longitudinal muscle in the intestine and colon. ICC emerged from the developing gut mesenchyme at week 9 to surround and closely appose the myenteric ganglia by week 11. By week 14, the intestine was invested with neural cells, longitudinal, circular and muscularis mucosae muscle layers, and an ICC network, giving the fetal gut a mature appearance.A.S.W. is funded by a PhD studentship awarded to A.J.B. by the Child Health Research Appeal Trust.     

Somatostatin in human enteric nerves

Summary Somatostatin-immunoreactive nerves and endocrine cells were localized by use of immunohistochemistry in human stomach, small and large intestine. The nature of the immunoreactivity in acid extracts of separated layers of intestine was determined with separation by high pressure liquid chromatography followed by detection with radioimmunoassay; authentic somatostatin-14 was found in the external musculature, which contains nerves, and in the submucosa and mucosa, which contain both nerve fibres and endocrine cells.The distribution of somatostatin nerves in the gastric antrum, duodenum, jejunum, ileum, ascending and sigmoid colon, and rectum is described. In the intestine many positive perikarya and fine varicose fibres were seen. Mucosal fibres formed a sub-epithelial plexus and a looser network in the lamina propria; this nerve supply was less dense in the large intestine. Submucous ganglia contained positive perikarya and terminals; many terminals formed pericellular baskets, mainly around non-reactive cells. A small number of nerve fibres were associated with submucosal blood vessels. The innervation of the circular and longitudinal muscle was sparse. Positive nerve terminals were seen in the myenteric plexus, although fewer than in the submucous ganglia; positive perikarya were scarce in myenteric ganglia. Somatostatin-immunoreactive nerves were found in the muscle layers and myenteric plexus of the gastric antrum, but were not detected in the antral mucosa and all layers of the gastric body.The distribution of human enteric somatostatin nerves is compared to that in small laboratory animals, and possible roles for these nerves are discussed.     

Urocortin and corticotropin-releasing factor receptor expression in the human colonic mucosa

Urocortin is a newly identified member of the CRF neuropeptide family. Urocortin has been found to bind with high affinity to CRF receptors. The present study investigated urocortin and CRF receptor expression in human colonic mucosa. Non-pathologic sections of adult colorectal tissues were obtained from patients with colorectal cancer at surgery. Urocortin expression was examined using immunohistochemistry and messenger (m) RNA in situ hybridization. Isolated lamina propria mononuclear cells (LPMC) and epithelial cells were also analyzed by flow cytometry for the characterization of urocortin-positive cells, and by RT-PCR for detection of urocortin, CRF, and CRF receptor mRNA. Urocortin peptide distribution at various stages of human development (n = 35, from 11 weeks of gestation to 6 years of age) was examined by immunohistochemistry using surgical and autopsy specimens. Immunoreactive urocortin and urocortin mRNA were predominantly detected in lamina propria macrophages. Urocortin peptide expression was detected from as early as three months of age, but not before birth or in neonates. Urocortin, CRF receptor type 1 and type 2 mRNA were detected in LPMC. CRF receptor type 2β mRNA, a minor isoform in human tissues, was also detected in LPMC, but at lower levels. Urocortin is locally synthesized in lamina propria macrophages and may act on lamina propria inflammatory cells as an autocrine/paracrine regulator of the mucosal immune system. The appearance of urocortin after birth indicates that the exposure to dietary intake and/or luminal bacteria after birth may contribute to the initiation of urocortin expression in human gastrointestinal tract mucosa.     

Calcitonin gene-related peptide: a sensory and motor neurotransmitter in the feline lower esophageal sphincter

The effect of calcitonin gene-related peptide (CGRP) on the feline lower esophageal sphincter (LES) was determined and correlated with its anatomic distribution as determined by immunohistochemistry. Intraluminal pressures of the esophagus and LES were recorded in anesthetized cats. In separate cats, gastroesophageal junctions were removed after locating the LES manometrically and stained for CGRP-like immunoreactivity (LI) and substance P-LI (SP-LI) by indirect immunohistochemistry. CGRP-LI in the LES was most prominent in large nerve fascicles between the circular and longitudinal muscle layers and only rarely seen in nerve fibers within the circular muscle. The myenteric plexus contained numerous CGRP-LI nerve fibers but cell bodies were not seen. Many CGRP-LI nerve fibers in the myenteric plexus and occasional varicose nerves in the circular muscle demonstrated colocalization with SP-LI. Colocalization of CGRP-LI with SP-LI was also seen in the perivascular nerves of the submucosal and intramural blood vessels and in varicose fibers in the lamina propria of the gastric fundic mucosa. In the esophagus, CGRP-LI nerves extended through the muscularis mucosa and penetrated the squamous epithelium to the lumen. CGRP, given intra-arterially caused a dose-dependent fall in basal LES pressure, with a threshold dose of 10(-8) g/kg (2.63 pmol/kg). At the maximal effective dose, 5 x 10(-6) g/kg (1.31 x 10(3) pmol/kg), CGRP produced 61.0 +/- 6.0% decrease in basal LES pressure. At this dose, mean systemic blood pressure fell by 40.9 +/- 7.8%. The LES relaxation induced by a submaximal dose of CGRP (10(-6) g/kg, 262.7 pmol/kg), 50.3 +/- 3.2% relaxation was partially inhibited by tetrodotoxin (26.9 +/- 10.8% relaxation, P less than 0.025). The inhibitory effect of CGRP was not affected by cervical vagotomy, hexamethonium, atropine, propranolol, or naloxone. The LES contractile response to the D90 of SP (5 x 10(-8) g/kg, 37.1 pmol/kg) was not altered by CGRP 10(-8) or 10(-6) g/kg and the CGRP relaxation effect was not altered by the threshold dose of substance P (5 X 10(-9) g/kg, 3.71 pmol/kg). CONCLUSIONS: (1) CGRP-LI is present at the feline LES and is primarily seen in large nerve fascicles which pass from the intermuscular plane and through the circular muscle layer to the submucosa and in mucosal nerves. (2) CGRP colocalizes with SP-LI in some varicose nerve fibers of the circular muscle of the esophagus, LES and fundus, in perivascular nerves of the submucosal and intramucosal blood vessels, and in nerves of the lamina propria of the gastric fundus. (3) The luminal penetration of CGRP-LI nerves in the squamous mucosa of the esophagus suggests a sensory func     

Membrane potential gradient is carbon monoxide-dependent in mouse and human small intestine

The aims of this study were to quantify the change in resting membrane potential (RMP) across the thickness of the circular muscle layer in the mouse and human small intestine and to determine whether the gradient in RMP is dependent on the endogenous production of carbon monoxide (CO). Conventional sharp glass microelectrodes were used to record the RMPs of circular smooth muscle cells at different depths in the human small intestine and in wild-type, HO2-KO, and W/W(V) mutant mouse small intestine. In the wild-type mouse and human intestine, the RMP of circular smooth muscle cells near the myenteric plexus was -65.3 +/- 2 mV and -58.4 +/- 2 mV, respectively, and -60.1 +/- 2 mV and -49.1 +/- 1 mV, respectively, in circular smooth muscle cells at the submucosal border. Oxyhemoglobin (20 microM), a trapping agent for CO, and chromium mesoporphyrin IX, an inhibitor of heme oxygenase, abolished the transwall gradient. The RMP gradients in mouse and human small intestine were not altered by N(G)-nitro-l-arginine (200 microM). No transwall RMP gradient was found in HO2-KO mice and W/W(V) mutant mice. TTX (1 microM) and 1H-[1,2,4-]oxadiazolo[4,3-a]quinoxalin-1-one (10 microM) had no effect on the RMP gradient. These data suggest that the gradient in RMP across the thickness of the circular muscle layer of mouse and human small intestine is CO dependent.     

Divergent fate and origin of neurosphere-like bodies from different layers of the gut

Enteric neural stem cells (ENSCs) are a population of neural crest-derived multipotent stem cells present in postnatal gut that may play an important role in regeneration of the enteric nervous system. In most studies, these cells have been isolated from the layer of the gut containing the myenteric plexus. However, a recent report demonstrated that neurosphere-like bodies (NLBs) containing ENSCs could be isolated from mucosal biopsy specimens from children, suggesting that ENSCs are present in multiple layers of the gut. The aim of our study was to assess whether NLBs isolated from layers of gut containing either myenteric or submucosal plexus are equivalent. We divided the mouse small intestine into two layers, one containing myenteric plexus and the other submucosal plexus, and assessed for NLB formation. Differences in NLB density, proliferation, apoptosis, neural crest origin, and phenotype were investigated. NLBs isolated from the myenteric plexus layer were present at a higher density and demonstrated greater proliferation, lower apoptosis, and higher expression of nestin, p75, Sox10, and Ret than those from submucosal plexus. Additionally, they contained a higher percentage of neural crest-derived cells (99.4 ± 1.5 vs. 0.7 ± 1.19% of Wnt1-cre:tdTomato cells; P < 0.0001) and produced more neurons and glial cells than those from submucosal plexus. NLBs from the submucosal plexus layer expressed higher CD34 and produced more smooth muscle-like cells. NLBs from the myenteric plexus layer contain more neural crest-derived ENSCs while those from submucosal plexus appear more heterogeneous, likely containing a population of mesenchymal stem cells.     

Corticotropin releasing factor—Distribution in rat intestine and role in neuroprotection

Aims of the present study were to describe the distribution of corticotropin releasing factor (CRF) immunoreactivity in rat small and large intestines, to quantify the percentage of CRF-immunoreactive (CRF-IR) enteric neurons, to reveal possible CRF immunoreactivity in cultured myenteric neurons from rat ileum and to examine if additions of CRF, urocortin 1 (Ucn1), CRF antagonist or vasoactive intestinal peptide (VIP) affect neuronal survival in vitro. Co-localization of CRF- and VIP-immunoreactivity was examined, as well as a possible interplay between CRF and VIP in neuroprotection. Further we wanted to elucidate if mast cells affect neuronal survival via CRF signaling.Networks of CRF-containing nerve cell bodies and fibers were detected in rat intestine. CRF-IR neurons contained to a high degree also VIP. A low number of cultured myenteric neurons was CRF-IR. CRF, Ucn1 or CRF-antagonist did not promote neuronal survival of cultured myenteric neurons, while VIP significantly enhanced neuronal survival. Simultaneous presence of CRF attenuated the VIP mediated increase in neuronal survival. Co-culturing neurons and mast cells resulted in a marked reduction in neuronal survival, not executed via CRF signaling pathways. Conclusion: CRF is present in enteric neurons and counteracts the neuroprotective effect of VIP in vitro.     

Calretinin-immunoreactive neurons and their projections in the guinea-pig colon

The distribution of nerve cells and fibres with immunoreactivity for the calcium-binding protein, calretinin, was studied in the distal colon of the guinea-pig. The projections of the neurons were determined by examining the consequences of lesioning the myenteric plexus. Calretinin-immunoreactive neurons comprised 17% of myenteric nerve cells and 6% of submucous nerve cells. Numerous calretinin-immunoreactive nerve fibres were located in the longitudinal and circular muscle, and within the ganglia of the myenteric and submucous plexuses. Occasional fibres were found in the muscularis mucosae, but they were very rare in the lamina propria of the mucosa. Lesion studies revealed that myenteric neurons innervated the underlying circular muscle and provided both ascending and descending processes that gave rise to varicose branches in myenteric ganglia. Calretinin-immunoreactive fibres also projected to the tertiary component of the myenteric plexus, and are therefore likely to be motor neurons to the longitudinal muscle. Varicose fibres that supplied the submucous ganglia appear to arise from submucous nerve cells. Arterioles of the submucous plexus were sparsely innervated by calretinin-immunoreactive fibres. The submucous plexus was the principal source of immunoreactive nerve fibres in the muscularis mucosae. This work shows that calretinin-IR reveals different neuronal populations in the large intestine to those previously reported in the small intestine.     

Neuroplasticity in the smooth muscle of the myenterically and extrinsically denervated rat jejunum

The objective of this study was to examine the effects of two different denervation procedures on the distribution of nerve fibers and neurotransmitter levels in the rat jejunum. Extrinsic nerves were eliminated by crushing the mesenteric pedicle to a segment of jejunum. The myenteric plexus and extrinsic nerves were eliminated by serosal application of the cationic surfactant benzyldimethyltetradecylammonium chloride (BAC). The effects of these two denervation procedures were evaluated at 15 and 45 days. The level of norepinephrine in whole segments of jejunum was initially reduced by more than 76% after both denervation procedures, but by 45 days the level of norepinephrine was the same as in control tissue. Tyrosine hydroxylase (nor-adrenergic nerve marker) immunostaining was absent at 15 days, but returned by 45 days. However, the pattern of noradrenergic innervating axons was altered in the segment deprived of myenteric neurons. Immunohistochemical studies showed protein gene product 9.5 (PGP 9.5)-immunoreactive fibers in whole-mount preparations of the circular smooth muscle in the absence of the myenteric plexus and extrinsic nerves. At 45 days, the number of nerve fibers in the circular smooth muscle increased. Vasoactive intestinal polypeptide (VIP)-immunoreactive fibers, a subset of the PGP 9.5 nerve fibers, were present in the circular smooth muscle at both time points examined. Choline acetyltransferase (CAT) activity and VIP and leucine enkephalin levels were measured in separated smooth muscle and submucosa-musosal layers of the denervated jejunum. VIP and leucine-enkephalin levels were no different from control in tissue that was extrinsically denervated alone. However, the levels of these peptides were elevated two-fold in the smooth muscle 15 and 45 days after myenteric and extrinsic denervation. In the submucosa-mucosa, VIP and leucine enkephalin levels also were elevated two-fold at 15 days, but comparable to control at 45 days. CAT activity was equal to control in the smooth muscle but elevated two-fold in the submucosa-mucosa at both times. These results provide evidence for innervation of the circular smooth muscle by the submucosal plexus. Moreover, these nerve fibers originating from the submucosal plexus proliferate in the absence of the myenteric plexus. Furthermore, the myenteric neurons appear to be essential for normal innervation of the smooth muscle by the sympathetic nerve fibers. It is speculated that the sprouting of the submucosal plexus induced by myenteric plexus ablation is mediated by increased production of trophic factors in the hyperplastic smooth muscle.     

What neurons hide behind calretinin immunoreactivity in the human gut?

Calretinin (CALR) is often used as an immunohistochemical marker for the histopathological diagnosis of human intestinal neuropathies. However, little is known about its distribution pattern with respect to specific human enteric neuron types. Prior studies revealed CALR in both myenteric and submucosal neurons, most of which colabel with choline acetyl transferase (ChAT). Here, we specified the chemical code of CALR-positive neurons in small and large intestinal wholemounts in a series of 28 patients. Besides other markers, we evaluated the labeling pattern of CALR in combination with vasoactive intestinal peptide (VIP). In colonic submucosa, CALR and VIP were almost completely colocalized in about three-quarters of all submucosal neurons. In the small intestinal submucosa, both the colocalization rate of CALR and VIP as well as the proportion of these neurons were lower (about one-third). In the myenteric plexus of both small intestine and colon, CALR amounted to 11 and 10 %, respectively, whereas VIP to 5 and 4 % of the whole neuron population, respectively. Colocalization of both markers was found in only 2 and 3 % of myenteric neurons, respectively. In section specimens, nerve fibers coreactive for CALR and VIP were found in the mucosa but not in the muscle coat. Summarizing the present and earlier results, CALR was found in at least one submucosal and two myenteric neuron populations. Submucosal CALR+/VIP+/ChAT± neurons innervate mucosal structures. Furthermore, CALR immunoreactivity in the myenteric plexus was observed in morphological type II (supposed primary afferent) and spiny type I (supposed inter- or motor-) neurons.