Recent advances in regenerative medicine to treat enteric neuropathies: use of human cells.

As current options for treating most enteric neuropathies are either non-effective or associated with significant ongoing problems, cell therapy is a potential attractive possibility to treat congenital and acquired neuropathies. Studies using animal models have shown that following transplantation of enteric neural progenitors into the bowel of recipients, the transplanted cells migrate, proliferate, and generate neurons that are electrically active and receive synaptic inputs. Recent studies have transplanted human enteric neural progenitors into the mouse colon and shown engraftment. In this article, we summarize the significance of these recent advances and discuss priorities for future research that might lead to the use of regenerative medicine to treat enteric neuropathies in the clinic.

Effects of tissue age, presence of neurones and endothelin-3 on the ability of enteric neurone precursors to colonize recipient gut: implications for cell-based therapies.

BACKGROUND Most enteric neurones arise from neural crest cells that originate in the post-otic hindbrain, and migrate into and along the developing gastrointestinal tract. There is currently great interest in the possibility of cell therapy to replace diseased or absent enteric neurones in patients with enteric neuropathies, such as Hirschsprung's disease. However, it is unclear whether neural crest stem/progenitor cells will be able to colonize colon (i) in which the mesenchyme has differentiated into distinct layers, (ii) that already contains enteric neurones or (iii) that lacks a gene expressed by the gut mesenchyme, such as endothelin-3 (Et-3). METHODS Co-cultures were used to examine the ability of enteric neural crest-derived cells (ENCCs) from E11.5 mouse gut to colonize a variety of recipient hindguts. KEY RESULTS Enteric neural crest-derived cells migrated and gave rise to neurones in E14.5 and E16.5 aneural colon in which the external muscle layers had differentiated, but they did not migrate as far as in younger colon. There was no evidence of altered ENCC proliferation, cell death or neuronal differentiation in older recipient explants. Enteric neural crest-derived cells failed to enter most recipient E14.5 and E16.5 colon explants already containing enteric neurones, and the few that did showed very limited migration. Finally, ENCCs migrated a shorter distance and a higher proportion expressed the pan-neuronal marker, Hu, in recipient E11.5 Et-3(-/-) colon compared to wild-type recipient colon. CONCLUSIONS & INFERENCES Age and an absence of Et-3 from the recipient gut both significantly reduced but did not prevent ENCC migration, but the presence of neurones almost totally prevented ENCC migration.

Dynamics of neural crest-derived cell migration in the embryonic mouse gut.

Neural crest-derived cells that form the enteric nervous system undergo an extensive migration from the caudal hindbrain to colonize the entire gastrointestinal tract. Mice in which the expression of GFP is under the control of the Ret promoter were used to visualize neural crest-derived cell migration in the embryonic mouse gut in organ culture. Time-lapse imaging revealed that GFP(+) crest-derived cells formed chains that displayed complicated patterns of migration, with sudden and frequent changes in migratory speed and trajectories. Some of the leading cells and their processes formed a scaffold along which later cells migrated. To examine the effect of population size on migratory behavior, a small number of the most caudal GFP(+) cells were isolated from the remainder of the population. The isolated cells migrated slower than cells in large control populations, suggesting that migratory behavior is influenced by cell number and cell-cell contact. Previous studies have shown that neurons differentiate among the migrating cell population, but it is unclear whether they migrate. The phenotype of migrating cells was examined. Migrating cells expressed the neural crest cell marker, Sox10, but not neuronal markers, indicating that the majority of migratory cells observed did not have a neuronal phenotype.

Enteric neural crest-derived cells and neural stem cells: biology and therapeutic potential.

The enteric nervous system arises from two regions of the neural crest; the vagal neural crest which gives rise to the vast majority of enteric neurones throughout the gastrointestinal tract, and the sacral neural crest which contributes a smaller number of cells that are mainly distributed within the hindgut. The migration of vagal neural crest cells into, and along the gut is promoted by GDNF, which is expressed by the gut mesenchyme and is the ligand for the Ret/GFRalpha1 signalling complex present on migrating vagal-derived crest cells. Sacral neural crest cells enter the gut after it has been colonized by vagal neural crest cells, but the molecular control of sacral neural crest cell development has yet to be elucidated. Under the influence of both intrinsic and extrinsic cues, neural crest cells differentiate into glia and different types of enteric neurones at different developmental stages. Recently, the potential for neural stem cells to form an enteric nervous system has been examined, with the ultimate aim of using neural stem cells as a therapeutic strategy for some gut disorders where enteric neurones are reduced or absent.

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.

Increased expression of p75NTR by neural crest-derived cells in vivo during mitosis.

The responses to neurotrophic factors may be modulated by changes in the levels of expression of their receptors. The aim of this study was to determine whether neural precursors in vivo show cell cycle phase-dependent changes in expression of p75NTR and Ret. Preparations of embryonic mouse gut were processed for Ret or p75NTR immunohistochemistry to identify neural crest-derived cells, and cells undergoing mitosis were identified using a fluorescent nucleic acid stain. Neural crest-derived cells undergoing mitosis showed over twice the levels of p75NTR immunoreactivity (pixel intensity measured on a confocal microscope) on their surface to that shown by nonmitotic neural crest cells. However, there was no significant difference between the levels of Ret on mitotic and nonmitotic neural crest cells.

Embryological origin of interstitial cells of Cajal.

Until recently, the embryological origin of the interstitial cells of Cajal (ICC) within the intestine was unclear. An origin from the neural crest or from the mesenchyme was considered possible because ICC possess some characteristics in common with neural crest-derived cells, and some characteristics in common with cells derived from the mesenchyme. Experiments in both mammalian and avian species, in which segments of embryonic gut were removed prior to the arrival of neural crest cells and grown in organ culture, have now shown that ICC do not arise from the neural crest. It appears that ICC and smooth muscle cells arise from common mesenchymal precursor cells. From mid-embryonic stages, ICC precursors express Kit, which is a receptor tyrosine kinase. Both ICC and many smooth muscle cell precursors initially express Kit, and then the cells destined to become smooth muscle cells down-regulate Kit and up-regulate the synthesis of myofilament proteins, whereas cells destined to differentiate into ICC maintain their expression of Kit. Adult mice with mutations that block the activity of Kit have disrupted arrays of ICC, whereas normal ICC are present until shortly after birth in such mice. It, therefore, appears that the Kit signalling pathway in not necessary for the embryonic development of ICC, but rather the post-natal proliferation of ICC.

Chromosomal aberrations in PARP(-/-) mice: genome stabilization in immortalized cells by reintroduction of poly(ADP-ribose) polymerase cDNA.

Depletion of poly(ADP-ribose) polymerase (PARP) increases the frequency of recombination, gene amplification, sister chromatid exchanges, and micronuclei formation in cells exposed to genotoxic agents, implicating PARP in the maintenance of genomic stability. Flow cytometric analysis now has revealed an unstable tetraploid population in immortalized fibroblasts derived from PARP(-/-) mice. Comparative genomic hybridization detected partial chromosomal gains in 4C5-ter, 5F-ter, and 14A1-C1 in PARP(-/-)mice and immortalized PARP(-/-)fibroblasts. Neither the chromosomal gains nor the tetraploid population were apparent in PARP(-/-) cells stably transfected with PARP cDNA [PARP(-/-)(+PARP)], indicating negative selection of cells with these genetic aberrations after reintroduction of PARP cDNA. Although the tumor suppressor p53 was not detectable in PARP(-/-) cells, p53 expression was partially restored in PARP(-/-) (+PARP) cells. Loss of 14D3-ter that encompasses the tumor suppressor gene Rb-1 in PARP(-/-) mice was associated with a reduction in retinoblastoma(Rb) expression; increased expression of the oncogene Jun was correlated with a gain in 4C5-ter that harbors this oncogene. These results further implicate PARP in the maintenance of genomic stability and suggest that altered expression of p53, Rb, and Jun, as well as undoubtedly many other proteins may be a result of genomic instability associated with PARP deficiency.

Expression of Ret-, p75(NTR)-, Phox2a-, Phox2b-, and tyrosine hydroxylase-immunoreactivity by undifferentiated neural crest-derived cells and different classes of enteric neurons in the embryonic mouse gut.

Cells of the enteric nervous system are derived from the neural crest. Probes to a number of molecules identify neural crest-derived cells within the gastrointestinal tract of embryonic mice prior to their differentiation into neurons and glial cells. However, it is unclear whether the different markers are identifying all neural crest-derived cells. In this study the distribution of p75(NTR)-immunoreactivity was compared with that of Ret-, Phox2a-, Phox2b-, and tyrosine hydroxylase (TH) in undifferentiated neural crest-derived cells in the E10.5-E13.5 mouse intestine. Neural crest-derived cells colonise the embryonic mouse gut in a rostral-to-caudal wave between E9.5-E14, and differentiation into enteric neurons also occurs in a rostral-to-caudal wave. Thus, the most caudal neural crest-derived cells within the gut are undifferentiated. These most caudal neural crest-derived cells co-expressed p75(NTR)-, Phox2b- and Ret-immunoreactivity; at E10.5 a sub-population was also TH-positive. The most caudal cells did not show Phox2a-immunoreactivity at any stage. However, a sub-population of cells, which was rostral to the undifferentiated neural crest-derived cells, was Phox2a-positive, and these are likely to be cells beginning to differentiate along a neuronal lineage. The expression of Ret-, Phox2a-, Phox2b- and p75(NTR)-immunoreactivity by two classes of enteric neurons that differentiate prior to birth was also examined. Nitric oxide synthase (NOS) neurons showed Phox2b and Ret immunoreactivity at all ages, and Phox2a and p75(NTR) immunoreactivity only transiently. Calcitonin gene-related peptide (CGRP) neurons showed Phox2b and Ret-immunoreactivity, but not Phox2a immunoreactivity. It is concluded that all undifferentiated neural crest-derived cells initially express Phox2b, Ret, and p75(NTR); a sub-population of these cells also expresses TH transiently. Those cells that are beginning to differentiate along a neuronal lineage maintain their expression of Phox2b and Ret, and they start to express Phox2a, but down-regulate p75(NTR); those cells that differentiate along a glial lineage down-regulate Ret and maintain their expression of p75(NTR). Dev Dyn 1999;216:137-152.

Diabetes does not alter the activity and localisation of nitric oxide synthase in the rat anococcygeus muscle.

Functional studies have revealed diabetes specifically impairs smooth muscle reactivity to nitric oxide in the rat anococcygeus muscle. The present study was conducted to examine whether concurrent prejunctional defects in nitrergic neurotransmission exist in anococcygeus muscles from diabetic rats. Nitric oxide synthase (NOS) activity was assessed by the conversion of 3H-L-arginine to 3H-L-citrulline in homogenates of anococcygeus muscles obtained from 8-week diabetic rats and control rats. NOS activity measured in all tissue samples was dependent on the presence of calcium (2 mM), NADPH (1 mM), tetrahydrobiopterin (100 microM) and flavin adenine dinucleotide (10 microM); however, removal of calmodulin (50 U/ml) did not reduce L-citrulline production. Both N(G)-nitro-L-arginine (100 microM) and N(G)-nitro-L-arginine methyl ester (100 microM) produced significant inhibition of enzyme activity. NOS activity measured in tissue samples from diabetic rats (369.6 +/- 75.9 fmol L-citrulline/mg protein) did not significantly differ from that measured in samples from control rats (423.9 +/- 110.6 fmol L-citrulline/mg protein). However, NOS activity measured after removal of the cofactor tetrahydrobiopterin, was significantly greater in samples from control rats than that from the diabetic group. NOS-immunoreactive and NADPH-diaphorase reactive nerve terminals were found to be sparsely distributed throughout longitudinal sections or whole mounts of anococcygeus muscles from both control and diabetic rats. Quantification of NADPH-diaphorase positive fibres intersecting transects of whole tissue mounts, revealed no significant difference in fibre number between the treatment groups. All NOS-immunoreactive fibres also showed vasoactive-intestinal-polypeptide immunoreactivity. In conclusion, the findings together provide no evidence to indicate that diabetes can induce prejunctional changes in NOS activity or localisation, concurrent with the reported postjunctional impairment in smooth muscle reactivity to nitric oxide, in the rat anococcygeus muscle.

Comparative genomic hybridization analysis of adrenocortical tumors of childhood.

Although several genes have been investigated in adrenal tumorigenesis, the genetic background of adrenocortical tumors (ACT) remains poorly characterized. In southern Brazil, the annual incidence of ACT is unusually high, ranging from 3.4-4.2/million children, compared with a worldwide incidence of 0.3/million children younger than 15 yr. Environmental factors have been implicated because the distribution of these tumors follows a regional, rather than a familial, pattern. However, decreased penetrance of a particular gene defect cannot be excluded. Because linkage or other traditional genetic analyses would not be appropriate to investigate the defect(s) associated with ACT in this population, we used comparative genomic hybridization (CGH) to screen for DNA sequence copy number changes in 9 nonfamilial ACT (6 carcinomas and 3 adenomas) from unrelated patients from this region. Six female (aged 10 months to 6 3/4 yr) and 3 male (1 1/12 to 3 1/4 yr) patients were studied. Three carcinomas were at stage I, 1 was at stage II, and another was at stage III. Two carcinomas had evidence of invasion of the vena cava, and 3 were more than 3 cm in size. All patients underwent surgical excision of their tumors; chemotherapy was administered to cancer patients. Currently, all patients are alive and in remission, with the exception of 1 patient with stage III cancer. High mol wt DNA was extracted from tumor tissue obtained at surgery and frozen at -70 C. This DNA was labeled and used for CGH according to standard procedures. Digital image analysis was performed to detect chromosomal gains or losses. CGH evaluation revealed extensive genetic aberrations in both adenomas and carcinomas; there were no significant differences relative to age, gender, size, or stage of the tumor (P > 0.1). Chromosomes and chromosomal regions 1q, 5p, 5q, 6p, 6q, 8p, 8q, 9q, 10p, 11q, 12q, 13q, 14q, 15q, 16, 18q, 19, and 20q demonstrated gains, whereas 2q, 3, 4, 9p, 11, 13q, 18, 20p, and Xq showed losses. The most striking finding was consistent copy number gain of chromosomal region 9q34 in 8 of the 9 tumors. We conclude that both benign and malignant ACT from southern Brazil show multiple genetic aberrations, including a consistent gain of chromosomal region 9q34. This genomic area may harbor genetic defects that predispose to ACT formation and are shared by the patients who were investigated in this study or are accumulated epigenetically under the influence of a common factor, such as an environmental mutagen.

A single rostrocaudal colonization of the rodent intestine by enteric neuron precursors is revealed by the expression of Phox2b, Ret, and p75 and by explants grown under the kidney capsule or in organ culture.

The colonization of the rodent gastrointestinal tract by enteric neuron precursors is controversial due to the lack of specific cellular markers at early stages. The transcription factor, Phox2b, is expressed by enteric neuron precursors (Pattyn et al. Development 124, 4065-4075, 1997). In this study, we have used an antiserum to Phox2b to characterize in detail the spatiotemporal expression of Phox2b in the gastrointestinal tract of adult mice and embryonic mice and rats. In adult mice, all enteric neurons (labeled with neuron-specific enolase antibodies), and a subpopulation of glial cells (labeled with GFAP antibodies), showed immunoreactivity to Phox2b. In embryonic mice, the appearance of Phox2b-immunoreactive cells was mapped during development of the gastrointestinal tract. At Embryonic Days 9.5-10 (E9.5-10), Phox2b-labeled cells were present only in the stomach, and during subsequent development, labeled cells appeared as a single rostrocaudal wave along the gastrointestinal tract; at E14 Phox2b-labeled cells were present along the entire length of the gastrointestinal tract. Ret and p75 have also been reported to label migratory-stage enteric neuron precursors. A unidirectional, rostral-to-caudal colonization of the gastrointestinal tract of embryonic mice by Ret- and p75-immunoreactive cells was also observed, and the locations of Ret- and p75-positive cells within the gut were very similar to that of Phox2b-positive cells. To verify the location of enteric neuron precursors within the gut, explants from spatiotemporally defined regions of embryonic intestine, 0.3-3 mm long, were grown in the kidney subcapsular space, or in catenary organ culture, and examined for the presence of neurons. The location and sequence of appearance of enteric neuron precursors deduced from the explants grown under the kidney capsule or in organ culture was very similar to that seen with the Phox2b, Ret, and p75 antisera. Previous studies have mapped the rostrocaudal colonization of the rat intestine by enteric neuron precursors using HNK-1 as a marker. In the current study, all HNK-1-labeled cells in the gastrointestinal tract of rat embryos showed immunoreactivity to Phox2b, but HNK-1 cells comprised only a small subpopulation of the Phox2b-labeled cells. In addition, in rats, Phox2b-labeled cells were present in advance of (more caudal to) the most caudal HNK-1-labeled cells by 600-700 microm in the hindgut at E15. We conclude that the neural crest cell population that arises from the vagal level of the neural axis and that populates the stomach, midgut, and hindgut expresses Phox2b, Ret, and p75. In contrast, the sacral-level neural crest cells that populate the hindgut either do not express, or show a delayed expression of, all of the known markers of vagal- and trunk-level neural crest cells.

The identification and chemical coding of cholinergic neurons in the small and large intestine of the mouse.

BACKGROUND: The recent availability of antisera to the vesicular acetylcholine transporter (VAChT) and choline acetyltransferase (ChAT) that demonstrate peripheral cholinergic neurons has made possible the anatomical identification of cholinergic neurons in the enteric nervous system. In this study, we localised cholinergic neurons in the mouse small and large intestine and identified which substances are found colocalised in the cholinergic neurons. METHODS: Immunohistochemical single and double staining techniques were used on whole mount preparations and frozen sections to examine the localisation and chemical coding of cholinergic neurons in the small and large intestine of the mouse. Cholinergic neurons were identified using antisera to ChAT or VAChT. RESULTS: In both the small and large intestine, numerous ChAT-immunoreactive nerve cell bodies were present in the myenteric and submucous ganglia, and ChAT- and VAChT-immunoreactive nerve terminals were abundant in the myenteric and submucous plexuses and the external muscle. Previous studies have identified two major classes of myenteric neurons in the small intestine of the mouse--those containing calretinin plus substance P, and those containing nitric oxide synthase (NOS) plus vasoactive intestinal peptide (VIP). Double-label studies showed that the vast majority of the calretinin/substance P neurons were cholinergic neurons, whereas only a small proportion of the NOS/VIP cells were cholinergic; the noncholinergic NOS/VIP neurons were motor neurons or interneurons, whereas the cholinergic NOS/VIP neurons appeared to be exclusively interneurons. In the small intestine, all of the 5-HT-loaded neurons and a subpopulation of the calbindin neurons were also cholinergic. In the large intestine, there was a pattern of overlaps similar to that found in the small intestine, except that in the large intestine approximately 25% of the calretinin cells were not cholinergic. Only approximately one third of the GABA-loaded neurons in the large intestine were cholinergic. CONCLUSIONS: Large subpopulations of motor neurons and interneurons in the mouse small intestine are cholinergic neurons.

Rescue of the HNF4 --> HNF1alpha pathway in hepatoma variant cells containing human chromosome 12.

Expression of liver-enriched trans-acting hepatocyte nuclear factors 1alpha (HNF1alpha) and 4 (HNF4) is correlated with the hepatic phenotype in cultured rat hepatoma cells. We have used a hepatoma variant cell line, H11, that specifically lacks the HNF4 --> HNF1alpha pathway as a model to understand mechanisms controlling hepatic gene expression. We have introduced randomly marked human chromosomes into H11 cells and have isolated a number of microcell hybrids that have rescued hepatic gene expression, including HNF4, HNF1alpha, and alpha1-antitrypsin. Chromosomal analysis of cell hybrids showed that the rescued hepatic phenotype correlated closely with the presence of human chromosome 12p sequences. Although the gene encoding HNF1alpha is located on chromosome 12q24, its retention was not required to rescue the hepatic phenotype. Thus, we suggest that a locus on human chromosome 12p plays an important role in maintenance of hepatic gene expression through activation of the HNF4 --> HNF1alpha pathway.

Development of nicotinic receptor clusters and innervation accompanying the change in muscle phenotype in the mouse esophagus.

During development, the external muscle of the mouse esophagus undergoes a transdifferentiation from smooth to striated muscle (Patapoutian et al. [1995] Science 270:1818-1821). We now report on the development of the innervation accompanying the change in phenotype of the external muscle of the mouse esophagus. The phenotype of the muscle was monitored by using light and electron microscopy. Nicotinic acetylcholine receptors were localised by using a fluorescence conjugate of alpha-bungarotoxin, and neural elements were localised by using antisera to synaptophysin (a synaptic vesicle protein that was used to label all nerve terminals), the vesicular acetylcholine transporter (VAChT), calcitonin gene-related peptide (CGRP), nitric oxide synthase (NOS), and vasoactive intestinal peptide (VIP). CGRP and VAChT were co-localised in the terminals of vagal motoneurons that innervate the external muscle, and NOS and VIP were co-localised in intrinsic (enteric) neurons, which provide some terminals that are associated with motor endplates. Cells exhibiting striations were first observed in the outer layers of the most rostral regions of the esophagus of embryonic day 15 (E15) mice. Clusters of nicotinic acetylcholine receptors were also first observed at the rostral end of the esophagus of E15 mice, and developed in a rostrocaudal progression that coincided with the appearance of striations within the muscle cells. Synaptophysin-, VAChT- and NOS-immunoreactive nerve terminals were present within the external muscle prior to the formation of receptor clusters, and their appearance did not follow any apparent rostrocaudal sequence. Surprisingly, not all of the receptor clusters at E15 had synaptophysin- and VAChT-immunoreactive nerve terminals closely associated with them. However, from E18 on, almost all of the clusters had synaptophysin-immunoreactive nerve terminals in close association. At late embryonic and early postnatal stages, there was a rostrocaudal gradient in the proportion of receptor clusters having VAChT-immunoreactive nerve terminals associated with them. Nerve terminals associated with nicotinic receptor clusters did not show detectable CGRP-immunoreactivity until one to two weeks after the appearance of synaptophysin- and VAChT-immunoreactivity. The NOS-immunoreactive neurons did not show detectable VIP-immunoreactivity until three days after NOS could be detected. These results show that the appearance of clusters of nicotinic receptors in the external muscle of the esophagus coincides with the expression of a striated muscle phenotype, but not with the presence of ingrowing nerve terminals. However, many of the receptor clusters that were observed first were apparently uninnervated.

The DAPI-3 amacrine cells of the rabbit retina.

In the rabbit retina, the nuclear dye, 4,6,diamidino-2-phenylindole (DAPI), selectively labels a third type of amacrine cell, in addition to the previously characterized type a and type b cholinergic amacrine cells. In this study, these "DAPI-3" amacrine cells have been characterized with respect to their somatic distribution, dendritic morphology, and neurotransmitter content by combining intracellular injection of biotinylated tracers with wholemount immunocytochemistry. There are about 100,000 DAPI-3 amacrine cells in total, accounting for 2% of all amacrine cells in the rabbit retina, and their cell density ranges from about 130 cells/mm2 in far-peripheral retina to 770 cells/mm2 in the visual streak. The thin varicose dendrites of the DAPI-3 amacrine cells form a convoluted dendritic tree that is symmetrically bistratified in S1/S2 and S4 of the inner plexiform layer. Tracer coupling shows that the DAPI-3 amacrine cells have a fivefold dendritic-field overlap in each sublamina, with the gaps in the arborization of each cell being occupied by dendrites from neighboring cells. The DAPI-3 amacrine cells consistently show the strongest glycine immunoreactivity in the rabbit retina and they also accumulate exogenous [3H]-glycine to a high level. By contrast, the AII amacrine cells, which are the best characterized glycinergic cells in the retina, are amongst the most weakly labelled of the glycine-immunopositive amacrine cells. The DAPI-3 amacrine cells costratify narrowly with the cholinergic amacrine cells and the On-Off direction-selective ganglion cells, suggesting that they may play an important role in movement detection.

Induction of nitric oxide synthase in the neointima induced by a periarterial collar in rabbits.

A Silastic collar placed around the common carotid artery of rabbits causes the formation, within 7 days, of an atheroma-like neointima containing cells with the appearance of synthetic-phenotype smooth muscle cells. Using immunohisto-chemistry, we detected the appearance of the cytokine-inducible form of nitric oxide synthase (iNOS, or isoform II) in the neointima of rabbits that had the collar in place for 7 or 14 days. This iNOS immunofluorescence collocalized with anti-smooth muscle myosin in the intima, indicating that it is expressed in smooth muscle cells, and iNOS was also present in a few endothelial cells in collared sections. There was no evidence of iNOS expression in the arterial wall before the neointima was apparent, that is, after only 2 days with the collar. The expression of endothelial NOS (eNOS, or isoform III) immunofluorescence was confined to the endothelial cells in control sections, as it was in collared sections with neointima at 7 and 14 days. Specific immunofluorescence for neuronal NOS (nNOS, or isoform I) was not observed in any sections. Our results suggest that nitric oxide is produced by the inducible isoform of NOS in modified smooth muscle cells of the developing neointima. Activity of iNOS might deprive the endothelium of substrate for nitric oxide production and might explain the compromised endothelium-dependent vasodilatation observed both in this model of atherosclerosis and in human coronary artery disease.

Projections of chemically identified myenteric neurons of the small and large intestine of the mouse.

The projections of different subpopulations of myenteric neurons in the mouse small and large intestine were examined by combining immunohistological techniques with myotomy and myectomy operations. The myotomies were used to examine the polarity of neurons projecting within the myenteric plexus and showed that neurons containing immunoreactivity for nitric oxide synthase (NOS), vasoactive intestinal peptide (VIP), calbindin and 5-HT projected anally, while neurons with substance P (SP)-immunoreactivity projected orally, in both the small and large intestine. Neurons containing neuropeptide Y (NPY)- and calretinin-immunoreactivity projected locally. In the large intestine, GABA-immunoreactive neurons projected both orally and anally, with more axons tending to project anally. Myectomy operations revealed that circular muscle motor neurons containing NOS/VIP/ +/-NPY and calretinin neurons projected anally both in the small and large intestine, while SP-immunoreactive circular muscle motor neurons projected orally. In the large intestine, GABA-IR circular muscle motor neurons projected both orally and anally. This study showed that although some neurons, such as the NOS/VP inhibitory motor neurons and interneurons, SP excitatory motor neurons and 5-HT interneurons had similar projections to those in other species, the projections of other chemical classes of neurons in the mouse intestine differed from those reported in other species.

Origin of interstitial cells of Cajal in the mouse intestine.

The interstitial cells of Cajal (ICC) are found in a number of different locations in the gastrointestinal tract, where they form close associations with both muscle cells and nerve terminals. In this study we examined the embryological origin of ICC in the mouse intestine to determine whether they arise from the neural crest or from the intestinal wall. Segments of intestine were removed from embryonic mice either before or after the arrival of neural crest cells (the precursors of enteric neurons and glial cells) and transplanted under the renal capsule of host (adult) mice and allowed to develop for 18-41 days. In the mouse intestine, antibodies to c-kit protein selectively label ICC at a variety of locations, and antibodies to the NK1 receptor (the receptor for substance P) labels ICC at the level of the deep muscular plexus in the small intestine and a subpopulation of enteric neurons in the large intestine. The presence of neurons in the explants was examined using antisera to neuron-specific enolase, substance P, and calretinin. In segments of small and large intestine explanted after the arrival of neural crest cells, immunoreactive neurons and c-kit- and NK1-immunoreactive ICC were present with a distribution similar to that seen in control tissue at a similar developmental age. In segments of large intestine explanted before the arrival of neural crest cells, neurons were not present; however, c-kit-immunoreactive ICC were present in these aneuronal explants, indicating that ICC do not arise from the neural crest. The source of ICC in mammals is therefore likely to be the mesenchyme of the gut.

Do vasoactive intestinal peptide (VIP)- and nitric oxide synthase-immunoreactive terminals synapse exclusively with VIP cell bodies in the submucous plexus of the guinea-pig ileum?

In the submucous plexus of the guinea-pig ileum, previous light-microscopic studies have revealed that vasoactive intestinal peptide (VIP)-immunoreactive and nitric oxide synthase (NOS)-immunoreactive terminals are found predominantly in association with VIP-immunoreactive nerve cell bodies. In this study, double-label immunohistochemistry at the light-microscopic level demonstrated co-localization of NOS-immunoreactivity and VIP-immunoreactivity in axon terminals in submucous ganglia. About 90% of nerve fibres with NOS-immunoreactivity or VIP-immunoreactivity were immunoreactive for both antigens; only about 10% of labelled varicosities contained only NOS-immunoreactivity or VIP-immunoreactivity. The VIP/NOS varicosities were more often seen in the central parts of the ganglia, close to the VIP-immunoreactive cell bodies. Ultrastructural immunocytochemistry with antibodies to VIP was used to determine if NOS/VIP terminals synapse exclusively with VIP-immunoreactive nerve cell bodies. We examined the targets of VIP-immunoreactive boutons in two submucous ganglia from different animals. Serial ultrathin sections were taken through the ganglia after they had been processed for VIP immunocytochemistry. For each cell body, the number of VIP inputs (synapses and close contacts) was determined. The number of VIP-immunoreactive synapses received by the cell bodies of submucous neurons varied from 0-4 and the number of VIP-immunoreactive close contacts varied from 3-10. There was no significant difference between VIP-immunoreactive nerve cell bodies and non-VIP nerve cell bodies in the number of VIP-immunoreactive synapses and close contacts they received. Thus, the implication from light microscopy that NOS/VIP terminals end predominantly on VIP nerve cells was not vindicated by electron microscopy.