Deletion of the mouse RegIIIbeta (Reg2) gene disrupts ciliary neurotrophic factor signaling and delays myelination of mouse cranial motor neurons.

A large number of cytokines and growth factors support the development and subsequent maintenance of postnatal motor neurons. RegIIIbeta, also known as Reg2 in rat and HIP/PAP1 in humans, is a member of a family of growth factors found in many areas of the body and previously shown to play an important role in both the development and regeneration of subsets of motor neurons. It has been suggested that RegIIIbeta expressed by motor neurons is both an obligatory intermediate in the downstream signaling of the leukemia inhibitory factor/ciliary neurotrophic factor (CNTF) family of cytokines, maintaining the integrity of motor neurons during development, as well as a powerful influence on Schwann cell growth during regeneration of the peripheral nerve. Here we report that in mice with a deletion of the RegIIIbeta gene, motor neuron survival was unaffected up to 28 weeks after birth. However, there was no CNTF-mediated rescue of neonatal facial motor neurons after axotomy in KO animals when compared with wild-type. In mice, RegIIIbeta positive motor neurons are concentrated in cranial motor nuclei that are involved in the patterning of swallowing and suckling. We found that suckling was impaired in RegIIIbeta KO mice and correlated this with a significant delay in myelination of the hypoglossal nerve. In summary, we propose that RegIIIbeta has an important role to play in the developmental fine-tuning of neonatal motor behaviors mediating the response to peripherally derived cytokines and growth factors and regulating the myelination of motor axons.

Schwann cell-derived Desert hedgehog controls the development of peripheral nerve sheaths.

We show that Schwann cell-derived Desert hedgehog (Dhh) signals the formation of the connective tissue sheath around peripheral nerves. mRNAs for dhh and its receptor patched (ptc) are expressed in Schwann cells and perineural mesenchyme, respectively. In dhh-/- mice, epineurial collagen is reduced, while the perineurium is thin and disorganized, has patchy basal lamina, and fails to express connexin 43. Perineurial tight junctions are abnormal and allow the passage of proteins and neutrophils. In nerve fibroblasts, Dhh upregulates ptc and hedgehog-interacting protein (hip). These experiments reveal a novel developmental signaling pathway between glia and mesenchymal connective tissue and demonstrate its molecular identity in peripheral nerve. They also show that Schwann cell-derived signals can act as important regulators of nerve development.

Dynamic actin-based epithelial adhesion and cell matching during Drosophila dorsal closure.

BACKGROUND: The adhesion of two epithelial sheets is a fundamental process that occurs throughout embryogenesis and during wound repair. Sealing of the dorsal epidermis along the midline of the Drosophila embryo provides a genetically tractable model to analyse the closure of such holes. Several studies indicate that the actin cytoskeleton plays a critical role in dorsal closure. Although many components of the signalling cascade directing this process have been identified, the precise cell-biological events upon which these signals act remain poorly described. RESULTS: By confocal imaging of living fly embryos expressing green fluorescent protein (GFP)-tagged actin, we found that dorsal closure relies on the activity of dynamic filopodia and lamellipodia that extend from front-row cells to actively zipper the epithelial sheets together. As these epithelial fronts approach one another, we observed long, thin filopodia, apparently 'sampling' cells on the opposing face. When the assembly of these actin-based protrusions was blocked (by interfering with the activities of Cdc42 and Jun N-terminal kinase signalling), the adhesion and fusion of opposing epithelial cells was prevented and their ability to 'sense' correct partners was also blocked, leading to segment misalignment along the midline seam. CONCLUSIONS: Dynamic, actin-based protrusions (filopodia and lamellae) are critical, both in the mechanics of epithelial adhesion during dorsal closure and in the correct 'matching' of opposing cells along the fusion seam.

Abnormalities in neuromuscular junction structure and skeletal muscle function in mice lacking the P2X2 nucleotide receptor.

ATP is co-released in significant quantities with acetylcholine from motor neurons at skeletal neuromuscular junctions (NMJ). However, the role of this neurotransmitter in muscle function remains unclear. The P2X2 ion channel receptor subunit is expressed during development of the skeletal NMJ, but not in adult muscle fibers, although it is re-expressed during muscle fiber regeneration. Using mice deficient for the P2X2 receptor subunit for ATP (P2X2(-/-)), we demonstrate a role for purinergic signaling in NMJ development. Whereas control NMJs were characterized by precise apposition of pre-synaptic motor nerve terminals and post-synaptic junctional folds rich in acetylcholine receptors (AChRs), NMJs in P2X2(-/-) mice were disorganized: misapposition of nerve terminals and post-synaptic AChR expression localization was common; the density of post-synaptic junctional folds was reduced; and there was increased end-plate fragmentation. These changes in NMJ structure were associated with muscle fiber atrophy. In addition there was an increase in the proportion of fast type muscle fibers. These findings demonstrate a role for P2X2 receptor-mediated signaling in NMJ formation and suggest that purinergic signaling may play an as yet largely unrecognized part in synapse formation.

Intrinsic versus extrinsic factors in determining the regeneration of the central processes of rat dorsal root ganglion neurons: the influence of a peripheral nerve graft.

The relative contribution of intrinsic growth capacity versus extrinsic growth-promoting factors in determining the capacity of transected dorsal root axons to regenerate long distances was studied. L4 dorsal root axons regenerating into 4-cm peripheral nerve grafts on transected dorsal roots were counted. Few dorsal root myelinated axons regenerated to the distal end of the grafts by 10 weeks unless the sciatic nerve was also crushed. Regeneration of unmyelinated axons was also increased by peripheral lesions. Crush or transection of the dorsal roots without grafting did not alter GAP-43 mRNA expression in L4 dorsal root ganglion (DRG) cells. Grafting a peripheral nerve onto the cut end of an L4 dorsal root doubled the number of DRG cells expressing high levels of GAP-43 mRNA after a delay of several weeks. Peripheral nerve crush at the time of nerve grafting resulted in a very rapid rise in GAP-43 mRNA expression, which then declined to a steady level, twice that of controls, by 7 weeks. Thus, the rapid increase in the number of DRG neurons expressing high levels of GAP-43 mRNA after peripheral but not central axotomy correlates with the regeneration of central axons through nerve grafts. Because GAP-43 mRNA is slowly upregulated in a subpopulation of sensory neurons in response to exposure of their central axons to a peripheral nerve environment, environments favourable for axonal growth may act by increasing the intrinsic growth response of neurons. Lack of intrinsic growth capacity may contribute to the failure of dorsal root axons to regenerate into the spinal cord.

Blood vessels in liver metastases from both sarcoma and carcinoma lack perivascular innervation and smooth muscle cells.

Hepatic arterial infusion (HAI) chemotherapy as treatment for human colorectal liver metastases is promising, but not entirely satisfactory. Improved drug delivery during HAI may be achieved by manipulating the different control mechanisms of normal versus tumour blood vessels. The peptidergic/aminergic innervation of vessels in normal liver and in two animal models of liver metastasis (Lister Hooded rat with syngeneic MC28 sarcoma; athymic (nude) rat with human HT29 carcinoma) was investigated to assess the suitability of these models for future pharmacological studies. Normal liver and metastases were studied immunohistochemically for the presence of protein gene product 9.5 (PGP), neuropeptide Y (NPY), tyrosine hydroxylase (TH), calcitonin gene-related peptide (CGRP), vasoactive intestinal polypeptide (VIP) and substance P (SP). Perivascular innervation was also examined by transmission electron microscopy. In Lister rat normal livers, perivascular immunoreactive nerve fibres containing PGP, NPY, TH, CGRP and SP were observed around the interlobular blood vessels near the hilum and in the portal tracts. The highest density was seen for PGP, followed in decreasing order, by NPY, TH, CGRP and SP. VIP-immunoreactive nerves were absent. No immunoreactive nerves were observed in the hepatic lobule. In athymic rat livers, the pattern of innervation was similar, except that SP immunoreactivity was more sparse. No perivascular immunoreactive nerves were observed in either MC28 or HT29 tumours. Electron microscopy confirmed the absence of perivascular nerves. Smooth muscle cells were not observed in tumour blood vessel walls. These results are comparable with previous observations on human liver metastases and suggest that the animal models may be suitable for pharmacological studies on vascular manipulation of HAI chemotherapy.

Evaluation of the rat thyroid cell strain FRTL-5 as an in-vitro bioassay system for thyrotrophin.

The cyclic AMP response to bovine TSH was characterized in a strain of rat thyroid follicular cells ( FRTL -5) maintained in continuous culture. Significant stimulation of intracellular cyclic AMP was attained at a TSH dose of 5 muu./ml. Cyclic AMP accumulation continued to increase, at higher TSH doses, with no evidence for attainment of a maximum level at the highest dose tested (5 mu./ml). The precision of TSH measurement was better than 10% over the range 50-5000 muu./ml, comparing favourably with that observed with analogous assays based on human cells, tissue slices or membrane preparations. Using sequential subcultures of FRTL -5 cells, the between-assay variation in response to a single dose of a standard preparation of bovine TSH (53/11; 370 muu./ml) was of the order of 20% which compared favourably with the between-assay variation observed with different cultures of human thyroid cells. Prolongation of the incubation of FRTL -5 cells with TSH to 3 h revealed a progressive increase in the extracellular accumulation of cyclic AMP. Addition of TSH to resting FRTL -5 cells resulted in a stimulation of inorganic iodide uptake with pronounced bell-shaped dose-response characteristics. Thus a maximum uptake was observed at a TSH dose of 100 muu./ml with a significant reduction at higher doses. Acute stimulation of cells with TSH (100 muu./ml) resulted in a rapid and marked alteration in cell morphology, with evidence of cellular retraction and surface ruffling.

Nitric oxide and human umbilical vessels: pharmacological and immunohistochemical studies.

Human umbilical vessels are devoid of nerves and therefore endothelial cells may play an important role in the control of fetoplacental blood flow. In this study we examined the pharmacological effects of various substances, known to produce endothelial-mediated vasodilation in many blood vessels, on the human umbilical artery and vein from legal terminations [mean gestational age, 15 (8-17) weeks; n = 12] and normal term vaginal deliveries [mean gestational age, 39 (38-41) weeks; n = 12]. Acetylcholine, adenosine 5'-triphosphate, the calcium ionophore A23187 and substance P had no effect on raised vascular tone, whereas sodium nitroprusside relaxed 5-hydroxytryptamine (5-HT) preconstricted, umbilical artery and vein from both early and late pregnancy. L-NG-Nitroarginine methyl ester (L-NAME) had no effect on basal tone or on high tone, after it was raised by 5-HT. Localization of nitric oxide synthase [NOS, type I (neuronal)] was examined in the same umbilical vessels using electron immunocytochemistry. No NOS-immunoreactive endothelial cells were observed in the umbilical vessels taken during early pregnancy. However, the percentage of NOS-immunoreactive endothelial cells in umbilical artery and vein from late pregnancy was 3 and 10 per cent, respectively. These results suggest that nitric oxide contributes little, if any, to the local control of umbilical blood flow throughout pregnancy, despite the presence of NOS-immunoreactivity in a subpopulation of endothelial cells in late pregnancy.

Peripheral nerve fibres regenerate through myenteric plexus.

The ability of myenteric glia and neurons to support peripheral nerve regeneration was tested by grafting pieces of muscularis externa 5 mm long from the distal colon of inbred CBA mice adjacent to the proximal stump of cut common peroneal nerves. By two weeks after operation many axons had invaded the plexus and after 3 weeks regeneration common peroneal nerve fibres could be identified in all parts of the plexus throughout the grafts. Some axonal profiles within the plexus appeared to be in the early stages of myelination by enteric glia. Axons surrounded by compact myelin were found at the periphery of ganglia, but the cells involved resembled Schwann cells and could not be positively identified as enteric glia. Profiles similar to those of regenerating axons were only very rarely seen in control experiments in which grafts were placed adjacent to intact common peroneal nerves. It is suggested that the cellular elements of the myenteric plexus can support peripheral nerve regeneration.

Schwann cells support extensive axonal growth into skeletal muscle implants in adult mouse brain.

The ability of striated muscle to support CNS axonal regeneration was tested by grafting pieces of the lateral rectus muscle of the orbit into the hippocampus or neocortex of adult inbred CBA mice. The mice were perfused with fixative 4-5 weeks after operation and ultrathin sections of the grafts examined by electron microscopy. Many axons were present in the grafts and some were traced into the surrounding brain tissue. Most axons were in contact with Schwann cells, or their processes, and both were often associated with basal lamina material left behind by degenerating muscle cells. A few axons and their accompanying Schwann cells were found in contact with the plasma membrane of muscle cells. Fenestrated capillaries were present in the grafts. It is suggested that Schwann cells form the substratum for axonal extension into muscle implants in the CNS, although other factors may contribute to the extensive axonal invasion of the tissue.

Satellite cells surrounding axotomised rat dorsal root ganglion cells increase expression of a GFAP-like protein.

Segments of sciatic nerve, 8-10 mm long, were removed from the left thigh of adult Sprague-Dawley rats. One day to 6 weeks after operation the animals were killed and the 4th and 5th lumbar dorsal root ganglia dissected out on both the operated and the unoperated side. A monoclonal and two polyclonal antibodies to glial fibrillary acidic protein (GFAP) were used to localize GFAP-like molecules in cryostat sections of the ganglia by means of indirect immunofluorescence. Expression of GFAP in satellite cells, demonstrable by the binding of polyclonal antibodies, had increased noticeably by three days post operation and remained at a high level throughout the remaining period of the experiment. Whereas on the unoperated side only about 15% of the neurons in the ganglia were surrounded by GFAP-positive satellite cells, on the operated side about 88% of the neurons were surrounded by GFAP-positive cells. Satellite cells could not be labelled with the monoclonal antibodies to GFAP.

An ultrastructural and immunocytochemical study of thoracic aortic endothelium in aged Sprague-Dawley rats.

The distribution of various vasoactive agents [nitric oxide synthase (NOS)- type I, endothelin-1 (ET-1), arginine-vasopressin (AVP), serotonin (5-HT), histamine and substance P (SP)] in the thoracic aortic endothelium of aged Sprague-Dawley rats was investigated using electron microscopic immunocytochemical methods. The aged thoracic aortic intima was characterized by a large number of leukocytes that adhered to the endothelium, an accumulation of a flake-like precipitate and clusters of leukocytes and smooth muscle cells (SMC) in the subendothelium. Age-associated alterations were also seen in the medial and adventitial layers of the vascular wall. An extensive vasa-vasorum was present in the adventitia from which leukocytes penetrated into perivascular tissue. Some vasa-vasorum showed mast cells adhered to perivascular pericytes. Immunocytochemistry showed about 70% endothelial cells (EC) with positive immunostaining for the brain isoform NOS-type I, compared to 10% in adult mature rats. About 10% of cells showed a positive immunoreaction for ET-1, which is about the same as for the mature adult thoracic aorta (8-9%). Subendothelial macrophages often showed positive immunostaining for antibodies against ET-1. The percentage of EC immunopositive to AVP, 5-HT, and histamine was 16-18, 15 and 12%, respectively compared to 5-8, 7-8 and 6% in mature adult rats. A few cells showed an immunopositive reaction for SP. In summary, the ageing vessel was characterized by a large number of leukocytes adhering to the endothelium and also by the presence of many macrophages and SMC in the subendothelial layer. The percentage of EC in rat thoracic aorta showing NOS immunostaining increased substantially from 10% in mature rats to 70% in aged rats. The percentage of EC immunopositive for AVP, 5-HT and histamine also increased about twofold compared to mature adult rats, while no changes were seen for ET-1.

Increased expression of NOS and ET-1 immunoreactivity in human colorectal metastatic liver tumours is associated with selective depression of constitutive NOS immunoreactivity in vessel endothelium.

The absence of perivascular nerves in tumour vessels suggests that endothelium derived vasoactive substances [nitric oxide (NO) and endothelin-1 (ET-1)] may be key factors in controlling tumour blood flow during tumour growth and metastasis. We have studied the ultrastructural distribution and immunoreactivity of different NO synthase (NOS) isoforms and ET-1 in human colorectal metastatic liver tumour tissues using pre-embedding peroxidase-anti-peroxidase and post-embedding immunoelectron microscopic triple gold labelling techniques. Dramatically lower NOS 1 immunoreactivity was observed in tumour vascular endothelium (1-3% and 15-20% in tumour and normal groups, respectively). As compared to control groups there were significantly less NOS3 immunopositive EC in metastatic tumour vessels (45-50% and 1-3% in normal and tumour groups, respectively). A striking rise in NOS2 was observed in tumour vessel endothelium (< 1% in normal and 65-70% in tumour vessel endothelium). ET-1 immunoreactivity levels were also significantly higher in tumour vessel endothelium (85-90% in tumour, 15-20% in normal group). This increased expression of NOS2 and ET-1 immunoreactivity was accompanied by the increased expression of three NOS isoforms and ET-1 immunoreactivity in liver parenchymal cells. These data suggest that metastatic tumour vessel endothelium is characterized by increased expression of NOS2 and ET-1 and by decreases in NOS1 and NOS3. These characteristics are associated with the overexpression of all three NOS isoforms and ET-1 immunoreactivity in non-vascular cells.

A comprehensive pathological survey of duodenal biopsies from dogs with diet-responsive chronic enteropathy.

BACKGROUND: The detailed pathological phenotype of diet-responsive chronic enteropathy (CE) and its modulation with dietary therapy remain poorly characterized. HYPOTHESIS/OBJECTIVES: Key mucosal lesions of diet-responsive CE resolve with dietary therapy. METHODS: This was a prospective observational study of 20 dogs with diet-responsive CE. Endoscopic duodenal biopsies collected before and 6 weeks after the start of a dietary trial were assessed by means of qualitative and quantitative histopathological, immunohistochemical, and ultrastructural criteria. Control duodenal biopsies were obtained from 10 healthy Beagle dogs on 1 occasion. RESULTS: Compared with control dogs, the CE dogs had higher villus stunting scores and higher overall WSAVA scores, a lower villus height-to-width ratio, and higher lamina propria density of eosinophils. The CE dogs also had ultrastructural lesions of the mitochondria and brush border. In common with other studies in which the disease and control populations are not matched for breed, age, sex, and environment, these comparisons should be interpreted with caution. Comparing biopsies collected at presentation and 6 weeks after starting the dietary trial, mean lamina propria mononuclear cell score and lamina propria densities of eosinophils and mononuclear cells decreased. Dietary therapy also improved ultrastructural lesions of the mitochondria and brush border, eliciting a decrease in intermicrovillar space and an increase in microvillus height. CONCLUSIONS AND CLINICAL IMPORTANCE: In dogs with diet-responsive CE, the remission of clinical signs with dietary therapy is associated with subtle decreases in lamina propria density of eosinophils and mononuclear cells, and resolution of ultrastructural lesions of the enterocyte.

In vitro three-dimensional modelling of human ovarian surface epithelial cells.

OBJECTIVES: Ninety percent of malignant ovarian cancers are epithelial and thought to arise from the ovarian surface epithelium (OSE). We hypothesized that biological characteristics of primary OSE cells would more closely resemble OSE in vivo if established as three-dimensional (3D) cultures. MATERIALS AND METHODS: OSE cells were cultured as multicellular spheroids (MCS) (i) in a rotary cell culture system (RCCS) and (ii) on polyHEMA-coated plastics. The MCSs were examined by electron microscopy and compared to OSE from primary tissues and cells grown in 2D. Annexin V FACS analysis was used to evaluate apoptosis and expression of extracellular matrix (ECM) proteins was analysed by immunohistochemical staining. RESULTS: On polyHEMA-coated plates, OSE spheroids had defined internal architecture. RCCS MCSs had disorganized structure and higher proportion of apoptotic cells than polyHEMA MCSs and the same cells grown in 2D culture. In 2D, widespread expression of AE1/AE3, laminin and vimentin were undetectable by immunohistochemistry, whereas strong expression of these proteins was observed in the same cells grown in 3D culture and in OSE on primary tissues. CONCLUSIONS: Physiological and biological features of OSE cells grown in 3D culture more closely resemble characteristics of OSE cells in vivo than when grown by classical 2D approaches. It is likely that establishing in vitro 3D OSE models will lead to greater understanding of the mechanisms of neoplastic transformation in epithelial ovarian cancers.

Focused ion beam milling and ultramicrotomy of mineralised ivory dentine for analytical transmission electron microscopy.

The use of focused ion beam (FIB) milling for preparation of sections of mineralised ivory dentine for transmission electron microscopy (TEM) is investigated. Ivory dentine is essentially composed of fibrillar type-I collagen and apatite crystals. The aim of this project is to gain a clearer understanding of the relationship between the organic and inorganic components of ivory dentine using analytical TEM, in order to utilise these analytical techniques in the context of common skeletal diseases such as osteoporosis and arthritis. TEM sections were prepared in both single and dual beam FIB instruments, using two standard lift-out techniques, in situ and ex situ. The FIB sections were systematically compared with sections prepared by ultramicrotomy, the traditional preparation route in biological systems, in terms of structural and chemical differences. A clear advantage of FIB milling over ultramicrotomy is that dehydration, embedding and section flotation can be eliminated, so that partial mineral loss due to dissolution is avoided. The characteristic banding of collagen fibrils was clearly seen in FIB milled sections without the need for any chemical staining, as is commonly employed in ultramicrotomy. The FIB milling technique was able to produce high-quality TEM sections of ivory dentine, which are suitable for further investigation using electron energy-loss spectroscopy (EELS) and energy-filtering TEM (EFTEM) to probe the collagen/apatite interface.

Alterations in purinoceptor expression in human long saphenous vein during varicose disease.

OBJECTIVES: Varicose veins are dilated tortuous veins of varying tone. Purinergic signalling is important in the control of tone and in mediating trophic changes in blood vessels. The expression of P2 receptors in control and varicose veins will be examined. METHODS: Purinergic signalling in circular and longitudinal smooth muscle of the human long saphenous vein was studied in control and varicose tissues using immunohistochemistry, organ bath pharmacology and electron microscopy. RESULTS: P2X1, P2Y1, P2Y2, P2Y4 and P2Y6 receptors were present on circular and longitudinal smooth muscle. Purine-mediated circular and longitudinal muscle contractions were weaker in varicose veins. Electron microscopy and immunohistochemistry findings support the view that smooth muscle cells change from the contractile to synthetic phenotype in varicose veins, associated with an upregulation of P2Y1 and P2Y2 receptors and a down regulation of P2X1 receptors. CONCLUSIONS: Down regulation of P2X1 receptors on the smooth muscle of varicose veins is associated with loss of contractile activity. Upregulation of P2Y1 and P2Y2 receptors is associated with a shift from contractile to synthetic and/or proliferative roles. The phenotype change in smooth muscle is associated with weakening of vein walls and may be a causal factor in the development of varicose veins.

Axonal regeneration through arterial grafts.

The left common peroneal nerves of adult inbred mice were severed and allowed to regenerate through the lumina of Y-shaped tubes comprising grafts of abdominal aorta and its bifurcation. Very little regeneration took place within the grafts unless the distal nerve stump was inserted into one limb of the Y-tube. Using syngeneic grafts virtually all the axons regenerating through the lumen grew down the limb of the Y-tube containing the distal nerve. Using non-syngeneic grafts, however, a substantial minority of the axons grew down the 'open' limb of the Y-tube. Axonal regeneration was more successful in non-syngeneic grafts.

Peripheral nerve regeneration through grafts of living and freeze-dried CNS tissue.

The ability of peripheral nerve fibres to regenerate through the central nervous system (CNS) extracellular matrix in the presence of CNS myelin debris was examined using living and freeze-dried optic nerve grafts. The grafts were placed end-to-end with the proximal stumps of severed common peroneal nerves of inbred mice. Within a 4 week period, regenerating peripheral nervous system fibres were found in only two of 14 living grafts. However axons always grew into freeze-dried grafts within one week, despite the presence of CNS myelin debris. The regenerating axons in freeze-dried grafts were accompanied by Schwann cells and were initially found associated with the inner aspect of the glial basal lamina. Although the extracellular matrix of the freeze-dried CNS tissue was subsequently reorganized by invading cells, it seems likely that neither the nature of the CNS extracellular matrix nor the presence of CNS myelin debris had a major inhibitory influence on peripheral nerve regeneration. It is suggested that the presence of living astrocytes covered by a basal lamina at the proximal end of the living optic nerve grafts may inhibit their penetration by regenerating axons.

Schwann cell migration through freeze-killed peripheral nerve grafts without accompanying axons.

Freeze-dried tibial nerve grafts were anastomosed to either the proximal stump or the distal stump of severed tibial nerves in adult inbred Fischer rats. In the case of grafts attached to the proximal stump the tibial nerve was ligated three times, the most distal ligature from the spinal cord being 1 cm from the site of anastomosis. In both types of experiment Schwann cells were, therefore, free to enter the initially acellular grafts without accompanying axons. The grafts were examined 17 days to 12 weeks after operation. Immunofluorescence for S-100 protein was used to evaluate the distance migrated by the Schwann cells and electron microscopy was used to examine the morphology of the cells which invaded the grafts. Schwann cell migration was similar from the proximal and distal stumps. The migrating Schwann cells formed columns which resembled bands of Bungner. They were found mainly, but not exclusively, inside the pre-existing basal lamina tubes left behind by the killed nerve fibres. Some Schwann cells secreted a thin, patchy basal lamina even though they lacked axonal contact. Schwann cell columns became partially compartmentalized by fibroblast processes. Myelin and other debris were removed most rapidly in those parts of the grafts penetrated by large numbers of Schwann cells. The maximum distance the Schwann cells penetrated into the grafts was 8.5 mm and this was achieved by 6 to 8 weeks after operation. This is about half the maximum distance migrated by Schwann cells accompanying regenerating axons through similar grafts.(ABSTRACT TRUNCATED AT 250 WORDS)