Spatial distribution of mast cells in chronic venous leg ulcers.

Chronic venous leg ulcers (CVUs) show chronic inflammation but different pathological changes occur in different parts of the ulcer. There is a lack of re-epithelialisation and defective matrix deposition in the ulcer base but epidermal hyperproliferation and increased matrix deposition in the surrounding skin. The role of mast cells in wound healing, inflammation, fibrosis and epidermal hyperproliferation has been extensively studied but less is known about their role in CVUs. In the present study, we investigated the distribution of mast cells in CVUs with specific consideration of the differences between the ulcer base and the skin surrounding the ulcer. Both histochemical and immunohistological methods were used to detect the mast cell marker tryptase in frozen sections of CVU biopsies. Mast cells were counted in the dermis of normal skin, in the ulcer base and in the skin surrounding the ulcer. Double immunofluorescence staining was used to study the location of mast cells in relation to blood vessels. In normal skin few mast cells were seen in the dermis but none in the epidermis. However in CVUs there was a significant increase in intact and degranulated mast cells in the surrounding skin and ulcer edge (184 per field, p<0.003) of CVUs and a significant reduction in the ulcer base (20.5 per field p<0.05) in comparison to normal skin (61 per field). In CVUs mast cells showed a characteristic location near the epithelial basement membrane whilst mast cell granules and phantom cells (mast cells devoid of granules) were predominantly seen in the epidermis. In the dermis, mast cells were seen associated with blood vessels. The marked increase in mast cells in the surrounding skin of CVUs and depletion of mast cells in the ulcer base could implicate mast cell mediators in the pathological changes in CVUs particularly in the epidermal and vascular changes occurring in the surrounding skin.

Expression of cyclooxygenase isoforms in normal human skin and chronic venous ulcers.

Chronic venous ulcers are an example of abnormal wound healing showing chronic inflammation which together with the underlying vascular pathology results in delayed healing. Prostaglandins are among the most important mediators of inflammation. They have proinflammatory effects, predominantly by affecting the vasculature. Cyclooxygenase (COX) is the rate-limiting enzyme in prostanoid synthesis. It is present in two isoforms: COX-1 (constitutive cyclooxygenase) which is produced in the body to maintain normal haemostatic functions, and COX-2 (inducible cyclooxygenase), which is induced during inflammation in response to cytokines. Using immunoenzymatic labelling and western blot analysis, this study has shown that both COX-1 and COX-2 were up-regulated in chronic venous leg ulcers by comparison with normal human skin. De novo appearance of COX-2 in chronic venous ulcers was demonstrated, which is not seen in normal human skin. The main cellular sources of both COX isoforms are macrophages and endothelial cells. COX-2 is also produced by mast cells and fibroblasts. A COX radioimmunoassay showed up-regulation of COX activity in chronic venous ulcers compared with normal skin (p<0.05). Up-regulation of COX-1 in chronic venous leg ulcers could produce prostacyclin, which contributes to angiogenesis. Thus, inhibition of COX-1 by non-steroidal anti-inflammatory drugs (NSAIDs) could increase the local ischaemia and hypoxia associated with chronic venous ulcers. On the other hand, up-regulation of COX-2 is most likely responsible for the persistent inflammation in chronic venous leg ulcers. COX-2 selective inhibitors could therefore be effective in the treatment of chronic venous ulcers.

Desmosomes are reduced in the mouse uterine luminal epithelium during the preimplantation period of pregnancy: a mechanism for facilitation of implantation.

Dynamic regulation of intercellular junctions is an essential aspect of many developmental, reproductive, and physiological processes. We have shown that expression of the desmosomal protein desmoplakin decreases in the luminal uterine epithelium during the preimplantation period of pregnancy in mice. By the time of implantation (between Days 4.5 and 5 of pregnancy), desmoplakin protein can barely be detected by SDS-PAGE and Western blotting, and by immunocytochemistry, it is restricted to well-spaced, punctate dots at the apicolateral junction. Using confocal XZ series and electron microscope quantitation, both the density and distribution of desmosomes along the lateral cell surfaces of luminal epithelial cells were observed to change during early pregnancy. On Day 1 of pregnancy, desmosomes were found at high density in the apicolateral junctional complex, being present here in 79% of ultrathin sections examined, whereas on Day 5, the density was much reduced (present in only 18% of ultrathin sections examined). Desmosomes were found along the lateral surfaces, at or below the level of the nucleus, in 15% of ultrathin sections examined on Day 1 of pregnancy but in only 1% on Day 5. Desmoplakin mRNA declined during the first 4-5 days of pregnancy, along with the protein, suggesting that these changes are controlled at the level of mRNA. This study shows that desmosomes are regulated during early pregnancy, and we propose that a reduction in desmosome adhesion facilitates penetration of the luminal epithelium by trophoblast cells at implantation.

Expression of nitric oxide synthase isoforms and arginase in normal human skin and chronic venous leg ulcers.

Chronic venous ulcers, an example of abnormal wound healing, show chronic inflammation with defective matrix deposition which together with the underlying vascular pathology, result in delayed healing. L-arginine is known to be metabolized by one of two pathways: nitric oxide synthase (NOS), producing nitric oxide (NO), or arginase, producing ornithine. NO is involved in many pathological conditions including vascular and inflammatory disorders. This study therefore investigated the distribution, level and activity of NOS and arginase in chronic venous ulcers in comparison with normal skin, using immunocytochemistry, western blotting, and enzyme assays. The results demonstrated an increased distribution of both NOS and arginase in chronic venous ulcer tissue compared with normal skin, with inflammatory cells and vascular endothelial cells as the main sources. These data were confirmed by western blot analysis, which showed increased levels of both enzymes in chronic venous ulcers. Moreover, there was significantly increased activity of both total NOS (p<0.04) and inducible NOS (p<0.05) in chronic venous ulcer tissue compared with normal skin, and significantly increased activity of arginase (p<0.01) in chronic venous ulcer tissue in comparison with normal skin. NO is known to combine with hydroxyl free radicals forming peroxynitrite, a potent free radical which causes tissue destruction. NO overexpression in chronic venous ulcers may be involved directly or indirectly (through production of peroxynitrite) in the pathogenesis and delayed healing of chronic venous ulcers, through its effects on vasculature, inflammation, and collagen deposition. Arginase is known to enhance matrix deposition. Thus, increased levels of arginase in chronic venous ulcers could contribute to the pathogenesis of lipodermatosclerosis associated with chronic venous insufficiency, predisposing to the formation of chronic venous ulcers and also to matrix cuff formation around blood vessels.

IL-1 alpha and IL-6 production by oral and skin keratinocytes: similarities and differences in response to cytokine treatment in vitro.

IL-1 alpha and IL-6 are important pro-inflammatory and immunomodulatory cytokines and their production by oral (OK) and skin keratinocytes (SK) was compared. OK and SK produced IL-1 alpha, but not IL-6, constitutively. TNF alpha stimulation increased IL-1 alpha production by both cell types and exhibited synergy with interferon gamma (IFN gamma), although the latter had no effect by itself. In contrast, both cell types produced IL-6 in response to TNF alpha, IFN gamma or IL-4, and IFN gamma and IL-4 exhibited synergy with TNF alpha. For all cytokines the levels of IL-6 production were greater for OK than SK and OK, but not SK, produced IL-6 in response to IL-1 alpha stimulation. In addition, the IL-6 response to IL-4 stimulation was more rapid for OK than SK. These observations may explain the similarities and differences in wound healing and immuno-inflammatory diseases affecting the skin and oral mucosa.

Epidermal and oral keratinocytes are induced to produce RANTES and IL-8 by cytokine stimulation.

RANTES, interleukin-8 (IL-8), and macrophage inflammatory protein-1-alpha (MIP-1 alpha) exhibit different and highly selective chemotactic activity for leukocytes. Resting cultured normal oral and skin keratinocytes produced little if any of these chemokines. Stimulation with 250-1,000 U/ml of tumor necrosis factor-alpha (TNF-alpha) or interferon-gamma (IFN-gamma) induced both cell types to produce RANTES. Protein levels peaked after 48 h and mRNA levels peaked after 8 h of stimulation. Used combination, TNF-alpha, and IFN-gamma synergistically increased mRNA and protein levels. Amounts of 100-1,000 U/ml of TNF-alpha also induced IL-8 production with peak mRNA levels after 4-24 h of stimulation and maximal protein production after 72 h or more. IL-8 production by oral keratinocytes was significantly greater than that by skin keratinocytes. Although IFN-gamma alone did not induce IL-8 production, it enhanced the effect of TNF-alpha on both cell types. Stimulation for 24 h with 100-1,000 U/ml of IL-alpha also induced IL-8 production by oral but not skin keratinocytes. No MIP-1 alpha production was detected under the conditions investigated. Keratinocyte production of RANTES and IL-8, under the influence of cytokines such as TNF-alpha or IFN-gamma, provides a mechanism for the selective accumulation of leukocytes into immunoinflammatory diseases of the skin and oral mucosa. Differences in their production may help to explain differences in the presentation of these diseases on the skin and oral mucosa.

Three dimensional analysis of the retinal vasculature using immunofluorescent staining and confocal laser scanning microscopy.

AIM: To undertake a qualitative and quantitative analysis in three dimensions of the human retinal vasculature. METHOD: Fixed and excised whole retinas were permeabilised and subjected to immunofluorescent staining for blood vessel components followed by confocal laser scanning microscopy. Single projection and stereoimages were constructed using computer software. XZ sections through the retina were constructed and the vasculature analysed using appropriate software. RESULTS: Immunofluorescent staining with no discontinuities was present in vessels of all sizes, the confocal images of the capillary network being free of out of focus blur at all depths. Quantitative analysis of XZ sections confirmed the qualitative impression of sharp delineation of the deep retinal capillary plexus, an absence of laminar arrangement of capillaries within the inner retina, and a truncated cone of capillaries around the foveal avascular zone (FAZ) wherein the superficial capillaries approached the FAZ more closely than those in the deeper retina. CONCLUSION: Immunofluorescent staining of the retina and confocal laser scanning microscopy were shown to be useful in analysing accurate three dimensional reconstructions of the normal retinal vasculature without affecting overall tissue architecture.

Venous ulcer fibroblasts compared with normal fibroblasts show differences in collagen but not fibronectin production under both normal and hypoxic conditions.

Previous immunocytochemical analysis showed that the base of venous ulcers was deficient in fibronectin compared with surrounding "normal" dermis. Here, we investigate whether impaired synthetic ability of ulcer fibroblasts could underlie this observation. Ulcer fibroblasts, established in culture from biopsies of the edge of chronic venous leg ulcers, were compared with normal fibroblasts grown from biopsies of site-and age-matched normal skin for their ability to synthesize matrix molecules. Collagen and fibronectin synthesis were measured following metabolic labeling, as collagenase susceptible counts and counts with gelatin affinity, respectively. More collagen was produced by normal fibroblasts than ulcer fibroblasts, both when the cells were cultured on plastic and in collagen gels. In fibronectin synthesis, however, there was no major difference between the two cell types on either substratum. The hypoxic environment to which ulcer fibroblasts are exposed may have caused the intrinsic differences in collagen synthesis by the two fibroblast types. When we tested the effect of culturing cells under hypoxic conditions, both cell types produced less collagen, especially normal fibroblasts grown in a collagen gel, but there was no effect of hypoxia on fibronectin synthesis. We conclude that venous ulcer edge-derived fibroblasts have an impaired ability to synthesize collagen in vitro, but synthesize fibronectin normally. Therefore, the low level of fibronectin found in venous ulcers is not likely to be due to the inability of ulcer cells to produce it or to the response to hypoxic conditions but may be due to the degradation of synthesized fibronectin by proteases present in these ulcers.

A role for HGF/SF in neural induction and its expression in Hensen's node during gastrulation.

It was previously shown (Roberts, C., Platt, N., Streit, A., Schachner, M. and Stern, C. D. (1991) Development 112, 959-970) that grafts of Hensen's node into chick embryos enhanced and maintain expression of the L5 carbohydrate in neighbouring epiblast cells, and that antibodies against L5 inhibit neural induction by such a graft. We now show that L5 is initially widely expressed in the epiblast, but as neural induction proceeds it gradually becomes confined to and up-regulated in the early neural plate. L5 can therefore be considered as a marker for cells that are competent to respond to neural induction. We also show that Hepatocyte Growth Factor/Scatter Factor (HGF/SF) promotes the expression of L5 by extraembryonic epiblast in collagen gels after overnight culture. Explants cultured for several days in the presence of HGF/SF, as well as explants of prospective neural plate, can differentiate into cells with neuronal morphology expressing neuronal markers. To investigate whether HGF/SF is expressed in the chick embryo at appropriate stages of development, we produced specific cDNA probes and used them for in situ hybridization. We find that at the primitive streak stage, HGF/SF is expressed specifically in Hensen's node. We therefore propose that HGF/SF plays a role during the early steps of neural induction, perhaps by inducing or maintaining the competence of the epiblast to respond to neural inducing signals.

A new in vitro model of murine mesoderm migration: the role of fibronectin and laminin.

Examination of the factors involved in primary mesodermal migration in the mouse has been complicated by the lack of a suitable in vitro model. We have developed a new culture system using primitive streak stage embryos denuded of primitive endoderm, which allows easy observation and manipulation of the outgrowing cells. The cells migrating away from these explants were shown by immunocytochemistry to express vimentin and an epitope of the I antigen recognised by the antibody C6, both of which are present on the newly emerged mesoderm and not on the embryonic ectoderm in sections of embryos in utero. Conversely, cytokeratin, stage-specific embryonic antigen 1 (SSEA-1), E-cadherin and desmoplakin are expressed by the embryonic ectoderm but lost during mesoderm formation in vivo. They are absent or expressed very weakly by the migrated cells in vitro. In addition, only explants of the ectoplacental cone (EPC) and visceral endoderm alone, expressed a carbohydrate epitope (recognised by monoclonal antibody BOO6), characteristic of the EPC and primitive endoderm in utero, but absent from mesoderm. Thus we conclude that the cells which outgrow in this system are indeed mesodermal in phenotype. We have confirmed the work of others in demonstrating the presence of fibronectin (FN) and laminin (LN) in the migratory path of the mesoderm, at the ectoderm-visceral endoderm interface. We also report that the beta 1 integrin subunit of the FN and LN receptor is expressed by mesodermal cells at this interface. Using our in vitro model we have examined the role of the extracellular matrix (ECM) in mesodermal migration. Mesodermal cells migrate further and faster on substrates coated with FN or LN, and this increased migration is abolished by appropriate blocking antibodies. We conclude that the ECM, in particular FN and LN, plays an important role in the migration of primary mesodermal cells during gastrulation in the mouse embryo.

HGF-SF: a neural inducing molecule in vertebrate embryos?

Neural induction is the process, during early embryonic development, by which cells of the mesoderm cause the overlying ectoderm cells to differentiate into neural structures, rather than epidermis. The phenomenon was discovered over 80 years ago in Hans Spemann's laboratory, and has since attracted much interest. However, we are still ignorant about the signals that elicit such a change in the direction of ectodermal differentiation, and about the mechanisms involved in the response of the ectoderm. Here, we report that HGF-SF can cause cultured chick ectodermal cells to become neural. We also discuss preliminary evidence suggesting that a homolog of this factor is expressed in Hensen's node, the inducing tissue, at about the stage at which neural induction occurs. We speculate that HGF-SF, or a related factor, could be a neural inducing signal during the early development of vertebrate embryos.

Epithelial scatter factor and development of the chick embryonic axis.

Scatter factor, a recently characterised protein secreted by certain embryonic fibroblasts, affects cultured epithelial by increasing cell motility, the breakdown of cell junctions and cell scattering. The process of gastrulation in higher vertebrate embryos, during which the primitive streak forms, involves an epithelial-to-mesenchymal transformation resembling the effects of the factor on cultured cells. The factor was applied locally to chick embryos, using both scatter-factor-secreting cell lines and inert carriers. We found that scatter factor can generate local supernumerary axial structures resembling primitive streak and/or neural plate and conclude that it may have primitive-streak and/or neural-inducing activity in chick embryos.

Limitation of substratum size alters cytoskeletal organization and behaviour of Swiss 3T3 fibroblasts.

Cell spreading and adhesion formation in Swiss 3T3 cells was studied on circular adhesive islands of size 400-500 microns 2 made by evaporating palladium through a mask onto an underlying non-adhesive surface. Cell spreading was limited since focal contacts were restricted to the palladium. On islands less than 2000 microns 2, focal contacts and actin bundles were arranged at the cell periphery. On islands less than 1000 microns 2, the size and number of focal contacts were reduced. Focal contacts may be important regulators of proliferation, but they do not seem to form a deterministic link between substratum contact and proliferative stimulus.

Effect of patterned surfaces of adhesive islands on the shape, cytoskeleton, adhesion and behaviour of Swiss mouse 3T3 fibroblasts.

A pattern of circular islands of adhesive substratum can be used to control cell shape and behaviour. We have shown previously that the proportion of Swiss 3T3 cells that synthesize DNA varies with the area of the island to which they are attached, within the range 500-5000 microns2. In this paper we investigate the cytoskeleton and adhesions of cells on islands using a variety of techniques including phalloidin staining and interference reflection microscopy. Islands of area 2000 microns2 or less constrain cell shape, and cause focal contacts and actin microfilament bundles to accumulate in a circle at the margin. These changes are most obvious in islands of about 1000 microns2, in which a complete ring of adhesion is sometimes formed in the periphery of the cell. This peripheral distribution is less common in cells on even smaller islands, and the focal contacts become smaller and less numerous. It is not yet clear whether any of these structural changes are associated directly with the proliferative stimulus due to contact with the substratum. However, we expect that the use of patterned substrata will contribute to the study of how cell shape and structure regulate many cell functions.

Antibacterial perineal washing for prevention of recurrent urinary tract infections.

Antibacterial ointment applied to the urethral meatus in females with recurrent urinary tract infections (UTIs) has been reported to decrease the incidence of UTIs. The value of perineal washing with the antibacterial lotion, hexachlorophene, in preventing urinary tract infections was determined in 56 females with recurrent UTIs. Of three groups of females one group used hexachlorophene perineal washings morning and night, another group used hexachlorophene perineal washings and an oral antibacterial daily (nitrofurantoin or trimethoprim-sulfamethoxazole), and the third group used the oral antibacterials daily alone. The infections per patient were 3.4 with hexachlorophene washings, 0.5 with hexachlorophene washing and oral antibacterials daily, and 0.9 with oral antibacterials alone. These results suggest that hexachlorophene perineal washing was not effective in preventing UTI in females.

Spontaneous rupture of hydronephrotic renal pelvis with massive hemorrhage.

Spontaneous rupture of a hydronephrotic renal pelvis with massive hemorrhage presents as a perplexing diagnostic challenge to the urologist. The preoperative distinction between this condition and spontaneous rupture of the kidney (Wunderlich disease) may prove uncertain or equivocal despite the application of exhaustive diagnostic modalities. Surgical intervention should follow initial stabilization, and nephrectomy has been the procedure of choice.

New evidence that growth in 3T3 cell cultures is a diffusion-limited process.

When a confluent culture of 3T3 cells is wounded, new growth occurs at the wound margins. This indicates that the suppression of growth in the intact confluent sheet is under local control, a phenomenon known as 'topoinhibition', and it has been suggested that intercellular contact is responsible. An alternative explanation for topoinhibition is possible, however, namely that a soluble factor necessary for growth is locally depleted from the medium by cells so that each cell in a confluent sheet normally receives an insufficient supply and its growth is inhibited. Here we show that the pattern of release from topoinhibition in a wounded culture can be distorted simply by inducing a gentle laminar flow of medium across the wound. Growth remains suppressed at the upstream margin of the wound despite the reduced level of intercellular contact at wound edges. We conclude that the signal for topoinhibition is carried by the flow as would be predicted if it were the local depletion of growth factor.

The ultrastructure of polygonal networks in chick embryonic cells in vitro.

Polygonal networks in cultured chick endoderm cells are ordered arrays of actin microfilaments situated just beneath the dorsal cell surface. Each strut is formed from a bundle of microfilaments and 5-7 bundles intersect at each node. Dense bodies are seen in nodes and some struts. At its periphery the network is attached to the substrate at the termini of long radial struts. Most of the network is resistant to detergent extraction. Sliding microfilaments can explain the observed behaviour of networks in live cells.

Cell-substrate contacts in cultured chick embryonic cells: an interference reflection study.

Cell-substrate contacts in explants of different regions of early chick tissues were investigated using the technique of interference reflection microscopy. All the explants spread as epithelial sheets. During initial spreading a peripheral zone of 2-3 cells formed broad contacts with the substrate. In spread explants some cells in the centre made broad substrate contacts. A mat of extracellular material containing fibronectin was found under the explants. Focal contacts and focal adhesions increased in number during culture, and stress fibres were associated with them. These changes in cell contacts appeared more quickly in some tissues than in others. After 24 h, explants of hypoblast and definitive endoblast could easily be distinguished but by 7 days they were very similar. In the absence of serum, specialized cell contacts developed more quickly; in higher concentrations of serum, more slowly. Confrontations between explants were also examined. The most conspicuous feature was that cells in invading explants normally underlapped invaded cells. Invasion from above by an unspread explant could occur even if the invaded explant had formed many focal adhesions.

Polygonal networks in living chick embryonic cells.

Regular polygonal networks have been found in explants and dissociated cells of early chick embryos. These networks are readily observable in live cells with phase-contrast optics thus allowing time-lapse cinemicroscopy. They consisted of a regular pattern of nodes and radiating struts found predominantly in the lamelliplasm of the free edges of the cells bordering explants. At the outer edge, the network was terminated by radial struts associated with substrate-attached retraction processes whilst toward the centre of the cells it faded out. The network was also associated with stress fibres running across the cell and with microextensions on the dorsal surface. Even within one cell the network varied in size. Time-lapse films showed that microvilli were protruded from the dorsal surface over the nodes. Although the cells containing the networks were poorly motile the network itself was a mobile structure. Many explants from regions differing in prospective fates developed these networks after 2-4 days in culture. They appeared earlier in the smaller less yolky cells of definitive endoblast and epiblast. Experiments with dissociated and reaggregated cells confirmed their occurrence mainly in free edges of cells. The relationship between these networks seen in living chick embryo cells and those seen in other cell types using immunofluorescent techniques is discussed and a mechanism is proposed for their formation.