Glucosaminylmuramyl dipeptide (GMDP) modulates endothelial cell activities in vitro but has no effect on angiogenesis in vivo.

OBJECTIVE AND DESIGN: The aim of the study was to evaluate the effects of GMDP on angiogenesis in vivo and as a modulator of human umbilical vein endothelial cell proliferation, cell surface antigen expression and cell adhesion in vitro. MATERIALS: Human umbilical vein endothelial cells (HUVEC), fertilized white leghorn chicken eggs, antibodies against adhesion molecules and glucosaminylmuramyl dipeptide (GMDP). TREATMENT: GMDP [0.01-100 micrograms/ml] applied to cell cultures for 6-72 h and to the chick chorioallantoic membrane (CAM) for four days. METHODS: Angiogenic activity of GMDP in vivo was assessed using the CAM assay; HUVEC proliferation was measured by tritiated thymidine incorporation and cell cycle studies; cell surface antigen expression by indirect immunofluorescence and flow cytometry; cell adhesion by quantification of [3H]-thymidine labeled leukocyte adherence to HUVEC monolayers. Statistical analysis was performed using one-way ANOVA and if necessary was followed by Duncan's multiple range test for variables. RESULTS: GMDP induced [3H]-thymidine incorporation in a concentration- and time-dependent manner (p < 0.003) and significantly increased the porportion of cells in the S phase of the cell cycle (p < 0.03). It weakly augmented the expression of ICAM-1 and CD31 but not adhesion of leukocytes to HUVEC monolayers GMDP was not angiogenic in the CAM assay. CONCLUSIONS: GMDP can modulate endothelial cell activity without the induction of angiogenesis in vivo which may have implications for its use as a therapeutic agent.

Adjuvancy and reactogenicity of N-acetylglucosaminyl-N-acetylmuramyl-dipeptide (GMDP) orally administered just prior to trivalent influenza subunit vaccine. A double-blind placebo-controlled study in nursing home residents.

One hundred and fifty-three nursing home residents received 0, 5, 25 or 50 mg N-acetylglucosaminyl-N-acetylmuramyl-dipeptide (GMDP) orally, and trivalent influenza subunit vaccine intramuscularly. One day after intervention, there was a strong increase of total leucocytes, monocytes and neutrophils in the groups receiving 25 or 50 mg GMDP. A GMDP dose dependent increase in systemic, but not in local, vaccine side-effects was observed. No significant differences in post-vaccination haemagglutination inhibiting serum antibody titres were observed between the four groups, indicating that oral administration of GMDP together with influenza vaccination, does not lead to a higher vaccine efficacy.

The occurrence of natural antibodies to minimal component of bacterial cell wall (N-acetylglucosaminyl-N-acetylmuramyl dipeptide) in sera from healthy humans.

ELISA assay showed that sera from each of 729 healthy donors contained antibodies to the minimal component of bacterial cell walls, N-acetylglucosaminyl-N-acetylmuramyl dipeptide (GMDP). Anti-GMDP antibody levels were determined for 686 sera which were classified into 3 groups: high (17.6%), medium (68.7%), and low responder (13.7%). Inhibition analysis performed on representative sera showed that a proportion contained specific anti-GMDP antibodies reacting only with GMDP (i.e., GMDP interaction with anti-GMDP antibodies was inhibited by GMDP only) whereas the remaining sera reacted both with GMDP and with the tetrasaccharide (GlcNAc-MurNAc)2 (i.e., GMDP interaction with anti-GMDP antibodies in the latter sera was inhibited by both GMDP and the tetrasaccharide). Inhibition analysis indicated, moreover, that the anti-GMDP antibodies contained in high-responder sera had higher affinity than those present in low-responder ones: GMDP inhibited the GMDP + anti-GMDP antibody interaction by 88.7% in the former sera vs. 53% in the latter. Sera contained both IgM and IgG antibodies to GDMP, but the mean level of anti-GMDP IgG antibodies in the high-responder sera was 30 times higher than in the low-responder ones.

In vitro cutaneous biotransformation of propranolol.

The metabolism of propranolol by human skin and by several cell preparations has been investigated in vitro. The major metabolites produced by human skin in organ culture and by keratinocytes were N-desisopropylpropranolol (DIP), propranolol glycol (GLY), and naphthoxylactic acid (NLA). Formation of GLY and NLA was linear with incubation time up to 6 d and was directly proportional to propranolol concentration. Fibroblasts and melanocytes also produced GLY and NLA, but appeared to have lower propranolol-biotransforming activity than keratinocytes. The three metabolites detected arise from side-chain oxidation of propranolol, and the use of specific enzyme inhibitors determined that monoamine oxidase and cytochrome P450 isozymes are involved in their formation. Aldehyde and alcohol dehydrogenases are also probably involved in the formation of NLA and GLY, but attempts to inhibit these enzyme systems were inconclusive, possibly due to the chemical instability of the intermediate aldehyde resulting from monoamine oxidase activity. No evidence was found for conjugation or ring oxidation by the skin or isolated cells. Induction of keratinocyte differentiation with Ca++ or phorbol ester treatment resulted in an increase of overall biotransformation and the NLA/GLY ratio.

The influence of heparin on the wound healing response to collagen implants in vivo.

The biologic response to fibrillar collagen (collagen) and fibrillar collagen plus heparin (collagen/heparin) implants have been compared in the rat subcutaneous and guinea pig dermal wound models. The reconstituted bovine dermal collagen implants were injected subcutaneously in rats at concentrations ranging from 18 to 30 mg/ml and in volumes ranging from 0.5 to 1.0 ml. The biologic response to the collagen implants alone was characterized by a transient invasion of a modest number of inflammatory cells within the first three days of implantation that was followed by limited fibroblast invasion into the peripheral 1/3 of the implant during the course of the next three to four weeks. Occasionally, blood vessels were observed to invade the peripheral regions of the implant. The degree (number) and extent (depth) of cell invasion were inversely related to initial collagen implant concentration. Addition of heparin (0.3-20 micrograms/mg collagen) to these implants resulted in a significant dose-dependent increase in the degree and extent of fibroblast invasion. Radiolabeling studies showed that the collagen and collagen/heparin implants were cleared from the subcutis at identical rates. Implantation of these formulations in a guinea pig dermal wound model was also performed, using a semi-occlusive wound dressing (Opsite) to maintain the implant in the wound site. The fibrillar collagen implant alone was pushed upward by developing granulation tissue at the base of the wound and served as a support for epidermal cell migration, proliferation, and differentiation as wound closure proceeded. The implant was slowly invaded and turned over as granulation tissue developed from the base and margins of the wound bed. The inclusion of heparin in these implants resulted in a significantly different pattern of wound healing. The collagen/heparin implants histologically presented a more broken-up or porous appearance following implantation, which was associated with a greater degree of penetration of developing granulation tissue into the implant itself as compared to the collagen implants. Radiolabeling studies revealed that clearance rates of implants with and without heparin from wound sites were similar, as noted in the rat subcutis. Laser doppler flowmetry studies suggested that the heparin--containing implants were more vascular than control wound sites or sites treated with collagen alone.

The preparation and physicochemical characterization of an injectable form of reconstituted, glutaraldehyde cross-linked, bovine corium collagen.

Pepsin-solubilized bovine corium collagen was purified, reconstituted, and treated with various levels of glutaraldehyde. Treatment of suspensions of fibrillar collagen with low concentrations of glutaraldehyde appeared to have little effect on the gross morphology of fibrils, as judged by electron microscopy, but did have a significant impact on their physicochemical stability. Fibrillar collagen treated with glutaraldehyde at a concentration equal to or greater than 0.0075% demonstrated significant decreases in neutral solubility at elevated temperatures as compared to noncross-linked controls. Differential scanning calorimetry provided a convenient and quantitative means to correlate increases in melting temperature with increases in glutaraldehyde treatment concentration. Fibrillar collagen cross-linked with glutaraldehyde concentrations as low as 0.0075% demonstrated a significantly greater resistance to proteolytic degradation than did noncross-linked fibrillar collagen samples. The residual, extractable aldehyde content of such preparations was between 1 and 3 ppm. Rheological measurements on such cross-linked suspensions demonstrated that they were non-Newtonian, shear-thinning fluids, and that they were two- to threefold more viscous than corresponding preparations of noncross-linked collagen.

Structural variability of collagen fibers in the calcareous axial rod of a sea pen.

Previous studies revealed that the organic matrix of the skeletal rod of the sea pen, Veretillum cynomorium, contained about 50% collagenous protein. The present ultrastructural study, based upon conventional staining methods, shows the existence of an abundant, longitudinally arranged nonbanded and fibrillar material separated by a reticular matrix. After incubation with 3 H-proline, labeling is specifically localized on the fibrillar material. Some fibers occasionally display a transverse striation with a period of 11 to 14 nm which can be associated with a chevron striation. Infrequently, some other fibers display a more distinct banding of 55 to 70 nm or even yield a checkerboard pattern. However, a majority of fibers remain without a regular structure comparable to the periodic striations observed in the collagen of other animals. After treatment with 1% PTA in 70% ethanol, all the fibers show a clear banding of 14 nm and some of them possess two types of striations. The same result is obtained on fibers mechanically dissociated and negatively stained. As these methods show a periodic banding pattern on all the fibers, it is likely that all the fibers (striated or not) observed after routine electron microscopy correspond to collagen material. This collagen appears to be both polymorphic and completely new in comparison to that which is characteristic of the mesoglea. The polymorphic aspect is compared to that obtained from vertebrate collagens.

ConA- and PNA-binding glycoproteins of human epidermis.

Concanavalia ensiformis agglutinin (ConA) and Arachis hypogaea agglutinin (PNA) binding glycoproteins in NP-40 extracts of human epidermis and epidermal cell preparations have been investigated by incubation of SDS-polyacrylamide gels with 125I-labeled lectins. In whole epidermis, 125I-ConA labels numerous glycoproteins, of which some appear to be restricted to the more differentiated upper layers of the epidermis and thus may represent markers of epidermal differentiation. These glycoproteins are not expressed by cells cultured on collagen. 125I-PNA labels a limited number of components, of which the major one is intercellular and/or extremely trypsin-sensitive. This material is expressed in low amounts after 13 days in culture.

Concanavalin (ConA)-reactive human epidermal glycoproteins.

The possibility that glycoconjugates play a role in epidermal differentiation has stimulated attempts at their biochemical identification. We have examined the Concanavalin (ConA)-binding glycoproteins present: in whole human epidermis, epidermal cells obtained by trypsinization of human skin, non dissociated layers (the most differentiated ones) remaining after preparation of these cells and human epidermal cell cultures. 125I-ConA was applied to NP-40 soluble molecules separated by SDS-PAGE. ConA labelling of the whole epidermal extract visualized numerous bands. That of trypsinized cells differed in the intensity of some bands and the absence of others. However most of the latter bands could be identified in the most differentiated layers. These glycoproteins were absent or poorly labelled in the cultures of keratinocytes. The restriction of some ConA-binding glycoproteins to the most differentiated epidermal layers, suggests that these molecules are involved in, or can be considered as markers of the human epidermal differentiative process.

Abnormal glycosylation of human fibronectin secreted in the presence of monensin.

Detailed studies of the effects of the ionophore monensin upon the glycosylation of secreted fibronectin have been carried out. Human fibroblasts in culture were incubated in 1 microM monensin for several hours, following which radiolabeled glucosamine or mannose was added to the cultures. Parallel incubation and labeling of control cultures were done. Labeled fibronectin was isolated from the culture media by gelatin-Sepharose chromatography, from cell surfaces by urea extraction, and from intracellular locations by cell lysis followed by immunoprecipitation. Detailed comparison of the glycopeptides released from fibronectin by pronase and of the oligosaccharides liberated by hydrazinolysis was carried out, particularly focusing on the secreted fibronectin, using gel filtration, high performance liquid chromatography, and concanavalin A chromatography, in conjunction with the use of endoglycosidase H and specific exoglycosidases. We demonstrate that fibronectin in the medium of monensin-treated cultures differs in its glycosylation pattern from the control fibronectin. High mannose oligosaccharides are abundant in the monensin-derived fibronectin, whereas the control protein contains primarily complex oligosaccharides. Monensin apparently does not alter the initial glycosylation of fibronectin since the high mannose oligosaccharides are present on both control and monensin-treated intracellular fibronectin. We suggest, therefore, that monensin, by impairing intracellular translocation through the Golgi region, allows incompletely processed forms of fibronectin to reach the cell surface and to be released into the culture medium.

Effects of the ionophore monensin on type II collagen and proteoglycan synthesis and secretion by cultured chondrocytes.

Detailed studies of the effects of the ionophore monensin upon avian chondrocyte ultrastructure, macromolecular synthesis, and macromolecular secretion have been carried out. Embryonic avian chondrocytes in suspension culture were incubated in concentrations of monensin ranging from 1 X 10(-7) to 1 X 10(-6) M for durations up to 8 h. Electron microscopy revealed that the treated chondrocytes developed abnormal Golgi structures and a markedly distended rough endoplasmic reticulum. Biochemical and immunoassay studies showed that while total protein synthesis was only slightly impaired by monensin, the ionophore had pronounced effects upon the secretion of both type II collagen and proteoglycans. These two macromolecules responded to monensin inhibition in a similar fashion and accumulated within the affected chondrocytes. The kinetics of response over the monensin concentration range used was virtually identical for type II collagen and proteoglycan. Undersulfation of proteoglycan, caused by monensin, was examined by ion exchange chromatography and analysis of the products of chondroitinase ABC digestion. The results indicated that undersulfation affected all glycosaminoglycan chains in a general fashion rather than affecting a specific population of chains.

Temporal and spatial distribution of fibronectin during development of the embryonic chick limb bud.

Indirect immunofluorescence has been used to study the distribution of fibronectin during the course of embryonic chick limb morphogenesis and differentiation. At all stages of development from 19 through 25, fibronectin is distributed throughout the non-differentiating mesenchymal tissue directly subjacent to the apical ectodermal ridge (AER), i.e. the mesenchyme extending 0.15 mm or so from the AER. Fibronectin is also distributed throughout the proximal condensing central core of the limb during the early stages of cartilage differentiation. In fact, fibronectin persists as a major component of the intercellular matrix in the central core of the limb following overt chondrogenic differentiation, since it is present as late as stage 27 throughout the well-differentiated cartilage rudiments of the radius, ulna, and humerus. The detection of fibronectin in differentiated cartilage is facilitated by pre-treatment of the sections with testicular hyaluronidase prior to immunofluorescent staining. In contrast to the chondrogenic central core of the limb, in the peripheral dorsal and ventral (myogenic) regions in which muscle differentiation is progressing, there is a progressive and striking diminution in fibronectin staining. By stage 27, little, if any, is present in the well-differentiated muscle primordia. Finally, at all stages of development, there is an accumulation of fibronectin at the ectodermal-mesenchymal interface, suggesting it is a component of embryonic limb basement membranes. On the basis of these observations, the possible role of fibronectin in limb cartilage and muscle differentiation and in other aspects of limb morphogenesis is discussed.

The phosphate content of human fibronectin.

The nature and distribution of the covalently bound phosphate of human fibronectin have been investigated. Phosphate was detected only as phosphoserine and is not present on asparagine-linked oligosaccharides. Fibroblast fibronectin labeled with 3H-amino acids and [32P]orthophosphate was subjected to limited tryptic digestion. Initial cleavages which reduced the molecule to nondisulfide-bonded polypeptides of Mr approximately 210,000-220,000 did not entail the loss of phosphate. Subsequent cleavage products included intermediate-sized 32P-phosphopeptides (43,000 and 37,000), but all 32P label became increasingly associated with lower molecular weight material (13,000 and 9,000) eventually terminating in a phosphopeptide (or peptides) of Mr approximately 6,000. The phosphorylated region(s) of human fibronectin are apparently not located at the extreme COOH-terminal region containing the interchain disulfide bond, but seem to be within a 40,000-50,000 segment near one end of the molecule. Cyanogen bromide cleavage of 32P-labeled fibronectin isolated from fibroblast culture medium produced one major phosphopeptide (Mr approximately 5,000) which co-purified through two chromatographic procedures with a CNBr phosphopeptide derived from plasma fibronectin. The phosphorylation of fibronectin is evidently restricted to a small region of the protein molecule, implying a very site-specific event. The phosphorylated region is present on human fibronectins obtained from plasma, fibroblast surfaces, and fibroblast culture medium, indicating that the role of the phosphate may be important to all forms of fibronectin.

Immunocytochemical localization of procollagen and fibronectin in human fibroblasts: effects of the monovalent ionophore, monensin.

The monovalent ionophore monensin inhibits the secretion of both procollagen and fibronectin from human fibroblasts in culture. The distribution of these proteins in control and inhibited (5 x 10(-7) M monensin) cells has been studied by immunofluorescence microscopy. In control cells, both antigens are present throughout the cytoplasm and in specific deposits in a region adjacent to the nucleus, which we identify as a Golgi zone by electron microscopy. Treatment of cells with monensin causes intracellular accumulation of procollagen and fibronectin, initially in the juxta-nuclear region and also subsequently in peripheral regions. Electron microscope studies reveal that in such cells the juxta-nuclear Golgi zone becomes filled with a new population of smooth-membraned vacuoles and that normal Golgi complexes are not found. Immunocytochemically detected procollagen and fibronectin are localized in the region of these vacuoles, whereas more peripheral deposits correspond to the dilated cisternae of rough endoplasmic reticulum, which are also caused by monensin. Procollagen and fibronectin are often codistributed in these peripheral deposits. Accumulation of exportable proteins in Golgi-related vacuoles is consistent with previous analyses of the monensin effect. The subsequent development of dilated rough endoplasmic reticulum also containing accumulated proteins may indicate that there is an additional blockade at the exit from the endoplasmic reticulum, or that the synthesized proteins exceed the capacity of the Golgi compartment and that their accumulation extends into the endoplasmic reticulum.

Kinetic studies of the intracellular transport of procollagen and fibronectin in human fibroblasts. Effects of the monovalent ionophore, monensin.

The monovalent ionophore, monensin, has been found to inhibit the secretion of both procollagen and fibronectin from human fibroblasts in cell culture. The kinetics of inhibition, as well as those for the release of inhibition, suggested that both proteins may be cotransported in the cell. In the present study, we have examined the intracellular translocation and release into the culture medium of procollagen and fibronectin, in the presence or absence of monensin. Pulse-chase studies were combined with subcellular fractionation of the fibroblasts to determine the rates of intracellular movement. We found that monensin did not significantly affect either the subcellular fractionation, or the distribution of organelle marker enzymes along the gradient, and that procollagen moved from a region of high buoyant density (primarily endoplasmic reticulum), through a mid-density region (primarily Golgi elements), to a region of low buoyant density before exiting from the cell. Monensin markedly decreased the rate of transit from one region to the other; kinetic analysis of the data showed a greater than 3-fold decrease in the rate constants for intracellular movement into the low buoyant density components and thence to the culture medium. The density gradient distribution of fibronectin was affected by monensin in similar fashion, indicating that it and procollagen probably follow the same intracellular route. Since monensin causes both proteins to accumulate in highest abundance in regions of the density gradient corresponding to Golgi and endoplasmic reticulum, the site of ionophore blockade appears to be within the Golgi apparatus. Thus, procollagen and fibronectin share a common intracellular route prior to entering the Golgi region of the cell.

Calcification of the collagenous axial skeleton of Veretillum cynomorium pall. (Cnidaria: Pennatulacea).

The axial skeletal rod of Veretillium cynomorium consists of a fibrillar collagenous matrix calcified with calcite. The present paper describes ultrastructural and crystallographic details of its organization and deposition. At the inferior end of the rod is a calcification gradient between the noncalcified tip and the rest of the axis. Initial mineral deposits, which are sometimes associated with cell debris, give rise to calcitic nodules which enlarge by the radical growth of several lobes. These nodules fuse and form the core of the axis. Subsequent increase in diameter of the rod involves the radial development of irregular columns of calcite which arise from the peripheral nodules. Mineral surfaces exhibit a distinctive microarchitecture which can be related to the predominantly c-axis parallel growth of the calcite. Particular attention is paid to the relationship between mineral and matrix. The collagen fibrils, embedded in the calcite but never impregnated with it, are not responsible for the initial nucleation of mineral. The crystallographic orientation of the calcite also appears to be independent of these fibrils.