Chitosan and chitosan sulfate have opposing effects on collagen-fibroblast interactions.

Soon after injury, hyaluronan is prominent in granulation tissue. As hyaluronan wanes, sulfated glycosaminoglycans predominate. The temporal relationship between the transition from unsulfated to sulfated glycosaminoglycans and the phenotypic changes in fibroblasts in the wound bed suggest that these two events are interrelated. This possibility was investigated using chitosan and its sulfated product as model compounds. The ability of cultured human foreskin fibroblasts to bind and to contract lattices of collagen, collagen-chitosan, and collagen-chitosan sulfate was determined. Fibroblast adherence to substrates after 24 hours was determined by the MTT assay at A570. Adherence to the collagen-chitosan substrate was markedly reduced (mean A570 +/- SD; 0.16 +/- 0.05, n = 6) (p < 0.01) compared to collagen alone (0.92 +/- 0.04) or to collagen-chitosan sulfate (0.84 +/- 0.05). Kinetics of contraction of lattices by enmeshed fibroblasts was determined by planimetric measurements, 0-48 hours after loosening the lattices. Contraction of the collagen-chitosan lattices (n = 5) was less at all time points than for the other two lattices. After 48 hours, the collagen- chitosan lattices contracted significantly (p < 0.01) less (30.0% +/- 4.4) compared to collagen alone (66.9% +/- 4.7) and collagen-chitosan sulfate (71.6% +/- 7.7). Scanning electron microscopy of the acellular lattices showed fibers of the collagen-chitosan mixture to be the thickest and with altered organization. These results show that chitosan sulfation markedly enhances fibroblast adhesion and promotes contraction of a collagen lattice compared to the unsulfated material. By analogy to the in vivo sequence of hyaluronan followed by sulfated glycosaminoglycans in wounds, the results suggest that glycosaminoglycan sulfation may be a contributing signal for phenotypic transformation during wound healing.

Radiochemical studies on contact lens soilation. II. Lens uptake of cholesteryl oleate and dioleoyl phosphatidylcholine.

Employing an artificial tear preparation composed of six proteins and six lipids as a deposit model, uptake of the lipids 3H-cholesteryl oleate and 14C-dioleoyl phosphatidylcholine was measured on contact lenses representative of the four FDA hydrogel groups and on select RGP lenses. Cholesteryl oleate uptake after 24 h at 37 degrees C generally was less than 1 microgram/lens although occasionally reaching 1-2 micrograms. DuraSoft 3 lenses (Group IV) accumulated the deposits in greater amounts (p = 0.04) with other lens groups not differing significantly from each other. Ionic DuraSoft 2 and 3 lenses bound more phosphatidylcholine (also < 1 microgram) than other lens groups, possibly reflecting an interaction between the positively charged choline residue and the negative surface of the lens. Lysozyme deposition, measured simultaneously with cholesteryl oleate, bound to a far greater extent to Group IV lenses (e.g., DuraSoft 3, mean surface deposit 279 micrograms) than to other lens types (p < 0.01). Multiple application of the artificial tear solution did not produce a statistically significant increase in cholesteryl oleate accumulation.

Anti-platelet action of nitric oxide and selective phosphodiesterase inhibitors.

Nitric oxide gas is a potent inhibitor of platelet aggregation, with an IC50 of 3.6 microM for rabbit platelets. Since the NO effect is mediated via increased cGMP, this in vitro study was undertaken to test the hypothesis that selective phosphodiesterase (PDE) inhibitors might enhance aggregation inhibition at lower NO concentrations. Because the cAMP-selective PDE III and the cGMP-selective PDE V are prominent in platelets, milrinone, a PDE III inhibitor, and zaprinast, a PDE V inhibitor, were tested alone and in the presence of NO for their effect on aggregation. Aggregometry was performed on rabbit platelet-rich plasma following addition of ADP as agonist. Milrinone alone gave an IC50 of 12.4 microM. With each agent set to give suboptimal inhibition of aggregation, the combination of milrinone (3-16 microM) and NO (2-10 microM) produced a greater effect than either agent alone. Zaprinast exhibited no effect on aggregation in concentrations up to 160 microM. However, adding zaprinast to 2 microM NO, which alone reduced aggregation approximately 30%, produced a marked synergism in the inhibitory effect up to and including no observable aggregation. These results indicate that elevation of either cAMP or cGMP is sufficient to inhibit platelet function. The platelet cAMP concentration appears high enough to be inhibitory when degradation is suppressed by milrinone. However, basal cGMP levels must be increased by NO before the zaprinast effect is observed.

Radiochemical studies on contact lens soilation. I. Lens uptake of 14C-lysozyme from simple and complex artificial tear solutions.

A systematic study has been made of the uptake of lysozyme by various contact lens materials. Lens uptake of 14C-methylated lysozyme was assessed using simple to complex artificial tear solutions. The data reflect prominent uptake by Group IV (ionic, high-water-content) hydrogel lenses, consistent with the literature. This includes protein on the lens surface and in the lens matrix, the former being estimated at about 33% of the total deposit for the DuraSoft 3 lenses used. Uptake appears to be contingent upon the nature of the lens material, the composition of the deposit model used, and the duration of lens exposure to the artificial tear solution.

Platelet and neutrophil distributions in pump oxygenator circuits. III. Influence of nitric oxide gas infusion.

The authors used quantitative gamma scintigraphy to evaluate the influence of nitric oxide gas on platelet and neutrophil deposition in Cobe Duo microporous oxygenators during cardiopulmonary bypass (CPB). The effects of nitric oxide gas on circulating platelet and neutrophil counts and platelet function also were assessed. Animals were prepared by standard methods. Cells were harvested, labeled (111 In platelet and 99mTc neutrophil), infused, and recirculated. Nitric oxide gas, a guanylate cyclase pathway promoter, was infused int he Duo gas port at 500 ppm (t = 0-60 min), increased to 1,000 (t = 60-80 min), and stopped (final, 10 min). Images were taken at 10-15 min intervals during CPB. Standard isotope image corrections were made. No differences between nitric oxide gas and control experiments were observed for flow, pressure, hematocrit, or replacement volume. Nitric oxide gas infusion significantly (p < 0.05) reduced both platelet adherence to the oxygenator and in vitro platelet aggregation. Neutrophil adhesion tended to be lower, and circulating platelet and neutrophil counts tended to be higher with nitric oxide gas infusion. Results of in vitro aggregometry studies using rabbit platelets indicate that the class V phosphodiesterase inhibitor zaprinast can strongly enhance the inhibitory effects of nitric oxide. The authors conclude nitric oxide gas is a promising platelet sparing agent in the setting of CPB.

Endothelial cells on Dacron vascular prostheses: adherence, growth, and susceptibility to neutrophils.

Human umbilical vein endothelial cells (HUVEC) on knitted and woven Dacron prostheses were compared with HUVEC on smooth surfaces (tissue culture polystyrene, PET film, and Natrix) with regard to adherence, growth, and susceptibility to injury by neutrophils (PMN). These are properties of importance for successful seeding or coating of prostheses. For prosthetic material of given macroscopic dimensions, more endothelial cells (EC) adhered than to smooth surfaces. However, the prostheses had a greater effective surface area as determined by the number of EC at confluency. When this parameter was taken into account, fewer EC were found adherent to prosthetic material per unit effective surface area than for the smooth surface substrates. Growth on prostheses was clearly inferior to that on smooth surfaces, and EC on prostheses were more susceptible to attack by activated PMN than on smooth surfaces. These differences may reflect the topographic differences in cells attached to fibers where they assume more distorted shapes by stretching to span fibers.

Inhibition of surface-induced platelet activation by nitric oxide.

This study was undertaken to determine whether the nitric oxide/platelet cyclic guanosine monophosphate (NO/cGMP) pathway might be used to reduce platelet activation by artificial surfaces. Because serotonin release (SR) is a platelet activation indicator, rabbit platelets in their own plasma (PRP) were labeled with 3H-serotonin. Labeled PRP was incubated with glass beads for 5-10 min, at 37 degrees C with gentle agitation, and SR was measured. PRP pretreatment with NO gas or nitroprusside + N-acetylcysteine inhibited SR 50%. Dose response studies indicate the existence of an optimal NO concentration above which its inhibitory effect is diminished. The guanylate cyclase inhibitor methylene blue attenuates the NO effect, implicating cGMP in NO mediated inhibition of surface induced platelet activation. Adult pigs were supported on a membrane oxygenator in an in vitro model of cardiopulmonary bypass (CPB). Introduction of NO gas into the oxygenator sweep gas at 500 ppm reduced platelet adherence to the oxygenator surfaces, increased circulating platelet counts, and decreased the rate of platelet aggregation (whole blood impedance platelet aggregometry) compared with the results of the control animals. Indications of NO toxicity were seen when the NO flow rate was increased to 1,000 ppm. These studies support the hypothesis that NO reduces platelet activation by artificial surfaces in clinical devices.

Comparison of tissue factor and prostacyclin production by human umbilical vein endothelial cells on Dacron vascular prostheses and Dacron smooth films.

The functional capacity of human umbilical vein endothelial cells (HUVEC) grown on Dacron (polyethylene terephthalate; PET) vascular prosthetic material was compared with the function of cells on smooth surfaced PET, tissue culture polystyrene (TCPS), and Natrix-coated TCPS. Prosthetic materials include two knitted fabrics (Bionit I and II) and two woven preparations (DeBakey Soft Woven and Extra Low Porosity). Two entities produced by HUVEC that influence blood coagulation were assessed: the procoagulant tissue factor (TF) and the anticoagulant prostacyclin (PGI2). Although TF activity was stimulated on all substrates by endotoxin (LPS), there was no difference among prostheses and no difference among smooth surface materials, but TF was reduced in cells on the prosthetic materials relative to those on smooth surface substrates. The reduced TF production by HUVEC on prosthetic material could be reversed by returning them to TCPS. In contrast, PGI2 production on prostheses was comparable to that on smooth surfaces for both stimulated and unstimulated cells. Stimulation with histamine (1 microM) gave a 2.4-fold increase in PGI2 whereas mellitin (10 micrograms/ml) increased production 12.5-fold. The differential response of HUVEC with regard to these two coagulation factors, one of which is secreted and the other membrane bound, may reflect the distorted shape of cells on fibers of the prosthesis.

Fabrication of resorbable microporous intravascular stents for gene therapy applications.

The authors have produced resorbable, microporous endoluminal stents from Poly-L-lactic acid (PLLA)/Poly epsilon-caprolactone (PCL) blends. Both helical and tube stent designs have been obtained by solvent casting and flotation-precipitation fabrication techniques. A range of PLLA/PCL blend ratios and process variables were employed to investigate their influence on mechanical properties, porosity, and degradation rate. Polymer blends with higher PLLA proportions exhibit higher elastic moduli and ultimate tensile strength, and lower elongation, porosity, and degradation rates than do materials with higher PCL content. Stents with suitable mechanical properties for deployment and support of the vessel wall were obtained. Poly(ethylene oxide) was incorporated into these devices using an acid swelling technique, opening the pore structure and improving the hydrophilic character, thereby enabling the uptake of recombinant adenoviral vectors. The 50:50 PLLA/PCL blended stents were impregnated with recombinant adenovirus (AdCMB beta Gal, encoding a nuclear localizing variant of Escherichia coli beta-galactosidase). Cultured CV-1 cells incubated with stents impregnated with the recombinant virus expressed nuclear localized beta-galactosidase activity, confirming that absorbed virus is released from the matrix in an infectious form, with kinetics suggesting that genetically enhanced endovascular devices of this design are feasible.

Endothelial cell binding to Dacron modified with polyethylene oxide and peptide.

Polyethylene oxide (PEO) was incorporated into the surface of Dacron (PET) vascular prosthetic material (crimped Bionit I, BNl) followed by covalent attachment of an endothelial cell (EC) adhesion peptide; Gly-Arg-Glu-Asp-Val-Tyr (GREDVY). This procedure provides the possibility of a surface selective for EC adherence. Optimal PEO incorporation with minimal fiber damage was achieved from 78% (vol) trifluoroacetic acid (TFA) as characterized by scanning electron microscopy, nuclear magnetic resonance, bromphenol blue staining, and tensile testing. By weight, there was 4.4 times as much 18.5kD PEO as 1.5kD PEO incorporated into PET, but there were 2.8 times as many molecules of low molecular weight PEO. Attachment of 125I-GREDVY to substrates increased as follows: BNJ < BNI - 18.5 kD PEO < BNI - 1.5 kD PEO. Polyethylene oxide size affected EC binding with high molecular weight material decreasing binding and low molecular weight PEO increasing attachment. GREDVY modification of BNI or either of the two BNI-PEOs gave small but consistently increased EC binding compared with the same preparation without GREDVY. In contrast, human fibroblasts cultured from umbilical vein showed decreased binding when GREDVY was attached to BNI or PEO modified BNI. These results indicate that selective EC binding to PET vascular prostheses can be achieved.

Dipeptidyl peptidase IV and aminopeptidase in burn wound exudates: implications for wound healing.

Two catalytically active proteases, dipeptidyl peptidase IV (DP IV) and aminopeptidase (AP), not previously reported as present in burn wound exudates, have been identified by substrate specificity and susceptibility to known enzyme inhibitors. The ratio of the two enzymes in exudates is significantly different from the ratio in plasma collected from the same patient during the same time interval, suggesting that measurement of exudate components may be more significant than plasma activities in evaluating local conditions in the wound. A number of biologically significant substances are DP IV substrates, and the list can be considerably extended by the sequential action of AP and DP IV. Some polypeptides are converted to their biologically active form by DP IV action, while others are degraded to inactive forms. Either action generates X-Pro dipeptides, which have a demonstrably beneficial effect on wound healing. Although not resolved by molecular sizing or anion exchange chromatography, DP IV and AP in a burn wound exudate were purified by affinity chromatography.

Proteolytic activity in burn wound exudates and comparison of fibrin degradation products and protease inhibitors in exudates and sera.

Proteolytic (caseinolytic) activity in burn wound exudates was screened over the range pH 5.3 to 8.4. Although activity was greatest at pH 8.4 in four of seven exudates, individual differences indicated that different proteases predominate in the local environment of the wound. Paired exudate and serum samples were compared with regard to fibrin degradation products and three protease inhibitors: antithrombin III, a1-protease inhibitor, and a2-antiplasmin. Fibrin degradation products concentration was higher in exudates than in paired sera, indicating the wound as the source of circulating fibrin degradation products rather than intravascular coagulation followed by fibrinolysis. In contrast, all three protease inhibitors exhibited higher concentrations in serum than in the paired exudate. The serum/exudate ratio for AT III differed significantly from that for a1-protease inhibitor and a2-antiplasmin, and the ratio of two inhibitors in serum differed from the ratio of the same two inhibitors in the exudate in two of three comparisons. These findings emphasize the importance of exudate examinations as a reflection of events in the wound itself. The importance of microenvironments is invoked to account for the significant exudate fibrin degradation products titers, which are seen despite the presence of antithrombin III, which could inhibit coagulation, and the presence of a2-antiplasmin, which could inhibit fibrin degradation.

Elastase and alpha 1-protease inhibitor in burn wound exudates.

By degrading antithrombin III, polymorphonuclear neutrophil (PMN) elastase can become a procoagulant. Because intravascular coagulation may accompany severe burn injury, this study examined burn wound exudates for PMN elastase and its physiologic inhibitor, plasma alpha 1-protease inhibitor (alpha 1-PI), as a step in evaluating their contributions to coagulopathy in patients with burns. Each of the nine exudates examined were inhibitory for PMN elastase. Chromatographic characterization of the inhibitor indicated that it was alpha 1-PI; its elution volume for four exudates was identical to that of pure alpha 1-PI. Confirmation of the inhibitor's identity was achieved by reaction of anti-alpha 1-PI antibody with each exudate and with inhibitory chromatographic fractions of exudates with the most inhibitory activity. Inhibitor potency, determined from dose-response curves against a standard PMN elastase activity, varied twentyfold among exudates. Only one exudate had catalytic activity with the PMN elastase substrate. Although this enzyme had elastase-like properties, it appeared to differ from PMN elastase. The presence of alpha 1-PI in the wound exudate suggests that this inhibitor may act to diminish fibrin formation from the level that might otherwise have been seen if excess elastase were free to degrade antithrombin III.

Human renal carcinoma: asparagine independence with asparaginase susceptibility in culture.

A human renal carcinoma cell line (Caki-1) was examined for asparagine (Asn) dependence and susceptibility to Escherichia coli asparaginase. Because this enzyme hydrolyzes glutamine (Gln) as well as Asn, even though at only 2-3% the rate, Asn- Gln+ and Asn- Gln- media were prepared. Only the former supported Caki growth. The Asn- Gln- medium was then repleted with Asn, Gln, or both. Although Asn repletion failed to promote growth, addition of Gln alone or the combination supported growth as well as complete medium. With [3H]leucine and [3H]mannose incorporation to indicate protein and glycoprotein synthesis, respectively, the Gln repleted medium supported these processes as well as complete medium. Asparaginase added to complete medium was highly toxic to the Caki cells, but this is a reflection of Gln depletion rather than Asn depletion.

Modulation of secreted plasminogen activator activity of human renal carcinoma cells by dimethylsulfoxide, butyrate and retinoate.

Dimethylsulfoxide, butyrate and retinoic acid, agents which induce differentiation of certain malignant cells, were examined for their effect on the activity of plasminogen activator (PA) of serumless conditioned medium (CM) of two human renal carcinoma cell lines. All three agents produced a decrease in PA activity. More than 90% of the PA was secreted in latent form, and this was not altered by the agents. Active PA components of Mr 52,000 and 93,000 were identified in cell secretions by zymography. In the presence of DMSO or butyrate the Mr 52,000 component was markedly reduced. Reversibility of the effect on PA was demonstrated for both DMSO and retinoic acid with activity of CM returning to control level after removal of the agent. That the effect was temporary agrees with most observations of the effects of these and similar agents on cells from solid tumors.

Characterization of plasminogen activator from two human renal carcinoma cell lines.

Plasminogen activator (PA) activity was identified in the conditioned medium of two human renal carcinoma cell lines, Cur and Caki-1. PA activity of medium, following chromatography on Con A-Sepharose, was divided into effluent and eluate fractions, the latter obtained after elution with methyl mannoside. The ratio of PA activity in effluent:eluate was 90:10 for Caki-1 and 60:40 for Cur. The PA of both effluent fractions and the Caki-1 eluate fraction was of the urokinase (UK) type. Identification rested on molecular weight determination by zymography (major component with Mr 52,000 and a less prominent component of 93,000), lack of binding to fibrin, inhibition by anti-UK antibodies, and lack of inhibitory effect of anti-tissue type PA (TPA) antibodies or the Erythrina trypsin inhibitor, which inhibits TPA but not UK. PA of the Cur eluate fraction gave a more complex pattern in that it bound significantly to fibrin (like TPA), was completely inhibited by both anti-UK and anti-TPA antibodies, but was unaffected by Erythrina trypsin inhibitor. These results raise the possibility of an unusual PA-like enzyme that immunologically cross reacts with anti-UK and anti-TPA. Most of the PA of both cell lines was secreted in a latent form that could be activated by trypsin treatment. The latency appears to result largely from secretion of urokinase proenzyme, which is consistent with the Mr 52,000 of the major PA species and the insensitivity to diisopropyl fluorophosphate inhibition prior to trypsin activation. However, in addition, a UK binding component was found in the conditioned medium, which produced an Mr 93,000 component by reaction with UK.

Effect of dimethyldioctadecylammonium bromide induced macrophages on malignant cell proliferation.

Murine peritoneal macrophages elicited by dimethyldioctadecylammonium bromide (DDA), which is a potent immunologic adjuvant, were examined for cytotoxic and growth inhibiting activity for malignant cells. DDA macrophages had no cytolytic activity for murine B16BL-6 melanoma or human SMS-SB pre-B leukemia cells even in the presence of up to 1 microgram bacterial endotoxin (lipopolysaccharide, LPS)/ml. However, they exhibited a variable inhibitory effect on the growth of several lines of leukemia cells. The number of SMS-SB and human NALL cells remained essentially static in the presence of DDA macrophages while they increased significantly when cultured with resident macrophages. In contrast, L1210 cells increased 5-8-fold in the presence of macrophages elicited either by DDA or the inflammatory agent proteose peptone (PP). Although DDA macrophages retarded L1210 growth relative to PP macrophages, both populations responded to LPS in a comparable dose dependent manner to become essentially cytostatic at 1 microgram LPS/ml.

Effect of dimethylsulfoxide and butyrate on 5'-nucleotidase of human renal carcinoma cells.

The effect of dimethylsulfoxide (DMSO) and butyrate, agents which induce differentiation of certain cancer cells, on membrane associated 5'-nucleotidase (5'-NT) of 2 human renal carcinoma cell lines (Cur and Caki) was investigated. Under a variety of conditions of agent addition, 5'-NT specific activity increased in Cur and decreased in Caki cells. This opposite response pattern was observed for assays performed on lysates at pH 9.0 and 7.4 and assays with intact cell monolayers, even under conditions of identical cellular growth inhibition. It is concluded that the cell lines responded in a fundamentally different way to the chemical agents. An increase in 5'-NT has correlated with cell maturation for a number of processes. The DMSO induced increase in Cur 5'-NT was dependent on protein synthesis.

Studies with a human plasma-derived immunosuppressive, anti-lymphoma factor.

A low-molecular-weight (1,400) factor isolated from a human plasma alpha-globulin concentrate by acid-salt dissociation and ultrafiltration inhibits proliferation of mitogen-stimulated T cells and L1210 leukemia cells. The factor (UM05R) inhibits DNA, RNA, and protein synthesis in sensitive cells, acts in G1 of the cell cycle, and appears to suppress mitogen-responsive T cells without an accessory cell requirement. UM05R activity is enhanced by known cAMP-elevating agents and by sulfhydryl compounds. The results of the present study are consistent with the hypothesis that the plasma-derived agent inhibits lympho-proliferation as a result of elevation of intracellular cAMP.