An in vitro system for studying osteointegration of dental implants utilizing cells grown on dense hydroxyapatite disks.

A proposed in vitro system is described where chick osteoblasts are cultured on the flat surfaces of dense, nonporous HA disks to facilitate the study of bone formation at the cell-HA interface. During early bone formation cell-coated HA disks were retrieved, fixed with buffered 2% glutaraldehyde, and embedded in epon/araldite. The underlying HA disks were demineralized in diluted acid, and the intact cell-HA interfaces were re-embedded and thin sectioned for routine transmission electron microscopy. Morphologic studies indicated that osteoblasts proliferated and formed nodules of cells on the surfaces of HA disks. With increasing time in culture, they deposited orthogonally packed collagen fibrils between the cell layers that were enveloped by electron-dense mineralized globules. Eventually, small spicules of mineralized HA formed along collagen fibrils. An electron-dense layer about 50 nm thick was observed on the surface of the HA disks. Biochemical studies indicated that cell proliferation, as judged by 3H-thymidine uptake, increased rapidly during the first 3 days, reached a maximum around 6 days, and then declined by 12 days in culture. AP activity and collagen synthesis, as determined by 3H-hydroxyproline formation, increased as cellular proliferation declined. Mineralization, as judged by 45Ca uptake and spicule formation, occurred, as expected, following the increase in AP activity and deposition of densely packed collagen fibrils. Thus, all morphological and biochemical parameters studied indicate that the proposed in vitro system is reproducible and can facilitate the study of the osteointegration of HA-coated implants.

The effects of chlorhexidine on the attachment and growth of cultured human periodontal cells.

The purpose of this investigation was to evaluate the effects of chlorhexidine (CHX) on the attachment and growth of human gingival fibroblasts and periodontal ligament cells using an in vitro system where periodontal cells were grown on root surfaces treated previously with CHX. Our results indicate that the attachment of periodontal cells onto root surfaces was not adversely affected when roots were treated for 15 minutes with up to 0.12% CHX. However, cell attachment and morphology were adversely altered with prior 0.2 to 2.0% CHX treatment. The cells appeared round and retracted from roots treated with 0.2% CHX. With 2% CHX treatment, the cells exhibited a foamy appearance in which most of the cytoplasm seemed to have been extracted from the cells. Although 0.12% CHX treatment did not adversely affect the attachment of periodontal cells onto roots, direct exposure to as little as 0.01% CHX caused a 90% reduction in 3H-thymidine incorporation by cultured gingival fibroblasts. We conclude that although 0.12% CHX did not inhibit the attachment of cultured periodontal cells to pretreated roots, direct exposure of cells to much lower concentrations of CHX (0.0025 to 0.01%) caused a dose-dependent inhibition of growth.

Fibroblastic growth and attachment on hydroxyapatite-coated titanium surfaces following the use of various detoxification modalities. Part I: Noncontaminated hydroxyapatite.

Hydroxyapatite-coated titanium alloy test strips were treated with chlorhexidine gluconate, stannous fluoride, citric acid, tetracycline HCl, polymyxin B, hydrogen peroxide, and a plastic Cavitron tip: untreated sterile strips served as controls. The strips were incubated with cultured human gingival and periodontal ligament fibroblasts. Image analysis of three photomicrographs of each test strip (original magnification x350) indicated that the tetracycline HCl treatment resulted in significantly greater cellular surface area coverage compared with the other treatments. Citric acid and the plastic Cavitron tip also stimulated cell attachment, although the results from the Cavitron tip were not significantly different from citric acid or the other treatment groups. The remainder of the modalities and the untreated cellular controls experienced similar cellular coverage.

Fibroblastic growth and attachment on hydroxyapatite-coated titanium surfaces following the use of various detoxification modalities. Part II: Contaminated hydroxyapatite.

This study evaluated the ability of various chemotherapeutic and mechanical modalities to detoxify endotoxin-contaminated hydroxyapatite-coated dental implant surfaces as determined by the early attachment and growth of human gingival fibroblasts. Hydroxyapatite-coated test strips were contaminated with purified outer membranes of Escherichia coli and treated with citric acid, hydrogen peroxide, stannous fluoride, chlorhexidine gluconate, tetracycline HCl, polymyxin B, a plastic sonic scaler tip, or left untreated (contaminated and sterile controls). Human gingival fibroblasts were then seeded onto the test strips and incubated for 48 hours. The citric acid-treated strips showed greater cell growth than the other treatments. The plastic sonic scaler tip and the polymyxin B-treated samples exhibited greater cell coverage than the sterile control specimens. The use of citric acid and/or a modified plastic sonic scaler tip may be a valuable adjunct when surgical repair of an ailing hydroxyapatite-coated dental implant is contemplated.

Alterations in cell morphology and cytoskeletal proteins in gingival fibroblasts exposed to a Bacteroides gingivalis extract.

A soluble sonic extract (SSE) from Bacteroides gingivalis caused a dose-dependent inhibition of gingival fibroblast growth, reduced cell attachment and altered cell morphology. Most of its cytotoxic activity was destroyed by heating, indicating that the factor(s) was a protein rather than endotoxin. Cells, grown in the presence of, or on, root surfaces pretreated with 100-200 micrograms SSE/ml, partially retracted from the substratum and exhibited extensive surface blebbing and finger-like protrusions. Immunofluorescent staining showed that the morphological effects of Bacteroides gingivalis SSE are directed specifically at actin stress fibers and not microtubules of the cytoskeleton. Exposure to the SSE resulted in a dramatic relocalization of the bulk of F-actin from a fibrous form to a non-aggregated diffuse form. Disorganization of actin stress fibres occurred at concentrations of SSE that inhibited cell growth, but preceded any observable changes in cell attachment or morphology. The microtubular network remained intact, although it stained less intensely than that of controls. By contrast, Bacteroides intermedius SSE did not significantly influence growth, alter cellular morphology or affect the two cytoskeletal proteins.

Suppression of fibroblast growth by Bacteroides gingivalis endotoxin is not reduced by serum lipoproteins.

Previous studies indicated that Bacteroides and E. coli endotoxins caused a dose-dependent inhibition of human fibroblast growth. However, these endotoxins were relatively weak inhibitors of growth. Since cells were grown with serum, we questioned whether serum lipoproteins, which lessen endotoxin cytotoxicity in vivo, reduced the growth inhibitory effects of endotoxins in vitro. To determine whether serum lipoproteins reduced the growth inhibitory effects of endotoxins, logarithmically growing human periodontal cells were incubated with endotoxin and high density (HDL) or low density lipoprotein (LDL). Neither HDL nor LDL significantly reduced the initial growth inhibitory effects of B. gingivalis or E. coli endotoxins, as judged by 3H-thymidine incorporation. Even with prolonged exposure of up to 10 days in culture, HDL did not ameliorate growth inhibition by endotoxin, as determined by direct cell counts. We conclude that serum lipoproteins do not provide any significant protective effects against fibroblast growth inhibition by endotoxins in vitro.

Growth inhibitory effects of dimethyl sulfoxide and dimethyl sulfone on vascular smooth muscle and endothelial cells in vitro.

The growth of bovine aortic smooth muscle and endothelial cells was studied after exposure to dimethyl sulfoxide (DMSO) or its major metabolite, dimethyl sulfone (DMSO2). Both compounds caused a dose-dependent inhibition of cell growth as determined by [3H]thymidine incorporation and by counting the number of cells with time of exposure in culture. The IC50 of DMSO (concentration which produces 50% inhibition of growth) was 1% for smooth muscle cells and 2.9% for endothelial cells. Similarly, the IC50 of DMSO2 was also 1% for smooth muscle cells, but was 1.8% for endothelial cells. After a 4-d exposure to either compound, the growth inhibition of smooth muscle cells was completely reversible at 1%, partially reversible at 2 to 3% and completely irreversible at 4%. By comparison, inhibition of endothelial cell growth was completely reversible up to 4% of either compound. It is concluded that the growth of smooth muscle cells was similarly inhibited by DMSO and DMSO2, but that smooth muscle cells were more susceptible than endothelial cells to the growth inhibitory effects of these compounds. In addition, DMSO2 was a more potent inhibitor of cell growth than DMSO and its growth inhibition was less reversible than that produced by DMSO.

Growth inhibitory effects of endotoxins from Bacteroides gingivalis and intermedius on human gingival fibroblasts in vitro.

Purified endotoxin or lipopolysaccharide from Bacteroides gingivalis and Bacteroides intermedius caused a similar dose-dependent inhibition of growth of cultured human gingival fibroblasts as determined by 3H-thymidine incorporation and direct cell count. Approximately 200 micrograms/ml endotoxin caused a 50% reduction in 3H-thymidine uptake of logarithmically growing cells. Inhibition of growth was similar in cultures of fibroblasts derived from either healthy or diseased human gingiva. When examining the change in cell number with time of exposure in culture, the rate of proliferation was significantly suppressed during the logarithmic phase of growth. However, the cells recovered so that the rate of proliferation, although reduced, was sufficient to produce a cell density similar to the control cells with prolonged culture. The endotoxins were characterized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The profiles of the Bacteroides endotoxins were different. B. gingivalis endotoxin showed a wide range of distinct bands indicating a heterogeneous distribution of molecular species. Endotoxin from B. intermedius exhibited a few discrete low molecular weight bands, but the majority of the lipopolysaccharides electrophoresed as a diffuse band of high molecular weight material. The apparent heterogeneity of the two Bacteroides endotoxins and the similarity in growth inhibitory capacity suggest that growth inhibitory effects of these substances cannot be attributed to any polysaccharide species of endotoxin.

The absorption, metabolism and excretion of dimethyl sulfoxide by rhesus monkeys.

The absorption and excretion of dimethyl sulfoxide (DMSO) were studied in Rhesus monkeys (Macaca mulatta) given daily oral doses of 3 gms DMSO/kg B.W. for 14 days. DMSO and its major metabolite, dimethyl sulfone (DMSO2), were measured in serum, urine and feces by gas-liquid chromatography. DMSO was absorbed rapidly, reached a steady state blood level after 1 day and then was cleared from blood within 72 hrs after ending treatment. Serum DMSO declined in a linear fashion on semilogarithmic coordinates as described by second order kinetics. It had a half-life of 16 hrs. DMSO2 appeared in blood within 2 hrs and reached a steady state concentration after 4 days of treatment. DMSO2 was cleared from blood about 120 hrs after DMSO administration was stopped. Its half-life in blood was calculated to be 38 hrs. Urinary excretion of unmetabolized DMSO and DMSO2 accounted for about 60% and 16%, respectively, of the total ingested dose. Neither DMSO nor DMSO2 was detected in fecal samples. However, when added to fecal samples, DMSO was degraded rapidly. Although dimethyl sulfide (DMS) was not measured, some DMSO was metabolized to this compound because of the particular sweetness of breath of the monkeys. We conclude that the absorption of DMSO by monkeys is similar to that for humans, but that its conversion to DMSO2 and urinary elimination are more rapid in monkeys.

Inhibitory effect of periodontally diseased root extracts on the growth of human gingival fibroblasts.

Cementum shavings obtained from periodontally diseased and nondiseased areas of 100 removed, single-rooted teeth were extracted with either pyrogen-free water (PFW) for 5 minutes, 1 M citric acid for 5 minutes or 45% phenol-PFW for 90 minutes at 65 degrees C. The extracts were membrane-filtered, dialyzed exhaustively versus PFW, lyophilized, weighed and then dissolved in complete growth medium. The phenol-water or citric acid extracts of cementum shavings from periodontally diseased roots were positive for endotoxin by the limulus lysate assay (LLA). Pyrogen-free water extracts of diseased or phenol-water extracts of nondiseased cementum shavings were negative, or only slightly positive, respectively, for endotoxin by LLA. Media containing the various extracts were added to logarithmically growing cultures of human gingival fibroblasts (HGF). Separate cultures of HGF were exposed to Escherichia coli endotoxin at concentrations of 50, 100, 250 and 500 micrograms/ml to determine the growth-inhibitory effects of a known endotoxin. Cell growth was analyzed by measuring the incorporation of tritiated thymidine into cells. Suppression of HGF growth from 30 to 49% by E. coli endotoxin was concentration-dependent and linear over the concentration range of endotoxin tested. Pyrogen-free water extracts of diseased (endotoxin negative) or phenol-water extracts of nondiseased cementum shavings (slightly endotoxin positive) did not effect HGF growth. However, citric acid or phenol-water extracts of diseased cementum shavings (highly endotoxin positive) significantly suppressed HGF growth 58% and 61%, respectively. These results indicate that citric acid is effective in removing cytotoxic substances, presumably endotoxin, from cementum shavings and suggest that citric acid treatment is effective clinically in detoxifying periodontally diseased root surfaces.

Regulation of low density lipoprotein receptors by plasma lipoproteins from patients with abetalipoproteinemia.

Despite an absence of low density lipoproteins (LDLs) and chylomicron remnants from plasma, the rates of cholesterol synthesis or the number of LDL receptors expressed on freshly isolated cells from patients with abetalipoproteinemia are not markedly increased. These observations suggest that other lipoprotein particles present in the plasma of patients with abetalipoproteinemia may regulate LDL receptor activity and the rates of cellular cholesterol synthesis in this disorder. In the present report we have studied the effects of lipoprotein fractions from the plasma of normal subjects, patients with abetalipoproteinemia, and a patient with dysbetalipoproteinemia on the binding, internalization, and degradation of 125I-labeled LDL (125I-LDL) by cultured human fibroblasts. LDL from normal subjects or the high density lipoprotein fraction HDL2 from the plasma of patients with abetalipoproteinemia effectively down-regulated LDL receptor activity (greater than 50% inhibition at 20 micrograms of protein per ml). HDL2 from the plasma of patients with abetalipoproteinemia also effectively reduced the binding, internalization, and degradation of 125I-LDL by cultured human fibroblasts. 125I-HDL2 from the plasma of patients with abetalipoproteinemia was bound, internalized, and degraded by cultured human fibroblasts; this process was competitively inhibited by unlabeled normal LDL or HDL2 from abetalipoproteinemic plasma and was 1/6th to 1/8th times as high when 125I-HDL2 was incubated with fibroblasts from a patient with receptor-negative homozygous familial hypercholesterolemia. We conclude that lipoproteins present in the HDL2 fraction of plasma from patients with abetalipoproteinemia (which are relatively rich in apoprotein E) are effective regulators of LDL receptor activity in normal human fibroblasts. These in vitro findings may explain why the in vivo rates of cholesterol synthesis and the number of LDL receptors expressed on freshly isolated cells from patients with abetalipoproteinemia are not markedly increased.

Regulation of colony-stimulating activity production. Interactions of fibroblasts, mononuclear phagocytes, and lactoferrin.

Neonatal skin fibroblasts were cultured in supernatants of peripheral blood monocytes that had been cultured with and without lactoferrin. Granulocyte-monocyte colony-stimulating activity (CSA) was measured in supernatants of the fibroblast cultures with normal T lymphocyte-depleted, phagocyte-depleted, low density bone marrow target cells in colony growth (colony-forming unit granulocyte/macrophage) assays. Monocyte-conditioned medium contained a nondialyzable factor that enhanced by 17-50-fold the production of CSA by fibroblasts. The addition of lactoferrin to monocyte cultures reduced the activity of this monokine by 75-100%. Lactoferrin did not inhibit CSA production by monokine-stimulated fibroblasts. We conclude that under appropriate conditions human fibroblasts are potent sources of CSA, that the production of CSA by these cells is regulated by a stimulatory monokine, and that the production and or release of the monokine is inhibited by lactoferrin, a neutrophil-derived putative feedback inhibitor of granulopoiesis. We propose that the major role of mononuclear phagocytes in granulopoiesis is played not by producing CSA, but by recruiting other cells to do so, and that in the steady state, feedback regulation of neutrophil production may occur as a result of a mechanism that inhibits the recruitment phenomenon.

Dimethyl sulfoxide as a cholesterol-lowering agent in cultured fibroblasts exposed to low density lipoproteins.

Confluent cultures of human skin fibroblasts were exposed to medium containing high levels of low density lipoproteins (LDL-cholesterol equivalent to 400 micrograms per ml) and 0 or 2% dimethyl sulfoxide (DMSO). The uptake and accumulation of cellular cholesterol from LDL were reduced significantly (30%) in the DMSO-treated cells as compared to the controls. The reduction in cellular sterol was due almost exclusively to a significant decrease (50%) in cholesterol ester accumulation. Incubation of cells with 125I-labelled LDL showed clearly that DMSO did not act by increasing the secretion of cholesterol from the cell, but rather by significantly decreasing the binding, internalization and degradation of exogenous LDL. De novo synthesis of cholesterol from [14C]acetate was measured and found to correlate inversely with cellular sterol levels in either control or DMSO-treated cells.

The mass uptake of cholesterol ester from low density lipoproteins by cultured smooth muscle and adventitial cells of human aortas.

Cultured human smooth muscle and adventitial cells were incubated with human serum and low density lipoprotein (LDL) to study the uptake and accumulation of cholesterol ester from exogenous LDL. The cellular total cholesterol varied with the amount of LDL cholesterol in the medium. The cholesterol ester content increased 4-fold after 2 hours of incubation. A 6-fold rise occurred by 24 hours and continued to 72 hours. The cholesterol ester of the adventitial cells was markedly depleted by incubation with abetalipoproteinemic serum or with a lipid-depleted plasma fraction. By the use of 14C-labeled LDL free cholesterol in the incubation medium, we calculated that some 70-80% of the total accumulated cholesterol ester after 24 hours of incubation was derived from LDL cholesterol ester, and only 20-30% was synthesized by the cells. These studies demonstrated conclusively that human cells greatly increase their cholesterol ester mass after incubation with LDL.

Inability of serum from abetalipoproteinemic subjects to stimulate proliferation of human smooth muscle cells and dermal fibroblasts in vitro.

Serum from two patients with abetalipoproteinemia, a rare disorder of lipid metabolism characterized by the absence of chylomicrons and very low density and low density lipoproteins, did not stimulate the proliferation and growth of human smooth muscle cells or dermal fibroblasts in vitro as effectively as normal serum. The growth-promoting activity of this serum was comparable to that observed for lipoprotein-deficient plasma from normolipidemic subjects. Although the mitogenic effect of abetalipoproteinemic serum was improved with supplementation of low density lipoproteins, it was still about half the activity achieved with normal serum. However, the growth-promoting activity of this serum was completely restored to normal levels after addition of a lysate of normal platelets. In contrast, the mitogenic activity of lipoprotein-deficient plasma remained unchanged after addition of a lysate from abetalipoproteinemic platelets, whereas a similar supplementation of normal platelets completely restored its growth-promoting activity to normal. Thus, the inability of abetalipoproteinemic serum to promote growth appears to be due both to a deficiency of a platelet-releasable growth factor(s) and to the absence of serum lipoproteins.

Biosynthesis of type I and III collagens by cultured smooth muscle cells from human aorta.

The synthesis of collagen by human aortic smooth muscle cells was studied after incubating the cells with [3H]proline and [3H]glycine for 48 hr. The culture medium and cells were lyophilized and then digested with cyanogen bromide (CNBr) in 70% (wt/vol) formic acid. The resultant peptides were subjected to ion exchange chromatography on CM-cellulose and gel filtration on agarose. On the basis of the molar ratios of the alpha1(I)-CB8 and alpha1(III)-CB8 peptides of the alpha1(I) and alpha1(III) chains, approximately one quarter of the collagen synthesized by these cells was identified as type III and three quarters as type I. These data indicate that the smooth muscle can synthesize at least two types of collagen found in the arterial wall.

The smooth muscle cell. III. Elastin synthesis in arterial smooth muscle cell culture.

Primate arterial smooth muscle cells and skin fibroblasts were examined for their ability to synthesize elastin in culture. In the presence of the lathyrogen beta-aminopropionitrile, the smooth muscle cells incorporate [3H]lysine into a lysyl oxidase substrate that was present in the medium and associated with the cell layer. A component having a mol wt of 72,000 and an electrophoretic mobility similar to that of authentic tropoelastin was isolated from the labeled smooth muscle cells by coacervation and fractionation with organic solvents. In the absence of beta-aminopropionitrile, long-term cultures of smooth muscle cells incorporated [14C]lysine into desmosine and isodesmosine, the cross-link amino acids unique to elastin. In contrast, no desmosine formation occurred in the fibroblast cultures. These characteristics demonstrate that arterial smooth muscle cells are capable of synthesizing both soluble and cross-lined elastin in culture.