D (+) Glucosamine and D(+) glucose were purchased from Sigma Chemical Co. (St. Louis, MO). L- (4,5 ³H) Leucine, ³H-Thymidine, (³⁵S) as sulfate in aqueous solutions and (¹²⁵I) were from Amersham Life Science Corp (Arlington Heights, IL). Heparinase, heparitinase and chondroitinase ABC were purchased from Seikagaku America Inc (Bethesda, MD).

LDL (d < 1.063), Lp(a) (d = 1.11) and HDL (d = 1.1–1.23) were isolated from fresh plasma by sequential ultracentrifugation. For some experiments ¹²⁵I-LDL radio-iodinated using iodine monochloride 5 was used.

Bovine aortic endothelial cells were isolated and cultured as described 20. The cells (5–20 passages) were grown in minimal essential medium containing 10% FBS (Life-Technologies, Gaithersburg, MD).

For ¹²⁵I-LDL transport experiments, endothelial cells were grown in tissue culture inserts (Falcon-0.3 μm pore size) to facilitate its access to the upper (luminal) and lower (subendothelial) surface of endothelial cells. The barrier function of the endothelial cell monolayer was examined as previously reported 16, using [³H]dextran (average mol wt 150,000) and [¹⁴C]albumin. Transport of these molecules from the apical to the basolateral side of the monolayers was < 5%/h, a rate similar to that reported by others. At the conclusion of each LDL transport experiment, the monolayers were stained with 2% toluidene blue to verify the uniformity of the monolayer. On the day of the experiment, ¹²⁵I-labeled LDL was added to the upper chamber and following incubation radioactivity in the lower-chamber medium and associated with the extracellular matrix was determined. The total amount of each lipoprotein transported across the monolayers (net transport) is the sum of these two measurements. To determine ¹²⁵I-labeled LDL associated with the subendothelial cell matrix, the cells were incubated for 10 min with medium containing 50 U/ml of heparin (Elkins-Sinn, Inc., Cherry Hill, NJ)

Endothelial PG was radiolabeled with (³⁵S) sulfate along with the indicated doses of glucosamine for 12–16 h. Cellular PG were assessed by removing the cells with Triton X-100/NH₄OH. Endothelial cells were grown in 24 or 48 well plates (Falcon: Beckton Dickinson, Lincoln Park, NJ). Subendothelial matrix (SEM) was prepared as described 17. Briefly, confluent monolayers of endothelial cells were washed three times with PBS and incubated for 10 min in a solution containing 0.1% TritonX-100 and 20 mM NH₄OH at room temperature. Detached cells were removed by washing four times with PBS. This procedure leaves the intact matrix attached to the surface of the well. Matrix PG was extracted with 6 M guanidine HCl for 4 h at 4°C. For enzyme treatments, SEM was incubated with a mixture of heparinase and heparitinase (1 U/ml each) or chondroitinase ABC (1–2 U/ml) for 3 h at 37°C.

To prepare GAG chains, 1-ml aliquots of purified cell surface PG were treated with 100 μl of 10 N NaOH for 18 h at 26°C with constant shaking and then neutralized with 10 N HCl 18. To remove core proteins, the samples were adjusted to 7 M urea and loaded on 1-ml DEAE mini column that was previously equilibrated with 3 bed volumes of 7 M urea, 0.1% Triton X-100, 0.2 M NaCl in 0.05 M Tris, pH7.2. Peak fractions were pooled and dialyzed against 10 mM Tris and 0.1% Triton X-100 to remove urea. To determine GAG size, 0.6 ml of the above protein free GAG were chromatographed on Sephacryl-200 (Pharmacia-Biotech) gel-filtration column that was previously calibrated with known molecular weight standards using 0.2 M NaCl as an eluate.

THP-1 cells were purchased from the American Type Culture Collection (Rockville, MD) and were grown in RPMI 1640 (Gibco-Laboratories, Grand Island, NY) containing 10% fetal bovine serum (Gemini Bioproducts Inc., Calabasas, CA).

Adhesion assay was done as described previously 19. Monocytes were incubated in leucine-free DMEM-BSA medium before labeling. 100 μCi of (³H) leucine was added to 1 × 10⁷ cells and incubated for another 2 h under cell culture conditions. The labeled cells were centrifuged at 800 rpm for 5 min to remove the unincorporated label. The cells were washed three times and re-suspended in DMEM-BSA and then added to endothelial cell monolayers or to SEM in 24 well plates (24 × 10⁵ cells/well). Binding was performed for 2 h at 37°C. Unbound monocytes were removed by washing four times with DMEM-BSA and the bound radioactivity was extracted with 0.5 N NaOH for 1 h at 37°C and then counted.

Rat aortic SMCs were cultured as previously described 20 in basal medium supplemented with growth factors, bFGF, hEGF and 5% fetal bovine serum (Clonetics Corp, San Diego, CA). SMCs were plated in low density (9 × 10⁴ cells/ well) in 6 or 12 well plates and cultured in the presence or absence of 30 mM glucose or 2.5 mM glucosamine for three days. The cells were then trypsinized and an aliquot of trypsinate was counted for the final cell number with hemacytometer. Net growth was assessed by subtracting the final cell number from the initial cell number.

In other experiments, cells grown under above conditions were labeled with (³H)-thymidine (5 μCi/ml) for 6 h and the cells were washed four times with DMEM-BSA to remove unincorporated label. The cells were then lysed by 0.5 N NaOH and the thymidine incorporation into the DNA was assessed.

To determine whether glucosamine increases endothelial HS production, aortic endothelial cell proteoglycans were labeled with ³⁵S sulfate. Glucosamine treatment increased ³⁵SO₄ incorporation into the media PG by 2 fold and into matrix PG by 3 fold (Figure 1A). Addition of glucose (30 mM) to the medium did not affect PG production. Enzymatic analysis showed that the increase was found to be exclusively in HSPG and glucosamine treatment did not affect CS/DS PG in endothelial cells (Figure 1B). Perlecan is the major HSPG secreted by endothelial cells, we therefore tested if the increase was in perlecan. Real time PCR analysis showed a 1.9 fold increase in perlecan mRNA (Figure 1C bars) and consistent with this immunoprecipitation analysis showed that media from glucosamine treated cells contain 2.3 fold perlecan in medium (Figure 1C line). Thus these data suggest that glucosamine primarily increased endothelial cell perlecan. Highly sulfated blocks of HS are referred to as heparin-like HS, which confer several biological properties to HS. To identify heparin-like HS GAG, matrix HSPG were subjected to heparitinase digestion, followed by low pH nitrous acid digestion 1821. Heparitinase-resistant and nitrous acid-sensitive HS was increased by two fold (2300 cpm in control versus 4550 cpm in glucosamine treated) suggesting that glucosamine treatment of endothelial cells increased heparin-like HS. Glucose or glucosamine did not significantly increase macrophage PG but glucosamine showed a moderate increase in SMC HSPG (Figure 2).

Subendothelial matrix HSPG is thought to play a key role in endothelial barrier function, however, whether decreased HSPG increases lipoprotein transport across endothelial monolayers is not known. We first tested whether removal of HSPG increases LDL transport across endothelium. Removal of HSPG by heparinase treatment led to a 2.1 fold increase in ¹²⁵I-LDL transport at 10 min. (Figure 3A). At 15 min and 30 min the increase in LDL transport was 57% and 36% higher than controls. Conversely, after glucosamine treatment, ¹²⁵I-LDL transport across the EC monolayers decreased by 15–22% at different time points in glucosamine treated cells (Figure 3B).

We previously showed that removal of HSPG increases monocyte retention in the subendothelial matrix 19. We determined THP-1 monocyte to the subendothelial matrix prepared from control, glucose and glucosamine-treated cells. Monocyte binding to the glucosamine stimulated endothelial cells was decreased by 52% compared to non-stimulated cells (Fig. 3C). In contrast monocyte binding to glucose-treated endothelial cells was slightly increased similar to previous observations (20%, not shown ref 22).

We next determined the effects of glucose and glucosamine on SMC proliferation. Sub-confluent SMC were incubated in media containing 30 mM glucose or 2.5 mM glucosamine for 48 h and net growth was determined. Glucose treatment did not alter SMC growth as assessed by cell number (Figure 4A) or thymidine incorporation (Figure 4B). In contrast, the cell number and ³H thymidine incorporation into the DNA was decreased by 60% by glucosamine treatment (4A and B). The major HSPG secreted by SMC is perlecan, which was known to negatively correlate with SMC growth 11. Immunoprecipitation analysis showed a 2.5 fold increase in perlecan protein in SMC treated with glucosamine (not shown). We next tested whether perlecan mediates the antiproliferative effect of glucosamine on SMC. Addition of an anti-perlecan antibody completely abolished the antiproliferative effect of glucosamine on SMC proliferation (Figure 5). These results suggest that glucosamine treatment increases perlecan production, which in turn modulates SMC growth in proliferating cells. Consistent with a perlecan-based mechanism, glucosamine did not inhibit endothelial cell growth (Figure 4C)

We tested glucosamine effects on HSPG and atherosclerosis. When given intraperitoneally (5 mg/kg once a day for 3 days) glucosamine increased ³⁵S sulfate incorporation into HSPG in liver (by 61%) and heart (by 82%) (Figure 6). To determine the effects of glucosamine on atherosclerosis, apoE null mice were treated with glucosamine at 5 mg/kg for 2 months. Plasma glucose and total cholesterol was not affected by glucosamine (not shown). Oil Red-O staining revealed 30% reduction in lesions in glucosamine treated group (p < 0.05) (Figure 7). These data suggest that macrovascular disease associated with hyperglycemia is unlikely due to glucosamine. In fact, glucosamine by inducing HSPG showed atheroprotective effects.