Increased protein O-GlcNAc modification inhibits inflammatory and neointimal responses to acute endoluminal arterial injury

Abstract

Inflammation plays a major role in vascular disease. We have shown that leukocyte infiltration and inflammatory mediator expression contribute to vascular remodeling after endoluminal injury. This study tested whether increasing protein O-linked-N-acetylglucosamine (O-GlcNAc) levels with glucosamine (GlcN) and O-(2-acetamido-2-deoxy-d-glucopyranosylidene) amino-N-phenylcarbamate (PUGNAc) inhibits acute inflammatory and neointimal responses to endoluminal arterial injury. Ovariectomized rats were treated with a single injection of GlcN (0.3 mg/g ip), PUGNAc (7 nmol/g ip) or vehicle (V) 2 h before balloon injury of the right carotid artery. O-GlcNAc-modified protein levels decreased markedly in injured arteries of V-treated rats at 30 min, 2 h, and 24 h after injury but returned to control (contralateral uninjured) levels after 14 days. Both GlcN and PUGNAc increased O-GlcNAc-modified protein levels in injured arteries compared with V controls at 30 min postinjury; the GlcN-mediated increase persisted at 24 h but was not evident at 14 days. Proinflammatory mediator expression increased markedly after injury and was reduced significantly (30–50%) by GlcN and PUGNAc. GlcN and PUGNAc also inhibited infiltration of neutrophils and monocytes in injured arteries. Chronic (14 days) treatment with GlcN reduced neointima formation in injured arteries by 50% compared with V controls. Acute GlcN and PUGNAc treatment increases O-GlcNAc-modified protein levels and inhibits acute inflammatory responses in balloon-injured rat carotid arteries; 14 day GlcN treatment inhibits neointima formation in these vessels. Augmenting O-GlcNAc modification of proteins in the vasculature may represent a novel anti-inflammatory and vasoprotective mechanism.

inflammation plays an important role in the pathogenesis of many forms of vascular disease, including responses to acute vascular injury. Previous studies, including our own, have shown that inflammatory mediator expression and leukocyte infiltration in injured vessels contribute to vascular remodeling after endoluminal injury (4, 7, 13, 21, 27, 32, 33, 43). We have shown that adhesion molecules [P-selectin, intercellular adhesion molecule (ICAM)-1 and vascular cell adhesion molecule (VCAM)-1)] and chemokines that are selective for neutrophils (cytokine-induced neutrophil chemoattractant (CINC)-2β and monocytes [monocyte chemotactic protein (MCP)-1] are expressed at high levels within 2 h after balloon injury of the rat carotid artery (32, 33). In turn, these mediators activate and stimulate migration of large numbers of granulocytes and monocytes into the arterial wall from the periadventitial tissues at 24 h postinjury (33, 43). These early inflammatory responses correlate with the extent of subsequent neointima formation (32, 33, 43). We have previously demonstrated that this cascade of events is inhibited by treatment with 17β-estradiol by an estrogen receptor-dependent mechanism (3, 42).

Glucosamine (GlcN) is an amino sugar that can stimulate O-linked-N-acetylglucosamine (O-GlcNAc) modification of proteins by increasing flux through the hexosamine biosynthesis pathway, thus increasing production of UDP-GlcNAc. UDP-GlcNAc is a substrate for O-GlcNAc transferase (OGT), which catalyzes the O-linked addition of GlcNAc to serine and threonine residues of nucleocytoplasmic proteins in higher eukaryotes (16, 19, 24). GlcN has anti-inflammatory effects in a variety of inflammatory models and cell types (9, 10, 12, 14, 17, 22, 23, 26, 31, 35, 47). GlcN has been shown to ameliorate parameters of disease in experimental models of adjuvant and rheumatoid arthritis in rat, to prolong cardiac allograft survival, and to suppress experimental autoimmune encephalomyelitis in mice (12, 23, 31, 47).

O-(2-acetamido-2-deoxy-d-glucopyranosylidene)amino-N- phenylcarbamate (PUGNAc) is a competitive inhibitor of the O-linked N-acetylglucosaminidase (O-GlcNAcase) that catalyzes the removal of O-GlcNAc from proteins (20). It has been shown that in vitro or in vivo administration of PUGNAc significantly increases O-GlcNAc levels in cardiac myocytes and intact heart and improves cardiac function after ischemia-reperfusion injury (8, 29, 23a). In a rat trauma-hemorrhage model, PUGNAc treatment deceased circulating interleukin (IL)-6 and tumor necrosis factor (TNF)-α levels and improved cardiac output, perfusion of kidney and brain, and survival rates (48). Thus there is proof of principle that PUGNAc treatment has cardioprotective and anti-inflammatory effects in some experimental settings.

The present study tested the hypothesis that systemic treatment with GlcN, which increases O-GlcNAc modification of proteins by increasing flux through the hexosamine biosynthesis pathway, and PUGNAc, which increases O-GlcNAc modification of proteins by inhibiting O-GlcNAcase, can inhibit acute inflammatory and neointimal responses to endoluminal arterial injury in the rat carotid artery.

METHODS

Animals

Ten-week-old female Sprague-Dawley rats were obtained from Charles River Breeding Laboratories and maintained at constant humidity (60 ± 5%), temperature (24 ± 1°C), and light cycle (6:00 A.M. to 6:00 P.M.) and fed a standard rat pellet diet ad libitum. All protocols were approved by the Institutional Animal Care and Use Committee at the University of Alabama at Birmingham and were consistent with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health.

Experimental Protocols

All rats were subjected to ovariectomy (OVX); 5 days later, rats were randomized to receive daily injections of GlcN (0.3 mg·g⁻¹·day⁻¹ ip; Sigma-Aldrich), PUGNAc (7 nmol/g ip; Sigma-Aldrich), or vehicle (saline, V). Two hours after the first injection of GlcN, PUGNAc or V, rats were anesthetized with 80 mg/kg ketamine and 5 mg/kg xylazine and subjected to balloon injury of the right common carotid artery (11). The uninjured left carotid artery served as a control.

Rats were killed with an overdose of pentobarbital sodium (100 mg/kg ip), and carotid arteries were removed and processed at 30 min postinjury (n = 4–5/group) for measurement of total O-GlcNAc-modified proteins by Western blot analysis (27–29); at 2 h postinjury (n = 5/group) for measurement of inflammatory mediator mRNA levels by real-time RT-PCR or measurement of O-GlcNAc-modified proteins; at 6 h postinjury (n = 6/group) for measurement of inflammatory mediator (32, 33) and O-GlcNAc-modified protein levels; and at 24 h postinjury (n = 5/group) for detection of infiltrating leukocytes and cell proliferation by immunohistochemistry (33, 43) or measurement of O-GlcNAc-modified proteins. No attempt was made to remove the adventitia or perivascular connective tissue from any artery.

Separate groups of rats (n = 9–10/group) received daily injections of GlcN (0.3 mg/g ip) or V. On day 14 of treatment, 2 h after the last GlcN or V injection, rats were killed with an overdose of pentobarbital sodium, as above, and perfused with 10% formalin at a pressure of 120 mmHg. Both carotid arteries were removed, fixed, embedded in paraffin, serially sectioned, and stained with Verhoeff's elastin stain for morphometric analysis of the vascular injury response (11). Arteries from additional groups of animals were assayed for O-GlcNAc-modified proteins by Western blot. Blood samples (200 μl) were obtained from tail veins of these rats at 7 and 13 days postinjury to measure plasma glucose levels under conscious and nonfasting conditions using the StanbioGlucose Liquicolor kit (Fisher Scientific). Blood samples (5 ml) were also collected by direct cardiac puncture for quantification of serum levels of insulin, total cholesterol (TC), triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C) at the time of death. Serum insulin concentrations were measured using a radioimmunoassay kit (Linco Research). Serum TC, TG, and HDL-C levels were determined by an automated Stanbio Sirrus* Clinical Chemistry Analyzer.

In a separate protocol, blood samples (500 μl) were drawn from a femoral artery catheter at 15 min and 1, 2, 6, and 24 h post-single injection of GlcN (0.3 mg·g⁻¹·day⁻¹ ip). Plasma GlcN levels were determined by HPLC (1, 34).

Experimental Methods

O-GlcNAc protein in carotid artery.

Total modified O-GlcNAc protein levels were determined as previously described via immunoblotting with the anti-O-GlcNAc antibody CTD110.6 (28–30). Solubilized protein (30 μg) was separated by 7.5% SDS-PAGE and transferred to a polyvinylidenedifluoride membrane. Blots were incubated with a 1:5,000 dilution of CTD110.6 (Covance) in 1% casein and PBS overnight at 4°C and then washed three times in PBS. The membranes were then incubated with a 1:10,000 dilution of horseradish peroxidase-conjugated goat anti-mouse IgM (Santa Cruz) in 1% casein and PBS for 1 h at room temperature. After further washing in PBS, the immunoblots were developed with enhanced chemiluminescence (Pierce), and the signal was recorded on X-ray film. Densitometric analysis was performed on the entire lane of each sample using Labworks Analysis Software (UVP), and the mean intensity was normalized to the control group. Protein loading was assessed by stripping the membranes and reprobing with anti-β-actin antibody (Sigma).

Real-time quantitative RT-PCR analysis of inflammatory mediators.

RNA was extracted from arteries, reverse transcribed to cDNA and amplified by PCR with specific primers, and quantified using the iCycler (Applied Biosystems) (31). Levels of specific mRNAs were normalized using ribosomal protein S9 mRNA, since expression of this housekeeping gene has been shown to remain stable in a rat carotid artery injury model (37). Results were standardized to the mean value of the OVX + V injured group.

Protein levels of cytokines/chemokines.

Arteries were homogenized with a glass homogenizer in 200 μl of lysis buffer with protease inhibitor cocktail (Pierce) and 40 μM PUGNAc. After centrifugation at 12,000 g for 15 min, supernatants were collected, and total protein concentration was quantified by the Coomassie Plus Protein Assay (Pierce). Protein levels of cytokines/chemokines were quantified by multiplexed sandwich enzyme-linked immunosorbent assay (Searchlight rat cytokine array; Pierce) (31) and normalized to total protein levels.

Immunohistochemical analysis of leukocyte infiltration and cell proliferation.

The avidin-biotin-peroxidase immunohistochemical technique was used to detect neutrophils and monocytes in paraffin-embedded sections of carotid artery using a Vector Laboratories kit (Burlingame) (32, 42). Rat neutrophils and monocytes were recognized by specific primary antibodies against HIS48 (Santa Cruz Biotechnology) and ED1 (Serotec), respectively. Proliferating cells were identified by anti-Ki67 antibody staining (Vector Laboratories). Apoptotic cells were identified by TUNEL staining with an ApopTag Peroxidase In Situ Apoptosis Detection Kit (Chemicon).

Morphometric analysis of carotid artery.

Morphometric analysis of representative cross-sectional photomicrographs from midthirds of injured and contralateral control carotid arteries was performed with a computer-based Bioquant II Morphometric system. At least 5 sections from the middle third of each vessel were examined, and the measurements were averaged for statistical analysis (11). The cross-sectional surface areas of the vessel within the external elastic lamina (total area), within the internal elastic lamina (intimal area), and within the lumen (lumen area) were measured. Neointima formation in the injured artery was expressed as the absolute area of neointima and the neointima-to-media ratio.

Statistical Analysis

Results were expressed as means ± SE. Data were evaluated by one-way or two-way ANOVA. When the overall F test result of ANOVA was significant, a multiple-comparison Dunnett's test was applied. Student's t-test was used in two-mean comparisons. Differences were reported as significant when p was <0.05.

RESULTS

Western blot analysis with the anti-O-GlcNAc antibody CTD110.6 showed multiple bands, indicating the wide variety of proteins modified by O-GlcNAc in both uninjured and injured arteries of V-treated rats (Fig. 1). At 30 min, 2 h, and 24 h postinjury, total O-GlcNAc levels decreased significantly in injured right carotid arteries compared with uninjured left carotid arteries in V-treated animals (Fig. 1). At 14 day postinjury, there were no significant differences in O-GlcNAc levels between injured and uninjured arteries in V-treated rats. Both GlcN and PUGNAc treatment increased O-GlcNAc levels significantly in injured carotid arteries at 30 min postinjury (P < 0.05) (Fig. 2). Densitometric analysis of immunoblots of 24-h injured arteries revealed significantly increased O-GlcNAc levels in GlcN-treated compared with V-treated rats [CTD/β-actin = 4.6 ± 1.1 (n = 4) vs. 1.6 ± 0.4 (n = 4) (P < 0.5)]. By 14 days postinjury, O-GlcNAc levels did not differ between the GlcN (2.4 ± 0.4; n = 3) and V (2.2 ± 0.7; n = 4) (P = not significant) groups. Neither treatment resulted in significant changes in O-GlcNAc levels in uninjured carotid arteries (data not shown).

Fig. 1.

Representative immunoblot of uninjured and injured carotid arteries from ovariectomized (OVX) + vehicle (V)-treated rats at 30 min, 2 h, 6 h, 24 h, and 14 days post-balloon injury with CTD110.6. Bar graph shows the mean ± SE intensities quantified by densitometric analysis of the immunoblots. Data are presented as ratio of intensities of CTD/β-actin. The sample size (no. of rats) is indicated. *P < 0.05 vs. uninjured left common carotid arteries (LCA) from V-treated rat, respectively.

Fig. 2.

Representative immunoblot of injured carotid arteries (CA) from OVX + V-, OVX + glucosamine (GlcN)-, and OVX + O-(2-acetamido-2-deoxy-d-glucopyranosylidene) amino-N-phenylcarbamate (PUGNAc)-treated rats at 30 min postballoon injury with CTD110.6. Bar graph shows the mean ± SE intensities quantified by densitometric analysis of the immunoblots. Data are presented as ratio of intensities of CTD/β-actin. The sample size (no. of rats) is indicated. #P < 0.05 vs. injured CA from OVX + V.

Real-time quantitative RT-PCR analysis of 2-h control and injured carotid arteries from OVX + V rats showed that all mediators were expressed at very low levels in uninjured vessels and increased markedly after injury (Fig. 3). GlcN and PUGNAc treatment resulted in significant reductions in mRNA levels of the adhesion molecules P-selectin (by 43 and 41%, respectively) and VCAM-1 (by 38 and 43%, respectively) and the neutrophil-selective chemokine CINC-2β (by 48 and 49%, respectively). GlcN treatment resulted in a significant 57% reduction in mRNA level of the monocyte-selective chemokine MCP-1, whereas PUGNAc had no significant effect.

Fig. 3.

Effects of GlcN and PUGNAc on mRNA expression of chemokines and adhesion molecules in injured and contralateral uninjured control arteries at 2 h postinjury. mRNA levels measured by real-time RT-PCR were first normalized using ribosomal protein S9 (RpS9) mRNA to correct for differences in total RNA loading and then standardized to the mean value of the OVX + V-injured group, which was assigned a value of 100. Results are expressed as means ± SE. CINC, cytokine-induced neutrophil chemoattractant; MCP, monocyte chemoattractant protein; VCAM, vascular cell adhesion molecule; ICAM, intercellular adhesion molecule. P < 0.05 vs. uninjured (*) and vs. OVX + V group (#).

Protein levels of CINC-2α and MCP-1 were quantified in 6-h injured and control injured arteries of GlcN- and V-treated rats (Fig. 4). In uninjured arteries, CINC-2α and MCP-1 were expressed at very low levels (6.3 pg/ml and 0.782 ng/mg protein, respectively). Expression of these chemokines was not affected by GlcN treatment in uninjured vessels. Expression of CINC-2α and MCP-1 was greatly amplified (19- and 24-fold, respectively) at 6 h after injury in V-treated rats. GlcN treatment significantly reduced CINC-2α and MCP-1 levels (by 35 and 23%, respectively) in injured arteries.

Fig. 4.

CINC-2α and MCP-1 in injured and uninjured CA from OVX + V and OVX + GlcN rats at 6 h after balloon injury by multiplexed sandwich enzyme-linked immunosorbent assay. Results are expressed as means ± SE. P < 0.05 vs. uninjured (*) and vs. injured arteries from the OVX + V group (#); n = 6/group.

Immunohistochemical staining of injured carotid arteries demonstrated large numbers of HIS48⁺ granulocytes and ED1⁺ monocytes in the adventitial domains of 24-h injured arteries of V-treated rats (305 ± 15 and 402 ± 25/mm², respectively); granulocyte and monocyte numbers were greatly reduced by GlcN and PUGNAc treatment (112 ± 41 and 150 ± 44/mm² for granulocytes and 215 ± 46 and 270 ± 38/mm² for monocytes, respectively, P < 0.05) (Fig. 5). Infiltrating leukocytes could not be detected in the media of injured arteries in either treatment group. Furthermore, infiltrating inflammatory leukocytes could not be detected in any domain of 14-day injured arteries in either treatment group.

Fig. 5.

Representative neutrophil and monocyte specific antibody-stained sections of uninjured LCA and injured right common carotid arteries (RCA) from OVX + V, OVX + GlcN, and OVX + PUGNAc rats at 24 h postinjury. Perfusion-fixed and sectioned arteries were immunostained using primary antibodies for rat neutrophils (HIS48) and monocytes (ED1). Results localize infiltrating leukocytes to the adventitial and perivascular regions of 24-h injured CA.

Ki67⁺ proliferating cells were found in the adventitial domain of 24-h injured arteries in both groups, and there was no difference in Ki67⁺ cell numbers between GlcN-treated and V-treated (110 ± 28 Ki67⁺ vs. 120 ± 16 Ki67⁺ cells/mm², respectively) (data not shown) groups. In 14-day injured arteries, there were rare Ki67⁺ cells in the leading edge of the neointimal domain in both treatment groups.

Apoptotic cells were also found mainly in the adventitial domain of 24-h injured arteries in both groups, and there was no difference in TUNEL-stained cell numbers between GlcN-treated and V-treated groups (data not shown). At 14 days postinjury, TUNEL-stained cells could not be detected in any domain of injured arteries in either treatment group.

Morphometric analysis showed that at 2 wk post-balloon injury, the neointimal areas and neointima-to-media area ratios of injured carotid arteries were significantly less in OVX rats treated with GlcN compared with those treated with V (Fig. 6). Neointimal areas and neointima-to-media ratios of injured carotid arteries from V-treated group were 0.1 ± 0.01 mm² and 0.86 ± 0.1, respectively, and were reduced in GlcN-treated rats to 0.05 ± 0.02 mm² and 0.43 ± 0.13, respectively. The medial areas were the same in both treatment groups. The intima-to-media ratios closely reflected the absolute intimal areas and were reduced by ∼50%.

Fig. 6.

Top: representative light micrographs of RCA from OVX + V and OVX + GlcN (0.3 mg·g⁻¹·day⁻¹ ip) rats at 14 days after balloon injury. Arterial sections were stained with Verhoff's elastin stain. Bar represents 200 μm. Bottom: effects of administration of GlcN on neointima area (top), media area (middle), and the neointima-to-media ratio (bottom) of injured CA in OVX rats at 14 days after CA balloon injury. Data are shown as means ± SE, and the sample size (no. of rats) is indicated. *P < 0.05 vs. OVX + V group.

As summarized in Table 1, GlcN treatment did not alter body weight compared with V treatment. Plasma glucose and serum insulin levels did not differ in GlcN-treated compared with V-treated groups, indicating that there was no evidence that insulin resistance or diabetes was induced by daily injection of GlcN. GlcN treatment had no effect on serum TC, TG, or HDL-C levels in OVX rats at 2 wks postinjury.

Table 1.

Effects of administration of GlcN for 2 wk on body weight, plasma glucose, serum insulin, TC, TG, and HDL-C levels after balloon injury of right common carotid arteries of ovariectomized female Sprague-Dawley rats

Groups	Body Wt, g			Glucose, mg/dl			Insulin, ng/dl	TC, mg/dl	TG, mg/dl	HDL-C, mg/dl
	0	1 wk	2 wk	0	1 wk	2 wk
OVX + V (n = 10)	351.9±11.2	340.7±8.6	345.4±9.8	NA	135.6±5.4	123.6±3.6	3.58±0.5	100.6±6.9	73.6±5.5	75.6±4.1
OVX + GlcN (n = 10)	356.9±13.1	350.6±10.9	349.3±10.9	NA	141.1±11	127.3±4.2	4.25±0.6	106.8±7.0	63.2±2.9	76.6±5.2

Results are means ± SE; n, no. of rats. TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; OVX, ovariectomy; V, vehicle; GlcN, glucosamine (0.3 mg·g⁻¹·day⁻¹ ip). Data for insulin, TC, TG, and HDL-C are for 2 wk. NA, not available.

Plasma GlcN concentration was 1.9 ± 0.3 mM at 15 min, 0.6 ± 0.1 mM at 1 h, and 0.05 ± 0.03 mM at 2 h after intraperitoneal injection of GlcN. At 6 and 24 h postinjection, plasma GlcN levels were undetectable (data not shown).

DISCUSSION

In this study, we have shown for the first time that GlcN and PUGNAc treatment in doses that produce significant increases in O-GlcNAc-modified proteins decreases expression of inflammatory mediators and infiltration of leukocytes at a very early time point following endoluminal carotid artery injury. We also demonstrated major decreases in O-GlcNAc-modified protein levels in injured arteries compared with uninjured contralateral controls at early time points (30 min to 24 h) after injury in V-treated rats, indicating that the stress of endoluminal injury was associated with depletion of the O-GlcNAc-modified protein pool. O-GlcNAc-modified protein returned to the same levels as in uninjured arteries by 14 day postinjury, indicating the dynamic nature of O-GlcNAc protein modification in response to acute vascular injury. Furthermore, chronic (14-day) treatment with daily intraperitoneal injections of GlcN in the same dose produced transient (2 h) elevations in plasma GlcN concentration and significantly inhibited neointima formation in injured arteries. GlcN-treated animals showed no evidence of insulin resistance or diabetes.

The finding that GlcN and PUGNAc treatment resulted in increases in O-GlcNAc levels in injured arteries compared with injured arteries of V controls suggests that the anti-inflammatory and vasoprotective effects of the interventions may be mediated via post-translational O-GlcNAcylation. Results from this study are consistent with the protection afforded by increasing O-GlcNAc levels in isolated cells and organs in other experimental settings and extend the potential application into the much more complex setting of an in vivo model of vascular injury. The mechanisms underlying this protection, including identification of specific protein targets, have yet to be elucidated.

The current finding of an anti-inflammatory effect in the setting of vascular injury adds to the large body of evidence that GlcN inhibits inflammation and related pathological processes in animal models and humans. GlcN has been shown to suppress expression of the proinflammatory mediators IL-6 and cyclooxygenase-2 in human chondrocytes (26, 35), to inhibit nuclear factor (NF)-κB activation and IL-1β bioactivity in rat chondrocytes (17), to downregulate TNF-α-induced expression of ICAM and inhibit nuclear translocation of the p65 subunit of NF-κB in human retinal pigment epithelial cells (10), to suppress neutrophil functions such as superoxide generation, phagocytosis, granule enzyme release, and chemotaxis (22), and to inhibit CD3-induced T cell activation (14). Furthermore, when employed in a murine model of cardiac allograft rejection, GlcN suppressed leukocyte activation and nearly doubled allograft survival, results that were similar to conventional immunotherapy (31).

Specific to cardiovascular stress, it has been shown that increasing O-GlcNAc levels in cardiac myocytes in vitro or in whole heart ex vivo or in vivo with either GlcN or PUGNAc is cardioprotective in the setting of ischemia-reperfusion injury (8, 15, 28–30, 23a). In neonatal rat cardiomyocytes, either GlcN or PUGNAc treatment significantly increased O-GlcNAc levels and improved cell viability, as well as reduced necrosis compared with untreated cells following in vitro ischemia-reperfusion (8). In the isolated perfused rat heart, administration of either GlcN or PUGNAc during reperfusion improved functional recovery and decreased tissue injury (28–30). In an in vivo ischemia-reperfusion mouse model, it has been shown that lethal oxidant stress of cardiac myocytes produced a time-dependent loss of cellular O-GlcNAc levels; PUGNAc treatment largely reversed this pathologic response and reduced myocardial infarct size following in vivo ischemia-reperfusion (23a). It has also been shown that both GlcN and PUGNAc improved cardiac function and organ perfusion and reduced the levels of circulating IL-6 and TNF-α in association with increased O-GlcNAc levels in the heart, liver, and kidney of rats subjected to trauma-hemorrhage (44, 48).

The mechanism by which GlcN and PUGNAc act as anti-inflammatory agents is likely linked to augmentation of O-GlcNAc-modified protein levels. O-GlcNAcylation is increasingly recognized as an important regulatory mechanism involved in signal transduction (18, 39–41). A number of studies have demonstrated that increased levels of O-GlcNAc are associated with increased cell survival in response to a range of different stress stimuli (37, 46). Increasing O-GlcNAcylation by overexpressing OGT has been shown to protect cells against experimental stressors and increase their survival (46), whereas decreasing O-GlcNAc levels by inhibiting glutamine-fructose 6-phosphate amidotransferase, which regulates the entry of glucose in the hexasamine biosynthesis pathway, has been shown to reduce cell survival following heat stress (37).

In addition to increasing O-GlcNAc levels, GlcN also increases UDP-GlcNAc levels, which is used for multiple N-glycosylation reactions that are involved in protein synthesis. Glucosamine 6-phosphate could potentially be metabolized to fructose 6-phosphate, thereby increasing glycolytic flux. Thus the protection against vascular injury seen here with GlcN treatment could potentially be mediated via a number of other pathways.

One potential limitation of this study is that we have not identified a specific pathway that is modulated by increased O-GlcNAc levels; however, activation of the NF-κB pathway is clearly implicated in the response to acute vascular injury in animal models (5, 6, 25). Furthermore, blockade of NF-κB with antisense p65, an NF-κB decoy, or by overexpression of inhibitory factor κBα has been shown to inhibit inflammatory responses in injured arteries (2, 36, 49). Of relevance to this study is that GlcN has been shown to inhibit NF-κB activation and expression of NF-κB-dependent cytokines such as IL-1β and IL-6 in rat and human chondrocytes (26, 35). GlcN also inhibited TNF-α-stimulated nuclear translocation of the p65 subunit of NF-κB and expression of ICAM-1 in human retinal pigment epithelial cells (10). These findings have led us to postulate that O-GlcNAc modification of components of the NF-κB signal pathway may contribute to the GlcN-mediated anti-inflammatory and vasoprotective effects seen here. Future studies will examine whether specific NF-κB signaling molecules are targets for O-GlcNAc modification and whether this is responsible for the vasoprotective effects of GlcN in the setting of acute vascular injury.

In conclusion, we have shown that, in the rat carotid artery balloon injury model, administration of both GlcN, a nontoxic and widely available nutritional supplement, and PUGNAc, an inhibitor of O-GlcNAcase, results in increased total protein O-GlcNAc modification and attenuates early inflammatory response in arteries subjected to acute endoluminal injury. Furthermore, daily administration of GlcN at a dose that does not cause insulin resistance or diabetes inhibits late vascular remodeling (neointima formation) in injured arteries. These data support the notion that increasing O-GlcNAc levels whether by activation of the hexosamine biosynthesis pathway or inhibition of O-GlcNAcase is cytoprotective. Thus acute augmentation of O-GlcNAcylation may provide a novel strategy for inhibiting acute vascular injury responses. Further experiments will investigate which specific proteins and pathways modified by O-GlcNAc are associated with the protection seen in our study.

GRANTS

This work was supported, in part, by National Heart, Lung, and Blood Institute grants HL-07457, HL-64614, and HL-75211 (S. Oparil), HL-68806 and HL-076175 (L. G. Nöt and J. C. Chatham), HL-067464 and HL-079364 (J. C. Chatham), and HL-080017 and HL-044195 (Y. F. Chen); by American Heart Association Greater Southeast Affiliate Grants 0455197B (Y. F. Chen) and 0425455B and 0765398B (D. Xing); and by a Fannie E. Ripple Foundation award (A. P. Miller).