α-Tocopherol Decreases Superoxide Anion Release in Human Monocytes Under Hyperglycemic Conditions Via Inhibition of Protein Kinase C-α

The human monocytic cell line THP-1 cells were obtained from American Type Culture Collection. Endotoxin-free, glucose-free RPMI-1640 media and fetal bovine serum (FBS) were purchased from Gibco BRL (Carlsbad, CA). Antibiotics, glutamine, phenylmethylsulfonyl fluoride, glucose, HEPES, protease inhibitor cocktail, Triton X-100, dithiothreitol, rotenone, aminooxyacetic acid (AOAC), α-cyano-4-hydroxycinnamic acid (4-HOCA), theonyl-trifluoroacetone (TTFA), and carbonyl cyanide m-chlorophenylhydrazone (CCCP) were from Sigma Chemical. Antibodies to PKC, PKC-α, PKC-βII, and p47phox were obtained from Santa Cruz Company. 2,2′,3,3′,4,4′-hexahydroxy-1,1′-biphenyl-6,6′-dimethanol dimethyl ether (HBDDE), apocynin, and diphenyleneiodonium chloride (DPI) were obtained from Calbiochem, and Eli Lilly kindly provided LY379196. Polyvinylidene difluoride (PVDF) membranes and Tris-glycine gels were from Invitrogen. Bicinchoninic acid (BCA) kit was obtained from Pierce. Enhanced chemiluminescence (ECL) and PKC activity kits were purchased from Amersham Pharmacia. Oligonucleotides were purchased from Integrated DNA Technologies.

THP-1 cells were maintained in endotoxin-free RPMI-1640 containing 5.5 mmol/l glucose, 50 μmol/l mercaptoethanol, 10% FBS, 2 mmol/l glutamine, 1 mmol/l sodium pyruvate, and 10 mmol/l HEPES and used for experiments between third and fifth passages. Cells were cultured (1 × 10⁶ cells/ml) for 3 days in either 5.5 mmol/l (normal glucose [NG]) or 15 mmol/l glucose (HG conditions), and as an osmotic control, 9.5 mmol/l mannitol was added along with NG. Cell viability, as determined by trypan blue exclusion, was >92%. AT (100 μmol/l) was added to cells with NG/HG, with daily changes in media. For inhibitor studies, 50 μmol/l HBDDE or 30–150 nmol/l LY379196 were added to cells at the end of the third day in HG conditions and incubated overnight, as per our preliminary data. Cells were also incubated with mitochondrial respiratory chain inhibitors, 100 μmol/l AOAC, 5 μmol/l rotenone, 10 μmol/l TTFA, and 0.5 μmol/l CCCP for 24 h along with 15 mmol/l glucose (5), and ROS was determined.

O₂⁻ production was measured by the superoxide dismutase (SOD)-inhibitable reduction of cytochrome C, as reported previously (2). Cells were incubated in RPMI-1640 without phenol red for 60 min at 37°C with or without SOD (100 μg/ml) and 80 μmol/l acetylated ferricytochrome C in a total volume of 1 ml. Results were expressed as nanomoles per minute per milligram cell protein.

ROS generation was detected using fluorescent probe, 5-(and-6)-carboxy-2′7′-dichlorodihydro-fluorescein diacetate (DCF-DA). Cells (1 × 10⁵ ml⁻¹) were loaded with 10 μmol/l DCF-DA, incubated for 1 h at 37°C, and analyzed at 0 and 60 min in CytoFlour multiwell reader. ROS production was determined by subtracting the values from initial intensities (0 min) and expressed per milligram cell protein.

PKC activity in THP-1 cells was determined by radioimmunoassay. It was based on the PKC-catalyzed transfer of the γ-phosphate group of ATP to a PKC-specific peptide. PKC activity was expressed as nanomoles of phosphate transferred per million cells.

At the end of culture, cells were lysed, and membrane fractions were isolated as described by Ceolotto et al. (18). Membrane proteins (10–30 μg) were resolved in 10% Tris-glycine gel, and blotting was performed with specific primary and secondary antibodies. Blots were visualized by ECL detection system.

Cells were incubated with oligodeoxynucleotides (ODNs) to PKC-α and PKC-βII along with HG conditions and added daily. The sequence for PKC-α isoenzyme-specific antisense oligonucleotide was 5′-CGC CGT GGA GTC GTT GCC CG-3′; the sense sequence was 5′-CGG GCA ACG ACT CCA CGG CG-3′ (7). The PKC-β antisense oligonucleotide was 5′-CGC AGC CGG GTC AGC ATC-3′; the sense sequence was 5′-GAT GGC TGA CCC GGC TGC G-3′ (19). The p47phox antisense oligonucleotide was 5′-TTT GTC TGG TTG TCT GTG GG-3′; the sense sequence was 5′-CCC ACA GAC AAC CAG ACA AA-3′ (20). All of the oligonucleotides were phosphorothioate modified and high-performance liquid chromatography purified. ODNs were added to the cells at the concentration of 2 μmol/l (7,19,20).

All experiments were performed at least three times in duplicate or triplicate. Experimental results are presented as the means ± SD. Paired t tests were used for data analysis, and significance was defined as P < 0.05.

Cells cultured in HG media showed a significant 40% increase of O₂⁻ production compared with NG (P < 0.01), which was inhibited by the addition of AT (P < 0.01) (Fig. 1A). Release of ROS, as assessed by DCF-DA staining, paralleled the O₂⁻ data (Fig. 1B). Mannitol (9.5 mmol/l) had no significant effect on ROS/O₂⁻ production (Fig. 1A and B).

Cells were cultured in HG conditions along with rotenone (an inhibitor of complex I), TTFA (an inhibitor of complex II), CCCP (an uncoupler of oxidative phosphorylation that abolishes the mitochondrial membrane proton gradient), AOAC (inhibitor of malate-aspartate shuttle), or 4-HOCA (inhibitor of glycolysis-derived pyruvate transport into mitochondria). There were no significant differences in ROS production when any of these inhibitors were used (Fig. 2). However, in human aortic endothelial cells (HAECs), 4-HOCA, TTFA, and CCCP significantly inhibited ROS release (P < 0.01) (Fig. 2).

Previous studies have shown that O₂⁻ release from human monocytes was mediated via activation of PKC (7); therefore, we tested PKC activity in THP-1 cells under HG conditions. PKC activity was significantly increased in HG conditions compared with NG, and AT enrichment of THP-1 cells significantly decreased PKC activity in HG conditions (Fig. 3A). Immunoblots of total PKC in membranes showed a significant increase in PKC translocation to the membranes in HG conditions compared with NG, whereas addition of AT inhibited the PKC translocation to membranes (Fig. 3B). NADPH oxidase is a major source of superoxide in phagocytes. The effect of hyperglycemia on p47phox translocation to the membrane was investigated. p47phox translocation was significantly increased in HG conditions when compared with NG, whereas in AT-treated cells, it was significantly reduced (Fig. 3B).

To elucidate the signaling pathway of O₂⁻ production, cells were incubated with inhibitors of PKC or NADPH oxidase. HG conditions significantly increased O₂⁻ release, whereas AT reduced it, as in previous experiments (P < 0.01). The addition of HBDDE significantly inhibited O₂⁻ release (P < 0.01), but the specific PKC-βII inhibitor (LY379196, 30 nmol/l) had no significant effect (Fig. 4). Also, 150 nmol/l LY379196 had no significant effect on O₂⁻ release. When NADPH oxidase inhibitors apocynin (30 μmol/l) and DPI (10 μmol/l) were used, O₂⁻ release was inhibited significantly (P < 0.001) (Fig. 4). These data suggest that monocytic O₂⁻ release is driven possibly by PKC-α, and NADPH oxidase is necessary for the monocytic O₂⁻ release. Western blots for PKC-α and -βII in membrane fractions are shown in Fig. 5, top and middle panels. HG conditions increased the expression of both PKC-α and -βII isoforms, whereas the addition of AT inhibited both. To determine whether PKC-α was involved in stimulating O₂⁻ release through p47 phox, membranes were blotted with anti-p47phox antibody. AT and the PKC-α inhibitor, but not the PKC-βII inhibitor, inhibited p47phox translocation to membranes (Fig. 5, bottom panel).

To further confirm that O₂⁻ release was mediated through PKC-α, cells were incubated with ODNs to PKC-α and -βII. Results showed that addition of antisense oligo to PKC-α significantly inhibited O₂⁻ release (P < 0.01), whereas antisense oligo to PKC-βII did not have any effect (Fig. 6). However both antisense oligos decrease PKC activity by 41% (P < 0.01) and inhibit the translocation (>90%) of respective isoform to the membrane. Similarly, antisense oligos to p47phox also inhibited O₂⁻ release (P < 0.001). To confirm that p47phox activation is mediated through PKC-α, Western blots were run for the membrane fractions and quantitated. p47phox translocation was significantly inhibited with the addition of antisense oligo to PKC-α but not with antisense oligo to PKC-β II (Fig. 7). In all ODN experiments, as a control, sense oligos to PKC-α and -β and p47phox were added and did not affect O₂⁻ release from monocytes under HG conditions.

In this study, we used THP-1 cells because they have many characteristics of human monocytes (21). We previously showed that monocytes from type 2 diabetic patients released increased O₂⁻ compared with those from matched control subjects, whereas AT supplementation decreased O₂⁻ production (2). The PKC isoenzyme family is required for human monocytic O₂⁻ release (22). PKC-α and -βII are primarily modulated by glucose in human monocytes (18,19). In this study, we show under HG conditions that O₂⁻ release and PKC activity were significantly increased. Li et al. (7) had shown that under NG, O₂⁻ release from monocytes is mediated via PKC-α. Recently, it has been shown that mitochondria are the major source for O₂⁻ in endothelial cells (ECs) under HG conditions (5). We show, for the first time, that various mitochondrial complex inhibitors did not inhibit ROS release from monocytes, whereas they decreased ROS release from HAECs (5). These findings suggest that there are different mechanisms mediating O₂⁻ production in different cells. In this regard, it should be emphasized that the monocyte is a classical phagocyte. The increased ROS produced under HG conditions could modify biomolecules, thus promoting vasculopathies. Because it is known that PKC activity is induced under hyperglycemia (6), we studied modulation of PKC isoforms in THP-1 cells under HG conditions. We show that high glucose induced both PKC-α and -βII, whereas AT reduced their expression. Our results are in partial agreement with the findings of Ganz and Seftel (23), in corpus cavernosum vascular smooth muscle cells, that PKC-βII but not PKC-α is increased with HG conditions, and AT inhibited it. An increase in PKC-α levels was not detected, probably because of the different cell type used. Ceolotto et al. (18) had shown that although there was an increase in both PKC-α and -βII in diabetic human monocytes, only the increase in PKC-βII was significant. It is not clear which of these isoforms mediates O₂⁻ release from human monocytes under hyperglycemia. Several laboratories have shown that PKC-dependent signaling is involved in the activation of NADPH oxidase and O₂⁻ production in neutrophils. Hence, we studied p47phox translocation to membranes. Our results also confirmed that there was increased p47phox translocation to membranes with HG conditions. This is well correlated with other reports that NADPH oxidase is involved in monocytic O₂⁻ release (11,24). The addition of AT reduced p47phox membrane translocation. This is supported by the studies of Cachia et al. (24) showing that under NG conditions, AT decreased PMA-induced O₂⁻ production in monocytes. However, although they studied the effect of AT under NG and reported decreased PKC activity, the effect of AT on the translocation of neither PKC-α nor -βII were studied. To study the involvement of PKC isoforms in glucose-induced O₂⁻ release and the mechanism of its inhibition by AT, we used PKC inhibitors as well as sense and antisense ODNs to both isoforms. HBDDE inhibited PKC-α and other isoforms nonspecifically and is not a specific inhibitor to PKC-α (25). The PKC-βII inhibitor did not have any effect on p47phox translocation. This suggests that monocytic O₂⁻ release is probably via PKC-α and not via PKC-βII, since HBDDE inhibits both PKC-α and -βII, and βII-specific inhibitor had no effect. To prove this, we used antisense ODNs. When cells were incubated with antisense to PKC-α, both O₂⁻ release and p47phox translocation to membranes were reduced, whereas the addition of antisense to PKC-β did not have any effect on monocytic O₂⁻ release, despite both ODNs decreasing PKC activity. Antisense oligos to p47phox further proved that NADPH oxidase is essential for monocytic superoxide production. This is in agreement with the study by Li et al. (7) showing that monocytic O₂⁻ release is mediated by PKC-α under euglycemia. We show for the first time that HG conditions induce PKC-α, which in turn activates p47phox translocation to membranes and induces O₂⁻ release. The antisense approach has proven quite successful in this study. Two factors likely contributing to the effectiveness of this approach are the use of monocytes as target cells and the careful selection and purity of the ODN.

The important mechanistic finding of our studies is that NADPH oxidase is activated via PKC-α by translocating p47phox to membranes under HG conditions, resulting in increased O₂⁻ release, although both PKC-α and -βII were increased by high glucose. We also show that AT inhibited these HG conditions-induced changes.

Our studies show that in monocytes, O₂⁻ release is derived predominantly through NADPH oxidase, and in EC it could be through mitochondria. The novelty of this study is that under HG conditions, PKC-α activation of NADPH oxidase triggers O₂⁻ release, and that AT decreases O₂⁻ release via inhibition of PKC-α, thus offering an explanation for the increased O₂⁻ release in diabetic monocytes. However, our findings in THP-1 cells need to be confirmed in diabetic monocytes. This amelioration of oxidative stress by AT could be beneficial in decreasing diabetic vascular complications and needs to be tested in clinical trials in diabetic patients.

This study was supported by National Institutes of Health Grant RO1 AT00005 and K24 AT00596.

We acknowledge the technical help of Veronica Cantu and Ronald Tankersley for manuscript preparation. We are thankful to Eli Lilly for providing LY379196.