Phosphorylation of p38 Mitogen-Activated Protein Kinase Downstream of Bax-Caspase-3 Pathway Leads to Cell Death Induced by High d-Glucose in Human Endothelial Cells

Human aortic endothelial cells (passage 3) and bovine aortic endothelial cells (passage 1) were obtained from Clonetics (San Diego, CA) and cultured in modified MCDB131 medium supplemented with 5% fetal calf serum, 50 μg/ml gentamicin sulfate, 50 ng/ml amphotericin-B, 10 ng/ml epidermal growth factor, and 1 mmol/l hydrocortisone in the standard fashion (27). Cells were incubated at 37°C in a humidified atmosphere of 95% O₂:5% CO₂ with medium changes every 2 days. These cells showed the specific characteristics of endothelial cells on the basis of immunohistochemical examination and morphologic observation. Briefly, human aortic endothelial cells tested positive for von Willebrand factor and for uptake of diacetylated LDL. All cells were used within passages 5–8.

An index of cell proliferation was determined using a WST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate) cell-counting kit, which is similar to the MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) assay (Wako, Osaka, Japan) (5,6,28). Endothelial cells were seeded onto uncoated 96-well tissue culture plates (Corning, NY). In the preparation of experiments for determination of cell count, the cells were grown to subconfluence. After subconfluence, the medium was changed to defined serum-free (DSF) medium. DSF medium was supplemented with insulin (5 × 10⁻⁷ mol/l), transferrin (5 mg/ml), and ascorbate (0.2 mmol/l), as previously described (29). The cells were then incubated overnight. On day 1, the medium was changed to fresh DSF medium with d-glucose (5–25 mmol/l). On day 3, an index of cell proliferation was determined using sulfonated tetrazolium salt, WST-1, because this compound produces a highly water-soluble formazan dye that makes the assay procedure easier to perform (28).

Cells were stained with Hoechst 33258 (10 μmol/l) and propidium iodide (PI) (10 μmol/l) for 10 min and analyzed under a nonconfocal fluorescence microscope (Olympus BX60) with excitation at 360 nmol/l, as previously reported (30). As Hoechst 33258 stains all nuclei and PI stains nuclei of cells with a disrupted plasma membrane, nuclei of viable, necrotic, and apoptotic cells were observed as blue intact nuclei, red round nuclei, and fragmented (or condensed) nuclei, respectively, under fluorescence microscopy. Individual nuclei were visualized at ×400 to distinguish the normal uniform nuclear pattern from the characteristic condensed coalesced chromatin pattern of apoptotic cells (30). Although chromatin undergoes condensation during mitosis, these cells can be readily distinguished from apoptotic cells by their uniform and equatorial pattern of chromatin condensation compared with the randomly coalesced pattern typical of apoptotic cells.

As an alternative assay of apoptosis, we used the measurement of cellular DNA fragmentation using a cellular DNA fragmentation enzyme-linked immunosorbent assay (ELISA) kit (Boehringer Mannheim, Mannheim, Germany) (31,32). We confirmed that an increase in apoptotic cells is associated by increased absorbance. Briefly, 10,000 apoptotic cells/well are reported to reflect an absorbance of 1.5 under the manufacturer’s recommended conditions. Regarding the sensitivity of DNA fragmentation ELISA, its results correlated well with the results of the conventional ³H-thymidine–based DNA fragmentation assay. In our experimental conditions, an increase in absorbance of 0.2 reflected an increase in cell number to 2,000 apoptotic cells/well.

Extent of cell death was assessed using a kit to measure released lactate dehydrogenase (LDH) activity from dead cells, because loss of cell membrane integrity was observed in both necrotic and apoptotic cells (33). LDH activity in the culture medium was determined using a commercially available kit (Wako) exactly as described by the manufacturer. In the preparation of experiments for the LDH assay, the cells were grown to subconfluence. After subconfluence, the medium was changed to DSF medium. The cells were then incubated overnight. On day 1, the medium was changed to fresh DSF medium containing d-glucose (5–25 mmol/l) or mannitol (25 mmol/l).

Western blotting was performed for analysis of bax, bcl-2, and MAPK kinase (MEK) kinase 1 (MEKK1) proteins. Endothelial cells were seeded onto 15-cm dishes (Corning). Endothelial cells were grown to confluence and made quiescent by incubation in DSF medium 24 h before treatment. After treatment, the cells were extracted with radioimmunoprecipitation assay (RIPA) buffer (50 mmol/l Tris-Cl, 0.15 mol/l NaCl, 0.1% SDS, 1% deoxycholate, 1% Triton-X, 2 mmol/l Na₃VO₄, and 10 mmol/l NaF). Samples containing 20 or 100 μg protein were run on 10 or 12.5% SDS polyacrylamide gels. Proteins were separated by SDS/PAGE, transferred to nitrocellulose membrane (Hybond ECL; Amersham), and incubated with an antibody to bax (anti-human rabbit IgG, 1:100; Santa Cruz), bcl-2 (anti-human mouse IgG, 1:25; Santa Cruz), and MEKK1 (anti-human rabbit IgG, 1:100; Santa Cruz) at 4°C overnight. Antibodies were diluted in 4% skimmed milk and 0.1% Tween 20 in phosphate-buffered saline (PBS). The membranes were then washed and incubated with a 1:2,000 dilution of mouse or rabbit Ig horseradish peroxidase–conjugated antibody (Amersham). Bound antibodies were detected by enhanced chemiluminescence (Amersham) and Hyperfilm-MP, Amersham). To quantify and compare levels of proteins, the density of each band was measured by densitometry (Shimazu, Kyoto, Japan). It was confirmed by staining with Coomassie brilliant blue R (CBB; Sigma) that amounts of loaded proteins were equal. Staining with CBB revealed identical amounts of protein in all samples for Western blotting (data not shown). Western blotting of α-tubulin using anti–α-tubulin antibody (anti-human mouse IgG, 1:100; Oncogene) was also performed to confirm that amounts of loaded proteins were equal. Western blotting was also performed for analysis of the phosphorylation of p38 MAPK. After treatment, the cells were extracted with lysis buffer (50 mmol/l Tris-Cl, 10 mmol/l EGTA, 10 mmol/l NaF, 150 nmol/l NaCl, 0.1% SDS, 1% deoxycorticosterone, and 1% Triton X-100). Proteins were separated by SDS/PAGE, transferred to nitrocellulose membranes (Hybond ECL, Amersham), and incubated with a polyclonal antibody to phosphospecific p38 MAPK (anti-human rabbit IgG, 1:1,000; New England BioLabs) or p38 MAPK (anti-human rabbit IgG, 1:1,000; New England BioLabs) at 4°C overnight.

Endothelial cells were lysed in PBS, resuspended in isotonic buffer A (200 mmol/l mannitol, 70 mmol/l sucrose, 1 mmol/l EGTA, and 1 mmol/l HEPES, pH 7.5) supplemented with protease inhibitors (1 mmol/l phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 10 μg/ml pepstatin A, 10 μg/ml soybean trypsin inhibitor, and 10 μg/ml aprotonin), and homogenized using a polytron homogenizer (Brinkmann Instruments) at setting 6.5 for 10 min. Nuclei and unbroken cells were separated at 120g for 5 min as the low-speed pellet (P1). This supernatant was centrifuged at 7,500g for 10 min to collect the heavy membrane (HM) pellet. This supernatant was centrifuged at 100,000g for 30 min to yield the light membrane (LM) pellet and final soluble fraction. Hypotonic lysis was performed by resuspending the cells in hypotonic buffer (10 mmol/l Tris, pH 7.5, 5 mmol/l MgCl₂, 1 mmol/l EGTA, 1 mmol/l dithiothreitol) and homogenized with a Dounce homogenizer. Samples containing 20 or 100 μg protein were run on 12.5% SDS polyacrylamide gels and analyzed as described above.

Cells were harvested after exposure to high d-glucose for the indicated periods of time and washed three times with PBS and then suspended in buffer containing 50 nmol/l Tris/HCl (pH 7.4), 1 mmol/l EDTA, and 10 mmol/l EGTA. After addition of 10 μmol/l digitonin, cells were incubated at 37°C for 10 min. They were centrifuged at 900g for 3 min, and the resulting supernatant (40 μg protein) was incubated with 50 μmol/l enzymesubstrate acetyl-l-aspartic-l-glutamic-l-valyl-l-asparatic acid-4-methyl-cou-mary-7-amide (Ac-DEVD-MCA) at 37°C for 1 h. Levels of released 7-amino-4-methylcoumarin were measured using spectrofluorometers (Hitachi F-3000 and F-2000) with excitation at 380 nm and emission at 460 nm. Excitation and emission slit width were adjusted to 10 and 20 mm, respectively. One unit was defined as the amount of enzyme required to release 0.22 nmol 7-amino-4-methylcoumarin per minute at 37°C.

Human aortic endothelial cells were transiently transfected with the plasmid pcDNA1 or a vector encoding a catalytically inactive mutant in which Lys432 was changed to Ala [MEKK1 (K432A)] (18). In addition, kinase-inactive p38 MAPK using the pcDNA3 vector, as described previously (34), was also transiently transfected. Transfection was performed using LipofectAMINE2000 and PLUS regent (Life Technologies).

SB 203580, a specific inhibitor of p38 MAPK, was obtained from Calbiochem. The tetrapeptide substrate for caspase-3, Ac-DEVD-MCA, was purchased from Peptide Institute (Osaka, Japan). The caspase inhibitors, benzyloxy-carbonyl-Val-Ala-Asp-fluoromethyl ketone z-VAD.FMK and z-DEVD.FMK, were obtained from Enzyme Systems Products (Dubin, CA). Hoechst 33258 (bisbenzimide) staining dye was obtained from Wako. PI was obtained from Molecular Probes (Eugene, OR).

All values are expressed as means ± SE. Analysis of variance with subsequent Bonferroni’s/Dunnet test was used to determine the significance of differences in multiple comparisons. Values of P < 0.05 were considered statistically significant.

First, we examined the effect of high d-glucose treatment on apoptosis in human endothelial cells. Consistent with previous reports (7,8,9), treatment of endothelial cells with a high concentration of d-glucose resulted in a significant decrease in cell number in a dose-dependent manner (d-glucose 5 mmol/l 0.483 ± 0.022 vs. d-glucose 25 mmol/l 0.416 ± 0.020, P < 0.01; mannitol 25 mmol/l 0.485 ± 0.024, l-glucose 25 mmol/l 0.480 ± 0.016 absorbance at optical density 450 nm). No significant changes were observed in cell number of endothelial cells treated with a high concentration of mannitol or l-glucose. Therefore, endothelial cell death was attributable to the cytotoxicity of high d-glucose concentration, but not to high osmolarity. After high d-glucose treatment, some cells started to become round and eventually detached from the plate and floated in the medium, leaving many holes in the sheet of confluent cells (data not shown). The floating cells could be recovered with the medium and neither attached to a new plate nor proliferated. In addition, cells treated with high d-glucose exhibited the characteristic features of cell shrinkage, membrane blebbing, and rounding, typical of apoptotic death, under phase-contrast microscopy. Thus, apoptosis was assessed by the morphologic features using double staining with Hoechst 33258 and PI under fluorescence microscopy (30). Simultaneous assessment of nuclear chromatic morphology by Hoechst 33258 staining (blue) and PI staining (red) verified that these cells eventually manifested typical apoptotic condensed and coalesced nuclei (Fig. 1A). Apoptotic cells were clearly observed in cells treated with high d-glucose, as shown in Fig. 1A. Cells with fragmented or condensed nuclei were classified further based on PI staining as “early” apoptotic cells with membrane integrity and “terminal” apoptotic cells without membrane integrity. These nuclear morphologic changes under fluorescence microscopy correspond well to necrosis and apoptosis defined by electron microscopy in endothelial cells induced by high glucose. As shown in Fig. 1B, treatment with high d-glucose significantly increased the number of apoptotic and necrotic cells (P < 0.05), while a high concentration of mannitol did not. A significant increase in apoptotic cells was also confirmed by DNA fragmentation ELISA (Fig. 1C) (P < 0.01), whereas mannitol as an osmotic negative control did not induce apoptosis. In LDH release, treatment of endothelial cells with high concentration of d-glucose also resulted in a significant increase in a time-dependent manner, as compared with a high concentration of mannitol (Fig. 1D) (P < 0.01).

Therefore, we further examined the molecular mechanisms of apoptosis induced by high d-glucose treatment. In particular, we focused on the expression of bcl-2 and bax proteins. Bcl-2 and bax are homologous proteins that have opposing effects on cell life and death, with bcl-2 serving to prolong cell survival and bax acting as an accelerator of apoptosis (35). The bcl-2 and bax proteins can form heterodimers in cells (36). Interestingly, d-glucose treatment significantly increased bax protein as assessed by Western blotting, whereas no significant change in bax protein was observed with mannitol treatment (Figs. 2A and B) (P < 0.05). In contrast, bcl-2 protein was not affected by high d-glucose treatment at 25 mmol/l (Figs. 2C and D). Thus, the ratio of bcl-2 to bax was significantly decreased in cells treated with high d-glucose (5 mmol/l d-glucose 100% vs. 25 mmol/l d-glucose 67.9 ± 8.6%, *P < 0.01; mannitol at 25 mmol/l 97.8 ± 10.2%).

Recent studies demonstrated that translocation of bax protein from the cytoplasm to the mitochondorial membrane is critical in apoptosis (37,38). Therefore, we also investigated the effect of high d-glucose treatment on the translocation of these apoptosis-related proteins. A substantial portion of bax protein was found consistently in the soluble S100 fraction representing the cytosol as well as the mitochondria-enriched HM fraction, as documented by mitochondrial markers (porin/VDAC, a marker of outer mitochondorial membrane) (Fig. 3A). Of particular importance, after high d-glucose treatment, most of bax protein moved from the cytosol to the HM fraction (Fig. 3A). In contrast, bcl-2 protein resided in the mitochondria-rich HM and LM fraction as well as the soluble fraction under normal glucose condition (Fig. 3B). Moreover, the distribution of bcl-2 protein was not altered by high d-glucose treatment (Fig. 3B).

It was recently reported that caspases are important regulators of apoptosis, and the release of cytochrome c from mitochondria activated caspase, especially caspase-3 (13). Interestingly, recent reports revealed that bax induces the release of cytochrome c and that bcl-2 prevents it (15,16). To further analyze the contribution of activation of caspase proteases in the death signal triggered by high d-glucose treatment, the proteolytic activity of caspases was measured using fluorogenic tetrapeptide substrates. Activity of caspase-3 proteases was significantly increased after high d-glucose treatment, and reached a peak at 24 h after exposure to high d-glucose in endothelial cells, as compared with mannitol (Fig. 4A). Thus, we next examined the effects of a specific caspase-3 inhibitor, z-DEVD, and a wide-spectrum caspase inhibitor, z-VAD, on the cell death induced by high d-glucose. Importantly, the addition of z-VAD as well as z-DEVD significantly attenuated the cell death induced by high d-glucose (P < 0.01) (Fig. 4B).

We also studied how signal transduction pathways, such as those controlled by kinases, modulate critical cellular functions such as cell growth, differentiation, and apoptosis. Three major kinase cascades that culminate in the activation of three different sets of MAPKs, extracellular signal–related kinase (ERK), JNK/SAPK (stress-activated protein kinase), and p38 MAPK, have recently been identified (39,40). Treatment with high d-glucose, as compared with normal glucose, significantly increased phosphorylated p38 MAPK assessed by Western blotting using phosphorylated specific antibody (Fig. 5A). p38 MAPK was markedly phosphorylated, with sustained stimulation (Fig. 5A), whereas mannitol did not change the level of phosphorylated p38 MAPK (data not shown). In addition, the specificity of phosphorylation of p38 MAPK was confirmed by the observation that no apparent change in total p38 MAPK by high d-glucose was observed (Fig. 5A). An osmotic control, mannitol, also did not affect total p38 MAPK in endothelial cells at normal and high concentrations. Importantly, administration of a specific inhibitor of p38 MAPK, SB 203580, completely attenuated endothelial cell death induced by high d-glucose (Fig. 5B) (P < 0.01). More importantly, administration of a specific inhibitor of p38 MAPK, SB 203580, completely attenuated endothelial cell death induced by high d-glucose in a dose-dependent manner, as assessed by DNA fragmentation and LDH release (Figs. 5C and D) (P < 0.01). In addition, transient overexpression of kinase-inactive p38 MAPK into endothelial cells also attenuated endothelial cell death in LDH releases (Fig. 5E) (P < 0.01). To confirm this result in another endothelial cell, we also checked the same experiment in bovine aortic endothelial cells. A specific inhibitor of p38 MAPK, SB 203580, and the overexpression of kinase-inactive p38 MAPK also attenuated cell death induced by high d-glucose in bovine aortic endothelial cells (control 1.00 ± 0.23, 25 mmol/l d-glucose 1.855 ± 0.163, 25 mmol/l d-glucose + SB203580 0.994 ± 0.034, and 25 mmol/l d-glucose + dominant-negative p38 MAPK 1.109 ± 0.022; P < 0.01) .

To further analyze the cross-talk between p38 MAPK and caspase proteases, we examined the effect of a specific caspase-3 inhibitor, z-DEVD, and a wide-spectrum caspase inhibitor, z-VAD, on the phosphorylation of p38 MAPK induced by high d-glucose treatment. p38 MAPK was phosphorylated after high d-glucose stimulation, whereas addition of z-VAD as well as z-DEVD significantly inhibited the phosphorylation of p38 MAPK at 12 h after high d-glucose treatment (Fig. 6A). However, at 30 min after high d-glucose treatment, these caspase inhibitors did not attenuate the phosphorylation of p38 MAPK (Fig. 6B). These results demonstrated that phosphorylation of p38 MAPK at late phase (6 h), but not acute phase (30 min), is due to caspase activation. More than likely, phosphorylation of p38 MAPK at acute phase may be due to other stimuli such as stress. For further analysis, we examined the effect of p38 MAPK phosphorylation on caspase-3 activation at 6 h after high d-glucose treatment. SB 203580 (30 μmol/l) and the overexpression of kinase-inactive p38 MAPK did not attenuate caspase-3 activation at 6 h after high d-glucose treatment (Fig. 6C). Together, SB 203580 and the overexpression of kinase-inactive p38 MAPK attenuated cell death induced by high d-glucose (Figs. 5C–E); p38 MAPK phosphorylation is located downstream of caspase.

Thus, we finally focused on MEKK1 to further examine the cross-talk between p38 MAPK and the bax-caspase proteases pathway, as MEKK1 is a 196-kDa enzyme that is involved in the regulation of apoptosis (41,42). MEKK1 is cleaved at Asp⁸⁷⁴ by caspases, and the cleaved kinase domain of MEKK1 itself stimulates caspase activity leading to apoptosis. Treatment of human aortic endothelial cells with high d-glucose induced cleavage of the 196-kDa MEKK1 protein at concentrations that maximize the apoptotic response indicated by the loss of full-length MEKK1 (Fig. 7A). Recent studies demonstrated that in HEK293 cells, the cleavage fragments of endogenous MEKK1 were not consistently detected with other cleaved proteins (42,43). Therefore, the total 196-kDa MEKK1 protein level is considered as the response to apoptotic stimuli. Interestingly, addition of z-VAD as well as z-DEVD significantly attenuated the cleavage of MEKK1 induced by high d-glucose. Thus, to test the hypothesis that the cleavage of MEKK1 contributes to the activation of p38 MAPK, we transfected a catalytically inactive mutant in which Lys⁴³² changed to Ala [MEKK1 (K432A)] into endothelial cells before high d-glucose treatment. Importantly, kinase-inactive MEKK1 also blocked the phosphorylation of p38 MAPK induced by high d-glucose (Fig. 7B). These data suggest that phosphorylation of p38 MAPK was located downstream of the caspase pathway mediated by MEKK1.

Although classical paradigms defining the pathophysiology of vascular disease have focused on the abnormal regulation of cell growth in response to growth factors, it has been recently postulated that the regulation of cell death by apoptosis may be another determinant of vascular structure and lesion formation (44). Therefore, vascular structure has been postulated to be determined in large part by a balance between cell growth and cell death by apoptosis. This process of vascular remodeling plays an important role in determining the natural history of vascular diseases (45). In response to a variety of stimuli and circumstances, cells have an intrinsic capacity to activate a gene-directed program that commits the cell to a suicidal death, described as apoptosis. It has become increasingly clear that the process of cell death by apoptosis is a relatively ubiquitous phenomenon observed in a variety of cell types, including endothelial cells (46). Disruption or dysfunction of endothelial cells, causing loss of multiple endothelium-derived substances (PGI₂, NO, CNP), has been hypothesized to play a pivotal role in the progression and/or development of vascular disease in diabetes. In this study, we demonstrated that high d-glucose treatment induced endothelial cell death through the induction of apoptosis, whereas mannitol and l-glucose, as controls for osmolarity, did not, which is consistent with previous findings (7,8,9). However, little is known about the molecular mechanisms on d-glucose–induced endothelial cell death. Apoptosis has been suggested to be largely controlled by 1) modulation of the activity of antiapoptotic and proapoptotic molecules (e.g., bcl-2 and bax) and 2) triggering of specific signaling pathways (e.g., MAPK cascade). Of importance, the present studies revealed a significant increase in bax, a proapoptotic factor, by high d-glucose treatment. Here, we demonstrated that apoptosis induced by high d-glucose may be attributable to an inappropriate increase in the ratio of bax to bcl-2 induced by high d-glucose. There were two supportive reports that in vivo bax accelerated death of retinal cells in hyperglycemia (47) and that hyperglycemic conditions increased the expression of bax as early as the preimplantation blastocyte stage in the mouse (48). In the latter report, the blastocysts from bax-deficient mice were protected from glucose-induced apoptosis. This result suggests that bax may be a key modulator of hyperglycemia-induced apoptosis and high-rate congenital malformations and spontaneous miscarriages induced by hyperglycemia at early stages after conception.

Recent studies have also documented that the subcellular localization of these molecules determines the fate of cells, in addition to classical diagrams defining the quantity. Bcl-2 and bax contain a hydrophobic segment at their COOH-terminal end (37,49) that is believed to serve as a membrane anchor. Subcellularly, bcl-2 protein is localized to the nuclear envelope, endoplasmic reticulum, and outer mitochondrial membrane (50). Recently, it was demonstrated that after delivery of death signals to cells in culture, bax moves to the mitochondria and other membrane sites and triggers a catastrophic transformation of mitochondrial function. However, no reports have revealed whether the translocation of bax protein could occur under physiological conditions such as high d-glucose. We demonstrated that after high d-glucose treatment, bax protein inserts into the mitochondrial membranes, while it is located in the cytosol and in association with intracellular membranes including mitochondria under normal glucose condition. Translocation of bax protein into the mitochondrial membrane was accompanied by a significant increase in caspase-3 activity. These results demonstrate that bax translocation from the cytosol to the mitochondrial membrane occurred under physiological conditions, thereby leading to pathological tissue destruction. As it is believed that cytosolic bax fails to homodimerize in the presence of protective levels of bcl-2, exploration of various stimuli of bax translocation might be important to understand the cell death system. Interestingly, we demonstrated that cell death mediated by bax targeting mitochondria could be inhibited by zVAD as well as Z-DEVD. Thus, caspase-3 may play a pivotal role in the process of endothelial cell death induced by high d-glucose. In contrast, the present study failed to demonstrate a change in the distribution of bcl-2 after high d-glucose treatment, whereas bcl-2 has been reported to prevent mitochondrial dysfunction (i.e., cytochrome c release and caspase protease activation).

Finally, we focused on signal transduction pathways, because these pathways regulate cell growth and death. The three MAPK (ERK, JNK/SAPK, and p38 MAPK) pathways have been implicated in the control of apoptosis. ERK is activated by mitogens and survival factors, whereas JNK/SAPK and p38 are stimulated by stress signals (39,40). Although previous findings indicated that the activation of p38 MAPK may play a pivotal role in the control of cell death (51,52), the cross-talk between p38 MAPK and caspase is not still clear. The present study demonstrated that phosphorylation of p38 MAPK induced by high d-glucose-mediated cell death and apoptosis. Interestingly, the sustained phosphorylation of p38 MAPK was observed after high d-glucose stimulation, whereas caspase inhibitors completely attenuated the sustained phosphorylation of p38 MAPK at 12 h, but not 30 min, after high d-glucose treatment. From these results, sustained phosphorylation of p38 MAPK is located downstream of activation of caspase proteases. Similar results about the cross-talk between p38 MAPK and the caspase protease pathway were recently reported by Berra et al. (52). More importantly, the inhibitor of p38 MAPK and the overexpression of kinase-inactive p38 MAPK attenuated endothelial cell death despite evidence of caspase activation.

These MAPK family members are in turn activated by upstream MAP/Efr kinases (MEKs) such as SAPK/ERK kinase-1 (SEK1, also known as MKK4) and MKK3/MKK6, which is upstream of JNK/SAPK and p38 MAPK. These MEKs are in turn activated by MEKKs, of which MEKK1, -2, -3, and -5 have been characterized to date (53,54). Although the identity of the MEKK upstream of the p38 MAPK pathway is unknown, previous studies provide evidence for the role of MEKK1 as an upstream activator of both p46 and p54 SAPK as well as p38 MAPK (55,56). On the other hand, MEKK1 is cleaved at Asp⁸⁷⁴ by caspases, and the cleaved kinase domain of MEKK1 itself stimulates caspase activity leading to apoptosis (41,42). Therefore, we focused on MEKK1 to further examine the cross-talk between p38 MAPK and the bax-caspase proteases pathway. Treatment of human aortic endothelial cells with high d-glucose induced MEKK1 cleavage, whereas the addition of z-VAD as well as z-DEVD significantly attenuated the cleavage. Importantly, kinase-inactive MEKK1 also blocked the phosphorylation of p38 MAPK induced by high d-glucose. Taken together, the present data clearly demonstrated that phosphorylation of p38 MAPK mediated by MEKK1 downstream of bax-caspase proteases determines the cell death and apoptosis induced by high d-glucose (Fig. 8). Cross-talk between the signal transduction system and the caspase pathway may determine the cellular fate, although the present study using in vitro condition may or may not mimic what happens in the in vivo diabetic state.

Overall, we demonstrated that high d-glucose, but not mannitol or l-glucose, induced aortic endothelial cell death through the induction of apoptosis. In addition, this study demonstrated that endothelial cell death and apoptosis induced by high d-glucose may be mediated by 1) caspase protease activation mediated by an appropriate increase and translocation of bax (modulating the activity of proapoptotic molecules) and 2) the phosphorylation of p38 MAPK mediated by MEKK1 downstream of caspase protease activity. These results demonstrate the molecular mechanism of high d-glucose–induced cell death, suggesting the possible mechanisms of glucose toxicity.

This work was partially supported by grants from the Japan Cardiovascular Research Foundation, a Japan Heart Foundation Research Grant, a Grant-in-Aid from Japan Promotion of Science, and a Grant-in-Aid from the Ministry of Education, Science, Sports and Culture of Japan.