Antiangiogenic Effect of KR-31372 by Apoptosis via Mediation of Mitochondrial K_ATP Channel Opening and the Phosphatase and Tensin Homolog Deleted from Chromosome 10 Phosphorylation

Angiogenesis is a process that involves enzymatic degradation and remodeling of the extracellular matrix, migration, and proliferation of capillary endothelial cells (Cao et al., 1996). Vascular endothelial growth factor (VEGF), as an endothelial cell-specific mitogen, promotes angiogenesis by stimulation of endothelial cell proliferation and inhibition of endothelial cell apoptosis (Kuzuya et al., 1999). Apoptosis, a programmed cell death, is critical during embryonal development and in tissue renewal (Cohen, 1993). Induction of endothelial cell apoptosis has recently attracted considerable interest as a potential target for antitumor therapy (Sgonc et al., 1998).

Phosphatase and tensin homolog deleted from chromosome 10 (PTEN) has dual specificity protein phosphatase (toward phospho-serine/threonine and phospho-tyrosine) and phosphoinositide 3-phosphatase activities (Myers et al., 1997; Maehama and Dixon, 1998) and antagonizes the phosphatidylinositol 3-kinase (PI3-K) pathway by catalyzing degradation of the phosphatidylinositol (3,4,5)-triphosphate [PI(3,4,5)P₃], which is generated by PI3-K, thereby leading to a control role in converting PI(3,4,5)P₃ to phosphatidylinositol (3,4)-diphosphate, an inactive state (Stambolic et al., 1998; Cantley and Neel, 1999). Reportedly, PTEN-induced-down-regulation of Bcl-2 protein requires reduction in serine/threonine kinase (Akt)/cAMP response element-binding protein signaling (Huang et al., 2001).

On the other hand, the existence of a mitochondrial ATP-sensitive potassium channels (mitoK_ATP) that are reversibly inactivated by ATP and inhibited by glibenclamide has been first demonstrated within the inner mitochondrial membrane by Inoue et al. (1991). Mitochondrial channels have higher sensitivity to opening by diazoxide, which exceeds the sensitivity of sarcolemmal channels by 2000-fold (Garlid et al., 1996). Glibenclamide, a potent and nonselective K_ATP channel blocker has been demonstrated to inhibit the mitoK_ATP in the heart and brain (Jaburek et al., 1998; Bajgar et al., 2001). 5-Hydroxydecanoic acid (5-HD) blocks the mitoK_ATP reconstituted in liposomes and isolated mitochondria, but it does not block cardiac type sarcolemmal K_ATP channels (Jaburek et al., 1998). There is, however, little information on the relationships between mitoK_ATP activation and regulation of PTEN phosphorylation in the angiogenesis in relation with apoptosis.

In the preliminary study, KR-31372 showed very weak vasodilator action despite having a benzopyran moiety in its structure, which is a striking contrast to levcromakalim, and showed a glibenclamide-inhibitable opening of K_ATP channel in isolated rat ventricular myocytes, suggestive of the K_ATP channel opener. Recently, KR-31372 exerted an inhibitory effect on the oxidized low-density lipoprotein-stimulated syntheses of [³H]thymidine incorporation and migrations of the cultured rat aortic smooth muscle cells (Kim et al., 2000). The antiangiogenic effect of KR-31372 was first demonstrated in rat sponge implant model, in that oral administration of KR-31372 (1 mg/kg for 7 days) significantly suppressed the neovascularization induced by angiotensin II and VEGF₁₆₅ (Kim et al., 2001).

In the present study, we designed to clarify the signal transduction mechanism by which KR-31372 exerts an inhibitory action on the angiogenesis in relation with apoptosis in the human umbilical vein endothelial cells (HUVECs) under treatment with and without 5-HD, a mitoK_ATP blocker. The results were compared with those of diazoxide, a specific mitoK_ATP opener. We examined the effect of KR-31372 against VEGF₁₆₅-induced increase in phosphorylated PTEN in the HUVECs and U87-MG cells, a PTEN-null glioblastoma cell line, transfected with expression vectors for sense PTEN in the absence and presence of 5-HD. To clarify whether KR-31372 increased the PTEN phosphorylation specifically in HUVECs, we used U-373 MG, a human brain glioblastoma cells for comparison with HUVECs.

Materials and Methods

Cell Cultures. HUVECs (ATCC CRL-1730, endothelial cell line derived from vein of human umbilical cord) were cultured in Kaighn's F-12K medium supplemented with 10% heat-inactivated fetal bovine serum, 0.1 mg/ml heparin sodium, 0.03 to 0.05 mg/ml endothelial cell growth supplement, and 1% antibiotics (100 U/ml penicillin and 100 μg/ml streptomycin). Cells were grown to confluence at 37°C in 5% CO₂ on 0.1% gelatin-coated culture dishes and used for experiments at no greater than passage 8. U-373 MG (KCLB 30017, human brain glioblastoma) and U87-MG (KCLB 30014, human brain PTEN-null glioblastoma) cells were cultured in Eagle's minimum essential medium (MEM) with 2 mM l-glutamine and 1.0 mM sodium pyruvate supplemented with 10% heat-inactivated fetal bovine serum.

Cell Proliferation Assay. HUVECs, seeded at a density of 1 × 10⁴ cells/well in 96-well plates, were incubated in growth media and allowed to attach for 24 h. Thereafter, cells were incubated for 3 h in Kaighn's F-12K medium containing 1% fetal bovine serum. Cells were exposed to KR-31372 (10^-8–10^-4 M) or diazoxide (10^-8–10^-4 M) for 3 h, and then were stimulated by VEGF₁₆₅ (10 ng/ml) for 48 h. After addition of 5-bromo-2′-deoxyuridine, cells were reincubated for 2 h for incorporation of 5-bromo-2′-deoxyuridine into the DNA of proliferating cells. The reaction product was quantified by measuring the absorbance at the 450 nm (reference wavelength 690 nm) using the ELISA reader (Bio-Tek Instruments, Winooski, VT).

Basal Tubular Formation Assays. The incubation of HUVECs were performed on 24-well plates that have been coated with 250 μl of growth factor-reduced Matrigel (10 mg of protein/ml) per well and polymerized for 30 min at 37°C. Cells were detached with 0.05% trypsin-EDTA solution, suspended in F-12K with 1% fetal bovine serum, and plated onto a layer of Matrigel at a density of 1 × 10⁵ cells/well. 5-HD was given to the cells 30 min before and after seeding. After 18 h, the culture plate was photographed (200×). The picture of the tubules was scanned and network area was determined using the GS-710 calibrated imaging densitometer (Bio-Rad, Hercules, CA).

Matrigel Plug Assay in Mice. Mice (C57BL/6) were injected subcutaneously with 0.5 ml of Matrigel containing VEGF₁₆₅ (5 ng/ml) with KR-31372 (10^-6 M) or diazoxide (10^-5 M). The injected Matrigel rapidly formed a single, solid gel plug. After 5 days, mice were sacrificed, and Matrigel plug was recovered, fixed with 10% formaldehyde/phosphate-buffered saline (pH 7.4), and embedded in paraffin and examined with hematoxylin and eosin stain. To quantify the formation of new blood vessel, the amount of hemoglobin was measured using the total hemoglobin kit (Sigma-Aldrich, St. Louis, MO). Four mice were used for one group, and the experiment was repeated twice.

DNA Fragmentation Assays. Cells were lysed in 1 ml of lysis buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM EDTA, 1% sodium dodecyl sulfate, and 0.5 mg/ml proteinase K). Digestion was continued for 1 to 3 h at 55°C, followed by addition of RNase A to 0.1 mg/ml and running dye (10 mM EDTA, 0.25% bromophenol blue, and 50% glycerol). Equivalent amounts of DNA (15–20 μg) were loaded into wells of 1.6% agarose gel and electrophoresed in 0.5× TAE buffer (40 mM Tris-acetate and 1 mM EDTA) for 2 h at 6 V/cm. DNA was visualized by ethidium bromide staining. Gel pictures were taken by UV transillumination with the Polaroid camera. Bands were quantified by the Molecular Analyst software using the Bio-Rad's Image Analysis System (Bio-Rad).

Western Blot Analyses. Cells were incubated for 24 h in the presence of 10 ng/ml VEGF₁₆₅. KR-31372 and diazoxide were applied 3 h, and 5-HD 30 min before experiment. For determination of Bcl-2, Bax protein, PTEN, phosphorylated PTEN, Akt, and phosphorylated Akt levels, cells were grown in 100-mm tissue culture dishes and treated with the indicated compounds. After washing, the cells were lysed in lysis buffer containing 50 mM Tris-Cl (pH 8.0), 150 mM NaCl, 0.02% sodium azide, 100 μg/ml phenylmethylsulflonyl fluoride, 1 μg/ml aprotinin, and 1% Triton X-100. After centrifugation at 12,000 rpm, 50 μg of total protein of each sample was loaded into 12% SDS-polyacrylamide gel electrophoresis gel, and transferred to nitrocellulose membrane (Amersham Biosciences, Inc., Piscataway, NJ). The blocked membranes were then incubated with the antibody of Bcl-2, Bax (Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), PTEN and phospho-PTEN (Ser380/Thr382/383), Akt, and phospho-Akt (Cell Signaling Technology, Inc., Beverly, MA).

Mitochondrial cytochrome c was prepared via following procedures. After washing cells (12 × 10⁶) with ice-cold phosphate-buffered saline, cell pellets were suspended in buffer (20 mM HEPES-KOH, pH 7.5, 10 mM KCl, 1.5 mM MgCl₂, 1 mM Na-EDTA, 1 mM Na-EGTA, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride) containing 250 mM sucrose. The cells were homogenized and then centrifuged twice at 750g for 10 min at 4°C. The harvested supernatants were centrifuged at 10,000g for 10 min at 4°C, and the resulting mitochondrial pellets were dissolved in 1× SDS sample buffer. Western blots were performed with the antibody of cytochrome c (Santa Cruz Biotechnology, Inc.). The immunoreactive bands were visualized using chemiluminescent reagent of the Super-Signal west dura extended duration substrate kit (Pierce Chemical, Rockford, IL). The signals of the bands were quantified using the calibrated imaging densitometer (GS-710; Bio-Rad). The protein concentration of the lysate was determined using the Bio-Rad DC assay kit (Bio-Rad).

Plasmid Construction. The expression of plasmid encoding the human PTEN protein was cloned by reverse transcription-polymerase chain reaction using the total RNA of SK-N-SH cells. Sequence analysis was performed to confirm the nucleotide sequences. The following sequences of oligodeoxynucleotides were used as primers containing linker recognizable by XhoI as underlined: 5′-GCGCTCGAGATGACAGCCATCAAAG-3′. Amplified 1264-bp fragments containing the human PTEN coding region were ligated into the XhoI site of pcDNA3.1 HisC (Invitrogen, Carlsbad, CA). pcDNA3.1-sPTEN is transcripted sense nucleotides.

DNA Transfection. U87-MG cells were seeded for 24 h before transfection in tissue culture dishes. At 50 to 70% confluence, the dishes were washed twice with Opti-MEM medium to remove the fetal bovine serum and a transfection cocktail containing 10 μg DNA and 10 μl of LipofectAMINE reagent (Invitrogen) per 100-mm dish was added. The medium was removed and then 7 ml of MEM medium containing 10% fetal bovine serum was added to each dish.

Drugs. VEGF₁₆₅ was purchased from the R & D Systems (Minneapolis, MN). (2R,3R,4S)-N″-Cyano-N-(6-nitro-3,4-dihydro-hydroxy-2-methyl-2-dimethoxymethyl-2H-1-benzopyran-4yl)-N′-benzyl guanidine (KR-31372) was donated from The Korea Research Institute of Chemical Technology (Daejon, Korea). 5-Hydroxydecanoic acid (sodium salt), endothelial cell growth supplement, and total hemoglobin kit were purchased from the Sigma-Aldrich (St. Louis, MO). 5-Bromo-2′-deoxyuridine kit was from the Roche Diagnostics (Mannheim, Germany). Matrigel was from the BD Biosciences Discovery Labware (Bedford, MA). KR-31372 and diazoxide were dissolved in dimethyl sulfoxide as a 10^-1 M stock solution. 5-Hydroxydecanoic acid was dissolved in distilled water as a 10^-1 M stock solution.

Statistics. The results are expressed as means ± S.E.M. Two-way repeated measures analysis of variance was used for the comparison of concentration-dependent changes in cell proliferation in response to agonists between inhibitor-treated and untreated groups. Statistical differences between groups were determined by paired or unpaired Student's t test or analysis of variance. P < 0.05 was considered significant.

Results

Tubular Formation. HUVECs plated on the Matrigel elongated and migrated to form a tubular network structure as shown in the morphological changes. When quantitatively estimated by measuring the length covered by the tubular network area using the image analysis program, both KR-31372 (10^-6 M) and diazoxide (10^-5 M) markedly suppressed the basal tubular formation to 53.5 ± 9.4% (P < 0.01) and 59.3 ± 8.7% (P < 0.01), respectively (Fig. 1, A and B). 5-HD (10^-5 M) applied 30 min before KR-31372 or diazoxide treatment significantly reversed the suppressed basal tubular formation induced by either KR-31372 or diazoxide. In the basal tubular formation, 5-HD (10^-5 M) alone showed little effect.

Matrigel Plug Assay in Mice. Five days after implantation, histological examination and hemoglobin content were estimated. As shown in Fig. 1, C and D, the neomicrovessel formation significantly increased in the Matrigel containing VEGF₁₆₅ (5 ng/ml), whereas it was not evident in the Matrigel without VEGF₁₆₅. Matrigel containing either KR-31372 (10^-6 M) or diazoxide (10^-5 M) significantly reduced the formation of VEGF₁₆₅-stimulated neomicrovessels, which was antagonized by 5-HD (10^-5 M). The neovascular formation was further confirmed by measurement of the hemoglobin content in the Matrigel. The hemoglobin content was elevated in the Matrigel containing 5 ng/ml VEGF₁₆₅ to 5.2 ± 0.6 g/dl (P < 0.001), which was markedly suppressed by treatment with 10^-6 M KR-31372 (1.6 ± 0.3 g/dl, P < 0.001) and 10^-5 M diazoxide (2.2 ± 0.4 g/dl, P < 0.05). Pretreatment with 5-HD (10^-5 M) significantly reversed both KR-31372- and diazoxide-induced reduction in hemoglobin content (Fig. 1D), indicating that KR-31372 as well as diazoxide elicit antiangiogenic activities in the in vivo experiment.

Cell Proliferation. VEGF₁₆₅ (1–20 ng/ml) concentration dependently increased DNA synthesis. After 48-h incubation, 10 ng/ml VEGF₁₆₅ increased DNA synthesis to 185.3 ± 7.7% of control cells (Fig. 2, inset). Cell proliferation was concentration dependently suppressed by simultaneous incubation with either KR-31372 or diazoxide (10^-8–10^-4 M, each), respectively. 5-HD (10^-5 M), when applied 30 min before KR-31372 or diazoxide treatment, significantly reversed the suppressed DNA synthesis induced by either KR-31372 or diazoxide (Fig. 2). After application of 5-HD (10^-5 M) in the absence of VEGF₁₆₅, the cells showed little effect on the proliferation of HUVECs (102.9 ± 3.7%).

Apoptotic Effect. Effect of KR-31372 was compared with diazoxide on the laddered feature of DNA fragmentation, which was pretreated with 10 ng/ml VEGF₁₆₅ in the HUVECs. Exposure of HUVECs to KR-31372 (10^-5 and 10^-4 M) and diazoxide (10^-5 M) induced prominent oligonucleosomal DNA fragmentation. Pretreatment with 5-HD (10^-5 M) strongly suppressed the DNA laddering induced by KR-31372 (10^-5 M) and diazoxide (10^-5 M), respectively (Fig. 3).

On the other hand, the effect of KR-31372 on the DNA fragmentation was identified in the U-373MG cells, naive U87-MG cells, and U87-MG cells of sPTEN. The cells were exposed to KR-31372 (10^-6–10^-4 M) for 3 h in the presence of VEGF₁₆₅ (10 ng/ml). KR-31372 (10^-6–10^-4 M) induced prominent oligonucleosomal DNA fragmentation in the U-373MG and U87-MG cells of sPTEN, but did not show DNA fragmentation in U87-MG lacking a wild-type PTEN (Fig. 4).

PTEN and Akt Phosphorylation. PTEN phosphorylation was concentration dependently increased by KR-31372 [1.82 ± 0.14-fold (P < 0.05) by 10^-6 M and 2.78 ± 0.31-fold (P < 0.01) by 10^-5 M KR-31372] when applied 3 h before VEGF₁₆₅ (10 ng/ml) in the HUVECs. Pretreatment with 5-HD (10^-5 M, P < 0.01) significantly suppressed the increased PTEN phosphorylation induced by KR-31372 (10^-5 M) to 0.88 ± 0.09-fold (P < 0.01). Diazoxide (10^-5 M) and 5-HD (10^-5 M) showed similar interactions as shown with KR-31372 and 5-HD (Fig. 5A). The similar results were also evident in the U-373MG cells, human brain glioblastoma cells (Fig. 5B). However, the PTEN protein expression was not altered by these agents.

Otherwise, PTEN protein was expressed in the HUVECs and U-373MG cells, but not in the naive U87-MG cells, whereas the Akt expression was well identified in the three cells. Introduction of PTEN cDNA (sense oligodeoxynucleotide) into the U87-MG cells, lacking a wild-type PTEN, caused a large increase in PTEN expression (Fig. 6A). In the U87-MG cells of sPTEN, both KR-31372 and diazoxide significantly increased the PTEN phosphorylation as shown in the HUVECs and U-373MG cells, and the phosphorylated Akt levels were, in contrast, significantly decreased by these agonists (Fig. 6B). The alterations induced by KR-31372 and diazoxide were well antagonized by 10^-5 M of 5-HD (Fig. 6, C and D). The levels of Akt and Akt phosphorylation were not altered by KR-31372 (10^-6–10^-4 M) and diazoxide (10^-5 M) in the U87-MG cells lacking a wild-type PTEN (data not shown).

Western Blot for Bcl-2, Bax, and Cytochrome c. Under treatment with VEGF₁₆₅ (10 ng/ml for 24 h), the Bcl-2 protein expression was markedly increased to 2.60 ± 0.54 relative density, which was strongly suppressed by KR-31372 (10^-6–10^-4 M) in a concentration-dependent manner (Fig. 7A). Diazoxide (10^-5 M) also significantly decreased the Bcl-2 protein level. In the presence of 5-HD (10^-5 M), both KR-31372- and diazoxide-induced inhibitions of Bcl-2 protein levels were significantly reversed (Fig. 7B).

In the presence of VEGF₁₆₅ (10 ng/ml), the basal levels of Bax protein and cytochrome c release from mitochondria were 0.77 ± 0.06 and 0.97 ± 0.08 relative densities, respectively. In contrast, both were largely elevated by pretreatment with KR-31372 (10^-6, 10^-5, and 10^-4 M) and diazoxide (10^-5 M). 5-HD (10^-5 M) significantly suppressed the increased Bax protein and cytochrome c release induced by either KR-31372 or diazoxide (Fig. 8, A and B).

Discussion

In the present study, the major findings were that KR-31372 as well as diazoxide markedly suppressed the in vitro basal tube formation in HUVECs and in vivo Matrigel-induced neovascularization in mice in association with augmentation of oligonucleosomal DNA fragmentation. Increased PTEN phosphorylation stimulated by KR-31372 and diazoxide was accompanied by decreased Akt phosphorylation and suppression of Bcl-2 expression in accordance with enhanced Bax protein level and cytochrome c release from mitochondria. All these variables were significantly reversed by 5-HD, a mitoK_ATP blocker. No previous study has defined the implication of mitoK_ATP opening- and PTEN phosphorylation-linked apoptotic mechanism in the antiangiogenesis.

The angiogenesis requires the triad processes: endothelial cell proliferation, migration, and local protease activity such as matrix metalloproteinases (Hanahan and Folkman, 1996). Overexpression of VEGF and its receptors is associated with chronic inflammation, tumor growth, and diabetic retinopathy (Williams, 1998). Of the various VEGF species, VEGF₁₆₅ is well characterized (Neufeld et al., 1999), and the expression of cell surface receptors for VEGF₁₆₅ was demonstrated in the endothelial cells (Millauer et al., 1993). Binding of VEGF to VEGF receptor-2 leads to the receptor phosphorylation and subsequent activation of PI3-K, phospholipase C-γ1, and Src family tyrosine kinases (Thakker et al., 1999).

PTEN was originally identified as a tumor suppressor gene based on its high frequency of mutation in a variety of tumors (Li et al., 1997). 3-Phosphoinositides are important substrates for PTEN both in vitro (Maehama and Dixon, 1998) and in vivo experiments (Sun et al., 1999). PTEN potently modulates VEGF-mediated signaling and function. Most recently, Huang and Kontos (2002) have provided evidence that adenovirus-mediated overexpression of a dominant negative PTEN mutant enhances VEGF-mediated endothelial cell proliferation and migration, whereas overexpression of wildtype PTEN, in contrast, shows inhibition of cell proliferation and chemotactic effects of VEGF. Our data showed for the first time that both KR-31372 and diazoxide exerted concentration-dependent increase in PTEN phosphorylation in the HUVECs, which was suppressed by 5-HD (10^-5 M, P < 0.01). PTEN has the property to restrain the PI3-K pathway by catalyzing degradation of the PI(3,4,5)P₃, which is generated by PI3-K (Stambolic et al., 1998; Cantley and Neel, 1999; Huang et al., 2001). It is known that overexpression of PI3-K and its downstream effector Akt mediate growth factor-induced neuronal survival, and Akt in turn up-regulates Bcl-2 promoter activity in association with Bcl-2 protein expression through enhanced cAMP response element-binding protein activation (Walton et al., 1999; Pugazhenthi et al., 2000). Therefore, it is speculated that activation of PTEN in the HUVECs may provide a signaling event that may induce a decrease in phosphorylated Akt and Bcl-2 expression followed by increased Bax protein and cytochrome c release from mitochondria, thereby enhancing cell apoptosis.

In our results, when the U87-MG cells, a glioblastoma cell line that lacks expression of wild type PTEN (Haas-Kogan et al., 1998), were transfected with expression vectors for sense PTEN, they exerted increased PTEN protein expression. These transfected cells also showed high phosphorylated PTEN accompanied by decreased Akt phosphorylation under treatment with KR-31372 and diazoxide, which was fully reversed by mitoK_ATP blocker 5-HD. Based on these facts, it is likely that KR-31372 has a role for regulation of PTEN phosphorylation as a mitoK_ATP opener.

In the previous results, glibenclamide reversed the KR-31372-induced suppression of the VEGF₁₆₅-stimulated chemotactic motility of HUVECs (Kim et al., 2003). At that time, they did not identify whether the action site of glibenclamide is on the sarcolemmal K_ATP channels or the mitoK_ATP channels. In the light of the present study, it seemed that the action of glibenclamide was ascribed to the inhibition of the mitoK_ATP (Jaburek et al., 1998; Bajgar et al., 2001). Tamura et al. (1998) have emphasized the importance of PTEN in the cell migration, in that overexpression of PTEN dephosphorylates FAK in vivo and in vitro, and reduces its tyrosine phosphorylation and inhibits integrin-mediated cell spreading, whereas antisense PTEN enhanced migration. Most recently, Kim et al. (2003) have demonstrated that KR-31372 significantly inhibited the KDR/Flk-1 tyrosine phosphorylation-linked extracellular signal-regulated kinase 1/2, p38 mitogen-activated protein kinase, and p125^FAK tyrosine phosphorylation, which were inhibited by glibenclamide, a K_ATP channel blocker. Considering that KR-31372 significantly suppressed the VEGF-stimulated triad of angiogenesis (DNA synthesis, migration, and MMP-2 release; data not shown) in HUVECs, it is likely predictable that KR-31372 might suppress the tube formation in HUVECs and Matrigel-induced neovascularization in mice via mediation of the mitoK_ATP opening, thereby leading to the antiangiogenesis.

Accumulating reports have shown that the mitoK_ATP is the receptor for cardioprotection by K_ATP channel openers and for 5-HD blockade and that 5-HD inhibits the mitoK_ATP, but not the sarcolemmal K_ATP channel (McCullough et al., 1991; Grover and Garlid, 2000). Garlid et al. (1996, 1997) showed that diazoxide was 1000 to 2000 times more potent in opening mitoK_ATP than in opening sarcolemmal K_ATP channels. Thus, it is intriguing that the mitoK_ATP openers, including diazoxide and KR-31372, increased the expression of PTEN phosphorylation and induced 5-HD-inhibitable apoptosis in the HUVECs, despite a number of studies demonstrating a cardioprotection by diazoxide and cromakalim as openers of mitoK_ATP (Grover and Garlid, 2000). Currently, it is not possible to reconcile the disparate results of diazoxide and KR-31372 between cardiac myocytes and HUVECs. In contrast to other cells, the endothelial cells possessed the unique properties. The electrochemical gradient or driving force for Ca²⁺ entry into endothelial cells is characteristically influenced by the membrane potential (Lückhoff and Busse, 1990), such that depolarization decreases, whereas hyperpolarization increases, Ca²⁺ entry. Resting membrane potentials in endothelial cells are thought to be controlled primarily by K⁺ channels; in particular, inwardly rectifying K⁺ channels (Nilius et al., 1997).

Although the data are not shown, mitochondrial membrane potential and intramitochondrial calcium levels were decreased by both KR-31372 and diazoxide in a concentration-dependent manner. These values were significantly reversed by treatment with 5-HD. Moreover, these reductions were accompanied by increase in Bax protein and increased cytochrome c release, indicating that KR-31372 and diazoxide augmented the apoptotic action via mitoK_ATP opening. Presently, it remains to be clarified how the mitoK_ATP opening triggers to increase the PTEN phosphorylation.

Apoptosis of endothelial cells occurs during the vascular regression in the process of scar formation (Desmouliere et al., 1995), atherosclerosis (Dimmeler et al., 1998), and progressive glomerulonephritis (Shimizu et al., 1997). The expression of Bcl-2 protein in the mitochondrial outer membrane prevents the association of the proapoptotic Bax protein with permeability transition pore and its pore forming activity, and then acts to inhibit cytochrome c release from mitochondria to cytosol (Kluck et al., 1997; Shimizu and Tsujimoto, 2000). The consequence of mitoK_ATP activation in the isolated rat heart mitochondria was emphasized by Holmuhamedov et al. (1998), in that K_ATP channel openers (pinacidil, cromakalim, and levcromakalim) induced mitochondrial membrane depolarization with an increase in the rate of mitochondrial respiration and consequently a decrease in ATP synthesis, these cascades resulting in enhanced release of cytochrome c from mitochondria.

Together, these findings provide a strong evidence to support that the proapoptotic effect of KR-31372 in the HUVECs is closely linked to 5-HD-sensitive mitoK_ATP opening and the up-regulation of PTEN phosphorylation and down-regulation of Akt phosphorylation and Bcl-2 protein expression, thereby leading to antiangiogenesis in the HUVECs.