INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

HIGH-MOLECULAR-WEIGHT KININOGEN (HK) is a plasma protein that was first identified as a precursor of the bioactive peptide bradykinin, a potent vasodilator that regulates many cardiovascular processes. It is now recognized that HK is a multifunctional protein that plays important roles in many pathophysiological processes, such as in fibrinolysis, thrombosis, and inflammation (11). HK is a 120-kDa single-chain glycoprotein consisting of six domains (designated as D1-D6, respectively) with a plasma concentration ~670 nM (11). HK can specifically and reversibly bind to endothelial cells in a Zn²⁺-dependent manner (9). HK bound to endothelial cells is a substrate of plasma kallikrein, which releases bradykinin from D4 through proteolytic cleavage. The remaining portion of the molecule [angiogenic inhibitor HK (HKa)], which contains a heavy chain and light chain linked together through a single disulfide bond, undergoes major conformational changes, and as a result, has acquired new properties (29). The endothelial cell surface is an important site for the generation of bradykinin and HKa, which in turn, may affect the physiology of endothelial cells. Whereas bradykinin has been intensively studied, the physiological implication of HKa generation is not clear. Recent studies from our laboratory (8, 15) and by other investigators (31) indicate that HKa may act as a naturally occurring angiogenic inhibitor in circulation. These findings represent a novel biological function of this molecule. Interestingly, it has recently been demonstrated that bradykinin stimulates angiogenesis (10, 21, 24). Therefore, HK is a precursor of both an angiogenic stimulator (bradykinin) and an inhibitor (HKa). Because of their vastly different half-lives in the circulation and their distinct cell surface receptors, HKa and bradykinin may play divergent roles in the regulation of angiogenesis.

Angiogenesis is the formation of new capillary from preexisting blood vessels, a process crucial for normal physiological events such as embryogenesis and wound healing. It also occurs under pathological conditions such as tumor development. Angiogenesis involves several steps, beginning with localized degradation of the basement membrane of the existing vessels by proteases bound to the endothelial cell membrane. This process is followed by the detachment of endothelial cells from adhesive proteins in the extracellular matrix (ECM) and migration into the perivascular space where endothelial cells proliferate. The new endothelial cells then form tubelike structures that eventually join and form new capillaries (12). This is a highly regulated process by both positive and negative regulators (1, 2). Many growth factors and cytokines stimulate angiogenesis. Basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF) are among the well-characterized angiogenic factors (1). The recent identification of several endogenous peptides with antiangiogenic activity provides important insight into how angiogenesis is negatively regulated. An emerging paradigm has been developed that certain proteolytic fragments of plasma or ECM proteins are naturally occurring angiogenesis inhibitors (2). Angiostatin derived from plasminogen and endostatin, a fragment of collagen XVIII, are prototypes of this group of polypeptides (4, 23). Generation of HKa from HK through proteolytic cleavage follows a similar model to the formation of angiostatin and endostatin.

Successful angiogenesis depends not only on endothelial cell proliferation but also critically relies on the preservation of endothelial cell viability of the new vessels. Therefore, inhibiting proliferation and/or inducing apoptosis among endothelial cells appear to be effective ways to abolish new vessel formation. These mechanisms may underline the efficiency of several peptide angiogenesis inhibitors (3, 13, 28). In a recent study (8), we have shown that HKa exhibited a potent antiangiogenic effect by inhibiting in vivo neovascularization. D5 resembles the effect of HKa; thus it has been named kininostatin (8). Following this initial report, we (15) further reported that HKa and D5 were able to induce apoptosis of proliferating endothelial cells, an event associated with a decreased expression of cyclin D1 and reduced cell proliferation. These results indicate that the proapoptotic effect of HKa and D5 may be related to their ability to interfere with cell cycle regulation. To test this hypothesis, we further investigated whether HKa affects the expression or activity of other cell cycle regulators. We report here that HKa-induced apoptosis was correlated with an increased expression of Cdc2 kinase and cyclin A. Cdc2 was initially characterized as a member of the cyclin-dependent protein kinase (Cdk) family, which plays a central role in the cell cycle regulation. However, several studies have shown that increased expression/activation of Cdc2 and certain members of the Cdk family were detected in apoptotic cells induced by several distinctive stimuli. These observations led to the hypothesis that certain cell cycle regulators are important components of the apoptotic pathway (19). Apoptosis of endothelial cells induced by the angiogenesis inhibitor HKa described in this study shows characteristics that fit this hypothesis.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Materials. HKa was purchased from Enzyme Research Laboratories (South Bend, IN), bFGF was purchased from Life Technologies (Grand Island, NY), and glutathione S-transferase (GST) and recombinant D5 (GST-D5) were prepared as previously described (15). Anti-Cdc2 bodies were from New England Biolabs (Beverly, MA). Antibodies against cyclin A and cyclin B were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). An annexin V-fluorescein isothiocyanate (FITC) apoptosis detecting kit was purchased from BD Biosciences (Palo Alto, CA), and a Cdc2 kinase assay kit was purchased from Promega (Madison, WI).

Cell culture. Human umbilical vein endothelial cells (HUVEC) and endothelial cell culture media were purchased from Clonetics (Walkersville, MD). The cells were maintained in endothelial cell growth medium (EGM; containing growth factors and 10% fetal calf serum) at 37°C in a humidified incubator (95% O₂-5% CO₂). Cells from passages 3 to 7 were used. Zn²⁺ is required for HKa and D5 binding to endothelial cells; therefore, the cell culture medium contained 15 µM ZnCl₂, which did not affect cell viability.

Apoptosis and cell viability analysis. Apoptotic cell death induced by D5 has been described in our previous report (15). In this study, HKa-induced apoptosis was assessed by nuclear morphology at the individual cell level and was then further quantified by flow cytometry. Briefly, HUVEC were seeded on gelatin-coated dishes at cell numbers that gave ~50% confluence. The cells were incubated in EGM medium for 3 h to allow for cell attachment, and then the medium was changed to endothelial cell basic medium (EBM, EGM minus growth factors and serum). The cells were treated with 10 ng/ml bFGF in the absence or presence of 200 nM HKa for the times indicated in the individual experiments. Cell morphological changes were examined under a microscope periodically during and after cell treatment. Apoptotic cell death was detected by nuclear fragmentation analysis with Hoechst-33285 nuclear staining as previously described (14). Apoptosis was further confirmed and quantified with an annexin V-FITC apoptosis detection kit. In this method, treated cells were washed twice with cold phosphate-buffered saline (PBS) and then resuspended in a binding buffer [10 mM HEPES (pH 7.4), 140 mM NaCl, 2.5 mM CaCl₂] at a concentration of 1 × 10⁶ cells/ml. The cells were stained with annexin V-FITC and propidium iodide (PI) according to the manufacturer's instructions and then analyzed with the use of a flow cytometer (FACScan, Becton Dickinson). For determination of the number of viable cells after treatment, the medium containing late apoptotic cells (detached cells) was removed. The viable cells attached to the dishes were washed with PBS and then released by brief trypsin/EDTA treatment. The cells were then collected and counted with a hemocytometer.

Cell cycle analysis. Cell proliferation was initiated by stimulating subconfluent HUVEC with bFGF. The number of cells at different cell cycle stages after treatment was determined by flow cytometry. The attached cells were trypsinized and combined with floating cells collected from medium. Cells were washed with PBS containing 1% FBS and then fixed with 80% ethanol at 4°C for 15 min. Fixed cells (1 × 10⁶ cells/ml) were stained with 20 µg/ml PI in PBS buffer containing 1% FBS, 0.05% Triton X-100, and 50 µg/ml RNase. After a 30-min incubation at room temperature, the percentage of cells in different phases of the cell cycle was determined by DNA content (PI intensity) from at least 7,500 cells with a FACScan flow cytometer (Becton Dickinson).

Cell lysate preparation, Western blot analysis, and Cdc2 kinase activity assay. After treatment, cells in the medium and attached to dishes were collected and combined. They were resuspended in an ice-cold lysis buffer containing 25 mM Tris · HCl (pH 7.5), 450 mM NaCl, 1 mM Na₃VO₄, 1% Triton X-100, 0.5 mM EDTA, 1 µM leupeptin, and 1 µg/ml aprotinin. Cell suspension was kept on ice for 20 min, and then the solution was clarified by centrifugation at 15,000 g for 15 min 4°C. The supernatant was designated as cell lysate and was used for Western blot and Cdc2 kinase activity analysis. Protein concentration of the cell lysate was determined using a protein assay kit (Pierce). Western blot analysis has been described in our previous study (15). Cdc2 kinase activity was determined according to the method described by Thomas et al. (27) by using the Promega SignaTECT Cdc2 kinase assay system. Briefly, Cdc2 kinase activity was determined with a kinase buffer containing 50 mM Tris · HCl (pH 7.5), 10 mM MgCl₂, 1 mM EGTA, 2 mM DTT, 40 mM -glycerolphosphate, 20 mM p-nitrophenylphosphate, 0.1 mM sodium vanadate, 25 µM biotinylated peptide substrate specific for Cdc2 (PKTPKKAKKL), and 50 µM ATP (containing ~1 µCi [-³²P]ATP). The reaction was initiated by adding 3 µg of protein sample and then incubated for 10 min at 37°C. The reaction was terminated by adding 12.5 µl of 7.5 M guanidine hydrochloride to the reaction mixture. An aliquot of 15 µl was spotted on the streptavidin-coated membrane. Radioactivity (in counts/min) was determined by scintillation counting of the membranes. Cdc2 kinase activity was calculated by subtracting the counts per minute obtained in the absence of substrate from the counts per minute in reactions containing the substrate.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

HKa upregulates expression of Cdc2 kinase and cyclin A. Cdc2 is a member of the Cdk family whose activity is regulated by cyclin A or cyclin B. We first examined the effect of HKa on the expression of Cdc2, cyclin A, and cyclin B in bFGF-stimulated proliferating endothelial cells by Western blot analysis. As shown in Fig. 1A, treatment of HUVEC with bFGF in the presence of HKa for 48 h (FGF + HKa), a condition where HKa and D5 have been shown to inhibit cell proliferation (15), dramatically increased the expression of Cdc2 and cyclin A compared with the control (FGF). The expression of cyclin B under the same conditions was not affected by HKa (data not shown). The effect of HKa on Cdc2 and cyclin A expression was further tested at two additional time points. The expression levels of Cdc2 and cyclin A at 12 h incubation were not affected by HKa treatment but were significantly increased after the cells were treated for 24 h (Fig. 1A, FGF + HKa). Whereas the expression of Cdc2 and cyclin A was apparently affected by HKa treatment, protein levels of Cdc2 and cyclin A in the control experiments remained relatively consistent at all time points tested (Fig. 1A, FGF). Actin was used as a control to show the protein loading on each lane in this experiment. We observed that recombinant D5 (GST-D5) also increased expression of Cdc2 and cyclin A in a similar pattern to HKa but is less potent than HKa (data not shown).

View larger version (37K): [in this window] [in a new window]   Fig. 1.   A: effect of cleaved high-molecular-weight kininogen (HKa) on the expression of Cdc2 and cyclin A. Human umbilical vein endothelial cells (HUVEC) were incubated in endothelial growth medium (EGM) for 3 h to allow cells for attachment. The cells were then incubated in endothelial cell basic medium (EBM) in the presence of 10 ng/ml basic fibroblast growth factor (bFGF) (FGF) as a control, or treated with 200 nM HKa in the presence of bFGF (FGF + HKa). Cell lysates were prepared from cells treated at the times indicated. Expression of Cdc2 and cyclin A (Cyc A) was detected with their specific antibodies by Western blot analysis. The blots were then reprobed with anti-actin antibodies to show the protein loading on each lane. B: effect of HKa on Cdc2 kinase activity. The cells were treated with bFGF alone (FGF) or bFGF in the presence of HKa (FGF + HKa) under the conditions described in A. Cdc2 activity was determined with a kinase assay kit as described in MATERIALS AND METHODS. Relative Cdc2 activity at each time point was expressed as fold of activation by comparing with the activity determined from bFGF-treated cells at 12 h (12 FGF). Results are means ± SE of three experiments.

HKa-induced apoptosis is correlated with increased expression of Cdc2 and cyclin A. To test whether HKa-induced expression of Cdc2 and cyclin A affected cell viability, bFGF- stimulated proliferating HUVEC were treated with HKa in the same time course as described in Fig. 1. Consistent with its inhibitory effect on cell proliferation (31), HKa treatment decreased the number of viable cells in a time-dependent manner (Fig. 2), a result inversely related to the increased expression of Cdc2 and cyclin A induced by HKa, as described in Fig. 1.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Effect of HKa on the viability of HUVEC. Cells were incubated in EGM medium for 3 h to allow cells for attachment. The cells were then incubated in EBM in the presence of 10 ng/ml bFGF (FGF) or in the presence of bFGF plus 200 nM HKa (FGF + HKa). Floating cells in the media were washed away after incubation for the times as indicated. Viable cells attached to the dishes were released with a trypsin/EDTA solution. The number of cells was determined with a hemocytometer. Results are means ± SE of four experiments.

View larger version (157K): [in this window] [in a new window]   Fig. 3.   Effect of HKa on the morphology of HUVEC. Cells were treated under the same conditions as described in Fig. 2. Cell morphology was examined under a microscope with a ×10 phase contrast objective lens and photographed with a charge-coupled device camera. Arrows indicate the cells showing apoptotic morphology.

View larger version (54K): [in this window] [in a new window]   Fig. 4.   Analysis of HKa-induced apoptosis by Hoechst-33285 nuclear staining. HUVEC were treated with bFGF alone (FGF) or bFGF in the presence of HKa (FGF + HKa) under the same conditions as described in Fig. 3 (48 h). The cells were fixed and stained with 10 µM Hoechst-33285 for 30 min and analyzed under a fluorescence microscope with excitation at 340 nm with the use of a × 20 objective lens. Arrows indicate fragmented (apoptotic) nuclei.

View larger version (43K): [in this window] [in a new window]   Fig. 5.   Analysis of HKa-induced apoptosis by annexin-fluorescein isothiocyanate (FITC) staining. Cells were treated under the same conditions as described in Fig. 3. The collected cells were stained with annexin-FITC apoptosis detection kit and analyzed by flow cytometry by the method described in the MATERIALS AND METHODS. The data shown are from a representative experiment repeated three times. PI, propidium iodide; FL1-H, fluorescence intensity for Annexin-FITC; FL2-H, fluorescence intensity for PI.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   Effect of HKa on the distribution of cells at different phases of cell cycle. A: 12 h; B: 24 h; C: 48 h. Cells were treated with bFGF alone (FGF) or in the presence of HKa (FGF + HKa). After incubation for the times indicated, the cells were detached with a trypsin/EDTA solution. They were then fixed, stained with PI, and analyzed for cell cycle profiles. Histograms of cell numbers vs. DNA content (PI staining intensity) were generated by flow cytometric analysis. Data shown are from a representative experiment repeated three times. G0/G1, cells at G0 and G1 phase; S, cells at S phase; G2/M, cells at G2/M phase; subG0/G1, cells with a subdiploid DNA content.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

HK is a relatively abundant plasma protein with a concentration of 670 nM. Previous studies (6, 10, 31) have shown that HKa at nanomolar concentrations inhibits cell proliferation and adhesion and induces apoptosis of endothelial cells in the presence of Zn²⁺. Recombinant domain 5 of HKa (GST-D5) partly resembles these effects of HKa, indicating that D5 is likely the active domain responsible for the antiangiogenic activity of HKa. bFGF has been used in most of the previous studies because it is a well-known angiogenic growth factor. However, it should be pointed out that the effect of HKa or D5 is not specific to bFGF because they also induce apoptosis of proliferating endothelial cells stimulated by VEGF, human growth factor, and platelet-derived growth factor (31) and inhibit neovascularization stimulated by VEGF in chicken chorioallantoic membrane assay (unpublished data). The present study further characterized the effect of HKa on cell viability and the cell cycle by using bFGF as an angiogenic growth factor.

The cell cycle is a precisely regulated cellular event primarily by Cdks and cyclins (17). Cdc2 has been known to play a central role in this process. Its kinase activity is regulated at multiple levels by its binding with cyclin A or cyclin B, by phosphorylation/dephosphorylation, and by Cdk inhibitors. The Cdc2 protein level normally remains relatively constant throughout the cell cycle (17). Several studies have shown that increased activity and/or expression of Cdc2 and cyclin A or cyclin B are found in cells in response to several apoptotic stimuli, such as tumor necrosis factor- (18), radiation (25), transforming growth factor-₁ (7), and microtubule damaging agents (5). These observations led to the hypothesis that the unscheduled expression/activation of Cdc2 and certain types of Cdk2, especially regulated by cyclin A or cyclin B, can result in activation of the apoptotic pathways. Therefore, these cell cycle regulators are also mediators of apoptosis (19). The results described in this study show the features that fit this hypothesis. Expression and activity of Cdc2 and cyclin A is significantly upregulated in HKa-treated endothelial cells at time points when HKa caused a significant amount of apoptosis, indicating that elevated Cdc2 kinase is implicated in HKa-induced apoptosis. Interestingly, an increase in Cdc2 kinase activity was also detected in control cells at 48-h incubation. However, the change of Cdc2 kinase activity under this condition can be different in nature from that caused by HKa because it did not affect cell viability, and was not associated with a change in Cdc2 protein level, which is normally constant during the cell cycle (17). It is not clear what caused the increase in Cdc2 activity at this time point in control cells, but it is likely a reflection of dynamic fluctuation of Cdc2 kinase activity during the normal cell cycle, a known phenomenon that can result from phosphorylation/dephosphorylation of Cdc2, by its association with cyclin A, cyclin B, or by Cdk inhibitors at different stages of the cell cycle.

Apoptosis caused by uncontrolled expression/activation of Cdc2 is often associated with cell cycle arrest. For example, transforming growth factor-₁-induced apoptosis of hepatoma cells is correlated with an accumulation of G2/M cells, a concurrence of increased Cdc2 activity (7). Similarly, paclitaxel (a microtubule damaging compound) stimulated accumulation and activation of Cdc2 during apoptosis of ovarian carcinoma cells. The elevated Cdc2 activity/expression in those cells was determined to be a consequence of cell cycle arrest at G2/M phase (5). Apparently, this is not the case of HKa-induced apoptosis of endothelial cells because HKa did not cause a detectable change in the proportion of cells at different phases of the cell cycle, although it markedly increased the activity and protein level of Cdc2. On the other hand, our results are similar to observations in apoptosis induced by gamma radiation (25) and by granzyme B (26), where cell cycle profiles were not affected by either treatment. The authors of those studies concluded that apoptosis occurs at all stages of the cell cycle and that elevated Cdc2 activity was not the result of cell cycle arrest. The same conclusion may apply to HKa-induced apoptosis of endothelial cells. Therefore, an upregulated expression and/or activation of Cdc2 seems to be a common feature among apoptosis induced by distinctive stimuli, but it may have a different consequence on cell cycle progression, and the mechanisms causing Cdc2 activation may differ depending on cell types as well as the nature of apoptotic insults. Whereas it is conceivable that the deregulated activity of Cdc2 may disrupt normal cell cycle progression and thus result in mitotic arrest, followed by apoptotic cell death, it is also possible that Cdc2 may directly couple with an apoptotic pathway without apparent effect on the cell cycle progression. In support of this view, a recent study (16) in neurons suggests that Cdc2 can directly phosphorylate the apoptotic protein BAD, thus induce apoptosis of these cells. Therefore, Cdc2 can be part of the apoptotic machinery.

In addition to the cellular and biochemical evidence, genetic studies have demonstrated that expression of the dominant negative mutants of Cdc2 or cyclin A can suppress apoptosis induced by tumor necrosis factor- and staurosporine (20), whereas expression of wild-type Cdc2 restored the sensitivity of FT210 cells (a Cdc2 mutant murine mammary cell line) to noscapine-induced apoptosis (30). These studies have established active roles of Cdc2 and cyclin A during apoptosis in these cells. However, it is sometimes controversial because other investigators (22) have reported that Cdc2 is not required in apoptosis in some cells such as thymocytes. The fact that a variety of unrelated agents can induce a similar response during apoptosis indicates that elevated expression and/or activation of Cdc2 is not unique to a specific apoptotic stimulus. As encountered in many cases regarding the role of Cdc2 during apoptosis, it is difficult to conclude from existing evidence whether the increased expression and activity of Cdc2 by HKa is a cause of cell death or a result of apoptotic response. Nevertheless, our results strongly suggest that HKa-induced apoptosis is closely related to the unregulated expression/activation of Cdc2 and cyclin A. This notion is consistent with our previous study (15), in which we have shown that HKa or D5 primarily induced apoptosis of proliferating endothelial cells stimulated by bFGF but not of the cells cultured in the absence of the growth factor. Zhang et al. (31) had a similar observation for HKa. It appears that proliferating endothelial cells may be more sensitive to the apoptotic effect of HKa than quiescent cells. Furthermore, we have shown that some mitotic cells are simultaneously identified as apoptotic cells (15). Taken together, we may conclude that HKa-induced apoptosis of endothelial cells is closely related to their proliferating state, which may explain the observation that HKa selectively induced apoptosis among proliferating cells.