Purpose: Molecular antagonists of the inhibitor of apoptosis protein survivin have shown promise as novel anticancer strategies for triggering tumor cell apoptosis, dysregulating mitotic progression, and inhibiting tumor growth in preclinical models. However, how survivin couples to the cell death machinery has remained elusive, and the relevant cellular targets of survivin antagonists have not been completely elucidated.
Experimental Design: Human umbilical vein and dermal microvascular endothelial cells were infected with replication-deficient adenoviruses encoding survivin (pAd-Survivin), green fluorescent protein (pAd-GFP), or a phosphorylation-defective survivin Thr34→Ala (pAd-T34A) dominant negative mutant. The effect of wild-type or mutant survivin was investigated on capillary network stability, endothelial cell viability, and caspase activation in vitro and on kinetics of tumor growth and development of angiogenesis in a breast cancer xenograft model in vivo. The cell death pathway initiated by survivin targeting was mapped with respect to cytochrome c release, changes in mitochondrial transmembrane potential, and apoptosome requirements using mouse embryonic fibroblasts deficient in Apaf-1 or caspase-9.
Results: Adenoviral transduction of endothelial cells with pAd-Survivin inhibited growth factor deprivation- or ceramide-induced apoptosis, reduced caspase-3 and -7 generation, and stabilized three-dimensional capillary networks in vitro. Conversely, expression of pAd-T34A caused apoptosis in umbilical vein and dermal microvascular endothelial cells and resulted in caspase-3 activity. Cell death induced by survivin targeting exhibited the hallmarks of mitochondrial-dependent apoptosis with release of cytochrome c and loss of mitochondrial transmembrane potential and was suppressed in Apaf-1 or caspase-9 knockout mouse embryonic fibroblasts. When injected in human breast cancer xenografts, pAd-T34A inhibited growth of established tumors and triggered tumor cell apoptosis in vivo. This was associated with a ∼60% reduction in tumor-derived blood vessels by quantitative morphometry of CD31-stained tumor areas, and appearance of endothelial cell apoptosis by internucleosomal DNA fragmentation in vivo.
Conclusions: Survivin functions as a novel upstream regulator of mitochondrial-dependent apoptosis, and molecular targeting of this pathway results in anticancer activity via a dual mechanism of induction of tumor cell apoptosis and suppression of angiogenesis.
The evasion from apoptosis, or programmed cell death, is an invariant molecular trait of human cancer (1), which may facilitate the acquisition of additional cancer traits (2) promoting resistance to therapy and disseminated disease. Cancer cells can achieve escape from apoptosis from allelic loss of cell death activators (3), deregulated overexpression of apoptosis inhibitors (4), or inactivating mutations in executioner molecules of cell death (5, 6, 7), i.e., caspases. The net result is an increased antiapoptotic threshold that allows cancer cells to bypass internal surveillance checkpoints, thrive in unfavorable microenvironments, and acquire an invasive phenotype (1, 2).
Among the regulators of apoptosis that may participate in cancer (4), interest has been recently focused on survivin (8). A member of the IAP5 gene family (9), survivin is expressed in most human tumors but is largely undetectable in normal differentiated tissues and correlates with reduced tumor cell apoptosis in vivo, abbreviated patient survival, accelerated rates of recurrences, and increased resistance to therapy (10). Molecular antagonists of the survivin pathway, including antisense, ribozymes, or dominant negative mutants, have shown efficacy in causing spontaneous apoptosis in tumor cells, enhancement of cell death stimuli, including chemotherapy and irradiation, and catastrophic mitotic defects (10). Consistent with the idea that manipulation of apoptotic pathways could provide novel cancer therapeutics (11), molecular targeting of survivin showed promising results in preclinical models in vivo, suppressing de novo tumor formation and inhibiting growth of established tumors, alone or in combination with other anticancer approaches (reviewed in Refs. 10 and 12). However, critical gaps in our understanding of the survivin pathway still exist that have hampered its full exploitation for cancer therapeutics. Specifically, how survivin couples to the cell death machinery (13) has not been fully elucidated, and the potential cellular targets of survivin antagonists have not been clearly defined (10). The complexity of the survivin pathway may extend beyond the tumor cell population, and increased survivin expression has been demonstrated in endothelial cells during the proliferative (14, 15) and remodeling (16, 17) phases of angiogenesis, potentially acting as a cytoprotective mechanism for these cells (18).
To conclusively credential the survivin pathway for cancer therapeutics, we used a phosphorylation-defective survivin Thr34→Ala dominant negative mutant (19) and mapped the link between survivin and the cell death machinery and its implications for tumor growth in vivo. We found that survivin functions as a novel upstream regulator of mitochondrial-dependent apoptosis and that this pathway is required for preservation of tumor cell viability as well as maintenance of tumor-associated angiogenesis in vivo.
MATERIALS AND METHODS
Cell Cultures and Antibodies.
HUVECs or DMVECs were purchased from Clonetics (San Diego, CA) and maintained in culture on gelatin-coated 6-well clusters in M199 medium supplemented with 20% FCS, 50 μg/ml endothelial cell growth supplement, 100 μg/ml heparin, 100 μg/ml penicillin, and 100 μg/ml streptomycin (all from Life Technologies, Inc., Grand Island, NY) in 5% CO2 at 37°C. The Tet-inducible human melanoma YUSAC-2 cell line expressing wild-type survivin or the phosphorylation-defective survivin Thr34→Ala mutant [survivin(T34A)] upon Tet removal (Tet-off system) was characterized previously (20). MEFs originated from wild-type, Apaf-1−/− (21), or caspase 9−/− (22) embryos have been described previously. MEFs were cultured in DMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 100 μg/ml streptomycin, and 0.05% β-mercaptoethanol. Breast carcinoma MCF-7 and kidney epithelial HEK293 cells were obtained from the American Type Culture Collection (Manassas, VA) and maintained in culture according to the supplier’s specifications. An affinity-purified rabbit antibody to full-length recombinant survivin (1:1,000; NOVUS Biologicals, Littleton, CO) was characterized previously (23). Rabbit polyclonal antibodies to caspase-3 (1:5,000) or caspase-7 (1:750) were from Transduction Laboratories or NOVUS Biologicals, respectively. Antibodies to GFP (1:1,000; Clontech, Palo Alto, CA), β-actin (1:10,000; Sigma, St. Louis, MO), cytochrome c (1:1,000; PharMingen, San Diego, CA), PARP (1:1,000; PharMingen), or Cox-4 (1:5,000; Clontech) were also used.
Replication-deficient adenoviruses encoding wild-type survivin (pAd-Survivin), GFP (pAd-GFP), or survivin Thr34→Ala mutant (pAd-T34A) were generated using the pAd-Easy system, as described previously (24). Viruses were propagated in HEK293 cells and purified by CsCl banding. With this protocol, no replication-competent adenovirus particles are generated (24). For adenoviral transduction, monolayers of proliferating endothelial cells or genetically modified MEFs were incubated with pAd-GFP, pAd-Survivin, or pAd-T34A at a multiplicity of infection of 50 in M199 medium plus 20% FCS for 10 h, washed with PBS (pH 7.4), and placed in fresh medium plus 20% FCS for 16 h at 37°C. Transduction efficiency (>95% of the cell population) was estimated by GFP fluorescence.
Three-dimensional Capillary Formation.
HUVEC monolayers (80% confluence) in 6-well clusters were incubated with pAd-GFP or pAd-Survivin at a multiplicity of infection of 50 for 8 h at 37°C followed by an additional 24-h incubation in complete medium at 37°C. Rat tail type I collagen (3 mg/ml; Becton Dickinson, Bedford, MA) in 0.1 volume of 10× DMEM was neutralized with sterile 1 m NaOH and kept on ice. HUVECs were added to the collagen suspension to a final concentration of 1 × 106 cells/ml collagen. Ten drops (0.1 ml each) of the HUVEC-collagen mixture were added to a 35-mm plate. Plates were placed in a humidified incubator at 37°C, and the HUVEC-collagen mixtures were allowed to gel for 10 min, after which 3 ml of M199 medium containing 20% FCS, 50 μg/ml endothelial cell growth supplement, 100 μg/ml heparin, 100 μg/ml penicillin, and 100 μg/ml streptomycin were added to each plate. Cells were allowed to form capillary-like vascular tubes over a 7-day culture in the presence of 16 nm PMA (Sigma). Addition of PMA results in a potent morphogenic effect promoting the formation of three-dimensional vascular tube-like structures, which closely mimic capillary formation in vivo via a protein kinase C-, mitogen-activated protein kinase-, and phosphatidylinositol 3′-kinase-dependent pathway (25). As determined in previous studies, PMA withdrawal under these conditions results in rapid regression of capillary structures and HUVEC apoptosis in vitro. In other experiments, capillary tube formation was assessed in HUVEC or DMVEC cultures in Matrigel (200 μl) in the presence of VEGF (50 ng/ml) and assessed by phase-contrast microscopy during a 48–72-h culture at 37°C. To quantify tube formation, cells were washed three times in PBS (pH 7.4) and snap frozen in OCT embedding compound. Cryostat sections of the gels (6 μm) were placed on poly-l-lysine-coated glass slides, fixed with acetone for 10 min at −20°C, air-dried, and stained with H&E. Sections were examined by phase-contrast microscopy, and the total vessel area (in 3 fields/slide in each experiment) was quantified using the NIH (Bethesda, MD) Image program.
Cell Viability and Apoptosis.
HUVECs transduced with pAd-GFP, pAd-Survivin, or pAd-T34A were incubated in 0% FCS for up to 72 h or, alternatively, treated with 25 μm C6 ceramide or the combination of TNF-α (10 ng/ml; Endogen, Woburn, MA) plus CHX (10 μg/ml; Sigma), for 6–12 h at 37°C (14). Cultures under the various conditions were analyzed for nuclear morphology of apoptosis after fixation in 4% paraformaldehyde containing 0.25% Triton X-100 for 10 min at 22°C, and staining of cell nuclei was performed with 6.5 μg/ml DAPI (Sigma) in 16% polyvinyl alcohol (Air Products and Chemicals, Allentown, PA) and 40% glycerol. The percentage of apoptotic cells was calculated by direct counting of nuclei with apoptotic morphology (condensed chromatin, fragmented DNA) in five independent high-power fields (×400; each field contained about 150 cells) in three independent experiments, using a Zeiss fluorescence microscope (26). For survivin targeting, exponentially growing HUVECs or DMVECs were transduced with pAd-GFP, pAd-survivin, or pAd-T34A; harvested after a 96-h culture at 37°C; and analyzed for DNA content by propidium iodide staining and flow cytometry. The hypodiploid (i.e., apoptotic) cell fraction was quantified using CELL Quest software (Becton Dickinson), as described previously (27). Apoptosis in transduced wild-type or Apaf-1 or caspase-9 knockout MEFs was also assessed by DNA content analysis and flow cytometry (27). For quantification of caspase activity, virally transduced HUVECs were lysed in 0.5% Triton X-100, 10 mm KCl, 1.5 mm MgCl2, 1 mm EDTA, 1 mm DTT, 20 mm HEPES, and protease inhibitors. Protein-normalized aliquots of the various cell extracts were separated by SDS-gel electrophoresis, transferred to nylon membranes (Millipore Corp.), and immunoblotted with antibodies to caspase-3, caspase-7, survivin, GFP, or β-actin, followed by chemiluminescence and autoradiography. In parallel experiments, detergent-solubilized HUVEC extracts were assayed for caspase-3-dependent hydrolysis of the fluorogenic substrate N-acetyl-Asp-Glu-Val-Asp-aldehyde (Ac-DEVD-AMC; PharMingen, San Diego, CA), with determination of fluorescence emissions on a spectrofluorometer with excitation wavelength of 360 nm and emission of 460 nm.
Characterization of Mitochondrial-dependent Apoptosis.
Tet-regulated YUSAC-2 cells (5 × 107 cells; Ref. 20) were washed in TD buffer containing 135 mm NaCl, 5 mm KCl, and 25 mm Tris-HCl (pH 7.6) and allowed to swell for 10 min in ice-cold hypotonic CaRSB buffer [10 mm NaCl, 1.5 mm CaCl2, 10 mm Tris-HCl (pH 7.5) and protease inhibitors]. Cells were Dounce-homogenized with 60 strokes, with addition of MS buffer [210 mm mannitol, 70 mm sucrose, 5 mm EDTA, and 5 mm Tris (pH 7.6)] to stabilize mitochondria (2 ml of 2.5× per 3 ml of homogenate). After removing nuclear contaminants by centrifugation at 3,000 rpm for 15 min on ice, the supernatants were layered over a 1–2 m sucrose step gradient [in 10 mm Tris (pH 7.6), 5 mm EDTA, 2 mm DTT, and protease inhibitors] and centrifuged at 110,000 × g for 30 min at 4°C. Mitochondria were collected at the 1–1.5 M interphase by lateral suction through the tube, washed in 4 volumes of MS buffer at 15,000 rpm, and suspended in a final volume of 200 μl of MS buffer. The top layer containing a cytosolic and free protein fraction was collected and used in parallel experiments. Samples from cytosolic or mitochondrial fractions harvested at various time intervals in the presence or absence of Tet were analyzed sequentially by Western blotting with antibodies to cytochrome c, PARP, Cox-4, or β-actin followed by chemiluminescence. To monitor changes in mitochondrial transmembrane potential, Tet− YUSAC-2 cells expressing wild-type survivin or survivin(T34A) were harvested at increasing time intervals (6–72 h) at 37°C, suspended in 600 μl of distilled deionized water, and homogenized. Cell homogenates were incubated with the green fluorescence dye JC-1 (10 μg/ml; Molecular Probes Inc., Eugene, OR) for 10 min in the dark, washed in ice-cold PBS, and analyzed by flow cytometry.
Breast Cancer Xenograft Model.
Female 6–8-week-old C.B.-17 SCID/beige mice (Taconic Farms, Germantown, NY) received s.c. injection (one injection in each flank) with 2.5 × 106 exponentially growing MCF-7 cells in 250 μl of sterile PBS (pH 7.4). Tumor growth was confined to local masses and did not affect animal survival over a 4-month observation period, as described previously (24). About 5 days after injection, tumors became palpable, and groups of three animals were randomized and distributed between treatment groups (6 tumors/group). Animals were treated with pAd-GFP or pAd-T34A by intratumoral injections (109 green fluorescence units in 50 μl distributed between 3 sites/tumor mass) on 2 consecutive days followed by 5 days of interval (24). Tumor volume was monitored by measuring tumors in the three dimensions with a caliper every other day for up to 14 days after tumors became palpable (total of two pAd injections). Adenoviral gene transfer in situ was assessed for GFP expression by fluorescence microscopy, as described previously (24). All experiments involving animals were approved by the institutional animal care and use committee.
Histological Assessment of Tumor-associated Angiogenesis.
Breast cancer tumors prepared as described above were harvested after 7 days of treatment, formalin-fixed, and paraffin-embedded. Five-μm tissue sections were cut, air-dried on glass slides, deparaffinized, rehydrated, and quenched with 1% hydrogen peroxide for 45 min at room temperature. For endothelial cell detection by CD31 (platelet-endothelial cell adhesion molecule-1) staining, a two-step antigen retrieval method was used with pressure cooking for 5 min in 10 mm sodium citrate buffer (pH 6.0) and tissue digestion with 0.05% trypsin at 37°C for 10 min. Tissue sections were blocked with 10% goat serum for 30 min at room temperature. A previously described rabbit polyclonal antibody to mouse CD31 [Sleet4; a kind gift of Dr. J. A. Madri, Yale School of Medicine, New Haven, CT (28, 29)] was applied for 14 h at 4°C at a 1:500 dilution in 0.5% goat serum. Naïve normal rabbit antiserum was used at an equivalent dilution and produced no staining (data not shown). Binding of the primary antibodies was detected using a biotinylated goat antirabbit secondary antibody, followed by the avidin-biotin-peroxidase system (NovaRed Peroxidase Substrate Kit; Vector Laboratories, Burlingame, CA) with 3-amino-9-ethyl carbazole (Vector Laboratories) as the chromophore. Total CD31 positivity was determined by measuring the percentage of positively stained areas within each field using a color-sensitive software imaging program based on MATLAB script (The MathWorks Inc., Natick, MA) and developed by B. Bourke in the laboratory of Prof. A. Sinusas (Yale School of Medicine, New Haven, CT). Quantitative image analysis was performed on 7 randomly selected fields/section (magnification, ×200), for each of 2 sections/tumor, for each of 4 tumors/treatment group (n = 56). In vivo apoptosis was determined by TUNEL (Zymed, San Francisco, CA) as described previously (24), except that antigen retrieval was carried out by tissue digestion in 0.05% trypsin at 37°C for 30 min.
The kinetics of tumor growth under the various conditions tested was analyzed by the unpaired two-tailed t test on a GraphPad Prism software package for Windows. A P of 0.05 was considered statistically significant.
Antiapoptotic Effect of Survivin in Endothelial Cells.
When deprived of nutrients, HUVECs transduced with pAd-GFP exhibited time-dependent loss of cell viability, with chromatin condensation and nuclear fragmentation. By 72 h of growth factor deprivation, ∼60% of the HUVEC population expressing pAd-GFP showed morphological signs of apoptosis (Fig. 1,A). In contrast, HUVECs transduced with pAd-Survivin were resistant to growth factor deprivation-induced apoptosis throughout a 72-h culture (Fig. 1,A). A potential effect of survivin on caspase activation and catalytic activity was investigated. Incubation of pAd-GFP-transduced HUVECs in low serum resulted in time-dependent generation of active caspase-3 and -7, by Western blotting with activation-specific antibodies (Fig. 1,B). This was associated with generation of caspase-3 catalytic activity by hydrolysis of the chromogenic substrate DEVD-AMC (Fig. 1,C). In contrast, HUVECs transduced with pAd-Survivin did not exhibit caspase activation by Western blotting (Fig. 1,B), and caspase-3 catalytic activity was largely suppressed in these cells (Fig. 1,C). In addition to growth factor deprivation, transduction of HUVECs with pAd-Survivin nearly completely suppressed ceramide-induced apoptosis, whereas cell death induced by the combination of TNF-α plus CHX was only minimally reduced (Fig. 1 D).
Survivin Promotes Capillary Formation in Vitro.
Cultivation of HUVECs or DMVECs in Matrigel pellets or rat collagen gels resulted in the formation of capillary-like tube structures by 48 h (Fig. 2,A) that exhibited extensive elongation and branching after a 72-h culture at 37°C (Fig. 2, B and C). In the absence of PMA as a stabilizing agent, HUVECs transduced with pAd-GFP exhibited only a negligible degree of capillary formation after a 7-day culture at 37°C. Treatment of pAd-GFP-treated HUVECs with PMA increased capillary length by ∼3-fold and stabilized tube formation by phase-contrast microscopy and H&E staining (Fig. 2, D and E). Transduction with pAd-Survivin prolonged capillary tube viability in the presence of PMA and supported the persistence of viable capillaries even in the absence of PMA throughout a 7-day culture at 37°C (Fig. 2, D and E).
Survivin Targeting Induces Endothelial Cell Apoptosis.
Transduction of exponentially proliferating HUVECs or DMVECs (expressing endogenous survivin) with a phosphorylation-defective survivin mutant pAd-T34A (20) resulted in induction of apoptosis in both cell types in the absence of other cell death stimuli by DNA content analysis and flow cytometry (Fig. 3,A). Consistent with genuine induction of apoptosis, survivin targeting by pAd-T34A resulted in de novo generation of active caspase-3 by Western blotting (Fig. 3,B). In contrast, HUVECs or DMVECs transduced with pAd-GFP or pAd-Survivin did not exhibit induction of apoptosis or caspase-3 cleavage (Fig. 3, A and B).
Survivin Targeting Initiates Mitochondrial-dependent Apoptosis.
To map the position of survivin in cell death pathway(s) (13), we used stably transfected YUSAC-2 melanoma cells in which expression of survivin(T34A) is conditionally induced upon withdrawal of Tet (Tet-off system; Ref. 20). Conditional expression of survivin(T34A) in YUSAC-2 cells resulted in rapid and progressive accumulation of mitochondrial cytochrome c in the cytosol by Western blotting (Fig. 4,A). In contrast, Tet-regulated expression of wild-type survivin in YUSAC-2 cells did not result in cytochrome c increase in the cytoplasm (Fig. 4,A). Conditional expression of survivin(T34A) was also associated with time-dependent loss of mitochondrial transmembrane potential, beginning at 6 h after Tet removal and steadily increasing over a 72-h time interval (Fig. 4,B). This coincided with time-dependent induction of apoptosis in YUSAC-2 cells by DNA content analysis (Fig. 4,B, inset), in agreement with previous observations (20). In contrast, Tet-regulated expression of wild-type survivin did not result in significant changes in mitochondrial transmembrane potential (Fig. 4,B) and did not affect YUSAC-2 cell viability throughout a 72-h culture (Fig. 4,B, inset). Consistent with genuine induction of apoptosis, Tet-regulated expression of survivin(T34A) resulted in time-dependent cleavage of Mr ∼115,000 proform PARP to an apoptotic Mr ∼85,000 fragment, whereas no PARP cleavage was observed in the presence of Tet (Fig. 4 C).
Genetic Requirements of Apoptosis Induced by Survivin Targeting.
To map more precisely the genetic requirements of apoptosis induced by survivin targeting, we used MEFs isolated from mouse embryos deficient in the initiating components of mitochondrial-dependent apoptosis, Apaf-1 and caspase-9 (13). Transduction of wild-type MEFs expressing endogenous survivin (data not shown) with pAd-T34A resulted in time-dependent induction of apoptosis by DNA content analysis and flow cytometry (Fig. 5, A and B). In contrast, apoptosis induced by pAd-T34A was completely suppressed in MEFs isolated from Apaf-1 or caspase-9 knockout embryos (Fig. 5, A and B). In control experiments, transduction with pAd-GFP did not reduce cell viability of wild-type, Apaf-1−/−, or caspase 9−/− MEFs (Fig. 5, A and B).
Cytoprotective Effect of Survivin on the Tumor Microenvironment.
Intratumoral injection of pAd-T34A in human MCF-7 breast cancer xenografts resulted in GFP expression in situ (Fig. 6,A, inset) and inhibition of tumor growth (Fig. 6,A), in agreement with previous observations (24). This was associated with massive induction of tumor cell apoptosis in vivo by internucleosomal DNA fragmentation (Fig. 6,B). In contrast, MCF-7 xenografts injected with pAd-GFP exhibited exponential tumor growth (Fig. 6,A) and no significant decrease in tumor cell viability by TUNEL staining (Fig. 6,B). Immunohistochemical analysis of endothelial cell-derived CD31 in MCF-7 xenografts injected with pAd-GFP revealed an extensive network of viable blood vessels, consistent with tumor-associated angiogenesis (Fig. 7,A). In contrast, intratumor injection of pAd-T34A resulted in a significant reduction in the number of viable blood vessels by CD31 staining (Fig. 7,A). As determined by quantitative morphometric analysis of CD31-stained areas, injection of pAd-T34A resulted in a ∼3-fold suppression of blood vessel density in MCF-7 xenografts, as compared with pAd-GFP (Fig. 7,B). Reduction in blood vessel density in pAd-T34A-injected tumors was associated with appearance of apoptosis in endothelial cells lining tumor-derived blood vessels by internucleosomal DNA fragmentation in vivo (Fig. 7 C).
In this study, we have positioned the survivin pathway as a novel regulatory mechanism of mitochondrial-dependent apoptosis. Secondly, we have shown that targeting survivin exerts anticancer activity by combining enhanced tumor cell apoptosis with suppression of tumor-associated angiogenesis in vivo.
Despite the considerable interest in survivin for its bifunctional role in cell viability and regulation of mitosis (10) and the dramatic exploitation of this pathway in human tumors (12), critical aspects of the survivin pathway have remained elusive. In particular, how survivin couples to the cell death machinery has not been conclusively elucidated. Earlier claims that this may involve suppression of caspase catalytic activity (30), similar to other antiapoptotic IAP proteins (9), were disputed on both functional (31) and structural grounds (32). More recent arguments favored an indirect model of survivin-mediated cell viability, in which interference with survivin expression/function resulted in cell death merely as a consequence of catastrophic cell division defects (33, 34), potentially involving mistargeting of Aurora B kinase (35).
Here, we obtained clear evidence to unambiguously rule out these speculations (34) and, for the first time, to firmly position the survivin pathway as a novel upstream regulator of mitochondrial-dependent apoptosis (13). Interference with survivin function using a phosphorylation-defective survivin Thr34→Ala mutant (20) caused all of the earliest cellular and biochemical events of mitochondrial-dependent apoptosis (13), including release of cytochrome c, loss of mitochondrial transmembrane potential, and cleavage of caspase substrates. This cell death response was abolished in MEFs deficient in the upstream apoptosome components, Apaf-1 and caspase-9 (13). This model for survivin function is consistent with the ability of survivin to reduce the generation of active caspase-3 and -7 (Fig. 1), rather than suppressing the activity of the mature enzymes, and with the previously reported physical interaction between survivin and the upstream mitochondrial initiator, caspase-9 (19). Several possibilities for how survivin could influence the upstream initiation of mitochondrial-dependent apoptosis could be envisioned, including its recently reported association with Smac/DIABLO (36), a mitochondrially released protein, which relieves the inhibitory function of IAP on caspase-9 activation (32). Clearly, the present data do not support the preliminary claims of Chen et al. (35) suggesting that survivin function could be recapitulated by its potential interaction with Aurora B kinase. Although critical for cell division, Aurora B is not believed to couple to the mitochondrial cell death machinery, and the phenotype of reduced spindle microtubule density and apoptosis induced by microinjection of antibodies to survivin (37) is quite distinct from the defect of microtubule bundling and astral microtubule extension observed after interference with Aurora B function (38, 39).
The Thr34→Ala mutation used here to map the survivin pathway abolishes a phosphorylation site for the main mitotic kinase p34cdc2-cyclin B1 (19) and has been used previously to interfere with the function of endogenous survivin, resulting in apoptosis (19) and anticancer activity in vitro and in vivo (24). Recently, Temme et al. (40) used an overexpression approach with a large, tetrameric DsRed-survivin fusion protein to investigate the subcellular distribution and function of wild-type survivin and survivin(T34A). In that study, expression of DsRed-survivin(T34A) caused mitotic defects, inhibited cell proliferation, and induced apoptosis (40), similar to the findings presented here. Surprisingly, however, Temme et al. (40) also reported that survivin(T34A) inhibited apoptosis in HeLa cells. The contradictory results of Temme et al. (40) may reflect their highly artificial overexpression system, which is unlikely to recapitulate the dynamic intracellular trafficking and association with multiple protein partners of endogenous survivin (23). Also contrary to the claims of Temme et al. (40) that mislocalization of survivin(T34A) may cause the observed cellular phenotype, published data have demonstrated that this survivin mutant exhibits accelerated degradation in vivo (41), suggesting that its dominant negative mode of action may involve dimerization with endogenous survivin and premature destruction of the heterocomplex.
Consistent with previous observations (14, 15, 17), expression of survivin in endothelial cells representative of different vascular beds resulted in a broad cytoprotective mechanism counteracting apoptosis, reducing the generation of active caspases, and preserving cellular survival. This translated in a productive proangiogenic response with stabilization of three-dimensional capillary networks in vitro. It was recently proposed that expression of survivin in the endothelium could have profound repercussions for tumor growth, reducing the effectiveness of metronomic, antiangiogenesis chemotherapy (18). The data presented here fit well with that model and demonstrate that survivin expression during angiogenesis may provide a pivotal advantage factor to maintain a florid blood supply during tumor growth. Previous studies have demonstrated that this may involve up-regulated survivin expression during the proliferative phase of angiogenesis as a transcriptional target of VEGF (14) as well as during the nonproliferative, remodeling of blood vessels contributed, among others, by angiopoietin-1 (16, 17). Although increased survivin expression in the transformed tumor cell population or angiogenic endothelium may involve distinct signaling pathways that may include loss of p53, activation of phosphatidylinositol 3′-kinase, or phosphorylation of signal transducers and activators of transcription 3 (14, 15, 16, 17, 42), the data presented here demonstrate that survivin provides a broad cytoprotective mechanism for the tumor microenvironment as a whole. This may explain, in part, the unfavorable outcome associated with survivin expression in embryologically disparate tumors and their propensity to progress to an invasive phenotype (10), as characterized recently in a transgenic mouse model of stepwise skin cancer (43). On the other hand, these data provide a rational basis for targeting survivin as a novel cancer therapeutic strategy aimed at lowering a general antiapoptotic threshold in tumor cells and favoring the collapse of tumor-associated angiogenesis.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Supported by NIH Grants HL54131, CA78810, and CA90917.
The abbreviations used are: IAP, inhibitor of apoptosis; GFP, green fluorescent protein; MEF, mouse embryonic fibroblast; HUVEC, human umbilical vein endothelial cell; DMVEC, dermal microvascular endothelial cell; PMA, phorbol 12-myristate 13-acetate; TUNEL, terminal deoxynucleotidyl transferase-mediated nick end labeling; Tet, tetracycline; VEGF, vascular endothelial growth factor; TNF-α, tumor necrosis factor α; CHX, cycloheximide; DAPI, 4′,6-diamidino-2-phenylindole; PARP, poly(ADP-ribose) polymerase; DEVD-AMC, Asp-Glu-Val-Asp-aldehyde.
We thank Dr. T. Mak for generously providing Apaf-1 and caspase-9 knockout MEFs.