Down-regulation of Glutathione and Bcl-2 Synthesis in Mouse B16 Melanoma Cells Avoids Their Survival during Interaction with the Vascular Endothelium*

B16 melanoma (B16M) cells with high GSH content show high metastatic activity. However, the molecular mechanisms linking GSH to metastatic cell survival are unclear. The possible relationship between GSH and the ability of Bcl-2 to prevent cell death was studied in B16M cells with high (F10) and low (F1) metastatic potential. Analysis of a Bcl-2 family of genes revealed that B16M-F10 cells, as compared with B16M-F1 cells, overexpressed preferentially Bcl-2 (∼5.7-fold). Hepatic sinusoidal endothelium-induced B16M-F10 cytotoxicity in vitro increased from ∼19% (controls) to ∼97% in GSH-depleted B16M-F10 cells treated with an antisense Bcl-2 oligodeoxynucleotide (Bcl-2-AS). l-Buthionine (S,R)-sulfoximine-induced GSH depletion or Bcl-2-AS decreased the metastatic growth of B16M-F10 cells in the liver. However, the combination of l-buthionine (S,R)-sulfoximine and Bcl-2-AS abolished metastatic invasion. Bcl-2-overexpressing B16M-F1/Tet-Bcl-2 and B16M-F10/Tet-Bcl-2 cells, as compared with controls, showed an increase in GSH content, no change in the rate of GSH synthesis, and a decrease in GSH efflux. Thus, Bcl-2 overexpression may increase metastatic cell resistance against oxidative/nitrosative stress by inhibiting release of GSH. In addition, Bcl-2 availability regulates the mitochondrial GSH (mtGSH)-dependent opening of the permeability transition pore complex. Death in B16M-F10 cells was sharply activated at mtGSH levels below 30% of controls values. However, this critical threshold increased to ∼60% of control values in Bcl-2-AS-treated B16M-F10 cells. GSH ester-induced replenishment of mtGSH levels (even under conditions of cytosolic GSH depletion) prevented cell death. Our results indicate that survival of B16M cells with high metastatic potential can be challenged by inhibiting their GSH and Bcl-2 synthesis.

The majority of metastatic cells entering microvascular beds are killed within the first hours and do not generate colonies. This failure, termed "metastatic inefficiency" (1), is due to mechanical trauma produced by blood flow (2), the inability of cancer cells to withstand deformation (3), cytotoxicity of locally released reactive oxygen and nitrogen species (4), and the lytic action of lymphocytes and macrophages (5). The B16 melanoma (B16M) 1 is a model widely used to study metastatic spread and tissue invasion (6). The liver is a common site for metastasis development, and we recently reported that GSH (␥-glutamylcysteinyl-glycine) protects B16M-F10 cells (with high metastatic potential) against nitrosative and oxidative stress in the hepatic microvasculature (4,6). In fact, multidrug and/or radiation resistance, which are characteristic features of malignant tumors, frequently associate with high GSH content in the cancer cells (7). B16M-F10 cell resistance to the HSE-induced cytotoxicity is highly dependent on GSH and GSH peroxidase (4). However, B16M-F10 cells cultured to low density (LD), with high GSH content, were more resistant to NO and H 2 O 2 than B16M cultured to high density (with ϳ25% of the GSH content found in LD cells) (4). NO-and H 2 O 2 -dependent cytotoxicity in B16M-F10 cells attached to cytokine (tumor necrosis factor (TNF)-␣ and IFN-␥)-activated HSE was ϳ18% in LD B16M-F10 cells and ϳ78% in high density B16M-F10 cells (4). HSE-induced tumor cytotoxicity in L-buthionine (S,R)-sulfoximine (BSO; a specific GSH synthesis inhibitor)-treated LD B16M cells was similar to that found in high density B16M cells (4), which suggested a direct involvement of GSH in protecting B16M cells against HSE-induced cytotoxicity. In addition, we also showed that metastatic growth can be implemented in B16M-F1 cells (with low metastatic potential) by using GSH ester, which directly increases their GSH content (8). Nevertheless, even after BSO-induced GSH depletion, a significant amount of LD B16M-F10 cells (ϳ32%) survived during in vitro interaction with the HSE (4). This is critical, since highly resistant metastatic cell subsets may likely be responsible for the explosive metastic growth that follows tissue invasion under in vivo conditions. Therefore, although high GSH content status is an important parameter for metastasis progression in B16M cells, other factor(s) must necessarily contribute to the survival of some cell subsets with high metastatic potential.
The expression of genes known to affect apoptosis (e.g. Bcl-2, p53, Fas, nitric-oxide synthetases, etc.) may affect tumor growth and possibly metastatic inefficiency (9). Takaoka et al. (10) observed that Bcl-2 overexpression in B16M cells enhanced pulmonary metastasis. In parallel, melanoma cells resistant to Fas-mediated apoptosis were found to be more likely to metastasize (11). Furthermore, although apoptotic H-ras and v-myc-transformed metastatic fibroblasts labeled with green fluorescent protein were observed in the lungs, in vitro induced Bcl-2 overexpression in these cells conferred resistance to apoptosis 24 -48 h after inoculation (12). Thus, it is plausible that regulation of cell death mechanisms influences metastatic growth, at least in the early stages after attachment to the vascular endothelium.
The proto-oncogene Bcl-2 and its antiapoptotic homologs are mitochondrial membrane permeabilization inhibitors (13) and participate in development of chemoresistance (14), whereas expression of pro-death genes (e.g. Bax or Bak) is often reduced in cancer cells (15). The thiol redox status (controlled by GSH) is one of the physiological effectors involved in regulating the mitochondrial permeability transition pore complex (16). The importance of GSH in regulating the ability of Bcl-2 to prevent apoptosis was first detected in GT1-7 neuronal cells, where Bcl-2-induced suppression of apoptosis required GSH (17). Later, different reports indicated that GSH depletion may have therapeutic use in sensitizing Bcl-2-overexpressing cells to apoptotic cell death (18 -21). Thus, possibly, GSH-dependent cancer cell survival within the microvasculature and Bcl-2-dependent cell death regulation are closely related mechanisms that cooperate, favoring survival of highly metastatic cell subsets. In this report, we examined this possibility by studying expression of pro-death and anti-death Bcl-2 genes in B16M cell lines with different metastatic potential and GSH contents. We found that Bcl-2 is preferentially overexpressed in B16M-F10 cells as compared with the low metastatic B16M-F1 cell line. Furthermore, treatment of B16M-F10 cells with Bcl-2 antisense oligodeoxynucleotide (Bcl-2-AS), when combined with GSH depletion, promoted massive metastatic cell death within microvessels and abrogated tissue invasion.
In Vivo Microscopy-Syngenic male C57BL/6J mice (9 weeks old) from IFFA Credo (L'Arbreole, France) were fed ad libitum on a stock laboratory diet (Letica, Barcelona, Spain) and kept on a 12-h light/12-h dark cycle with the room temperature maintained at 22°C. Procedures involving animals were in compliance with the national and international laws and policies (64,65). In vivo microscopy to follow metastatic cell dynamics within the liver was performed, as previously described (6), using calcein-AM (Molecular Probes, Poortgebouw, Leiden, The Netherlands)-labeled B16M cells. The total number of calcein-AM-labeled cells per hepatic lobule was recorded in 10 different lobules per liver at 15-min intervals for a 6-h period. The microscope was an Eclipse E600FN, providing transillumination or epiillumination, equipped for video microscopy using a digital DXM 1200 camera (Nikon, Tokyo, Japan).
GSH and GSSG-GSH was measured by the glutathione S-transferase reaction and GSSG by high pressure liquid chromatography (8).
Cellular Bcl-2 protein levels were also analyzed by flow cytometry using an EPICS PROFILE II (Coulter Electronics, Hialeah, FL), tuned at 488 nm and 250 milliwatts. Cellular suspensions were diluted to ϳ250,000 cells/ml. Primary monoclonal anti-Bcl-2 from mouse followed by biotin-conjugated goat antimouse IgG and phycoerythrin-labeled streptavidin (Sigma) were used. Samples of 10,000 cells were measured for forward angle light and integrated side scatters, and fluorescent light emissions (expressed as logarithms of the integrated fluorescent light). Electronic noise and cellular debris were eliminated from the calculations by ignoring in all measurements the values of the forward angle light scatter/integrated side scatter biparametric plot between 0 and 40 (resolution of 256 channels). Bcl-2 protein levels were expressed as arbitrary fluorescence units (⌬FL1). Cell viability in these experiments was determined with propidium iodide (final concentration 10 M; Molecular Probes).
Bcl-xL Analysis-Cellular Bcl-xL protein levels were analyzed by flow cytometry, as described above for Bcl-2, but using monoclonal anti-Bcl-xL antibodies from mouse (Sigma).
B16M-HSE Adhesion and Cytotoxicity Assays-Hepatic sinusoidal cells were isolated and identified as previously described (4). To quantify damage to B16M cells during their in vitro adherence to the HSE, B16M cells were resuspended in culture medium containing 2 g of calcein-AM/ml and incubated for 15 min at 37°C (calcein-AM up to 10 g/ml did not affect B16M or HSE cell viability; not shown). After washing, the cells were resuspended in HEPES-buffered DMEM without phenol red at a concentration of 2 ϫ 10 6 cells/ml. Labeled B16M cells were added to HSE (0.1 ml/well) and also to plastic or collagencoated (0.1 g/ml collagen/well). Fluorescence of the added number of cells in each well was determined using a Fluoroskan Ascent FL (Labsystems, Manchester, UK). The number of adhering tumor cells was quantified by arbitrary fluorescence units based on the percentage of the initial number of B16M cells added to the HSE culture (4). The plates were then incubated at 37°C, and 20 min later, wells were washed three times with fresh medium and read for fluorescence. B16M cells damage was assessed by plate-scanning fluorimetry of intracellular calcein fluorescence. Plates were red using a 485/22-nm excitation filter and a 530/25-nm emission filter. Relative tumor cytotoxicity was determined as follows: (fluorescence from B16M cells adhered to cellu-lar substrate)/(fluorescence from B16M cells adhered to collage-coated substrate) ϫ 100. An identical approach was used to evaluate possible damage to endothelial cells during their in vitro adherence to B16M cells, and we found that the percentage of HSE viability was 99 -100% in all cases. Cell integrity was assessed by trypan blue exclusion and by measuring lactate dehydrogenase activity (27) released to the extracellular medium (6).
Measurement of H 2 O 2 , Nitrite, and Nitrate-The assay of H 2 O 2 production was based on the H 2 O 2 -dependent oxidation of the homovanillic acid (3-methoxy-4-hydroxyphenylacetic acid) to a highly fluorescent dimer (2,2Ј-dihydroxydiphenyl-5,5Ј-diacetic acid), which is mediated by horseradish peroxidase (4). Total NO x (NO 2 Ϫ plus NO 3 Ϫ ) determinations were made by monitoring NO evolution (chemiluminescence detection) from a measured sample placed into a boiling VCl 3 /HCl solution (which reduces both NO 2 Ϫ and NO 3 Ϫ to NO) (4). Experimental Metastases-Hepatic metastases were produced by intravenous injection (portal vein) into anesthetized mice (Nembutal, 50 mg/kg intraperitoneally) of 4 ϫ 10 5 viable B16 cells suspended in 0.2 ml of DMEM. Mice were cervically dislocated 10 days after B16M inoculation. The livers were fixed with 10% formaldehyde in PBS (pH 7.4) for 24 h at 22°C and then paraffin-embedded. Metastasis density (mean number of foci/100 mm 3 of liver detected in 15 10 ϫ 10-mm 2 sections/ liver) and metastasis volume (mean percentage of liver volume occupied by metastasis) were determined as previously described (28).
Compartmentation of B16M Cells-Cultured B16M cells were harvested by exposure to 0.02% EDTA (5 min at 37°C) and then washed twice and resuspended in ice-cold Krebs-Henseleit bicarbonate medium (pH 7.4). Cytosolic and mitochondrial compartments were separated as recently described (29).
Western Blots-Cultured cells were harvested, as indicated above, and then washed twice in ice-cold Krebs-Henseleit bicarbonate medium (pH 7.4). Whole cell extracts were made by freeze-thaw cycles in a buffer containing 150 mM NaCl, 1 mM EDTA, 10 mM Tris-HCl, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml aprotinin, and 1 g/ml pepstatin, pH 7.4. Fifty g of protein (as determined by the Bradford assay) were boiled with Laemmli buffer and resolved in 12.5% SDS-polyacrylamide gel electrophoresis. Proteins were transferred to a nitrocellulose membrane and subjected to Western blotting with specific anti-Fas antibodies (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Blots were developed using horseradish peroxidase-conjugated secondary antibody and enhanced chemiluminescence (ECL system; Amersham Biosciences).
Cell Death Analysis-Apoptotic and necrotic cell death were distinguished by using fluorescence microscopy (23). For this purpose, isolated cells were incubated with Hoescht 33342 (10 M; which stains all nuclei) and propidium iodide (10 M; which stains nuclei of cells with a disrupted plasma membrane) for 3 min and analyzed using a Diaphot 300 fluorescence microscope (Nikon, Tokyo, Japan) with excitation at 360 nm. Nuclei of viable, necrotic, and apoptotic cells were observed as blue round nuclei, pink round nuclei, and fragmented blue or pink nuclei, respectively. About 1,500 cells were counted each time. DNA strand breaks in apoptotic cells were assayed by using a direct terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling assay (Roche Applied Science) and fluorescence microscopy following the manufacturer's methodology.
Quantitative Determination of the Mitochondrial Membrane Potential-Measurements of the mitochondrial membrane potential were performed by the uptake of the radiolabeled lipophilic cation methyltriphenylphosphonium (TPMP), which enables small changes in potential to be determined (31). Briefly, B16M cells (2 ϫ 10 6 ) were incubated at 37°C, for 60 min, in 1 ml of DMEM supplemented as mentioned above but including 1 M TPMP, 250 nCi of [ 3 H]TPMP (Amersham Biosciences), and 1 M sodium tetraphenylboron. After incubation, the cells were pelleted by centrifugation (1,000 ϫ g for 5 min), 100 l of supernatant was removed, the pellet was resuspended in 100 l of 10% Triton X-100, and the radioactivity in the supernatant and pellet was measured using an LKB Wallace 8100 LSC liquid scintillation counter with quench corrections. Nonspecific TPMP binding was corrected as previously described (31). Energization-dependent TPMP uptake was expressed as an accumulation ratio in units of ((TPMP/mg of protein)/ (TPMP/l of supernatant)) (32).
Cytochrome c Release-Tumor cells were washed twice with phosphate buffered saline, and the pellet was suspended in ice-cold homogenization buffer (2 ϫ 10 6 cells/ml of buffer; 20 mM HEPES, pH 7.5, 250 mM sucrose, 1 mM MgCl 2 , 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, and 10 g/ml each leupeptin, aprotinin, and pepstatin A). The cells were homogenized with a Dounce homogenizer. After centrifugation at 2,500 ϫ g for 5 min at 4°C, the supernatants were centrifuged at 100,000 ϫ g for 30 min at 4°C. The resulting supernatant was used as the soluble cytosolic fraction. Proteins were quantified by the Bradford method (33), separated by SDS-PAGE, transferred to nitrocellulose membranes, and probed with anti-cytochrome c (Pharmingen, San Diego, CA). Bands were quantitated using a Bio-Rad GS-670 imaging densitometer.
Caspase 3 Assay-This activity was measured by using a highly sensitive colorimetric substrate, N-acetyl-Asp-Glu-Val-Asp p-nitroanilide following the manufacturer's instructions (Calbiochem). Briefly, cells were lysed in lysis buffer (50 mM HEPES (pH 7.4), 100 mM NaCl, 0.1% (v/v) CHAPS, 1 mM dithiothreitol, and 0.1 mM EDTA) on ice for 10 min, then centrifuged at 10,000 ϫ g for 10 min at 4°C. Equal volumes of the supernatants were added to equal volumes of assay buffer (50 mM HEPES (pH 7.4), 100 mM NaCl, 0.1% (v/v) CHAPS, 10 mM dithiothreitol, 0.1 mM EDTA, and 10% glycerol) and incubated at 37°C for 10 min. Then freshly prepared acetyl-Asp-Glu-Val-Asp p-nitroanilide (200 M) was added to the mixture, and A 405 was monitored every 20 min for 3 h at room temperature. Control cultures without cell lysates were used as controls. Enzyme activity was calculated, using the manufacturer's formulas, as pmol/min.
Expression of Results and Statistical Significance-Data were analyzed by one-way ANOVA or unpaired t tests where appropriate (SPSS 9.0 software for Windows; SPSS Inc., Chicago, IL). The homogeneity of the variances was analyzed by the Levene test. The null hypothesis was accepted for all of the values of the tests in which the F value was nonsignificant at p Ͼ 0.05. The data for which the F value was significant were examined by Tukey's test at p Ͻ 0.05.

Effect of GSH Depletion on B16M Cell Arrest and Viability in
the Liver Sinusoids-Calcein-labeled B16M cells, which present a green fluorescent cytoplasm, were arrested in the liver sinusoids within a few minutes after intraportal inoculation (Table I) (6). The number of arrested B16M-F1 or -F10 cells was similar (Table I) and constant along the time (no significant differences were found within the 15-240 min postinjection) (not shown). As previously reported (6), the number of arrested cells was not altered by depleting tumor cell GSH or by treating mice with lipopolysaccharide (LPS) (to preactivate the endothelium) before B16M inoculation (Table I). In untreated mice (Table I), ϳ25 and 93% of arrested, non-BSO-pretreated B16M-F1 and B16M-F10 cells, respectively, appeared as round bright fluorescent cells of well delineated profile (nondamaged "intact" cells, since no fluorescence diffusion from their cytoplasm to their neighboring tissue was observed). The rest of the cells appeared as irregularly shaped fluorescent cells with a spread of diffuse fluorescence surrounding them, which stained the contiguous hepatic tissue within the initial 15 min postarrest, and were considered as damaged with cytoplasmic leakage. BSO-induced GSH depletion decreased the percentage of arrested intact B16M-F1 and B16M-F10 cells (Table I). However, we found an interesting difference between both variant cell lines. Although non-BSO-treated B16M-F1 and BSO-treated B16M-F10 cells had similar GSH values (see the caption to Table I), the percentage of intact arrested cells was very different (ϳ25% F1 versus 64% F10 cells) (Table I). LPS, which preactivates the HSE (8), decreased the percentage of intact arrested B16M-F1 and B16M-F10 cells. However, in LPS-treated animals, the difference between non-BSO-treated B16M-F1 and BSO-treated B16M-F10 was still evident (Table I). These results indicated that GSH-independent factors should contribute to the survival of highly metastatic cell subsets.
During interaction between endothelial and cancer cells, NO and H 2 O 2 released by the endothelium have been postulated as the main inducers of metastatic cell cytotoxicity (4). Nevertheless, it is also possible that one or several apoptotic inducers contribute to the rapid death mechanism observed within the microvasculature. The TNF, Fas, and TNF-related apoptosisinducing ligand (TRAIL) receptors are highly specific mediators of apoptotic signaling (e.g. Ref. 34). Fas-and/or TNFmediated apoptosis can be modulated by the intracellular GSH content (e.g. Ref. 35). Thus, it is plausible that, during metastatic cell interaction with the vascular endothelium, rapid apoptotic signaling from Fas and/or TNF could also contribute to tumor cytotoxicity in vivo. Human and mouse melanoma cells express Fas ligand but practically no Fas, which may contribute to their immune privilege (36). In fact, we did not detect Fas by Western blot analysis in B16M-F1 or in B16M-F10 cells (although it was clearly detectable in mouse hepatocytes, used as positive controls; not shown). On the other hand, serum TNF activity remained Ͻ50 units/ml in control or in LPS-treated (0.5 mg/kg 6 h before inoculation) B16M-F1-and B16M-F10-bearing mice 15, 60, or 240 min after intraportal inoculation. In vitro, 50 units of rmTNF-␣/ml did not induce cytostatic or cytotoxic effects in growing B16M-F10 or B16M-F1 cells (not shown). This is not surprising, since it is known that different tumor-bearing animals and cancer patients may show very low or even undetectable levels of TNF in the blood (e.g. Ref. 37). Nevertheless, the absence of detectable levels of TNF may be due to very rapid formation of a complex between TNF and its receptor (37). To evaluate this possibility, we administered intravenously 250 g of anti-mTNF-␣ antibody (controls received an identical volume of 100 l of sterile, endotoxin-free saline) 15 min before inoculation of the tumor cells (a dose capable of maintaining serum TNF activity below 50 units/ml even if 25,000 units of rmTNF-␣ are administered intravenously 5 min after injecting the antibody). In these B16M-bearing mice treated with anti-mTNF-␣ antibody, the numbers of arrested and intact B16M-F1 and -F10 cells were similar to the control values displayed in Table I (not shown). These facts rule out a significant role of TNF on metastatic cell cytotoxicity in vivo. Furthermore, as occurs with other highly metastatic tumors, B16M is a TRAIL-resistant tumor (e.g. Ref. 38). Therefore, based on these facts and on our previous data (4), we conclude that endothelium-derived NO/H 2 O 2 appear fundamental as a direct cytotoxic mechanism for metastatic cells.
To prove that GSH is directly involved in regulating metastatic cell survival, we tested the effect of GSH ester (which enters the cell and delivers free GSH (e.g. Ref. 8). As shown in Fig. 1, GSH ester treatment did not affect survival of B16M-F10 cells, which have high GSH levels in the absence of GSH ester treatment. However, in B16M-F1 cells GSH ester increased their GSH content by ϳ4-fold (reaching GSH values similar to those found in control B16M-F10 cells) and the percentage of arrested intact cells (3-5-fold) (Fig. 1). However, even with GSH ester treatment, the percentage of cell survival was different, comparing F1 and F10 cells (Fig. 1). This fact suggested again the possible involvement of other regulatory factor(s).

Expression of Bcl-2 Genes in B16M Variant Cell Lines with Different Metastatic Potential and GSH Contents-GSH and
Bcl-2 levels may be interrelated within the mechanism of cancer cell resistance against different death-inducing factors (see the Introduction). Reverse transcriptase-PCR expression analysis of a Bcl-2 family of genes revealed that B16M-F10 cells, as compared with the low metastatic F1 cell variant, overexpressed anti-death Bcl-2 genes: Bcl-2 (ϳ5.7-fold) and Bcl-xL (ϳ2.3-fold) in particular (Table II). This is in agreement with previous data showing that overexpression of Bcl-2 or Bcl-xL is  Table I; not shown) and percentage of intact cells were calculated at 60 min postinjection. LPS was administrated as indicated in the legend to Table I. Before inoculation into mice, B16M cells were cultured for 24 h in the absence or in the presence of GSH ester (0.5 mM; added 1 h before harvesting the cells). GSH contents before inoculation were as follows: 11 Ϯ 3 and 39 Ϯ 7 nmol GSH/10 6 B16M-F1 cells cultured, respectively, in the absence or presence of GSH ester; 35 Ϯ 5 and 47 Ϯ 6 nmol GSH/10 6 B16M-F10 cells cultured, respectively, in the absence or presence of GSH ester. Data are means Ϯ S.D. for five or six different experiments. The significant test (Student's unpaired t test) refers, for all groups, to the comparison in the absence or in the presence of GSH ester (*, p Ͻ 0.01) and also to the difference between F1 and F10 cells (ϩ, p Ͻ 0.01).

TABLE I
In vivo time course of B16M-F1 and -F10 cell arrest and viability within the hepatic microvasculature Number of arrested cells and percentage of intact cells displayed were calculated at 60 min postinjection (no significant differences were found when measurements were performed at 15, 30, 120, 180, or 240 min postinjection, not shown). LPS-treated mice received a single intravenous injection of 0.5 mg/kg 6 h before B16M inoculation (6). Before inoculation into mice, B16M cells were cultured for 24 h in the absence or in the presence of BSO (0.2 mM, added 4 h after plating). GSH contents before inoculation were 38 Ϯ 4 and 10 Ϯ 3 nmol of GSH/10 6 B16M-F10 cells cultured, respectively, in the absence or presence of BSO; 13 Ϯ 2 and 5 Ϯ 1 nmol of GSH/10 6 B16M-F1 cells cultured, respectively, in the absence or presence of BSO. Data are means Ϯ S.D. for five or six different experiments. The significant test (Student's unpaired t test) refers, for all groups, to the comparison in the absence or in the presence of BSO (*, p Ͻ 0.01) and also to the difference between LPS-treated mice and untreated mice ( ϩ , p Ͻ 0.01). common in many types of malignant tumors, including melanoma, and is believed to increase resistance to chemotherapy (39).

GSH Depletion and Antisense Bcl-2 Oligodeoxynucleotides Increase Endothelium-induced Cytotoxicity in B16M Cells with High Metastatic Potential-The involvement of anti-death
Bcl-2 gene overexpression in the mechanism of metastatic cell survival was first examined in vitro. As shown in Table III, HSE-induced cytotoxicity on GSH-depleted B16M-F10 cells is high (ϳ74%). However, when BSO-induced GSH depletion was combined with Bcl-2-AS, metastatic cell death during interaction with the HSE rose to ϳ97% (Table III). Bcl-xL-AS did not increase the percentage of death in GSH-depleted B16M-F10 cells (Table III).

GSH and Bcl-2 Levels Regulate Survival and Growth of B16M Cells with High Metastatic Potential in Vivo-
The relevance of our in vitro observations was directly tested by measuring metastatic activity of B16M-F10 cells with different GSH and Bcl-2 contents. As shown in Table IV, BSO-induced GSH depletion or treatment with Bcl-2-AS, before inoculation, decreased the liver volume occupied by B16M-F10 cells from ϳ 26% (control) to 7 or 18%, respectively (Table IV). However, the combination of BSO and Bcl-2-AS completely abolished B16M-F10 metastatic growth (Table IV). These results indicate that GSH and Bcl-2 are critical molecular targets that regulate survival and growth of highly metastatic B16M cells.
Regulation of GSH Levels by Bcl-2 in B16M Cells-Bcl-2 overexpression in B16M-F1 cells was associated with an increase in their intracellular GSH levels (Table V). This fact suggests, in agreement with previous reports (40), a positive correlation between Bcl-2 and GSH and perhaps a role of GSH in regulating the ability of Bcl-2 to suppress apoptosis. However, treatment of B16M-F10 cells with Bcl-2-AS, which caused a decrease in Bcl-2 levels to less than 10% of controls, did not alter significantly their GSH content, although it significantly decreased the percentage of arrested intact cells (Table IV). We further investigated this apparent paradox by measuring GSH homeostasis and its relationship with Bcl-2 levels in B16M variant cell lines. As shown in Table V, Bcl-2 overexpression associated with an increase in GSH content in both B16M-F1 and -F10 cells, whereas glutathione disulfide (GSSG) levels decreased. The rates of GSH synthesis in control and Bcl-2overexpressing B16M cells were similar (Table V). However, GSH efflux was significantly reduced in Bcl-2-overexpressing B16M cells (Table V).
The decrease in GSSG correlated with lower rates of H 2 O 2 generation (Table V), which is in agreement with previous observations showing that Bcl-2 overexpression decreases generation of reactive oxygen species (17).
To investigate whether Bcl-2 is directly involved in inhibit-ing GSH efflux from intact cells or if the mechanism depends on other(s) protein(s) or peptide(s) (e.g. nuclear GSH, via a Bcl-2dependent mechanism, could regulate redox-sensitive transcription factors (41)), we measured GSH release from B16M-F10/Tet-Bcl-2 cells loaded by electroporation (Bio-Rad system; 1.0 kV/cm with a time constant of 50 ms) (29) with monoclonal anti-Bcl-2 antibodies from mice (Sigma) (see e.g. Ref. 42 for technical details). Control rates of GSH efflux from B16M-F10 or from B16M-F10/Tet-Bcl-2 cells were similar to those reported in Table V (not shown). However, GSH release increased to 2.4 Ϯ 0.2 nmol of GSH/10 6 cells ϫ h (n ϭ 6) (measured for 20 min starting 5 min after electroporation) in B16M-F10/Tet-Bcl-2 cells loaded with anti-Bcl-2 antibodies (no effect on GSH release was observed when B16M cells were loaded with a nonspecific IgG; not shown). These results suggested that Bcl-2 directly inhibits the transport system(s) and that, in consequence, Bcl-2 overexpression may increase metastatic cell resistance against oxidative/nitrosative stress by preserving intracellular GSH. Further work will be necessary to elucidate the molecular mechanisms channeling GSH release from B16M cells, among which, for example, an L-methionine-sensitive system (43) and the multidrug resistance-associated proteins (44) are potential candidates.

Mitochondrial GSH Depletion and the Molecular Activation of Cell Death in B16M Cells with High Metastatic Potential-
Previous studies in Ehrlich ascites tumor cells (23) showed that mitochondrial GSH depletion, which is not synthesized by mitochondria but taken up from the cytosol (see Ref. 29 and references therein), can cause a fall in mitochondrial membrane potential, opening of the permeability transition pore complex, and the release of proapoptotic molecular signals. Bcl-2 overexpression prevented the GSH depletion-associated increase in mitochondrial membrane permeability, although only when mtGSH levels remained above a critical threshold

TABLE III Effect of in vitro BSO-induced GSH depletion and treatment with Bcl-2-AS and Bcl-xL-AS on HSE-induced B16M-F10 cytotoxicity
Twenty-four-h cultured HSE cells (ϳ2.5 ϫ 10 5 cells/well) were cocultured with B16M-F10 cells (ϳ5.0 ϫ 10 5 cells/well; precultured for 24 h). Twenty min after the B16M-F10 addition to the HSE, the plates were washed as described under "Experimental Procedures." The ratio of tumor cells adhering to the HSE was ϳ1:1. TNF-␣ (100 units/ml) and IFN-␥ (50 units/ml), used as potent activators of NO and H 2 O 2 generation by the HSE (4), were added to the cocultures when all tumor cells present were attached to the HSE. NO x and H 2 O 2 accumulation in the culture medium, during the first 3 h of incubation (where tumor and endothelial cell viability was Ͼ95% in all cases; not shown), was 5.5 Ϯ 0.9 and 59 Ϯ 15 nmol/10 6 cells (similar to values previously reported) (4). Where indicated, BSO (0.2 mM) was incubated for 24 h before coculturing B16M and endothelial cells. GSH levels in controls ("None" under B16M-F10 pretreatment) were 34 Ϯ 7 nmol/10 6 cells and 12 Ϯ 6 nmol/10 6 (23). Thus, we investigated whether mtGSH depletion in Bcl-2-depleted B16M-F10 could be the mechanism involved in the high metastatic cell lethality observed during in vitro and in vivo interaction with the vascular endothelium. As shown in Table VI, NO and H 2 O 2 decreased B16M-F10 viability to ϳ50% of control values under conditions of BSO-induced GSH depletion and to ϳ2% in GSH-and Bcl-2-depleted cells. However, in both cases, cytGSH and mtGSH levels were decreased to ϳ15 and 60% of control values, respectively. To investigate the role of mtGSH levels in regulating activation of mitochondrion-based cell death mechanisms, we used diethylmaleate (DEM; a thiol-depleting ␣,␤-unsaturated carbonyl compound) (29) and TNF-␣ (which interferes with electron flow in the mitochondria (45) and increases production of reactive oxygen species (46)) to induce mtGSH depletion in B16M-F10 cells growing in vitro. As shown in Fig. 2, during DEM-and TNF-␣-induced progressive mtGSH decrease, apoptotic death in B16M-F10 cells (see the legend to Fig. 2) was sharply activated at mtGSH levels below 30% of control values. However, this critical threshold increased to ϳ60% of control values in Bcl-2-AS-treated B16M-F10 cells (Fig. 2). This result indicates that mtGSH-dependent opening of the PTPC is regulated by Bcl-2 availability. To prove that GSH is directly involved in this mechanism, we used GSH ester and found that mtGSH replenishment prevented the molecular activation of apoptosis in B16M-F10 cells treated with DEM plus TNF-␣ and Bcl-2-AS (Table VII). This effect was similar in B16M-F10 cells treated with DEM plus TNF-␣, Bcl-2-AS, and GSH ester even when only cytGSH levels where maintained low (ϳ10 -15% of control values) by using monochlorobimane as in Ref. 29 (not shown). This proves, as previously indicated (29), that the GSH pool relevant for cell survival is the mitochondrial one.

DISCUSSION
The Bcl-2 family members are characterized by the presence of one or several Bcl-2 homology domains and include pro-death and anti-death proteins (47). Bcl-2 itself is an anti-death protein, and its overexpression has been linked to cancer development, metastatic growth, and chemotherapy resistance (9 -12, 40). GSH protects highly metastatic B16M cells against nitrosative and oxidative stress within blood vessels (4,6,29) and, in addition, promotes metastatic growth (8). Nevertheless we found that different B16M cell lines, containing similar GSH levels, showed different rates of survival after in vivo interaction with the HSE (Table I). This fact suggested the involvement of additional factors favoring metastatic cell survival. Interestingly, a recent report indicated that development of resistance to GSH depletion-induced cytotoxicity in CC531 colon carcinoma cells associated with increased expression of Bcl-2 (19). Moreover, different reports showed that increased Bcl-2 levels associated with a concomitant increase in the intracellular GSH content. Thus, a possible link between Bcl-2 and GSH in blocking metastatic cell death was investigated.
Bcl-2 was preferentially overexpressed in B16M-F10 cells as compared with the low metastatic F1 cell variant (Table II). Regulation of Bcl-2 protein levels may include transcriptional and post-transcriptional control, protein translocation, and protein-protein interactions (34). However, in cancer cells, whereas some reports show evidence for post-transcriptional down-regulation of Bcl-2 (e.g. Ref. 48), others demonstrate an overproduction of the Bcl-2 protein on the basis on increased Bcl-2 mRNA levels (e.g. Ref. 49). Nevertheless, the excellent work by Schiavone et al. (50) may help to explain this apparent paradox, since the destabilizing potential of the Bcl-2 mRNA adenine-and uracil-rich element can be regulated by different mechanisms: half-life of the mRNA of Bcl-2 in Jurkat cells is prolonged by protein kinase C stimulation but shortened by C (2)-ceramide addition, strongly supporting the view that Bcl-2 mRNA stability plays a physiological role in modulating Bcl-2 levels. Regarding our experimental model, B16M-F10 cells, as compared with the low metastatic F1 cell variant, overex- Different superscript letters within a column indicate differences, p Ͻ 0.01.
TABLE V Bcl-2-induced inhibition of GSH efflux from B16M cells B16M cells were cultured for 24 h. Some flasks were used for GSH, GSSG, Bcl-2, and H 2 O 2 determination, and others were used to measure rates of GSH synthesis or efflux. H 2 O 2 production was calculated from a 6-h period of culture starting 18 h after plating. For GSH synthesis, cells were harvested, washed twice, and resuspended in ice-cold Krebs-Henseleit bicarbonate medium (pH 7.4) and incubated (5 mg dry weight/ml) (28) in 10-ml Erlenmeyer flasks (final volume 2 ml) for 60 min, at 37°C, in the presence of amino acid precursors for GSH synthesis (5 mM L-Gln, 2 mM Gly, 1 mM L-Ser, 1 mM N-acetylcysteine). Glucose (5 mM) and bovine serum albumin (2%) were always present. GSH synthesis was calculated from total GSH content at 0, 20, 40, and 60 min of incubation. GSH efflux was calculated from the contents of total glutathione (GSH ϩ GSSG) and GSSG in the culture medium at 0, 30, 60, and 120 min of incubation (starting 24 h after seeding). GSH efflux corresponded practically to GSH, since GSSG was, in all conditions, ϳ1-3% of the total glutathione found in the extracellular space (not shown). A one-way ANOVA was performed for comparison among groups. B16M   pressed the Bcl-2 gene by ϳ5.7-fold (Table II), whereas Bcl-2 protein levels were ϳ6-fold higher in B16M-F10 cells than in the F1 variant (Table V). These results indicate a very good correlation between increased gene expression and increased protein content, which, in these cells, minimizes the effect of post-transcriptional regulation steps.
In vitro HSE-induced B16M-F10 cytotoxicity was almost 100% when GSH-depleted metastatic cells were treated with Bcl-2-AS (Table III). These results appear to be in agreement with a recent report showing that GSH depletion enforces the mitochondrial permeability transition and causes cell death in HL60 cells that overexpress Bcl-2 (20). Furthermore, when BSO-and Bcl-2-AS-pretreated B16M-F10 cells where inoculated intravascularly into mice, the number of intact arrested cells on the HSE decreased by ϳ98% and the very small number of metastatic cell survivors (probably bearing molecular damages) (29) did not form detectable colonies (Table IV). Moreover, using HMB45 or S100 monoclonal antibodies (51) as markers of melanocytic tumors, no B16M-F10 cells could be detected within the microvasculature or the hepatic parenchyma 5 or 10 days after inoculation (not shown), which suggests that probably intravascular granulocytes and/or the Kupffer cells, present in the metastatic microenvironment (not shown), eliminate those few survivors.
The thiol redox state (controlled by GSH) is one of the endogenous effectors involved in regulating the PTPC, and, in consequence, thiol oxidation may be a causal factor in the mitochondrion-based mechanism that leads to cell death (16,23). Indeed, diamide, which forms disulfide bridges between sulfydryls and may overcome the ability of Bcl-2 to prevent PTPC opening, causes rapid apoptosis (52). Mitochondrial GSH is the only defense against peroxides generated from the electron transport chain (53). PTPC opening by itself causes dissipation of the mitochondrial membrane potential, loss of mtRNA, cessation of the import of cytosolic proteins, release of Ca 2ϩ and GSH from the matrix, mitochondrial uncoupling, an increase in reactive oxygen species generation, and, in consequence, GSH oxidation (16). Thus, several factors may be cause or consequence of the mitochondrial permeability transition. Our present results link mtGSH depletion with the fall of the mitochondrial membrane potential and the opening of the PTPC. We found that Bcl-2 levels influence the critical threshold of mtGSH levels that must be reached to activate release of proapoptotic molecules (below 30% of control values in B16M-F10 cells containing Bcl-2). However, in Bcl-2-AS-treated B16M-F10 cells, a small decrease in mtGSH was enough to activate the mitochondrion-based death mechanism ( Fig. 2 and Table VII). Our results prove that mtGSH depletion facilitates opening of the PTPC and that Bcl-2 modulates this mechanism in metastatic cells. Whether GSH may also regulate cell death mechanisms at other steps is unknown. The fact that Bcl-2 overexpression may cause redistribution of GSH to the nucleus (54) and the fact that transcription factors (e.g. NF-B, AP-1, and p53) are sensitive to redox changes affecting their DNA binding domains (55) suggest the involvement of GSH as a transcriptional regulator. Moreover, glutathionylation of proteases (e.g. Ref. 56) could also affect, for example, caspase activity. Finally, it is also possible that mtGSH depletion to extremely low values may cause a bioenergetic catastrophe (16) and change the mode of cell death to necrosis.
The present contribution identifies GSH and Bcl-2 as critical molecular targets to challenge survival of cancer cells with high metastatic potential. But can this approach be useful against metastatic cancers in vivo? Recently, we reported that B16M-F10 cells that survive after interaction with the HSE (invasive cells, iB16M) showed an impairment of the mitochondrial system for GSH uptake and, in addition, a decreased activity of respiratory complexes II, III, and IV, less O 2 consumption and ATP levels, higher O 2 . and H 2 O 2 production, and lower mitochondrial membrane potential (29). This is important, since mitochondria do not synthesize GSH and mtGSH depletion facilitates mitochondrial membrane permeabilization, PTPC opening, and the release of apoptosis-inducing molecular signals (Table VII). Indeed, iB16M cells with low mtGSH levels were highly susceptible to TNF-␣-induced oxidative stress and death (29). Further, to improve its efficacy, nontoxic TNF-␣ doses can be combined with other cytokines (e.g. IFN-␥ (57)), with thiol-depleting agents such as BSO or diethylmaleate (7), or with an L-glutamine-enriched diet to facilitate an L-glutamate-induced inhibition of GSH transport into tumor mitochondria (23). On the other hand, Bcl-2 antisense therapy using, for example, G3139, an 18-base phosphorothioate oligonucleotide complementary to the first six codons of the Bcl-2 mRNA, selectively and specifically inhibits Bcl-2 expression and promotes apoptosis in different human and murine cancer cell lines (58). Moreover, systemic administration of G3139 to Shionogi tumor-bearing mice led to a rapid decrease of tumor size (higher when chemotherapy was simultaneously administrated), whereas the oligonucleotide did not affect Bcl-2 expression in normal organs (25,59). G3139-induced tumor regression without dose-limiting toxicity was also observed in other tumors (e.g. lymphoma, melanoma, or gastric cancers) (58). Furthermore, synergism of the G3139 and anticancer drugs has been also shown in different tumors (60 -62). Interestingly, treating severe combined immunodeficient mice with G3139 and dacarbazine, the most effective chemotherapeutic agent for human melanoma, led to a complete ablation of the tumor in three of six animals (63). In conclusion, (a) Bcl-2 regulates metastatic cell GSH and, through this mechanism, metastatic cell capacity for survival, invasion, and growth, and (b) GSH depletion and Bcl-2 antisense therapy, which can be combined with chemotherapy and/or radiotherapy, may have therapeutic applications against metastatic tumors. VII Mitochondrial GSH depletion and the molecular activation of apoptosis in B16M-F10 cells DEM and rmTNF-␣ were added to cultured B16M-F10 cells as indicated in the legend to Fig. 2. GSH ester (0.5 mM) was added 1 h before DEM addition. Pretreatment of B16M-F10 cells with the ODN was performed as indicated under "Experimental Procedures." Measurements were performed 5 h after rmTNF-␣ addition in non-Bcl-2-AS-treated cells and 1 h after rmTNF-␣ addition in Bcl-2-AS-treated cells (see Fig. 2). Data are means Ϯ S.D. for five or six different experiments. A one-way ANOVA was performed for comparison among different groups.