Vanadate Facilitates Interferon α-mediated Apoptosis That Is Dependent on the Jak/Stat Pathway*

Type I interferon (IFN)-dependent inhibition of cell growth can occur either in the absence or presence of apoptosis. The mechanisms that determine whether or not cells undergo apoptosis after exposure to IFN-α are not clear. This study shows that a variety of cell lines that display growth inhibition but not apoptosis in response to IFN-α will undergo programmed cell death when low concentrations of the protein-tyrosine phosphatase inhibitor vanadate are added with IFN-α. In contrast, the combination of tumor necrosis factor-α with vanadate did not trigger apoptosis in these cells. Caspase-3 activity was detected only in cells exposed to IFN-α and vanadate but not to IFN-α or vanadate alone. The ability of IFN-α and vanadate to induce apoptosis did not require expression of p53 and was blocked by N-acetyl-l-cysteine. Activation of the Jak/Stat pathway and expression of IFN-inducible genes was not altered by incubation of cells with IFN-α and vanadate compared with IFN-α alone. However, mutant cells lacking Stat1, Stat2, Jak1, or Tyk2, or cells expressing kinase inactive Jak1 or Tyk2 did not undergo apoptosis in the presence of IFN-α and vanadate. These results suggest that IFN-α stimulation of Stat-dependent genes is necessary, but not sufficient, for this cytokine to induce apoptosis. Another signaling cascade that involves the activity of a protein-tyrosine phosphatase and/or the generation of reactive oxygen species may play an important role in promoting IFN-α-induced apoptosis.

Type I interferons (IFNs) 1 are a family of secreted polypeptides composed of three major subtypes (IFN-␣, IFN-␤, and IFN-) that exert many biological functions including antiviral and antigrowth activities, and modulation of immune responses (1,2). Because of its antiproliferative and antiviral properties, type I IFNs are currently used for the treatment of several forms of cancer and viral diseases. The biological effects of type I IFNs are mediated at least in part by the expression of a set of cellular genes induced upon IFNs binding to their specific receptor, which is composed of at least two subunits, IFNAR1 and IFNAR2 (3). Each receptor subunit interacts with a distinct member of the Jak family of tyrosine kinases. IF-NAR1 associates with Tyk2 whereas IFNR2 binds Jak1 (3). The binding of IFNs to their receptors causes receptor dimerization and activation of Jak kinases by transphosphorylation, which in turn tyrosine phosphorylate the receptor chains. Latent transcription factors termed Stats bind to the phosphorylated receptor chains, become tyrosine-phosphorylated, and dimerize via their Src homology 2 domains. Phosphorylated Stat dimers translocate to the nucleus and bind specific sequences found in the promoters of IFN-inducible genes (4). Stat1 and Stat2 together with the DNA binding subunit p48 form the IFN-stimulated response element (ISRE) binding transcription factor complex ISGF3 (5,6). Type I IFNs can also activate Stat1 and Stat3 homo-and heterodimers and Stat2 and Stat4 heterodimers, which bind to ␥ interferon activation sequences (GAS) like palindromic sequences (5,7).
The molecular mechanisms by which type I IFNs exert their antigrowth effects are diverse and not clearly understood, but do require signaling mediated by IFN activation of the Jak/Stat pathway. However, transmission of IFN-␣ signals in various cell lines can lead to decreased rates of proliferation in the presence or absence of apoptosis, suggesting that other signaling pathways may play a role in the final response (8,9). For example, IFN-␣-mediated apoptosis in H9 and U266 cells (a T-cell lymphoma and a myeloma cell line, respectively) occurs in the absence of cell cycle arrest. In contrast, the B-cell lymphoma, Daudi, only exhibits cell growth arrest at the G 0 /G 1 transition after IFN-␣ treatment (8 -10), while, in the leukemic T cell line Jurkat, IFN-␣ slows the rate of growth without causing cell cycle arrest or apoptosis (11). Effector molecules of the cell cycle have been reported to be regulated by type I IFNs. IFN-␣ induces the expression of the cyclin-dependent kinase inhibitors p21 Cip1 and Ink4, resulting in decreased cyclin-dependent kinase activity and inhibition of cell cycle progression (12,13). Other reports indicate that IFN-␣ decreases DNA binding activity of the E2F transcription factor and causes hypophosphorylation of the tumor suppressor protein, Rb (14). Moreover, IFN-␣ can inhibit or induce the expression of the anti-apoptotic protein Bcl-2 and the proto-oncogene/transcription factor c-myc (15,16). However, Bcl-2 and c-Myc proteins and the tumor suppressor p53 might not be key determinants for the apoptotic process induced by IFN-␣ in certain cell lines (9,17).
The role of the Jak/Stat signaling pathway or a second parallel signaling cascade in type I IFN-mediated apoptosis remains to be elucidated. We report here that vanadate in combination with IFN-␣ can promote apoptosis in transformed cell lines that only exhibit antigrowth activity in the presence of only type I IFNs. IFN-␣and vanadate-stimulated pro-grammed cell death does not appear to require p53, but reactive oxygen species (ROS) play a role in this process, in addition to components of the Jak/Stat signaling pathway. These results suggest that IFN-␣-mediated apoptosis requires inhibition of a protein-tyrosine phosphatase(s) and generation of ROS to activate another signaling pathway(s) that allows this cytokine to stimulate programmed cell death.

EXPERIMENTAL PROCEDURES
Cells and Cell Culture-The human leukemic T-cell line Jurkat, subclone E6 was cultured in RPMI 1640 medium supplemented with 10% heat inactivated fetal calf serum (Life Technologies, Inc.), 2 mM L-glutamine, penicillin (10 units/ml) and streptomycin (10 g/ml) at 37°C and 5% CO 2 . The human fibrosarcoma 2fTGH cells, and mutant cell lines derived from this clone deficient in Jak1 (U4A), Tyk2 (U1A), Stat1 (U3A), and Stat2 (U6A) (a gift from G. Stark, Cleveland Clinic Foundation, OH), and the human cell lines U251, HeLa, and Hep 3B were cultured in high glucose Dulbecco's modified Eagle's medium containing 10% fetal calf serum and antibiotics at 37°C. Reconstituted 2fTGH mutant lines were maintained in the presence of 450 g/ml G418.
Cell Stimulation and Preparation of Cell Lysates-Adherent cells at 30% confluence and Jurkat cells (10 6 cells/ml) were incubated with medium, IFN-␣, vanadate, or the combination of IFN-␣ plus vanadate at 37°C for the indicated times. All samples were washed once with cold phosphate-buffered saline and resuspended in lysis buffer containing 1% Triton X-100, 50 mM Tris, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 10 mM ␤-glycerophosphate. Lysates were vortexed, incubated on ice for 10 min, and insoluble material cleared by centrifugation at 12,000 rpm for 10 min at 4°C.
Immunoblot Analysis-Proteins (30 g of whole cell extract) were separated on 8% SDS-PAGE gels and transferred to polyvinylidene difluoride membrane (Millipore, Bedford, MA). Membranes were immunoblotted with the indicated antibodies. Immunoblots were developed using horseradish peroxidase-conjugated secondary antibodies (Zymed Laboratories Inc., San Francisco, CA) and ECL (Amersham Pharmacia Biotech).
Measurement of Apoptosis by Annexin V Staining-Adherent cells plated in six-well plates or suspension cells were diluted to 1 ϫ 10 5 cells/ml. Cells were left untreated or stimulated with IFN-␣, vanadate, or IFN-␣ plus vanadate, followed by incubation at 37°C. Cells were collected, washed in phosphate-buffered saline, and resuspended in 100 l of binding buffer (10 mM HEPES, pH 7.4, 140 mM NaCl, and 2.5 mM CaCl 2 ). Adherent cells were harvested by trypsinization, and detached cells in medium were also collected. Cells were then incubated with 5 l of annexin V-FITC (PharMingen, Palo Alto, CA) and 5 l of propidium iodide (50 g/ml) (Sigma) for 10 min at room temperature in the dark, followed by addition of 400 l of binding buffer. Cells were collected ungated (10,000 events) and analyzed by two-color flow cytometry using a FACScan (Becton-Dickinson, San Jose, CA). Data were further analyzed using CellQuest (Becton-Dickinson) to discriminate between apoptotic versus necrotic cells.
Measurement of Caspase-3 Activity-Cells were plated in 60-mm dishes at 40% confluence and stimulated as described above, and cell extracts were prepared. Caspase-3 activity was assayed in a 150-l reaction mixture with a caspase-3-specific fluorogenic substrate peptide (Ac-DEVD-AMC) (PharMingen, Palo Alto, CA). The substrate peptide (100 M) was incubated with 20 g of cell extract in 20 mM Hepes, pH 7.5, 10% glycerol, 2 mM dithiothreitol at 37°C for 2 h. Fluorescence was measured using a Cytofluor 2350 spectrofluorometer plate reader with an excitation of 380 nm and an emission wavelength of 460 nm (Millipore, Marlborough, MA).

Apoptosis Can Be Induced in Cells by Treatment with IFN-␣
and Vanadate-In order to understand the signaling events associated with type I IFN-mediated apoptosis in certain cellular contexts, we searched for conditions that would allow IFN-␣ to trigger programmed cell death in cells where IFN-␣ normally exerts only antiproliferative responses. Since proteintyrosine phosphatases (PTPases) modulate both type I IFN signaling and also can play a role in cell survival (20 -22), we tested whether vanadate (a PTPase inhibitor) alters the ability of this cytokine to induce an apoptotic response. Human fibrosarcoma 2fTGH cells were incubated with different doses of IFN-␣ in the absence or presence of vanadate (5 M) for 48 h, and apoptosis was measured by annexin-V staining (Fig. 1A). Although treatment of cells with IFN-␣ or vanadate alone did not increase the number of apoptotic cells compared with unstimulated cells, the combination of IFN-␣ and vanadate caused cells to apoptose. Maximal numbers of apoptotic cells were observed when cells were incubated in 5-10 ϫ 10 3 units/ml IFN-␣ (ϳ40%) in the presence of 5 M vanadate. These concentrations of IFN-␣ are also required to achieve maximal antiproliferative actions in 2fTGH cells (data not shown). Similar results were also obtained when we measured apoptosis by terminal dUTP nick-end labeling assay (data not shown). To explore whether the combination of IFN-␣ and vanadate stimulates apoptosis in other cells, HeLa cells (a human cervical carcinoma) and Jurkat cells (a T-cell leukemia) (Fig. 1, B and C) were treated with this combination of agents. Incubation of both cell lines with IFN-␣ and vanadate induced programmed cell death whereas neither compound alone was effective. Unlike 2fTGH cells, HeLa and Jurkat cells required a higher dose of vanadate (25-50 M) to undergo apoptosis in the presence of IFN-␣. These results suggest that the activities of a PTPase(s) might play an important role in determining whether IFN-␣ causes programmed cell death.

Induction of Apoptosis by IFN-␣ but Not by TNF-␣ in the Presence of Vanadate-
To determine if the sensitizing effect of vanadate to IFN-␣-mediated programmed cell death was a global response to all cytokines known to stimulate apoptosis, we tested if vanadate could enhance TNF-␣-induced apoptosis of 2fTGH cells. Treatment of cells with TNF-␣ in combination with actinomycin D has been reported to promote apoptosis in these cells (23). However, TNF-␣ alone does not stimulate cell death. Although incubation of 2fTGH cells with vanadate and IFN-␣ significantly stimulated cells to apoptose, the combination of TNF-␣ with vanadate failed to induce apoptosis in these cells (Fig. 2). Incubation of cells with TNF-␣ and IFN-␣ also did not induce apoptosis (data not shown). Similar observations were made in HeLa and Jurkat cells incubated with TNF-␣ and vanadate (data not shown). Thus, the ability of vanadate to prime cells to undergo programmed cell death does not apply to all pro-apoptotic stimulators.
Priming of 2fTGH Cells with High Dose of Vanadate Is Sufficient to Induce IFN-␣-mediated Apoptosis-To determine whether cells needed to be cultured in the continuous presence of vanadate for IFN-␣ to stimulate apoptosis, 2fTGH cells were incubated with IFN-␣ and various doses of vanadate for 6 h, followed by treatment with either medium or IFN-␣ alone for 40 h (Fig. 3A, lanes 1-6). Although incubation of cells with IFN-␣ and either 5 or 20 M vanadate for 6 h, followed by 40 h of IFN-␣ alone, did not significantly increase the number of apoptotic cells, priming with a higher dose of vanadate (50 M) and IFN-␣ significantly enhanced apoptosis in the continuous presence of IFN-␣ (lanes 2 and 4 versus lane 6). We also found that 6-h priming of cells with 50 M vanadate alone followed by a 40-h incubation with IFN-␣ in the absence of vanadate allowed cells to apoptose (Fig. 3A, lanes 7 and 8). It should be noted that priming of 2fTGH cells for less than 6 h did not induce IFN-␣-mediated apoptosis. Additionally, cells did not apoptose when cultured with IFN-␣ and vanadate for 6 h, followed by low dose vanadate alone for 40-h (data not shown). At this moment, we cannot determine whether this is a transcriptionally or translationally dependent effect. The RNA polymerase inhibitor actinomycin D or the protein synthesis inhibitor cyclohexamide, even when used at low concentrations for 48 h, were both toxic to cells. It thus appears that incubation of cells with vanadate for a short period (i.e. 6 h) initiates a condition that is irreversible with regard to the ability of IFN-␣ to stimulate programmed cell death.
The Antioxidant NAC Blocks the IFN-␣-dependent Priming Apoptotic Effect of Vanadate-Vanadate has recently been reported to generate ROS, which play a role in the induction of programmed cell death in certain cell lines (24). To investigate the possible involvement of ROS in IFN-␣-inducible apoptosis initiated by priming of cells with high dose vanadate for 6 h, 2fTGH cells were left untreated or pretreated with the free radical scavenger NAC (10 mM) for 30 min, incubated with high dose vanadate (50 M) for 6 h followed by stimulation with medium, or IFN-␣ alone for 40-h. NAC was continuously present in the medium of 2fTGH cells for the duration of the experiment (Fig. 3B). Vanadate priming of cells incubated without or with NAC showed a basal level of 5% apoptotic cells. Pretreatment of 2fTGH cells with NAC significantly abrogated the apoptotic effects of vanadate followed by IFN-␣ compared with vanadate-primed cells stimulated with IFN-␣ alone (compare 11% versus 32% apoptotic cells, respectively). Cells primed with vanadate followed by IFN-␣ and NAC treatment were still protected from IFN-␣-induced apoptosis (data not shown). These results suggest that generation of ROS induced by vanadate and/or IFN-␣ and vanadate play a role in sensitizing cells to IFN-␣-mediated apoptosis.
Caspase-3 Activity Is Induced by Treatment of Cells with IFN-␣ plus Vanadate-Caspase-3 activation is one of the hallmarks and commitment steps of programmed cell death (25,26). To address whether induction of apoptosis by IFN-␣ and vanadate correlated with activation of caspases, we measured the activity of caspase-3 in 2fTGH cells, using a fluorogenic peptide substrate. As shown in Fig. 4, neither IFN-␣ nor vanadate stimulated caspase-3 activity in these cells. In contrast, the combination of the two compounds led to the activation of caspase-3 after only 24 h. Maximal caspase-3 activity was detected between 36 and 48 h, which preceded the onset of maximal apoptosis that occurs after 2 days of treatment. In a similar manner, maximal caspase-3 activity was observed in HeLa cells treated with IFN-␣ and vanadate for 24 and 30 h (data not shown). We also found that treatment of 2fTGH cells with the combination of IFN-␣ plus vanadate or either treatment alone did not increase caspase-3 RNA levels compared with unstimulated cells when examined up to 24 h (data not shown). This observation indicates that the onset of apoptosis induced by IFN-␣ and vanadate coincides with caspase-3 activity.

Activation of Stat1 and Expression of IFN-stimulated Genes (ISGs) by IFN-␣ Are Not Enhanced by Incubation of Cells with
Vanadate-Prolonged activation of Jaks and Stats has been reported to correlate with the ability of IFN-␣ to induce cell cycle arrest in sensitive Daudi cells compared with Daudi cells that are resistant to the antiproliferative effects of this cytokine (27). It is known that incubation of cells with high concentrations of vanadate can stimulate tyrosine phosphorylation of Stats (28). Therefore, we wanted to determine whether priming with vanadate would cause a change in the magnitude or duration of IFN-␣-stimulated tyrosine phosphorylation of Stats that could account for the apoptotic effects. 2fTGH cells were left untreated or incubated with IFN-␣, low dose vanadate, or the combination of both compounds for the indicated times, and cell extracts were prepared. The kinetics and magnitude of Stat1 tyrosine phosphorylation was analyzed by immunoblot analysis using a phosphotyrosine-specific Stat1 antibody. Tyrosine phosphorylation of Stat1 induced by IFN-␣ was not enhanced by incubation of cells with vanadate over the course of 6 h (Fig. 5). Interestingly, the low concentrations of vanadate used in these experiments did not induce Stat1 tyrosine phosphorylation (Fig. 5, lanes 2, 5, 8, and 11). We did not detect enhanced tyrosine phosphorylation of Stat1 when shorter term stimulations were performed between 5 and 30 min or when cells were pretreated with low dose vanadate followed by IFN-␣ stimulation (data not shown). We also examined tyrosine phosphorylation of Stat3 and obtained the same results observed with Stat1 in 2fTGH and HeLa cells (data not shown). Since serine phosphorylation of Stat1 is required for maximal transcriptional response of ISGs (29), we analyzed several such RNAs using RNase protection assays (Fig. 6). Vanadate did not augment or prolong expression of ISGs that are activated by ISGF3 (ISG15 and ISG54) via an ISRE element or by Stat1 homodimers (GBP, IRF) via an IFN-␥ response (GAS) element. The level of expression of ISGs at 24 h induced by vanadate and IFN-␣ stimulation were not different from those detected at 18 h (data not shown). We also examined whether vanadate could enhance protein expression levels of the IFN-␣/␤-inducible protein PKR in 2fTGH cells since this ISG plays a role in sensitizing cells to undergo apoptosis (30). Stimulation of 2fTGH cells with IFN-␣ plus vanadate did not augment PKR protein expression when compared with IFN-␣ treatment alone over the course of 40 h (data not shown). This indicates that apoptosis induced by IFN-␣ and vanadate likely requires an additional signal besides ISGs.
Selective Tyrosine Phosphorylation of Proteins Induced by Low Concentrations of Vanadate-Because stimulation of 2fTGH cells with vanadate did not appear to enhance IFN-␣induced gene expression of known ISGs, we wanted to determine whether vanadate alone might induce tyrosine phosphorylation of target proteins that might play a role in IFN-␣-induced apoptosis. 2fTGH cells were incubated with low dose vanadate (5 M) over the course of 30 min, and tyrosine-phosphorylated proteins were detected by immunoblot analysis. Fig. 7 shows the appearance of two distinct proteins that became tyrosinephosphorylated upon vanadate stimulation as early as 15 min. The identity of these proteins is unknown; however, we know these proteins are not Stat1, Stat2, and Tyk2 (data not shown). Stimulation of 2fTGH cells with IFN-␣ plus vanadate did not increase the tyrosine phosphorylation content of these proteins or the appearance of new tyrosine-phosphorylated proteins (data not shown). We also could not detect inducible protein tyrosine phosphorylation by IFN-␣ in other cell lines where this cytokine induces apoptosis (H9 and U266 cell) or cell growth arrest (Jurkat) (data not shown). The fact that increases in tyrosine phosphorylation were not detected in IFN-␣-stimulated cells is not surprising because the ability of the antiphosphotyrosine antibody to recognize all tyrosine-phosphorylated proteins in total cell extracts is limited. Our results indicate that vanadate modulates the tyrosine phosphorylation of at least two distinct proteins that might be important in sensitizing cells to IFN-␣-mediated apoptosis.
Defective Apoptosis in the Absence of Jak/Stat Signaling Components-The experiments shown in Figs. 5 and 6 indicate that alterations in IFN-␣ stimulation of the Jak/Stat pathway are probably not sufficient to account for the ability of IFN-␣ and vanadate to stimulate programmed cell death in a variety of cells. However, we wanted to know whether components of the Jak/Stat cascade were necessary for the apoptotic effects of IFN-␣. To address this issue, we used 2fTGH mutant lines that do not express Stat1, Stat2, Jak1, or Tyk2. Cells were stimulated with IFN-␣ plus vanadate or either agent alone and assayed for apoptosis by annexin V staining after 48 h. In the   FIG. 4. Caspase-3 activity is induced by treatment of cells with IFN-␣ plus vanadate. 2fTGH cells were stimulated with medium, IFN-␣ (5000 units/ml), vanadate (5 M), or the combination of IFN-␣ plus vanadate for the indicated times. Cell extracts were prepared, and caspase-3 activity was measured using Ac-DEVD-AMC peptide as a substrate. This is a representative experiment of two that were performed. Results are presented as mean Ϯ S.D. of duplicate samples. absence of any of these proteins, IFN-␣ and vanadate failed to stimulate apoptosis (Fig. 8). Elevated annexin V staining was detected in Stat1Ϫ/Ϫ and Stat2Ϫ/Ϫ cells incubated with vanadate; however, the reason why this occurs is not clear. To determine if apoptotic function could be restored in these cell variants, the wild type forms of these proteins were expressed in the appropriate deficient mutant cell lines (Fig. 9). Expression of the missing protein restored IFN-␣-and vanadatestimulated apoptosis. We also examined whether the kinase activity of either Jak1 or Tyk2 was required for the apoptotic response by reconstituting the variants with kinase-inactive forms of these enzymes (Fig. 9, C and D). Neither kinaseinactive Tyk2 nor Jak1 restored IFN-␣-plus vanadate-induced programmed cell death in these lines. The fact that kinaseinactive Tyk2 did not restore the apoptotic response is particularly interesting since it is well documented that IFN-␣-stimulated ISGs occur in Tyk2Ϫ/Ϫ cells that express kinase-dead Tyk2 (31).
Phosphorylation of tyrosine 701 (Tyr-701) is required for Stat1 to dimerize, translocate to the nucleus, and bind DNA. In addition, serine 727 (Ser-727) must also be phosphorylated for Stat1 to function as a maximal activator of transcription (29). Although a substitution of Tyr-701 with phenylalanine (Y701F) prevented restoration of the apoptotic response, a mutation on Ser-727 to alanine on Stat1 (S727A) and Stat1-␤, which is truncated at the carboxyl terminus and contains Tyr-701 but not Ser-727, both restored apoptosis to similar levels as the wild type Stat1 (Fig. 9A).
IFN-␣-plus Vanadate-induced Apoptosis Does Not Require p53-The tumor suppressor protein p53 is a critical modulator in the onset of apoptosis following DNA damage; however, not all apoptotic pathways require p53 (32,33). Induction of apoptosis by IFN-␣ alone in certain hematopoietic cell lines occurs through a p53-independent pathway (9). Recently, vanadate has been reported to induce apoptosis by activating p53 activity through the generation of ROS (34). To address the involvement of p53 in IFN-␣-and vanadate-mediated induced apoptosis, we used the following human cell lines: Hep 3B (a p53 negative hepatocellular carcinoma), HeLa (a HPV positive cervical carcinoma with decreased p53 activity), and U251 (a p53 mutant glioma) (Fig. 10). Treatment of each cell type with IFN-␣ plus vanadate (ϳ23% apoptotic cells) but not with IFN-␣ or vanadate alone stimulated programmed cell death. These results indicate that apoptosis induced in cells by IFN-␣ plus vanadate occurs in a p53-independent manner.

DISCUSSION
The Jak/Stat pathway is required for the antiproliferative actions of IFN-␣; however, its role if any in determining whether IFN-␣ will inhibit cell cycle progression, either by arrest or slowing without accumulation in a single phase, or induce programmed cell death, is unclear. These studies begin to address this question by examining cell lines that are growth-inhibited by IFN-␣ but do not apoptose when exposed to this cytokine. Treatment of cells with vanadate allowed IFN-␣ to stimulate apoptosis, whereas vanadate alone had no effect. Vanadate is a well known specific inhibitor of PTPases, so presumably its effects are associated with altering the cellular activity of these enzymes; however, this needs to be further substantiated. The effects of vanadate are not global, since TNF-␣, a known activator of apoptosis, could not substitute for IFN-␣ to stimulate apoptosis in the cell lines we examined.
Type I IFNs activate the Jak/Stat pathway resulting in the transcription of genes responsible for the biological effects of this cytokine. Expression of Jak1 and Tyk2 are required for the actions of type I IFNs (35,36). In the absence of Stat2, Stat1 becomes weakly tyrosine-phosphorylated and transcription of IFN-stimulated genes (ISGs) is defective (37,38). To address whether components of the Jak/Stat pathway were needed to induce apoptosis, we examined 2fTGH cell variants that do not express Jak1, Tyk2, Stat1, or Stat2. We found that expression of these signaling components was required since reconstitution of these lines with the missing component restored apoptosis induced by the combination of IFN-␣ and vanadate. Although Tyk2 devoid of kinase activity has been shown to sustain expression of ISGs by IFN-␣/␤ (31), expression of a kinase inactive Tyk2 in Tyk2Ϫ/Ϫ cells did not restore the apoptotic action of IFN-␣ and vanadate (Fig. 9D). This observation, in conjunction with the finding that vanadate does not alter the magnitude and duration of ISG induction, suggests that other events must occur in addition to activation of the classical Jak/Stat pathway, to allow for an apoptotic response.
For Stat1 to be fully active, it must be phosphorylated on Tyr-701 to dimerize and translocate to the nucleus, and phosphorylated on Ser-727 to maximally activate gene transcription (4,29). Although the antiproliferative effects of IFN-␣ and IFN-␥ are impaired when Stat1 is not expressed (39), some of the activities of these cytokines are maintained, such as stimulation of c-myc expression (40). Stat1 has also been shown to be necessary for the expression of caspases (23). In this context, Stat1 does not need to be tyrosine-phosphorylated. Stat1 Ser-727 is required to induce apoptosis by the combination of TNF-␣ and actinomycin D and to restore expression of caspase-3 (23). In contrast to TNF-␣-mediated apoptosis, mutation of Tyr-701 of Stat1 abrogated IFN-␣-mediated apoptosis, and phosphorylation of Ser-727 was not essential for the activation of apoptosis by IFN-␣ and vanadate. The onset of apoptosis induced by vanadate plus IFN-␣ correlated with enhanced caspase-3 activity (Fig. 4).
Different groups have tried to characterize the mechanisms that regulate the antigrowth properties of IFN-␣. Some studies show that IFN-␣-mediated cell growth arrest is associated with prolonged activation of the Jak/Stat pathway (27). We find that a change in the duration or magnitude of IFN-␣-induced ISG transcription is unlikely to be sufficient for the apoptotic ac- tions of this cytokine (Fig. 6). A domain in the IFN receptor subunit, IFNAR2, that has been reported to regulate the antiproliferative actions of type I IFNs may interact with a PTPase (41). It is not known whether this domain is required for the apoptotic actions of IFN-␣ and vanadate or whether vanadate alone can regulate the function of the type I IFN receptor. Our data infer that the activities of PTPase(s) act as regulatory switches that determine whether IFN-␣ can induce an apoptotic response. It is also possible that a PTPase(s) might be associated with a specific domain in type I IFN receptors that regulates apoptosis. Perhaps, cells that are susceptible to IFN-␣-mediated apoptosis may retain these PTPases in an inactive state, and only after an additional signal is provided, through IFN-␣ binding its receptor, will cells be programmed for destruction. The observation that incubation of cells with higher concentrations of vanadate for only 6 h is sufficient to allow IFN-␣ to induce programmed cell death 40 h later (Fig. 3) suggests that the event initiated by vanadate is stable and/or irreversible. The fact that free radical scavengers could in part abrogate the priming apoptotic effect of vanadate suggests that vanadate may not only be inhibiting the activities of PTPases, but also contributes to IFN-␣-mediated induced apoptosis through the generation of ROS. Several reports have suggested a role for ROS in modulating IFN activity. These effects have not been clearly associated with an IFN-regulated signaling cascade or a well defined biological action of these cytokines (42,43). Although p53 plays a pivotal role in the activation of many apoptotic pathways and in the generation of ROS by vanadate, we found that activation of the apoptotic signaling network induced by IFN-␣ plus vanadate occurs independently of p53. We have also examined several other signaling cascades that are activated by IFN-␣, including activation of Erk1/2, and p38 mitogen-activated protein kinase (44 -46), but vanadate does not seem to alter the ability of IFN-␣ to regulate these enzymes (data not shown). IFN-␣ and vanadate treatment of cells also does not appear to alter cell survival or death promoting signals such as Akt or Bcl-2 (data not shown). It is notable that immunoblots with anti-phosphotyrosine antibodies reveal two proteins that are selectively tyrosine-phosphorylated in cells incubated with low concentrations of vanadate (Fig. 7). These tyrosine-phosphorylated proteins are not Stat1, Stat2, or Tyk2 but may provide another molecular handle to gain further insights into the key cellular events that contribute to the ability of IFNs to inhibit cell growth by inducing programmed cell death. Such information will have significant implications with regard to the therapeutic effects of this cytokine.