Inhibition of signal transducer and activator of transcription 3 activity results in down-regulation of Survivin following irradiation

Signal transducer and activator of transcription 3 (Stat3) was originally identified as an acute-phase response factor that responds to stimulation by interleukin-6. One of seven members of the STAT family of proteins, it transduces cytokine and growth factor receptor signaling from the cytoplasm to the nucleus, activating expression of many targeted genes (1, 2), and also plays an important role in several pathologic events in oncogenesis, including cell transformation, cell cycle progression, cell survival, and angiogenesis (3, 4). In these processes, persistent transcriptional activity of Stat3 results from phosphorylation of a single tyrosine residue (Tyr⁷⁰⁵) by overactive receptor-associated tyrosine kinases, such as Janus-activated kinase family members or non–receptor-associated tyrosine kinases like SRC (5, 6). Constitutive Stat3 activation occurs frequently in a variety of human tumor cell lines, including solid tumors, such as breast, gastric, and prostate cancer along with several leukemias and lymphomas (6–9). A gene therapy vector in murine melanoma tumors in nude mice originally validated the potential targeting of Stat3 in cancer therapy, and strategies inhibiting this protein have been shown to promote cell death in several human cancer cell lines (10–17).

There is mounting evidence that Stat3 may mediate the expression of Survivin, the smallest member of the inhibitor of apoptosis gene family and a suppressor of caspases. Survivin is overexpressed in most human cancers (18) and plays a critical role as a factor mediating radiation resistance (19). Recent studies have shown that activation of Stat3 supports cell survival in association with Survivin expression in gastric cancer cells, whereas inhibition of Stat3 decreases Survivin expression and induces apoptosis in primary effusion lymphoma and breast cancer cells (9, 12, 20, 21). Like novel strategies targeting Stat3, several studies have shown that down-regulation of Survivin through methods like ASO, dominant-negative mutants, cyclin-dependent kinase inhibitors, and ribozymes can be effective at inhibiting tumor growth and enhancing more traditional therapeutics in human tumor cell models (22–26).

Radiation induces expression of a variety of transcription factors, such as activator protein, Sp-1, p53, and nuclear factor-κB in the rat brain (27). We have previously found that 3 Gy significantly reduced both protein and mRNA levels of Survivin in human umbilical vein endothelial cells (HUVEC) but not in tumor cell lines (28). There is much to be known, however, regarding regulation of Stat3 by irradiation. In the present study, we examined how Stat3 activity affects Survivin expression following irradiation in HUVECs and MDA-MB-231 breast cancer cells. We also investigated whether inhibition of Survivin and Stat3 through the use of dominant-negative mutants could radiosensitize breast cancer cells.

HUVECs were cultured in endothelial basal medium-2 supplemented with endothelial growth medium MV single aliquots (Bio Whittaker, Walkersville, MD). MDA-MB-231 cells were cultured in BMEM (DMEM, Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37°C and humidified 5% CO₂. Primary human mammary epithelial cells overexpressing wild-type or Y705F mutant Stat3 were provided by Dr. George R. Stark (New York University School of Medicine, New York, NY) and grown in Mammary Endothelial Cell Growth Medium (Cell Applications, Inc., San Diego, CA) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37°C and humidified 5% CO₂ (29).

Adenoviruses pAdCMVpLPA(−)Xop-SSP (vector control), Stat3-C (constitutively active Stat3), and mutant Stat3Y705F were provided by Dr. Beverly L. Davidson (Gene Transfer Vector Core, University of Iowa). The procedures of the infection were as described previously (30). Cells (0.5 × 10⁶) were plated into 10-cm culture dish 2 days before infection. The cells were infected with adenoviruses overexpressing a control empty vector (Stat3-C) or Stat3Y705F, respectively, at a multiplicity of infection of 100 plaque-forming units/cell. After 5 hours of incubation, the culture medium was removed and replaced with fresh medium. The cells were harvested and analyzed after 24 hours.

Cells (0.5 × 10⁶) were treated with various doses of radiation and collected at various time points. The cells were harvested and washed with ice cold PBS twice before the addition of lysis buffer [20 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 20 mmol/L EDTA, 1% NP40, 50 mmol/L NaF, 1 mmol/L Na₃VO₄, 1 mmol/L NaMO₄, and cocktail inhibitor (Sigma, St. Louis, MO; 5 μL/mL)]. Protein concentration was quantified by the Bio-Rad (Richmond, CA) method. Equal amounts of protein were loaded into each well and separated by 10% SDS-PAGE gel followed by transfer onto polyvinylidene difluoride membranes (Bio-Rad). Membranes were blocked by use 5% nonfat dry milk in PBS-T for 1 hour at room temperature. The blots were then incubated with the rabbit Survivin (R&D Systems, Minneapolis, MN), phospho-Stat3 (Tyr⁷⁰⁵, Cell Signaling, Beverly, MA and Ser⁷²⁷, MBL, Woburn, MA), and Stat3 (BD Transduction Laboratories, Lexington, KY) antibodies. Goat anti-rabbit IgG secondary (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA) was incubated for 45 minutes at room temperature. Immunoblots were developed by using the enhanced chemiluminescence detection system (Amersham, Arlington Heights, IL) and autoradiography.

Cells (3 × 10⁶) were harvested and washed with cold PBS and resuspended in 50 μL of lysis buffer [10 mmol/L HEPES (pH 7.9), 1.5 mmol/L MgCl₂, 10 mmol/L KCl, 0.5 mmol/L DTT, and cocktail inhibitor (Sigma; 5 μL/mL)]. The cells were allowed to swell on ice for 10 minutes, after which the cells were resuspended in 30 μL of lysis buffer containing 0.05% NP40. The tube was then was vigorously mixed on a vortex machine thrice for 10 seconds, and the homogenate was centrifuged at 250 × g for 10 minutes to pellet the nuclei. The nuclear pellet was resuspended in 40 μL of ice-cold nuclear extraction buffer [5 mmol/L HEPES (pH 7.9), 26% glycerol (v/v), 1.5 mmol/L MgCl₂, 0.2 mmol/L EDTA, 422 mmol/L NaCl, 0.5 mmol/L DTT, and cocktail inhibitor (Sigma; 5 μL/mL)], incubated on ice for 30 minutes with intermittent mixing, and centrifuged at 24,000 × g for 20 minutes at 4°C. The nuclear extract was either used immediately or stored at −80°C. Electrophoretic mobility shift assay was done by incubating 5 μg of nuclear extract for 20 minutes at room temperature with 17.5 fmol of ³²P-end labeled 24-mer double-stranded STAT3 oligonucleotide (5′-GATCCTTCTGGGATTTCCTAGATC-3′; Santa Cruz Biotechnology) in binding buffer [10 mmol/L Tris-HCl (pH 7.9), 50 mmol/L NaCl, 1 mmol/L MgCl₂, 0.5 mmol/L DTT, 0.5 mmol/L EDTA, 4% glycerol] containing 50 μg/mL of poly(deoxyinosinic-deoxycytidylic acid). The DNA-protein complex form was separated from free oligonucleotide on 4% nondenaturing polyacrylamide gel using 0.5× Tris-borate EDTA buffer [44.5 mmol/L Tris-HCl (pH 8), 44.5 mmol/L boric acid, 1 mmol/L EDTA]. For gel mobility supershift assay, 1 μg of anti-Stat3 antibody was incubated with nuclear extracts on ice for 30 minutes before Stat3 (C20; Santa Cruz Biotechnology) binding reaction.

MDA-MB-231 cells were infected with adenoviruses overexpressing a control empty vector LacZ, DN-Survivin, or DN-Stat3, respectively, at a multiplicity of infection of 100 plaque-forming units/cell for 4 hours. The transfected cells were radiated with 0, 2, 4, or 6 Gy, given from a cobalt source. After 8 to 10 days, cells were fixed for 15 minutes with 3:1 methanol/acetic acid and stained for 15 minutes with 0.5% crystal violet (Sigma) in methanol. After staining, colonies were counted using a cutoff of 50 viable cells. Surviving fraction was calculated as (mean colony counts / cells inoculated) × plating efficiency, where plating efficiency was defined as mean colony counts divided by cells inoculated for nonirradiated controls.

HUVECs were infected with adenoviruses overexpressing the control empty vector LacZ, DN-Survivin, or DN-Stat3, respectively, at a multiplicity of infection of 100 plaque-forming units/cell for 4 hours. Medium was changed, and cells were treated with 3 Gy. Cells were trypsinized and counted. They were seeded at 48,000 per well on 24-well plates coated with 300 μL of Matrigel (BD Biosciences, San Jose, CA). These cells undergo differentiation into capillary-like tube structures and were periodically observed by microscope. One day later, cells were stained with H&E, and photographs were taken via microscope. The average number of tubes for three separate microscopic fields (×100) and representative photographs were taken.

As noted above, Stat3 activation has been shown to be associated with Survivin expression in various malignant cell lines. We previously showed that Survivin is transcriptionally down-regulated by radiation in HUVECs (28). To determine whether Stat3 regulates the expression of Survivin in HUVECs, we transduced HUVECs with three different adenovirus vectors. We transduced HUVECs with a constitutively active Stat3 vector (Stat3-C) that is capable of dimerizing and constitutively binding DNA without undergoing phosphorylation of the Tyr⁷⁰⁵ site (31). A second adenovirus vector, a dominant-negative mutant form of Stat3, inhibits Stat3 activity. We also transduced a control empty vector [pAdCMVpLPA(−)Xop-SSP]. We found the expression of Survivin to be positively correlated with the level of phospho-Stat3 Tyr⁷⁰⁵ (Fig. 1). The level of Survivin protein detected by Western immunoblot was increased by 85% in the cells with constitutively active Stat3-C and decreased in the DN-Stat3–expressing mutant cells in comparison with the control. Consistently, phospho-Stat3 Tyr⁷⁰⁵ was increased in the constitutively active Stat3-C cells and decreased in DN-Stat3 relative to control. The measured total Stat3 level was elevated in cells with both DN-Stat3 and Stat3-C activity because the antibody employed exhibited equal binding affinity for both DN-Stat3 and constitutively active Stat3-C.

After showing that Stat3 was positively correlated with Survivin expression in endothelial cells, we next sought to investigate the effects of irradiation on Stat3 activation and DNA-binding capability in these cells (Fig. 2A–C). In our Western immunoblots, we set the ratio of phosphorylated Stat3 to total Stat3 of our control to 1. The ensuing data represent the relative difference of phosphorylated Stat3 to total Stat3 from the control. In HUVECs, phospho-Stat3 Tyr⁷⁰⁵ was reduced by 73% following irradiation at 30 minutes. Levels of phospho-Stat3 Tyr⁷⁰⁵ remained depressed at 60 minutes but had mostly recovered by 240 minutes (Fig. 2A). Irradiation did not, however, significantly affect the level of phospho-Stat3 Ser⁷²⁷. There was a small increase in phosphorylation at this site following irradiation (Fig. 2B).

Stats induce gene expression by physically binding with specific sequences in the promoter regions of responsive genes. We used electrophoretic mobility shift assay to determine whether DNA-binding activity of Stat3 in HUVECs is reduced following irradiation. We observed that Stat3 DNA binding was markedly reduced at 30 minutes following irradiation, was still decreased at 60 minutes, and had returned to baseline after 240 minutes (Fig. 2C).

Although irradiation down-regulates Survivin in HUVECs, we previously showed that 3 Gy failed to down-regulate Survivin in MDA-MB-231 breast and other cancer cell lines (28). We used Western immunoblotting and electrophoretic mobility shift assay to investigate the level of phosphorylated Stat3 (Tyr⁷⁰⁵ and Ser⁷²⁷) and DNA binding activity of Stat3 following 3 Gy in breast cancer cells. Again, we set the ratio of phosphorylated Stat3 to total Stat3 in our control to 1 with the subsequent data representing the relative difference of phosphorylated Stat3 to total Stat3 from the control. Phospho-Stat3 Tyr⁷⁰⁵ was unchanged at 30 minutes following irradiation (Fig. 3A). There was a slight decrease at 60 and 240 minutes. No significant change in phospho-Stat3 Ser⁷²⁷ levels was detected following irradiation at 30 and 60 minutes, although an 11% increase was noted at 240 minutes (Fig. 3B). Figure 3C shows no differences in Stat3 DNA-binding activity accompanied following irradiation of MDA-MB-231 cells.

To determine whether dephosphorylation of Stat3 Tyr⁷⁰⁵ is essential for radiation-induced down-regulation of Survivin, we subjected primary human mammary epithelial cells overexpressing either wild-type Stat3 or mutant Y705F Stat3 to 3 Gy. Mutant Y705F Stat3 cannot be phosphorylated on residue 705. Survivin expression was determined by Western blotting following irradiation. The ratio of Survivin to actin in the control was set to 1, and the data represented relative changes in this ratio from the control. Following irradiation, Survivin levels were decreased by nearly 70% in the wild-type cells at 6 hours and remained depressed at 12 hours. In contrast, Survivin levels were unchanged by irradiation in the mutant Y705F Stat3 cells at both 6 and 12 hours, respectively (Fig. 4).

To investigate the effects of inhibition of Stat3 and Survivin on MDA-MB-231 breast cancer cells subjected to varying doses of radiotherapy, we did clonogenic assays. Cells transfected with DN-Stat3 treated with irradiation downshifted the survival curve compared with control with a dose-enhancing ratio of 1.18 (Fig. 5). Cells transfected with DN-Survivin were also sensitized to radiotherapy with a dose-enhancing ratio of 1.26. Inhibition of both Survivin and Stat3 together resulted in the greatest radiation sensitization of MDA-MB-231 breast cancer cells after 2, 4, and 6 Gy. The dose-enhancing ratio for this last assay was 1.65, a 2- to 3-fold reduction of cell viability compared with inhibition of either Survivin or Stat3 individually.

After detecting the potential efficacy of DN-Stat3 and DN-Survivin in increasing radiation sensitivity of breast cancer cells (Fig. 5), we next sought to determine what effect inhibition of these proteins would have in combination with irradiation on angiogenesis. We subjected HUVECs infected with DN-Stat3, DN-Survivin, and LacZ control vectors to irradiation, stained the cells, and examined their differentiation into capillary-like tube structures under the microscope. Without irradiation, DN-Survivin and DN-Stat3 vectors each individually decreased differentiation of HUVECs into tubules, but DN-Survivin and DN-Stat3 together were more effective. Inhibition of Stat3 or Survivin with their dominant-negative mutants even more markedly decreased angiogenesis along with 3 Gy. However, the combination of DN-Survivin, DN-Stat3, and irradiation resulted in maximum inhibition of angiogenesis, decreasing the number of tube structures formed by ∼7-fold relative to treatment with 3 Gy alone.

Previously, we showed that irradiation (3 Gy) significantly reduced both protein and mRNA levels of Survivin in HUVECs by decreasing the promoter activity and down-regulating transcription of the Survivin gene. This down-regulating effect of radiation was not observed in malignant cell lines (3). In this study, we further explored the mechanism of how radiation induces down-regulation of Survivin to better understand the different molecular response of HUVECs and malignant cells to radiation. It is generally accepted that COOH-terminal tyrosine phosphorylation of Stat3 is required for its dimerization, nuclear translocation, and ability to bind DNA, whereas phosphorylation of Ser⁷²⁷ enhances Stat3-mediated transcriptional activity (32, 33). We have shown for the first time, to our knowledge, that dephosphorylation of the Tyr⁷⁰⁵ site of Stat3 is required for radiation induced down-regulation of Survivin. Furthermore, we have shown that phospho-Stat3 Tyr⁷⁰⁵ may mediate radiation resistance in breast cancer cells by up-regulating Survivin.

Stat3 has previously been shown to regulate Survivin in various cell lines. Gritsko et al. found that Stat3 activation and elevated Survivin levels occurred concurrently in high-risk breast tumors that were resistant to neoadjuvant chemotherapy, highlighting the importance of better understanding the relationship and regulation of these two proteins (20). Here, we showed that expression of Survivin was positively correlated with the level of phospho-Stat3 Tyr⁷⁰⁵ in HUVECs (Fig. 1). We detected the level of Survivin to be increased by nearly 2-fold in cells with constitutively active Stat3-C relative to control. This increase in Survivin paralleled the increase in phospho-Stat3 Tyr⁷⁰⁵ in cells expressing constitutively active Stat3-C. We showed that 3 Gy caused over a 70% decrease in the ratio of phospho-Stat3 Tyr⁷⁰⁵ to total Stat3 relative to control in HUVECs (Fig. 2). In contrast to what was observed in HUVECs, irradiation caused no change in phospho-Stat3 Tyr⁷⁰⁵ at 30 minutes and only a small, insignificant reduction at 60 and 240 minutes in the malignant MDA-MB-231 breast cancer cell line (Fig. 3). These data corresponded with our electrophoretic mobility shift assay results. There was diminished Stat3 DNA-binding activity following irradiation in HUVECs at 30 and 60 minutes. In contrast to HUVECs, no change in the DNA-binding activity of Stat3 in MDA-MB-231 breast cancer cells was detected. We have previously shown that 3 Gy does not alter Survivin levels in MDA-MB-231 breast cancer cells (28). We were subsequently able to confirm that dephosphorylation of phospho-Stat3 Tyr⁷⁰⁵ is essential for radiation induced down-regulation of Survivin because 3 Gy failed to down-regulate Survivin in cells expressing Stat3 with a mutant Tyr⁷⁰⁵ phosphorylation site. Of note, Mahboubi et al. previously found that Stat3 was required for interleukin-11 induced up-regulation of Survivin expression in HUVECs (34). Our results are consistent with their work, and we show here the novel observation that phosphorylation of the Tyr⁷⁰⁵ site of Stat3 regulates Survivin expression in response to irradiation.

Previously, UV B light has been shown to dephosphorylate phospho-Stat3 Tyr⁷⁰⁵ in cultured keratinocytes (35). These authors showed that vanadate-treated keratinocytes were resistant to UV B–induced apoptosis. Vanadate is a phosphatase inhibitor, suggesting that a tyrosine phosphatase was responsible for UV B–induced dephosphorylation and inactivation of Stat3. Our group is currently conducting ongoing studies directed at identifying the tyrosine phosphatase that mediates the dephosphorylation of phospho-Stat3 Tyr⁷⁰⁵ by irradiation in HUVECs.

In brief, our data provide a molecular understanding of why HUVECs respond to radiotherapy, whereas MDA-MB-231 breast cancer cells do not. We showed in MDA-MB-231 breast cancer cells that like Survivin expression, Stat3 activity (as measured by phospho-Stat3 Tyr⁷⁰⁵ levels and ability to bind DNA) does not decrease in response to radiation therapy. This is in stark contrast to the drop in Stat3 activity following 3 Gy in HUVECs we show here, as well as the decrease in Survivin following irradiation in HUVECs we have previously shown (28). Furthermore, we show that phospho-Stat3 Tyr⁷⁰⁵ activity is required to down-regulate Survivin. These results are consistent with prior studies showing Survivin's regulation by Stat3, as well as the role of these two proteins in promoting cell proliferation and resistance to apoptosis.

Having established that dephosphorylation of phospho-Stat3 Tyr⁷⁰⁵–mediated down-regulation of Survivin following irradiation, we hypothesized that inhibiting both Survivin and Stat3 together would promote radiation sensitivity in MDA-MB-231 breast cancer cells. In prior studies, independent inhibition of Stat3 or Survivin using various techniques, including dominant-negative mutant proteins, has promoted apoptosis in various malignant and nonmalignant cell lines. Although DN-Survivin and DN-Stat3 proteins were each effective at promoting radiation sensitivity, to our knowledge, this is the first report of successful targeting of both Stat3 and Survivin to promote radiation sensitization.

We provided further support for the potential targeting of Survivin and Stat3 together to promote radiation sensitization in an endothelial cell angiogenesis model. Vascular endothelial growth factor is produced in greater-than-normal amounts by cancer cells. It binds to tyrosine kinase receptors on endothelial cells and stimulates growth of blood vessels. Along with hypoxia-inducible factor-1, Stat3 has been shown to be one of the most important transcription factors that regulate vascular endothelial growth factor (36, 37). Moreover, Stat3 participates in cell signaling involving the basic fibroblast growth factor receptor, which also leads to blood vessel proliferation and migration (6). Prior studies have also established the importance of Survivin in tumor formation and extension through angiogenesis (26, 32), and vascular endothelial growth factor, in turn, has been shown to regulate Survivin in endothelial cells (38). Here, we showed that inhibition of Survivin and Stat3 with their dominant-negative mutants in combination with irradiation resulted in a severalfold greater inhibition of angiogenesis than radiotherapy alone.

It is noteworthy that in our electrophoretic mobility shift assay experiment (Fig. 2C), we observed that the DNA-binding activity of Stat3, although still depressed, seemed to recover more quickly at 60 minutes than had the level of phospho-Stat3 Tyr⁷⁰⁵. It is possible that the action of protein inhibitors of activated STATs, which interact with STATs and block DNA binding, could be altered by irradiation in HUVECs, resulting in mildly increased DNA binding despite a still low level of tyrosine phosphorylation (6). We were not surprised to detect similar levels of Survivin expression in human primary mammary epithelial cells expressing Stat3 with mutated Tyr⁷⁰⁵ and wild-type Stat3 before 3 Gy (Fig. 4). Besides STAT activation, other non cell cycle–dependent pathways have been shown to up-regulate Survivin, including growth factor receptor signaling, phosphatidylinositol 3-kinase/Akt signaling, and Ras expression (32).

Although unlike the dramatic change seen in phospho-Stat3 Tyr⁷⁰⁵ secondary to irradiation in HUVECs, we did detect a small trend of increased phosphorylation at the Ser⁷²⁷ site of Stat3 over time in both HUVECs and MDA-MB-231 breast cancer cells following 3 Gy. Phosphorylation of Ser⁷²⁷ of Stat3 can be caused by stressors, such as UV irradiation, hyperosmolality, or exposure to inflammatory mediators like tumor necrosis factor-α and lipopolysaccharide, and may actually decrease the rate of tyrosine phosphorylation in some instances (33). Several groups have shown that serine phosphorylation of Stat3 has no influence on DNA-binding activity of the protein (39, 40). Furthermore, phosphorylation of Stat3 Ser⁷²⁷ does not enhance tyrosine phosphorylation (33). Our findings in both HUVECs and breast cancer cells of a small increase in phospho-Stat3 Ser⁷²⁷ over time following 3 Gy occurring independently of tyrosine phosphorylation correlates well with these prior reports, as does our data showing that radiation-induced down-regulation of Stat3 DNA-binding activity is dependent on dephosphorylation of phospho-Stat3 Tyr⁷⁰⁵.

In conclusion, this study notably showed that irradiation stimulates dephosphorylation of phospho-Stat3 Tyr⁷⁰⁵ in endothelial but not breast cancer cells. Dephosphorylation of Stat3 Tyr⁷⁰⁵ is an essential step in down-regulating Survivin. These results suggest that the persistence of phospho-Stat3 Tyr⁷⁰⁵ may lead to radiation resistance in breast cancer cells. Therefore, targeting Stat3 and Survivin with dominant-negative mutant proteins may sensitize breast cancer and its vasculature to radiation therapy.