Sage Journals: Discover world-class research

Abstract

Aims: Mitochondrial thioredoxin reductase (Txnrd2) is a central player in the control of mitochondrial hydrogen peroxide (H₂O₂) abundance by serving as a direct electron donor to the thioredoxin-peroxiredoxin axis. In this study, we investigated the impact of targeted disruption of Txnrd2 on tumor growth. Results: Tumor cells with a Txnrd2 deficiency failed to activate hypoxia-inducible factor-1α (Hif-1α) signaling; it rather caused PHD2 accumulation, Hif-1α degradation and decreased vascular endothelial growth factor (VEGF) levels, ultimately leading to reduced tumor growth and tumor vascularization. Increased c-Jun NH2-terminal Kinase (JNK) activation proved to be the molecular link between the loss of Txnrd2, an altered mitochondrial redox balance with compensatory upregulation of glutaredoxin-2, and elevated PHD2 expression. Innovation: Our data provide compelling evidence for a yet-unrecognized mitochondrial Txnrd-driven, regulatory mechanism that ultimately prevents cellular Hif-1α accumulation. In addition, simultaneous targeting of both the mitochondrial thioredoxin and glutathione systems was used as an efficient therapeutic approach in hindering tumor growth. Conclusion: This work demonstrates an unexpected regulatory link between mitochondrial Txnrd and the JNK-PHD2-Hif-1α axis, which highlights how the loss of Txnrd2 and the resulting altered mitochondrial redox balance impairs tumor growth as well as tumor-related angiogenesis. Furthermore, it opens a new avenue for a therapeutic approach to hinder tumor growth by the simultaneous targeting of both the mitochondrial thioredoxin and glutathione systems. Antioxid. Redox Signal. 22, 938–950.

Introduction

During tumor growth, an increase in the levels of intracellular reactive oxygen species (ROS) is commonly observed (30). However, protective enzymatic antioxidant systems are upregulated, simultaneously (6). Accordingly, these antioxidant systems have been suggested as targets for redox-dependent anti-cancer treatment (47).

Innovation

Investigations regarding the role of reactive oxygen species (ROS) in PHD2 stabilization and thus hypoxia-inducible factor-1α (Hif-1α) signaling in tumor development have yielded controversial results. Taking advantage of well-defined knockout tumor models, evidence is provided that the chronic leakage of mitochondrial ROS triggers PHD2 stabilization in a JNK-dependent manner that, ultimately, leads to impaired tumor angiogenesis and tumor growth. This study highlights the relevance of mitochondrial thioredoxin reductase as a yet-unrecognized critical player in tumor development.

In mammalian cells, the major intracellular sites for ROS production are the mitochondria, in which leakage of electrons generates superoxide that is readily dismutated by superoxide dismutase to hydrogen peroxide (H₂O₂). To maintain ROS at nonpathological levels, cells are equipped with highly efficient antioxidant redox networks. With regard to mitochondrial redox homeostasis, the peroxiredoxin III (Prx III)/mitochondrial thioredoxin (Txn2)/mitochondrial thioredoxin reductase (Txnrd2) axis is considered the most important H₂O₂-scavenging system (14, 45), as corroborated by reverse genetics in mice: Mice lacking Prx III are viable but show increased sensitivity to ROS-generating toxins (29); genetic deletion of either Txn2 or Txnrd2 causes embryonic lethal phenotypes during mid-gestation and fetal development, respectively (10, 36).

Hypoxia is another classical hallmark of tumor growth and metastasis (21). The individual components of hypoxic signaling downstream of restricted oxygen and nutrient supply are well established and comprise the inactivation of Hypoxia-inducible factor prolyl hydroxylase (PHD) and subsequent stabilization of Hypoxia-inducible factor-1α (Hif-1α), culminating in increased transcription of Hif-1α target genes (41). Via these target genes, Hif-1α is involved in a myriad of cellular processes, including typical features favoring tumor growth such as angiogenesis or cell proliferation (44, 53). Interestingly, Hif-1α can also be stabilized under nonhypoxic conditions in response to various hormones (19), growth factors, and cytokines (23), and many of these stimuli induce ROS production as a part of their signaling cascade. Although the role of hypoxia in tumor growth is well established, conclusive data on the contribution of cellular ROS to hypoxic signaling are scarce and even conflicting. Previously, H₂O₂ was viewed as an inevitable and toxic by-product of aerobic life due to its potentially deleterious side effects. A number of studies, however, suggest that H₂O₂ plays an important role in cell signaling by regulating a diverse set of physiological processes (12). Previous reports demonstrated that Hif-1α stabilization and activation of its target genes can also be induced by exogenous H₂O₂ or other oxidative stressors (4, 8, 15, 34). Furthermore, mitochondrial ROS production has also been shown to stabilize Hif-1α (5, 20, 34), and Hif-1α DNA-binding activity appears to be sensitive to oxidizing reagents (51). On the other hand, short pre-exposure of cells to H₂O₂ selectively prevents hypoxia-induced Hif-1α binding by preventing the accumulation of Hif-1α protein, whereas treatment of hypoxic cell extracts with H₂O₂ has no effect on Hif-1α binding (24). It appears that ROS can interfere with the Hif-1α pathway at the level of PHDs by oxidizing the essential co-factor ferrous iron [Fe(II)] to the inactive ferric form [Fe(III)] (17, 38). However, to date, the precise link between ROS and hypoxic signaling has remained elusive.

To address whether the PrxIII/Txn2/Txnrd2 axis contributes to tumor growth/viability by maintaining redox homeostasis and controlling hypoxic signaling, we generated transformed cells with a targeted deficiency in the mitochondrial ROS-regulating enzyme Txnrd2.

In our experiments, we could show that loss of Txnrd2 in tumor cells results in a modified cellular redox balance. This imbalance leads to a sustained upregulation of PHD2 and failure to stabilize Hif-1α even in response to appropriate stimuli, ultimately culminating in the inhibition of tumor vascularization and growth. We additionally identified c-Jun NH2-terminal Kinase (JNK) as a hitherto unknown participant in ROS-mediated hypoxic signaling.

Results

Transformation and characterization of Txnrd2-deficient mouse embryonic fibroblasts

The gene targeting strategy for the disruption of Txnrd2 is depicted in Figure 1A (10). First, mouse embryonic fibroblasts (MEFs) were isolated from E12.5 embryos. Genotyping and immunoblotting was performed to confirm the deletion of Txnrd2 (Fig. 1B). To maintain H₂O₂ emission from the mitochondria at cell-tolerable levels, Txnrd2^−/− MEFs responded with a strong compensatory upregulation of the mitochondria-located oxidoreductase glutaredoxin-2 (grx2) (n=5, Fig. 1C). However, despite this compensatory response, Txnrd2^−/− cells still displayed higher ROS levels than wild-type (WT) cells (mean fluorescence intensity: 8.6±2.1 in Txnrd2^+/+ vs. 25.2±7.6 in Txnrd2^−/− cells, n=10, Fig. 1D).

FIG. 1.

Targeted loss of Txnrd2 and its effects on cellular redox balance. (A) Schematic outline of the generation of the conditional Txnrd2 knockout (KO) allele (10). (B) Expression of Txnrd2 in MEFs was analyzed on the genomic (top panels), mRNA (middle panels) as well as at the protein level (bottom panels). Equal loading was confirmed by Aldolase or β-Actin expression.(C) Immunoblot revealed that grx2 expression levels are strongly upregulated in Txnrd2^−/− MEFs. (D) Total cellular ROS levels were determined by flow cytometry using the fluorogenic probe CellROX Deep Red Reagent. Note that despite the enhanced levels of grx2 expression, Txnrd2-deficient MEFs accumulate more ROS than wild-type (WT) cells. (E) Equal expression of c-Myc and Ha-ras in transformed Txnrd2^+/+ and Txnrd2^−/− cells was confirmed by immunoblotting. (F) The proliferation rate of transformed Txnrd2^−/− MEFs (KO) is similar to that of WT cells. (G) Transformed Txnrd2^−/− cells are sensitive to cell death induced by GSH depletion in contrast to WT cells. Data in (D) are represented as mean±SD, data in (F, G) as mean±SEM. ***p<0.001. grx2, glutaredoxin-2; MEFs, mouse embryonic fibroblasts; ROS, reactive oxygen species; RT-PCR, real-time-polymerase chain reaction; SD, standard deviation.

Immortalized MEFs were transformed by lentiviral transduction of the proto-oncogene c-Myc and the oncogene Ha-ras^V12 and plated in soft agar (28). We established transformed cell lines from individually picked clones. This was deemed necessary as a basis for the in vivo tumor cell implantation studies. Equal expression of c-Myc and Ha-ras^V12 was confirmed by immunoblotting (Fig. 1E). Transformation of cells did not have any impact on cellular ROS levels (mean fluorescence intensity: 9.9±1.6 in Txnrd2^+/+ vs. 20.7±4.9 in Txnrd2^−/− , n=10, data not shown). However, we observed an alignment of in vitro proliferation of the transformed WT and knockout (KO) clones under baseline cell culture conditions (n=6, Fig. 1F).

Since the Prx III-dependent release of H₂O₂ from the mitochondria is not only under the control of the Trx2-Txnrd2 node but can also be altered by grx2 in a glutathione (GSH)-dependent manner (22), we tested how Txnrd2-deficient cells would react to an additional loss of GSH. Indeed, when Txnrd2 ^−/− cells were treated with L-buthionine sulfoximine (BSO), a highly specific and irreversible inhibitor of the GSH-synthesizing enzyme γ-glutamylcysteine synthetase, cell survival and proliferation were considerably impaired (n=6, Fig. 1G).

Targeted loss of Txnrd2 substantially impairs tumor growth in vivo

Having been characterized, transformed cells were subcutaneously implanted into the retral flank of C57BL/6 mice. Although single cell-derived transformed fibroblasts showed no difference in proliferation rates in vitro (n=6, Fig. 1F), the tumor mass of tumors formed by Txnrd2-deficient cells was substantially reduced by ∼50% (mean 0.8±0.16 g, median 0.6 g) after 11 days in contrast to tumors from Txnrd2-expressing cells (mean 1.7±0.34 g, median 1.57 g, n=25, Fig. 2A). Analysis of tumor growth at different time points indicated that the difference in the tumor mass was most pronounced on day 11 (day 2, n=3; day 3, n=9; day 4, n=3; day 6, n=9, day 8, n=3; day 11, n=25, Fig. 2B).

FIG. 2.

Loss of Txnrd2 results in a strong reduction in tumor growth. (A) Subcutaneous injection of single-cell-derived transformed cells into C57BL/6 mice revealed a substantial reduction of tumor growth in Txnrd2 null tumors. (B) A longitudinal determination of tumor mass indicated that tumor weight differences were most pronounced at day 11. (C, D) While the number of cells positive for cleaved caspase-3 (apoptotic cells) was not different among the groups at day 11 (C), the number of proliferating, Ki-67-positive cells was strongly reduced in Txnrd2 null tumors (D). ***p<0.001 Data are represented as mean±SD. n.s., not significant. Scale bar=100 μm.

Histological analysis was performed to elucidate the underlying reasons for the attenuated tumor growth. Quantification of cells positive for cleaved caspase-3, an apoptosis marker, revealed no difference (Txnrd2^+/+ 0.08±0.1 vs. Txnrd2^−/− 0.07±0.05 positive cells per region of interest, [Fig. 2C]). The assessment of necrotic areas within the tumors (n=6) also revealed that the smaller tumor volume was not a result of necrotic cell loss (percentage of the necrotic area relative to the entire tumor area: Txnrd2 ^+/+ tumors: 20%±6.5% vs. Txnrd2 ^−/− tumors: 14.5%±7.2%, data not shown). However, Txnrd2-deficient tumors immunostained for the proliferation marker Ki-67 showed a significantly lower ratio of proliferating to nonproliferating cells (proliferation index [PI]) than WT tumors (0.39±0.19 vs. 0.68±0.14, n=6 tumors per group, Fig. 2D).

Txnrd2^−/− tumors are highly susceptible to pharmacologic GSH deprivation

In our initial in vitro experiments, we could show that depletion of GSH by BSO resulted in the death of Txnrd2-deficient cells (Fig. 1G). Since both Txnrd2 and grx2 contribute to the reduction of Prx III (22), thereby controlling H₂O₂ emission from the mitochondria, we assessed whether inhibition of the de novo GSH synthesis may represent a pharmacological rationale to further reduce the growth of Txnrd2-deficient tumors in vivo. First, we confirmed that Txnrd2-deficient tumors also expressed significantly higher levels of grx2 than WT tumors (n=3 tumors per group, Supplementary Fig. 1A; Supplementary Data are available online at www.liebertpub.com/ars). Subsequently, BSO was administered via the drinking water, which proved highly efficient. Glutathione levels dropped sharply, as shown by high-performance liquid chromatography (HPLC) analysis (Fig. 3A). Reduced glutathione (GSH) (0.0007±0.0002 μmol/μg protein) and oxidized glutathione (GSSG) levels (0.0008±0.0003 μmol/μg protein) were considerably diminished in Txnrd2-deficient tumors compared with untreated Txnrd2 ^−/− tumors when BSO was administered via the drinking water (GSH 0.015±0.003; GSSG 0.003±0.001 μmol/μg protein) (n=8 tumors per group, Fig. 3A, B). The decrease of the GSH/GSSG ratio in BSO-treated Txnrd2 null tumors indicates elevated levels of oxidative stress (Fig. 3C, n=8 tumors per group). Most importantly, manipulation of in vivo glutathione levels by BSO caused a further reduction of Txnrd2-deficient tumor mass by ∼38% (0.52±0.14 g) compared with untreated Txnrd2^−/− tumors (0.84±0.27 g, n=8, Fig. 3D).

FIG. 3.

Loss of Txnrd2 renders tumors susceptible to GSH depletion. (A) Pharmacological glutathione depletion by BSO was confirmed by HLPC in tumor samples. (B) BSO treatment efficiently depleted tumors from glutathione as determined by measuring the levels of GSSG by HLPC. (C) The decrease of the GSH/GSSG ratio in BSO-treated Txnrd2 null tumors indicates elevated levels of oxidative stress. (D) BSO treatment of tumor-bearing mice via the drinking water caused a further reduction in growth of Txnrd2 null tumors. Data in (A, B) are represented as mean±SD, data in (C, D) as mean±SEM. *p<0.05, ***p<0.001. BSO, L-buthionine sulfoximine.

Loss of Txnrd2 delays the angiogenic switch and impairs tumor angiogenesis

The tumor-surrounding vascular network of Txnrd2-deficient tumors was less prominent compared with WT tumors (Fig. 4A). In line, quantification of tumor vessels demonstrated significantly fewer vessels per Txnrd2-deficient tumor (143±13.9) compared with WT tumors (239±30.43, n=6, Fig. 4B).

FIG. 4.

Loss of Txnrd2 delays the angiogenic switch and reduces tumor vascularization. (A) The tumor-surrounding vascular network was less prominent in the Txnrd2 null tumors at 3 days after tumor inoculation. Dotted circles outline the tumor tissue. (B) Quantification of the vessels by CD31 staining confirmed the morphological finding illustrated in (A). (C) Analysis of key-angiogenic factors revealed decreased Hif-1α as well as VEGF-A protein expression in Txnrd2 null tumor tissues at days 3 and 6 of tumor growth. *p<0.05, **p<0.01, ***p<0.001. Left panel: representative immunoblots from day 3; middle and right panels: semi-quantitative assessment of immunoblots. (D) Transformed Txnrd2^−/− cells failed to stabilize Hif-1α protein after serum starvation. Left panel: representative immunoblot; right panel: semi-quantitative assessment of immunoblots. All data are represented as mean±SD. Hif-1α, hypoxia-inducible factor-1α; VEGF-A, vascular endothelial growth factor A.

Reduced expression of key angiogenic players in tumors derived from Txnrd2^−/− cells

The delayed angiogenic switch and the diminished vascularization could be attributed to a substantial reduction of Hif-1α protein levels in Txnrd2-deficient tumors (day 3: 38%±2%; day 6: 41%±7%) compared with tumors expressing Txnrd2 (day 3, day 6: arbitrarily set to 100%) at both time points (n=4 per time point, Fig. 4C). In accordance, vascular endothelial growth factor A (VEGF-A) levels were also significantly reduced at day 3 (18%±2%) in these tumors compared with Txnrd2 WT tumors (arbitrarily set to 100%, Fig. 4C). This phenotype was still present on day 6 of tumor growth (Txnrd2^+/+ : arbitrarily set to 100% vs. Txnrd2^−/− : 59%±2%). An ELISA specific for mouse VEGF confirmed that Txnrd2-deficient tumors contained less VEGF-A (n=4 per time point, day 3: 0.19±0.06 pg/μg protein; day 6: 0.09±0.05 pg/μg protein) than WT tumors (day 3: 0.48±0.24 pg/μg protein day 6: 0.12±0.05 pg/μg protein) (data not shown).

Hif-1α stabilization in response to serum withdrawal and hypoxia is impaired

Under normal cell culture conditions, Hif-1α expression was barely detectable in either cell line (Txnrd2^+/+ : arbitrarily set to 100% vs. Txnrd2^−/− : 50%±16.7%, n=6, Fig. 4D). However, Hif-1α expression substantially increased in WT cells but not in Txnrd2-deficient cells after serum withdrawal (Txnrd2^+/+ : 4830%±2640% vs. Txnrd2^−/− : 70%±17.3%) (n=6, Fig. 4D). A similar effect was also observed under hypoxic conditions (n=4, Supplementary Fig. S1C).

To investigate whether Hif-1α is regulated at the transcriptional and/or the translational level, Hif-1α mRNA expression levels were assessed by quantitative real-time-polymerase chain reaction (qRT-PCR). Under normal cell culture conditions, Hif-1α mRNA levels were comparable (Txnrd2^+/+ : arbitrarily set to 100% vs. Txnrd2^−/− : 110%±26%, n=5, data not shown). Starvation resulted in a moderate increase in WT cells, whereas Txnrd2-deficient cells failed to upregulate Hif-1α mRNA levels (Txnrd2^+/+ : 190%±78% vs. Txnrd2^−/− : 90%±35%). These data indicate that stabilization of the majority of Hif-1α occurs at the protein level since the small change of Hif-1α transcripts was unlikely to account for the increased levels of Hif-1α protein in WT cells.

Increased PHD2 expression promotes Hif-1α protein degradation in Txnrd2^−/− cells and tumors

Next, we examined some upstream regulators of Hif-1α protein stability, including HIF prolyl hydroxylases. Interestingly, even under baseline conditions, PHD2 protein levels were elevated in Txnrd2^−/− cells compared with control cells (Txnrd2^+/+ : arbitrarily set to 100% vs. Txnrd2^−/− : 322.2%±99.8%, n=4, Fig. 5A). Vice versa, knockdown of PHD2 using siRNA increased Hif-1α protein levels in both WT and Txnrd2-deficient MEFs, thereby substantiating that PHD2 is the main regulator of Hif-1α protein stability (n=3, Supplementary Fig. S1D).

FIG. 5.

PHD2 protein levels are highly increased in Txnrd2 null cells. (A) PHD2 protein levels are strongly increased in Txnrd2^−/− cells, both under normal cell culture conditions and in response to starvation. Data are represented as mean±SEM. (B) Nuclear translocation of the p65 subunit of NFκB is diminished in Txnrd2^−/− cells as shown by immunocytochemistry. (C) An NFκB-specific transcription factor assay confirmed the data shown in (B). *p<0.05, **p<0.01, ***p<0.001. Data are represented as mean±SD.

Serum starvation had no additional impact on PHD2 protein expression in either of the cell lines (n=4, Fig. 5A). Consistent with these in vitro findings, Txnrd2 null tumors also displayed augmented PHD2 expression (n=3, Supplementary Fig. S1B).

To examine whether the mRNA levels were also increased, we performed qRT-PCR. Indeed, Txnrd2^−/− cells showed statistically significant enhanced levels of PHD2 mRNA (n=11, WT arbitrarily set to 100% vs. KO 242%±55%, data not shown).

Apart from affecting Hif-1α stabilization, PHD2 is also a well-known regulator of nuclear factor-κB (NF-κB) activity (11). Indeed, immunocytochemistry revealed reduced nuclear translocation of the p65 NF-κB subunit in Txnrd2 null cells (Fig. 5B). Quantification of nuclear translocation of p65 confirmed this observation (n=6, Fig. 5C).

Enhanced JNK phosphorylation leads to increased PHD2 protein levels

Intrigued by the findings that Txnrd2 deletion impairs tumor growth via an increase in PHD2 expression, we endeavored to unmask a previously unknown link between Txnrd2 and PHD2. Since JNK is known to be regulated by the cellular redox status (27), we investigated its phosphorylation status. Indeed, JNK phosphorylation was increased in immortalized Txnrd2^−/− cells in contrast to their WT counterparts (n=5, Fig. 6A) and this effect could be attributed to elevated ROS as the phosphorylation status could be reduced by addition of polyethylene glycol-conjugated catalase (PEG-catalase) (n=4, Fig. 6B). Substantiating our hypothesis that ROS-induced, sustained JNK activation leads to elevated PHD2 expression, PEG-catalase treatment also blunted PHD2 expression (n=4, Fig. 6B). To provide a molecular link between mitochondrial ROS and increased JNK-phosphorylation, we measured PTP activity. ROS have been implicated in this pathway through inhibition of the counteracting JNK phosphatases (27), and measurements of protein tyrosine phosphatase (PTP) activity, indeed, demonstrated a reduced activity in the Txnrd2-deficient cells (n=7, Fig. 6F).

FIG. 6.

Deletion of Txnrd2 increases PHD2 expression via sustained JNK phosphorylation. (A) Representative immunoblot showing the substantially augmented presence of phosphorylated JNK (pJNK) in Txnrd2 null MEFs. (B) PEG-catalase treatment of Txnrd2^−/− cells not only lowered the phosphorylation of JNK but also reduced the overall PHD2 protein levels. (C) Knockdown of Txnrd2 in the bona fide tumor cell line LLC1 led to a subsequent increase of PHD2 and pJNK protein levels. (D) The JNK inhibitor SP600125 reduced phosphorylated JNK as well as PHD2 protein levels in both transformed Txnrd2^−/− cells and the Txnrd2-deficient tumor cell line, LLC1. (E) siRNA-mediated knockdown of JNK in Txnrd2 ^−/− MEFs also led to a reduction in PHD2 expression levels. (F) PTP-activity is significantly reduced in Txnrd2^−/− cells compared with WT cells. *p<0.05. JNK, c-Jun NH2-terminal Kinase; MEFs, mouse embryonic fibroblasts; PEG-catalase, polyethylene glycol-conjugated catalase; PTP, protein tyrosine phosphatase.

To investigate whether the Txnrd2-dependent PHD2 regulation was a general mechanism and not a cell type-specific effect, we manipulated Txnrd2 expression levels in the murine lung tumor cell line LLC1. Using endoribonuclease-prepared small interfering RNA (esiRNA), we were able to decrease Txnrd2 protein levels by ∼90% compared with scrambled esiRNA-treated cells (n=4, Fig. 6C). This knockdown led to a subsequent increase in pJNK and PHD2 levels in Txnrd2-deficient LLC1 cells. Furthermore, treating Txnrd2 KO MEFs and LLC1 cells with the JNK inhibitor SP600125 caused both a decrease of basal pJNK levels and a strong decrease in PHD2 expression, which is in accordance with the proposed mechanism (n=5, Fig. 6D). These data demonstrate that this Txnrd2-pJNK-PHD2 axis is also a relevant pathway for bona fide tumor cells. Finally, knockdown of JNK by siRNA significantly attenuated PHD2 expression (control siRNA arbitrarily set to 100%, siRNA: JNK 22.6%±3.7%, PHD2 53.2%±6.3%, n=13, exemplarily shown in Fig. 6E), further substantiating our findings that JNK acts as an upstream participant in this pathway.

Expression of Txnrd2 in human cancer tissue

To demonstrate that our findings are also of relevance for human tumors, the expression of Txnrd2 was studied by immunohistochemistry in human cancer tissue derived from three different types of tumors: colon carcinoma (n=10), hepatocellular carcinoma (n=8), as well as lung adenocarcinoma (n=8). In colon carcinomas as well as in hepatocellular carcinomas, we could observe an abundant expression of Txnrd2. Lung adenocarcinomas also expressed Txnrd2, however to a much lesser extent than in the colon and liver samples (Supplementary Fig. S2). These observations are in line with previously published reports (9, 49) as well as with the human protein atlas (48).

Discussion

Elevated levels of ROS have been detected in almost all cancers, where they promote many aspects of tumor development and progression, including DNA damage, activation of inflammatory pathways, receptor tyrosine kinase signaling, and stabilization of Hif-1α (30). On the other hand, several oncogenes were recently discovered to act as active inducers of NF-E2-related factor-2 (Nrf2), which, in turn, promotes an ROS detoxification program required for tumor initiation (13). Moreover, most tumor cells also express increased levels of antioxidants to detoxify ROS, suggesting that a delicately balanced elevation of intracellular ROS levels is required for cancer cell function (30, 40).

Since the PrxIII-Trx2-Txnrd2 node is considered the most important route for mitochondrial H₂O₂ removal (14, 45) and we observed an abundant expression of Txnrd2 in a variety of human tumors (Supplementary Fig. S2), we employed genetically engineered Txnrd2-deficient cells to study the effects of an altered mitochondrial redox balance on tumor signaling and growth. Indeed, our model enabled us to unveil how a chronic change of this balance affects Hif signaling and tumor growth via a previously unknown pathway. The cornerstones of this pathway encompass a ROS-dependent activation of JNK after the loss of Txnrd2, elevated PHD2 protein and mRNA expression, a failure to stabilize Hif-1α in response to hypoxia and other stressors, and decreased VEGF-A levels. In vivo we could observe a delayed angiogenic switch, attenuated tumor vascularization, and, ultimately, reduced growth of tumors lacking Txnrd2.

Unexpectedly and in contrast to previous results (5, 8, 17), loss of Txnrd2 followed by a compensatory increase in grx2 expression, as well as moderately increased ROS levels, did not stabilize Hif-1α. Numerous studies have demonstrated that addition of exogenous ROS leads to the stabilization of Hif-1α protein and activation of Hif target genes such as VEGF-A (4, 8, 15). In accordance, tumor cell-derived ROS can regulate angiogenesis and tumor growth through VEGF (52) and exposure to antioxidants has been demonstrated to delay tumorigenesis by reducing HIF levels (16). Given the previous reports, we expected that the Txnrd2 deletion and coincident further elevation of ROS levels would lead to a stabilization of the Hif-1α protein through PHD2 inactivation; for instance, by oxidizing PHD-bound Fe(II) to Fe(III) (17, 38), and/or the cysteine residue 201 within the catalytic domain of PHD2, thereby further inhibiting PHD2 activity (35, 37). In this study, we observed quite the contrary: Hif-1α protein levels were reduced not only in Txnrd2^−/− tumors but also in Txnrd2^−/− MEFs, particularly following certain stressors such as hypoxia or serum starvation. The diminished Hif-1α levels could be attributed to high PHD2 expression, which leads to hydroxylation of key proline residues of Hif-1α and subsequent targeting of Hif-1α to the proteasome (25, 26). Considering the existing reports in the literature, these findings indicate that, apparently, PHD2 is differentially regulated upon a ROS challenge. In contrast to an acute and possibly nonphysiological, exogenous bolus administration of ROS that inhibits PHD2 activity, a compensatory expression of the mitochondria-located oxidoreductase grx2 is induced in Txnrd2-deficient cells, thereby keeping the emission of ROS from the mitochondria in a cell-tolerable, however moderately elevated range. This continuous and modest elevation of endogenous ROS in Txnrd2-deficient cells appears to result in an adaptive upregulation of PHD2 protein levels, possibly as a protective measure against chronically increased Hif-1α expression. A similar adaptive mechanism has also been observed during chronic hypoxia. Acute hypoxia impairs PHD2 activity, whereas chronic hypoxia inhibits mitochondrial respiration, thus restoring oxygen availability and leading to PHD2 over-activation (18).

In pursuit of a molecular link between mitochondrial ROS and increased PHD2 expression, we investigated the redox-sensitive JNK signaling pathway. ROS-induced JNK phosphorylation via Mitogen-activated protein kinase kinase 4 (MKK4) can be initiated by the liberation of the apoptosis signal-regulating kinase 1 (ASK-1) from its endogenous suppressor, reduced thioredoxin 1 (42). Furthermore, ROS have been implicated in this pathway through inhibition of JNK phosphatases (27) and we could indeed measure a reduced activity of PTPs in the Tnxrd2-deficient cells. Furthermore, we could show that JNK phosphorylation was markedly increased in Txnrd2^−/− cells as well as in the established mouse cancer cell line, LLC1, when Txnrd2 was knocked down. Administration of the JNK inhibitor SP600125 and knockdown of JNK by siRNA substantiated our hypothesized mechanism, as this led to reduced PHD2 protein levels, thus confirming the yet-unrecognized link between JNK and PHD2 expression. Currently, it remains to be shown how exactly PHD2 levels are regulated by activated JNK.

Taken together, our in vitro observations serve as a plausible foundation for the restricted growth of Txnrd2-null tumors. First, JNK signaling has been shown to suppress tumor growth (33, 50). Second, tumor angiogenesis is, at least partly, regulated by PHD2 (7), and Txnrd2-null tumors express reduced Hif-1α as well as reduced VEGF-A protein levels. Finally, PHD2 has additional Hif-independent tumor suppressor functions (7), which are also in agreement with the reduced tumor growth observed in our model.

We previously showed that primary Txnrd2^−/− fibroblasts are highly sensitive to GSH depletion-induced cell death (10). As demonstrated here, this apparently also holds true for immortalized cells and tumors derived thereof, most likely due to a block of the compensatory GSH-dependent Prx III reduction by grx2. For future studies targeting antioxidants to efficiently combat cancer, we therefore suggest the concomitant targeting of at least two key antioxidant systems.

In conclusion, our data describe a hitherto unrecognized pathway for PHD2 activation via JNK, originating from an altered mitochondrial redox balance. Phosphorylated JNK leads to a sustained expression of PHD2 in a yet unknown manner, ultimately leading to the impairment of Hif-1α signaling. While our data oppose previous results on PHD2 inhibition by ROS, this should be rationalized on the basis of variations in the models regarding radicals, sources of ROS, as well as their concentrations and the duration in which they are present.

Although our study only scratches the surface regarding the relevance of Txnrd2 for human cancer biology, we hope it will motivate other groups to study ROS-controlling systems in this context. Finally, our findings highlight the need for future anti-cancer strategies to consider the fine tuning of intracellular ROS signaling to efficiently shift cells from ROS-dependent tumor growth-promoting events toward ROS-induced cell death signaling.

Materials and Methods

Cell lines and reagents

Txnrd2^+/+ and Txnrd2^−/− MEFs were isolated from embryos at embryonic day E12.5 from breeding of heterozygous Txnrd2 mice. Genotyping and cell culturing was performed as previously described (10). Primer sequences are listed in Supplementary Tables S1 and S2. Primary MEFs (passage number<10) were immortalized by serial passaging (passage number>10) as described (39). LLC1 cells were obtained from ATCC (Teddington, United Kingdom). To induce Hif-1α protein expression, cells were either cultured in serum-free medium for 4 h or kept in a hypoxic chamber (1% O₂) for 6 h.

Generation of transformed Txnrd2^+/+ and Txnrd2^−/− cells

For the transformation of cells, MEFs were co-transduced with c-Myc and Ha-ras^V12-expressing lentiviruses as described (31). Soft agarose assays were performed to select successfully transformed cells. After 7 days of growth in soft agarose, single-cell-derived colonies were isolated, expanded, and analyzed for expression of Txnrd2, c-Myc, and Ha-ras^V12. To assess the efficacy of transformation with c-Myc and Ha-ras^V12, cells were analyzed for green fluorescent protein (GFP) expression using BD FACSort (Beckton Dickinson, Heidelberg, Germany). Data were analyzed using the CellQuest software (Beckton Dickinson) and WinMDI software, version 2.9 (Scripps Research Institute, La Jolla, CA).

Proliferation assay

Cell proliferation and cytotoxicity were measured using the electronic cell sensor array technology xCELLigence (Roche, Mannheim, Germany) as described (32). BSO (Sigma-Aldrich, Munich, Germany) was used at a final concentration of 10 μM to inhibit the de novo synthesis of GSH.

Measurement of cellular ROS levels

ROS generation was assessed using CellROX Deep Red Reagent (Life Technologies, Darmstadt, Germany) according to the manufacturer's recommendations. Fluorescence intensity was analyzed by flow cytometry using a Gallios 2/8 flow cytometer (Becton Coulter, Krefeld, Germany) at an excitation wavelength of 633 nm and emission collected with a 660 band-pass detector. Cells were gated and analyzed using the Gallios Data acquisition and analysis software (Becton Coulter).

Measurement of GSH concentration in cells and tumors

Total glutathione content in cells was analyzed by performing a modified assay based on a method first described by Tietze (46). The concentration of total glutathione in cells was calculated as μM/mg protein and referenced to a 10 mM GSH calibration solution (Sigma-Aldrich). Reduced and oxidized glutathione concentrations (GSH and GSSG) in tumor samples were measured by isocratic HPLC as previously described (31).

Quantitative real-time PCR

The expression of Hif-1α, PHD2, and Txnrd2 was investigated by RT-PCR/qRT-PCR as described (31). Expression was normalized to β-actin or aldolase expression. Primer sequences are listed in Supplementary Table S2.

Hematoxylin-eosin staining/immunohistochemistry/immunofluorescence and image analysis

Paraformaldehyde-fixed (4% w/v, in phosphate-buffered solution [PBS]) and paraffin-embedded tumor material was cut in 5 μm-thick sections and stained as described (10). Hematoxylin-eosin (H&E) staining, immunohistochemistry, and immunofluorescence staining were performed as previously described (2). Antibodies are listed in Supplementary Table S3. As control, primary antibodies were omitted during staining. Slides for peroxidase staining were incubated with peroxidase-conjugated streptavidin (Vectastain KIT ABC; Vector Laboratories, Linaris, Wertheim-Bettingen, Germany). Thereafter, slides were incubated with Vector^® DAB kit or AEC kit (Vector Laboratories). Sections were counterstained with hematoxylin. Sections for immunofluorescence were counterstained with 4,6-diamidino-2-phenylindole ([DAPI], Hard Set™ Mounting Medium; Vector Laboratories). Immunostaining was analyzed using the Olympus BX41 microscope in combination with the CAMEDIA C-5050 digital camera and the software Olympus DP-Soft v3.2 (Olympus, Tokio, Japan). For the quantification of tumor vascularization, tumor cell proliferation, tumor cell apoptosis, and necrosis, three representative sections per tumor were analyzed (n=6 tumors) and micrographs were obtained from five random regions of interest from all three sections. Altogether, the analysis of these 15 random microscopic visual fields per tumor was performed in a blinded fashion by two researchers using a 20× objective. Total blood vessels were determined by counting the number of CD31-positive blood vessels. Proliferation was studied by counting Ki-67 positive and -negative cells. The ratio of proliferative to nonproliferative cells was expressed as PI. Apoptosis was assessed by counting caspase-3 positive cells. For the determination of necrosis, the percentage of the necrotic area relative to the entire tumor area was measured on H&E stained tumor sections.

Immunoblotting

Immunoblotting was performed as previously described (1). Primary antibodies are listed in Supplementary Table S4. Detection of Txnrd2 expression was achieved with a Txnrd2-specific antibody as described (31). Nuclear translocation of Hif-1α in cells was analyzed with the NE-PER^® Nuclear and Cytoplasmic Extraction Reagent (Fisher Scientific GmbH, Schwerte, Germany). Equal loading was confirmed with a lamin-specific antibody. To block H₂O₂, cells were treated with PEG-catalase (50 U/ml; Sigma-Aldrich) for 18 h.

To analyze the levels of phosphorylated JNK, samples were lysed with buffer containing sodium fluoride (0.5 mM) and Na₃VO₄ (0.5 mM) (Sigma Aldrich). All blots were stripped and reprobed with either β-actin or glyceraldehyde 3-phosphate dehydrogenase (GAPDH). To block JNK activation, the specific JNK-inhibitor SP600125 (#420119, 50 μM; Merck Millipore, Schwalbach, Germany) was applied to the cell culture medium, either for 8 h or overnight. Application of dimethyl sulfoxide (DMSO) served as a control.

PHD2 knockdown in Txnrd2^+/+ and Txnrd2^−/− MEFS

MEFs were transfected with siMax siRNA (Eurofins Genomics, Ebersberg, Germany) using the following sequences: PHD2 siRNA 5′-GUG GAG GUA UUC UUC GAA UTT-3′; scramble siRNA 5′-UUC UUC GAA CGU GUC ACG UTT-3′. 10⁵ cells were seeded onto 12-well plates and incubated with 5 nM siRNA and HiPerfect transfection reagent (Qiagen, Hilden, Germany) in full medium for 50 h. Subsequently, cells were washed in PBS and proteins were extracted by cell lysis. Knockdown of PHD2 was confirmed by immunoblot.

Txnrd2 knockdown in LLC1 tumor cells

LLC1 cells were transfected with MISSION^® esiRNA targeting mouse Txnrd2 (EMU067221; Sigma-Aldrich) according to the manufacturer's instructions. Cells were harvested, lysed for 72 h after transfection, and blotted against designated antibodies as described earlier.

JNK knockdown in Txnrd2 ^−/− MEFs

MEFs were transfected with FlexiTube siRNA (Qiagen) against both JNK1 (GS26419) and JNK2 (GS26420). Reverse transfection was performed in 24-well plates according to the manufacturer's advice using 5 μl HiPerFect transfection reagent and 50 nM siRNA for 70,000 cells for 72 h. Subsequently, cells were lysed and knockdown of JNK was confirmed by immunoblot.

Enzyme-linked immunosorbent assay

To measure mouse VEGF in cell culture supernatants and tumor tissue, the Mouse VEGF Quantikine Immunoassay (R&D Systems, Wiesbaden-Nordenstedt, Germany) was employed according to the manufacturer's recommendations. Absorbance was measured at a wavelength of 450 nm using the ELISA plate reader Infinite F200 (Tecan, Crailsheim, Germany). Total protein concentration in supernatants and tumor homogenates was measured with the BCA™ Protein Assay Kit (Pierce; Fisher Scientific).

Subcutaneous xenograft tumor model

Transformed Txnrd2^−/− and Txnrd2^+/+ cells (4×10⁶ cells dissolved in 200 μl sterile PBS) were injected subcutaneously into the retral flank of C57BL/6 mice. Tumors were collected over a period of 11 days. All animal experiments were performed in compliance with the German Animal Welfare Law and had been approved by the institutional committee on animal experimentation and the government of Upper Bavaria.

Treatment of tumor-bearing mice with BSO

For the in vivo pharmacological studies, 4×10⁶ transformed Txnrd2-deficient cells in 200 μl sterile PBS were injected subcutaneously into the retral flank of C57BL/6 mice. Tumors were allowed to settle for 3 days. From day 3 onward, BSO (20 mM) was administered via the drinking water for 8 days. BSO-containing water was exchanged every third day. At day 11, mice were sacrificed and tumor mass and volume was analyzed.

Measurement of PTP activity

PTP activity in total cell lysates was measured toward a ³²P-labeled phosphotyrosine substrate (AEEEIpYGEFEA KKKK) as previously described (43). WT and Txnrd2^−/− fibroblasts were starved overnight in phenol-red free DMEM (Gibco, Stockholm, Sweden) with 1% fetal calf serum. Cells were washed in 20 mM Hepes pH 7.4 (Sigma-Aldrich, Stockholm, Sweden) and lysed in degassed lysis buffer (50 mM sodium acetate, 150 mM NaCl, 1% NP-40, 25 μg/ml aprotinin, 25 μg/ml leupeptin, 250 U/ml catalase, and 125 U/ml superoxide dismutase [Calbiochem, Solna, Sweden]). Lysates were cleared by centrifugation at 21,000 g at 4°C for 5 min. A fraction of the lysate was mixed with deoxygenated 25 mM imidazole pH 7.4 (Sigma-Aldrich), the PTP substrate (kind gift from Carl Henrik Heldin) was added, and the mixture was incubated for 20 min. The dephosphorylation reaction was stopped by addition of a charcoal mixture (0.9M HCl, 90 mM sodium pyrophosphate, 2 mM NaH₂PO₄, 4% vol/vol NoritA; Sigma-Aldrich) and centrifuged at 21,000 g for 3 min. The supernatant containing released ³²P (PerkinElmer, Upplands Väsby, Sweden) was transferred to scintillation vials (Sigma-Aldrich) with EcoScint A (National Diagnostics, BioNordika, Stockholm, Sweden), and radioactivity was measured using a Wallac Winspectral 1414 Liquid Scintillation Counter (PerkinElmer). The radioactive signal reflects the activity of the reduced PTPs and was normalized toward cell lysate treated with the reducing agent dithiothreitol (DTT; Sigma-Aldrich) that represents the maximal activity when all PTPs are reduced.

Determination of NFκB activation

Translocation of the NFκB subunit p65 to the nucleus was analyzed by immunocytochemistry in formaldehyde-fixed MEFs as described (3). In brief, cells were grown to ∼80% confluence on glass coverslips coated with collagen G (10 μg/ml; Biochrom AG, Berlin, Germany). Subsequently, cells were fixed with 3.7% formaldehyde, permeabilized with 0.3% Triton X-100, and incubated with a p65-specific antibody. Slides were mounted with DAPI-containing Hard Set Mounting Medium (Vector Laboratories) and analyzed using the confocal laser scanning microscope Leica TCS SP5 (Leica Microsystems, Wetzlar, Germany).

To further substantiate the results from immunocytochemistry, activation of the NFκB subunit p65 was measured using the colorimetric NFκB p65 EZ-TFA transcription factor assay according to the manufacturer's instructions (Millipore GmbH, Schwalbach, Germany).

Statistical analysis

Statistical analysis was performed using SigmaPlot^© 11.0 software (Jandel GmbH, Erkrath, Germany). Experimental values are expressed as mean±standard deviation, unless stated otherwise. Numbers of independent experiments are indicated in the corresponding results section. Statistically significant differences between groups were calculated by Student's t-test or analysis of variance followed by Bonferroni's correction. Non-Gaussian distributed data were analyzed by the nonparametric Kruskal–Wallis test for nonpaired data. p<0.05 was considered significant.

Footnotes

Acknowledgments

The authors thank Dora Kiesl, Dorothee Gössel, Heidi Förster, Kasia Stefanowski, and Matthias Semisch for their excellent technical assistance.

This work was supported by the Deutsche Forschungsgemeinschaft (DFG) priority Program SPP 1190 to H.B. and M.C., the Friedrich-Baur Stiftung to H.B. and T.F., a DFG grant (CO 291/2-3) to M.C., a Marie Curie Actions Fellowship of the FP7 People Program to J.K., and a fellowship from Alexander von Humboldt-Stiftung to J.P.F.A.

Author Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Arner

and Holmgren

. The thioredoxin system in cancer-introduction to a thematic volume of Seminars in Cancer Biology. Semin Cancer Biol, 16: 419, 2006.

Beck

, Acker

, Wiessner

, Allegrini

, and Plate

. Expression of angiopoietin-1, angiopoietin-2, and tie receptors after middle cerebral artery occlusion in the rat. Am J Pathol, 157: 1473–1483, 2000.

Belaiba

, Bonello

, Zahringer

, Schmidt

, Hess

, Kietzmann

, and Gorlach

. Hypoxia up-regulates hypoxia-inducible factor-1alpha transcription by involving phosphatidylinositol 3-kinase and nuclear factor kappaB in pulmonary artery smooth muscle cells. Mol Biol Cell, 18: 4691–4697, 2007.

Brauchle

, Funk

, Kind

, and Werner

. Ultraviolet B and H2O2 are potent inducers of vascular endothelial growth factor expression in cultured keratinocytes. J Biol Chem, 271: 21793–21797, 1996.

Brunelle

, Bell

, Quesada

, Vercauteren

, Tiranti

, Zeviani

, Scarpulla

, and Chandel

. Oxygen sensing requires mitochondrial ROS but not oxidative phosphorylation. Cell Metab, 1: 409–414, 2005.

Cairns

, Harris

, and Mak

. Regulation of cancer cell metabolism. Nat Rev Cancer, 11: 85–95, 2011.

Chan

, Kawahara

, Sutphin

, Chang

, Chi

, and Giaccia

. Tumor vasculature is regulated by PHD2-mediated angiogenesis and bone marrow-derived cell recruitment. Cancer Cell, 15: 527–538, 2009.

Chandel

, McClintock

, Feliciano

, Wood

, Melendez

, Rodriguez

, and Schumacker

. Reactive oxygen species generated at mitochondrial complex III stabilize hypoxia-inducible factor-1alpha during hypoxia: a mechanism of O2 sensing. J Biol Chem, 275: 25130–25138, 2000.

Choi

, Kim

, Baek

, Kim

, Lee

, and Kim

. Overexpression of mitochondrial thioredoxin reductase and peroxiredoxin III in hepatocellular carcinomas. Anticancer Res, 22: 3331–3335, 2002.

10.

Conrad

, Jakupoglu

, Moreno

, Lippl

, Banjac

, Schneider

, Beck

, Hatzopoulos

, Just

, Sinowatz

, Schmahl

, Chien

, Wurst

, Bornkamm

, and Brielmeier

. Essential role for mitochondrial thioredoxin reductase in hematopoiesis, heart development, and heart function. Mol Cell Biol, 24: 9414–9423, 2004.

11.

Cummins

, Berra

, Comerford

, Ginouves

, Fitzgerald

, Seeballuck

, Godson

, Nielsen

, Moynagh

, Pouyssegur

, and Taylor

. Prolyl hydroxylase-1 negatively regulates IkappaB kinase-beta, giving insight into hypoxia-induced NFkappaB activity. Proc Natl Acad Sci U S A, 103: 18154–18159, 2006.

12.

D'Autreaux

and Toledano

. ROS as signalling molecules: mechanisms that generate specificity in ROS homeostasis. Nat Rev Mol Cell Biol, 8: 813–824, 2007.

13.

DeNicola

, Karreth

, Humpton

, Gopinathan

, Wei

, Frese

, Mangal

, Yu

, Yeo

, Calhoun

, Scrimieri

, Winter

, Hruban

, Iacobuzio-Donahue

, Kern

, Blair

, and Tuveson

. Oncogene-induced Nrf2 transcription promotes ROS detoxification and tumorigenesis. Nature, 475: 106–109, 2011.

14.

Drechsel

and Patel

. Respiration-dependent H2O2 removal in brain mitochondria via the thioredoxin/peroxiredoxin system. J Biol Chem, 285: 27850–27858, 2010.

15.

Duyndam

, Hulscher

, Fontijn

, Pinedo

, and Boven

. Induction of vascular endothelial growth factor expression and hypoxia-inducible factor 1alpha protein by the oxidative stressor arsenite. J Biol Chem, 276: 48066–48076, 2001.

16.

Gao

, Zhang

, Dinavahi

, Li

, Xiang

, Raman

, Bhujwalla

, Felsher

, Cheng

, Pevsner

, Lee

, Semenza

, and Dang

. HIF-dependent antitumorigenic effect of antioxidants in vivo . Cancer Cell, 12: 230–238, 2007.

17.

Gerald

, Berra

, Frapart

, Chan

, Giaccia

, Mansuy

, Pouyssegur

, Yaniv

, and Mechta-Grigoriou

. JunD reduces tumor angiogenesis by protecting cells from oxidative stress. Cell, 118: 781–794, 2004.

18.

Ginouves

, Ilc

, Macias

, Pouyssegur

, and Berra

. PHDs overactivation during chronic hypoxia “desensitizes” HIFalpha and protects cells from necrosis. Proc Natl Acad Sci U S A, 105: 4745–4750, 2008.

19.

Gorlach

, Diebold

, Schini-Kerth

, Berchner-Pfannschmidt

, Roth

, Brandes

, Kietzmann

, and Busse

. Thrombin activates the hypoxia-inducible factor-1 signaling pathway in vascular smooth muscle cells: role of the p22(phox)-containing NADPH oxidase. Circ Res, 89: 47–54, 2001.

20.

Guzy

, Hoyos

, Robin

, Chen

, Liu

, Mansfield

, Simon

, Hammerling

, and Schumacker

. Mitochondrial complex III is required for hypoxia-induced ROS production and cellular oxygen sensing. Cell Metab, 1: 401–408, 2005.

21.

Hanahan

and Weinberg

. Hallmarks of cancer: the next generation. Cell, 144: 646–674, 2011.

22.

Hanschmann

, Lonn

, Schutte

, Funke

, Godoy

, Eitner

, Hudemann

, and Lillig

. Both thioredoxin 2 and glutaredoxin 2 contribute to the reduction of the mitochondrial 2-Cys peroxiredoxin Prx3. J Biol Chem, 285: 40699–40705, 2010.

23.

Hellwig-Burgel

, Rutkowski

, Metzen

, Fandrey

, and Jelkmann

. Interleukin-1beta and tumor necrosis factor-alpha stimulate DNA binding of hypoxia-inducible factor-1. Blood, 94: 1561–1567, 1999.

24.

Huang

, Arany

, Livingston

, and Bunn

. Activation of hypoxia-inducible transcription factor depends primarily upon redox-sensitive stabilization of its alpha subunit. J Biol Chem, 271: 32253–32259, 1996.

25.

Ivan

, Kondo

, Yang

, Kim

, Valiando

, Ohh

, Salic

, Asara

, Lane

, and Kaelin

Jr . HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science, 292: 464–468, 2001.

26.

Jaakkola

, Mole

, Tian

, Wilson

, Gielbert

, Gaskell

, Kriegsheim

, Hebestreit

, Mukherji

, Schofield

, Maxwell

, Pugh

, and Ratcliffe

. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science, 292: 468–472, 2001.

27.

Kamata

, Honda

, Maeda

, Chang

, Hirata

, and Karin

. Reactive oxygen species promote TNFalpha-induced death and sustained JNK activation by inhibiting MAP kinase phosphatases. Cell, 120: 649–661, 2005.

28.

Land

, Parada

, and Weinberg

. Tumorigenic conversion of primary embryo fibroblasts requires at least two cooperating oncogenes. Nature, 304: 596–602, 1983.

29.

, Shoji

, Takano

, Nishimura

, Aoki

, Takahashi

, Goto

, Kaifu

, Takai

, and Obinata

. Increased susceptibility of MER5 (peroxiredoxin III) knockout mice to LPS-induced oxidative stress. Biochem Biophys Res Commun, 355: 715–721, 2007.

30.

Liou

and Storz

. Reactive oxygen species in cancer. Free Radic Res, 44: 479–496, 2010.

31.

Mandal

, Schneider

, Kolle

, Kuhlencordt

, Forster

, Beck

, Bornkamm

, and Conrad

. Loss of thioredoxin reductase 1 renders tumors highly susceptible to pharmacologic glutathione deprivation. Cancer Res, 70: 9505–9514, 2010.

32.

Mannes

, Seiler

, Bosello

, Maiorino

, and Conrad

. Cysteine mutant of mammalian GPx4 rescues cell death induced by disruption of the wild-type selenoenzyme. FASEB J, 25: 2135–2144, 2011.

33.

Manning

and Davis

. Targeting JNK for therapeutic benefit: from junk to gold?. Nat Rev Drug Discov, 2: 554–565, 2003.

34.

Mansfield

, Guzy

, Pan

, Young

, Cash

, Schumacker

, and Simon

. Mitochondrial dysfunction resulting from loss of cytochrome c impairs cellular oxygen sensing and hypoxic HIF-alpha activation. Cell Metab, 1: 393–399, 2005.

35.

Mecinovic

, Chowdhury

, Flashman

, and Schofield

. Use of mass spectrometry to probe the nucleophilicity of cysteinyl residues of prolyl hydroxylase domain 2. Anal Biochem, 393: 215–221, 2009.

36.

Nonn

, Williams

, Erickson

, and Powis

. The absence of mitochondrial thioredoxin 2 causes massive apoptosis, exencephaly, and early embryonic lethality in homozygous mice. Mol Cell Biol, 23: 916–922, 2003.

37.

Nytko

, Maeda

, Schlafli

, Spielmann

, Wenger

, and Stiehl

. Vitamin C is dispensable for oxygen sensing in vivo . Blood, 117: 5485–5493, 2011.

38.

Pan

, Mansfield

, Bertozzi

, Rudenko

, Chan

, Giaccia

, and Simon

. Multiple factors affecting cellular redox status and energy metabolism modulate hypoxia-inducible factor prolyl hydroxylase activity in vivo and in vitro . Mol Cell Biol, 27: 912–925, 2007.

39.

Parrinello

, Samper

, Krtolica

, Goldstein

, Melov

, and Campisi

. Oxygen sensitivity severely limits the replicative lifespan of murine fibroblasts. Nat Cell Biol, 5: 741–747, 2003.

40.

Perera

and Bardeesy

. Cancer: when antioxidants are bad. Nature, 475: 43–44, 2011.

41.

Pugh

and Ratcliffe

. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med, 9: 677–684, 2003.

42.

Saitoh

, Nishitoh

, Fujii

, Takeda

, Tobiume

, Sawada

, Kawabata

, Miyazono

, and Ichijo

. Mammalian thioredoxin is a direct inhibitor of apoptosis signal-regulating kinase (ASK) 1. EMBO J, 17: 2596–2606, 1998.

43.

Sandin

, Dagnell

, Gonon

, Pernow

, Stangl

, Aspenstrom

, Kappert

, and Ostman

. Hypoxia followed by re-oxygenation induces oxidation of tyrosine phosphatases. Cell Signal, 23: 820–826, 2011.

44.

Semenza

. HIF-1 and tumor progression: pathophysiology and therapeutics. Trends Mol Med, 8: S62–S67, 2002.

45.

Stanley

, Sivakumaran

, Shi

, McDonald

, Lloyd

, Watson

, Aon

, and Paolocci

. Thioredoxin reductase-2 is essential for keeping low levels of H(2)O(2) emission from isolated heart mitochondria. J Biol Chem, 286: 33669–33677, 2011.

46.

Tietze

. Enzymic method for quantitative determination of nanogram amounts of total and oxidized glutathione: applications to mammalian blood and other tissues. Anal Biochem, 27: 502–522, 1969.

47.

Trachootham

, Alexandre

, and Huang

. Targeting cancer cells by ROS-mediated mechanisms: a radical therapeutic approach?. Nat Rev Drug Discov, 8: 579–591, 2009.

48.

Uhlen

, Oksvold

, Fagerberg

, Lundberg

, Jonasson

, Forsberg

, Zwahlen

, Kampf

, Wester

, Hober

, Wernerus

, Bjorling

, and Ponten

. Towards a knowledge-based Human Protein Atlas. Nat Biotechnol, 28: 1248–1250, 2010.

49.

Valdman

, Haggarth

, Cheng

, Lopez-Beltran

, Montironi

, Ekman

, and Egevad

. Expression of redox pathway enzymes in human prostatic tissue. Anal Quant Cytol Histol, 31: 367–374, 2009.

50.

Vivanco

, Palaskas

, Tran

, Finn

, Getz

, Kennedy

, Jiao

, Rose

, Xie

, Loda

, Golub

, Mellinghoff

, Davis

, Wu

, and Sawyers

. Identification of the JNK signaling pathway as a functional target of the tumor suppressor PTEN. Cancer Cell, 11: 555–569, 2007.

51.

Wang

, Jiang

, and Semenza

. Effect of altered redox states on expression and DNA-binding activity of hypoxia-inducible factor 1. Biochem Biophys Res Commun, 212: 550–556, 1995.

52.

Xia

, Meng

, Liu

, Rojanasakul

, Wang

, and Jiang

. Reactive oxygen species regulate angiogenesis and tumor growth through vascular endothelial growth factor. Cancer Res, 67: 10823–10830, 2007.

53.

Zhong

, De Marzo

, Laughner

, Lim

, Hilton

, Zagzag

, Buechler

, Isaacs

, Semenza

, and Simons

. Overexpression of hypoxia-inducible factor 1alpha in common human cancers and their metastases. Cancer Res, 59: 5830–5835, 1999.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.07 MB

0.32 MB

Knockout of Mitochondrial Thioredoxin Reductase Stabilizes Prolyl Hydroxylase 2 and Inhibits Tumor Growth and Tumor-Derived Angiogenesis

Abstract

Introduction

Results

Transformation and characterization of Txnrd2-deficient mouse embryonic fibroblasts

Targeted loss of Txnrd2 substantially impairs tumor growth in vivo

Txnrd2−/− tumors are highly susceptible to pharmacologic GSH deprivation

Loss of Txnrd2 delays the angiogenic switch and impairs tumor angiogenesis

Reduced expression of key angiogenic players in tumors derived from Txnrd2−/− cells

Hif-1α stabilization in response to serum withdrawal and hypoxia is impaired

Increased PHD2 expression promotes Hif-1α protein degradation in Txnrd2−/− cells and tumors

Enhanced JNK phosphorylation leads to increased PHD2 protein levels

Expression of Txnrd2 in human cancer tissue

Discussion

Materials and Methods

Cell lines and reagents

Generation of transformed Txnrd2+/+ and Txnrd2−/− cells

Proliferation assay

Measurement of cellular ROS levels

Measurement of GSH concentration in cells and tumors

Quantitative real-time PCR

Hematoxylin-eosin staining/immunohistochemistry/immunofluorescence and image analysis

Immunoblotting

PHD2 knockdown in Txnrd2+/+ and Txnrd2−/− MEFS

Txnrd2 knockdown in LLC1 tumor cells

JNK knockdown in Txnrd2 −/− MEFs

Enzyme-linked immunosorbent assay

Subcutaneous xenograft tumor model

Treatment of tumor-bearing mice with BSO

Measurement of PTP activity

Determination of NFκB activation

Statistical analysis

Footnotes

Acknowledgments

Author Disclosure Statement

Abbreviations Used

References

Supplementary Material

Txnrd2^−/− tumors are highly susceptible to pharmacologic GSH deprivation

Reduced expression of key angiogenic players in tumors derived from Txnrd2^−/− cells

Increased PHD2 expression promotes Hif-1α protein degradation in Txnrd2^−/− cells and tumors

Generation of transformed Txnrd2^+/+ and Txnrd2^−/− cells

PHD2 knockdown in Txnrd2^+/+ and Txnrd2^−/− MEFS

JNK knockdown in Txnrd2 ^−/− MEFs