Hypoxia upregulates osteopontin expression in NIH-3T3 cells via a Ras-activated enhancer

Abstract

Osteopontin (OPN) is a secreted phosphoglycoprotein that has been linked to tumor progression and survival in several solid tumors, including head and neck cancers. Previous studies showed that OPN expression is induced by tumor hypoxia, and its plasma levels can serve as a surrogate marker for tumor hypoxia and treatment outcome in head and neck cancer patients. In this study, we investigate the transcriptional mechanism by which hypoxia enhances OPN expression. We found that OPN is induced in head and neck squamous cell carcinoma (HNSCC) cell lines and in NIH3T3 cells by hypoxia at both mRNA and protein levels in a time-dependent manner. Actinomycin D chase experiments showed that hypoxic induction of OPN was not due to increased mRNA stability. Deletion analyses of the mouse OPN promoter regions indicated that a ras-activated enhancer (RAE) located at −731 to −712 relative to the transcription start site was essential for hypoxia-enhanced OPN transcription. Using electrophoretic mobility shift assays with the RAE DNA sequence, we found that hypoxia induced sequence-specific DNA-binding complexes. Furthermore, hypoxia and ras exposure resulted in an additive induction of OPN protein and mRNA levels that appeared to be mediated by the RAE. Induction of OPN through the RAE element by hypoxia is mediated by an Akt-kinase signaled pathway as decreasing Akt levels with dominant negative constructs resulted in inhibition of OPN induction by hypoxia. Taken together, these results have identified a new hypoxia responsive transcriptional enhancer that is regulated by Akt signaling.

Main

Hypoxia is a common feature in solid tumors (Vaupel et al., 2001). It is an important microenvironmental factor that decreases the effectiveness of conventional chemotherapy and radiotherapy (Brown and Giaccia, 1998), increases malignant tumor progression (Teicher et al., 1990; Young and Hill, 1990; Graeber et al., 1996), enhances tumor cell invasion and metastasis (Le et al., 2004), and is prognostic for tumor control by conventional treatment modalities (Brizel et al., 1996; Hockel et al., 1996). Accordingly, identification of molecular effectors of hypoxia will permit the development of molecular targeted therapies that is critical for improved therapeutic management of cancer patients (Giaccia et al., 2003). The cellular response to hypoxia involves a series of metabolic and biosynthetic events that are essential for adaptation to a hypoxic environment, including the induction of a number of physiologically important genes (Koong et al., 2000; Zhu et al., 2002b; Denko et al., 2003; Semenza, 2003). These include genes involved in metabolism, angiogenesis, cell cycle regulation, tissue invasion, and apoptosis, many of which are regulated by the hypoxia inducible factor-1α (HIF-1α) (Harris, 2002; Denko et al., 2003). Recent studies have also shown that osteopontin (OPN) is also a hypoxia inducible gene (Sodhi et al., 2001; Le et al., 2003).

OPN is a highly phosphorylated and glycosylated protein secreted into the extracellular matrix by a variety of cell types, including osteoblasts, epithelial cells, endothelial cells, macrophages, and smooth muscle cells (Denhardt et al., 2001a, 2001b). OPN has been shown to be multifunctional, with activities in cell migration, cell survival, inhibition of calcification, regulation of immune cell function, and tumor cell metastasis (Denhardt et al., 1995; Weber and Cantor, 1996; Liaw et al., 1998). Increased levels of OPN protein have been linked to many disease states including atherosclerosis, transformation, tumor metastasis (Denhardt et al., 2001a; Furger et al., 2001), and renal diseases (Xie et al., 2001). Previous studies from this laboratory have identified a link between OPN expression and tumor hypoxia in several cancer cell lines. We also found that elevated plasma OPN levels correlated with tumor hypoxia and poor treatment outcomes in head and neck cancer patients (Le et al., 2003). However, the mechanism governing the induction of OPN by hypoxia has yet to be elucidated.

The regulation of OPN expression under normoxic conditions may differ in different cell types. While the OPN promoter contains various motifs including a purine-rich sequence, an ets-like sequence, glucocorticoid-, vitamin D-, and 12-O-tetradecanoylphorbol-13-acetate (TPA)-response elements (Craig et al., 1989; Noda et al., 1990; Denhardt and Guo, 1993; Koszewski et al., 1996), it does not possess any known hypoxic response elements. Guo et al. (1995) identified a 9-nucleotide region, IndexTermGGAGGCAGG, which bound a protein whose apparent abundance was enhanced in metastatic cell lines and in ras-transformed NIH-3T3 cells relative to non-transformed cells. This sequence was termed ras-activated enhancer (RAE). However, isolation of this RAE-binding protein (s) has been elusive.

To investigate the molecular mechanism by which hypoxia promotes transcriptional regulation of the mouse OPN gene, we evaluated hypoxic regulation of OPN expression in NIH-3T3 cells and performed a detailed analysis of the mouse OPN promoter region in these cells. We found that the RAE is essential for hypoxic induction of OPN transcription. Furthermore, we investigated the hypoxic regulation of OPN in cells that overexpress activated H-ras and or its dominant negative constructs. The results of these studies indicate that the hypoxic induction of OPN is mediated through the RAE elements and is enhanced in ras transformed cell. In addition, hypoxic regulation of OPN appeared to be mediated via the Akt protein. Taken together, these studies suggest a mechanism for hypoxic induction of OPN expression that involves the ATK protein and the RAE element.

Results

Hypoxia up-regulates OPN in HNSCC and in NIH-3T3 cells

To determine whether hypoxia regulates OPN expression in head and neck squamous cell carcinoma (HNSCC) cell lines (SCC4 and SCC22B), cells were exposed to hypoxia for 0 or 24 h. Real-time polymerase chain reaction (PCR) showed that OPN mRNA was induced about threefold at 24 h of hypoxia exposure in SCC4 and SCC22B cells (Figure 1a). Immunoblot analysis showed that OPN protein was induced by hypoxia in a time-dependent manner in SCC22B cells (Figure 1b). Similarly, OPN mRNA was induced in a time-dependent manner in NIH-3T3 cells. Induction of OPN mRNA was first observed at 8 h of hypoxia exposure and reached the highest level (threefold increase) at 24 h (Figure 2a). This observation was validated by Northern blot analysis (Figure 2b), in which total RNA from NIH-3T3 cells treated with 1α, 25-dihydroxyvitamin D3 (10 nM), was used as a positive control (Noda et al., 1990; Koszewski et al., 1996). Immunoblot studies showed that OPN protein was increased by hypoxia in a similar time sequence as observed for OPN mRNA (Figure 2c) in NIH-3T3 cells. Taken together, these data indicated that OPN is a hypoxia responsive gene in human HNSCC cells and in NIH-3T3 cells.

Figure 1
figure1

Hypoxia increases OPN mRNA and protein expression in HNSCC cells. (a) Quantitative real-time PCR analysis showing increased OPN mRNA expression in SCC4 and SCC22B cells after 24 h of hypoxia exposure in relation to normoxic controls. OPN expression was first normalized to that of β-actin and then expressed in relation to normoxic levels, which were arbitrarily set at 100% for SCC4 cells. Values (mean±s.e.m., n=3) **P<0.01 compared to normoxia. Nmx, normoxia; Hpx, hypoxia. (b) Immunoblot showing induction of OPN protein in SCC22B cells after a time course of hypoxia. Blots were stripped and reprobed with anti-β-actin to show equal loading

Figure 2
figure2

Hypoxia increases OPN mRNA and protein expression in NIH-3T3 cells. (a) A time course study showing increasing OPN mRNA expression under hypoxia relative to the normoxic control by quantitative real-time PCR analysis. OPN expression was expressed in the same way as in Figure 1a. Values (mean±s.e.m., n=3). *P<0.05, **P<0.01 compared to 0 h. (b) Northern blot study confirming induction of OPN mRNA after 24 h of hypoxia. RNA from cells exposed to 24 h of 25-Dihydroxyvitamin D3 (VitD3, 10 nM) were used as positive controls. Blots were probed for OPN mRNA and β-actin mRNA. (c) Immunoblot showing induction of OPN protein after a time course of hypoxia. Blots were stripped and reprobed with anti-β-actin to show equal loading. (d) Effect of hypoxia on OPN mRNA stability. Actinomycin D (Act D, 5 μg/ml) was added to NIH-3T3 cells after 16 h initial exposure to normoxia (N) or hypoxia (H). Real-time PCR was performed at 0, 2, 4, and 6 h from the initiation of actinomycin D treatment. OPN mRNA expression was expressed in the same way as in Figure 1a

Hypoxic induction of OPN mRNA is not through increased mRNA stability

To further clarify whether hypoxic induction of OPN mRNA is a result of enhanced gene transcription or mRNA stability, we examined the rate of OPN mRNA decay following inhibition of RNA polymerase II-dependent transcription by actinomycin D (Harrold et al., 1991). NIH-3T3 cells were exposed to normoxia or hypoxia for 16 h and were then treated with actinomycin D to arrest transcription. Cells were harvested at distinct time points over a 6 h period and OPN mRNA levels were analysed by real-time RT–PCR. Estimated half-lives for OPN mRNA obtained from three independent experiments were as follows: under normoxia, 3.94±0.21 h; after 16 h hypoxia treatment, 4.11±0.26 h (P>0.05) (Figure 2d). Thereby, hypoxia did not significantly prolong the half-life of OPN mRNA. These results suggested that the increase in OPN mRNA levels in NIH-3T3 cells by hypoxia is mediated via transcriptional induction rather than via an increase in mRNA stability.

Identification of a hypoxia responsive region in the OPN promoter

To gain insight into hypoxia regulation of OPN transcription, we transiently transfected luciferase reporter constructs containing 0.168–0.962 kb of OPN upstream sequence ranging from −882 to +79 (Figure 3a), as described previously (Guo et al., 1995), and analysed the activity of each promoter fragment under either normoxia or hypoxia. The construct p6, spanning the −777 to +79 region, showed the highest level of activity in response to hypoxia (2.93-fold increase relative to normoxia) (Figure 3b). Constructs p1–p5 and p7 were all capable of stimulating moderate expression of the reporter gene under hypoxia due to the presence of an AP-1 element in the basal promoter segment (construct p1, −88/+79) (Guo et al., 1995). This AP-1 activity may account for the small increase in the luciferase activity in constructs p1−p5 under hypoxia (Ausserer et al., 1994). Extension of the promoter fragment up to −882 (construct p7) resulted in a decrease in the transcriptional activity under both normoxia and hypoxia when compared to that of construct p6, suggesting the presence of a repressor element. Overall, these promoter analyses suggest that a 177 base-pair DNA fragment located at −777 to −600 relative to the transcription start site is the essential promoter element for hypoxia-enhanced OPN expression. Interestingly, construct p3 showed high activity both under normoxic and hypoxic conditions, an observation not previously noted by Guo et al. (1995) who used β-galactosidase-based reporter constructs. The reasons for this difference between the two studies are unknown; however, we also observed similar elevated activity both under normoxia and hypoxia for construct p3 in a different cell line (bEND3 cells, data not shown).

Figure 3
figure3

Identification of a region in the mouse OPN promoter responsible for hypoxic induction of gene expression. (a) Schematic representations of different OPN promoter fragments used to drive expression of the luciferase reporter gene as described (Guo et al., 1995). The location of RAE is shown in constructs p6 and p7. (b) Relative luciferase activities for different constructs under normoxia and 24 h of hypoxia in NIH-3T3 cells transiently transfected with an empty vector (v) or a vector containing different promoter fragments as described in (a). Fragment p6 showed highest hypoxic induction under hypoxia (*P<0.05). Promoter activity was measured in relative light units per μg of protein lysate. Results are expressed as mean activity±s.d. (n=3). Nmx, normoxia; Hpx, hypoxia

RAE is required for hypoxia induction of OPN transcription in NIH-3T3 cells

Within the 177 bp segment, there are several potential binding sites for transcription factors, including VDRE (vitamin D3 responsive element), AP-1, AP-3, AP5, CRE, and an RAE (Hijiya et al., 1994; Guo et al., 1995). To define the binding site(s) that is crucial for hypoxia induction of OPN, variant fragments of this segment and the RAE (−740 to −712) were inserted into the p1 construct (Figure 4a). Luciferase activity assays using the derived reporter constructs and the p1 construct showed that the RAE-containing constructs (A, B, and D) had a higher fold of hypoxic induction relative to normoxia (2.5−2.8-fold increase, P<0.01), especially when compared to nonRAE-containing constructs (C and p1, 1.3−1.7-fold increase, P>0.05, Figure 4b). Furthermore, cells transfected with construct Mut (a mutant version of the RAE) showed lower luciferase activity in comparison with those transfected with the wild-type RAE (Figure 4b). These results indicated that the RAE binding site is important for hypoxic induction of OPN transcription in NIH-3T3 cells.

Figure 4
figure4

RAE is required for hypoxic induction of OPN transcription in NIH-3T3 cells. (a) Schematic representation of different fragments within the 177-bp sequence (−777 to −600) responsible for hypoxic induction of OPN. Fragments with or without an RAE binding site were subcloned into p1 construct, which contains the basal promoter of OPN (−88 to +79). (b) Fold induction of relative luciferase activities by hypoxia (in relation to normoxic controls) in NIH-3T3 cell transiently transfected with different luciferase reporter constructs containing the promoter fragments described in A (P<0.01 for constructs A, B, and D vs P1, P>0.05 for construct C and Mut vs P1). Promoter activity was measured in relative light units per μg of protein lysate and expressed as mean fold inductions±s.d. (n=3)

To confirm the involvement of the RAE in hypoxic regulation of OPN, whole cell extracts from NIH-3T3 cells that were treated for 24 h in either hypoxia or normoxia were analysed by EMSA using 32P-labeled oligonucleotide probes containing the RAE binding sequence (−731 to −712). In addition, since the RAE was previously identified to be ras-activated (Guo et al., 1995), extracts from T24 H-ras-transformed 3T3 cells (PAP2) and ras-v12 transiently transfected NIH-3T3 cells were used as positive controls (Figure 5, lanes 3 and 4). Two major and four minor complexes were detected, of which complex B appeared to be Ras- and hypoxia-specific (Figure 5b; lanes 1–4). Binding specificity in complex B was demonstrated by a gradual reduction and elimination of the band intensity in the presence of increasing concentrations of unlabeled wild-type RAE oligonucleotides (Figure 5b, lane 5, 50-fold excess; lane 6, 100-fold excess; lane 7, 200-fold excess), whereas it is minimally affected by the addition of 100-fold excess of unlabeled mutant competitors (lanes 8–11). In addition, complex B failed to form when a 32P labeled mutant RAE probe was incubated with 10-fold excess of unlabeled wild-type RAE oligonucleotides (Figure 5c, lane 13). These data suggest that the RAE binding site interacts with certain protein (s) under hypoxia and in ras-transformed cells to form complex B and this interaction is important for hypoxic upregulation of OPN expression.

Figure 5
figure5

Analysis of DNA–protein complexes formed with RAE by EMSA. (a) DNA sequence of the wild-type and a mutated version of RAE used in EMSA. (b) EMSA experiments were performed using whole cell extracts from NIH-3T3 cells and radiolabeled wild-type RAE probes (wt). Samples: lane 1, untreated normoxic NIH-3T3 cells (negative control); lane 2, normoxic PAP2 cells (T24 H-ras-transformed 3T3 cells, positive control); lane 3, NIH-3T3 cells transiently transfected with ras-v12 (positive control); lanes 4–11, NIH-3T3 cells treated with hypoxia for 24 h. For the experiments in lanes 1–4, no competitor was added; lanes 5–7, unlabeled wild-type RAE oligonucleotides (w) were used as competitors (lane 5 (50-fold), lane 6 (100-fold), lane 7 (200-fold)); lanes 8–11, unlabeled mutant RAE oligonucleotides (m) were used as competitors. (c) EMSA showing failure of complex B formation under hypoxia when labeled mutant RAEs were used as baits and 10-fold excess unlabeled wild-type RAEs were used as competitors (lane 13). Lane 12 is the control showing complex B formation under hypoxia when labeled RAEs were used as baits and unlabeled mutant RAEs were used as competitors (similar conditions as used in lane 11, (b)

Relationship between hypoxia and ras regulation of OPN expression

The expression of OPN correlates with the extent of ras activation (Craig et al., 1990; Chambers et al., 1992). To investigate the relationship between hypoxia and ras regulation of OPN, we compared the hypoxic and normoxic expression of OPN mRNA and protein in NIH-3T3 cells that were manipulated to either transiently overexpress ras-V12 (NIH-3T3-ras) or stably overexpress H-ras (PAP2 cells). Real-time PCR and immunoblot studies showed that hypoxia increased OPN mRNA and protein expression at least twofold over the basal induction by ras activation, suggesting an additive interaction between the two mechanisms of OPN regulation (Figure 6a and b). Luciferase reporter assays confirmed that the additive induction of OPN mRNA by hypoxia and ras appeared to be mediated through the RAE element in both PAP2 cells and NIH-3T3 cells that transiently overexpress ras (Figure 7a and b). Hypoxia exposure resulted in a 2.5–3.0-fold increase in luciferase activity in RAE containing promoter constructs (constructs A and D) but much less so in those without RAE or with mutant RAE constructs (p1 and Mut). Again, a small increase in luciferase activity was noted for constructs p1 and mutant, presumably due to the presence of an AP-1 site in the basal promoter contained in these constructs. As such an increase was not noted in the 2 × RAE constructs that lacked the AP-1 site (Figure 7d). In addition, suppression of ras activity by transient transfection of the ras-N17 dominant negative construct in NIH-3T3 cells abrogated hypoxia induction of OPN expression by luciferase reporter assay (Figure 7c). Since the AP-1 binding site in the OPN minimal promoter plays a partial role in Ras regulation of OPN (Guo et al., 1995), to further investigate the role of RAE on OPN regulation, we generated a 2 × RAE construct using the pGL3 promoter vector that does not contain an AP-1 site. As shown in Figure 7d, luciferase activity from the 2 × RAE construct was elevated by hypoxia and/or ras-v12, and was diminished by ras-N17. Such induction was not noted for the mutant 2 × RAE construct. Thus, ras exerts a positive effect on hypoxia induction of OPN transcription through the RAE element on the OPN promoter.

Figure 6
figure6

Effect of ras and hypoxia on OPN mRNA and protein expression in NIH-3T3 cells. (a) Quantitative real time RT–PCR analysis of OPN mRNA expression in NIH-3T3, NIH-3T3-rasV12, or PAP2 cells under normoxia or hypoxia for 8 or 24 h. OPN expression was expressed in the same way as in Figure 1a. Values (mean±s.e.m., n=3). *P<0.05, **P<0.01. Nmx, normoxia; Hpx, hypoxia. (b) Immunoblot analysis of OPN protein expression in NIH-3T3, or PAP2 cells under normoxia or hypoxia for 24 h (left), which was quantified by computer densitometry and normalized to the expression of β-actin (right). N, normoxia; H, hypoxia

Figure 7
figure7

Additive induction of OPN expression by hypoxia and Ras is mediated via the RAE and requires functional Ras. (a) Luciferase reporter assays in NIH-3T3 cells that were transiently transfected with OPN promoter constructs and either ras-V12 or an empty vector. The cells were allowed to recover for 24 h, exposed to normoxia or hypoxia for 24 h, and assayed for luciferase activity (normalized to μg of protein lysate). Results are expressed as mean activity±s.d. (n=3). (b) PAP2 cells were transfected with OPN promoter constructs, allowed to recover for 24 h, exposed to normoxia (Nmx) or hypoxia (Hpx) for 24 h, and assayed for luciferase activity (normalized to μg of protein lysate). Results are expressed as mean activity±s.d. (n=3). (c) Luciferase reporter assays in NIH-3T3 cells that were transiently transfected with OPN promoter constructs and either ras-N17 dominant-negative constructs or an empty vectors. The cells were allowed to recover for 24 h, exposed to normoxia or hypoxia for 24 h, and assayed for luciferase activity (normalized to μg of protein lysate). Results are expressed as mean activity±s.d. (n=3). (c) Luciferase reporter assays in NIH-3T3 cells that were transiently transfected with OPN promoter constructs and either ras-N17 or an empty vector. The cells were allowed to recover for 24 h, exposed to normoxia or hypoxia for 24 h, and assayed for luciferase activity (normalized to μg of protein lysate). Results are expressed as mean activity±s.d. (n=3). (d) Luciferase reporter assays in NIH-3T3 cells that were transiently transfected with OPN promoter constructs and either ras-V12, ras-N17 or an empty vector. The cells were allowed to recover for 24 h, exposed to normoxia or hypoxia for 24 h, and assayed for luciferase activity. Results are expressed as mean activity±s.d. (n=3). Nmx, normoxia; Hpx, hypoxia

Hypoxic induction of OPN is mediated via the Akt protein

Zhang et al. (2003) have previously shown that overexpression of the Akt kinase in breast cancer cells induced OPN mRNA through an Akt-responsive element that is located in the same region as the hypoxia-responsive promoter domain (between base −600 and base −777). Therefore, we investigated the relationship between Akt and hypoxia in the regulation of OPN through promoter constructs. Transient overexpression of Akt resulted in induction of OPN expression by luciferase reporter assay of the 2 × RAE, which does not contain the AP-1 site, under normoxia and to a greater degree under hypoxia (Figure 8). More importantly, suppression of Akt level by transient transfection of the PI3 kinase dominant-negative mutant Δp85 in NIH-3T3 cells abrogated hypoxia induction of OPN expression by luciferase reporter assay of the 2 × RAE (Figure 8). These results suggest that hypoxia induction of OPN transcription is mediated via the Akt kinase.

Figure 8
figure8

Hypoxic induction of OPN is mediated via the Akt protein. Luciferase reporter assays in NIH-3T3 cells that were transiently transfected with OPN promoter constructs and either pWZLneo-Akt or Δp85 or an empty vector. The cells were allowed to recover for 24 h, exposed to normoxia or hypoxia for 24 h, and assayed for luciferase activity (normalized to μg of protein lysate). Results are expressed as mean activity±s.d. (n=3). Nmx, normoxia; Hpx, hypoxia

Discussion

The regulation of OPN by hypoxia and reoxygenation was first reported over a decade ago in primary cultures of renal proximal tubule epithelial (PTE) cells (Hwang et al., 1994). Biologically, OPN signaling has been proposed to protect cells from hypoxia- and reoxygenation-induced injury (Denhardt et al., 2001a). Hwang et al. (1994) noted that OPN expression was increased by hypoxia and reoxygenation in PTE cells in young but not aged donors. In addition, PTE cells from aged kidneys were more susceptible to cell death under hypoxic conditions than those from young kidneys, suggesting that the diminished ability of cells from old kidneys to increase OPN expression after hypoxia-reoxygenation may contribute to their increased susceptibility to oxidant injury. In support of this concept, OPN null mice exhibited ischemia-induced renal dysfunction that was twice as pronounced as that observed in wild-type mice. Furthermore, the addition of OPN resulted in cytoprotection of proximal tubular cells subjected to hypoxia in in vitro studies (Noiri et al., 1999). OPN has also been implicated to play an important role in hypoxia-induced proliferation of mesangial cells (Sodhi et al., 2000). Overall, increased OPN expression appears to exert a survival benefit in epithelial cells exposed to hypoxia and reoxygenation and induction of OPN seems to be an adaptive response of cells to redox microenvironmental stresses to improve their chance of survival. While the role of OPN in cell adaptation to hypoxia is documented, little is known about the hypoxic regulation of this protein.

The results presented in this report demonstrate that the induction of OPN was dependent on a nucleotide sequence in its 5′-flanking region that was previously identified as an RAE. Using activated ras-transformed NIH-3T3 cells, we found that hypoxia and ras induction of OPN was additive and was mediated through the RAE DNA sequence. Therefore, the induction of OPN by hypoxia is transcriptionally regulated by a new regulatory element.

Molecular adaptation to hypoxia usually depends on the binding of the HIF-1 transcription factor to cognate response elements in oxygen-regulated genes (Ebert and Bunn, 1998). Although HIF-1-responsive elements appear to be necessary for transcriptional induction of many hypoxia-responsive genes, a number of different transcriptional binding sites have also been shown to be involved in hypoxic regulation of gene expression. For example, the AP-1, AP-2, SP-1, and CRE (cyclic AMP response element) binding sites together with the HIF-1-responsive element are important for hypoxic upregulation of the vascular endothelial growth factor (VEGF) (Ebert and Bunn, 1998; Pollmann et al., 2001). The OPN promoter is responsive to many stimuli, including various forms of physical stress, cytokines, growth factors and hormones through different DNA-binding sites (Weber, 2001; Denhardt et al., 2003). Within the OPN promoter, multiple recognition sites for transcription factors have been mapped (Weber, 2001; Denhardt et al., 2003). Among them, the RAE is an enhancer element that conferred an increase in both OPN transcriptional activity in ras-transformed NIH-3T3 cells and metastatic potential of transformed cells (Guo et al., 1995). In this study, functional analysis of the mouse OPN promoter in NIH-3T3 cells showed that the RAE was not only an RAE but also a hypoxia-responsive element.

Ras activation has been reported to be synergistic with hypoxia in the regulation of other significant hypoxia regulated genes such as VEGF (Mazure et al., 1997). Ras also represents a critical junction in the downstream transmission of signals from growth factor receptors and stress-inducing agents (Lowy and Willumsen, 1993). Here, we show ras and hypoxia can operate in concert with each other to increase OPN expression. As activating ras mutations and hypoxia are the common features of many solid tumors, they can both lead to enhanced OPN expression in these tumors.

As mentioned previously, Zhang et al. (2003) showed that overexpression of the Akt kinase could induce OPN mRNA level and the Akt-responsive promoter domain was mapped to the same region as the hypoxia-responsive region (between base −600 and base −777). Teramoto et al. (2003) have reported that both Akt and OPN were induced when ras-V12 are stably transfected in NIH-3T3 cells. These data suggest that the Akt kinase may play a role in OPN regulation by hypoxia and ras. Our data suggest that hypoxic induction of OPN expression is mediated through RAS and the Akt kinase and the following pathway can be suggested for OPN regulation by hypoxia and ras (Figure 9). Briefly, hypoxia exposure and/or ras activation results in the activation of the Akt kinase, which in turn activates an unknown transcriptional factor that binds to the RAE and turns on OPN expression. In addition, our data suggest that RAS is partially downstream of hypoxia as transient inactivation of RAS activity abrogated hypoxic induction of OPN. The identity of the transcription factor(s) involved in hypoxia and ras regulation of OPN has yet to be elucidated and is a subject of active investigation in several laboratories.

Figure 9
figure9

Suggested pathway for hypoxia regulation of OPN expression

In summary, we have identified a region in the OPN promoter that is important for both ras and hypoxic regulation of OPN expression. We also identified a potential pathway for hypoxia induction of OPN that involves the activation of the Akt kinase. Our results suggest the first link between ras transformation, hypoxia, Akt, and the RAE binding element.

Materials and methods

Cell lines and cell culture

Human cell lines SCC22B from University of Michigan, SCC4 and NIH-3T3 cells from ATCC, and a T24 H-ras-transformed NIH-3T3 cell line PAP2 (Guo et al., 1995) were used in this study. These cells were maintained in Dulbecco's minimal essential medium supplemented with 10% fetal bovine serum.

Hypoxia treatment

Hypoxic cultures were maintained in a hypoxia chamber (<0.2% O2) (Sheldon Corp., Cornelius, OR, USA) (Hammond et al., 2002) for 0 to 24 h at 37°C in humidified 95% N2/5% CO2, and normoxic cultures were maintained in 95% air/5% CO2 for similar duration.

Real-time PCR and Northern blotting

Total RNA from cell cultures was extracted, reverse transcribed, and quantified by real-time PCR as described (Zhu et al., 2003). Mouse OPN-specific primers were IndexTerm5′-CTCAGAAGCAGAATCTC-3′ (forward) and IndexTerm5′-ATGGTCTCCATCGTCATCAT-3′ (reverse); mouse β-actin-specific primers IndexTerm5′-ACCGAGCGTGGCTACAGCTT-3′ (forward) and IndexTerm5′-TCAGGCAGCTCATAGCTCTT-3′ (reverse) were used as controls. Quantitative RT–PCR analysis was performed with a fluorescent dye, SYBR Green, using ABI PRISM 7900 Sequence Detection System (Applied Biosystem, Foster City, CA, USA). The thermal profile was 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 10 s and 60°C for 1 min. Data were analysed with the use of ABI PRISM 7900 Sequence Detection System 2.0. All reactions were repeated at least 3 times to ensure reproducibility. OPN expression for each time point was first normalized to that of β-actin, whose expression was unchanged with hypoxic treatment. Hypoxic OPN levels were then expressed in relation to normoxic levels, which were arbitrarily set at 100% for ease of interpretation and comparison.

Northern blotting was carried out as described previously (Zhu et al., 1999). Briefly, total RNA from treated and untreated cells was extracted using RNeasy Mini Kits (Qiagen, Valencia, CA, USA), according to the manufacturer’s instructions. Of total RNA, 15 μg from each sample was fractionated on 1% agarose gels with 10 mM sodium phosphate pH 7.0 and transferred to Hybond-N nylon membranes (Amersham Biosciences, Piscataway, NJ, USA). Filters were stained with methylene blue and then hybridized with a probe specific for mouse OPN mRNA. 32P-Radiolabeled mouse OPN cDNA (ATCC) was prepared using rediprime II random primer labeling system (Amersham Biosciences). The radioactive signal was detected using a Bio-Rad phosphorimaging system.

Actinomycin D chase experiments

To assess the effect of hypoxia on OPN mRNA stability, NIH-3T3 cells were exposed to normoxia or hypoxia for 16 h as described above. Actinomycin D (5 μg/ml) was added and cells were incubated for varying lengths of time, up to 6 h. To exclude the effect of reoxygenation, hypoxic cells were maintained in a hypoxic chamber until RNA extraction. Total cellular RNA was extracted at 0, 2, 4, and 6 h after the addition of actinomycin D, and Real-time RT–PCR was performed as described above. Relative levels of remaining OPN mRNA (normalized to β-actin) were reported for each time point.

Western blotting

For immunoblotting, cells were lysed in 1 × sample buffer (2% SDS, 100 mM DTT, 60 mM Tris (pH 6.8), and 10% glycerol) and boiled for 5 min (Zhu et al., 2002a). Protein concentrations were determined using BC protein assay (Bio-Rad, Hercules, CA, USA). In all, 15 μg protein samples were loaded on 10 or 12% sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) gels and blotted onto polyvinylidene fluoride (PVDF) membranes (Bio-Rad), which were probed with primary antibody at 4°C overnight and secondary antibody at room temperature for 1 h; signals were detected by chemiluminescence (Roche, Indianapolis, IN, USA). Primary antibodies were goat anti-OPN (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and mouse anti-β-actin (Sigma, St Louis, MO, USA).

Plasmid construction

Different 5′-flanking regions of mouse OPN promoter starting from −882 to the noncoding part of the first exon (+79) were subcloned into the pXP2 luciferase reporter vector (ATCC), as previously described (Guo et al., 1995). A synthetic oligonucleotide identical to the OPN sequence from −740 to −712 and the mutant versions of this sequence (Denhardt et al., 2003), and PCR fragments from −777 to −600, from −740 to −600, or from −695 to −600 were inserted upstream of the basal luciferase construct pOpn (−88/+79) Luc. In addition, a synthetic oligonucleotide contains two tandem repeats of the OPN promoter sequence from −740 to −712 and the mutant version of this sequence, designated as 2 × RAE and M, respectively, were inserted into pGL3-promoter luciferase reporter vector. The correct clones were identified by restriction digestion and confirmed by DNA sequencing.

Transfection and luciferase activity assays

Luciferase reporter constructs were transiently transfected into NIH-3T3 cells by FuGENE 6 transfection reagent (Roche). A construct encoding green fluorescent protein (pEGFP, Clontech, Palo Alto, CA, USA) was used to verify transfection efficiency (Zhu et al., 2003). Luciferase activity was determined by using luciferase assay reagent (Promega, Madison, WI, USA) in a Monolight 2010 Luminometer (Analytical Luminescence Laboratory, Sparks, MD, USA), as previously described (Chan et al., 2002). In the experiments to define the relationship between hypoxia and Ras regulation of OPN expression, the constitutively active and dominant negative mutants of ras (ras-v12 and ras-N17, Clontech, Palo Alto, CA, USA) were transfected with or without luciferase reporter constructs into NIH-3T3 cells. After 24 h, the cells were subjected to hypoxia treatment for 24 h and were colleted for total RNA, protein extraction, or for luciferase assays. In the experiments to define the relationship between hypoxia and Akt regulation of OPN expression, the constitutively active Akt (pWZLneo-Akt) (Kohn et al., 1996) and the PI3 kinase dominant negative mutant (Δp85) (Bedogni et al., 2004) were used.

Preparation of whole-cell extracts and mobility shift assay

Whole-cell extracts were prepared as described previously (Choi et al., 1990; Huang et al., 1994). Briefly, cell pellets were quickly frozen in liquid nitrogen and then thawed on ice for 5–10 min in cell lysis buffer (Choi et al., 1990). Cells were lysed at 4°C by passage through a 27-gauge needle 20 times, followed by centrifugation at 12 000 g for 15 min at 4°C. The supernatants were stored at −80°C. Protein concentrations were determined by Bradford protein assays (Bio-Rad). Extracts and mobility shift assay (EMSA) was performed as described (Guo et al., 1995). Binding was tested in 20 μl of solution by incubating 15 μg of nuclear extract with 17.5 fmol of [γ-32P]ATP-radiolabeled RAE probe (Guo et al., 1995; Denhardt et al., 2003) in the presence or absence of the competitor, as indicated below, in reaction buffer (20 mM HEPES (pH 7.9), 100 mM KCl, 0.5 mM DTT, 0.2 mM EDTA, 20% glycerol, 150 ng poly(dI-dC), 0.05% NP40) for 10 min at room temperature and subsequently 30 min on ice. The complex was separated on a 8% nondenaturing polyacrylamide gel in 0.3 × Tris-borate EDTA buffer for 2.5 h at 250 V. Gels were dried and visualized by using the PhosphorImager™ system (Molecular Dynamics, Sunnyvale, CA, USA). For competition assays, unlabeled DNA (1.75 pmol) was incubated with nuclear extracts at room temperature for 10 min before the addition of the labeled probes.

References

  1. Ausserer WA, Bourrat-Floeck B, Green CJ, Laderoute KR and Sutherland RM . (1994). Mol. Cell. Biol., 14, 5032–5042.

  2. Bedogni B, O’Neill MS, Welford SM, Bouley DM, Giaccia AJ, Denko NC and Powell MB . (2004). Cancer Res., 64, 2552–2560.

  3. Brizel DM, Scully SP, Harrelson JM, Layfield LJ, Dodge RK, Charles HC, Samulski TV, Prosnitz LR and Dewhirst MW . (1996). Cancer Res., 56, 5347–5350.

  4. Brown JM and Giaccia AJ . (1998). Cancer Res., 58, 1408–1416.

  5. Chambers AF, Behrend EI, Wilson SM and Denhardt DT . (1992). Anticancer Res., 12, 43–47.

  6. Chan DA, Sutphin PD, Denko NC and Giaccia AJ . (2002). J. Biol. Chem., 277, 40112–40117.

  7. Choi H, Lin Z, Li B and Liu A . (1990). J. Biol. Chem., 265, 18005–18011.

  8. Craig AM, Bowden GT, Chambers AF, Spearman MA, Greenberg AH, Wright JA, McLeod M and Denhardt DT . (1990). Int. J. Cancer, 46, 133–137.

  9. Craig AM, Smith JH and Denhardt DT . (1989). J. Biol. Chem., 264, 9682–9689.

  10. Denhardt DT, Giachelli CM and Rittling SR . (2001a). Annu. Rev. Pharmacol. Toxicol., 41, 723–749.

  11. Denhardt DT and Guo X . (1993). FASEB J., 7, 1475–1482.

  12. Denhardt DT, Lopez CA, Rollo EE, Hwang SM, An XR and Walther SE . (1995). Ann. NY Acad. Sci., 760, 127142.

  13. Denhardt DT, Mistretta D, Chambers AF, Krishna S, Porter JF, Raghuram S and Rittling SR . (2003). Clin. Exp. Metastasis, 20, 77–84.

  14. Denhardt DT, Noda M, O’Regan AW, Pavlin D and Berman JS . (2001b). J. Clin. Invest., 107, 1055–1061.

  15. Denko NC, Fontana LA, Hudson KM, Sutphin PD, Raychaudhuri S, Altman R and Giaccia AJ . (2003). Oncogene, 22, 5907–5914.

  16. Ebert BL and Bunn HF . (1998). Mol. Cell. Biol., 18, 4089–4096.

  17. Furger KA, Menon RK, Tuckl AB, Bramwelll VH and Chambers AF . (2001). Curr. Mol. Med., 1, 621–632.

  18. Giaccia A, Siim BG and Johnson RS . (2003). Nat. Rev. Drug Discov., 2, 803–811.

  19. Graeber TG, Osmanian C, Jacks T, Housman DE, Koch CJ, Lowe SW and Giaccia AJ . (1996). Nature, 379, 88–91.

  20. Guo X, Zhang YP, Mitchell DA, Denhardt DT and Chambers AF . (1995). Mol. Cell. Biol., 15, 476–487.

  21. Hammond EM, Denko NC, Dorie MJ, Abraham RT and Giaccia AJ . (2002). Mol. Cell. Biol., 22, 1834–1843.

  22. Harris AL . (2002). Nat. Rev. Cancer, 2, 38–47.

  23. Harrold S, Genovese C, Kobrin B, Morrison SL and Milcarek C . (1991). Anal. Biochem., 198, 19–29.

  24. Hijiya N, Setoguchi M, Matsuura K, Higuchi Y, Akizuki S and Yamamoto S . (1994). Biochem. J., 303, 255–262.

  25. Hockel M, Schlenger K, Aral B, Mitze M, Schaffer U and Vaupel P . (1996). Cancer Res., 56, 4509–4515.

  26. Huang L, Zhang H, Bae S and Liu A . (1994). J. Biol. Chem., 269, 30718–30725.

  27. Hwang SM, Wilson PD, Laskin JD and Denhardt DT . (1994). J. Cell Physiol., 160, 61–68.

  28. Kohn AD, Summers SA, Birnbaum MJ and Roth RA . (1996). J. Biol. Chem., 271, 31372–31378.

  29. Koong AC, Denko NC, Hudson KM, Schindler C, Swiersz L, Koch C, Evans S, Ibrahim H, Le QT, Terris DJ and Giaccia AJ . (2000). Cancer Res., 60, 883–887.

  30. Koszewski NJ, Reinhardt TA and Horst RL . (1996). J. Steroid Biochem. Mol. Biol., 59, 377–388.

  31. Le QT, Denko NC and Giaccia AJ . (2004). Cancer Metastasis Rev., 23, 293–310.

  32. Le QT, Sutphin PD, Raychaudhuri S, Yu SC, Terris DJ, Lin HS, Lum B, Pinto HA, Koong AC and Giaccia AJ . (2003). Clin. Cancer Res., 9, 59–67.

  33. Liaw L, Birk DE, Ballas CB, Whitsitt JS, Davidson JM and Hogan BL . (1998). J. Clin. Invest., 101, 1468–1478.

  34. Lowy DR and Willumsen BM . (1993). Annu. Rev. Biochem., 62, 851–891.

  35. Mazure NM, Chen EY, Laderoute KR and Giaccia AJ . (1997). Blood, 90, 3322–3331.

  36. Noda M, Vogel RL, Craig AM, Prahl J, DeLuca HF and Denhardt DT . (1990). Proc. Natl. Acad. Sci. USA, 87, 9995–9999.

  37. Noiri E, Dickman K, Miller F, Romanov G, Romanov VI, Shaw R, Chambers AF, Rittling SR, Denhardt DT and Goligorsky MS . (1999). Kidney Int., 56, 74–82.

  38. Pollmann C, Huang X, Mall J, Bech-Otschir D, Naumann M and Dubiel W . (2001). Cancer Res., 61, 8416–8421.

  39. Semenza GL . (2003). Nat. Rev. Cancer, 3, 721–732.

  40. Sodhi CP, Batlle D and Sahai A . (2000). Kidney Int., 58, 691–700.

  41. Sodhi CP, Phadke SA, Batlle D and Sahai A . (2001). Diabetes, 50, 1482–1490.

  42. Teicher BA, Holden SA, al-Achi A and Herman TS . (1990). Cancer Res., 50, 3339–3344.

  43. Teramoto H, Malek RL, Behbahani B, Castellone MD, Lee NH and Gutkind JS . (2003). Oncogene, 22, 2689–2697.

  44. Vaupel P, Thews O and Hoeckel M . (2001). Med. Oncol., 18, 243–259.

  45. Weber GF . (2001). Biochim. Biophys. Acta., 1552, 61–85.

  46. Weber GF and Cantor H . (1996). Cytokine Growth Factor Rev., 7, 241–248.

  47. Xie Y, Sakatsume M, Nishi S, Narita I, Arakawa M and Gejyo F . (2001). Kidney Int., 60, 1645–1657.

  48. Young SD and Hill RP . (1990). J. Natl. Cancer Inst., 82, 371–380.

  49. Zhang G, He B and Weber GF . (2003). Mol. Cell. Biol., 23, 6507–6519.

  50. Zhu Y, Hon T, Ye W and Zhang L . (2002a). Cell Growth Differ., 13, 431–439.

  51. Zhu Y, Hon T and Zhang L . (1999). Biochem. Biophys. Res. Commun., 258, 87–93.

  52. Zhu Y, Sun Y, Mao XO, Jin KL and Greenberg DA . (2002b). Neuroscience, 110, 191–198.

  53. Zhu Y, Sun Y, Xie L, Jin K, Sheibani N and Greenberg DA . (2003). Stroke, 34, 2483–2488.

Download references

Acknowledgements

This work was supported by the USPHS Grant CA 67166 (YZ, AJG, QTL) and the Damon Runyon–Lilly Clinical Investigator Award (ACK).

Author information

Correspondence to Quynh-Thu Le.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • Akt
  • hypoxia
  • osteopontin
  • ras
  • regulation

Further reading