|Home | About | Journals | Submit | Contact Us | Français|
Hexavalent chromium (Cr(VI)) promotes lung injury and pulmonary diseases through poorly defined mechanisms that may involve the silencing of inducible protective genes. The current study investigated the hypothesis that Cr(VI) actively signals through a signal transducer and activator of transcription 1 (STAT1)–dependent pathway to silence nickel (Ni)–induced expression of vascular endothelial cell growth factor A (VEGFA), an important mediator of lung injury and repair. In human bronchial airway epithelial (BEAS-2B) cells, Ni-induced VEGFA transcription by stimulating an extracellular regulated kinase (ERK) signaling cascade that involved Src kinase–activated Sp1 transactivation, as well as increased hypoxia-inducible factor-1α (HIF-1α) stabilization and DNA binding. Ni-stimulated ERK, Src, and HIF-1α activities, as well as Ni-induced VEGFA transcript levels were inhibited in Cr(VI)-exposed cells. We previously demonstrated that Cr(VI) stimulates STAT1 to suppress VEGFA expression. In BEAS-2B cells stably expressing STAT1 short hairpin RNA, Cr(VI) increased VEGFA transcript levels and Sp1 transactivation. Moreover, in the absence of STAT1, Cr(VI), and Ni coexposures positively interacted to further increase VEGFA transcripts. This study demonstrates that metal-stimulated signaling cascades interact to regulate transcription and induction of adaptive or repair responses in airway cells. In addition, the data implicate STAT1 as a rate limiting mediator of Cr(VI)-stimulated gene regulation and suggest that cells lacking STAT1, such as many tumor cell lines, have opposite responses to Cr(VI) relative to normal cells.
Chronic inhalation of hexavalent chromium (Cr(VI)) and nickel (Ni) are well-known environmental and occupational hazards that promote pulmonary diseases (Bright et al., 1997; Kelleher et al., 2000; Leikauf, 2002; O'Flaherty, 1998; Rice et al., 2001). Mixed exposures are common in metal industries (Antonini et al., 2004, 2007; Bright et al., 1997; Novey et al., 1983; Takemoto et al., 1991) and human epidemiological studies have associated mixed exposures with increased risks of lung diseases compared with exposure to the individual metals (Antonini et al., 2004; Bright et al., 1997; Hisatomi et al., 2006; Kuo et al., 2006; Takemoto et al., 1991). However, there is limited mechanistic understanding of the cellular and molecular pathogenic actions of the individual metals or their interactions in airway epithelial cell injury and repair processes.
Vascular endothelial growth factor A (VEGFA) expression is well established in vascular biology for its role in angiogenesis and vascularization. However, its role in the airway epithelium and lung injury repair remains unclear. Epithelial cells are the major source of VEGFA in the lung (Acarregui et al., 1999; Christou et al., 1998) and VEGFA is crucial for normal lung development (Acarregui et al., 1999) and airway epithelial cell proliferation (Brown et al., 2001). Yet, its role in lung injuries and disease remains controversial (Mura et al., 2004). Elevated VEGFA promotes pulmonary edema in the initial phase of inflammatory diseases, such as asthma, chronic bronchitis, and acute lung injury (ALI) (Kanazawa et al., 2003; McColley et al., 2000; Mura et al., 2004). However, in later phases, increased VEGFA levels are associated with the resolution of ALI (Thickett et al., 2002). VEGFA levels are decreased in patients with bronchopulmonary dysplasia, emphysema, and during ischemia and reperfusion correlating to epithelial cell damage (Fehrenbach et al., 2003; Kanazawa et al., 2003; Lassus et al., 1999). Moreover, in both in vivo skin and in vitro lung epithelial cell models, VEGFA promotes wound repair and elicits anti-apoptotic responses (Boussat et al., 2000; Frank et al., 1995; Roberts et al., 2007).
The main inducers of VEGFA expression are hypoxia (Namiki et al., 1995) and other stimuli that stabilize hypoxia-inducible protein-1α (HIF-1α), which transactivates the VEGFA promoter (Semenza, 2000). These stimuli include a variety of metals, including Ni (Andrew et al., 2001; Davidson et al., 2006; Ouyang et al., 2005; Salnikow et al., 1999). Ni stabilizes HIF-1α primarily by direct inhibition of the prolyl hydroxylases that mark HIF-1α for proteosomal degradation (Davidson et al., 2006). However, Ni can also signal through extracellular regulated kinase (ERK) and phosphoinositide 3-kinase (PI3K) to increase both HIF-1α stability and its transactivation of genes (Andrew et al., 2001; Ouyang et al., 2005). HIF-1α cooperates with other transcription factors (e.g., Sp1, activator protein 1 [AP-1], signal transducer and activator of transcription 3 [STAT3]) (Gray et al., 2005; Richard et al., 1999; Salnikow et al., 2002) and Sp1 transactivation is required for full induction of the VEGFA promoter (Pages and Pouyssegur, 2005). Unlike STAT3, STAT1, stimulated by interferons (IFNs), is a negative regulator of VEGFA (Gimeno et al., 2005; Tanabe et al., 2005) and the loss of STAT1 results in IFN stimulating transcriptional activation of the VEGFA promoter (Battle et al., 2006; Huang et al., 2002; von Marschall et al., 2003). Once STAT1 is activated, it forms the complex, interferon-stimulated gene factor 3 (ISGF3) that inhibits IFN-β–stimulated genes by interfering with the assembly of the transcriptional machinery on the promoters (Laver et al., 2008; Zhao et al., 2007).
Exposure to Cr(VI) silences the induction of protective genes induced by polycyclic aromatic hydrocarbons and metals by activating signaling pathways that alter transcriptional activity (Majumder et al., 2003; O'Hara et al., 2006; Shumilla et al., 1999; Wei et al., 2004). We previously reported that Cr(VI) stimulates STAT1 phosphorylation and nuclear translocation in human airway epithelial cells (O'Hara et al., 2007). Therefore, the current studies examined the hypothesis that Cr(VI) suppresses Ni-induced VEGFA expression by stimulating STAT1-dependent gene repression.
Human bronchial epithelial cells (BEAS-2B) (ATCC, Manassas, VA) and the stable BEAS-2B cell lines expressing scrambled negative control (shNC) or STAT1 short hairpin RNA (shRNA) (shSTAT1) were cultured on a matrix of 0.01 mg/ml of human fibronectin (Invitrogen, Carlsbad, CA), 0.029 mg/ml Vitrogen 100 (COHESION, Inc., Palo Alto, CA), and 0.01 mg/ml bovine serum albumin (BSA) (Invitrogen) in LHC-9 medium (Invitrogen). The parental BEAS-2B cells were cultured in LHC-9 media and the stable transfected cell lines were cultured in the same medium supplemented with 75 μg/ml G418 (Sigma-Aldrich, St Louis, MO). The cells were maintained at 37°C under an atmosphere of 5% CO2 as described previously (Barchowsky et al., 1998). Experiments were performed on 1-day postconfluent cells unless otherwise indicated. Medium was changed 12–16 h prior to all experiments. Under these conditions, the BEAS-2B responses to chromium were similar to responses in primary human bronchial epithelial cells grown in air/liquid interface cultures (O'Hara et al., 2007).
Human STAT1 shRNA sequences were designed using the Insert Design Tool for pSilencer Vectors at www.ambion.com and was subcloned into the pSilencer 4.1-CMV neo vector (Applied Biosystems, Foster City, CA) using BamHI and HindIII sites. The resultant clones were sequenced at the Genomics and Proteomics Core Laboratories at the University of Pittsburgh to verify the presence of the shRNA. A negative control expressing scrambled shRNA in the pSilencer 4.1-CMV neo vector was purchased from Applied Biosystems. BEAS-2B cells were transfected with either the scrambled shRNA or the STAT1 shRNA using Lipofectamine and PLUS reagents (Invitrogen) according to the manufacturer's instructions. Transfected cells were selected with G418 (75 μg/ml) (Sigma-Aldrich) and three cell lines were generated for each shRNA. STAT1 knockdown was confirmed by Western analysis.
Cr(VI) and NiSO4 (Ni) solutions were prepared fresh from potassium dichromate and nickel(II) sulfate hexahydrate (Sigma-Aldrich), respectively. Cells were treated with 5μM Cr(VI) or 200μM Ni for the indicated times. For coexposures, Cr(VI) was added to cells 2 h prior to adding Ni. There was no medium change or Cr(VI) washout between additions. These exposure levels are relevant to occupational exposures and are based on our previous demonstration of effective increases in cell signaling for gene expression changes without cytotoxicity. In certain experiments, cells were pretreated with 10μM U0126, 20μM SB203580, or 10μM PP2 for 30 min, or 1μM wortmannin for 90 min. To inhibit protein synthesis, cells were pretreated with 10 μg/ml cycloheximide (Sigma-Aldrich) for 5 min.
Total RNA was isolated using Trizol reagent (Invitrogen) and quantified by measuring absorbance (260 nm). Total RNA was reverse transcribed and the cDNA was amplified by real-time PCR (MJ Research Opticon 2; BioRad Laboratories, Hercules, CA), as described previously (O'Hara et al., 2006). The following primers sets were used: VEGFA (forward 5′-CTTGCCTTGCTGCTCTACCT-3′; reverse 5′-GCAAGGCCCACAGGGATTTT-3′), and ribosomal protein L13A (RPL13A) (forward 5′-CGAGGTTGGCTGGAAGTACC-3′; reverse 5′-ATTCCAGGGCAACAATGGAG-3′). Gene expression was quantified using standard curves for the respective cDNA products. All changes in VEGFA cDNA levels were normalized to changes in RPL13A. Data are presented as mean ± SEM of fold control.
Conditioned medium was collected and stored at −80°C until use. VEGFA(165 and 121) content in the medium was analyzed using specific enzyme-linked immunosorbent assay (ELISA) kits obtained from R&D Systems (Minneapolis, MN) according to the manufacturer's instructions. VEGFA protein release was normalized to total cellular protein from each individual sample.
Cells were rinsed twice in stop buffer (10mM Tris-HCl, pH 7.4, 10mM ethylenediaminetetraacetic acid [EDTA], 5mM ethylene glycol-bis(β-aminoethyl ether)tetraacetic acid, 0.1M NaF, 0.2M sucrose, 100μM sodium orthovanadate, 5mM pyrophosphate) and scraped in boiling lysis buffer (20mM Tris, pH 7.5, 1% sodium dodecyl sulfate [SDS], 100μM sodium orthovanadate, and supplemented with protease inhibitors) or modified RIPA buffer (50mM Tris-HCl, pH 7.6, 150mM NaCl, 1mM EDTA, 10mM NaF, 1% Triton X-100, 0.1% SDS, and supplemented with protease inhibitors and sodium orthovanadate) for analysis of HIF-1α or phosphorylated ERK, respectively. Protein concentrations were determined by the absorbance (595 nm) after the addition of Coomassie blue dye (Thermo-Fisher Scientific, Pittsburgh, PA) using BSA as a reference standard.
Cells were rinsed twice in stop buffer and scraped in modified RIPA buffer. Lysates were incubated for 30 min on ice and then sonicated three times at 5-s intervals. Lysates were centrifuged at 13,000 × g for 10 min at 4°C and the supernatants were collected. Equal amounts of protein were incubated with the antibody against total Src overnight at 4°C on a rotating platform. Protein A/G beads (Thermo-Fisher Scientific, Pittsburgh, PA) were added and incubated for an additional 3 h at 4°C. The beads were collected by centrifugation at 13,000 × g for 1 min, rinsed three times with modified RIPA buffer, suspended in 2× sample buffer, and boiled for 5 min.
To determine protein abundance, total cell lysates were resolved by SDS-PAGE and transferred to PVDF membranes (Millipore, Billerica, MA). Membranes were blocked and incubated overnight with primary antibodies at 4°C. HIF-1α antibody was obtained from BD Biosciences (San Jose, CA). Total STAT1, phospho-Tyr416-Src family kinase, total Src, phospho-p44/42 MAPK (pERK1/2), and total p44/42 MAPK (ERK1/2) were obtained from Cell Signaling Technology, Inc. (Danvers, MA). β-Actin was obtained from Sigma-Aldrich. Antibody binding was detected with horseradish peroxidase–conjugated antibodies (GE Healthcare, Piscataway, NJ) and enhanced chemiluminescence (PerkinElmer, Boston, MA). Reactive bands were quantified using ImageJ software (http://rsweb.nih.gov/ij/index/html). Data are presented as mean ± SEM of fold control.
Cells (70–80% confluence) were transfected with HRE-luc or Sp1-luc plasmid reporter constructs and enhanced green fluorescence protein (eGFP) plasmid using Lipofectamine and PLUS reagents (Invitrogen) according to the manufacturer's instructions. HRE-luc and Sp1-luc plasmids were kind gifts from Konstantin Salnikow and have been described previously (Salnikow et al., 1999). Cells were lysed and luciferase assays were performed as described previously (O'Hara et al., 2006). Data are presented as mean ± SEM of fold change over control.
One-way ANOVA was used to determine whether the mean of each treatment was different from the untreated cells (control). Dunnett's or Tukey's multiple comparisons post hoc tests were used to determine significant differences between the mean of each group. All statistics were performed using GraphPad Prism version 5 (GraphPad Software, San Diego, CA). Data are represented as mean ± SEM or as fold control.
Cr(VI) often suppresses gene inducibility, including induction of protective genes in the lung and metal-stimulated gene induction in cultured lung cells (O'Hara et al., 2006). The data in Figure 1 are consistent with this observation and confirm that Cr(VI) represses both Ni-induced VEGFA mRNA and protein release in BEAS-2B cells, but has no effect on basal expression.
To identify the mechanism for the negative interaction between Cr(VI) and Ni in the induction of VEGFA, we first characterized the Ni-stimulated signaling cascades leading to this induction. Ni activates both ERK- and HIF-1α–signaling pathways in BEAS-2B cells (Andrew et al., 2001) and both pathways are implicated in VEGFA transactivation (Duyndam et al., 2001; Gao et al., 2002; Ouyang et al., 2005). To identify the role of the ERK relative to other kinases in Ni-induced VEGFA expression, BEAS-2B cells were pretreated with inhibitors of ERK (U0126), p38 (SB203580), PI3K (wortmannin), or Src family kinases (SFK) (PP2) prior to exposure to Ni. Inhibition of ERK and SFK prevented Ni-induced VEGFA mRNA level whereas inhibiting p38 and PI3K had no effect (Figs. 2A and 2B). Note that there is no statistical difference between VEGFA transcript levels in the presence of Ni plus U0126 or PP2 relative to U0126 or PP2 alone. This would suggest that the two inhibitors are equally effective in providing a complete or at least equal block of Ni-induced VEGFA. ERK appeared to be upstream of SFK activation, because Ni-stimulated Src phosphorylation was absent in cells pretreated with U0126 (Fig. 2C). As demonstrated for Ni induction of the SEPRINE1 promoter (Andrew et al., 2001), ERK was required for maximal Ni-stimulated HIF-1α stabilization and transactivation of VEGFA (Fig. 3A). In addition, Src was also found to be essential for Ni induction of the gene (Fig. 2B), but not for Ni-stimulated HIF-1α stabilization (Fig. 3B). These data indicate that both HIF-1α and Src are required for Ni-induced VEGFA mRNA expression and that they are divergent pathways downstream of ERK. The VEGFA promoter contains numerous response elements that might be targets of ERK signaling, including Sp1 (Curry et al., 2008; Pages and Pouyssegur, 2005; Reisinger et al., 2003). A role for Sp1 in Ni-induced transactivation was demonstrated by Ni stimulating the activity of a Sp1-driven luciferase reporter construct by 3.764 ± 0.6853 fold compared with untreated cells. Together, these data suggest that the ERK-regulated pathways activated by Ni are both capable of functional gene activation that converge to induce the VEGFA promoter.
Exposure to Cr(VI) had no effect on the basal ERK or Src phosphorylation states (Fig. 4), and Cr(VI) did not affect basal HIF-1α protein levels or the activity of hypoxic response element (HRE)–driven luciferase reporter construct (Fig. 5). In contrast, Cr(VI) inhibited Ni-stimulated ERK and Src activation (Fig. 4), HIF-1α protein expression (Fig. 5A), and transactivation of the HRE reporter construct (Fig. 5B). The pattern of inhibition is consistent with Cr(VI) preventing Ni from activating ERK, the upstream kinase of the divergent signaling pathways.
STAT1 negatively regulates VEGFA induction (Battle et al., 2006) and transactivation by Sp1 (Laver et al., 2008). To investigate a role for Cr(VI)-stimulated STAT1 in repressing VEGFA inducibility, we generated BEAS-2B cell lines stably expressing either scrambled (shNC) or STAT1 (shSTAT1) shRNA. Western blot analysis in Figure 6A verifies STAT1 protein knockdown in shSTAT1 cells relative to shNC cells. In contrast to the response of parental BEAS-2B or the shNC cells, VEGFA mRNA levels increased in shSTAT1 cells within 4 h of Cr(VI) exposure (Fig. 6C). In addition, a 2 h pretreatment of shSTAT1 cells with Cr(VI)-enhanced VEGFA mRNA levels induced by Ni (Fig. 6C), whereas the same treatment in shNC cells resulted in the same Cr(VI) inhibition of VEGFA inducibility observed in the parental BEAS-2B cells (Figs. 1A and and6B).6B). In addition, Cr(VI) had no effect on Ni-induced HIF-1α protein in the shSTAT1 cells relative to its effect in shNC cells (Fig. 7A). Finally, Cr(VI) increased the activity of the Sp1-driven luciferase reporter construct in shSTAT1 cells, whereas there was no Cr(VI) effect on luciferase activity in the shNC cells (Fig. 7B). The STAT1-containing complex, ISGF3, is known to indirectly inhibit IFN-β–inducible genes most likely by activating or increasing expression of an inhibitory protein (Laver et al., 2008; Zhao et al., 2007). To examine this hypothesis that Cr(VI) stimulates ISGF3 induction of an inhibitor, parental BEAS-2B cells were exposed to Cr(VI) in the presence or absence of cycloheximide to inhibit protein synthesis. Data in Figure 6D confirm the effect of on basal VEGFA expression in control cells and demonstrates that Cr(VI) increased VEGFA mRNA levels when protein synthesis was inhibited. These data suggest that Cr(VI) signals through STAT1 to repress VEGF mRNA expression and that STAT1 may exert its inhibitory effects by increasing expression of protein repressor of VEGFA induction.
Environmental and occupational exposure to Cr(VI) and Ni cause pulmonary diseases (Kelleher et al., 2000; Leikauf, 2002; O'Flaherty, 1998; Rice et al., 2001) and although epidemiological studies associate exposure to metal mixtures with exacerbated lung injury (Antonini et al., 2004; Bright et al., 1997; Hisatomi et al., 2006; Kuo et al., 2006; Takemoto et al., 1991), there are few studies investigating the molecular interaction of these metals. Cr(VI) readily enters airway cells through anion channels where it is rapidly reduced to Cr(III) (reviewed in, Salnikow and Zhitkovich, 2008) and exerts effects on both cell signaling processes and DNA. Exposure to Cr(VI) rarely affects the expression of constitutive genes (Hamilton and Wetterhahn, 1989; Wei et al., 2004), but silences inducible gene expression in vivo and in vitro (Hamilton and Wetterhahn, 1989; Hamilton et al., 1998; O'Hara et al., 2006; Wetterhahn and Hamilton, 1989). Although there are substantial amounts of data in the literature suggesting that Cr(VI)-induced DNA adducts and reactive oxygen species generated through the reduction of Cr(VI) are responsible for gene silencing (Hamilton et al., 1998; O'Brien et al., 2003; Reynolds et al., 2004; Wetterhahn and Hamilton, 1989), there is also ample evidence suggesting that Cr(VI) exerts epigenetic effects in silencing gene induction through altering transcriptional complexes (Majumder et al., 2003; O'Hara et al., 2006; Shumilla et al., 1999; Wei et al., 2004). In the present studies, we examined the effect of Cr(VI) on basal and Ni-induced signal transduction and VEGFA transcript levels in a pertinent lung target cell using concentrations relevant to occupational exposures. These concentrations are not cytotoxic in this cell model ([O'Hara et al., 2007] and data not shown), but promote both positive and negative signaling effects (O'Hara et al., 2006, 2007). These studies are the first to identify signaling through STAT1 as an essential regulatory mechanism for Cr(VI)-stimulated gene suppression. These data may also explain the discrepancies observed in other studies where Cr(VI) induced VEGFA in cancer cells whose functional STAT1 signaling status is questionable (Gao et al., 2002).
A major stimulus of Ni-induced VEGFA mRNA and protein is HIF-1α stabilization (Andrew et al., 2001; Ouyang et al., 2005; Salnikow and Zhitkovich, 2008). However, HIF-1α signaling could not be the dominant or sole pathway for Ni-induced VEGFA in the BEAS-2B cell model, because inhibiting Src completely blocked VEGFA mRNA without affecting HIF-1α stability (Figs. 2B and and3B).3B). Although ERK, p38, PI3K, and Src have all been implicated as upstream kinases in VEGFA induction in response to a variety of stimuli (Eliceiri et al., 1999; Gao et al., 2002; Ouyang et al., 2005), our data indicates Ni signals through ERK and Src to induce VEGFA (Figs. 2A and 2B). Ni-stimulated ERK phosphorylation within 5 min of exposure in contrast to Src phosphorylation which required a 30-min exposure to Ni (Fig. 2 and data not shown) indicating that ERK is upstream of Src. Given that ERK is a serine/threonine kinase and the temporal lag in Src activation, it is likely that there are intermediate steps between the two kinases. Investigating there signaling events is beyond the scope of this paper and it is unlikely that these signaling steps are important targets of Cr(VI). Furthermore, the upstream dominance of ERK was demonstrated by U0126 inhibiting both Src phosphorylation (Fig. 2C) and HIF-1α stabilization (Fig. 3A). HIF-1α cooperates with other transcription factors to transactivate a number of different promoters (Andrew et al., 2001; Barchowsky et al., 2002; Salnikow et al., 2002) and the data are consistent with cooperation with Sp1 to produce full ERK-mediated induction of VEGFA (Pages and Pouyssegur, 2005) following Ni exposure. Although the data implicate ERK as the primary divergence point for the cooperating Ni-stimulated signaling pathways, these studies did not investigate the mechanism for Ni-stimulated phosphorylation of ERK or potential interactions of Ni and Cr(VI) signaling on upstream kinases. In addition, Ni activation of ERK may be dependent on cell type or cell culture conditions because others fail to observe Ni stimulation of ERK when basal levels are high or in transformed airway epithelium (Ke et al., 2008).
Cr(VI) completely blocked both Ni-induced ERK and Src phosphorylation and VEGFA mRNA expression (Figs. 1 and and2).2). We previously demonstrated that Cr(VI) had an inhibitory effect on both ERK and Src activities (O'Hara et al., 2003) and these data suggest that Cr(VI) interferes with their activation by Ni to prevent VEGFA induction. However, Cr(VI) only partially inhibited Ni-induced HIF-1α protein stabilization (Fig. 5A), because Ni-induced HIF-1α stabilization mostly results from a direct inhibitory effect on the proline hydroxylases marking the protein for ubiquitination and degradation (Davidson et al., 2006). However, Ni-stimulated activity of the HRE reporter construct was completely inhibited by Cr(VI) (Fig. 5B). The HRE reporter construct does not contain additional cis elements that might cooperate with HIF-1α binding to the HRE sites. Thus, the data are consistent with previous observations that phosphorylation by ERK-mediated signaling cascades is required for full transactivation potential of Ni-activated HIF-1α (Andrew et al., 2001). Moreover, Cr(VI) had a greater effect on Ni-activated ERK and Src compared with HIF-1α indicating that ERK-mediated Src signaling and Sp1 transactivation are more crucial for Ni-induced VEGFA.
STAT1 is a member of the STAT transcription factor family. Although STAT3 has been implicated as an upstream inducer of VEGFA (Gray et al., 2005), STAT1 is required for IFN-mediated inhibition of VEGFA (Battle et al., 2006; Rosewicz et al., 2004; von Marschall et al., 2003). Cr(VI) activates STAT1 phosphorylation and nuclear translocation in parental BEAS-2B cells; although the mechanism of this activation remains uncharacterized (O'Hara et al., 2007). The current findings support the hypothesis that STAT1 activation is essential for Cr(VI) to repress gene expression. In addition, they also demonstrate that STAT1 activation blocks the Cr(VI) signals that induce VEGFA in shSTAT1 cells or that positively interact with Ni signaling (Figs. 6B and 6C). It is likely that STAT1 suppresses the signaling for activation of factors that cooperate with HIF-1α or sequesters these factors by STAT1-induced protein complexes to limit Cr(VI) induction of VEGFA in the parental cells. Recent evidence suggests that IFN-β activated a STAT1-containing ISGF3 complex to inhibit Sp1 recruitment or binding of Sp1 and coactivators (e.g., CBP, p300) to inducible promoters (Laver et al., 2008; Zhao et al., 2007). The exact mechanism of this inhibition is uncharacterized. Neither ISGF3 nor any of the individual protein components bind directly to the promoters to block transcription, but rather, it is possible that STAT1 stimulates an inhibitory protein that interferes with the binding of Sp1. In the absence of protein synthesis, Cr(VI)-induced VEGFA mRNA expression (Fig. 6D); suggesting that a protein induced by STAT1 transcriptional complexes represses VEGFA induction. The identity of this protein remains unknown, but is being actively investigated. Because Cr(VI) stimulates Sp1 in the absence of STAT1 (Fig. 7B) and Sp1 activation is essential for VEGFA expression, it is possible that the induced inhibitory protein represses Sp1-dependent transactivation and induction of the VEGFA.
The antiviral and antiproliferative effects of IFN-α and IFN-β are limited or reversed to proliferative responses in cells that lack STAT1 (Huang et al., 2002; Tanabe et al., 2005). STAT1 is deleted in a number of cancers (Battle et al., 2006; Tanabe et al., 2005) and without STAT1 activation, IFN stimulation of STAT3 and 5 or their constitutive activation provides a cell survival advantage and can cause transformation (Huang et al., 2002; Tanabe et al., 2005; Xi et al., 2003). Others reported that Cr(VI) stabilizes HIF-1α protein and induces VEGFA in prostate cancer cells (Gao et al., 2002) and we show that Cr(VI) has opposite effects on HIF-1α protein and VEGFA in shNC and shSTAT1 cell lines (Figs. 6B and 6C and 7A). Cr(VI) also has opposite effects on HMOX1 inducibility in lung cell lines that differ in transformation status (Dubrovskaya and Wetterhahn, 1998; O'Hara et al., 2006). Thus, an implication of these findings is that, as with IFN responses, Cr(VI) has opposite signaling effects and pathogenic actions in normal cells compared with transformed cells that lack functional STAT1. In the normal cells, Cr(VI) would limit inducibility of protective genes or genes involved in injury repair. In transformed cells, Cr(VI) alone or in combination with other stimuli like Ni might promote proliferation and tumor growth by increasing VEGFA or other growth factor expression.
In summary, this study has identified STAT1 activation as an essential and pivotal mechanism in Cr(VI)-stimulated gene regulation. In normal cells exposed to metal mixtures, Cr(VI) interferes with Ni-stimulated ERK signaling to compromise the induction of protective genes which may exacerbate pulmonary diseases. Loss of STAT1-mediated transcriptional repression unmasks Cr(VI)-stimulated gene induction and enhancement of the Ni response. Thus, the data support the conclusion that pathogenic actions of Cr(VI) signaling in lung cells and interaction of this signaling with that of other metals in mixtures depend on the STAT1 status of the airway cells.
National Institutes of Health (ES10638 and HL06569).