Role of HIF Signaling on Tumorigenesis in Response to Chronic Low-Dose Arsenic Administration
http://www.100md.com
《毒物学科学杂志》
Department of Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma 73190
Center for Environmental Health Sciences and Department of Pharmacology and Toxicology, Dartmouth Medical School, Hanover, New Hampshire 03755
ABSTRACT
Trivalent inorganic arsenic (arsenite, arsenic trioxide, As(III)) is a primary contaminant of groundwater supplies worldwide. As(III), marketed as trisenox, is also an FDA-approved agent to treat cancer It has been previously shown by our laboratory that As(III) administered at doses lower than a therapeutic anticancer dose results in an increase in tumor formation and blood vessel density of tumors. In this work it was found that chronic administration of As(III) approaching the EPA action level of 10 ppb, given in the drinking water of mice 5 weeks prior to B16-F10 melanoma implantation, increased the growth rate of primary tumors and the number of metastases to the lung. Further, levels of arsenic in the tumor and lung were found to be much greater than those in the blood and similar to pro-angiogenic As(III) doses. Levels of hypoxia inducible factor-1 (HIF-1) and vascular endothelial growth factor (VEGF) surrounding the blood vessels in the tumors of the As(III)-treated mice were also found to be increased. Exposure of isolated B16-F10 tumor cells to chronic (3 or 7 day) but not acute (4 h) low-dose As(III) was found to increase HIF-1 expression and secretion of VEGF. Finally, coadministration of an inhibitor of HIF (YC-1) or a VEGFR-2 kinase inhibitor (SU5416) was found to antagonize the pro-angiogenic effects of low-dose As(III). Together, these results suggest that chronic exposure to low-dose As(III) could stimulate growth of tumors through a HIF-dependent stimulation of angiogenesis.
Key Words: arsenic; angiogenesis; hypoxia inducible factor; vascular endothelial growth factor; tumorigenesis.
INTRODUCTION
Arsenite (As(III)) is a major natural contaminant of drinking water throughout the world. As(III) is also an FDA-approved anticancer agent for leukemias and is in clinical trials for several solid tumor types. In the United States, the Environmental Protection Agency (EPA) lowered the action level arsenic from 50 parts per billion (ppb) to 10 ppb in 2001, due to take effect in 2006 (http://www.epa.gov/safewater/arsenic.html). As(III) exposure is associated with an increased incidence of a number of disease states, including vascular conditions such as Blackfoot disease (Tseng, 2002; Yu et al., 2002), diabetes (Tseng et al., 2002), hypertension (Rahman, 2002), and cancers such as those of the lung, bladder, and skin (Tchounwou et al., 2003). With respect to its vascular effects, As(III) has been shown to be anti-angiogenic at higher doses such as those used in the treatment of cancer (Davison et al., 2002; Roboz et al., 2000), whereas it has also been shown that lower, more environmentally relevant doses of As(III) are in fact pro-angiogenic (Kao et al., 2003; Soucy et al., 2003). It has also been shown by our group (Soucy et al., 2003) and others (reviewed in Rossman, 2003) that As(III) is tumorigenic.
Tumor-related angiogenesis is a very complex process involving several different cell types including tumor cells, endothelial cells, smooth muscle cells, or pericytes surrounding the endothelial cells and fibroblasts/keratinocytes. During angiogenesis, these cell types release and respond to multiple growth factors and digestive enzymes that facilitate cell growth, destabilization of existing vessels sprouting of new vessels, and restabilization of nascent vessels (reviewed in Auguste et al., 2005). Vascular endothelial growth factor (VEGF) has been shown to be a critical factor in tumor-related angiogenesis and is a primary therapeutic target for the treatment of cancer through angiogenesis inhibition (reviewed in Shinkaruk et al., 2003). VEGF is primarily regulated through hypoxia inducible factor (HIF), a heterodimeric transcription factor complex. As the name implies, HIF levels are induced by low oxygen conditions, principally through the inhibition of the proteasome-mediated degradation of HIF (reviewed in Covello and Simon, 2004). HIF signaling is also regulated by nitric oxide (NO) and other reactive oxygen species (ROS) such as superoxide and hydroxy free radical (reviewed in Chun et al., 2002). In addition to VEGF, HIF regulates a large panel of genes involved in angiogenesis (e.g., plasminogen activator inhibitor-1), RBC formation (erythropoietin, heme oxygenase-1), glucose regulation (glucose transporter-1, phosphoglucokinase-1), and NO signaling (inducible nitric oxide synthase). As(III) has been shown to induce levels of HIF-1, the regulated subunit of HIF (Duyndam et al., 2001) as well as levels of VEGF (Duyndam et al., 2003; Kao et al., 2003) in isolated cells.
Since As(III) exposure is associated with the formation of several ROS types (Hei and Filipic, 2004; Kumagai and Pi, 2004; Shi et al., 2004) and increases in NO signaling (Kumagai and Pi, 2004), it was reasoned that low-dose As(III) might be acting through HIF signaling to stimulate tumor-related angiogenesis. It has been previously shown by our laboratory that, when As(III) is given at doses five- or ten-fold lower than those used to treat cancer (0.5 and 1 mg/kg As(III) given biweekly), it can stimulate the growth of implanted tumors (Soucy et al., 2003). It is hypothesized that chronic exposure to environmental doses of As(III) leads to tumor growth through the stimulation of angiogenesis via HIF-mediated cellular signaling.
In this work, implanted tumors in mice exposed to As(III) for 5 weeks at levels approaching the drinking water action level of 10 ppb were shown to grow faster than those of animals not receiving As(III). In addition, the number of metastases to the lungs was found to be higher in animals receiving As(III). Localized levels of HIF-1 and VEGF surrounding the blood vessels were increased in tumors of animals receiving low-dose As(III). In isolated tumor cells, smooth muscle cells (SMCs), and endothelial cells, HIF-1 and VEGF levels were induced by chronic (72 h or 7 day) low-dose As(III) but not by acute (4 h) As(III). Finally, inhibitors of HIF-1 and VEGF attenuated the effect of As(III) on blood vessel density, implying that these two proteins are functionally involved in the angiogenic effect of low-dose As(III).
MATERIALS AND METHODS
Chemicals and cell culture reagents.
All chemicals were purchased from EMD Biosciences (San Diego, CA) unless otherwise noted. Sodium arsenite was purchased from Sigma Chemical, St. Louis, MO. Cell culture reagents were purchased from InVitrogen (Carlsbad, CA).
Cell culture.
B16-F10 mouse melanoma cells (Soucy et al., 2003), HMEC-1 human microvascular endothelial cells (Ades et al., 1992) (obtained from the Centers for Disease Control), and human J82 bladder tumor cells (American Type Culture Collection), were grown in Dulbecco's modified Eagle media supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), pyruvate, glutamine (Mediatech, Herndon, VA), and penicillin/streptomycin (Hyclone) in a humidified environment with 5% carbon dioxide added. H500Z human bladder smooth muscle cells (a kind gift of Dr. Bradley Kropp, OUHSC Department of Urology) were grown in M199 media (InVitrogen) with 10% fetal bovine serum (Hyclone), pyruvate, glutamine, and penicillin/streptomycin, 1% vitamins, 1 MEM % nonessential amino acid solution (InVitrogen, Grand Island, NY).
Chorioallantoic membrane (CAM) assay of angiogenesis.
The CAM assay that was used is based upon the original procedure by Jakob et al. (1978) with modifications made by our laboratory (Marks et al., 2002; Shi et al., 2003; Soucy et al., 2003). Fertile leghorn chicken embryos (CBT farms, Chestertown, MD), 24 h after laying, were incubated for 3–4 days with rocking and then placed horizontally for a minimum of 4 h at 37.2°C. Embryos were then cracked into 100 x 20 mm Petri dishes and allowed to grow at 37.2°C for one additional day. As(III) +/– YC-1 or SU5416 were placed at saturation onto 1.2-mm-diameter cellulose nitrate discs (Whatman) and onto a blood vessel (small arteriole; metarteriole) distal from the developing embryo. As(III) +/– inhibitors were replaced onto the disc at 24 and 48 h, and then at 72 h post-treatment, digital brightfield images were collected. The disc was centered in the visual field at 16x magnification, then removed using a camel's hair brush, and images were collected. Vascular density in the area of the disc was counted by overlaying a 4 x 4 grid on the 16x images and counting vessels per grid square in five grids. Statistical analysis was accomplished using unpaired two-tailed Student's t-test. Typically, 4–5 individual CAMs were used for each treatment and all experiments were repeated at least twice.
Mouse tumor studies.
All animal studies were conducted in accordance with AALAC approved guidelines using a protocol approved by IACUC at the University of Oklahoma. Six-week-old male NCr nu/nu mice (NCI APA breeding stock, Frederick MD) were housed in ventilated cages with autoclaved AIN-7A arsenic-characterized chow (Research Diets, New Brunswick, NJ), autoclaved water and autoclaved bedding. Following 1 week of acclimation, 0, 10, 50, or 200 ppb sodium arsenite was added to the drinking water of appropriate animals for 5 weeks with water changed biweekly. At 5 weeks after initial As(III) administration, 5 x 105 B16-F10 (GFP) cells in 100 μl PBS/l mM EDTA were injected subcutaneously into the external surface at the base of the right ear of all animals. Tumor measurement began 7 days after implantation and continued biweekly until tumors reached approximately 8–10% of animal body weight. The length (L) and width (W) of tumors were measured using Vernier calipers (Mitutoyo, Kawasaki Kanagawa, Japan), and tumor volume was calculated using the formula L x W2.
After sacrifice, the tumor and lungs were excised, fixed for 8 h in neutral buffered formalin, embedded in paraffin, sectioned, and stained by either H&E or standard immunohistochemical methods as previously described (Soucy et al., 2003). Sections were deparaffinized, incubated in 6% peroxide in 80% ethanol for 20 min, incubated in 0.1 M sodium citrate to retrieve the antigens, heated until boiling, and then heated for 10 additional min on 30% power using a 700 W microwave oven. Sections were then blocked using Vectastain kits (Vector Laboratories, Burlingame, CA), exposed to antibodies against GFP (Clontech; 1:150 dilution) to examine metastases, HIF-1 (Novus Biologicals monoclonal antibody H1alpha67, 1:50) or mouse VEGF (Abcam ab2992; 1:150) overnight at 4°C; then a secondary antibody and ABC reagent were added, and sections were stained with VIP peroxidase substrate (Vector Laboratories) and counterstained with Gomori Trichrome (Harleco EMD, San Diego, CA).
ICPMS arsenic measurement.
For these experiments, tumor-bearing mice described above were anesthetized using halothane, and then blood was removed, using heart puncture, to a heparin tube. Portions of tumor, lung, and heart were removed and snap-frozen in liquid nitrogen and stored at –30°C until analysis. Prior to analysis, the samples were digested with concentrated nitric acid in closed Teflon vessels in a microwave oven applying the following procedure: 200 mg of sample (blood, tumor, or organ) was weighed accurately into a 7-ml Teflon vial, and 1.00 ml of ultrapure concentrated nitric acid was added. The vial was closed and placed with up to 26 other samples in a microwave oven. A power of 300 W was applied for 10 min to a temperature of 130°C, which was held for 10 additional min. The resulting clear green/yellow solution was transferred to a 5-ml acid-cleaned polypropylene vial; the vial was capped and stored at 4°C until ICP-MS analysis. The arsenic contents of the digested samples were determined using hydride-generation inductively coupled plasma mass spectrometry (HG-ICPMS). Hydride generation was applied to eliminate interference from chloride, seen in the form of ArCl+ on the As+ signal (both seen of m/z = 75 in the mass spectrum). Applying hydride generation, the arsenic was converted to AsH3 and removed from the sample matrix (and chloride) before introduction to ICP-MS. General matrix interferences were eliminated by the use of standard addition calibration. As quality control, two blanks and one certified reference material were analyzed for every eight samples in order to monitor contamination and method precision and accuracy. Trace elements in whole blood and CRM 185R trace elements in bovine liver were used for quality control of blood and tumor/organs, respectively. The method detection limit is approximately 10 pg/ml in the sample solution corresponding to 0.25 ng/g (ppb) in blood and organs.
VEGF ELISA.
VEGF levels were assessed by ELISA using a rabbit polyclonal anti-VEGF antibody (Abcam, Cambridge, MA ). Supernatants from cells (100 μl) were added to a high binding white ELISA plate (COSTAR, Corning, NY), together with a standard curve of recombinant mouse VEGF (0.1–10,000 pg/ml) (Biosource International, Camarillo, CA) overnight at 4°C. The wells were washed with TBS, and anti-VEGF antibody (1:1000 dilution of 1 mg/ml stock) was added for 2 h at room temperature (RT). The wells washed again, and 1:2000 anti-rabbit HRP conjugate (Pierce, Rockford, IL) was added for 45 min at RT. The wells were washed again with TBS, then an ECL ELISA substrate (Pierce) was added to wells for 5 min, and the chemiluminescence was measured using a plate reader (Fluostar Optima, BMG Labtechnologies, NC). Data were normalized to the standard curve to yield values in pg/ml VEGF and graphed as percentage of untreated control. Significance among treatments was determined by unpaired two-tailed Student's t-test and two-way ANOVA with Bonferroni's post-test used to compare two dose-response curves, with the null hypothesis rejected when p < 0.05.
Western blot.
Equal amounts of lysates (20–50 μg) were separated by SDS-PAGE and transferred to nitrocellulose. Transfer was assessed by staining the membrane with 0.1% India ink in TBS with 0.1% Tween-20. Membranes were blocked 2 h in StartingBlock (TBS) Blocking Buffer (Pierce) with 0.1% Tween-20 (TBST). Membranes were washed in TBST and exposed to anti-HIF-1 (Chemicon: 1:500) antibody overnight at 4°C or 2 h at RT with rocking. Membranes were washed in TBST, secondary anti-rabbit HRP-conjugate (Pierce) diluted in StartingBlock (TBS) Blocking Buffer was added at 1:2500 dilution for 1 h at RT, the membranes were washed again in TBST, and then ECL reagent (Pierce) was added for 5 min, and images were captured using a CCD digital darkroom (UVP, Upland, CA).
Construction of the HRE-SEAP reporter construct.
To create the HRE-SEAP reporter, a double-stranded sequence containing the HRE (in bold) and flanking sequence from the human VEGF-165 gene (Liu et al., 1995) (5'AGATCTCCACAGTGCATACGTGGGCTCCAACACGTGGCTCTTCAGGAATTC-3'; upper strand) were created by annealing two complementary oligonucleotides with unique restriction sites (underlined) together. This sequence was then placed into the multiple cloning site of an empty SEAP reporter vector (Clontech, Palo Alto, CA). This construct was confirmed by sequencing using the T7 site of the vector.
Transfection of B16-F10 cells with the HRE-SEAP reporter construct.
Stable transfectants of the above HRE-SEAP reporter gene were made using liposome-mediated transfection using Fugene reagent (Roche, Indianapolis, IN) and selection under G418 pressure. A viable single clone colony was cultured and used for the SEAP assay.
Secreted alkaline phosphatase (SEAP) reporter assay.
B16-F10 cells transfected with the HRE-SEAP reporter gene and cultured in 96-well plates (Corning, NY) were exposed to 0.01 or 0.1 μM As(III) alone or with YC-1 (7.5 μM ) and a SEAP assay performed as previously described (Cullen and Malim, 1992). Briefly, 40 μl supernatant was removed from cells and plated into a 96-well plate, 6 μl 10x TBS added, and samples were incubated at 65°C for 30 min to inactivate endogenous alkaline phosphatase. Samples were placed on ice for 10 min. Then a chemifluorescent alkaline phosphatase substrate (CDP Star; Perkin-Elmer, Boston, MA) was added for 5 min, and images were captured by a CCD digital darkroom (UVP) at various times of exposure. Data were graphed as percentage of untreated control.
Statistics.
All analysis was done using Prism 4.0 (GraphPad Software, San Diego, CA). One-way ANOVA was conducted, with multiple comparisons tested by Bonferroni's post-hoc test at a significance level of 0.05 or greater. Tumor volumes were compared using repeated measures two-way ANOVA with treatments and days after implantation as independent variables with multiple comparisons tested by Dunnett's post-hoc test at a significance level of 0.05 or greater.
RESULTS
Chronic Low-Dose As(III) Exposure through Drinking Water Increases Tumor Growth and Metastasis
It has previously been shown by our laboratory that low-dose As(III) stimulates angiogenesis and tumorigenesis when administered IP in biweekly doses, similar to those used for anticancer treatment (Soucy et al., 2003). The hypothesis that chronic low-dose As(III) in drinking water stimulates tumor growth was tested using a classical mouse syngeneic tumor model, the B16 melanoma. Animals were exposed to 10, 50, or 200 ppb As(III) in their drinking water for 5 weeks; then B16-F10 (GFP) tumor cells were implanted into the skin around the ear of the mice (Barkhordar et al., 1995). All doses of As(III), including the U.S. EPA action level of 10 ppb (Fig. 1A, diamonds, and Fig. 1B, second bar from left) and the old (grandfathered) standard of 50 ppb (Fig. 1A, open squares, and Fig. 1B, second bar from right) resulted in increases in growth of implanted B16-F10 (GFP) tumors. To examine whether low-dose As(III) affected the metastasis of the lung-colonizing B16-F10 (GFP) cells (Fidler et al., 1976), lungs from the tumor-bearing mice were subjected to immunohistochemical staining for the GFP epitope transfected into the tumor cells (Soucy et al., 2003). Both 10 and 200 ppb As(III) in the drinking water significantly increased the average number of B16-F10 (GFP) metastases (Fig. 1C). In addition, the size of the metastases in animals exposed to 10 or 50 ppb As(III) was greater than in animals receiving no As(III) in their drinking water (data not shown). Thus chronic exposure to low-dose As(III) is capable of increasing growth of both primary tumors and secondary metastases in an in vivo syngeneic tumor model.
Arsenic Blood and Organ Levels
It was next determined what the levels of arsenic were in the blood and tissues from the tumor-bearing mice described above. Organs were snap-frozen, and arsenic levels were measured by ICPMS as previously described (Schmeisser et al., 2004). It is striking how low the levels of arsenic in the blood (Fig. 2, left-hand grouping) are as compared to the various organs (Fig. 2, middle and right groupings) and that, in most of the organs, As(III) levels increase incrementally with dose. It is also interesting that in the skin-tumor (melanoma surrounded by skin layer), the average values of the arsenic organ levels were similar to the dose (10, 50, and 200 ppb). In addition, these levels of arsenic in the tumors were within the angiogenic range found in the chorioallantoic angiogenesis assay in our previous paper (Soucy et al., 2003).
Chronic Low-Dose As(III) Stimulates HIF-1 and VEGF Levels Surrounding the Blood Vessels in Implanted Tumors
In our previous paper (Soucy et al., 2003) it was shown that low-dose (0.01–1 μM) As(III) increases blood vessel density in B16-F10 tumors in addition to increasing angiogenesis in two separate in vivo assays. VEGF is arguably the most important regulator of angiogenesis (McColl et al., 2004), and HIF is a primary regulator of VEGF (Adelman et al., 1999). High-dose As(III) (50–100 μM; 3750–7500 ppb) has been shown to induce HIF-1, the inducible subunit of HIF (Duyndam et al., 2001), whereas low-dose As(III) has been shown not to induce HIF-1 in smooth muscle cells (Soucy et al., 2004). To examine whether HIF signaling could be altered in the tumors of mice receiving chronic low-dose As(III), tumors were removed, tissue lysates made and a Western blot against HIF-1 completed. Barely detectable HIF-1 levels were observed in whole tumor lysates, with no change in HIF-1 levels in response to any dose of As(III) (Fig. 3A). To determine whether localized changes in HIF-1 could be occurring, sections of primary melanoma tumors were subjected to immunohistochemical staining. The results of representative sections against HIF-1 are shown in Figure 3B (top row) and against VEGF, a major angiogenic protein controlled in part through HIF signaling (Adelman et al., 1999) (Fig. 3B, bottom row). Chronic low-dose As(III) stimulated the expression of both HIF-1 and VEGF, but there was no apparent dose-dependent effect. Tumors of animals exposed to chronic low-dose As(III) in their drinking water had substantially increased HIF-1 staining; however, this staining was limited to areas just surrounding the blood vessels (Fig. 3B, top panel). Similarly, VEGF staining was also only observed close to the blood vessels in response to chronic low-dose As(III).
Chronic Low-Dose As(III) Stimulates HIF-1, HRE Transactivation, and VEGF Levels in Isolated B16-F10 Tumor Cells
It was next desired to see whether chronic low-dose As(III) exposure resulted in altered HIF-1 levels in isolated B16-F10 cells. Chronic (72 h or 7 day) exposure to 0.01 μM (7.5 ppb) As(III) resulted in significant increases in HIF-1 protein levels, whereas higher doses of As(III) resulted in decreased HIF-1 levels (Fig. 4A). In contrast, acute (4 h) exposure to As(III), from 0.01 to 10 μM, did not result in any increase in HIF-1 levels (data not shown).
It was next examined whether these changes in HIF-1 would result in increases in the HIF-regulated protein, VEGF, a critical regulator of angiogenesis. As(III) exposure (4 h), followed by 20 h in As(III)-free media, resulted in a small increase in VEGF (Fig. 4B, left grouping), particularly at the higher doses of As(III), as determined by ELISA. Chronic (72 h or 7 day) exposure to low-dose (0.01 μM) As(III) (Fig. 4B, middle and right groupings) resulted in a more pronounced increase in VEGF as compared with acute As(III) exposure, whereas chronic high-dose As(III) (10 μM for 72 h or 0.1μM for seven days) resulted in decreased VEGF levels. These VEGF levels again correlate well to the pro- and anti-angiogenic effects of low (0.01–1 μM) and high (>3.33 μM) dose As(III) observed in the CAM assay in our previous paper (Soucy et al., 2003).
Because VEGF could be regulated through a factor other than HIF, we investigated the effect of chronic low-dose As(III) on HIF-mediated transactivation through its DNA binding element, the hypoxia response element (HRE). In these experiments, the B16-F10 cells were transfected with two HREs linked to a secreted alkaline phosphatase (SEAP) reporter gene and stable transfectant clones isolated. Both 4 h cobalt chloride (CoCl2), a known HIF inducer (Wang and Semenza, 1993) and seven days of either 0.01 μM or 0.1 μM As(III) resulted in significant increases in HRE reporter activity (Fig. 4C). Further, a small molecule inhibitor of HIF signaling, 7.5 μM YC-1 (Chun et al., 2001) given together with As(III) attenuated the HRE induction of both 0.01 and 0.1 μM As(III) (Fig. 4C, second bar from right and rightmost bar, respectively).
Chronic Low-Dose As(III) Stimulates HIF-1 and VEGF Levels in Human Tumor, Smooth Muscle, and Endothelial Cells
It was next sought to determine whether the HIF signaling changes in response to chronic low-dose As(III) were specific to the B16-F10 cells. Because arsenic exposure is associated with bladder cancer (Steinmaus et al., 2000), isolated human bladder tumor cells, human bladder smooth muscle cells (SMCs), and human endothelial cells were exposed to 0.01 or 0.1 μM As(III) for seven days. Seven day exposure to 0.01 μM As(III) resulted in increased HIF-1 levels in all three cell types, whereas 0.1 μM As(III) resulted in increased HIF-1 only in the SMCs (Fig. 5A). In similar fashion, seven day exposure to 0.01 μM As(III) resulted in substantial increases in VEGF levels in all three cell types, whereas 0.1 μM As(III) resulted in increased VEGF levels only in the SMCs. (Fig. 5B).
HIF and VEGF Antagonists Attenuate the Angiogenic Effects of Low-dose As(III)
To examine whether HIF and VEGF were functionally involved in the angiogenic response to chronic low-dose As(III), the HIF inhibitor YC-1 (Chun et al., 2001) and a VEGF receptor-2 kinase inhibitor, SU5416 (Fong et al., 1999) were given together with As(III) in the chicken embryo chorioallantoic membrane (CAM) angiogenesis assay. Given alone, 10 μM YC-1 had little effect on blood vessel density in the CAM assay whereas 10 μM SU5416 alone resulted in a small (20%) reduction in blood vessel density (data not shown). As has been shown before by our laboratory (Soucy et al., 2003), 0.33 μM As(III) given alone resulted in a significant increase in blood vessel density (Fig. 6B and C). When given together with 0.33 μM As(III), either YC-1, the HIF-1 inhibitor, or SU5416, the VEGFR-2 kinase inhibitor, resulted in significant reductions in blood vessel density as compared to As(III) alone (Fig. 6B and C). These results were also observed with 5 μM 2-methoxyestradiol (2-ME2), another HIF-1 inhibitor (Mabjeesh et al., 2003) with antimicrotubule activity (D'Amato et al., 1994), although 2-ME2 given alone caused significant decreases in vessel density (data not shown). When 10 μM As(III) was applied to the CAM, a significant decrease in blood vessel density was observed (Fig. 6A and B), as shown previously (Soucy et al., 2003). When YC-1 was added to 10 μM As(III) (Fig. 6B, rightmost grouping), less of a decrease in blood vessel density was observed, with vessel density approaching that of an untreated CAM.
DISCUSSION
For the first time, this work shows that As(III) given chronically at environmental levels in the drinking water of animals can enhance both primary and secondary tumor growth in a syngeneic tumor model (Fig. 1). In addition, arsenic levels in the tumors are greater than in the blood (Fig. 2) and reach levels previously shown to be pro-angiogenic in angiogenesis assays (Soucy et al., 2003). HIF-1 and VEGF levels are increased in the resulting tumors of As(III)-exposed mice, but that the expression of these proteins is limited to areas surrounding blood vessels (Fig. 3). Further, chronic, but not acute, exposure to low-dose As(III) results in increased HIF-1 expression, HRE transactivation and VEGF release in isolated B16 cells (Fig. 4. This increase in HIF signaling in response to chronic low-dose As(III) is not limited to B16 mouse tumor cells, but is also observed in human bladder tumor cells, human bladder SMCs and in human endothelial cells (Fig. 5). Finally, we show that inhibitors of HIF and VEGF are capable of disrupting the angiogenic effects of As(III) in the CAM assay (Fig. 6). Taken together, these data suggest a role for HIF signaling in tumor-related angiogenesis in response to chronic low-dose As(III).
The finding that levels as low as 50 ppb As(III) given in drinking water result in significant increases in primary tumor growth (Fig. 1A and 1B) and levels as low as 10 ppb As(III) result in an increased number of metastases (Fig. 1C) suggests that environmental levels of arsenic are capable of causing significant physiological changes. Because these levels fall between the new EPA action level of 10 ppb arsenic due to take effect in 2006 and the current standard of 50 ppb, it is clear that further studies both in the laboratory and the field are not only warranted but critical to better understand the complexity of the effects of low-dose arsenic in the public health setting.
The finding that only chronic exposure (three or seven day but not 4 h) to low-dose As(III) was capable of stimulating HIF-1 expression suggests a cellular buildup of a substance capable of stimulating this protein. Because As(III), even at low-doses is known to generate cellular reactive oxygen species (ROS) (Barchowsky et al., 1999) and because HIF is known to be regulated by ROS (reviewed in Michiels et al., 2002; Page et al., 2002), ROS are a likely candidate as a stimulator of HIF. In addition, other metals like nickel have been shown to induce HIF signaling via ROS (Andrew et al., 2001). Preliminary studies in our laboratory suggest that the addition of a superoxide dismutase scavenger (MnTBAP; 50 μM) together with a pro-angiogenic dose of 0.33 μM As(III) reduces the increase in blood vessel density as compared to As(III) alone in the CAM assay (data not shown). This hypothesis is being explored further in our laboratory in isolated tumor cells and in a mouse tumor model. The further ability of the HIF and VEGF inhibitors to reverse the angiogenic effects of both low- and high-dose As(III) suggests a functional role for this signaling pathway in both the pro- and perhaps the anti-angiogenic effects of low-dose and high-dose As(III), respectively (Fig. 6).
The slight induction of VEGF in response to 4 h As(III) (Fig. 4B, first grouping of bars) in the absence of HIF-1 induction suggests that alternate HIF isoforms such as HIF-2 or some other As(III)-induced mechanism might be playing a role in these cells. VEGF might also be regulated by HIF-1–independent pathways such as MAP kinase (Duyndam et al., 2002) or Akt (Gao et al., 2004) in response to acute (4 h) As(III) exposure. The restriction of both HIF-1 and VEGF expression to areas adjacent to the blood vessels suggested that HIF signaling might be cell specific. However, since chronic low-dose As(III) induced HIF-1 expression in three other cell types (Fig. 5A), it is unlikely that HIF-1 induction is cell specific. Therefore, we speculate that limited diffusion of As(III) from the blood vessel to the tumor may be responsible for the localized induction of HIF-1 (Fig. 3). The finding that As(III) distributes more to the tumor than to the blood (Fig. 2) indicates the existence of a diffusion gradient between the blood and the internal portion of the tumor. This diffusion gradient may explain the localized increases in HIF-1 and VEGF (Fig. 3).
In conclusion, these studies demonstrate that environmentally relevant levels of As(III) given chronically in drinking water can enhance growth of both primary tumors and metastases in an animal model. In addition, chronic but not acute exposure to low-dose As(III) was found to stimulate HIF signaling in tumors in vivo, as well as in isolated tumor, smooth muscle, and endothelial cells. Finally, inhibitors of HIF and VEGF were found to interrupt the angiogenic effects of As(III), suggesting a role for this pathway in the vascular response to chronic low-dose As(III). These data add to the growing mass of knowledge suggesting that chronic exposure to environmental levels of arsenic can result in profound effects on the vasculature and on disease states dependent upon neovascularization, such as cancer.
NOTES
1 These two authors contributed equally to this work.
ACKNOWLEDGMENTS
This work was supported by startup funds and by an Elsa U. Pardee Foundation award (M.A.I.). Conflict of interest: none disclosed.
REFERENCES
Adelman, D. M., Maltepe, E., and Simon, M. C. (1999). Multilineage embryonic hematopoiesis requires hypoxic ARNT activity. Genes Dev. 13, 2478–2483.
Ades, E. W., Candal, F. J., Swerlick, R. A., George, V. G., Summers, S., Bosse, D. C., and Lawley, T. J. (1992). HMEC-1: Establishment of an immortalized human microvascular endothelial cell line. J. Invest. Dermatol. 99, 683–690.
Andrew, A. S., Klei, L. R., and Barchowsky, A. (2001). Nickel requires hypoxia-inducible factor-1 alpha, not redox signaling, to induce plasminogen activator inhibitor-1. Am. J. Physiol. Lung Cell. Mol. Physiol. 281, L607– L615.
Auguste, P., Lemiere, S., Larrieu-Lahargue, F., and Bikfalvi, A. (2005). Molecular mechanisms of tumor vascularization. Crit. Rev. Oncol. Hematol. 54, 53–61.
Barchowsky, A., Klei, L. R., Dudek, E. J., Swartz, H. M., and James, P. E. (1999). Stimulation of reactive oxygen, but not reactive nitrogen species, in vascular endothelial cells exposed to low levels of arsenite. Free Radic. Biol. Med. 27, 1405–1412.
Barkhordar, R. A., Berston, E. D., and Ramos, D. M. (1995). Cervical lymph node metastasis: Model for study of head/neck melanoma. Eur. J. Cancer B Oral Oncol. 31B, 49–52.
Chun, Y. S., Kim, M. S., and Park, J. W. (2002). Oxygen-dependent and -independent regulation of HIF-1alpha. J. Korean Med. Sci. 17, 581–588.
Chun, Y. S., Yeo, E. J., Choi, E., Teng, C. M., Bae, J. M., Kim, M. S., and Park, J. W. (2001). Inhibitory effect of YC-1 on the hypoxic induction of erythropoietin and vascular endothelial growth factor in Hep3B cells. Biochem. Pharmacol. 61, 947–954.
Covello, K. L., and Simon, M. C. (2004). HIFs, hypoxia, and vascular development. Curr. Top. Dev. Biol. 62, 37–54.
Cullen, B. R., and Malim, M. H. (1992). Secreted placental alkaline phosphatase as a eukaryotic reporter gene. Methods Enzymol. 216, 362–368.
D'Amato, R. J., Lin, C. M., Flynn, E., Folkman, J., and Hamel, E. (1994). 2-Methoxyestradiol, an endogenous mammalian metabolite, inhibits tubulin polymerization by interacting at the colchicine site. Proc. Natl. Acad. Sci. U.S.A. 91, 3964–3968.
Davison, K., Mann, K. K., and Miller, W. H., Jr. (2002). Arsenic trioxide: Mechanisms of action. Semin. Hematol. 39, 3–7.
Duyndam, M. C., Hulscher, T. M., Fontijn, D., Pinedo, H. M., and Boven, E. (2001). Induction of vascular endothelial growth factor expression and hypoxia-inducible factor 1alpha protein by the oxidative stressor arsenite. J. Biol. Chem. 276, 48066–48076.
Duyndam, M. C., Hulsher, S. T., Van Der Wall, E., Pinedo, H. M., and Boven, E. (2003). Evidence for a role of p38 kinase in hypoxia-inducible factor 1-independent induction of vascular endothelial growth factor expression by sodium arsenite. J. Biol. Chem. 278, 6885–6895.
Fidler, I. J., Gersten, D. M., and Budmen, M. B. (1976). Characterization in vivo and in vitro of tumor cells selected for resistance to syngeneic lymphocyte-mediated cytotoxicity. Cancer Res. 36, 3160–3165.
Fong, T. A., Shawver, L. K., Sun, L., Tang, C., App, H., Powell, T. J., Kim, Y. H., Schreck, R., Wang, X., Risau, W., et al. (1999). SU5416 is a potent and selective inhibitor of the vascular endothelial growth factor receptor (Flk-1/KDR) that inhibits tyrosine kinase catalysis, tumor vascularization, and growth of multiple tumor types. Cancer Res. 59, 99–106.
Gao, N., Shen, L., Zhang, Z., Leonard, S. S., He, H., Zhang, X. G., Shi, X., and Jiang, B. H. (2004). Arsenite induces HIF-1alpha and VEGF through PI3K, Akt and reactive oxygen species in DU145 human prostate carcinoma cells. Mol. Cell. Biochem. 255, 33–45.
Hei, T. K., and Filipic, M. (2004). Role of oxidative damage in the genotoxicity of arsenic. Free Radic. Biol. Med. 37, 574–581.
Jakob, W., Jentzsch, K. D., Mauersberger, B., and Heder, G. (1978). The chick embryo choriallantoic membrane as a bioassay for angiogenesis factors: Reactions induced by carrier materials. Exp. Pathol. (Jena) 15, 241–249.
Kao, Y. H., Yu, C. L., Chang, L. W., and Yu, H. S. (2003). Low concentrations of arsenic induce vascular endothelial growth factor and nitric oxide release and stimulate angiogenesis in vitro. Chem. Res. Toxicol. 16, 460–468.
Kumagai, Y., and Pi, J. (2004). Molecular basis for arsenic-induced alteration in nitric oxide production and oxidative stress: Implication of endothelial dysfunction. Toxicol. Appl. Pharmacol. 198, 450–457.
Liu, Y., Cox, S. R., Morita, T., and Kourembanas, S. (1995). Hypoxia regulates vascular endothelial growth factor gene expression in endothelial cells. Identification of a 5' enhancer. Circ. Res. 77, 638–643.
Mabjeesh, N. J., Escuin, D., LaVallee, T. M., Pribluda, V. S., Swartz, G. M., Johnson, M. S., Willard, M. T., Zhong, H., Simons, J. W., and Giannakakou, P. (2003). 2ME2 inhibits tumor growth and angiogenesis by disrupting microtubules and dysregulating HIF. Cancer Cell 3, 363–375.
Marks, M. G., Shi, J., Fry, M. O., Xiao, Z., Trzyna, M., Pokala, V., Ihnat, M. A., and Li, P. K. (2002). Effects of putative hydroxylated thalidomide metabolites on blood vessel density in the chorioallantoic membrane (CAM) assay and on tumor and endothelial cell proliferation. Biol. Pharm. Bull. 25, 597–604.
McColl, B. K., Stacker, S. A., and Achen, M. G. (2004). Molecular regulation of the VEGF family—inducers of angiogenesis and lymphangiogenesis. APMIS 112, 463–480.
Michiels, C., Minet, E., Mottet, D., and Raes, M. (2002). Regulation of gene expression by oxygen: NF-kappaB and HIF-1, two extremes. Free Radic. Biol. Med. 33, 1231–1242.
Page, E. L., Robitaille, G. A., Pouyssegur, J., and Richard, D. E. (2002). Induction of Hypoxia-inducible Factor-1alpha by Transcriptional and Translational Mechanisms. J. Biol. Chem. 277, 48403–48409.
Rahman, M. (2002). Arsenic and hypertension in Bangladesh. Bull. World Health Organ. 80, 173.
Roboz, G. J., Dias, S., Lam, G., Lane, W. J., Soignet, S. L., Warrell, R. P., Jr., and Rafii, S., (2000). Arsenic trioxide induces dose- and time-dependent apoptosis of endothelium and may exert an antileukemic effect via inhibition of angiogenesis. Blood 96, 1525–1530.
Rossman, T. G. (2003). Mechanism of arsenic carcinogenesis: An integrated approach. Mutat. Res. 533, 37–65.
Schmeisser, E., Goessler, W., Kienzl, N., and Francesconi, K. A. (2004). Volatile analytes formed from arsenosugars: Determination by HPLC-HG-ICPMS and implications for arsenic speciation analyses. Anal. Chem. 76, 418–423.
Shi, H., Hudson, L. G., and Liu, K. J. (2004). Oxidative stress and apoptosis in metal ion-induced carcinogenesis. Free Radic. Biol. Med. 37, 582–593.
Shi, J., Xiao, Z., Ihnat, M. A., Kamat, C., Pandit, B., Hu, Z., and Li, P. K. (2003). Structure-activity relationships studies of the anti-angiogenic activities of linomide. Bioorg. Med. Chem. Lett. 13, 1187–1189.
Shinkaruk, S., Bayle, M., Lain, G., and Deleris, G. (2003). Vascular endothelial cell growth factor (VEGF). an emerging target for cancer chemotherapy. Curr. Med. Chem. Anti-Canc. Agents 3, 95–117.
Soucy, N. V., Ihnat, M. A., Kamat, C. D., Hess, L., Post, M. J., Klei, L. R., Clark, C., and Barchowsky, A. (2003). Arsenic stimulates angiogenesis and tumorigenesis in vivo. Toxicol. Sci. 76, 271–279.
Soucy, N. V., Klei, L. R., Mayka, D. D., and Barchowsky, A. (2004). Arsenic stimulates angiogenesis and tumorigenesis in vivo. Chem. Res. Toxicol. 17, 555–563.
Steinmaus, C., Moore, L., Hopenhayn-Rich, C., Biggs, M. L., and Smith, A. H. (2000). Arsenic in drinking water and bladder cancer. Cancer Invest. 18, 174–182.
Tchounwou, P. B., Patlolla, A. K., and Centeno, J. A. (2003). Carcinogenic and systemic health effects associated with arsenic exposure—a critical review. Toxicol. Pathol. 31, 575–588.
Tseng, C. H. (2002). An overview on peripheral vascular disease in blackfoot disease-hyperendemic villages in Taiwan. Angiology 53, 529–537.
Tseng, C. H., Tseng, C. P., Chiou, H. Y., Hsueh, Y. M., Chong, C. K., and Chen, C. J. (2002). Epidemiologic evidence of diabetogenic effect of arsenic. Toxicol. Lett. 133, 69–76.
Wang, G. L., and Semenza, G. L. (1993). Desferrioxamine induces erythropoietin gene expression and hypoxia-inducible factor 1 DNA-binding activity: Implications for models of hypoxia signal transduction. Blood 82, 3610–3615.
Yu, H. S., Lee, C. H., and Chen, G. S. (2002). Peripheral vascular diseases resulting from chronic arsenical poisoning. J. Dermatol. 29, 123–130.(Chandrashekhar D. Kamat, )
Center for Environmental Health Sciences and Department of Pharmacology and Toxicology, Dartmouth Medical School, Hanover, New Hampshire 03755
ABSTRACT
Trivalent inorganic arsenic (arsenite, arsenic trioxide, As(III)) is a primary contaminant of groundwater supplies worldwide. As(III), marketed as trisenox, is also an FDA-approved agent to treat cancer It has been previously shown by our laboratory that As(III) administered at doses lower than a therapeutic anticancer dose results in an increase in tumor formation and blood vessel density of tumors. In this work it was found that chronic administration of As(III) approaching the EPA action level of 10 ppb, given in the drinking water of mice 5 weeks prior to B16-F10 melanoma implantation, increased the growth rate of primary tumors and the number of metastases to the lung. Further, levels of arsenic in the tumor and lung were found to be much greater than those in the blood and similar to pro-angiogenic As(III) doses. Levels of hypoxia inducible factor-1 (HIF-1) and vascular endothelial growth factor (VEGF) surrounding the blood vessels in the tumors of the As(III)-treated mice were also found to be increased. Exposure of isolated B16-F10 tumor cells to chronic (3 or 7 day) but not acute (4 h) low-dose As(III) was found to increase HIF-1 expression and secretion of VEGF. Finally, coadministration of an inhibitor of HIF (YC-1) or a VEGFR-2 kinase inhibitor (SU5416) was found to antagonize the pro-angiogenic effects of low-dose As(III). Together, these results suggest that chronic exposure to low-dose As(III) could stimulate growth of tumors through a HIF-dependent stimulation of angiogenesis.
Key Words: arsenic; angiogenesis; hypoxia inducible factor; vascular endothelial growth factor; tumorigenesis.
INTRODUCTION
Arsenite (As(III)) is a major natural contaminant of drinking water throughout the world. As(III) is also an FDA-approved anticancer agent for leukemias and is in clinical trials for several solid tumor types. In the United States, the Environmental Protection Agency (EPA) lowered the action level arsenic from 50 parts per billion (ppb) to 10 ppb in 2001, due to take effect in 2006 (http://www.epa.gov/safewater/arsenic.html). As(III) exposure is associated with an increased incidence of a number of disease states, including vascular conditions such as Blackfoot disease (Tseng, 2002; Yu et al., 2002), diabetes (Tseng et al., 2002), hypertension (Rahman, 2002), and cancers such as those of the lung, bladder, and skin (Tchounwou et al., 2003). With respect to its vascular effects, As(III) has been shown to be anti-angiogenic at higher doses such as those used in the treatment of cancer (Davison et al., 2002; Roboz et al., 2000), whereas it has also been shown that lower, more environmentally relevant doses of As(III) are in fact pro-angiogenic (Kao et al., 2003; Soucy et al., 2003). It has also been shown by our group (Soucy et al., 2003) and others (reviewed in Rossman, 2003) that As(III) is tumorigenic.
Tumor-related angiogenesis is a very complex process involving several different cell types including tumor cells, endothelial cells, smooth muscle cells, or pericytes surrounding the endothelial cells and fibroblasts/keratinocytes. During angiogenesis, these cell types release and respond to multiple growth factors and digestive enzymes that facilitate cell growth, destabilization of existing vessels sprouting of new vessels, and restabilization of nascent vessels (reviewed in Auguste et al., 2005). Vascular endothelial growth factor (VEGF) has been shown to be a critical factor in tumor-related angiogenesis and is a primary therapeutic target for the treatment of cancer through angiogenesis inhibition (reviewed in Shinkaruk et al., 2003). VEGF is primarily regulated through hypoxia inducible factor (HIF), a heterodimeric transcription factor complex. As the name implies, HIF levels are induced by low oxygen conditions, principally through the inhibition of the proteasome-mediated degradation of HIF (reviewed in Covello and Simon, 2004). HIF signaling is also regulated by nitric oxide (NO) and other reactive oxygen species (ROS) such as superoxide and hydroxy free radical (reviewed in Chun et al., 2002). In addition to VEGF, HIF regulates a large panel of genes involved in angiogenesis (e.g., plasminogen activator inhibitor-1), RBC formation (erythropoietin, heme oxygenase-1), glucose regulation (glucose transporter-1, phosphoglucokinase-1), and NO signaling (inducible nitric oxide synthase). As(III) has been shown to induce levels of HIF-1, the regulated subunit of HIF (Duyndam et al., 2001) as well as levels of VEGF (Duyndam et al., 2003; Kao et al., 2003) in isolated cells.
Since As(III) exposure is associated with the formation of several ROS types (Hei and Filipic, 2004; Kumagai and Pi, 2004; Shi et al., 2004) and increases in NO signaling (Kumagai and Pi, 2004), it was reasoned that low-dose As(III) might be acting through HIF signaling to stimulate tumor-related angiogenesis. It has been previously shown by our laboratory that, when As(III) is given at doses five- or ten-fold lower than those used to treat cancer (0.5 and 1 mg/kg As(III) given biweekly), it can stimulate the growth of implanted tumors (Soucy et al., 2003). It is hypothesized that chronic exposure to environmental doses of As(III) leads to tumor growth through the stimulation of angiogenesis via HIF-mediated cellular signaling.
In this work, implanted tumors in mice exposed to As(III) for 5 weeks at levels approaching the drinking water action level of 10 ppb were shown to grow faster than those of animals not receiving As(III). In addition, the number of metastases to the lungs was found to be higher in animals receiving As(III). Localized levels of HIF-1 and VEGF surrounding the blood vessels were increased in tumors of animals receiving low-dose As(III). In isolated tumor cells, smooth muscle cells (SMCs), and endothelial cells, HIF-1 and VEGF levels were induced by chronic (72 h or 7 day) low-dose As(III) but not by acute (4 h) As(III). Finally, inhibitors of HIF-1 and VEGF attenuated the effect of As(III) on blood vessel density, implying that these two proteins are functionally involved in the angiogenic effect of low-dose As(III).
MATERIALS AND METHODS
Chemicals and cell culture reagents.
All chemicals were purchased from EMD Biosciences (San Diego, CA) unless otherwise noted. Sodium arsenite was purchased from Sigma Chemical, St. Louis, MO. Cell culture reagents were purchased from InVitrogen (Carlsbad, CA).
Cell culture.
B16-F10 mouse melanoma cells (Soucy et al., 2003), HMEC-1 human microvascular endothelial cells (Ades et al., 1992) (obtained from the Centers for Disease Control), and human J82 bladder tumor cells (American Type Culture Collection), were grown in Dulbecco's modified Eagle media supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), pyruvate, glutamine (Mediatech, Herndon, VA), and penicillin/streptomycin (Hyclone) in a humidified environment with 5% carbon dioxide added. H500Z human bladder smooth muscle cells (a kind gift of Dr. Bradley Kropp, OUHSC Department of Urology) were grown in M199 media (InVitrogen) with 10% fetal bovine serum (Hyclone), pyruvate, glutamine, and penicillin/streptomycin, 1% vitamins, 1 MEM % nonessential amino acid solution (InVitrogen, Grand Island, NY).
Chorioallantoic membrane (CAM) assay of angiogenesis.
The CAM assay that was used is based upon the original procedure by Jakob et al. (1978) with modifications made by our laboratory (Marks et al., 2002; Shi et al., 2003; Soucy et al., 2003). Fertile leghorn chicken embryos (CBT farms, Chestertown, MD), 24 h after laying, were incubated for 3–4 days with rocking and then placed horizontally for a minimum of 4 h at 37.2°C. Embryos were then cracked into 100 x 20 mm Petri dishes and allowed to grow at 37.2°C for one additional day. As(III) +/– YC-1 or SU5416 were placed at saturation onto 1.2-mm-diameter cellulose nitrate discs (Whatman) and onto a blood vessel (small arteriole; metarteriole) distal from the developing embryo. As(III) +/– inhibitors were replaced onto the disc at 24 and 48 h, and then at 72 h post-treatment, digital brightfield images were collected. The disc was centered in the visual field at 16x magnification, then removed using a camel's hair brush, and images were collected. Vascular density in the area of the disc was counted by overlaying a 4 x 4 grid on the 16x images and counting vessels per grid square in five grids. Statistical analysis was accomplished using unpaired two-tailed Student's t-test. Typically, 4–5 individual CAMs were used for each treatment and all experiments were repeated at least twice.
Mouse tumor studies.
All animal studies were conducted in accordance with AALAC approved guidelines using a protocol approved by IACUC at the University of Oklahoma. Six-week-old male NCr nu/nu mice (NCI APA breeding stock, Frederick MD) were housed in ventilated cages with autoclaved AIN-7A arsenic-characterized chow (Research Diets, New Brunswick, NJ), autoclaved water and autoclaved bedding. Following 1 week of acclimation, 0, 10, 50, or 200 ppb sodium arsenite was added to the drinking water of appropriate animals for 5 weeks with water changed biweekly. At 5 weeks after initial As(III) administration, 5 x 105 B16-F10 (GFP) cells in 100 μl PBS/l mM EDTA were injected subcutaneously into the external surface at the base of the right ear of all animals. Tumor measurement began 7 days after implantation and continued biweekly until tumors reached approximately 8–10% of animal body weight. The length (L) and width (W) of tumors were measured using Vernier calipers (Mitutoyo, Kawasaki Kanagawa, Japan), and tumor volume was calculated using the formula L x W2.
After sacrifice, the tumor and lungs were excised, fixed for 8 h in neutral buffered formalin, embedded in paraffin, sectioned, and stained by either H&E or standard immunohistochemical methods as previously described (Soucy et al., 2003). Sections were deparaffinized, incubated in 6% peroxide in 80% ethanol for 20 min, incubated in 0.1 M sodium citrate to retrieve the antigens, heated until boiling, and then heated for 10 additional min on 30% power using a 700 W microwave oven. Sections were then blocked using Vectastain kits (Vector Laboratories, Burlingame, CA), exposed to antibodies against GFP (Clontech; 1:150 dilution) to examine metastases, HIF-1 (Novus Biologicals monoclonal antibody H1alpha67, 1:50) or mouse VEGF (Abcam ab2992; 1:150) overnight at 4°C; then a secondary antibody and ABC reagent were added, and sections were stained with VIP peroxidase substrate (Vector Laboratories) and counterstained with Gomori Trichrome (Harleco EMD, San Diego, CA).
ICPMS arsenic measurement.
For these experiments, tumor-bearing mice described above were anesthetized using halothane, and then blood was removed, using heart puncture, to a heparin tube. Portions of tumor, lung, and heart were removed and snap-frozen in liquid nitrogen and stored at –30°C until analysis. Prior to analysis, the samples were digested with concentrated nitric acid in closed Teflon vessels in a microwave oven applying the following procedure: 200 mg of sample (blood, tumor, or organ) was weighed accurately into a 7-ml Teflon vial, and 1.00 ml of ultrapure concentrated nitric acid was added. The vial was closed and placed with up to 26 other samples in a microwave oven. A power of 300 W was applied for 10 min to a temperature of 130°C, which was held for 10 additional min. The resulting clear green/yellow solution was transferred to a 5-ml acid-cleaned polypropylene vial; the vial was capped and stored at 4°C until ICP-MS analysis. The arsenic contents of the digested samples were determined using hydride-generation inductively coupled plasma mass spectrometry (HG-ICPMS). Hydride generation was applied to eliminate interference from chloride, seen in the form of ArCl+ on the As+ signal (both seen of m/z = 75 in the mass spectrum). Applying hydride generation, the arsenic was converted to AsH3 and removed from the sample matrix (and chloride) before introduction to ICP-MS. General matrix interferences were eliminated by the use of standard addition calibration. As quality control, two blanks and one certified reference material were analyzed for every eight samples in order to monitor contamination and method precision and accuracy. Trace elements in whole blood and CRM 185R trace elements in bovine liver were used for quality control of blood and tumor/organs, respectively. The method detection limit is approximately 10 pg/ml in the sample solution corresponding to 0.25 ng/g (ppb) in blood and organs.
VEGF ELISA.
VEGF levels were assessed by ELISA using a rabbit polyclonal anti-VEGF antibody (Abcam, Cambridge, MA ). Supernatants from cells (100 μl) were added to a high binding white ELISA plate (COSTAR, Corning, NY), together with a standard curve of recombinant mouse VEGF (0.1–10,000 pg/ml) (Biosource International, Camarillo, CA) overnight at 4°C. The wells were washed with TBS, and anti-VEGF antibody (1:1000 dilution of 1 mg/ml stock) was added for 2 h at room temperature (RT). The wells washed again, and 1:2000 anti-rabbit HRP conjugate (Pierce, Rockford, IL) was added for 45 min at RT. The wells were washed again with TBS, then an ECL ELISA substrate (Pierce) was added to wells for 5 min, and the chemiluminescence was measured using a plate reader (Fluostar Optima, BMG Labtechnologies, NC). Data were normalized to the standard curve to yield values in pg/ml VEGF and graphed as percentage of untreated control. Significance among treatments was determined by unpaired two-tailed Student's t-test and two-way ANOVA with Bonferroni's post-test used to compare two dose-response curves, with the null hypothesis rejected when p < 0.05.
Western blot.
Equal amounts of lysates (20–50 μg) were separated by SDS-PAGE and transferred to nitrocellulose. Transfer was assessed by staining the membrane with 0.1% India ink in TBS with 0.1% Tween-20. Membranes were blocked 2 h in StartingBlock (TBS) Blocking Buffer (Pierce) with 0.1% Tween-20 (TBST). Membranes were washed in TBST and exposed to anti-HIF-1 (Chemicon: 1:500) antibody overnight at 4°C or 2 h at RT with rocking. Membranes were washed in TBST, secondary anti-rabbit HRP-conjugate (Pierce) diluted in StartingBlock (TBS) Blocking Buffer was added at 1:2500 dilution for 1 h at RT, the membranes were washed again in TBST, and then ECL reagent (Pierce) was added for 5 min, and images were captured using a CCD digital darkroom (UVP, Upland, CA).
Construction of the HRE-SEAP reporter construct.
To create the HRE-SEAP reporter, a double-stranded sequence containing the HRE (in bold) and flanking sequence from the human VEGF-165 gene (Liu et al., 1995) (5'AGATCTCCACAGTGCATACGTGGGCTCCAACACGTGGCTCTTCAGGAATTC-3'; upper strand) were created by annealing two complementary oligonucleotides with unique restriction sites (underlined) together. This sequence was then placed into the multiple cloning site of an empty SEAP reporter vector (Clontech, Palo Alto, CA). This construct was confirmed by sequencing using the T7 site of the vector.
Transfection of B16-F10 cells with the HRE-SEAP reporter construct.
Stable transfectants of the above HRE-SEAP reporter gene were made using liposome-mediated transfection using Fugene reagent (Roche, Indianapolis, IN) and selection under G418 pressure. A viable single clone colony was cultured and used for the SEAP assay.
Secreted alkaline phosphatase (SEAP) reporter assay.
B16-F10 cells transfected with the HRE-SEAP reporter gene and cultured in 96-well plates (Corning, NY) were exposed to 0.01 or 0.1 μM As(III) alone or with YC-1 (7.5 μM ) and a SEAP assay performed as previously described (Cullen and Malim, 1992). Briefly, 40 μl supernatant was removed from cells and plated into a 96-well plate, 6 μl 10x TBS added, and samples were incubated at 65°C for 30 min to inactivate endogenous alkaline phosphatase. Samples were placed on ice for 10 min. Then a chemifluorescent alkaline phosphatase substrate (CDP Star; Perkin-Elmer, Boston, MA) was added for 5 min, and images were captured by a CCD digital darkroom (UVP) at various times of exposure. Data were graphed as percentage of untreated control.
Statistics.
All analysis was done using Prism 4.0 (GraphPad Software, San Diego, CA). One-way ANOVA was conducted, with multiple comparisons tested by Bonferroni's post-hoc test at a significance level of 0.05 or greater. Tumor volumes were compared using repeated measures two-way ANOVA with treatments and days after implantation as independent variables with multiple comparisons tested by Dunnett's post-hoc test at a significance level of 0.05 or greater.
RESULTS
Chronic Low-Dose As(III) Exposure through Drinking Water Increases Tumor Growth and Metastasis
It has previously been shown by our laboratory that low-dose As(III) stimulates angiogenesis and tumorigenesis when administered IP in biweekly doses, similar to those used for anticancer treatment (Soucy et al., 2003). The hypothesis that chronic low-dose As(III) in drinking water stimulates tumor growth was tested using a classical mouse syngeneic tumor model, the B16 melanoma. Animals were exposed to 10, 50, or 200 ppb As(III) in their drinking water for 5 weeks; then B16-F10 (GFP) tumor cells were implanted into the skin around the ear of the mice (Barkhordar et al., 1995). All doses of As(III), including the U.S. EPA action level of 10 ppb (Fig. 1A, diamonds, and Fig. 1B, second bar from left) and the old (grandfathered) standard of 50 ppb (Fig. 1A, open squares, and Fig. 1B, second bar from right) resulted in increases in growth of implanted B16-F10 (GFP) tumors. To examine whether low-dose As(III) affected the metastasis of the lung-colonizing B16-F10 (GFP) cells (Fidler et al., 1976), lungs from the tumor-bearing mice were subjected to immunohistochemical staining for the GFP epitope transfected into the tumor cells (Soucy et al., 2003). Both 10 and 200 ppb As(III) in the drinking water significantly increased the average number of B16-F10 (GFP) metastases (Fig. 1C). In addition, the size of the metastases in animals exposed to 10 or 50 ppb As(III) was greater than in animals receiving no As(III) in their drinking water (data not shown). Thus chronic exposure to low-dose As(III) is capable of increasing growth of both primary tumors and secondary metastases in an in vivo syngeneic tumor model.
Arsenic Blood and Organ Levels
It was next determined what the levels of arsenic were in the blood and tissues from the tumor-bearing mice described above. Organs were snap-frozen, and arsenic levels were measured by ICPMS as previously described (Schmeisser et al., 2004). It is striking how low the levels of arsenic in the blood (Fig. 2, left-hand grouping) are as compared to the various organs (Fig. 2, middle and right groupings) and that, in most of the organs, As(III) levels increase incrementally with dose. It is also interesting that in the skin-tumor (melanoma surrounded by skin layer), the average values of the arsenic organ levels were similar to the dose (10, 50, and 200 ppb). In addition, these levels of arsenic in the tumors were within the angiogenic range found in the chorioallantoic angiogenesis assay in our previous paper (Soucy et al., 2003).
Chronic Low-Dose As(III) Stimulates HIF-1 and VEGF Levels Surrounding the Blood Vessels in Implanted Tumors
In our previous paper (Soucy et al., 2003) it was shown that low-dose (0.01–1 μM) As(III) increases blood vessel density in B16-F10 tumors in addition to increasing angiogenesis in two separate in vivo assays. VEGF is arguably the most important regulator of angiogenesis (McColl et al., 2004), and HIF is a primary regulator of VEGF (Adelman et al., 1999). High-dose As(III) (50–100 μM; 3750–7500 ppb) has been shown to induce HIF-1, the inducible subunit of HIF (Duyndam et al., 2001), whereas low-dose As(III) has been shown not to induce HIF-1 in smooth muscle cells (Soucy et al., 2004). To examine whether HIF signaling could be altered in the tumors of mice receiving chronic low-dose As(III), tumors were removed, tissue lysates made and a Western blot against HIF-1 completed. Barely detectable HIF-1 levels were observed in whole tumor lysates, with no change in HIF-1 levels in response to any dose of As(III) (Fig. 3A). To determine whether localized changes in HIF-1 could be occurring, sections of primary melanoma tumors were subjected to immunohistochemical staining. The results of representative sections against HIF-1 are shown in Figure 3B (top row) and against VEGF, a major angiogenic protein controlled in part through HIF signaling (Adelman et al., 1999) (Fig. 3B, bottom row). Chronic low-dose As(III) stimulated the expression of both HIF-1 and VEGF, but there was no apparent dose-dependent effect. Tumors of animals exposed to chronic low-dose As(III) in their drinking water had substantially increased HIF-1 staining; however, this staining was limited to areas just surrounding the blood vessels (Fig. 3B, top panel). Similarly, VEGF staining was also only observed close to the blood vessels in response to chronic low-dose As(III).
Chronic Low-Dose As(III) Stimulates HIF-1, HRE Transactivation, and VEGF Levels in Isolated B16-F10 Tumor Cells
It was next desired to see whether chronic low-dose As(III) exposure resulted in altered HIF-1 levels in isolated B16-F10 cells. Chronic (72 h or 7 day) exposure to 0.01 μM (7.5 ppb) As(III) resulted in significant increases in HIF-1 protein levels, whereas higher doses of As(III) resulted in decreased HIF-1 levels (Fig. 4A). In contrast, acute (4 h) exposure to As(III), from 0.01 to 10 μM, did not result in any increase in HIF-1 levels (data not shown).
It was next examined whether these changes in HIF-1 would result in increases in the HIF-regulated protein, VEGF, a critical regulator of angiogenesis. As(III) exposure (4 h), followed by 20 h in As(III)-free media, resulted in a small increase in VEGF (Fig. 4B, left grouping), particularly at the higher doses of As(III), as determined by ELISA. Chronic (72 h or 7 day) exposure to low-dose (0.01 μM) As(III) (Fig. 4B, middle and right groupings) resulted in a more pronounced increase in VEGF as compared with acute As(III) exposure, whereas chronic high-dose As(III) (10 μM for 72 h or 0.1μM for seven days) resulted in decreased VEGF levels. These VEGF levels again correlate well to the pro- and anti-angiogenic effects of low (0.01–1 μM) and high (>3.33 μM) dose As(III) observed in the CAM assay in our previous paper (Soucy et al., 2003).
Because VEGF could be regulated through a factor other than HIF, we investigated the effect of chronic low-dose As(III) on HIF-mediated transactivation through its DNA binding element, the hypoxia response element (HRE). In these experiments, the B16-F10 cells were transfected with two HREs linked to a secreted alkaline phosphatase (SEAP) reporter gene and stable transfectant clones isolated. Both 4 h cobalt chloride (CoCl2), a known HIF inducer (Wang and Semenza, 1993) and seven days of either 0.01 μM or 0.1 μM As(III) resulted in significant increases in HRE reporter activity (Fig. 4C). Further, a small molecule inhibitor of HIF signaling, 7.5 μM YC-1 (Chun et al., 2001) given together with As(III) attenuated the HRE induction of both 0.01 and 0.1 μM As(III) (Fig. 4C, second bar from right and rightmost bar, respectively).
Chronic Low-Dose As(III) Stimulates HIF-1 and VEGF Levels in Human Tumor, Smooth Muscle, and Endothelial Cells
It was next sought to determine whether the HIF signaling changes in response to chronic low-dose As(III) were specific to the B16-F10 cells. Because arsenic exposure is associated with bladder cancer (Steinmaus et al., 2000), isolated human bladder tumor cells, human bladder smooth muscle cells (SMCs), and human endothelial cells were exposed to 0.01 or 0.1 μM As(III) for seven days. Seven day exposure to 0.01 μM As(III) resulted in increased HIF-1 levels in all three cell types, whereas 0.1 μM As(III) resulted in increased HIF-1 only in the SMCs (Fig. 5A). In similar fashion, seven day exposure to 0.01 μM As(III) resulted in substantial increases in VEGF levels in all three cell types, whereas 0.1 μM As(III) resulted in increased VEGF levels only in the SMCs. (Fig. 5B).
HIF and VEGF Antagonists Attenuate the Angiogenic Effects of Low-dose As(III)
To examine whether HIF and VEGF were functionally involved in the angiogenic response to chronic low-dose As(III), the HIF inhibitor YC-1 (Chun et al., 2001) and a VEGF receptor-2 kinase inhibitor, SU5416 (Fong et al., 1999) were given together with As(III) in the chicken embryo chorioallantoic membrane (CAM) angiogenesis assay. Given alone, 10 μM YC-1 had little effect on blood vessel density in the CAM assay whereas 10 μM SU5416 alone resulted in a small (20%) reduction in blood vessel density (data not shown). As has been shown before by our laboratory (Soucy et al., 2003), 0.33 μM As(III) given alone resulted in a significant increase in blood vessel density (Fig. 6B and C). When given together with 0.33 μM As(III), either YC-1, the HIF-1 inhibitor, or SU5416, the VEGFR-2 kinase inhibitor, resulted in significant reductions in blood vessel density as compared to As(III) alone (Fig. 6B and C). These results were also observed with 5 μM 2-methoxyestradiol (2-ME2), another HIF-1 inhibitor (Mabjeesh et al., 2003) with antimicrotubule activity (D'Amato et al., 1994), although 2-ME2 given alone caused significant decreases in vessel density (data not shown). When 10 μM As(III) was applied to the CAM, a significant decrease in blood vessel density was observed (Fig. 6A and B), as shown previously (Soucy et al., 2003). When YC-1 was added to 10 μM As(III) (Fig. 6B, rightmost grouping), less of a decrease in blood vessel density was observed, with vessel density approaching that of an untreated CAM.
DISCUSSION
For the first time, this work shows that As(III) given chronically at environmental levels in the drinking water of animals can enhance both primary and secondary tumor growth in a syngeneic tumor model (Fig. 1). In addition, arsenic levels in the tumors are greater than in the blood (Fig. 2) and reach levels previously shown to be pro-angiogenic in angiogenesis assays (Soucy et al., 2003). HIF-1 and VEGF levels are increased in the resulting tumors of As(III)-exposed mice, but that the expression of these proteins is limited to areas surrounding blood vessels (Fig. 3). Further, chronic, but not acute, exposure to low-dose As(III) results in increased HIF-1 expression, HRE transactivation and VEGF release in isolated B16 cells (Fig. 4. This increase in HIF signaling in response to chronic low-dose As(III) is not limited to B16 mouse tumor cells, but is also observed in human bladder tumor cells, human bladder SMCs and in human endothelial cells (Fig. 5). Finally, we show that inhibitors of HIF and VEGF are capable of disrupting the angiogenic effects of As(III) in the CAM assay (Fig. 6). Taken together, these data suggest a role for HIF signaling in tumor-related angiogenesis in response to chronic low-dose As(III).
The finding that levels as low as 50 ppb As(III) given in drinking water result in significant increases in primary tumor growth (Fig. 1A and 1B) and levels as low as 10 ppb As(III) result in an increased number of metastases (Fig. 1C) suggests that environmental levels of arsenic are capable of causing significant physiological changes. Because these levels fall between the new EPA action level of 10 ppb arsenic due to take effect in 2006 and the current standard of 50 ppb, it is clear that further studies both in the laboratory and the field are not only warranted but critical to better understand the complexity of the effects of low-dose arsenic in the public health setting.
The finding that only chronic exposure (three or seven day but not 4 h) to low-dose As(III) was capable of stimulating HIF-1 expression suggests a cellular buildup of a substance capable of stimulating this protein. Because As(III), even at low-doses is known to generate cellular reactive oxygen species (ROS) (Barchowsky et al., 1999) and because HIF is known to be regulated by ROS (reviewed in Michiels et al., 2002; Page et al., 2002), ROS are a likely candidate as a stimulator of HIF. In addition, other metals like nickel have been shown to induce HIF signaling via ROS (Andrew et al., 2001). Preliminary studies in our laboratory suggest that the addition of a superoxide dismutase scavenger (MnTBAP; 50 μM) together with a pro-angiogenic dose of 0.33 μM As(III) reduces the increase in blood vessel density as compared to As(III) alone in the CAM assay (data not shown). This hypothesis is being explored further in our laboratory in isolated tumor cells and in a mouse tumor model. The further ability of the HIF and VEGF inhibitors to reverse the angiogenic effects of both low- and high-dose As(III) suggests a functional role for this signaling pathway in both the pro- and perhaps the anti-angiogenic effects of low-dose and high-dose As(III), respectively (Fig. 6).
The slight induction of VEGF in response to 4 h As(III) (Fig. 4B, first grouping of bars) in the absence of HIF-1 induction suggests that alternate HIF isoforms such as HIF-2 or some other As(III)-induced mechanism might be playing a role in these cells. VEGF might also be regulated by HIF-1–independent pathways such as MAP kinase (Duyndam et al., 2002) or Akt (Gao et al., 2004) in response to acute (4 h) As(III) exposure. The restriction of both HIF-1 and VEGF expression to areas adjacent to the blood vessels suggested that HIF signaling might be cell specific. However, since chronic low-dose As(III) induced HIF-1 expression in three other cell types (Fig. 5A), it is unlikely that HIF-1 induction is cell specific. Therefore, we speculate that limited diffusion of As(III) from the blood vessel to the tumor may be responsible for the localized induction of HIF-1 (Fig. 3). The finding that As(III) distributes more to the tumor than to the blood (Fig. 2) indicates the existence of a diffusion gradient between the blood and the internal portion of the tumor. This diffusion gradient may explain the localized increases in HIF-1 and VEGF (Fig. 3).
In conclusion, these studies demonstrate that environmentally relevant levels of As(III) given chronically in drinking water can enhance growth of both primary tumors and metastases in an animal model. In addition, chronic but not acute exposure to low-dose As(III) was found to stimulate HIF signaling in tumors in vivo, as well as in isolated tumor, smooth muscle, and endothelial cells. Finally, inhibitors of HIF and VEGF were found to interrupt the angiogenic effects of As(III), suggesting a role for this pathway in the vascular response to chronic low-dose As(III). These data add to the growing mass of knowledge suggesting that chronic exposure to environmental levels of arsenic can result in profound effects on the vasculature and on disease states dependent upon neovascularization, such as cancer.
NOTES
1 These two authors contributed equally to this work.
ACKNOWLEDGMENTS
This work was supported by startup funds and by an Elsa U. Pardee Foundation award (M.A.I.). Conflict of interest: none disclosed.
REFERENCES
Adelman, D. M., Maltepe, E., and Simon, M. C. (1999). Multilineage embryonic hematopoiesis requires hypoxic ARNT activity. Genes Dev. 13, 2478–2483.
Ades, E. W., Candal, F. J., Swerlick, R. A., George, V. G., Summers, S., Bosse, D. C., and Lawley, T. J. (1992). HMEC-1: Establishment of an immortalized human microvascular endothelial cell line. J. Invest. Dermatol. 99, 683–690.
Andrew, A. S., Klei, L. R., and Barchowsky, A. (2001). Nickel requires hypoxia-inducible factor-1 alpha, not redox signaling, to induce plasminogen activator inhibitor-1. Am. J. Physiol. Lung Cell. Mol. Physiol. 281, L607– L615.
Auguste, P., Lemiere, S., Larrieu-Lahargue, F., and Bikfalvi, A. (2005). Molecular mechanisms of tumor vascularization. Crit. Rev. Oncol. Hematol. 54, 53–61.
Barchowsky, A., Klei, L. R., Dudek, E. J., Swartz, H. M., and James, P. E. (1999). Stimulation of reactive oxygen, but not reactive nitrogen species, in vascular endothelial cells exposed to low levels of arsenite. Free Radic. Biol. Med. 27, 1405–1412.
Barkhordar, R. A., Berston, E. D., and Ramos, D. M. (1995). Cervical lymph node metastasis: Model for study of head/neck melanoma. Eur. J. Cancer B Oral Oncol. 31B, 49–52.
Chun, Y. S., Kim, M. S., and Park, J. W. (2002). Oxygen-dependent and -independent regulation of HIF-1alpha. J. Korean Med. Sci. 17, 581–588.
Chun, Y. S., Yeo, E. J., Choi, E., Teng, C. M., Bae, J. M., Kim, M. S., and Park, J. W. (2001). Inhibitory effect of YC-1 on the hypoxic induction of erythropoietin and vascular endothelial growth factor in Hep3B cells. Biochem. Pharmacol. 61, 947–954.
Covello, K. L., and Simon, M. C. (2004). HIFs, hypoxia, and vascular development. Curr. Top. Dev. Biol. 62, 37–54.
Cullen, B. R., and Malim, M. H. (1992). Secreted placental alkaline phosphatase as a eukaryotic reporter gene. Methods Enzymol. 216, 362–368.
D'Amato, R. J., Lin, C. M., Flynn, E., Folkman, J., and Hamel, E. (1994). 2-Methoxyestradiol, an endogenous mammalian metabolite, inhibits tubulin polymerization by interacting at the colchicine site. Proc. Natl. Acad. Sci. U.S.A. 91, 3964–3968.
Davison, K., Mann, K. K., and Miller, W. H., Jr. (2002). Arsenic trioxide: Mechanisms of action. Semin. Hematol. 39, 3–7.
Duyndam, M. C., Hulscher, T. M., Fontijn, D., Pinedo, H. M., and Boven, E. (2001). Induction of vascular endothelial growth factor expression and hypoxia-inducible factor 1alpha protein by the oxidative stressor arsenite. J. Biol. Chem. 276, 48066–48076.
Duyndam, M. C., Hulsher, S. T., Van Der Wall, E., Pinedo, H. M., and Boven, E. (2003). Evidence for a role of p38 kinase in hypoxia-inducible factor 1-independent induction of vascular endothelial growth factor expression by sodium arsenite. J. Biol. Chem. 278, 6885–6895.
Fidler, I. J., Gersten, D. M., and Budmen, M. B. (1976). Characterization in vivo and in vitro of tumor cells selected for resistance to syngeneic lymphocyte-mediated cytotoxicity. Cancer Res. 36, 3160–3165.
Fong, T. A., Shawver, L. K., Sun, L., Tang, C., App, H., Powell, T. J., Kim, Y. H., Schreck, R., Wang, X., Risau, W., et al. (1999). SU5416 is a potent and selective inhibitor of the vascular endothelial growth factor receptor (Flk-1/KDR) that inhibits tyrosine kinase catalysis, tumor vascularization, and growth of multiple tumor types. Cancer Res. 59, 99–106.
Gao, N., Shen, L., Zhang, Z., Leonard, S. S., He, H., Zhang, X. G., Shi, X., and Jiang, B. H. (2004). Arsenite induces HIF-1alpha and VEGF through PI3K, Akt and reactive oxygen species in DU145 human prostate carcinoma cells. Mol. Cell. Biochem. 255, 33–45.
Hei, T. K., and Filipic, M. (2004). Role of oxidative damage in the genotoxicity of arsenic. Free Radic. Biol. Med. 37, 574–581.
Jakob, W., Jentzsch, K. D., Mauersberger, B., and Heder, G. (1978). The chick embryo choriallantoic membrane as a bioassay for angiogenesis factors: Reactions induced by carrier materials. Exp. Pathol. (Jena) 15, 241–249.
Kao, Y. H., Yu, C. L., Chang, L. W., and Yu, H. S. (2003). Low concentrations of arsenic induce vascular endothelial growth factor and nitric oxide release and stimulate angiogenesis in vitro. Chem. Res. Toxicol. 16, 460–468.
Kumagai, Y., and Pi, J. (2004). Molecular basis for arsenic-induced alteration in nitric oxide production and oxidative stress: Implication of endothelial dysfunction. Toxicol. Appl. Pharmacol. 198, 450–457.
Liu, Y., Cox, S. R., Morita, T., and Kourembanas, S. (1995). Hypoxia regulates vascular endothelial growth factor gene expression in endothelial cells. Identification of a 5' enhancer. Circ. Res. 77, 638–643.
Mabjeesh, N. J., Escuin, D., LaVallee, T. M., Pribluda, V. S., Swartz, G. M., Johnson, M. S., Willard, M. T., Zhong, H., Simons, J. W., and Giannakakou, P. (2003). 2ME2 inhibits tumor growth and angiogenesis by disrupting microtubules and dysregulating HIF. Cancer Cell 3, 363–375.
Marks, M. G., Shi, J., Fry, M. O., Xiao, Z., Trzyna, M., Pokala, V., Ihnat, M. A., and Li, P. K. (2002). Effects of putative hydroxylated thalidomide metabolites on blood vessel density in the chorioallantoic membrane (CAM) assay and on tumor and endothelial cell proliferation. Biol. Pharm. Bull. 25, 597–604.
McColl, B. K., Stacker, S. A., and Achen, M. G. (2004). Molecular regulation of the VEGF family—inducers of angiogenesis and lymphangiogenesis. APMIS 112, 463–480.
Michiels, C., Minet, E., Mottet, D., and Raes, M. (2002). Regulation of gene expression by oxygen: NF-kappaB and HIF-1, two extremes. Free Radic. Biol. Med. 33, 1231–1242.
Page, E. L., Robitaille, G. A., Pouyssegur, J., and Richard, D. E. (2002). Induction of Hypoxia-inducible Factor-1alpha by Transcriptional and Translational Mechanisms. J. Biol. Chem. 277, 48403–48409.
Rahman, M. (2002). Arsenic and hypertension in Bangladesh. Bull. World Health Organ. 80, 173.
Roboz, G. J., Dias, S., Lam, G., Lane, W. J., Soignet, S. L., Warrell, R. P., Jr., and Rafii, S., (2000). Arsenic trioxide induces dose- and time-dependent apoptosis of endothelium and may exert an antileukemic effect via inhibition of angiogenesis. Blood 96, 1525–1530.
Rossman, T. G. (2003). Mechanism of arsenic carcinogenesis: An integrated approach. Mutat. Res. 533, 37–65.
Schmeisser, E., Goessler, W., Kienzl, N., and Francesconi, K. A. (2004). Volatile analytes formed from arsenosugars: Determination by HPLC-HG-ICPMS and implications for arsenic speciation analyses. Anal. Chem. 76, 418–423.
Shi, H., Hudson, L. G., and Liu, K. J. (2004). Oxidative stress and apoptosis in metal ion-induced carcinogenesis. Free Radic. Biol. Med. 37, 582–593.
Shi, J., Xiao, Z., Ihnat, M. A., Kamat, C., Pandit, B., Hu, Z., and Li, P. K. (2003). Structure-activity relationships studies of the anti-angiogenic activities of linomide. Bioorg. Med. Chem. Lett. 13, 1187–1189.
Shinkaruk, S., Bayle, M., Lain, G., and Deleris, G. (2003). Vascular endothelial cell growth factor (VEGF). an emerging target for cancer chemotherapy. Curr. Med. Chem. Anti-Canc. Agents 3, 95–117.
Soucy, N. V., Ihnat, M. A., Kamat, C. D., Hess, L., Post, M. J., Klei, L. R., Clark, C., and Barchowsky, A. (2003). Arsenic stimulates angiogenesis and tumorigenesis in vivo. Toxicol. Sci. 76, 271–279.
Soucy, N. V., Klei, L. R., Mayka, D. D., and Barchowsky, A. (2004). Arsenic stimulates angiogenesis and tumorigenesis in vivo. Chem. Res. Toxicol. 17, 555–563.
Steinmaus, C., Moore, L., Hopenhayn-Rich, C., Biggs, M. L., and Smith, A. H. (2000). Arsenic in drinking water and bladder cancer. Cancer Invest. 18, 174–182.
Tchounwou, P. B., Patlolla, A. K., and Centeno, J. A. (2003). Carcinogenic and systemic health effects associated with arsenic exposure—a critical review. Toxicol. Pathol. 31, 575–588.
Tseng, C. H. (2002). An overview on peripheral vascular disease in blackfoot disease-hyperendemic villages in Taiwan. Angiology 53, 529–537.
Tseng, C. H., Tseng, C. P., Chiou, H. Y., Hsueh, Y. M., Chong, C. K., and Chen, C. J. (2002). Epidemiologic evidence of diabetogenic effect of arsenic. Toxicol. Lett. 133, 69–76.
Wang, G. L., and Semenza, G. L. (1993). Desferrioxamine induces erythropoietin gene expression and hypoxia-inducible factor 1 DNA-binding activity: Implications for models of hypoxia signal transduction. Blood 82, 3610–3615.
Yu, H. S., Lee, C. H., and Chen, G. S. (2002). Peripheral vascular diseases resulting from chronic arsenical poisoning. J. Dermatol. 29, 123–130.(Chandrashekhar D. Kamat, )