Arsenic Stimulates Angiogenesis and Tumorigenesis In Vivo

Nicole V. Soucy*, Michael A. Ihnat{dagger}, Chandrashekhar D. Kamat{dagger}, Linda Hess{dagger}, Mark J. Post{ddagger}, Linda R. Klei*, Callie Clark{dagger} and Aaron Barchowsky*,{dagger},1

* Departments of Pharmacology and Toxicology and {ddagger} Physiology, Dartmouth Medical School, Hanover, New Hampshire 03755; {dagger} Department of Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma 73190

Received June 27, 2003; accepted August 21, 2003


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION:
 REFERENCES
 
Trivalent inorganic arsenic (arsenite, arsenic trioxide, As[III]) is currently being used to treat hematologic tumors and is being investigated for treating solid tumors. However, low concentrations of As(III) stimulate vascular cell proliferation in cell culture, although this has not been confirmed in vivo. Therefore, the hypothesis that As(III) enhances blood vessel growth (angiogenesis) and tumorigenesis was tested in two in vivo models of angiogenesis and a model of tumor growth. In the first, arsenite caused a dose-dependent increase in vessel density in a chicken chorioallantoic-membrane (CAM) assay. The threshold As(III) concentration for this response was 0.033 µM and inhibition of vessel growth was observed at concentrations greater than 1 µM. Mouse Matrigel implants were used to test the angiogenic effects of As(III) in an adult mammalian system. Mice were injected with 0.8–80 µg/kg As(III)/day over a three-week period. During the last two weeks, Matrigel plugs were placed on the abdominal wall. Low and high doses of As(III) were synergistic with fibroblast growth factor-2 (FGF-2) in increasing vessel density in the Matrigel assay, while a middle dose had no effect. To test the effects of As(III) on tumor growth, GFP-labeled B16-F10 mouse melanoma cells were implanted in nude mice, which subsequently received biweekly injections of 0.5–5.0 mg/kg As(III). Significant tumor growth and lung metastasis was seen in all animals, with the largest tumors occurring in animals treated with lower doses of As(III). These studies support the hypothesis and indicate that induction of angiogenesis, enhanced tumor growth, and metastasis are potential dose-dependent toxic side effects of arsenic therapies.

Key Words: arsenic; angiogenesis; Matrigel; chicken chorioallantoic membrane; melanoma.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION:
 REFERENCES
 
Arsenicals have been used over the millennia to treat a variety of medical conditions ranging from infectious disease to cancer. Trivalent arsenic (As(III)) is currently FDA-approved for treating chronic or acute leukemias. It is also used in clinical trials to evaluate its efficacy for treating multiple myeloma and a variety of solid tumors (Anderson et al., 2002Go; Davison et al., 2002Go; Murgo, 2001Go; Waxman et al., 2001Go). Treatments consist of daily slow infusions of 0.15–0.3 mg/kg arsenic trioxide (0.76–1.32 µmol/kg) over a 30 to 60 day period. Although the pharmacokinetics of arsenic is difficult to determine, these therapies yield circulating levels of roughly 1–7 µmol/l. However, the mechanisms and specificity for the antitumor effects of these levels of As(III) remain unresolved.

Arsenic therapies and environmental exposures have been associated with significant cardiovascular effects. While the underlying causes are controversial, careful measures must be taken to avoid cardiac arrhythmias in patients receiving As(III) infusions to treat leukemias (Barbey, 2001Go; Ohnishi et al., 2000Go; Singer, 2001Go; Unnikrishnan et al., 2001Go). This is due to the propensity of As(III) to cause slow, progressive prolongation of the QT interval, which can result in potentially fatal Torsades de Pointes (Ohnishi et al., 2000Go; Unnikrishnan et al., 2001Go). In addition, prolonged injection of As(III) was recently shown to produce cardiac toxicities in mice (Li, Y. et al., 2002Go). Environmental exposures are associated with increased incidence of ischemic heart disease, arrhythmias, hypertension, and peripheral vascular disease (Engel et al., 1994Go; Lewis et al., 1999Go; Tseng et al., 2003Go; Wang et al., 2002Go). The mechanisms for these vascular effects are unknown. However, depending on the dose, As(III) stimulates clonal expansion of vascular smooth muscle, which may contribute to blood vessel occlusion (Engel et al., 1994Go; Wang et al., 2002Go). Cell culture-based studies have demonstrated differential, concentration-dependent effects of As(III) on vascular cells ranging from increased proliferation to increased apoptosis (Barchowsky et al., 1996Go; Roboz et al., 2000Go).

Angiogenesis, the development of new blood vessels from existing ones, is fundamental to growth and invasion in both solid and hematologic tumors (Herbst et al., 2002Go; Jung et al., 2002Go; Orpana et al., 2002Go; Papetti et al., 2002Go; Yang et al., 2002Go). One hypothesis for the antitumor effects of As(III) is that it inhibits angiogenesis by reducing expression of vascular endothelial cell growth factor (VEGF) or by being directly toxic to endothelial cells in high doses (Lew et al., 1999Go; Roboz et al., 2000Go). As discussed in several recent reviews, this hypothesis and the impetus for using arsenic as an antiangiogenic agent to treat solid tumors are based primarily on observations made in three studies (Anderson et al., 2002Go; Davison et al., 2002Go; Waxman et al., 2001Go). The first study demonstrated that injecting mice with high doses of As(III) (10 mg/kg) acutely collapses the tumor vasculature and sensitizes the tumor to radiation therapy (Lew et al., 1999Go). The second study demonstrated that clinically relevant concentrations of As(III) inhibited log-phase growth of endothelial cells in culture and VEGF expression in leukemic cells (Roboz et al., 2000Go). The relevance of the endothelial cell model used is questionable since endothelial cells in vivo are generally in close contact with matrix and not in log-phase growth. These studies have recently been repeated with slightly lower doses of As(III) and increased angiogenesis or endothelial cell growth and tube formation in three-dimension was found (Kao et al., 2003Go). These findings confirm an earlier report which demonstrated that As(III) causes dose-dependent endothelial cell proliferation in confluent, but not in log-phase cultures (Barchowsky et al., 1996Go). Finally, arsenic has been demonstrated to induce expression of angiogenic vascular endothelial cell growth factor (VEGF) in endothelial cells (Kao et al., 2003Go) and cervical cancer cells (Duyndam et al., 2001Go, 2003Go), but not in leukemic cells (Roboz et al., 2000Go). These conflicting results suggest that the effects of As(III) on the angiogenic process are dose-, cell-, and context-specific.

There have been no reports on the effects of As(III) on the angiogenic process in vivo. Therefore, the following studies were conducted to investigate the hypothesis that As(III) causes dose-dependent stimulation of blood vessel growth and thereby contributes to tumor growth. In the studies presented below, avian and mammalian in vivo models of angiogenesis were used to establish the dose-dependent capacity of As(III) to promote or inhibit the growth of new blood vessels. The range of As(III) doses used in these studies approaches the levels that are used clinically to treat leukemias and that are being evaluated for treating solid tumors. The data demonstrate complex kinetics with increased angiogenesis at both low and higher doses of As(III). A solid-tumor model in mice was used to demonstrate that instead of reducing tumor size, lower doses of injected As(III) increased both growth and metastasis. The data suggest that As(III) may not be a reasonable therapy for treating solid tumors, since low concentrations may enhance tumor growth, and high therapeutic levels may not limit this growth.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION:
 REFERENCES
 
Chorioallantoic membrane (CAM) assay of angiogenesis.
The CAM assay used was modified from the procedures of Sheu et al.(1998)Go and Brooks et al.(1998)Go. In this assay, fertile leghorn chicken eggs were allowed to grow until 10 days of incubation. Microbial testing disks were saturated with increasing concentrations of sodium arsenite (Fisher Scientific, Pittsburgh, PA) solutions, then placed onto the CAM, after breaking a small hole in the superior surface of the eggshell. After 24 h, the CAMs were perfused with 4% paraformaldehyde containing 0.05% triton X-100 for 10 s, further fixed for one min in 4% paraformaldehyde, placed onto petri dishes, and a digitized image taken using a dissecting microscope (Nikon USA) and a CCD imaging system (Scion Corporation, Frederick, MD). For each sample value, a 0.1 x 0.1-cm grid was added to the digital CAM images and the average number of vessels within 5–7 grids was counted as a measure of vascularity. The proangiogenic factors human VEGF-165 and FGF-2 (100 ng each; Peprotech, Rocky Hill, NJ) were used in combination as a positive control for angiogenesis and resulted in a 180–220% increase in blood vessel density in all experiments (data not shown).

Mouse Matrigel assay for angiogenesis.
Matrigel plug assays in mice were performed in accordance with institutional guidelines for animal safety and welfare, as previously described (Colorado et al., 2000Go; Swift et al., 1999Go). Normal male mice (C57BL/6NCr) 6–8 weeks of age and weighing ~20 g were obtained from the National Cancer Institute (Frederick, MD) and allowed to acclimate for 3–4 days prior to the start of treatments. Fresh sodium arsenite stock solutions (5 mmol/l) were prepared every other day and kept at 4°C. Dosing solutions of 0.162, 1.62, and 16.2 µg/ml were diluted in sterile H2O, and 0.1 ml was administered ip daily for 3 weeks to groups of four animals per dose. The average weight of the mice was 20 g, which yields a dose range of 0.8–80 µg/kg/day. The rationale for this dosing range was to provide daily injections of 0.125 to 12.5 nmol of As(III)/animal/day. Assuming a 2.5 ml blood volume, this would cause circulating levels of 0.05 to 5.0 µmol/l, which are relevant to anti-leukemic therapies and are comparable to the range of As(III) concentrations used in the CAM assays. Control animals received 0.1 ml of sterile water ip daily. Following the first week of As(III) injections, the mice were anesthetized with 0.2 ml avertin (35.3 mmol/l 2,2,2 tribromomethanol; Aldrich, Milwaukee, WI) dissolved in 1:40 2-methyl-2-butanol (Fisher Chemical Co, Fairlawn, NJ):H20). Matrigel (0.3–0.4 ml; Fischer Scientific, Pittsburgh, PA) was then injected between the skin and abdominal muscle of each mouse. The Matrigel was injected alone or supplemented with 50 ng/ml or 250 ng/ml of recombinant FGF-2. This range of concentrations of As(III) yielded no overt signs of toxicity, either in behavior or body weight over the course of the three-week study. At necropsy, the animals exhibited no signs of infection at the site of either the Matrigel or arsenic injections.

Histopathology.
Matrigel plugs were removed two weeks after implantation and were fixed in 10% neutral buffered formalin prior to paraffin processing. Sections were stained with H&E and assessed microscopically for blood vessel formation. Blood vessels were counted in ten non-overlapping fields of each section at 400 x magnification. The number of the vessels in all 10 fields from a single animal is summed and the data are reported as the mean sum ± SD for each treatment group.

Preparation of B16-F10 (GFP) cells.
B16-F10 mouse melanoma cells were purchased from American Type Tissue Collection (Manassas, VA) and maintained in Dulbecco’s minimal essential media (DMEM) containing 10% FBS and penicillin/streptomycin. To enhance the ability to detect metastatic tumor cells, stable transfectants that express green fluorescent protein (GFP) were prepared. The pLNCX-GFP vector was prepared by removing the CMV immediate-early gene promoter, EGFP protein, and the polyadenylation signal from the pEGFP-C1 vector (Clontech) and inserting this into the multiple cloning site of pLNCX retroviral vector (Clontech). This vector was transfected into PT67 packaging cells (Clontech), viral titer determined, and high-titer virus-containing supernatant used to infect B16-F10 cells overnight. Cells were then grown in G418 selection media for one week, resulting in GFP-expressing cells isolated using fluorescence automated cell sorting (FACS), and clones were selected by limiting dilutions. A single clone designated B16-F10 (GFP), to be used in the mouse tumor experiments, was selected for its high GFP fluorescence and for a doubling time similar to that of the parental cells.

Mouse tumor studies.
All animal studies were conducted in accordance with AALAC-approved guidelines using a protocol approved by the 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 chow, water, and bedding. Following one week of acclimation, 5 x 105 B16-F10 (GFP) cells in 100 µl PBS/1 µmol/l EDTA were injected subcutaneously into the external surface at the base of the right ear of mice. At days 7, 11, 14, 18, 21, and 25, post-implantation, animals received injections of 0, 0.5, 1, or 5 mg/kg arsenic trioxide (Sigma Chemical), ip, in 100 µl sterile water for injection. These higher levels of As(III) produce steady-state blood levels in the mice that are comparable to those seen in leukemic patients receiving As(III) therapy (Lallemand-Breitenbach et al., 1999Go). Animals were assessed for the presence of tumors every other day, and tumor measurement began 10 days after implantation and continued biweekly until tumors reached approximately 8–10% of animals’ body weights. The length (L) and width (W) of tumors were measured using Vernier calipers (Mitutoyo, Kawasaki Kanagawa, Japan) and tumor volume calculated using the formula L x W2. After sacrifice, the tumor and lungs were excised, fixed overnight in neutral buffered formalin, embedded in paraffin, and sectioned.

Immunohistochemical staining.
Mouse tumor sections were de-paraffinized using traditional protocols and then incubated in 6% peroxide for 20 min to further minimize melanin coloration. The sections were then permeabilized using 0.1 mM CaCl2, 0.1 mM MgCl2, and 0.1 mM MnCl2 containing 1% Triton X-100 in PBS for 15 min at room temperature. Antigen retrieval was accomplished in 0.1 M sodium citrate brought to boiling in a microwave oven at high power and then heating was continued at 30% power for 10 min. Tissues were blocked with rabbit serum (VectaStain ABC rabbit kit; Vector Labs, Burlingame, CA), and antibody against Tie-2 (Santa Cruz; sc9026; 1:500 dilution) was added for 4 h at room temperature. Secondary antibody, conjugated to horseradish peroxidase, and Vector® VIP peroxidase substrate were added according to protocol (Vector ABC). Sections were then counterstained with Gomori trichrome stain (Harleco, Gibbstown, NJ).

Statistics.
All analysis was done using Prism 3.0. (GraphPad Software, San Diego, CA). One-way ANOVA was conducted, with multiple comparisons tested by the Bonferroni post hoc test at a significance level of p >= 0.05. Tumor volumes were compared using two-way ANOVA, with treatments and days after implantation as independent variables.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION:
 REFERENCES
 
As(III) Induces Angiogenesis in the Chick CAM
The hypothesis that low levels of As(III) promote angiogenesis was examined in the in utero CAM assay. Chick CAMs were exposed to filter-paper disks containing increasing concentrations of As(III) for 6–48 h. The vascularity was then determined in each membrane. Vascularity, or blood vessel density, increased significantly in response to 0.033–1 µmol/l As(III). Above 1 µmol/l, As(III) caused significant injury and decreased vessel density (Fig. 1Go). At the highest concentrations of As(III), only pools of blood are observed instead of vessels (Fig. 1Go). Depending on the dose, the angiogenic effects of As(III) were observed as early as 6 h after As(III) treatment and were more pronounced at 48 than at 24 h (data not shown). There was no overt effect of the As(III) on embryonic development or viability.



View larger version (70K):
[in this window]
[in a new window]
 
FIG. 1. As(III) induced blood vessel formation in the CAM assay: Disks containing the indicated concentrations of As(III) were applied to chorioallantoic membranes of 10-day-old chicken embryos for 24 h. Membranes were imaged (A) and average blood vessel density was determined for each sample, as described in Materials and Methods. Data in B are expressed as the mean ± SD vessel density per treatment as a percentage of vehicle-treated controls; **p < 0.05. These data are representative of three separate experiments.

 
As(III) Stimulates Angiogenesis in the Mouse Matrigel Model
The CAM assay examines angiogenesis in the context of a rapidly developing embryonic system that is a mixture of vasculogenesis and angiogenesis. Therefore, the mouse Matrigel model of angiogenesis was used to investigate whether As(III) is capable of stimulating angiogenesis in an adult mammalian model. For these experiments, arsenite was given in daily ip injections because of ease of administration and based on previous reports of animal studies that mimicked therapeutic daily arsenite administration. The images in Figure 2Go illustrate the morphology of the Matrigel plugs following two weeks of incubation in control and As(III)-treated animals. The main morphological differences seen in the Matrigel from As(III) injected animals were dose-dependent increases in cellular infiltrates and luminal structures containing red blood cells (see arrows). Only these structures were counted as blood vessels, to avoid confusion with other structures such as lymph vessels or artifacts. In these studies, the Matrigel was primed with a threshold concentration of FGF-2, to induce a minimal rate of angiogenesis. As a control, separate groups of animals were injected with Matrigel that lacked FGF-2. As(III) did not increase blood vessel density in plugs unless FGF-2 was added (Figs. 3A Go and 3BGo). However, in the presence of FGF-2, As(III) caused a multiphasic effect of on blood vessel density (Fig. 3BGo). Low and high doses of As(III) synergistically increased the number of blood vessels, while a middle dose was no different from control. It should be noted that only luminal vessels containing fixed red blood cells were counted in these assays. Thus, while this method is quantitative, it may have underestimated the number of capillaries present in the Matrigel plugs (Claffey et al., 2001Go) and the results may reflect enhanced remodeling.



View larger version (61K):
[in this window]
[in a new window]
 
FIG. 2. Morphological features of Matrigel plugs in C57-BL6 mice: Mice were treated with daily ip injections of sodium arsenite for three weeks. Matrigel plugs containing the indicated amount of FGF-2 were established in the abdominal wall for the later two weeks, harvested, and sectioned. The images are hematoxylin/eosin stains of cross sections of the plugs at the border with the abdominal muscle at 400 x magnification. The arrows point to luminal structures that contain red blood cells, which were counted to produce the graphical data in Figure 3Go. These experiments have been reproduced three times.

 


View larger version (13K):
[in this window]
[in a new window]
 
FIG. 3: Effect of As(III) on blood vessel formation in Matrigel plugs: Matrigel plugs harvested from mice injected with the indicated amounts of As(III) were harvested, sectioned, and stained as indicated in Figure 2Go. The numbers of luminal vessels containing fixed red cells in 10 separate fields of each section were summed. The data are presented as the mean ± SD vessels in cross sections from four or eight animals in each treatment group in A or B, respectively. It is important to note that the data in B are the composite of two separate experiments with four animals per group in both experiments. Significant differences from groups receiving control injections are designated by *p < 0.05 or **p < 0.001.

 
As(III) Alters Solid Tumor Growth Properties
To test for differential, dose-dependent effects of As(III) on the growth of tumors, biweekly As(III) injections were started in nude mice seven days after the subcutaneous implantation of GFP-labeled B16-F10 mouse melanoma cells. The highest dose of As(III) used was given in a biweekly antiangiogenic, "metronomic" dosing schedule comparable to current therapeutic strategies (Hanahan et al., 2000Go; Li et al., 2002Go). Surface measurement indicated that significant tumor growth occurred in all of the animals in the study, regardless of treatment group (Fig. 4Go). Animals that received either 0.5 or 1.0 mg/kg of As(III) had significantly (p < 0.05) larger tumor volumes when compared with untreated animals. Staining of paraffin sections of the 28-day tumors with Tie-2, an endothelial-specific antibody (Peters et al., 1998Go), indicated that the blood vessel size and density showed changes, which depended on the dose of As(III). As seen in Figure 5AGo, the blood vessels in tumors of untreated animals were typically small and diffuse. In animals treated with 0.5 mg/kg As(III), there was an increase in both the size and the number of blood vessels in resulting tumors (Fig. 5BGo). In contrast, the blood vessel content of tumors of animals treated with 5 mg/kg of As(III) was similar to that of untreated animals (Fig. 5DGo). The B16-F10 mouse melanoma is known to preferentially metastasize to the lung, due to selective adhesion molecules found on the lung endothelium (Goetz et al., 1996Go). Therefore, GFP-positive masses in the mouse lungs were counted as a measure of metastasis. As seen in Figure 6Go, all doses of As(III) tended to increase the number of metastases seen in the lungs of tumor-bearing mice.



View larger version (18K):
[in this window]
[in a new window]
 
FIG. 4. Effects of bi-weekly As(III) administration on the growth of B16-F10 (GFP) primary tumors: B16-F10 (GFP) tumor cells (5 x 105) were implanted subcutaneously in the outer ears of male NCr nu/nu mice. After seven days, the mice were treated with biweekly ip injections of the indicated doses of As(III). Tumor size was measured using Vernier calipers on the indicated days, and tumor volumes were calculated as in Materials and Methods. Data represent the average tumor volumes at days after implantation. Significant growth was observed in all groups (p < 0.05, n = 6–10 mice) and significant differences between treatment groups (p < 0.05, as determined by two-way ANOVA) and controls are designated by *. These data are representative of two separate experiments.

 


View larger version (75K):
[in this window]
[in a new window]
 
FIG. 5. Effect of As(III) on tumor blood vessels. At 28 days, the mice presented in Figure 4Go were sacrificed and their tumors were embedded in paraffin. All animals received tumors and the treatment of the tumors is given below each panel. Immunohistochemical staining for endothelial cell Tie-2 was performed as described in Materials and Methods. Tie-2-containing endothelial cells stained purple with Vector® VIP peroxidase substrate, and the sections were counterstained with Gomori trichrome, which stained tumor pale blue/green. Images were obtained with 40 x magnification on an Olympus AX70 microscope fitted with a SPOT CCD camera. The images are representative of sections of tumors from 6–10 animals in each treatment group. The epidermis (EPI), tumor (T), and some blood vessels (BV) are labeled for orientation.

 


View larger version (13K):
[in this window]
[in a new window]
 
FIG. 6. Effects of As(III) administration on B16-F10(GFP) tumor metastasis: The lungs from the animals presented in Figure 4Go were examined for GFP-positive masses. The data are presented as the mean ± SD number of metastases found in each treatment group. Significant differences between the groups, *p < 0.05.

 

    DISCUSSION:
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION:
 REFERENCES
 
There are no previous reports of the effects of As(III) on the angiogenic process in vivo. Others have examined effects of arsenic on endothelial cell growth and tube formation in cell-culture models and have found mixed results (Barchowsky et al. 1996Go, 1999Go; Kao et al., 2003Go; Roboz et al., 2000Go). The differences reported in the cell culture models may relate to the concentrations of As(III) used or to the context of the cells when exposed (Barchowsky et al., 1996Go; Kao et al., 2003Go; Roboz et al., 2000Go). A consistent finding is that As(III) is toxic to endothelial cells in culture when the threshold of 5 µmol/l is exceeded, especially when the cells are in the log phase of growth (Barchowsky et al., 1996Go, 1999Go; Deneke, 1992Go; Roboz et al., 2000Go). Despite certain limitations, the CAM and Matrigel assays examine angiogenesis in a more appropriate context than the cell culture models. Thus, the results obtained in the current study more firmly demonstrate that the vascular effects of As(III) are dose-dependent and context-specific. In the three-dimensional context of matrix and surrounding cells, low doses of As(III) synergize with angiogenic factors to potently stimulate new blood vessel formation (Figs. 1–Go3Go). Enhancement of the angiogenic potential of the blood vessels following exposure to lower doses of As(III) may be clinically significant to tumorigenesis and as a risk when As(III) is used to treat solid tumors.

The angiogenic process in normal tissue and in disease has been extensively reviewed (Carmeliet et al., 2000Go; Papetti and Herman, 2002Go; Shacter et al., 2002Go). In general, endothelial cells and the vasculature are very stable in normal adult tissues until wounded. When stimulated, angiogenesis is highly ordered and tightly regulated by the release of positive and negative factors that include several soluble peptides, growth factors, inflammatory mediators, cell–cell and cell-matrix interactions, and hemodynamic effects (Papetti et al., 2002Go; Shacter et al., 2002Go). Thus, modeling angiogenesis in cell culture compromises the ability to investigate many potential mechanisms for the pro-angiogenic effects of specific compounds such as As(III). For example, previous studies from this laboratory demonstrated that As(III) directly stimulated postconfluent endothelial cells to proliferate (Barchowsky et al., 1996Go, 1999Go). It is likely that these proliferative signals are generated by stimulation of a growth pathway involving low level generation of reactive oxygen species following As(III)-induced activation of vascular NADPH oxidase and tyrosine kinases (Barchowsky et al., 1999Go; Smith et al., 2001Go). However, in vivo proliferative and angiogenic stimuli are more likely to arise from cells surrounding the endothelial cells and not the endothelial cells themselves (Papetti et al., 2002Go). Thus, monoculture of endothelial cells may not model the full complement of effects for agents that promote angiogenesis by stimulating release of factors from pericytes, and vascular smooth muscle, epithelial, and tumor cells.

One of the difficulties with As(III) as a therapeutic agent is that the dose dependence of its cellular and molecular effects are nonlinear and are cell- or context-specific. It is apparent from recent reports that the potential for As(III) to stimulate tumor cells to release angiogenic factors depends on the tumor type and the As(III) concentration (Duyndam et al., 2001Go; Kao et al., 2003Go; Roboz et al., 2000Go; Roboz and Roboz, 2000Go) demonstrated that five µmol/l As(III) inhibited VEGF release from a human leukemic cell line and inhibited the proliferation of subconfluent human endothelial cells. In contrast, As(III) increased VEGF expression in both human umbilical veins and in cervical cancer cells (Duyndam et al., 2001Go). The studies presented in Figures 4 Goand 5Go suggest that low doses of As(III) promote growth and metastasis of the B16-F10 tumors in vivo. The highest dose of 5 mg/kg As(III) biweekly trends towards less tumor growth (Fig. 4Go), but is the most significant for lung metastasis (Fig. 6Go). The difference in tumor growth appears to result from differential effects on blood vessel density and tumor morphology. Low-dose As(III) greatly increases the density of blood vessels in the tumors, and tumor size is increased (Fig. 5BGo). Blood vessel size was decreased to near control levels by the highest dose of As(III), but the number of vessels was still high (Fig. 5CGo). At high dose, the tumors were of a looser morphology and there were more metastases. Thus, there may have been more tumor masses invading into tissue at the highest dose of As(III), but this dose may have retarded the growth of the invading blood vessels and melanoma cells.

The data in Figure 3Go suggest that dose is an extremely important consideration and indicate that As(III) produces nonlinear effects. The general implication is that there is no threshold for As(III) effects, since it is quite easy to arrive at doses that are both pro- and antiangiogenic. The clinical implication of these nonlinear effects is that there may be only a narrow window of blood levels that will provide both antiangiogenic and antitumor effects. In humans, the pharmacokinetics of a single dose of 10 mg arsenic trioxide IV shows a peak blood level of approximately 7 µmol/l within four h of ip administration (Shen et al., 1997Go). Blood levels fall, however, to 1 µmol/l at around 10 h after arsenic administration, and reach approximately 0.3 µmol/l at 24 h after administration. The peak levels from this ip dose could cause endothelial cell toxicity (Barchowsky et al., 1996Go; Roboz et al., 2000Go) and do appear to decrease blood vessel density in the CAM assay (Fig. 1Go). In contrast, the highest dose used in the mouse Matrigel study of approximately 0.08 mg/kg per day produced significant angiogenesis (Fig. 3Go) and higher biweekly doses of 0.5 and 1.0 mg/kg seemed to enhance tumor growth (Fig. 4Go). Taken together, these data suggest that caution should be taken when dosing patients with solid tumors that are dependent on blood vessel growth.

The complexity of the As(III) actions was indicated by the observation that As(III) had no effect in the Matrigel assays unless a threshold level of FGF-2 was added to the plugs (Fig. 3Go). It is common that systemically administered agents fail to produce angiogenesis in this assay when FGF-2 is absent or until inflammation is present (Claffey et al., 2001Go; Shacter et al., 2002Go). In contrast, As(III) added alone increased blood vessel density in the CAM assay (Fig. 1Go) and tumors tended to be larger in response to lower doses of As(III) (Fig. 4Go). In these latter cases, synergy may be provided by the high background level of angiogenic and remodeling factors normally present in developing systems or in tumors. The major significance derived from these studies is that As(III) facilitates the angiogenic process, but may not act as a direct promoter. For many years, arsenic has been associated with various types of cancer, including skin cancers (Hughes, 2002Go). Historically, however, it has been very difficult to observe a direct carcinogenic effect of arsenic (Hughes, 2002Go; Rossman et al., 2002Go). Instead, arsenic is now thought to be a cocarcinogen when given with known carcinogens such, as UV radiation (Rossman et al., 2002Go). The data in the current study indicate that low doses of As(III) are capable of inducing angiogenesis, a process critical to the growth of most tumors. This provides another means by which low-dose arsenite can enhance tumor growth and may help to explain the cocarcinogenicity of this compound.

The mechanisms for As(III)-induced angiogenesis are unclear. As(III)-induced proliferative responses in a variety of cultured cell types, including vascular cells, are associated with increased Src-tyrosine-kinase activity (Barchowsky et al., 1999Go; Simeonova et al., 2002Go), enhanced growth factor responses (Kao et al., 2003Go; Simeonova et al., 2002Go; Trouba et al., 1999Go), increased MAP kinase activity (Bode et al., 2002Go; Trouba et al., 2000Go) of and activation of transcription factors, such as AP-1 and NF-{kappa}B (Barchowsky et al., 1996Go, 1999Go; Bode et al., 2002Go; Drobna et al., 2003Go; Simeonova et al., 2002Go). Most recently, Kao et al.(2003)Go demonstrated that concentrations of arsenite below 5 µM induced angiogenesis and increased the expression of the angiogenic mitogen, VEGF. It is known that VEGF and FGF-2 synergize to induce pathological angiogenesis in vivo (Lindner et al., 2001Go). While synergy between an As(III)-induced mitogenic response, such as increased VEGF expression, and FGF-2 is a plausible mechanism, many more studies will be required to define this pathway as rate limiting in vivo. In addition, few studies have examined sustained signaling events that might manifest during the one-week As(III) priming period used in the studies shown in Figures 2Go and 3Go, which allowed for synergistic increases in angiogenesis. It is not unreasonable to speculate that these long-term signaling or phenotypic changes contribute to the etiology of diseases caused by chronic environmental arsenic exposures.

In conclusion, these studies present the dose-dependent effects of As(III) in two in vivo models of angiogenesis. Depending on the systemic level, arsenic can promote or reduce angiogenesis. Further, As(III) may synergize with other factors to promote cardiovascular remodeling and tumor growth. These effects occur at levels of arsenic that are relevant to both cancer therapies and environmental exposures. Thus, the implication of the results of these studies is that caution should be used when treating solid tumors with arsenic. Dosing at subtherapeutic levels may actually enhance tumor growth and tumoristatic doses may enhance tissue remodeling. For environmental exposures, the ability of As(III) to induce angiogenesis may represent an axis for the progression of vascular disease. For clinical use as an anticancer agent, As(III)-induced angiogenesis may represent an unacceptable toxic side effect.


    ACKNOWLEDGMENTS
 
This work was supported by the Superfund Basic Research Program (ES07373) and the services of the Norris Cotton Cancer Center. The Ruth Estrin Goldberg Cancer Center Research Memorial to M.A.I. provided additional support.


    NOTES
 
1 To whom correspondence should be sent at the present address: Department of Environmental and Occupational Health, University of Pittsburgh, 3343 Forbes Avenue, Room 205 FORBL, Pittsburgh, PA 15260. Fax: (412) 383-2123. Email: abarchowsky{at}ceoh.pitt.edu. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION:
 REFERENCES
 
Anderson, K. C., Boise, L. H., Louie, R., and Waxman, S. (2002). Arsenic trioxide in multiple myeloma: Rationale and future directions. Cancer J. 8, 12–25.[ISI][Medline]

Barbey, J. T. (2001). Cardiac toxicity of arsenic trioxide. Blood 98, 1632–1634.[Free Full Text]

Barchowsky, A., Dudek, E. J., Treadwell, M. D., and Wetterhahn, K. E. (1996). Arsenic induces oxidant stress and NF-{kappa}B activation in cultured aortic endothelial cells. Free Radic. Biol. Med. 21, 783–790.[CrossRef][ISI][Medline]

Barchowsky, A., Roussel, R. R., Klei, L. R., James, P. E., Ganju, N., Smith, K. R., and Dudek, E. J. (1999). Low levels of arsenic trioxide simulate proliferative signals in primary vascular cells without activating stress effector pathways. Toxicol. Appl. Pharmacol. 159, 65–75.[CrossRef][ISI][Medline]

Bode, A. M., and Dong, Z. (2002). The paradox of arsenic: Molecular mechanisms of cell transformation and chemotherapeutic effects. Crit. Rev. Oncol. Hematol. 42, 5–24.[ISI][Medline]

Brooks, P. C., Silletti, S., von Schalscha, T. L., Friedlander, M., and Cheresh, D. A. (1998). Disruption of angiogenesis by PEX, a noncatalytic metalloproteinase fragment with integrin binding activity. Cell 92, 391–400.[ISI][Medline]

Carmeliet, P., and Jain, R. K. (2000). Angiogenesis in cancer and other diseases. Nature 407, 249–257.[CrossRef][ISI][Medline]

Claffey, K. P., Abrams, K., Shih, S. C., Brown, L. F., Mullen, A., and Keough, M. (2001). Fibroblast growth factor-2 activation of stromal cell vascular endothelial growth factor expression and angiogenesis. Lab. Invest. 81, 61–75.[ISI][Medline]

Colorado, P. C., Torre, A., Kamphaus, G., Maeshima, Y., Hopfer, H., Takahashi, K., Volk, R., Zamborsky, E. D., Herman, S., Sarkar, P. K., Ericksen, M. B., Dhanabal, M., Simons, M., Post, M., Kufe, D. W., Weichselbaum, R. R., Sukhatme, V. P., and Kalluri, R. (2000). Anti-angiogenic cues from vascular basement membrane collagen. Cancer Res. 60, 2520–2526.[Abstract/Free Full Text]

Davison, K., Mann, K. K., and Miller, W. H., Jr. (2002). Arsenic trioxide: Mechanisms of action. Semin. Hematol. 39, 3–7.[CrossRef]

Deneke, S. M. (1992). Induction of cystine transport in bovine pulmonary artery endothelial cells by sodium arsenite. Biochim. Biophys. Acta 1109, 127–131.[ISI][Medline]

Drobna, Z., Jaspers, I., Thomas, D. J., and Styblo, M. (2003). Differential activation of AP-1 in human bladder epithelial cells by inorganic and methylated arsenicals. FASEB J. 17, 67–69.[Free Full Text]

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 1{alpha} protein by the oxidative stressor arsenite. J. Biol. Chem. 276, 48066–48076.[Abstract/Free Full Text]

Duyndam, M. C., Hulscher, S. T., van der, W. 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.[Abstract/Free Full Text]

Engel, R. R., Hopenhayn-Rich, C., Receveur, O., and Smith, A. H. (1994). Vascular effects of chronic arsenic exposure: A review. Epidemiol. Rev. 16, 184–208.[ISI][Medline]

Goetz, D. J., el-Sabban, M. E., Hammer, D. A., and Pauli, B. U. (1996). Lu-ECAM-1-mediated adhesion of melanoma cells to endothelium under conditions of flow. Int. J. Cancer 65, 192–199.[CrossRef][ISI][Medline]

Hanahan, D., Bergers, G., and Bergsland, E. (2000). Less is more, regularly: Metronomic dosing of cytotoxic drugs can target tumor angiogenesis in mice. J. Clin. Invest. 105, 1045–1047.[Free Full Text]

Herbst, R. S., Hidalgo, M., Pierson, A. S., Holden, S. N., Bergen, M., and Eckhardt, S. G. (2002). Angiogenesis inhibitors in clinical development for lung cancer. Semin. Oncol. 29, 66–77.[CrossRef][ISI][Medline]

Hughes, M. F. (2002). Arsenic toxicity and potential mechanisms of action. Toxicol. Lett. 133, 1–16.[CrossRef][ISI][Medline]

Jung, Y. D., Ahmad, S. A., Liu, W., Reinmuth, N., Parikh, A., Stoeltzing, O., Fan, F., and Ellis, L. M. (2002). The role of the microenvironment and intercellular crosstalk in tumor angiogenesis. Semin. Cancer Biol. 12, 105–112.[CrossRef][ISI][Medline]

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.[CrossRef][ISI][Medline]

Lallemand-Breitenbach, V., Guillemin, M. C., Janin, A., Daniel, M. T., Degos, L., Kogan, S. C., Bishop, J. M., and de The, H. (1999). Retinoic acid and arsenic synergize to eradicate leukemic cells in a mouse model of acute promyelocytic leukemia. J. Exp. Med. 189, 1043–1052.[Abstract/Free Full Text]

Lew, Y. S., Brown, S. L., Griffin, R. J., Song, C. W., and Kim, J. H. (1999). Arsenic trioxide causes selective necrosis in solid murine tumors by vascular shutdown. Cancer Res. 59, 6033–6037.[Abstract/Free Full Text]

Lewis, D. R., Southwick, J. W., Ouellet-Hellstrom, R., Rench, J., and Calderon, R. L. (1999). Drinking-water arsenic in Utah: A cohort mortality study. Environ. Health Perspect. 107, 359–365.[ISI][Medline]

Li, D., Du, C., Lin, Y., and Wu, M. (2002). Inhibition of growth of human nasopharyngeal cancer xenografts in SCID mice by arsenic trioxide. Tumorigenesis 88, 522–526.

Li, Y., Sun, X., Wang, L., Zhou, Z., and Kang, Y. J. (2002). Myocardial toxicity of arsenic trioxide in a mouse model. Cardiovasc. Toxicol. 2, 63–73.[CrossRef][Medline]

Lindner, V., and Maciag, T. (2001). The putative convergent and divergent natures of angiogenesis and arteriogenesis. Circ. Res. 89, 747–749.[Free Full Text]

Murgo, A. J. (2001). Clinical trials of arsenic trioxide in hematologic and solid tumors: Overview of the National Cancer Institute Cooperative Research and Development Studies. Oncologist 6(Suppl. 2), 22–28.[Abstract/Free Full Text]

Ohnishi, K., Yoshida, H., Shigeno, K., Nakamura, S., Fujisawa, S., Naito, K., Shinjo, K., Fujita, Y., Matsui, H., Takeshita, A., Sugiyama, S., Satoh, H., Terada, H., and Ohno, R. (2000). Prolongation of the QT interval and ventricular tachycardia in patients treated with arsenic trioxide for acute promyelocytic leukemia. Ann. Intern. Med. 133, 881–885.[Abstract/Free Full Text]

Orpana, A., and Salven, P. (2002). Angiogenic and lymphangiogenic molecules in hematological malignancies. Leuk.. Lymphoma 43, 219–224.[CrossRef][ISI][Medline]

Papetti, M., and Herman, I. M. (2002). Mechanisms of normal and tumor-derived angiogenesis. Am. J. Physiol. 282, C947–970.[ISI]

Peters, K. G., Coogan, A., Berry, D., Marks, J., Iglehart, J. D., Kontos, C. D., Rao, P., Sankar, S., and Trogan, E. (1998). Expression of Tie2/Tek in breast tumour vasculature provides a new marker for evaluation of tumour angiogenesis. Br. J. Cancer 77, 51–56.[ISI][Medline]

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.[Abstract/Free Full Text]

Rossman, T. G., Uddin, A. N., Burns, F. J., and Bosland, M. C. (2002). Arsenite cocarcinogenesis: An animal model derived from genetic toxicology studies. Environ. Health Perspect. 110(Suppl. 5), 749–752.

Shacter, E., and Weitzman, S. A. (2002). Chronic inflammation and cancer. Oncology 16,217–226, 229.[Medline]

Shen, Z. X., Chen, G. Q., Ni, J. H., Li, X. S., Xiong, S. M., Qiu, Q. Y., Zhu, J., Tang, W., Sun, G., Yang, K. Q., Chen, Y., Zhou, L., Fang, Z. W., Wang, Y. T., Ma, J., Zhang, P., Zhang, T., Chen, S. J., Chen, Z., and Wang, Z. Y. (1997). Use of arsenic trioxide (As2O3) in the treatment of acute promyelocytic leukemia (APL): II. Clinical efficacy and pharmacokinetics in relapsed patients. Blood 89, 3354–3360.[Abstract/Free Full Text]

Sheu, J. R., Fu, C. C., Tsai, M. L., and Chung, W. J. (1998). Effect of U-995, a potent shark cartilage-derived angiogenesis inhibitor, on antiangiogenesis and antitumor activities. Anticancer Res. 18, 4435–4441.[ISI][Medline]

Simeonova, P. P., Wang, S., Hulderman, T., and Luster, M. I. (2002). c-Src-dependent activation of the epidermal growth factor receptor and mitogen-activated protein kinase pathway by arsenic. Role in carcinogenesis. J. Biol. Chem. 277, 2945–2950.[Abstract/Free Full Text]

Singer, J. W. (2001). Cardiac toxicity of arsenic trioxide. Blood 98, 1633–1634.[ISI][Medline]

Smith, K. R., Klei, L. R., and Barchowsky, A. (2001). Arsenite stimulates plasma membrane NADPH oxidase in vascular endothelial cells. Am. J. Physiol. 280, L442–449.[ISI]

Swift, M. E., Kleinman, H. K., and DiPietro, L. A. (1999). Impaired wound repair and delayed angiogenesis in aged mice. Lab. Invest. 79, 1479–1487.[ISI][Medline]

Trouba, K. J., Glanzer, J. G., and Vorce, R. L. (1999). Wild-type and Ras-transformed fibroblasts display differential mitogenic responses to transient sodium arsenite exposure. Toxicol. Sci. 50, 72–81.[Abstract]

Trouba, K. J., Wauson, E. M., and Vorce, R. L. (2000). Sodium arsenite-induced dysregulation of proteins involved in proliferative signaling. Toxicol. Appl. Pharmacol. 164, 161–170.[CrossRef][ISI][Medline]

Tseng, C. H., Chong, C. K., Tseng, C. P., Hsueh, Y. M., Chiou, H. Y., Tseng, C. C., and Chen, C. J. (2003). Long-term arsenic exposure and ischemic heart disease in arseniasis-hyperendemic villages in Taiwan. Toxicol. Lett. 137, 15–21.[CrossRef][ISI][Medline]

Unnikrishnan, D., Dutcher, J. P., Varshneya, N., Lucariello, R., Api, M., Garl, S., Wiernik, P. H., and Chiaramida, S. (2001). Torsades de pointes in 3 patients with leukemia treated with arsenic trioxide. Blood 97, 1514–1516.[Abstract/Free Full Text]

Wang, C. H., Jeng, J. S., Yip, P. K., Chen, C. L., Hsu, L. I., Hsueh, Y. M., Chiou, H. Y., Wu, M. M., and Chen, C. J. (2002). Biological gradient between long-term arsenic exposure and carotid atherosclerosis. Circulation 105, 1804–1809.[Abstract/Free Full Text]

Waxman, S., and Anderson, K. C. (2001). History of the development of arsenic derivatives in cancer therapy. Oncologist 6(Suppl. 2), 3–10.[Abstract/Free Full Text]

Yang, R., and Han, Z. C. (2002). Angiogenesis in hematologic malignancies and its clinical implications. Int. J. Hematol. 75, 246–256.[ISI][Medline]