1α,25-Dihydroxyvitamin D3 Inhibits Angiogenesis In Vitro and In Vivo
Abstract—Modulation of angiogenesis is now a recognized strategy for the prevention and treatment of pathologies categorized by their reliance on a vascular supply. The purpose of this study was to evaluate the effect of 1α,25-dihydroxyvitamin D3 [1,25(OH)2D3], the active metabolite of vitamin D3, on angiogenesis by using well-characterized in vitro and in vivo model systems. 1,25(OH)2D3 (1×10−9 to 1×10−7 mol/L) significantly inhibited vascular endothelial growth factor (VEGF)-induced endothelial cell sprouting and elongation in vitro in a dose-dependent manner and had a small, but significant, inhibitory effect on VEGF-induced endothelial cell proliferation. 1,25(OH)2D3 also inhibited the formation of networks of elongated endothelial cells within 3D collagen gels. The addition of 1,25(OH)2D3 to endothelial cell cultures containing sprouting elongated cells induced the regression of these cells, in the absence of any effect on cells present in the cobblestone monolayer. Analysis of nuclear morphology, DNA integrity, and enzymatic in situ labeling of apoptosis-induced strand breaks demonstrated that this regression was due to the induction of apoptosis specifically within the sprouting cell population. The effect of 1,25(OH)2D3 on angiogenesis in vivo was investigated by using a model in which MCF-7 breast carcinoma cells, which had been induced to overexpress VEGF, were xenografted subcutaneously together with MDA-435S breast carcinoma cells into nude mice. Treatment with 1,25(OH)2D3 (12.5 pmol/d for 8 weeks) produced tumors that were less well vascularized than tumors formed in mice treated with vehicle alone. These results highlight the potential use of 1,25(OH)2D3 in both the prevention and regression of conditions characterized by pathological angiogenesis.
Angiogenesis, the formation of new blood vessels from an existing vascular bed, is of fundamental importance in several physiological and pathological processes, including atherosclerosis, diabetic retinopathy, psoriasis, and tumor growth and metastasis.1 2 The angiogenic process is characterized by a series of stages: (1) degradation of the vascular basement membrane, (2) sprouting, elongation, migration, and proliferation of endothelial cells, (3) association of endothelial cells into new tubular channels, and (4) synthesis of a new basement membrane, recruitment of perivascular cells, and vessel maturation.3 These processes are tightly controlled through a balance of positive and negative regulatory factors.2
Vascular endothelial growth factor (VEGF) is a key mediator of angiogenesis.4 5 This growth factor stimulates endothelial cell proliferation, sprouting, migration, and morphogenesis, acting via 2 high-affinity class II receptor tyrosine kinases, VEGF-R1 (flt-1) and VEGF-R2 (KDR/flk-1).5 VEGF is also a survival factor for newly formed vessels6 7 and is essential for embryonic development, because mice deficient in VEGF, VEGF-R1, or VEGF-R2 die in utero of vascular defects.4 5
The reliance of many pathological conditions on angiogenesis has led to the proposal that such diseases can be controlled by use of antiangiogenic compounds that inhibit or regress newly forming blood vessels.2 The effect of 1α,25-dihydroxyvitamin D3 [1,25(OH)2D3] on angiogenesis is unclear, having been reported to decrease8 or have no effect9 10 on endothelial cell proliferation, to have no effect on the formation of capillary tubelike structures in vitro,11 and to inhibit angiogenesis in vivo.12
The purpose of the present study was to examine the effect of 1,25(OH)2D3 on the proliferation, sprouting, and elongation of endothelial cells and on their subsequent association into multicellular cords and tubelike structures by using well-characterized in vitro model systems.13 14 15 The effect of 1,25(OH)2D3 on tumor growth and angiogenesis in vivo was investigated in a tumor model by using cells that had been induced to overexpress VEGF.16 Accordingly, we demonstrate that 1,25(OH)2D3 inhibits angiogenesis both in vitro and in vivo. Furthermore, we show that 1,25(OH)2D3 specifically induces the regression of preformed endothelial cell sprouts, multicellular cords, and networks of elongated endothelial cells in vitro, and we demonstrate the involvement of apoptosis in this process.
Materials and Methods
Bovine aortic endothelial cells (BAECs) were cultured in 10% FCS-MEM on gelatin-coated dishes.13 Recombinant human VEGF (VEGF165, 20 ng/mL, from Dr D. Ogilvy, AstraZeneca Pharmaceuticals, Macclesfield, UK) and 1,25(OH)2D3 (1×10−9 to 1×10−7 mol/L, Roche) were added when the medium was changed (every 2 days). Control cultures received vehicle alone (10 μL ethanol/mL medium). Duplicate dishes were used in every experiment.
BAECs were plated in 10% FCS-MEM. After 24 hours (day 1), VEGF was added with or without 1,25(OH)2D3 in 2% FCS-MEM or serum-free MEM. Proliferation was assessed by counting cell numbers on days 1 and 613 and by measuring [3H]thymidine incorporation into trichloroacetic acid–insoluble material.17
Angiogenesis In Vitro
Confluent BAECs were incubated in 2% FCS-MEM containing VEGF with or without 1,25(OH)2D3 for 14 days. Angiogenesis was quantified by using a computerized image analysis system (Quantimet 600). Two cross-sectional diameters were assessed, field-by-field (20 fields per dish). Elongated sprouting cells were delineated digitally, and the area occupied by these cells was calculated and expressed as percentage of the total field area. The effect of 1,25(OH)2D3 on preformed endothelial cell sprouts was investigated by incubating confluent endothelial cells with VEGF for 4 days before the addition of 1,25(OH)2D3.
The effect of 1,25(OH)2D3 on later stages of angiogenesis was investigated by culturing cells within 3D collagen gels.13 After 2 hours in 10% FCS-MEM, the cells were cultured in 2% FCS-MEM with or without 1,25(OH)2D3 for 5 days. To determine the effect of 1,25(OH)2D3 on preformed networks of elongated endothelial cells within collagen gels, 1,25(OH)2D3 was added after 4 days of culture in 10% FCS-MEM.
Apoptosis was assessed by using the In Situ Cell Death Detection Kit (Boehringer-Mannheim). Cell nuclei were examined by fluorescence microscopy after staining with acridine orange. The integrity of DNA extracted from cells cultured on gelatin-coated dishes in the presence of VEGF and 1,25(OH)2D3 was assessed.18
In Vivo Assays
Female BALB/c nu/nu mice, aged 6 weeks (Charles River Ltd, Margate, UK), received 2 dorsal subcutaneous implants under halothane anesthesia: a slow release pellet of estradiol (1 mg) and a 14-day micro-osmotic pump (Alza), which released either 12.5 pmol/d 1,25(OH)2D3 in propylene glycol (4 mice) or vehicle alone (4 mice). Six days later, each mouse received, under fluothane anesthesia, a subcutaneous implant in each flank of 107 MCF7 breast cancer cells transfected with VEGF plus 107 MDA-435S breast cancer cells (from Dr R. Bicknell, University of Oxford, Oxford, UK).16 Pumps were replaced after 14 days. After 28 days, both 1,25(OH)2D3 and estradiol were provided by weekly subcutaneous injections in propylene glycol. Tumor growth was monitored weekly. After 8 weeks, mice were killed by an excess of anesthetic, and the tumors were harvested. Histochemical staining of frozen tumor sections and assessment of microvessel density was performed.16 MEC13.3 (rat anti-mouse CD31) was kindly provided by Dr A. Vecchi, Milano, Italy. Experiments were performed in accordance with UK Home Office Guidelines.
Differences between continuous variables were examined by the Student t test. Multiple comparisons of continuous data were examined by 1-way ANOVA with post hoc comparisons using the method of Bonferroni. Comparisons of proportionality were performed by calculation of the z statistic. The Mann-Whitney U test was used to compare microvessel density in different tumor samples.
1,25(OH)2D3 Inhibits VEGF-Induced Endothelial Cell Proliferation
1,25(OH)2D3 was added to semiconfluent BAECs cultured in the presence of VEGF, and the effects of these factors on cell proliferation were determined. VEGF stimulates endothelial cell proliferation (P<0.01) (Figure 1⇓ and Online Table 1, available at http://www.circresaha.org). 1,25(OH)2D3 inhibited VEGF-induced endothelial cell proliferation, and this inhibition was dose dependent [P<0.05 at 1×10−7 mol/L 1,25(OH)2D3] (Figure⇓ 1 and Online Table 1, available at http://www.circresaha.org). Final cell counts in cultures incubated with VEGF and 1×10−7 mol/L 1,25(OH)2D3 were equivalent to those incubated in control medium (Figure 1⇓).
1,25(OH)2D3 Inhibits Angiogenesis In Vitro
To determine the effect of 1,25(OH)2D3 on the early stages of growth factor–induced angiogenesis in vitro, VEGF was added to confluent BAECs in the presence and absence of 1,25(OH)2D3, and the formation of elongated sprouting endothelial cells was investigated and quantified by image analysis (Figures 2⇓ and 3⇓). Control endothelial cells remained as monolayers of polygonal cells, characteristic of the “resting” phenotype of quiescent endothelial cells (Figure 2A⇓). A small number of single, elongated, sprouting cells appeared in the control cultures as the experiment progressed; however, these cells remained isolated and few in number. VEGF induced the formation of sprouting elongated cells underneath the cobblestone monolayer that combined to form networks of interconnected multicellular cords (Figure⇑ 2B). The simultaneous addition of 1,25(OH)2D3 and VEGF significantly inhibited the morphogenetic effect of this growth factor in a dose-dependent manner [P<0.05 at 1×10−9 mol/L 1,25(OH)2D3; P<0.01 at 1×10−8 to 1×10−7 mol/L 1,25(OH)2D3] (Figures 2 and 3). The few sprouting cells that were present were less elongated than those formed in the absence of 1,25(OH)2D3, and these cells remained isolated and did not combine into networks (Figure⇑ 2C). The addition of 1,25(OH)2D3 (1×10−7 mol/L) alone inhibited the formation of endothelial cell sprouts in control medium (P<0.01) but had no effect on the morphology of the cells present in the cobblestone monolayer (Figures 2D and 3).
To investigate further the effect of 1,25(OH)2D3 on endothelial cell morphogenesis, cells were plated within collagen gels. Under these conditions, endothelial cells elongate, migrate, and self-associate into tubelike structures in the absence of cell proliferation.13 Control BAECs plated in collagen gels adopted an elongated morphology and associated with each other to form networks of cells (Figure 4A⇓). In contrast, cells cultured in the presence of 1,25(OH)2D3 remained rounded (Figure 4B⇓).
1,25(OH)2D3 Induces Regression of Existing Sprouting Endothelial Cells
The effect of 1,25(OH)2D3 on preexisting elongated sprouting cells was investigated. Confluent endothelial cells were induced to sprout in the presence of VEGF. Cultures were then incubated in the presence or absence of 1,25(OH)2D3 without growth factor. In control cultures, networks of elongated sprouting cells were still present 6 days after the withdrawal of VEGF (Figure 5A⇓). Incubation of cells with 1,25(OH)2D3 resulted in a dose-dependent regression of the sprouting cells. Moreover, the sprouting cells present in these cultures were less elongated and heavily vacuolated compared with control cultures (Figure⇑ 5B).
The effect of 1,25(OH)2D3 on endothelial cell morphogenesis in collagen gels was also investigated. The addition of 1,25(OH)2D3 resulted in the regression of the elongated sprouting cells to rounded cells (Figure⇑ 5D). In contrast, control cells remained elongated (Figure⇑ 5C).
1,25(OH)2D3 Specifically Induces Apoptosis of Sprouting Endothelial Cells
The induction of apoptosis by 1,25(OH)2D3 was investigated by 3 methods. First, DNA strand breaks were detected by using an in situ cell death assay, terminal deoxynucleotidyl transferase–mediated dUTP nick end-labeling (TUNEL) (Figures 6A⇓ through 6C). Limited regression of sprouting cells was observed in control cultures. Nuclei of sprouting endothelial cells were generally negative for DNA fragmentation, as indicated by low TUNEL staining (Figure 6A⇓). The addition of 1,25(OH)2D3 (1×10−7 mol/L) to VEGF-stimulated cultures resulted in a high number of nuclei that were positive for DNA fragmentation. Stained nuclei were specifically associated with the population of sprouting endothelial cells; no staining was observed in the cobblestone monolayer (Figures 6B and 6C). Second, examination of the integrity of the DNA isolated from cells cultured in the presence of 1,25(OH)2D3 demonstrated the presence of a DNA ladder indicative of internucleosomal cleavage to 180- to 200-bp integers in these samples (Figure 6D⇓). Third, the nuclear morphology of cells incubated in the presence or absence of 1,25(OH)2D3 was assessed (Figures 6E⇓ and 6F⇓). Homogeneous populations of sprouting cells were obtained by plating the cells within collagen gels. Cells incubated in the presence or absence of 1,25(OH)2D3 were collected13 and stained with acridine orange. The number of apoptotic cells, as assessed by the presence of chromatin condensation and fragmentation of nuclei (Figure 6F⇓), was determined, and the results were expressed as a percentage of the total number of cells present. Cells incubated with 1,25(OH)2D3 had significantly more apoptotic nuclei (12.0%) than did control cells (3.75%) (z=4.34, P<0.0001).
1,25(OH)2D3 Inhibits Angiogenesis In Vivo
The effect of 1,25(OH)2D3 on the formation of microvessels in vivo was examined by using a model in which MCF-7 breast carcinoma cells that had been induced to overexpress VEGF121 (VEGF transfectants) were xenografted subcutaneously together with MDA-435S breast carcinoma cells into nude mice.16
Tumors were formed in all of the mice injected with VEGF transfectants. Immunohistochemistry using a rat anti-CD31 antibody was used to assess the vascularity of the tumors produced in these mice. Clusters of small capillaries were present in all tumors (Figure 7⇓). However, treatment with 1,25(OH)2D3 produced tumors that appeared less vascularized than tumors formed in mice treated with vehicle alone (P=0.079, Figure 8⇓) (compare Figures 7A⇓ and 7B⇓ with Figures 7C⇓ and 7D⇓). In addition, large capillaries were evident in tumors formed in control animals (Figures 7E⇓ and 7F⇓); these were never observed in tumors formed after treatment with 1,25(OH)2D3. Tumor volume was not significantly altered, and the proportion of MCF7 and MDA-435S cells present in the tumors was not different in mice treated with 1,25(OH)2D3 compared with those treated with vehicle alone (assessed as described,16 results not shown).
We have demonstrated that 1,25(OH)2D3 has antiangiogenic properties in vivo and in vitro that are mediated by a direct effect on “activated” endothelial cells. First, using in vitro model systems, we have demonstrated that 1,25(OH)2D3 (1×10−9 to 1×10−7 mol/L) markedly inhibits endothelial cell sprouting and morphogenesis and has a small, but significant, inhibitory effect on VEGF-induced endothelial cell proliferation. Second, we have shown that 1,25(OH)2D3 induces the regression of existing endothelial cell sprouts and elongated cells and have established the involvement of apoptosis in this process. Finally, we have demonstrated that 1,25(OH)2D3 (at a dose of 12.5 pmol/d) appears to inhibit the formation of new blood vessels in tumor xenografts induced to overexpress VEGF. This dose, which does not induce hypercalcemia in mice, has previously been shown to stimulate calcium transport in weanling rats and is of the same order of magnitude, on a weight basis, as the therapeutic replacement dose in humans.
The majority of studies on the effect of 1,25(OH)2D3 on cell growth and differentiation report an inhibition of proliferation coupled with an induction of differentiation, with responses occurring in the concentration range 1×10−9 to 1×10−7 mol/L.19 20 However, the response of endothelial cells to 1,25(OH)2D3 has produced conflicting results. One study concluded that 1,25(OH)2D3 inhibits serum-induced endothelial cell proliferation.8 In contrast, other studies found that 1,25(OH)2D3 had no effect on serum-induced endothelial cell growth,9 10 consistent with our data. We have now extended these studies and demonstrate that 1,25(OH)2D3 causes a small, but significant, inhibition of endothelial cell proliferation induced by VEGF and that this inhibition is dose dependent.
The most marked effect on endothelial cells was on the induction (and maintenance) of the elongated sprouting phenotype in these cells and on the subsequent association of these elongated cells into networks. Endothelial cell sprouting and morphogenesis is fundamental to the angiogenic process.3 Endothelial cells contained within developing vessel buds form vascular sprouts in vivo, undergoing morphogenic alterations involving cellular extension and the formation of elongated projections. We demonstrate that the simultaneous addition of 1,25(OH)2D3 and VEGF to confluent endothelial cells results in a dose-dependent inhibition of the sprouting induced when VEGF is added alone. No disruption of the cobblestone monolayer of endothelial cells was observed when 1,25(OH)2D3 was added to cultured cells. To investigate the effect of 1,25(OH)2D3 on the elongation of endothelial cells and their association into networks in the absence of any effect on proliferation, cells were plated within collagen gels.13 Using this model system, we have confirmed that both sprouting and elongation of the cells and their association into networks are inhibited by 1,25(OH)2D3.
The inhibition of endothelial cell spouting and morphogenesis by 1,25(OH)2D3 is consistent with the antiangiogenic effect of 1,25(OH)2D3 reported with use of the chick chorioallantoic membrane.12 The use of in vitro model systems in the present study has extended this latter study to establish a direct effect of 1,25(OH)2D3 on endothelial cells. Lansink et al11 have recently demonstrated that high concentrations of 1,25(OH)2D3 (1×10−5 to 1×10−6 mol/L) have little effect on the formation of capillary tubelike structures by human dermal microvascular endothelial cells cultured on fibrin matrices in the presence of basic fibroblast growth factor and tumor necrosis factor-α. These seemingly divergent results may be explained by differences in the nature of the model system (fibrin versus collagen), the stimulatory factors used (VEGF versus basic fibroblast growth factor and tumor necrosis factor-α), and the concentrations of 1,25(OH)2D3 tested. However, it is noteworthy that 1,25(OH)2D3 (1×10−9 to 1×10−7 mol/L) inhibits the formation of elongated sprouting cells in BAEC cultures stimulated with hepatocyte growth factor (D.J. Mantell, A.E. Canfield, unpublished data, 2000).
The inhibitory effects of 1,25(OH)2D3 on growth factor–stimulated proliferation and morphogenesis can be explained through a number of mechanisms. First, growth factors can modulate the expression of the nuclear vitamin D receptor.21 Second, 1,25(OH)2D3 may regulate phospholipase C production by the cells,22 which, in turn, may modulate signal transduction by receptors with tyrosine kinase activity, including VEGF-R1 and VEGF-R2. Finally, 1,25(OH)2D3 can modulate the expression of growth factor receptors.23 However, it has been shown that 1,25(OH)2D3 has no effect on the expression of VEGF receptor genes by endothelial cells,10 suggesting that the inhibition of VEGF-induced endothelial cell proliferation observed in the present study is not due to a downregulation of VEGF receptors by 1,25(OH)2D3.
Importantly, we demonstrate that 1,25(OH)2D3 also induces the regression of elongated sprouting cells in vitro by inducing apoptosis specifically within this cell population. 1,25(OH)2D3 induces apoptosis of cancer cell lines in vitro24 and in vivo.25 Previous studies have shown that 1,25(OH)2D3 upregulates clusterin and cathepsin B.24 In addition, 1,25(OH)2D3 downregulates the antiapoptotic bcl-2 protein while upregulating p53 expression, resulting in active cell death.26 VEGF prevents apoptosis in cultured endothelial cells by inducing the expression of the antiapoptotic proteins Bcl-2 and A1.27 Therefore, it will be of particular interest to determine the effect of 1,25(OH)2D3 on the expression of antiapoptotic and proapoptotic proteins by endothelial cells.
The induction of apoptosis of cancer cells in vivo has been suggested as the mechanism behind the regression of tumors observed when analogues of 1,25(OH)2D3 are administered.25 However, the demonstration in the present study, which just falls short of statistical significance, that low doses of 1,25(OH)2D3 appear to inhibit the formation of new blood vessels in an in vivo model of VEGF-induced tumor angiogenesis (P=0.079) suggests that regression of the vascular supply by the induction of apoptosis in growth factor–stimulated endothelial cells may be another mechanism by which 1,25(OH)2D3 causes tumor regression. Furthermore, large capillaries were never observed in tumors formed after treatment with 1,25(OH)2D3, suggesting that 1,25(OH)2D3 may also inhibit vessel growth and maturation. The fact that we did not detect any differences in tumor volumes may reflect either the low dose of 1,25(OH)2D3, the duration of the experiments, and/or the number of animals used. The use of high doses of 1,25(OH)2D3 was avoided because of the risk of inducing hypercalcemia in the mice. But the finding that even this low dose of 1,25(OH)2D3 (which is within the normal range) inhibited VEGF-induced angiogenesis in vivo raises the possibility that 1,25(OH)2D3 analogues with less calcemic effects could be used even more successfully in this model. Furthermore, these data are consistent with our clinical study showing that serum 1,25(OH)2D3 levels are inversely related to disease progression in patients with metastatic breast cancer.28
In summary, we demonstrate that 1,25(OH)2D3 inhibits specific stages of the angiogenic process. The direct effects of 1,25(OH)2D3 on endothelial cell proliferation and morphogenesis observed with the use of these in vitro model systems provide evidence consistent with the antiangiogenic effects of 1,25(OH)2D3 shown in vivo. Furthermore, the induction of apoptosis specifically within the activated, angiogenic, endothelial cell population suggests that this may be the mechanism behind the observed antiangiogenic effects of 1,25(OH)2D3. These results suggest that 1,25(OH)2D3 (or analogues of this hormone) may be of use in the prevention of conditions involving pathological angiogenesis and may also be of use in the therapeutic regression of such conditions characterized by aberrant angiogenesis.
The financial support of the University of Manchester Bequest Fund, Pearl Assurance, and Association for International Cancer Research are gratefully acknowledged. We also thank Adele Poole for excellent technical assistance.
Reprint requests to Dr A.E. Canfield, Wellcome Trust Centre for Cell Matrix Research, Department of Medicine, University of Manchester, 2.205 Stopford Building, Oxford Road, Manchester M13 9PT, UK.
- Received April 28, 2000.
- Accepted June 8, 2000.
- © 2000 American Heart Association, Inc.
Pepper MS. Manipulating angiogenesis: from basic science to the bedside. Arterioscler Thromb Vasc Biol. 1997;17:605–619.
Neufeld G, Cohen T, Genrinovitch S, Poltorak Z. Vascular endothelial growth factor (VEGF) and its receptors. FASEB J. 1999;13:9–22.
Gerber HP, Hillan KJ, Ryan AM, Kolawski J, Keller GA, Rangell L, Wright BD, Radtke F, Aguet M, Ferarra N. VEGF is required for growth and survival in neonatal mice. Development. 1999;126:1149–1159.
Merke J, Milde P, Lewicka S, Hugel U, Klaus G, Mangelsdorf DJ, Haussler MR, Rauterberg EW, Ritz EJ. Identification and regulation of 1,25-dihydroxyvitamin D3 receptor activity and biosynthesis of 1,25-dihydroxyvitamin D3: studies in cultured bovine aortic endothelial cells and human dermal capillaries. J Clin Invest. 1989;83:1903–1915.
Lansink M, Koolwijk P, Hinsbergh V, Kooistra T. Effect of steroid hormones and retinoids on the formation of capillary-like tubular structures of human microvascular endothelial cells in fibrin matrices is related to urokinase expression. Blood. 1998;92:927–938.
Schor AM, Schor SL, Allen TD. Effects of culture conditions on the proliferation, morphology and migration of bovine aortic endothelial cells. J Cell Sci. 1983;62:267–285.
Zhang H-T, Craft P, Scott PAE, Ziche M, Weich HA, Harris AL, Bicknell R. Enhancement of tumour growth and vascular density by transfection of vascular endothelial growth factor into MCF-7 human breast carcinoma cells. J Natl Cancer Inst. 1995;87:213–219.
Pullan S, Wilson J, Metcalfe A, Edwards GM, Goberdhan N, Tilly J, Hickman JA, Dive C, Streuli CH. Requirement of basement membrane for the suppression of programmed cell death in mammary epithelium. J Cell Sci. 1996;109:631–642.
Stern PH. 1,25 Dihydroxyvitamin D3 interactions with local factors in bone remodelling. In: Feldman D, Glorieux FH, Pike JW, eds. Vitamin D. San Diego, Calif: Academic Press; 1997:341–352.
Koga M, Eisman JA, Sutherland RL. Regulation of epidermal growth factor receptor levels by 1,25-dihydroxyvitamin D3 in human breast cancer cells. Cancer Res. 1988;48:2734–2739.
Gerber H-P, Dixit V, Ferrara N. Vascular endothelial growth factor induces expression of the antiapoptotic proteins Bcl-2 and A1 in vascular endothelial cells. J Biol Chem. 1998;273:13313–13316.