Abstract
Although tumors can activate vascular endothelial growth factor (VEGF)promoter in host stromal cells, the relative contribution to VEGF production of host versus tumor cells and the resulting vascular response have not been quantitated to date. To this end, we implanted VEGF−/− and wild-type (WT) embryonic stem (ES)cells in transparent dorsal skin windows in severe combined immunodeficient mice. VEGF−/− ES cell-derived tumors produced ∼50% of VEGF compared with the WT tumors, suggesting significant contribution of host stromal cells. To discern the hypoxia-induced hypoxia inducible factor (HIF)-1α → hypoxia response element (HRE) → VEGF signaling cascade, we also examined tumors derived from HIF-1α−/− and HRE−/−ES cells. As expected, the VEGF protein level in HIF-1α−/− ES tumors was intermediate between VEGF−/− and WT ES cell tumors. Surprisingly,HRE−/− ES tumors produced the same level of VEGF as the VEGF−/− ES tumors, suggesting a critical role of HRE in tumor cell VEGF production. Angiogenesis in these tumors was proportional to their VEGF levels (VEGF−/− ≈HRE−/− < HIF-1α−/− < WT). In contrast, vascular permeability, leukocyte-endothelial adhesion, and tumor growth were reduced in VEGF−/− and HRE−/− tumors but were comparable in HIF-1α−/− and WT tumors. This discrepancy suggests that different intracellular signaling pathways may be involved in each of these functions of VEGF. More importantly,these data suggest that host cells are active players in tumor angiogenesis and growth and need to be taken into account in the design of any therapeutic strategy.
Introduction
VEGF,4a multifunctional cytokine, was originally discovered as a tumor-secreted protein that promotes vascular permeability(1). Subsequently, it was reported to promote angiogenesis by inducing migration and proliferation of endothelial cells (2, 3). The effect of VEGF on leukocyte adhesion to the endothelium was discovered recently (4). VEGF production is up-regulated by hypoxia, acidosis, and hypoglycemia (2). Hypoxia induces HIF-1α expression, which then mediates a series of transcriptional responses. The cognate DNA recognition site of HIF-1αis HRE (2, 5). HIF-1α binds to HRE of target genes such as VEGF, erythropoietin, and glycolytic enzymes. The binding of HIF-1α to HRE in the VEGF promoter is a predominant enhancer of VEGF production (5). VEGF protein binds to VEGF receptors on endothelial cells, and these mediate its physiological functions. We have shown recently that VEGF can be produced by both cancer cells and host-stromal cells in a tumor(6). However, the relative contribution of these different cell types to tumor-VEGF and the resulting vascular response have not been quantitated to date.
A variety of genetic and pharmacological strategies have been used to test the causal relationships between VEGF and its putative physiological functions. These include overexpression of VEGF in host cells (7) or neoplastic cells (8); targeted deletion of VEGF gene (5, 9, 10, 11);introduction of sense or antisense VEGF gene constructs into neoplastic cells (12, 13); superfusion with recombinant VEGF (14); incorporation of VEGF in a gel(15); treatment with anti-VEGF antibody (16);and modulation of VEGF expression by oncogene, hormone(17), or other microenvironmental factors (2, 3). All of these approaches have collectively provided powerful insight into the role of VEGF in angiogenesis, vascular permeability,leukocyte rolling and adhesion, and tumor growth. However, none of these studies have simultaneously measured all of these parameters in the same tissue at different levels of host versus tumor VEGF. Therefore, the relative dependence of these functions on VEGF levels is unknown. An answer to this important question requires tumors that have controlled VEGF levels and the techniques to simultaneously monitor and quantify all of these parameters in vivo.
In this study, VEGF levels in ES cells were differentially modulated by targeted deletion of three members of the hypoxia → VEGF cascade: HIF-1α, HRE of the VEGF, and VEGF. VEGF expression in the resulting ES cells in culture varied as follows: WT (HIF-1α+/+ ≈HRE+/+ ≈VEGF+/+) > HIF-1α−/− ≈HRE−/− > VEGF−/− = 0. Using tumors derived from these ES cells expressing decreasing levels of VEGF and using intravital microscopy to measure vascular parameters noninvasively and continuously, we show that different VEGF functions such as angiogenesis, vascular permeability, and L/E interaction have distinct VEGF dose dependencies. In addition, VEGF levels in the resulting tumors showed that host cells in VEGF−/− tumors produced significant amount of VEGF (50% of WT tumor), and that HRE,but not HIF-1α deletion, abrogates VEGF production by ES cells in vivo.
Materials and Methods
ES Cell Lines.
VEGF−/−, VEGF-HRE−/−(referred to as HRE−/−), and HIF-1α−/− ES cells were generated by targeted deletion as described previously (10, 18) 5. ES clones with a randomly integrated gene targeting vector, which survived high G418 selection, were used as WT controls(VEGF+/+, HRE+/+, and HIF-1α+/+, respectively).
Animal and Tumor Models.
Mutant or WT cells (2 × 105cells) were implanted into dorsal skinfold chambers in SCID mice (6–8 weeks of age; 25–35 g) as described previously (18, 19). Chambers were checked for tumor growth every other day. Twelve to 14 days after cell implantation, macroscopically visible tumors were observed in the chambers. Measurements of angiogenesis, hemodynamics,vascular permeability, and L/E interaction were performed 16–26 days after implantation in size-matched tumors.
Measurement of Tumor Growth.
The tumor-bearing chambers of unanesthetized mice were placed in a polycarbonate tube and observed by intravital microscopy (Axioplan;Zeiss, Oberkochen, Germany; Ref. 19). After capturing an image of the tumor, its surface area was calculated as π × a × b (a and b are major and minor axes). The thickness of the tumors, c, was measured by a caliper, and the volume was calculated as π/6 × a × b × c. These measurements were made on days 7, 14, 21, and 28 in at least five animals/group.
Measurements of Angiogenesis and Hemodynamics.
Animals bearing tumors in the dorsal chambers were anesthetized by s.c. injection of a mixture of 75 mg of ketamine hydrochloride (Parke-Davis,Morris Plains, NJ) and 25 mg of xylazine (Fermetia, Kansas City, MO)per kg body weight and observed by intravital microscopy(19). For measurement of RBC velocity and vessel diameters, 10 mg/ml FITC-labeled dextran solution(Mr 2,000,000; Sigma Chemical Co., St. Louis, MO) was injected i.v. via a tail vein cannula to illuminate blood vessels. Epi-illumination was performed using a 100-W mercury lamp equipped with a fluorescence filter for FITC (excitation, 525–555 nm; emission, 580–635 nm). An intensified charge-coupled device video camera (C2400–88; Hamamatsu Photonics K.K., Hamamatsu, Japan) was used to visualize microvessels in five random areas of each tumor. Functional vascular density (an index of angiogenesis) was measured as the total length of perfused vessels per unit area of observation field(19). RBC velocity was measured by the four-slit method(Microflow system, model 208C, videophotometer version; IPM, San Diego,CA; Ref. 19). Vessel diameter was measured by an image-shearing device (digital video image shearing monitor, model 908;IPM; Ref. 19). Vascular volume density of perfused vessels per unit area was calculated from vascular density and vessel diameter. Mean blood flow rates and shear rates of individual vessels were calculated using vessel diameter and mean RBC velocity as described previously (19).
Microvascular Permeability Measurement.
Mice were injected with a bolus (100 μl) of 1%tetramethylrhodamine-labeled BSA (Molecular Probes, Eugene, OR) in saline via the tail vein. Fluorescence intensity of the tumor tissue was measured every 2 min for a total of 20 min by a photomultiplier in a well-perfused area using a ×20 objective lens. The microvascular permeability to albumin was then calculated as described previously(16).
L/E Interaction Measurements.
Leukocyte rolling and adhesion in tumor vessels were measured as described previously (20). Briefly, mice were injected with a bolus (20 μl) of 0.1% rhodamine 6-G (Molecular Probes) in saline via the tail vein to facilitate visualization of leukocytes. One hundred μm of nonbranching segments with diameters ranging between 13 and 30 μm were selected for the measurements. The number of rolling(Nr) and adhering (Na) leukocytes, and total leukocyte flux (Nt) in 30 s was measured. Leukocyte rolling (%) count was calculated as 100 × Nr/Nt. The density of adhering leukocytes(cells/mm2) was calculated as 106 × Na/(π × D × 100 μm), where D is the diameter of a given vessel.
ELISA for VEGF in ES Cell Culture and Tumor Tissue.
ES cells were cultured in DMEM supplemented with 10% FCS (Progen Industries Ltd., Brisbane, Australia) and 500 units/ml leukemia inhibitory factor (ESGRO; Life Technologies, Inc., Rockville, MD). The supernatant of ES cell culture medium exposed to both normoxia and hypoxia (20 and 1% oxygen, respectively) for 24 h was collected. Measurements were made on samples collected from at least three different dishes. To determine tissue VEGF levels, tumors were excised from the dorsal skinfold chambers and frozen in liquid nitrogen. The tissues were homogenized, and the protein was extracted with 50 mm Tris-HCl buffer (1 ml, pH 7.4) containing 0.25% Triton X-100, 0.5 m EDTA, 0.1% NP40, and protease inhibitor mixture (Complete protease inhibitors; Roche Biochemicals,Indianapolis, IN). The total protein level in each group was determined by a standard protein assay. The tissue extracts were diluted with 50 mm Tris-HCl to yield samples with the same protein concentration. Fifty μl of each sample were used to determine VEGF levels. Immunoreactive VEGF was quantified using a sandwich ELISA(Quantikine M Mouse VEGF Immunoassay kit; R & D System, Minneapolis,MN) according to the manufacturer’s recommended protocol.
Statistics.
Results are presented as mean ± SE. Values of different genotypes of ES cell-derived tumors were compared using a Mann-Whitney U test (StatView; Abacus, Berkeley, CA). Significance was assumed at P < 0.05.
Results and Discussion
ES Cell-derived Tumors Allowed Us to Study the Function of Embryonic Lethal Genes.
Mutant mice with VEGF,HRE6, or HIF-1α gene knocked out are embryonic lethal on day E8.5 to E9, and thus, it is difficult to study angiogenesis and physiological functions in adult tissue. On the other hand, ES cells have the capacity to form teratomas when implanted into immunodeficient mice (Fig. 1). These ES cells were used to determine the effect of complete and partial deletion of VEGF expression on vascular response and solid tumor formation in vivo.
Host Stromal Cells Produce a Significant Amount of VEGF.
Recently, using VEGF promoter-driven GFP transgenic mice we found that host fibroblasts migrate into murine tumors grown in dorsal skin chambers and express strong VEGF promoter activity(6). In addition, endothelial cells of angiogenic vasculature and infiltrating lymphocytes in tumors are reported to be a source of VEGF (21, 22). The use of additional sequences in VEGF promoter construct may detect other type of cells that produce VEGF in tumors (23, 24). Therefore, the first goal of this study was to measure the relative contribution of host stromal versus tumor cells to VEGF production. Indeed, we found that VEGF levels in VEGF−/− ES cell-derived tumors are ∼200 pg/mg protein (Table 1) in contrast to undetectable7 levels in the supernatant of VEGF−/− ES cell cultures. Host stromal cells in VEGF−/− ES tumors produce VEGF protein ∼50%of that in WT tumors. Although we recognize the possibility that the anti-VEGF antibody used in ELISA8 detects nonspecific proteins in the tissue and the contribution of host cells may vary in different tumors, these findings support the importance of the production of angiogenic growth factors by host cells in tumor angiogenesis.
VEGF protein levels were undetectable7in conditioned media from VEGF−/− cells and reduced in conditioned medium from HIF-1α−/−cells (normoxia, 67 ± 16; hypoxia, 42 ± 10 pg/ml) and HRE−/− cells (normoxia,38 ± 14; hypoxia, 43 ± 15 pg/ml) as compared with WT cells (normoxia, 203 ± 32; hypoxia,312 ± 32 pg/ml), both under normoxia and hypoxia,respectively. In fact, hypoxia-induced VEGF up-regulation was completely abolished in HIF-1α−/− and HRE−/− ES cells in culture. Although a similar pattern was observed in ES cell-derived teratomas, VEGF protein levels in HRE−/− tumors are comparable with VEGF−/− tumors (Table 1). These data suggest that hypoxia-driven signaling of the HIF-1α−/−-HRE pathway plays a predominant role in the induction of VEGF in these tumor tissues. Although ES cells were incapable of inducing VEGF under hypoxia in vitro in the absence of HIF-1α, other members of HIF families, such as HIF-2/EPAS, also bind to HRE and thus can mediate hypoxia-induced gene expression in vivo(5). Our data suggest redundancy in hypoxia-inducible transcriptional factors but not in the responsive element in the VEGF promoter in vivo.
Angiogenesis Is Proportional to Tissue Levels of VEGF.
We investigated the dose dependency of angiogenesis on VEGF using tumors expressing different levels of VEGF. Vascular density and diameter were analyzed from digitized FITC fluorescence images of angiogenic vessels (Fig. 1). Loss of HIF-1α impaired angiogenesis, as exhibited by reduced tumor vascular density and diameter compared with WT tumors (Figs. 1 and 2; Table 1). VEGF−/− and HRE−/− tumors exhibited a further decrease in vascular density. Collectively, angiogenesis followed a trend similar to VEGF levels in vivo such that VEGF−/− ≈ HRE−/− < HIF-1α−/− < WT. Our current data, coupled with the existing data from the literature(1, 3, 8, 12, 13, 16), indicate that tumor angiogenesis increases in a dose-dependent manner in response to endogenous tumor VEGF production.
Blood Perfusion Is Proportional to VEGF Levels.
We also determined blood flow rates in individual vessels as a function of RBC velocity and vessel diameters (20). Blood flow in tumors was disorganized and often sluggish, and the vascular structure was heterogeneous. As a result, there was no correlation between RBC velocity and vessel diameter in tumors. Furthermore, loss of HIF-1α,HRE, or VEGF did not significantly change the mean centerline RBC velocity in the perfused vessels (Fig. 2; Table 1). Reduced vessel diameters and equivalent RBC velocities in HIF-1α−/−, HRE−/−,and VEGF−/− tumors compared with WT tumors resulted in decreased blood flow rates and increased shear rates. Overall tissue perfusion was further reduced in HIF-1α−/−, HRE−/−,and VEGF−/− tumors relative to WT tumors because of the reduction in blood flow rates in individual vessels and vessel density in tissue. Thus, tissue perfusion of ES cell-derived tumors varied as follows: WT (HIF-1α+/+ ≈HRE+/+ ≈VEGF+/+) > HIF-1α−/− > HRE−/− ≈ VEGF−/−.
Vascular Permeability Is Not Reduced in HIF-1α−/−Tumors Despite Decreased VEGF Level.
Vascular permeability to albumin was approximately the same in WT and HIF-1α−/− tumors (Fig. 2; Table 1). In contrast, VEGF−/− and HRE−/− tumors exhibited significantly reduced vascular permeability compared with HIF-1α−/−and WT tumors. These findings are consistent with our recent measurement of vascular permeability to albumin in normal and tumor vessels at various doses of exogenous human VEGF (14). Dose-dependent increases in vascular permeability were observed in normal vessels up to 100 ng/ml VEGF, and no further increases were observed between concentrations of 100 and 1000 ng/ml VEGF. On the other hand, vascular permeability of LS174T human colon carcinoma xenografts was significantly higher than permeability in normal tissue,and exogenous VEGF (10–1000 ng/ml) did not alter the tumor vascular permeability. We also found that inhibition of endogenous VEGF in the same tumor resulted in a significant decrease in vascular permeability(16). Taken together, these findings suggest that vascular permeability reaches a plateau at a certain VEGF level in both normal tissue and tumor tissue.
L/E Interaction Is Reduced Only in the Tumors with Lowest VEGF Level.
We reported that VEGF in tumor interstitial fluid increases intercellular adhesion molecule-1, vascular cell adhesion molecule-1,and E-selectin expression in human umbilical vein endothelial cells(4). Furthermore, transgenic mice overexpressing VEGF under the control of the K14 promoter showed increased L/E interaction and angiogenesis in vivo (7). In the present study, we found that L/E interaction was approximately the same in WT and HIF-1α−/− tumors. In contrast,VEGF−/− and HRE−/−tumors exhibited significantly reduced L/E interactions compared with HIF-1α−/− and WT tumors (Fig. 2). These findings suggest that L/E interaction reaches a plateau at a certain tissue VEGF level similar to vascular permeability and support the link between angiogenesis and L/E interaction (25).
Growth Rate of HIF-1α−/− Tumors Is Comparable with WT Tumors Despite Reduced Angiogenesis.
One week after implantation of these ES cells, a red spot was observed in chambers in all genotypes, suggesting a host inflammatory response. At 12–14 days after implantation, ES cell implants developed into tumors with different growth rates, depending on the genotype (Fig. 3). HIF-1α−/− tumors grew at the same rate as WT tumors for 4 weeks after cell implantation, the end point of this study. Both WT and HIF-1α−/− tumors reached a size of ∼75 mm2 in area and 120 mm3 in volume between 21 and 28 days, at which point they filled the entire dorsal skin windows. This is consistent with our previous finding that HIF-1α mediates apoptosis by inducing p53 (18). In contrast, both HRE−/−and VEGF−/− tumors grew at a significantly slower rate than WT and HIF-1α−/− tumors in both area and volume. The growth rate of HRE−/−tumors analyzed by volume was between WT and HIF-1α−/− tumors and VEGF−/− tumors during the observation period(Fig. 3). Collectively, these data suggest that prediction of tumor growth rate on the basis of VEGF levels or angiogenesis alone is simplistic and must take into account other molecules/parameters that govern proliferation and apoptosis of cells.
Limitations and Conclusions.
Although the ES tumor system provides powerful insight into the functions of the gene of interest in vivo (9, 18), there are some limitations: (a) ES cell-derived teratomas may be different from malignant solid tumors (the latter usually have multiple gene mutations); (b) targeting a specific gene in ES cells does not preclude associated recombination events, and thus, cells with a specific gene mutation may alter some properties unrelated to the gene of interest. In fact, different HIF-1α mutation studies showed somewhat different responses (18, 26, 27); and (c)post-transcriptional regulation is as important as transcriptional regulation in many genes such as VEGF for mRNA stability. Targeted gene deletion strategy may not address posttranscriptional regulation.
Despite these limitations, our study suggests that stepwise manipulation of the HIF-1α → HRE → VEGF signaling cascade differentially impacts angiogenesis, vascular permeability, and L/E interaction. Different signaling pathways may be involved in these different functions of VEGF. The binding of HIF-1α to HRE in the VEGF promoter is a major pathway that leads to the induction of VEGF, contributing to its physiological function in tumors. The deletion of the VEGF gene in tumor cells did not completely block tumor VEGF production, angiogenesis, or tumor growth. Thus, host-derived VEGF may be enough to maintain angiogenesis and tumor growth albeit at slower rate. This finding has significant implications for various antiangiogenic therapies for solid tumors(28).
Angiogenic vessels in ES cell-derived tumors. Left column, low power images of various ES cell-derived tumors were visualized by intravital microscopy using transillumination and a green filter. Significant reduction in vasculature was seen in HRE−/− and VEGF−/− ES tumors. Hemorrhagic spots were observed in HIF-1α−/−, HRE−/−,and VEGF−/− ES cellderived tumors. Right column, high power images of ES cell-derived tumors were visualized by fluorescence microscopy using FITC-dextran as vascular contrast agent. A significant reduction in vessel density and diameter was observed in HIF-1α−/−, HRE−/−, and VEGF−/− ES tumors compared with the WT ES tumor.
Angiogenic vessels in ES cell-derived tumors. Left column, low power images of various ES cell-derived tumors were visualized by intravital microscopy using transillumination and a green filter. Significant reduction in vasculature was seen in HRE−/− and VEGF−/− ES tumors. Hemorrhagic spots were observed in HIF-1α−/−, HRE−/−,and VEGF−/− ES cellderived tumors. Right column, high power images of ES cell-derived tumors were visualized by fluorescence microscopy using FITC-dextran as vascular contrast agent. A significant reduction in vessel density and diameter was observed in HIF-1α−/−, HRE−/−, and VEGF−/− ES tumors compared with the WT ES tumor.
Vascular functions in ES cell-derived tumors. Physiological parameters in ES tumors were determined by intravital microscopy with the use of appropriate fluorescence tracers. Data are shown as a relative value (%) to the corresponding WT ES tumors. A, functional vascular density. B, vessel diameter. C, mean blood flow rate of individual vessels. D, vascular permeability. E, leukocyte rolling. F, leukocyte adhesion. ∗, P < 0.05 as compared with corresponding WT tumors. #, P < 0.05 as compared with HIF-1α−/− tumors. †, P < 0.05 as compared with HRE−/− tumors. Bars, SE.
Vascular functions in ES cell-derived tumors. Physiological parameters in ES tumors were determined by intravital microscopy with the use of appropriate fluorescence tracers. Data are shown as a relative value (%) to the corresponding WT ES tumors. A, functional vascular density. B, vessel diameter. C, mean blood flow rate of individual vessels. D, vascular permeability. E, leukocyte rolling. F, leukocyte adhesion. ∗, P < 0.05 as compared with corresponding WT tumors. #, P < 0.05 as compared with HIF-1α−/− tumors. †, P < 0.05 as compared with HRE−/− tumors. Bars, SE.
Tumor growth of ES cell-derived tumors. A,tumor surface area animals was measured on days 7, 14, 21, and 28. B, tumor volume was calculated from tumor surface area and thickness on days 7, 14, 21, and 28. WT (n = 6), HIF-1α−/− (n = 6), HRE−/− (n = 5), and VEGF−/− (n = 8). ∗, P < 0.05 as compared with corresponding WT tumors. #, P < 0.05 as compared with HIF-1α−/− tumors. †, P < 0.05 as compared with HRE−/− tumors. Bars, SE.
Tumor growth of ES cell-derived tumors. A,tumor surface area animals was measured on days 7, 14, 21, and 28. B, tumor volume was calculated from tumor surface area and thickness on days 7, 14, 21, and 28. WT (n = 6), HIF-1α−/− (n = 6), HRE−/− (n = 5), and VEGF−/− (n = 8). ∗, P < 0.05 as compared with corresponding WT tumors. #, P < 0.05 as compared with HIF-1α−/− tumors. †, P < 0.05 as compared with HRE−/− tumors. Bars, SE.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Supported by Outstanding Investigator Grant R35-CA56591 from National Cancer Institute (to R. K. J.) and a grant from the Alexander and Margaret Stewart Trust (to D. F.). This work was presented at 91st Annual Meeting of the American Association for Cancer Research, April 1–5, 2000, San Francisco.
The abbreviations used are: VEGF, vascular endothelial growth factor; ES, embryonic stem; L/E,leukocyte-endothelial; HIF, hypoxia-inducible factor; HRE, hypoxia response element; WT, wild type; SCID, severe combined immunodeficient.
B. Oosthuyse, L. Moons, H. Beck, J. V. Dorpe, P. Hellings, M. Gorselink, D. Nuyens, S. Heymans, G. Theilmeier, M. Dewerchin, V. Laudenbach, P. Vermylen, T. Acker, A. Damert, N. Cashman,H. Fujisawa, M. R. Drost, W. Robberecht, R. Sciot, F. Bruyninckx, P. Gressens, K. H. Plate, F. Lupu, J. M. Herbert, D. Collen, and P. Carmeliet. Targeted deletion of the hypoxia response element in the VEGF promoter causes adult motor neuron degeneration, submitted for publication.
Thirty to 50% of HRE−/− mice survive for more than a year after birth and develop motor neuron defects (see footnote 5).
Detection limit of VEGF ELISA was 7.8 pg/ml.
Murine VEGF ELISA kit from R&D does not cross-react with murine placental growth factor.
VEGF levels and physiological parameters in ES cell tumorsa
Tumor type . | VEGF level in tissue (pg/mg protein) . | Angiogenesis . | . | Hemodynamics . | . | . | Permeability (10 cm/s) . | L/E interaction . | . | . | |||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | Vascular density (cm/cm2) . | Diameter (μm) . | RBC velocity (μm/s) . | Blood flow (pl/s) . | Shear rate (1/s) . | . | Rolling (%) . | Total influx (cells/30 s) . | Adhesion (cells/mm2) . | |||||
WT(HIF-1α) | 486 ± 57 | 150 ± 35 | 22 ± 2 | 146 ± 5 | 40 ± 6 | 39 ± 6 | 2.6 ± 0.7 | 37 ± 5 | 17 ± 3 | 150 ± 19 | |||||
HIF-1α | 292 ± 22b | 112 ± 12b | 16 ± 1b | 140 ± 5 | 24 ± 3b | 46 ± 5 | 2.7 ± 0.8 | 37 ± 5 | 16 ± 1 | 166 ± 20 | |||||
WT(HRE) | 396 ± 76 | 147 ± 5.5 | 23 ± 2 | 155 ± 12 | 52 ± 8 | 42 ± 11 | 3.1 ± 0.7 | 35 ± 2 | 17 ± 2 | 167 ± 14 | |||||
HRE | 183 ± 36b,c | 71 ± 15b,c | 15 ± 2b | 143 ± 10 | 19 ± 4b | 59 ± 13 | 1.8 ± 0.5b | 26 ± 3b | 14 ± 1b | 87 ± 8b | |||||
WT(VEGF) | 406 ± 32 | 157 ± 34 | 23 ± 1 | 141 ± 6 | 41 ± 5 | 35 ± 2 | 2.7 ± 0.5 | 37 ± 6 | 19 ± 2 | 148 ± 15 | |||||
VEGF | 199 ± 19b,c | 63 ± 16b,c | 15 ± 2b | 143 ± 10 | 17 ± 2b | 59 ± 13b | 1.3 ± 0.1b | 18 ± 2b,c | 12 ± 3b | 64 ± 10b,c |
Tumor type . | VEGF level in tissue (pg/mg protein) . | Angiogenesis . | . | Hemodynamics . | . | . | Permeability (10 cm/s) . | L/E interaction . | . | . | |||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | Vascular density (cm/cm2) . | Diameter (μm) . | RBC velocity (μm/s) . | Blood flow (pl/s) . | Shear rate (1/s) . | . | Rolling (%) . | Total influx (cells/30 s) . | Adhesion (cells/mm2) . | |||||
WT(HIF-1α) | 486 ± 57 | 150 ± 35 | 22 ± 2 | 146 ± 5 | 40 ± 6 | 39 ± 6 | 2.6 ± 0.7 | 37 ± 5 | 17 ± 3 | 150 ± 19 | |||||
HIF-1α | 292 ± 22b | 112 ± 12b | 16 ± 1b | 140 ± 5 | 24 ± 3b | 46 ± 5 | 2.7 ± 0.8 | 37 ± 5 | 16 ± 1 | 166 ± 20 | |||||
WT(HRE) | 396 ± 76 | 147 ± 5.5 | 23 ± 2 | 155 ± 12 | 52 ± 8 | 42 ± 11 | 3.1 ± 0.7 | 35 ± 2 | 17 ± 2 | 167 ± 14 | |||||
HRE | 183 ± 36b,c | 71 ± 15b,c | 15 ± 2b | 143 ± 10 | 19 ± 4b | 59 ± 13 | 1.8 ± 0.5b | 26 ± 3b | 14 ± 1b | 87 ± 8b | |||||
WT(VEGF) | 406 ± 32 | 157 ± 34 | 23 ± 1 | 141 ± 6 | 41 ± 5 | 35 ± 2 | 2.7 ± 0.5 | 37 ± 6 | 19 ± 2 | 148 ± 15 | |||||
VEGF | 199 ± 19b,c | 63 ± 16b,c | 15 ± 2b | 143 ± 10 | 17 ± 2b | 59 ± 13b | 1.3 ± 0.1b | 18 ± 2b,c | 12 ± 3b | 64 ± 10b,c |
n = 3 each for VEGF ELISA; n = 4–5 each for all physiological parameters.
P < 0.05 as compared with corresponding WT tumors.
P < 0.05 as compared with HIF-1α−/− tumors.
Acknowledgments
We thank Drs. Chae-Ok Yun and Yi Chen for help with molecular analysis, Dr. Emmanuelle di Tommaso for help with histological analysis, Julia Kahn for dorsal chamber preparation, and Drs. Donald G. Buerk, Yves Boucher, Kevin Burton, Carla Mouta Carreira, Leo E. Gerweck and Ananth Kadambi for helpful comments on the manuscript.