- Email: [email protected]
Effects of microenvironmental extracellular pH and extracellular matrix proteins on angiostatin's activity and on intracellular pH
General Pharmacology 35 (2002) 277 – 285
Effects of microenvironmental extracellular pH and extracellular matrix proteins on angiostatin’s activity and on intracellular pH Miriam L. Wahla,*, Derrick S. Grantb a
Department of Biochemistry and Molecular Pharmacology, Thomas Jefferson University, 233 South, 10th Street, Room 226, Philadelphia, PA 19107, USA b Cardeza Foundation, Department of Medicine, Thomas Jefferson University, Philadelphia, PA 19107, USA Received 1 September 2001; accepted 1 October 2001
Abstract Antiangiogenic agents target migratory and proliferative endothelial cells (EC) in the process of forming new vessels, resulting in growth inhibition or cell death. Here we have shown that the antiangiogenic activity of angiostatin on EC is enhanced in culture when the microenvironmental extracellular pH (pHe) is reduced to levels similar to that of many tumors. In a migration/scratch assay and during tube formation, angiostatin in combination with reduced pHe synergistically resulted in an increased EC death—an effect not seen with either stimulus individually. Lowering of pHe decreased intracellular pH (pHi), and a further lowering of pHi occurred when low pHe was combined with angiostatin. These data suggest that low pHe plays a role in the relative specificity and efficacy of angiostatin for tumor neovasculature and indicate roles for both pHe and pHi in the mechanism of angiostatin action. A receptor for angiostatin, the a-subunit of ATP synthase, was found on the surface of EC. We show that cell surface receptor distribution is increased on Matrigel, a basement-like matrix, as opposed to fibronectin or RGD peptide substrates, and redistributed to a more punctuate appearance at low pHe. Furthermore, positive cell surface histochemical staining for a-ATP synthase was blocked by preincubation with angiostatin. These data indicate that substrate and pHe are critical parameters in the evaluation of this antiangiogenic substance, and probably for others as well. D 2002 Elsevier Science Inc. All rights reserved. Keywords: Extracellular acidification; Angiogenesis; Angiostatin; Intracellular pH regulation; Fluorescent indicator dye
1. Introduction Blood vessel formation is essential for most tumor growths (Folkman, 1990; Folkman and Ingber, 1992), and antiangiogenic drugs such as angiostatin have been shown to reduce tumor vascularization by blocking endothelial cells (EC) proliferation and migration (Lucas et al., 1998; O’Reilly et al., 1994; Wu et al., 1997). The exact Abbreviations: angiostatin, rh-angiostatin, human recombinant angiostatin1 (Entremed, Rockville, MD); BCECF-AM, bis carboxyethyl-carboxyl fluorescein-acetoxymethylester; Cl/HCO3, chloride/bicarbonate exchanger; DAPI, 406-diamidino-2-phenylindole; dilactate DPBS, Dulbecco’s phosphate-buffered saline; EBM, endothelial basal medium; EC, endothelial cells; ECGS, endothelial cell growth supplement; pHe, extracellular pH; FAK, focal adhesion kinase; MCT, H+-linked monocarboxylate transporter; HUVEC, human umbilical vein endothelial cells; pHi,, intracellular pH; NHE-1, sodium – proton exchanger isoform 1. * Corresponding author. Tel.: +1-215-503-7867; fax: +1-215-923-9162. E-mail address: [email protected] (M.L. Wahl).
mechanism of action of angiostatin and the basis for its selectivity (within tumor beds) have not yet been elucidated. EC initiate the process of neovascularization by invading the extracellular matrix, proliferating, migrating, and then forming capillary tubes (Auerbach et al., 1991; Folkman and Klagsbrun, 1987; Griffioen and Molema, 2000). One component of the process is thought to be the elaboration of proangiogenic growth factors (Auerbach et al., 1991; Folkman and Klagsbrun, 1987; Griffioen and Molema, 2000). Tumor cells are ischemic and hypoxic (Tannock and Rotin, 1989; Wike-Hooley et al., 1984), and one consequence of ischemia and hypoxia is low extracellular pH. Therefore, we tested this parameter for its bearing on the efficacy of angiostatin. Tumor-associated angiogenesis is a complex multistep process under the control of positive and negative soluble and insoluble factors. A mutual stimulation occurs between tumor and vascular cells by paracrine mechanisms (Rak et
0306-3623/02/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved. PII: S 0 3 0 6 - 3 6 2 3 ( 0 1 ) 0 0 11 5 - X
278
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
al., 1996). Because angiogenesis is a major prerequisite for progression of malignant tumors, and because the development of a vascular supply and stromal support are essential for tumor growth, tumor cells require a critical number of EC to switch to the angiogenic phenotype and form new vessels (Folkman and Shing, 1992), resulting in an increase in vessel density. There is compelling evidence that acquisition of the angiogenic phenotype is a common pathway for tumor progression, and that active angiogenesis is associated with other molecular mechanisms leading to tumor progression (Folkman, 1985, 1990). In contrast, this overgrowth of the tumor mass leads to a depletion of nutrients locally and an increase in tumor necrosis. This imbalance in tumor cell proliferation and death results in a microenvironment that has regions of hypoxia and high CO2 content with an accompanying overall decrease in pHe (Owen, 1996; Wike-Hooley et al., 1984). This relationship has been demonstrated for many types of tumors, such as invasive breast cancer (Weidner et al., 1991), non small cell lung cancer (Macchiarini et al., 1992), and prostate carcinoma (Weidner et al., 1993). To investigate factors that influence angiogenesis and to gain a more fundamental understanding of the cellular process involved in the development of new blood microvessels, a number of models of angiogenesis, which can be used to assess the local tumor microenvironment, have been developed. It has been clearly demonstrated that proangiogenic EC activity requires tumor matrix proteins for EC migration and proliferation, leading to neovascularization of the tumor mass (Auersperg et al., 1987; Canfield et al., 1986; Haralabopoulos et al., 1994). We tested angiostatin’s activity while providing cells with matrix proteins during experiments. Since angiostatin binding has been shown to disrupt FAK signaling (Claesson-Welsh et al., 1998) and ion transporters are located in FAK plaques (Grinstein et al., 1993), we tested the hypothesis that angiostatin binding may interfere with pHi regulation. Since pH regulation is also affected by substrate attachment (Plopper et al., 1995; Schwartz et al., 1991), we compared cells on fibronectin or fibronectin-like RGD peptides with cells on Matrigel, which is a mixture of tumor basement membrane proteins. In our systems, we were able to test effects of both angiostatin and extracellular acidification in substrateattached EC under conditions of physiologic bicarbonate and CO2. The changes in intracellular pH that we have been able to attribute to angiostatin’s activity make a large leap towards the unraveling of the mechanism by which this drug kills EC, and are the first such measurements made during exposure to an antiangiogenic agent. Furthermore, substrate affects the degree of expression of a-ATP synthase expression on the cells’ surface, and extracellular pH affects its distribution and probably thus its activity. These parameters need to be carefully considered in the evaluation of antiangiogenic agents.
2. Materials and methods 2.1. Materials Human recombinant angiostatin (Rh-angiostatin) was supplied by Entremed (Rockville, MD). Matrigel, Matrisperse, and endothelial cell growth supplement (ECGS) were purchased from Collaborative Biomedicals (Bedford, MA), and additional tumor stromal proteins (Matrigel-like) were produced as previously described (Grant et al., 1989). Bis carboxyethyl-carboxyl fluorescein-acetoxymethylester (BCECF-AM) was obtained from Molecular Probes (Eugene, OR). Serum-free endothelial basal medium (EBM) was obtained from Clonetics (Biowhittaker, Walkersville, MD). L-glutamine, penicillin/streptomycin, gentamicin and 0.5% fungizone, and trypsin/EDTA (0.05%: 0.53 mM) were from GibcoBRL (Grand Island, NY). Heparin sulfate was obtained from Fisher Scientific (Pittsburgh, PA). Six-well plates were obtained from Nalge Nunc International (Rochester, NY). The Leukostat/Hema 3 kit was obtained from Biomedical Sciences (Swedesboro, NJ). 2.2. Media For tube and migration/scratch assays using human umbilical vein endothelial cells (HUVEC), culture media used in the proliferation assay contained M199 media supplemented with 10% defined calf serum, 0.05% ECGS, 0.5% L-glutamine, 0.013% heparin sulfate, 0.5% penicillin/ streptomycin, 0.05% gentamycin and 0.5% fungizone (Grant et al., 1995), which has a bicarbonate concentration of 26 mM and a pHe = 7.3 under 5% CO2. Low pH (6.7) medium was prepared by reducing the bicarbonate to 7 mM and supplementing with 19 mM NaCl to keep the osmolarity constant (6.7 B maintained at 5% CO2) or using 26 mM medium in a 17% CO2 incubator (6.7 C). Data using each method were the same, so results from one 6.7 B and two 6.7 C experiments were averaged together. 2.3. HUVEC migration/scratch assay HUVEC were plated at a density of 1 106 cells/ml on a thin layer of 1:3 Matrigel, and allowed to adhere for a period of 24 h. Then, a scratch of 2 mm diameter was made in each plate, and the plates were incubated at pHe 7.3 or 6.7 with or without 30 mg/ml angiostatin for 6 h. Then intracellular pH or wound diameter was measured in parallel series of separate plates. For scratch diameter quantitation, plates were fixed and stained with a Leukostat kit. Measurements of scratch widths were done using an upright NIKON microscope attached to NIH image analysis system. 2.4. Acridine orange staining to assay for apoptosis HUVEC were plated at 70,000 cells in 500 ml of normal HUVEC medium in a four-well slide coated with diluted
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
(1:100) Matrigel. The slides were incubated at 37C to allow cells to attach and spread on the culture surface. The cells were incubated in culture medium either at pH 7.3 or 6.7 for 24 h. For a positive control, the medium was removed and replaced with medium lacking ECGS. Angiostatin was then added to the medium at 0, 30, or 100 mg/ml and incubated with the cells for 24 h at 37C. The medium was removed and 250 ml of diluted acridine orange (1:5 in Dulbecco’s phosphate-buffered saline or DPBS2 + ) was added and incubated with the cells for 5 min. The acridine orange was then removed, and the cells were fixed with Histochoice fixative (without alcohol) for 10 min at 4C, then washed for 15 min at room temperature. The cells were washed additionally five times with DPBS. Slides were mounted for viewing under the microscope and cell numbers measured using NIH image analysis of positively stained nuclei. 2.5. Histochemistry staining for a-subunit of ATP synthase HUVEC were plated in eight-well Labtek slides at a density of 70% confluency. Multiwell slides were first coated either with RGD peptides, fibronectin, gelatin, or Matrigel before plating the EC. Once the cells had been attached and spread, angiostatin was added to some of the wells at a concentration of 30 mg/ml in complete medium and incubated at 37C for 4 –6 h. In some experiments, the pHe was lowered to 6.7 in a high CO2 incubator. Cells were washed with Medium 199 and then with DPBS, fixed at 4C with Histochoice (without alcohol for nonpermeabilized staining) for 10 min, and then blocked with 5% BSA. For permeabilized cells, we used Histochoice with alcohol. Sometimes cells were further permeabilized with 0.1% Triton X-100 (Fisher Scientific) in DPBS. The cells were incubated with a polyclonal antibody to bovine a-ATP synthase (1:50 in DPBS) for 1 h at 37C. Cells were washed with DPBS and incubated in a Cy3 secondary antibody (1:1000 in DPBS; Sigma) for 30 min at 37C. 406-diamidino-2-phenylindole (DAPI) staining was done at the same time to identify the EC nuclei. After washing in DPBS, cells were mounted with SlowFade (Molecular Probes) and staining was observed in a NIKON upright fluorescence microscope. 2.6. HUVEC tube formation on Matrigel HUVEC culture medium is described in media above. HUVEC were pretreated for 24 h with angiostatin (0 –20 mg/ ml) and media at pHe 7.3 or 6.7 and transferred to six-well plates. These plates were coated with ice-cold Matrigel (300 ml/well of 15 mg/ml) and allowed to polymerize in a 37C incubator for 30 min. The angiostatin-pretreated cells were then harvested by incubation with trypsin/EDTA (0.05%: 0.53 mM), neutralized with their own culture medium, centrifuged, then resuspended in culture medium and added to each Matrigel-coated well at 50,000 cells/well.
279
The cells were then incubated for an additional 20 h in the presence or absence of angiostatin at 37C and 5% CO2 allowing for tube formation. Cells were fixed and stained with a Leukostat Kit and quantitated with a NIH image analysis system with the particle analysis macro to measure the square millimeters per visual field occupied by tubes at 15 magnification. This assay was done in triplicate for each experimental variable. 2.7. Intracellular pH measurements in HUVEC tubes and monolayers For determination of pHi, cells were plated on 35-mm microwell plastic dishes with 18-mm glass coverslips glued to the underside of 1-cm-diameter holes in the center of each dish (Mattek, Ashland, MA). For HUVEC tubes and wounds, cells were plated onto Matrigel, thin enough to focus through on the microscope stage. Cells were incubated for 4 min with 5 mM BCECF-AM, as previously described (Owen et al., 1995, 1998; Wahl et al., 1996, 1997, 2000). For the sprouts, BCECF-AM was injected via Hamilton syringe through the Matrigel in order to load embedded sprouts that were closest to the glass coverslip. Cells were postincubated in growth medium for 20 min at 37C, 5% CO2 to complete hydrolysis of the dye ester and allow recovery time from the dye loading. This procedure had to be done with extreme care in the continued presence of angiostatin, to avoid loss of the effect, indicating that the binding to its receptor is not of high affinity. Immediately after dye loading and recovery, each plate was then mounted on the microscope stage at 37C under humidified air containing 5% CO2. The experiments were performed in complete growth medium at 37C, modified as needed for each experimental protocol, under a flowing humidified atmosphere which contained 5% CO2 or 17% CO2, as indicated. Spectral analysis and calibration of pH measurement by whole fluorescence spectra were essentially the same as described previously for adherent cells containing BCECF (Owen et al., 1995, 1998; Wahl et al., 2000). Each mean pHi is calculated from five consecutive spectra on three to four fields of cells, shown with standard deviation. We have used an optical system that copes with the complete medium (containing serum and phenol red), dense substrates, and the need for bicarbonate buffering (under 5% CO2), as well as the necessity of cells to remain attached to substrate during pH measurements. These features are required to mimic physiologic and/or pathologic conditions in studies of EC. The system uses a fluorescent indicator dye that exhibits spectral shifts as a function of pH. The intracellular pH is calculated from the complete fluorescence spectrum (Owen et al., 1995), rather than from a ratio of two points in the spectrum as is commonly done (Grynkiewicz et al., 1985), and is thus resistant to artifacts and is more accurate than studies done on cells in suspension and under less physiologic conditions (Wahl et al., 1996).
280
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
3. Results Angiostatin was evaluated at normal and low pHe in a scratch/migration assay and during tube formation (differ-
entiation), both of which are established model systems used to study angiogenesis. In both models, angiostatin had an effect on cell proliferation and/or survival, which was dependent on pHe. Furthermore, angiostatin had an
Fig. 1. Graph and fluorescence images from HUVEC wound assay done with cells incubated in normal and low pH media ± angiostatin. (A) Graph shows average intracellular pH measurements in the HUVEC monolayer in confluent areas and at the scratch edge, done with and without 0 – 30 mg/ml angiostatin. The graph shows a significant decrease in pHi at scratch edge and in confluent monolayer following incubation at pHe = 6.7. This decrease in pHi was further increased in the presence of 30 mg/ml angiostatin. Panels (B) to (E) show fluorescence images of HUVEC monolayers in normal confluent zones or scratch edges following acridine orange staining to identify apoptotic cells. (B) At pHe 7.3 both in a confluent area and at a wound edge, there are no dying cells. (C) At pHe 6.7, there are a few (10 – 15%) dying cells at the wound edge, observed as round and hyperchromatic. (D) The addition of 30 mg/ml angiostatin at pHe 7.3 only slightly increases the number of hyperchromatic cells. (D) When the pHe is lowered to 6.7, 30 mg/ml angiostatin results in an increase to 40% apoptotic cells. (E) The combination of pHe 6.7 and 30 mg/ml angiostatin resulted in approximately 30% apoptotic cells away from the wound edge as well.
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
effect on pHi, but this effect was essentially absent unless pHe was reduced. Angiostatin was found to be active at a dose of 20 –30 mg/ml if it was in the presence of a reduced pHe. By comparison, at normal extracellular pH, this study and others have found that dose levels three times higher, or, alternatively, longer time frames, are required in order to demonstrate angiostatin activity
281
(Claesson-Welsh et al., 1998; Funaki et al., 1998; Nelson et al., 2000). This is shown in the migration/scratch assay (Fig. 1) and during tube formation, as a dose response (Fig. 2). We propose that the differences in dose level and duration in various laboratories may both relate to differences in pHe, which can vary as a function of time in culture.
Fig. 2. Dose response to angiostatin at normal and low pHe in the tube formation assay. Angiostatin was evaluated at a range of 0 – 20 mg/ml at pHe 7.3 and 6.7, as described in Materials and Methods. Micrographs of tube formation show tube network in bas relief in controls without angiostatin. With increasing amount of angiostatin, no significant change is observed at pHe 7.3, whereas the tube network is broken up at pH 6.7 with as little as 10 mg/ml angiostatin.
282
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
Fig. 3. Cells were plated on RGD fibronectin or Matrigel, and the histochemistry was performed to evaluate distribution of a-subunit of ATP synthase. Cells were either permeabilized or not permeabilized prior to immunofluorescence. Surface staining (unpermeabilized cells) for a-subunit of ATP synthase is low or nonexistent on cells plated on RGD or fibronectin, but cell surface labeling is strong on Matrigel (left panels). When cells were permeabilized, positive staining for a-subunit of ATP synthase was observed in all cells on the three matrices, typical of mitochondrial localization (right hand panels).
In the migration/scratch assay (Fig. 1), a 2-mm scratch was made through a confluent HUVEC monolayer. The effect of angiostatin on the cell migration and proliferation
needed to close the scratch was observed with and without extracellular acidification. Angiostatin had no effect on scratch diameter in plates which were kept at pHe 7.3
Fig. 4. The detection of the a-subunit of ATP synthase on HUVEC following 30 mg/ml angiostatin pretreatment. The blocking and redistribution of a-subunit of ATP synthase are seen with low pH incubation, and blocking of receptor by preincubation with angiostatin.
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
(50% closure without angiostatin and 55% with 30 mg/ml angiostatin). At pHe 6.7, there was a 75% closure without angiostatin, and 50% in the presence of 30 mg/ml angiostatin (data not shown). These data clearly suggested that there is a decrease in the migratory ability of the EC. Plates of cells used in the migration/scratch assay were loaded with BCECF for measurements of intracellular pH. Marked changes in pHi were observed with acidification and angiostatin, as shown in Fig. 1. At normal pHe, angiostatin alone had no effect on pHi. Acidification of the medium by 0.6 U in pH in the absence of angiostatin resulted in a decrease in pHi by about 0.3 U. This degree of pH homeostasis is extremely typical of most cell lines and was seen on cells both from the scraped edge and within the monolayer. In the presence of angiostatin, the same acidification produced a decrease in pHe of 0.5– 0.6 U, suggesting some loss of control over intracellular pH regulation. Plates that were wounded were also incubated with acridine orange to measure cell death (apoptosis) (Erenpreisa et al., 1997; Lucas et al., 1998; Whitacre and Berger, 1997). In Fig. 1 (bottom panel), bright cells were usually rounded and potentially dead. There was a slight (5– 10%) increase in the number of bright cells at a scratch edge as a result of lowering pHe from 7.3 to 6.7. The ef-fect of angiostatin at pHe 7.3 was negligible. The combination of low pHe and angiostatin produced an increase (up to a level of 30 – 40%) in the number of hyperchromatic (bright, potentially apoptotic) cells at or near the scratch edge. In order to test the effect of pHe on EC differentiation, HUVEC were plated on polymerized Matrigel (tumor basement membrane proteins) at either pHe 7.3 or 6.7. It has previously been shown that angiostatin inhibits differentiation (EC tube formation) at a concentration (IC50) of 100 mg/ml for 16 h (Nelson et al., 2000). Pretreated HUVEC were plated with or without angiostatin at either a pHe 7.3 or 6.7 for 20 h to allow tube formation. Tubes were then stained, images were converted to relief images, and area occupied by tubes was quantitated, resulting in the tube areas shown. Fig. 2, a dose – response experiment, showed that tube formation was noticeably inhibited by 20 mg/ml angiostatin when extracellular pH was acidified to pH 6.7, whereas the drug had no effect at pH 7.3. The IC50 for inhibition of tube formation by angiostatin at pHe 7.3 was 80 mg/ml, whereas at pHe 6.7 it was reduced to 20 mg/ml. Tube formation was affected approximately threefold by 20 mg/ml angiostatin at low pHe. High dose levels overcame the need for low pHe, so that no pH effect on tube formation was observed when 100 mg/ml angiostatin was used (data not shown). Parallel experiments to measure pHi on the HUVEC produced effects that were in agreement with the findings in the other systems. Angiostatin at 30 mg/ml had no effect on pHi when pHe was 7.3. However, pHi decreased by approximately 0.3 U when pHe was 6.7 (Wahl et al., 2002). Fig. 3 shows the effect of substrate on the cell surface staining of the a-subunit of ATP synthase. This enzyme
283
usually is found in mitochondria, as observed in the righthand panels that shows cells permeabilized with detergent or ethanol. The left-hand panels show positive staining for ATP synthase on the cells’ surface in the presence of a Matrigel substratum, but not in the presence of RGD peptides, which simulate the integrin-clustering amino acid sequence of fibronectin, or in the presence of native human fibronectin itself. Matrigel is a mixture of basement membrane and growth factor proteins typically found in the Englebreth –Holm –Swarm (EHS) tumor stroma, and therefore, the experiments show a dependence on receptor expression of tumor-like microenvironmental stimuli. In Fig. 4, cells plated on dilute Matrigel substratum as compared at pHe 7.3 and 6.7. The staining is different at low pH, and in fact is more punctate. This distribution is in line with possible distribution of FAK plaques, and may therefore increase the likelihood of perturbation of pH regulation via interaction of angiostatin’s receptor with FAK.
4. Discussion In the interest of obtaining data that could be compared to work done by others in vivo, the effects of angiostatin at different pHe and on pHi were measured in EC, which retained their attachment to tumor stromal proteins. We have shown that angiostatin lowered pHi in EC under conditions of low pHe, resulting in a decrease in cell migration and an increase in cell death. This corroborates work showing that proangiogenic vessel growth requires matrix for the EC migration and proliferation, which leads to neovascularization of the tumor mass (Auersperg et al., 1987; Canfield et al., 1986). Low pHe alone (without angiostatin) has also been shown to have an effect on endothelial vessel sprouting in two angiogenesis model systems. In one (Burbridge et al., 1999), a delay was measured in three-dimensional sprout outgrowth at pHe 6.9 vs. 7.4, which could be overcome by exogenous growth factors. We have confirmed these observations in our aortic sprout model; in addition, we have shown that angiostatin takes advantage of this situation by further lowering the pHi, resulting in EC cell death. This is an important feature that helps to understand angiostatin’s efficacy and specificity in vivo in a low pHe tumor environment. Another study (D’Arcangelo et al., 1999) demonstrated that low pH reduced apoptotic death; they also concluded that low pHe inhibited proangiogenic behavior in bovine EC. Regulation of pHi has also been shown to be dependent on cell attachment in fibroblasts (Frisch et al., 1996; Grinstein et al., 1993; Plopper et al., 1995; Schwartz et al., 1991) and in other cells that require substrate attachment for growth (Schwartz, 1993; Wahl et al., 1997). Folkman’s laboratory has provided evidence that angiostatin causes an aberrant activation of focal adhesion kinase (FAK) plaques in EC (Claesson-Welsh et al., 1998). It has
284
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
been shown (Plopper et al., 1995) that adhesion-dependent survival is regulated by FAK, and FAK expression has been correlated positively with motility in melanoma cells (Akasaka et al., 1995). FAK plaques are organizing centers that mediate effects on cytoskeleton and other intracellular events, some of which appear to involve alteration in proton transport (Plopper et al., 1995). In fibroblasts, proton transport has been implicated in cell migration as well (Grinstein et al., 1993). If angiostatin-dependent aberrant activation of FAK plaques in EC were to compromise the ability of transporters to regulate intracellular pH, this could be part of the mechanism by which angiostatin kills EC. This mechanism would be inherently sensitive to low extracellular pH because low pHe activates two of the major proton transporters, the sodium – proton exchanger isoform 1 (NHE-1) (Counillon and Pouyssegur, 2000) and the H + -linked monocarboxylate transporters (MCT) (Halestrap and Price, 1999). The relative activities of these transporter families in EC have not yet been characterized. Evidence (Moser et al., 1999, 2001) has been provided that the a-subunit of ATP synthase or a homologue of it can be detected on the extracellular surface of the plasma membrane and that it binds to angiostatin. The complete ATP synthase protein is composed of many subunits in a large, complex structure found in mitochondria, where it functions in proton transport. As was originally pointed out by Gillies (1999), the proton gradient required to synthesize extracellular ATP on the cell surface would be contrary to the inward direction of the gradient in acidic parts of a tumor (Gerweck et al., 1991; Wahl et al., 1997). It is possible that the a-subunit of ATP synthase is interacting with another proton transporter in a novel way, or alternatively, that all or part of the ATP synthase complex is functioning as an ATPase in this location. The discovery that moderate doses of angiostatin may be far more effective against EC in an acidic extracellular milieu may help explain why angiostatin has been found to have different effects on physiological and pathological angiogenesis (Folkman, 1997). Tumors are often distinguished from most of the nonpathological examples of angiogenesis by the relative hypoxia and acidosis that frequently characterize the tumor extracellular microenvironment. This is because acidosis is often a corollary of hypoxia and reliance on glycolytic metabolism, together with poor, irregular vascularity, impedes removal of catabolites, including lactic and carbonic acids (Yamagata and Tannock, 1996). The data in this report focus on the effects of reduced pHe on angiostatin activity. The present study indicates that pH should be accorded more attention as a parameter in in vitro assays. For example, if angiostatin requires several days in culture to produce an effect, it will be important to monitor medium pH carefully. The development of medium acidity could easily produce an apparent dependence on elapsed time that was, in reality, a dependence on a gradually decreasing pHe.
In conclusion, these studies describe the critical interplay between the extracellular matrix and pH microenvironment of the tumor bed and antiangiogenic drugs during neovascularization. These parameters should be evaluated for other antitumor drugs currently under development.
Acknowledgments We thank Michael Zahaczewsky for expert technical assistance. The rh-angiostatin was a generous gift from Entremed. This work has been supported by NIH grant P01 CA56690 (M.L.W.) funds from Merck (West Point, PA) (D.S.G.) and Translational Grant Funding from Entremed (D.S.G.).
References Akasaka, T., vanLeeuwen, R.L., Yoshinaga, I.G., Mihm, M.C., Byers, H.R., 1995. Focal adhesion kinase (p125FAK) expression correlates with motility of human melanoma cell lines. J. Invest. Dermatol. 105, 104 – 108. Auerbach, R., Auerbach, W., Polakowski, I., 1991. Assays for angiogenesis: a review. Pharmacol. Ther. 51 (1), 1 – 11. Auersperg, N., Pawson, T., Worth, A., Weinmaster, G., 1987. Modifications of tumor histology by point mutations in the v-fps oncogene: possible role of extracellular matrix. Cancer Res. 47 (23), 6341 – 6348. Burbridge, M.F., West, D.C., Atassi, G., Tucker, G.C., 1999. The effect of extracellular pH on angiogenesis in vitro. Angiogenesis 3, 281 – 288. Canfield, A.E., Schor, A.M., Schor, S.L., Grant, M.E., 1986. The biosynthesis of extracellular-matrix components by bovine retinal endothelial cells displaying distinctive morphological phenotypes. Biochem. J. 235 (2), 375 – 383. Claesson-Welsh, L., Welsh, M., Ito, N., Ananda-Apte, B., Soker, S., Zetter, B., O’Reilly, M., Folkman, J., 1998. Angiostatin induces endothelial cell apoptosis and activation of focal adhesion kinase independently of the integrin-binding motif RGD. Proc. Natl. Acad. Sci. USA 95, 5579 – 5583. Counillon, L., Pouyssegur, J., 2000. The expanding family of eucaryotic Na + /H + exchangers. J. Biol. Chem. 275, 1 – 4. D’Arcangelo, D., Facchiano, F., Barlucchi, L.M., Melillo, G., Illi, B., Testolin, L., Gaetano, C., Capogrossi, M.C., 1999. Acidosis inhibits endothelial cell apoptosis and function and induces basic fibroblast growth factor and vascular endothelial growth factor expression. Circ. Res. 86, 312 – 318. Erenpreisa, J., Freivalds, T., Roach, H., Alston, R., 1997. Apoptotic cell nuclei favour aggregation and fluorescence quenching of DNA dyes. Histochem. Cell Biol. 108, 67 – 75. Folkman, J., 1985. Tumor angiogenesis. Adv. Cancer Res. 43, 175 – 203. Folkman, J., 1990. What is the evidence that tumors are angiogenesis dependent? J. Natl. Cancer Inst. 82, 4 – 6. Folkman, J., 1997. Regulation of Angiogenesis. Birkaeuser Verlag, Basel, pp. 1 – 8. Folkman, J., Ingber, D., 1992. Inhibition of angiogenesis. Semin. Cancer Biol. 3, 89 – 96. Folkman, J., Klagsbrun, M., 1987. Angiogenic factors. Science 235, 442 – 447. Folkman, J., Shing, Y., 1992. Angiogenesis. J. Biol. Chem. 267, 10931 – 10934. Frisch, S.M., Vuori, K., Ruoslahti, E., Chan-Hui, P.-Y., 1996. Control of adhesion-dependent cell survival by focal adhesion kinase. J. Cell Biol. 134, 793 – 799. Funaki, N.O., Tanaka, J., Imamura, M., 1998. Quantitative analysis of
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285 alpha-fetoprotein mRNA in circulating peripheral blood of patients with hepatocellular and a-fetoprotein-producing gastric carcinomas. Life Sci. 62, 1973 – 1984. Gerweck, L.E., Rhee, J.G., Koutcher, J.A., Song, C.W., Urano, M., 1991. Regulation of pH in murine tumor and muscle. Radiat. Res. 126, 206 – 209. Gillies, R.J., 1999. Angiostatin’s partners. Science 288, 434. Grant, D.S., Kleinman, H.K., Martin, G.R., 1989. The role of basement membrane in vascular development. Ann. N.Y. Acad. Sci. 588, 61 – 72. Grant, D.S., Kinsella, J.L., Kibbey, M.C., LaFlamme, S., Burbelo, P.D., Goldstein, A.L., Kleinman, H.K., 1995. Matrigel induces thymosin b4 gene in differentiating endothelial cells. J. Cell Sci. 108. Griffioen, A.W., Molema, G., 2000. Angiogenesis: potentials for pharmacologic intervention in the treatment of cancer, cardiovascular diseases and chronic inflammation. Pharmacol. Rev. 52, 237 – 268. Grinstein, S., Woodside, M., Waddell, T.K., Downey, G.P., Orlowski, J., Pouyssegur, J., Wong, D.C.P., Foskett, J.K., 1993. Focal localization of the NHE-1 isoform of the Na + /H + antiport: assessment of effect on intracellular pH. EMBO J. 12 (13), 5209 – 5218. Grynkiewicz, G., Poenie, M., Tsien, R.Y., 1985. A new generation of Ca2 + indicators with greatly improved fluorescence properties. J. Biol. Chem. 260, 3440 – 3450. Halestrap, A.P., Price, N.T., 1999. The proton-linked monocarboxylate transporter (MCT) family: structure, function, and regulation. Biochem. J. 343, 281 – 299. Haralabopoulos, G.C., Grant, D.S., Kleinman, H.K., Lelkes, P.I., Papaioannou, S.P., Maragoudakis, M.E., 1994. Inhibitors of basement membrane collagen synthesis prevent endothelial cell alignment in Matrigel in vitro and angiogenesis in vivo. Lab. Invest. 71 (4), 575 – 582. Lucas, R., Holmgren, L., Garcia, I., Jimenez, B., Mandriota, S.J., Borlat, F., Sim, B.K., Wu, Z., Grau, G.E., Shing, Y., Soff, G.A., Bouck, N., Pepper, M.S., 1998. Multiple forms of angiostatin induce apoptosis in endothelial cells. Blood 92 (12), 4730 – 4741. Macchiarini, P., Fontanini, G., Hardin, M.J., Squartini, F., Angeletti, C.A., 1992. Relation of neovascularisation to metastasis of non-small-cell lung cancer. Lancet 340, 145 – 146. Moser, T.L., Stack, M.S., Asplin, I., Enghild, J.J., Hojrup, P., Everitt, L., Hubchak, S., Schnaper, H.W., Pizzo, S.V., 1999. Angiostatin binds ATP synthase on the surface of human endothelial cells. Proc. Natl. Acad. Sci. USA 96, 2811 – 2816. Moser, T.L., Kinan, D.J., Ashley, T.A., Roy, J.A., Goodman, M.D., Misro, U.K., Cheek, D.J., Pizzo, S., 2001. Endothelial cell surface F1 – Fo ATP synthase is active in ATP synthesis and is inhibited by angiostatin. Proc. Natl. Acad. Sci. 98, 6656 – 6661. Nelson, B.J., Vu, H.A., Hanson, A.D., Fortier, A.H., Sim, B.K.L., 2000. Recombinant human angiostatin protein inhibits both migration and tube formation of human endothelial cells. Proc. Am. Assoc. Cancer Res. 41, 485. O’Reilly, M.S., Holmgren, L., Shing, Y., Chen, C., Rosenthal, R.A., Lane, W.S., Cao, Y., Sage, E.H., Folkman, J., 1994. Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell 79, 315 – 328. Owen, C.S., 1996. Dependence of proton generation on aerobic or anaerobic metabolism and implications for tumor pH. Int. J. Hyperthermia 12, 495 – 499.
285
Owen, C.S., Wahl, M.L., Leeper, D.B., Perry, H.D., Bobyock, S.B., Russel, M., Woodward, W., 1995. Accurate whole spectrum measurements of intracellular pH and [Na + ]. J. Fluoresc. 5, 329 – 335. Owen, C.S., Wahl, M.L., Pooler, P.M., Coss, R.A., Leeper, D.B., 1998. Temporal association between alterations in proton extrusion and low pH adaptation. Int. J. Hyperthermia 14, 227 – 232. Plopper, G.E., McNamee, H.P., Dike, L.E., Bojanowski, K., Ingber, D.E., 1995. Convergence of integrin and growth factor receptor signaling pathways within the focal adhesion complex. Mol. Biol. Cell 6, 1349 – 1365. Rak, J., Filmus, J., Kerbel, R.S., 1996. Reciprocal paracrine interactions between tumour cells and endothelial cells ‘angiogenesis progression’ hypothesis. Eur. J. Cancer 32A, 2438 – 2450. Schwartz, M.A., 1993. Signaling by integrins: implications for tumorigenesis. Cancer Res. 53, 1503 – 1506. Schwartz, M.A., Lechene, C., Ingber, D.E., 1991. Insoluble fibronectin activates the Na + /H + antiporter by clustering and immobilizing integrin a v b 1, independent of cell shape. Proc. Nat. Acad. Sci. USA 88, 7849 – 7853. Tannock, I., Rotin, D., 1989. Acid pH in tumors and its potential for therapeutic exploitation. Cancer Res. 49, 4373 – 4384. Wahl, M.L., Coss, R.A., Bobyock, S.B., Leeper, D.B., Owen, C.S., 1996. Thermotolerance and intracellular pH in two Chinese hamster cell lines adapted to growth at low pH. J. Cell. Physiol. 166, 438 – 445. Wahl, M.L., Bobyock, S.B., Leeper, D.B., Owen, C.S., 1997. Effects of 42C hyperthermia on intracellular pH in ovarian carcinoma cells during acute acidification or chronic exposure to low extracellular pH. Int. J. Radiat. Oncol., Biol., Phys. 39, 205 – 212. Wahl, M.L., Pooler, P.M., Briand, P., Leeper, D.B., Owen, C.S., 2000. Intracellular pH regulation in a nonmalignant and a derived malignant human breast cell line. J. Cell. Physiol. 183, 373 – 380. Wahl, M., Owen, C., Grant, D., 2002. Low dose angiostatin induces intracellular acidosis and anoikis cell death in endothelial cells at tumor-like low pH. Endothelium (in press). Weidner, N., Semple, J.P., Welch, W.R., Folkman, J., 1991. Tumor angiogenesis and metastasis-correlation in invasive breast carcinoma. N. Engl. J. Med. 324, 1 – 8. Weidner, N., Carroll, P.R., Flax, J., Blumenfeld, W., Folkman, J., 1993. Tumor angiogenesis correlates with metastasis in invasive prostate carcinoma. Am. J. Pathol. 143, 409 – 410. Whitacre, C.M., Berger, N.A., 1997. Factors affecting topotecan-induced programmed cell death, adhesion protects cells from apoptosis and impairs cleavage of poly(ADP-ribose)polymerase. Cancer Res. 57, 2157 – 2163. Wike-Hooley, J., Haveman, J., Reinhold, H.S., 1984. The relevance of tumor pH in the treatment of malignant disease. Radiother. Oncol. 2, 342 – 366. Wu, Z., O’Reilly, M.S., Folkman, J., Shing, Y., 1997. Suppression of tumor growth with recombinant murine angiostatin. Biochem. Biophys. Res. Commun. 236 (3), 651 – 654. Yamagata, M., Tannock, I., 1996. The chronic administration of drugs that inhibit the regulation of intracellular pH: in vitro and anti-tumour effects. Br. J. Cancer 73, 1328 – 1334.
Effects of microenvironmental extracellular pH and extracellular matrix proteins on angiostatin’s activity and on intracellular pH Miriam L. Wahla,*, Derrick S. Grantb a
Department of Biochemistry and Molecular Pharmacology, Thomas Jefferson University, 233 South, 10th Street, Room 226, Philadelphia, PA 19107, USA b Cardeza Foundation, Department of Medicine, Thomas Jefferson University, Philadelphia, PA 19107, USA Received 1 September 2001; accepted 1 October 2001
Abstract Antiangiogenic agents target migratory and proliferative endothelial cells (EC) in the process of forming new vessels, resulting in growth inhibition or cell death. Here we have shown that the antiangiogenic activity of angiostatin on EC is enhanced in culture when the microenvironmental extracellular pH (pHe) is reduced to levels similar to that of many tumors. In a migration/scratch assay and during tube formation, angiostatin in combination with reduced pHe synergistically resulted in an increased EC death—an effect not seen with either stimulus individually. Lowering of pHe decreased intracellular pH (pHi), and a further lowering of pHi occurred when low pHe was combined with angiostatin. These data suggest that low pHe plays a role in the relative specificity and efficacy of angiostatin for tumor neovasculature and indicate roles for both pHe and pHi in the mechanism of angiostatin action. A receptor for angiostatin, the a-subunit of ATP synthase, was found on the surface of EC. We show that cell surface receptor distribution is increased on Matrigel, a basement-like matrix, as opposed to fibronectin or RGD peptide substrates, and redistributed to a more punctuate appearance at low pHe. Furthermore, positive cell surface histochemical staining for a-ATP synthase was blocked by preincubation with angiostatin. These data indicate that substrate and pHe are critical parameters in the evaluation of this antiangiogenic substance, and probably for others as well. D 2002 Elsevier Science Inc. All rights reserved. Keywords: Extracellular acidification; Angiogenesis; Angiostatin; Intracellular pH regulation; Fluorescent indicator dye
1. Introduction Blood vessel formation is essential for most tumor growths (Folkman, 1990; Folkman and Ingber, 1992), and antiangiogenic drugs such as angiostatin have been shown to reduce tumor vascularization by blocking endothelial cells (EC) proliferation and migration (Lucas et al., 1998; O’Reilly et al., 1994; Wu et al., 1997). The exact Abbreviations: angiostatin, rh-angiostatin, human recombinant angiostatin1 (Entremed, Rockville, MD); BCECF-AM, bis carboxyethyl-carboxyl fluorescein-acetoxymethylester; Cl/HCO3, chloride/bicarbonate exchanger; DAPI, 406-diamidino-2-phenylindole; dilactate DPBS, Dulbecco’s phosphate-buffered saline; EBM, endothelial basal medium; EC, endothelial cells; ECGS, endothelial cell growth supplement; pHe, extracellular pH; FAK, focal adhesion kinase; MCT, H+-linked monocarboxylate transporter; HUVEC, human umbilical vein endothelial cells; pHi,, intracellular pH; NHE-1, sodium – proton exchanger isoform 1. * Corresponding author. Tel.: +1-215-503-7867; fax: +1-215-923-9162. E-mail address: [email protected] (M.L. Wahl).
mechanism of action of angiostatin and the basis for its selectivity (within tumor beds) have not yet been elucidated. EC initiate the process of neovascularization by invading the extracellular matrix, proliferating, migrating, and then forming capillary tubes (Auerbach et al., 1991; Folkman and Klagsbrun, 1987; Griffioen and Molema, 2000). One component of the process is thought to be the elaboration of proangiogenic growth factors (Auerbach et al., 1991; Folkman and Klagsbrun, 1987; Griffioen and Molema, 2000). Tumor cells are ischemic and hypoxic (Tannock and Rotin, 1989; Wike-Hooley et al., 1984), and one consequence of ischemia and hypoxia is low extracellular pH. Therefore, we tested this parameter for its bearing on the efficacy of angiostatin. Tumor-associated angiogenesis is a complex multistep process under the control of positive and negative soluble and insoluble factors. A mutual stimulation occurs between tumor and vascular cells by paracrine mechanisms (Rak et
0306-3623/02/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved. PII: S 0 3 0 6 - 3 6 2 3 ( 0 1 ) 0 0 11 5 - X
278
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
al., 1996). Because angiogenesis is a major prerequisite for progression of malignant tumors, and because the development of a vascular supply and stromal support are essential for tumor growth, tumor cells require a critical number of EC to switch to the angiogenic phenotype and form new vessels (Folkman and Shing, 1992), resulting in an increase in vessel density. There is compelling evidence that acquisition of the angiogenic phenotype is a common pathway for tumor progression, and that active angiogenesis is associated with other molecular mechanisms leading to tumor progression (Folkman, 1985, 1990). In contrast, this overgrowth of the tumor mass leads to a depletion of nutrients locally and an increase in tumor necrosis. This imbalance in tumor cell proliferation and death results in a microenvironment that has regions of hypoxia and high CO2 content with an accompanying overall decrease in pHe (Owen, 1996; Wike-Hooley et al., 1984). This relationship has been demonstrated for many types of tumors, such as invasive breast cancer (Weidner et al., 1991), non small cell lung cancer (Macchiarini et al., 1992), and prostate carcinoma (Weidner et al., 1993). To investigate factors that influence angiogenesis and to gain a more fundamental understanding of the cellular process involved in the development of new blood microvessels, a number of models of angiogenesis, which can be used to assess the local tumor microenvironment, have been developed. It has been clearly demonstrated that proangiogenic EC activity requires tumor matrix proteins for EC migration and proliferation, leading to neovascularization of the tumor mass (Auersperg et al., 1987; Canfield et al., 1986; Haralabopoulos et al., 1994). We tested angiostatin’s activity while providing cells with matrix proteins during experiments. Since angiostatin binding has been shown to disrupt FAK signaling (Claesson-Welsh et al., 1998) and ion transporters are located in FAK plaques (Grinstein et al., 1993), we tested the hypothesis that angiostatin binding may interfere with pHi regulation. Since pH regulation is also affected by substrate attachment (Plopper et al., 1995; Schwartz et al., 1991), we compared cells on fibronectin or fibronectin-like RGD peptides with cells on Matrigel, which is a mixture of tumor basement membrane proteins. In our systems, we were able to test effects of both angiostatin and extracellular acidification in substrateattached EC under conditions of physiologic bicarbonate and CO2. The changes in intracellular pH that we have been able to attribute to angiostatin’s activity make a large leap towards the unraveling of the mechanism by which this drug kills EC, and are the first such measurements made during exposure to an antiangiogenic agent. Furthermore, substrate affects the degree of expression of a-ATP synthase expression on the cells’ surface, and extracellular pH affects its distribution and probably thus its activity. These parameters need to be carefully considered in the evaluation of antiangiogenic agents.
2. Materials and methods 2.1. Materials Human recombinant angiostatin (Rh-angiostatin) was supplied by Entremed (Rockville, MD). Matrigel, Matrisperse, and endothelial cell growth supplement (ECGS) were purchased from Collaborative Biomedicals (Bedford, MA), and additional tumor stromal proteins (Matrigel-like) were produced as previously described (Grant et al., 1989). Bis carboxyethyl-carboxyl fluorescein-acetoxymethylester (BCECF-AM) was obtained from Molecular Probes (Eugene, OR). Serum-free endothelial basal medium (EBM) was obtained from Clonetics (Biowhittaker, Walkersville, MD). L-glutamine, penicillin/streptomycin, gentamicin and 0.5% fungizone, and trypsin/EDTA (0.05%: 0.53 mM) were from GibcoBRL (Grand Island, NY). Heparin sulfate was obtained from Fisher Scientific (Pittsburgh, PA). Six-well plates were obtained from Nalge Nunc International (Rochester, NY). The Leukostat/Hema 3 kit was obtained from Biomedical Sciences (Swedesboro, NJ). 2.2. Media For tube and migration/scratch assays using human umbilical vein endothelial cells (HUVEC), culture media used in the proliferation assay contained M199 media supplemented with 10% defined calf serum, 0.05% ECGS, 0.5% L-glutamine, 0.013% heparin sulfate, 0.5% penicillin/ streptomycin, 0.05% gentamycin and 0.5% fungizone (Grant et al., 1995), which has a bicarbonate concentration of 26 mM and a pHe = 7.3 under 5% CO2. Low pH (6.7) medium was prepared by reducing the bicarbonate to 7 mM and supplementing with 19 mM NaCl to keep the osmolarity constant (6.7 B maintained at 5% CO2) or using 26 mM medium in a 17% CO2 incubator (6.7 C). Data using each method were the same, so results from one 6.7 B and two 6.7 C experiments were averaged together. 2.3. HUVEC migration/scratch assay HUVEC were plated at a density of 1 106 cells/ml on a thin layer of 1:3 Matrigel, and allowed to adhere for a period of 24 h. Then, a scratch of 2 mm diameter was made in each plate, and the plates were incubated at pHe 7.3 or 6.7 with or without 30 mg/ml angiostatin for 6 h. Then intracellular pH or wound diameter was measured in parallel series of separate plates. For scratch diameter quantitation, plates were fixed and stained with a Leukostat kit. Measurements of scratch widths were done using an upright NIKON microscope attached to NIH image analysis system. 2.4. Acridine orange staining to assay for apoptosis HUVEC were plated at 70,000 cells in 500 ml of normal HUVEC medium in a four-well slide coated with diluted
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
(1:100) Matrigel. The slides were incubated at 37C to allow cells to attach and spread on the culture surface. The cells were incubated in culture medium either at pH 7.3 or 6.7 for 24 h. For a positive control, the medium was removed and replaced with medium lacking ECGS. Angiostatin was then added to the medium at 0, 30, or 100 mg/ml and incubated with the cells for 24 h at 37C. The medium was removed and 250 ml of diluted acridine orange (1:5 in Dulbecco’s phosphate-buffered saline or DPBS2 + ) was added and incubated with the cells for 5 min. The acridine orange was then removed, and the cells were fixed with Histochoice fixative (without alcohol) for 10 min at 4C, then washed for 15 min at room temperature. The cells were washed additionally five times with DPBS. Slides were mounted for viewing under the microscope and cell numbers measured using NIH image analysis of positively stained nuclei. 2.5. Histochemistry staining for a-subunit of ATP synthase HUVEC were plated in eight-well Labtek slides at a density of 70% confluency. Multiwell slides were first coated either with RGD peptides, fibronectin, gelatin, or Matrigel before plating the EC. Once the cells had been attached and spread, angiostatin was added to some of the wells at a concentration of 30 mg/ml in complete medium and incubated at 37C for 4 –6 h. In some experiments, the pHe was lowered to 6.7 in a high CO2 incubator. Cells were washed with Medium 199 and then with DPBS, fixed at 4C with Histochoice (without alcohol for nonpermeabilized staining) for 10 min, and then blocked with 5% BSA. For permeabilized cells, we used Histochoice with alcohol. Sometimes cells were further permeabilized with 0.1% Triton X-100 (Fisher Scientific) in DPBS. The cells were incubated with a polyclonal antibody to bovine a-ATP synthase (1:50 in DPBS) for 1 h at 37C. Cells were washed with DPBS and incubated in a Cy3 secondary antibody (1:1000 in DPBS; Sigma) for 30 min at 37C. 406-diamidino-2-phenylindole (DAPI) staining was done at the same time to identify the EC nuclei. After washing in DPBS, cells were mounted with SlowFade (Molecular Probes) and staining was observed in a NIKON upright fluorescence microscope. 2.6. HUVEC tube formation on Matrigel HUVEC culture medium is described in media above. HUVEC were pretreated for 24 h with angiostatin (0 –20 mg/ ml) and media at pHe 7.3 or 6.7 and transferred to six-well plates. These plates were coated with ice-cold Matrigel (300 ml/well of 15 mg/ml) and allowed to polymerize in a 37C incubator for 30 min. The angiostatin-pretreated cells were then harvested by incubation with trypsin/EDTA (0.05%: 0.53 mM), neutralized with their own culture medium, centrifuged, then resuspended in culture medium and added to each Matrigel-coated well at 50,000 cells/well.
279
The cells were then incubated for an additional 20 h in the presence or absence of angiostatin at 37C and 5% CO2 allowing for tube formation. Cells were fixed and stained with a Leukostat Kit and quantitated with a NIH image analysis system with the particle analysis macro to measure the square millimeters per visual field occupied by tubes at 15 magnification. This assay was done in triplicate for each experimental variable. 2.7. Intracellular pH measurements in HUVEC tubes and monolayers For determination of pHi, cells were plated on 35-mm microwell plastic dishes with 18-mm glass coverslips glued to the underside of 1-cm-diameter holes in the center of each dish (Mattek, Ashland, MA). For HUVEC tubes and wounds, cells were plated onto Matrigel, thin enough to focus through on the microscope stage. Cells were incubated for 4 min with 5 mM BCECF-AM, as previously described (Owen et al., 1995, 1998; Wahl et al., 1996, 1997, 2000). For the sprouts, BCECF-AM was injected via Hamilton syringe through the Matrigel in order to load embedded sprouts that were closest to the glass coverslip. Cells were postincubated in growth medium for 20 min at 37C, 5% CO2 to complete hydrolysis of the dye ester and allow recovery time from the dye loading. This procedure had to be done with extreme care in the continued presence of angiostatin, to avoid loss of the effect, indicating that the binding to its receptor is not of high affinity. Immediately after dye loading and recovery, each plate was then mounted on the microscope stage at 37C under humidified air containing 5% CO2. The experiments were performed in complete growth medium at 37C, modified as needed for each experimental protocol, under a flowing humidified atmosphere which contained 5% CO2 or 17% CO2, as indicated. Spectral analysis and calibration of pH measurement by whole fluorescence spectra were essentially the same as described previously for adherent cells containing BCECF (Owen et al., 1995, 1998; Wahl et al., 2000). Each mean pHi is calculated from five consecutive spectra on three to four fields of cells, shown with standard deviation. We have used an optical system that copes with the complete medium (containing serum and phenol red), dense substrates, and the need for bicarbonate buffering (under 5% CO2), as well as the necessity of cells to remain attached to substrate during pH measurements. These features are required to mimic physiologic and/or pathologic conditions in studies of EC. The system uses a fluorescent indicator dye that exhibits spectral shifts as a function of pH. The intracellular pH is calculated from the complete fluorescence spectrum (Owen et al., 1995), rather than from a ratio of two points in the spectrum as is commonly done (Grynkiewicz et al., 1985), and is thus resistant to artifacts and is more accurate than studies done on cells in suspension and under less physiologic conditions (Wahl et al., 1996).
280
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
3. Results Angiostatin was evaluated at normal and low pHe in a scratch/migration assay and during tube formation (differ-
entiation), both of which are established model systems used to study angiogenesis. In both models, angiostatin had an effect on cell proliferation and/or survival, which was dependent on pHe. Furthermore, angiostatin had an
Fig. 1. Graph and fluorescence images from HUVEC wound assay done with cells incubated in normal and low pH media ± angiostatin. (A) Graph shows average intracellular pH measurements in the HUVEC monolayer in confluent areas and at the scratch edge, done with and without 0 – 30 mg/ml angiostatin. The graph shows a significant decrease in pHi at scratch edge and in confluent monolayer following incubation at pHe = 6.7. This decrease in pHi was further increased in the presence of 30 mg/ml angiostatin. Panels (B) to (E) show fluorescence images of HUVEC monolayers in normal confluent zones or scratch edges following acridine orange staining to identify apoptotic cells. (B) At pHe 7.3 both in a confluent area and at a wound edge, there are no dying cells. (C) At pHe 6.7, there are a few (10 – 15%) dying cells at the wound edge, observed as round and hyperchromatic. (D) The addition of 30 mg/ml angiostatin at pHe 7.3 only slightly increases the number of hyperchromatic cells. (D) When the pHe is lowered to 6.7, 30 mg/ml angiostatin results in an increase to 40% apoptotic cells. (E) The combination of pHe 6.7 and 30 mg/ml angiostatin resulted in approximately 30% apoptotic cells away from the wound edge as well.
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
effect on pHi, but this effect was essentially absent unless pHe was reduced. Angiostatin was found to be active at a dose of 20 –30 mg/ml if it was in the presence of a reduced pHe. By comparison, at normal extracellular pH, this study and others have found that dose levels three times higher, or, alternatively, longer time frames, are required in order to demonstrate angiostatin activity
281
(Claesson-Welsh et al., 1998; Funaki et al., 1998; Nelson et al., 2000). This is shown in the migration/scratch assay (Fig. 1) and during tube formation, as a dose response (Fig. 2). We propose that the differences in dose level and duration in various laboratories may both relate to differences in pHe, which can vary as a function of time in culture.
Fig. 2. Dose response to angiostatin at normal and low pHe in the tube formation assay. Angiostatin was evaluated at a range of 0 – 20 mg/ml at pHe 7.3 and 6.7, as described in Materials and Methods. Micrographs of tube formation show tube network in bas relief in controls without angiostatin. With increasing amount of angiostatin, no significant change is observed at pHe 7.3, whereas the tube network is broken up at pH 6.7 with as little as 10 mg/ml angiostatin.
282
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
Fig. 3. Cells were plated on RGD fibronectin or Matrigel, and the histochemistry was performed to evaluate distribution of a-subunit of ATP synthase. Cells were either permeabilized or not permeabilized prior to immunofluorescence. Surface staining (unpermeabilized cells) for a-subunit of ATP synthase is low or nonexistent on cells plated on RGD or fibronectin, but cell surface labeling is strong on Matrigel (left panels). When cells were permeabilized, positive staining for a-subunit of ATP synthase was observed in all cells on the three matrices, typical of mitochondrial localization (right hand panels).
In the migration/scratch assay (Fig. 1), a 2-mm scratch was made through a confluent HUVEC monolayer. The effect of angiostatin on the cell migration and proliferation
needed to close the scratch was observed with and without extracellular acidification. Angiostatin had no effect on scratch diameter in plates which were kept at pHe 7.3
Fig. 4. The detection of the a-subunit of ATP synthase on HUVEC following 30 mg/ml angiostatin pretreatment. The blocking and redistribution of a-subunit of ATP synthase are seen with low pH incubation, and blocking of receptor by preincubation with angiostatin.
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
(50% closure without angiostatin and 55% with 30 mg/ml angiostatin). At pHe 6.7, there was a 75% closure without angiostatin, and 50% in the presence of 30 mg/ml angiostatin (data not shown). These data clearly suggested that there is a decrease in the migratory ability of the EC. Plates of cells used in the migration/scratch assay were loaded with BCECF for measurements of intracellular pH. Marked changes in pHi were observed with acidification and angiostatin, as shown in Fig. 1. At normal pHe, angiostatin alone had no effect on pHi. Acidification of the medium by 0.6 U in pH in the absence of angiostatin resulted in a decrease in pHi by about 0.3 U. This degree of pH homeostasis is extremely typical of most cell lines and was seen on cells both from the scraped edge and within the monolayer. In the presence of angiostatin, the same acidification produced a decrease in pHe of 0.5– 0.6 U, suggesting some loss of control over intracellular pH regulation. Plates that were wounded were also incubated with acridine orange to measure cell death (apoptosis) (Erenpreisa et al., 1997; Lucas et al., 1998; Whitacre and Berger, 1997). In Fig. 1 (bottom panel), bright cells were usually rounded and potentially dead. There was a slight (5– 10%) increase in the number of bright cells at a scratch edge as a result of lowering pHe from 7.3 to 6.7. The ef-fect of angiostatin at pHe 7.3 was negligible. The combination of low pHe and angiostatin produced an increase (up to a level of 30 – 40%) in the number of hyperchromatic (bright, potentially apoptotic) cells at or near the scratch edge. In order to test the effect of pHe on EC differentiation, HUVEC were plated on polymerized Matrigel (tumor basement membrane proteins) at either pHe 7.3 or 6.7. It has previously been shown that angiostatin inhibits differentiation (EC tube formation) at a concentration (IC50) of 100 mg/ml for 16 h (Nelson et al., 2000). Pretreated HUVEC were plated with or without angiostatin at either a pHe 7.3 or 6.7 for 20 h to allow tube formation. Tubes were then stained, images were converted to relief images, and area occupied by tubes was quantitated, resulting in the tube areas shown. Fig. 2, a dose – response experiment, showed that tube formation was noticeably inhibited by 20 mg/ml angiostatin when extracellular pH was acidified to pH 6.7, whereas the drug had no effect at pH 7.3. The IC50 for inhibition of tube formation by angiostatin at pHe 7.3 was 80 mg/ml, whereas at pHe 6.7 it was reduced to 20 mg/ml. Tube formation was affected approximately threefold by 20 mg/ml angiostatin at low pHe. High dose levels overcame the need for low pHe, so that no pH effect on tube formation was observed when 100 mg/ml angiostatin was used (data not shown). Parallel experiments to measure pHi on the HUVEC produced effects that were in agreement with the findings in the other systems. Angiostatin at 30 mg/ml had no effect on pHi when pHe was 7.3. However, pHi decreased by approximately 0.3 U when pHe was 6.7 (Wahl et al., 2002). Fig. 3 shows the effect of substrate on the cell surface staining of the a-subunit of ATP synthase. This enzyme
283
usually is found in mitochondria, as observed in the righthand panels that shows cells permeabilized with detergent or ethanol. The left-hand panels show positive staining for ATP synthase on the cells’ surface in the presence of a Matrigel substratum, but not in the presence of RGD peptides, which simulate the integrin-clustering amino acid sequence of fibronectin, or in the presence of native human fibronectin itself. Matrigel is a mixture of basement membrane and growth factor proteins typically found in the Englebreth –Holm –Swarm (EHS) tumor stroma, and therefore, the experiments show a dependence on receptor expression of tumor-like microenvironmental stimuli. In Fig. 4, cells plated on dilute Matrigel substratum as compared at pHe 7.3 and 6.7. The staining is different at low pH, and in fact is more punctate. This distribution is in line with possible distribution of FAK plaques, and may therefore increase the likelihood of perturbation of pH regulation via interaction of angiostatin’s receptor with FAK.
4. Discussion In the interest of obtaining data that could be compared to work done by others in vivo, the effects of angiostatin at different pHe and on pHi were measured in EC, which retained their attachment to tumor stromal proteins. We have shown that angiostatin lowered pHi in EC under conditions of low pHe, resulting in a decrease in cell migration and an increase in cell death. This corroborates work showing that proangiogenic vessel growth requires matrix for the EC migration and proliferation, which leads to neovascularization of the tumor mass (Auersperg et al., 1987; Canfield et al., 1986). Low pHe alone (without angiostatin) has also been shown to have an effect on endothelial vessel sprouting in two angiogenesis model systems. In one (Burbridge et al., 1999), a delay was measured in three-dimensional sprout outgrowth at pHe 6.9 vs. 7.4, which could be overcome by exogenous growth factors. We have confirmed these observations in our aortic sprout model; in addition, we have shown that angiostatin takes advantage of this situation by further lowering the pHi, resulting in EC cell death. This is an important feature that helps to understand angiostatin’s efficacy and specificity in vivo in a low pHe tumor environment. Another study (D’Arcangelo et al., 1999) demonstrated that low pH reduced apoptotic death; they also concluded that low pHe inhibited proangiogenic behavior in bovine EC. Regulation of pHi has also been shown to be dependent on cell attachment in fibroblasts (Frisch et al., 1996; Grinstein et al., 1993; Plopper et al., 1995; Schwartz et al., 1991) and in other cells that require substrate attachment for growth (Schwartz, 1993; Wahl et al., 1997). Folkman’s laboratory has provided evidence that angiostatin causes an aberrant activation of focal adhesion kinase (FAK) plaques in EC (Claesson-Welsh et al., 1998). It has
284
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285
been shown (Plopper et al., 1995) that adhesion-dependent survival is regulated by FAK, and FAK expression has been correlated positively with motility in melanoma cells (Akasaka et al., 1995). FAK plaques are organizing centers that mediate effects on cytoskeleton and other intracellular events, some of which appear to involve alteration in proton transport (Plopper et al., 1995). In fibroblasts, proton transport has been implicated in cell migration as well (Grinstein et al., 1993). If angiostatin-dependent aberrant activation of FAK plaques in EC were to compromise the ability of transporters to regulate intracellular pH, this could be part of the mechanism by which angiostatin kills EC. This mechanism would be inherently sensitive to low extracellular pH because low pHe activates two of the major proton transporters, the sodium – proton exchanger isoform 1 (NHE-1) (Counillon and Pouyssegur, 2000) and the H + -linked monocarboxylate transporters (MCT) (Halestrap and Price, 1999). The relative activities of these transporter families in EC have not yet been characterized. Evidence (Moser et al., 1999, 2001) has been provided that the a-subunit of ATP synthase or a homologue of it can be detected on the extracellular surface of the plasma membrane and that it binds to angiostatin. The complete ATP synthase protein is composed of many subunits in a large, complex structure found in mitochondria, where it functions in proton transport. As was originally pointed out by Gillies (1999), the proton gradient required to synthesize extracellular ATP on the cell surface would be contrary to the inward direction of the gradient in acidic parts of a tumor (Gerweck et al., 1991; Wahl et al., 1997). It is possible that the a-subunit of ATP synthase is interacting with another proton transporter in a novel way, or alternatively, that all or part of the ATP synthase complex is functioning as an ATPase in this location. The discovery that moderate doses of angiostatin may be far more effective against EC in an acidic extracellular milieu may help explain why angiostatin has been found to have different effects on physiological and pathological angiogenesis (Folkman, 1997). Tumors are often distinguished from most of the nonpathological examples of angiogenesis by the relative hypoxia and acidosis that frequently characterize the tumor extracellular microenvironment. This is because acidosis is often a corollary of hypoxia and reliance on glycolytic metabolism, together with poor, irregular vascularity, impedes removal of catabolites, including lactic and carbonic acids (Yamagata and Tannock, 1996). The data in this report focus on the effects of reduced pHe on angiostatin activity. The present study indicates that pH should be accorded more attention as a parameter in in vitro assays. For example, if angiostatin requires several days in culture to produce an effect, it will be important to monitor medium pH carefully. The development of medium acidity could easily produce an apparent dependence on elapsed time that was, in reality, a dependence on a gradually decreasing pHe.
In conclusion, these studies describe the critical interplay between the extracellular matrix and pH microenvironment of the tumor bed and antiangiogenic drugs during neovascularization. These parameters should be evaluated for other antitumor drugs currently under development.
Acknowledgments We thank Michael Zahaczewsky for expert technical assistance. The rh-angiostatin was a generous gift from Entremed. This work has been supported by NIH grant P01 CA56690 (M.L.W.) funds from Merck (West Point, PA) (D.S.G.) and Translational Grant Funding from Entremed (D.S.G.).
References Akasaka, T., vanLeeuwen, R.L., Yoshinaga, I.G., Mihm, M.C., Byers, H.R., 1995. Focal adhesion kinase (p125FAK) expression correlates with motility of human melanoma cell lines. J. Invest. Dermatol. 105, 104 – 108. Auerbach, R., Auerbach, W., Polakowski, I., 1991. Assays for angiogenesis: a review. Pharmacol. Ther. 51 (1), 1 – 11. Auersperg, N., Pawson, T., Worth, A., Weinmaster, G., 1987. Modifications of tumor histology by point mutations in the v-fps oncogene: possible role of extracellular matrix. Cancer Res. 47 (23), 6341 – 6348. Burbridge, M.F., West, D.C., Atassi, G., Tucker, G.C., 1999. The effect of extracellular pH on angiogenesis in vitro. Angiogenesis 3, 281 – 288. Canfield, A.E., Schor, A.M., Schor, S.L., Grant, M.E., 1986. The biosynthesis of extracellular-matrix components by bovine retinal endothelial cells displaying distinctive morphological phenotypes. Biochem. J. 235 (2), 375 – 383. Claesson-Welsh, L., Welsh, M., Ito, N., Ananda-Apte, B., Soker, S., Zetter, B., O’Reilly, M., Folkman, J., 1998. Angiostatin induces endothelial cell apoptosis and activation of focal adhesion kinase independently of the integrin-binding motif RGD. Proc. Natl. Acad. Sci. USA 95, 5579 – 5583. Counillon, L., Pouyssegur, J., 2000. The expanding family of eucaryotic Na + /H + exchangers. J. Biol. Chem. 275, 1 – 4. D’Arcangelo, D., Facchiano, F., Barlucchi, L.M., Melillo, G., Illi, B., Testolin, L., Gaetano, C., Capogrossi, M.C., 1999. Acidosis inhibits endothelial cell apoptosis and function and induces basic fibroblast growth factor and vascular endothelial growth factor expression. Circ. Res. 86, 312 – 318. Erenpreisa, J., Freivalds, T., Roach, H., Alston, R., 1997. Apoptotic cell nuclei favour aggregation and fluorescence quenching of DNA dyes. Histochem. Cell Biol. 108, 67 – 75. Folkman, J., 1985. Tumor angiogenesis. Adv. Cancer Res. 43, 175 – 203. Folkman, J., 1990. What is the evidence that tumors are angiogenesis dependent? J. Natl. Cancer Inst. 82, 4 – 6. Folkman, J., 1997. Regulation of Angiogenesis. Birkaeuser Verlag, Basel, pp. 1 – 8. Folkman, J., Ingber, D., 1992. Inhibition of angiogenesis. Semin. Cancer Biol. 3, 89 – 96. Folkman, J., Klagsbrun, M., 1987. Angiogenic factors. Science 235, 442 – 447. Folkman, J., Shing, Y., 1992. Angiogenesis. J. Biol. Chem. 267, 10931 – 10934. Frisch, S.M., Vuori, K., Ruoslahti, E., Chan-Hui, P.-Y., 1996. Control of adhesion-dependent cell survival by focal adhesion kinase. J. Cell Biol. 134, 793 – 799. Funaki, N.O., Tanaka, J., Imamura, M., 1998. Quantitative analysis of
M.L. Wahl, D.S. Grant / General Pharmacology 35 (2002) 277–285 alpha-fetoprotein mRNA in circulating peripheral blood of patients with hepatocellular and a-fetoprotein-producing gastric carcinomas. Life Sci. 62, 1973 – 1984. Gerweck, L.E., Rhee, J.G., Koutcher, J.A., Song, C.W., Urano, M., 1991. Regulation of pH in murine tumor and muscle. Radiat. Res. 126, 206 – 209. Gillies, R.J., 1999. Angiostatin’s partners. Science 288, 434. Grant, D.S., Kleinman, H.K., Martin, G.R., 1989. The role of basement membrane in vascular development. Ann. N.Y. Acad. Sci. 588, 61 – 72. Grant, D.S., Kinsella, J.L., Kibbey, M.C., LaFlamme, S., Burbelo, P.D., Goldstein, A.L., Kleinman, H.K., 1995. Matrigel induces thymosin b4 gene in differentiating endothelial cells. J. Cell Sci. 108. Griffioen, A.W., Molema, G., 2000. Angiogenesis: potentials for pharmacologic intervention in the treatment of cancer, cardiovascular diseases and chronic inflammation. Pharmacol. Rev. 52, 237 – 268. Grinstein, S., Woodside, M., Waddell, T.K., Downey, G.P., Orlowski, J., Pouyssegur, J., Wong, D.C.P., Foskett, J.K., 1993. Focal localization of the NHE-1 isoform of the Na + /H + antiport: assessment of effect on intracellular pH. EMBO J. 12 (13), 5209 – 5218. Grynkiewicz, G., Poenie, M., Tsien, R.Y., 1985. A new generation of Ca2 + indicators with greatly improved fluorescence properties. J. Biol. Chem. 260, 3440 – 3450. Halestrap, A.P., Price, N.T., 1999. The proton-linked monocarboxylate transporter (MCT) family: structure, function, and regulation. Biochem. J. 343, 281 – 299. Haralabopoulos, G.C., Grant, D.S., Kleinman, H.K., Lelkes, P.I., Papaioannou, S.P., Maragoudakis, M.E., 1994. Inhibitors of basement membrane collagen synthesis prevent endothelial cell alignment in Matrigel in vitro and angiogenesis in vivo. Lab. Invest. 71 (4), 575 – 582. Lucas, R., Holmgren, L., Garcia, I., Jimenez, B., Mandriota, S.J., Borlat, F., Sim, B.K., Wu, Z., Grau, G.E., Shing, Y., Soff, G.A., Bouck, N., Pepper, M.S., 1998. Multiple forms of angiostatin induce apoptosis in endothelial cells. Blood 92 (12), 4730 – 4741. Macchiarini, P., Fontanini, G., Hardin, M.J., Squartini, F., Angeletti, C.A., 1992. Relation of neovascularisation to metastasis of non-small-cell lung cancer. Lancet 340, 145 – 146. Moser, T.L., Stack, M.S., Asplin, I., Enghild, J.J., Hojrup, P., Everitt, L., Hubchak, S., Schnaper, H.W., Pizzo, S.V., 1999. Angiostatin binds ATP synthase on the surface of human endothelial cells. Proc. Natl. Acad. Sci. USA 96, 2811 – 2816. Moser, T.L., Kinan, D.J., Ashley, T.A., Roy, J.A., Goodman, M.D., Misro, U.K., Cheek, D.J., Pizzo, S., 2001. Endothelial cell surface F1 – Fo ATP synthase is active in ATP synthesis and is inhibited by angiostatin. Proc. Natl. Acad. Sci. 98, 6656 – 6661. Nelson, B.J., Vu, H.A., Hanson, A.D., Fortier, A.H., Sim, B.K.L., 2000. Recombinant human angiostatin protein inhibits both migration and tube formation of human endothelial cells. Proc. Am. Assoc. Cancer Res. 41, 485. O’Reilly, M.S., Holmgren, L., Shing, Y., Chen, C., Rosenthal, R.A., Lane, W.S., Cao, Y., Sage, E.H., Folkman, J., 1994. Angiostatin: a novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell 79, 315 – 328. Owen, C.S., 1996. Dependence of proton generation on aerobic or anaerobic metabolism and implications for tumor pH. Int. J. Hyperthermia 12, 495 – 499.
285
Owen, C.S., Wahl, M.L., Leeper, D.B., Perry, H.D., Bobyock, S.B., Russel, M., Woodward, W., 1995. Accurate whole spectrum measurements of intracellular pH and [Na + ]. J. Fluoresc. 5, 329 – 335. Owen, C.S., Wahl, M.L., Pooler, P.M., Coss, R.A., Leeper, D.B., 1998. Temporal association between alterations in proton extrusion and low pH adaptation. Int. J. Hyperthermia 14, 227 – 232. Plopper, G.E., McNamee, H.P., Dike, L.E., Bojanowski, K., Ingber, D.E., 1995. Convergence of integrin and growth factor receptor signaling pathways within the focal adhesion complex. Mol. Biol. Cell 6, 1349 – 1365. Rak, J., Filmus, J., Kerbel, R.S., 1996. Reciprocal paracrine interactions between tumour cells and endothelial cells ‘angiogenesis progression’ hypothesis. Eur. J. Cancer 32A, 2438 – 2450. Schwartz, M.A., 1993. Signaling by integrins: implications for tumorigenesis. Cancer Res. 53, 1503 – 1506. Schwartz, M.A., Lechene, C., Ingber, D.E., 1991. Insoluble fibronectin activates the Na + /H + antiporter by clustering and immobilizing integrin a v b 1, independent of cell shape. Proc. Nat. Acad. Sci. USA 88, 7849 – 7853. Tannock, I., Rotin, D., 1989. Acid pH in tumors and its potential for therapeutic exploitation. Cancer Res. 49, 4373 – 4384. Wahl, M.L., Coss, R.A., Bobyock, S.B., Leeper, D.B., Owen, C.S., 1996. Thermotolerance and intracellular pH in two Chinese hamster cell lines adapted to growth at low pH. J. Cell. Physiol. 166, 438 – 445. Wahl, M.L., Bobyock, S.B., Leeper, D.B., Owen, C.S., 1997. Effects of 42C hyperthermia on intracellular pH in ovarian carcinoma cells during acute acidification or chronic exposure to low extracellular pH. Int. J. Radiat. Oncol., Biol., Phys. 39, 205 – 212. Wahl, M.L., Pooler, P.M., Briand, P., Leeper, D.B., Owen, C.S., 2000. Intracellular pH regulation in a nonmalignant and a derived malignant human breast cell line. J. Cell. Physiol. 183, 373 – 380. Wahl, M., Owen, C., Grant, D., 2002. Low dose angiostatin induces intracellular acidosis and anoikis cell death in endothelial cells at tumor-like low pH. Endothelium (in press). Weidner, N., Semple, J.P., Welch, W.R., Folkman, J., 1991. Tumor angiogenesis and metastasis-correlation in invasive breast carcinoma. N. Engl. J. Med. 324, 1 – 8. Weidner, N., Carroll, P.R., Flax, J., Blumenfeld, W., Folkman, J., 1993. Tumor angiogenesis correlates with metastasis in invasive prostate carcinoma. Am. J. Pathol. 143, 409 – 410. Whitacre, C.M., Berger, N.A., 1997. Factors affecting topotecan-induced programmed cell death, adhesion protects cells from apoptosis and impairs cleavage of poly(ADP-ribose)polymerase. Cancer Res. 57, 2157 – 2163. Wike-Hooley, J., Haveman, J., Reinhold, H.S., 1984. The relevance of tumor pH in the treatment of malignant disease. Radiother. Oncol. 2, 342 – 366. Wu, Z., O’Reilly, M.S., Folkman, J., Shing, Y., 1997. Suppression of tumor growth with recombinant murine angiostatin. Biochem. Biophys. Res. Commun. 236 (3), 651 – 654. Yamagata, M., Tannock, I., 1996. The chronic administration of drugs that inhibit the regulation of intracellular pH: in vitro and anti-tumour effects. Br. J. Cancer 73, 1328 – 1334.