Collateral Expression of Proangiogenic and Tumorigenic Properties in Intestinal Epithelial Cell Variants Selected for Resistance to Anoikis

Abstract

Although in vitro anchorage-independent growth is widely used as a marker of cell transformation, the biological implications of this trait are poorly understood. We previously demonstrated that enforced anchorage-independent growth of a nontumorigenic, immortalized epithelial cell line (IEC-18) in multicellular spheroid culture results in massive apoptotic cell death. This death process, termed anoikis, is prevented by expression of transforming oncogenes, which also confer tumorigenic competence. This study examines whether acquisition of an anoikis-resistant phenotype is causally related to the tumorigenic capacity of transformed epithelial cells. Parental IEC-18 cells were subjected to 10 cycles of selection for survival in speroid culture. Unlike parental cells, the resulting anoikis-resistant variants (AR1.10 and AR2.10) formed relatively large tumors in nude mice. Both anoikis-resistant sub-lines displayed upregulated expression of vascular endothelial growth factor (VEGF), a potent angiogenesis stimulator. VEGF121 overexpression alone did not induce tumorigenic conversion of parental IEC-18 cells, which remained highly susceptible to anoikis. We postulate that both anoikis-resistance and angiogenic-competence contribute to tumor formation. Development of anoikis-resistance can be then viewed as a precondition for expression of the tumorigenic phenotype. Our results suggest that even when angiogenesis is not a rate limiting factor (e.g. in vitro) the selective pressures of solid tumor-like, 3-dimensional growth conditions favoring anoikis resistance result in collateral induction of a proangiogenic phenotype.

Introduction

Anchorage-independent growth, discovered almost 35 years ago, remains a widely accepted in vitro hallmark and surrogate marker of malignant transformation and tumorigenesis [1,2]. Indeed, the distinct ability of tumor cells to grow in semisolid media, suspension or multicellular spheroid culture has been frequently linked to aberrant cellular mitogenesis, i.e., anchorage-independent growth [3], and more recently, to abnormal control of physiologic cell death processes, i.e., anchorage independent survival [4,5]. The latter is of particular importance for epithelial cells, for which attachment to basement membranes and proper uni- or multilayered cytoarchitecture appear to be an important cell survival mechanism. Withdrawal or disruption of such a survival signal, e.g., by placing epithelial cells in 3-dimensional culture, is associated with precipitous induction of what is currently known as adhesion regulated programmed cell death (ARPCD) [6] or death of homelessness, for which the term anoikis has been coined recently [7]. Anoikis is believed to act as an essential safeguard mechanism that helps maintain proper cell number, geographic architecture, and position in epithelial organs. It is not entirely clear how this form of apoptosis is executed and controlled. Some recent molecular analyses have implicated activation of c-jun terminal kinase (JNK1), MEKK, and caspase 8 [8]. However, some of these results have also been disputed [9].

Anoikis can be aborted, or mitigated, by overexpression of proto-oncogenic proteins, such as Bcl-2 or v-Src, or by mutant oncogenes, such as H-ras [6]. The latter type of cellular transformation is now known to exert this survival function through activation of phosphoinositol 3′OH kinase (PI3K) and protein kinase B (PKB/Akt) [11] a pathway that can change the phosphorylation status of some apoptosis-effector genes, such as BAD [12] or expression levels of others, such as BAK [13]. It is likely that some of these genetic and signaling changes may be relevant for anchorage-independent growth and in vivo survival of cells derived from spontaneous tumors, as well as for the ability of single cancer cells or small aggregates to survive in the blood-stream or in ectopic organ sites after extravasation. However poorly understood, the molecular defects underlying anchorage independence in general, and resistance to anoikis in particular, appear to be quite heterogenous, even within tumors of the same origin. For example, somatic gene knockout studies have demonstrated that inactivation of the mutant K-ras allele reproducibly leads to a marked decrease in soft agar colony formation by 2 independent colorectal cancer cell lines [14]. Although this finding suggests that oncogenic ras may be essential for abnormal 3-dimensional growth capacity [14], it is also true that K-ras mutations are detected in only about 50% of colorectal carcinoma cases and in a proportion of the respective cancer cell lines in culture, whereas virtually all such cell lines are capable of some degree of anchorage independent growth and survival. It is therefore possible that resistance to anoikis may represent a generic phenotype that is strongly selected for, regardless of whether tumor progression is driven by one particular molecular pathway (e.g., activating ras mutations) or another.

Despite the strong correlation between the resistance of transformed cells to anoikis and their ability to form tumors in vivo, direct evidence for a causal relationship between the two characteristics has not been firmly established. This is because available cancer cells usually express a multitude of functional abnormalities such as resistance to various apoptosis-inducing stimuli, deregulated mitogenic activity, reduced growth requirements, and expression of proangiogenic properties, all of which are difficult to separate functionally from each other. Indeed, different aspects of the transformed phenotype may become rate limiting at different stages of tumor growth and progression. For example, anoikis would probably severely limit early stages of 3-dimensional expansion of tumor cells because they would be forced to lose their contact with the basement membrane, whereas angiogenesis would be required at later times to allow tumor growth beyond the microscopic size of 1 to 2 mm in diameter [15,16].

In order to elucidate some of these interrelationships in a more direct manner, we designed an in vitro selection protocol, which theoretically would specifically favor the acquisition of resistance of epithelial cells to anoikis. For this purpose, we used an immortalized (but nontumorigenic) rat intestinal epithelial cell line, IEC-18, which displays some intrinsic heterogeneity as to the kinetics of cell death induced under anchorage-independent growth conditions in spheroid culture. We showed that sequential exposure of IEC-18 cells to such 3-dimensional culture conditions spontaneously leads to the derivation of rare cellular variants with a markedly reduced susceptibility to anoikis. This exposure was also sufficient for acquisition of tumorigenic competence by these cells in vivo and was associated with the collateral and spontaneous upregulation of VEGF, a potent stimulator of angiogenesis. These observations raise the possibility that during epithelial tumorigenesis, selective pressures exerted by 3-dimensional growth conditions may favor survival pathways, the activation of which is coupled with the simultaneous ability to induce tumor angiogenesis.

Materials and Methods

Cells and Culture Conditions

IEC-18 cells were obtained and cultured as described previously [6]. Briefly, all cell lines were maintained in monolayer culture in growth medium containing αMEM base supplemented with 5% fetal bovine serum (Gibco BRL, Gaithersburg, MD), 4 mmol/L l-glutamine, 20 mmol/L glucose, 10 µg/mL insulin, (Sigma Chemical Co, St. Louis, MO). VEGF₁₂₁ overexpressing cell lines were generated by transfection of IEC-18 cells with the pcDNAI expression vector containing the respective cDNA coding sequence by using a lipofectin reagent (Gibco BRL), as indicated by the manufacturer. Transfectants 18V1 and 18V7, expressing various amounts of VEGF₁₂₁, were selected in 400 µg/mL of G418 (Gibco BRL) and periodically exposed to the drug afterward. For selection of cells in spheroid culture, single-cell suspensions were prepared and plated at a concentration of 1–2 x 10⁵ cells per well in 24 well plates (Nunc, Denmark) precoated with a thin layer of 1% Seaplaque agarose (FMC Corp, Rockland, ME). After 4 to 7 days of incubation (longer for later selections) the mostly dead and dying IEC-18 cells contained in the 24 wells were transferred to a 100-mm (Nunc) tissue culture dish and left to recover in the regular growth medium. After the surviving clones were expanded, the cells were plated again in agarose-coated dishes, and the cycle was repeated 9 times.

Cell Survival Assays

The characteristics of apoptotic death of IEC-18 cells in spheroid culture, its abrogation in oncogenic ras transfectants (e.g., the ras-3 and ras-4 clones of IEC-18), have been described in detail [6]. For the purpose of the present study we have used a commercially available version of the MTS assay (Promega, Madison, WI) to evaluate cumulative cell survival. Briefly, the cells were plated in spheroid culture at 10⁴ per well of the 96 well plate precoated with 2% poly(2-hydroxyethyl methacrylate) (Sigma). At the time points indicated, 100 µL of a mixture containing MTS, phenazine methosulfate (Promega, Madison, WI), reagent, and Hank's balanced salt solution (300:15:700) was added to the wells and the absorbance at 490 nm recorded after 1 and 3.5 hours. The average absorbance from multiple wells containing spheroids was normalized to absorbance of corresponding monolayer cultures or to day zero in time course experiments. Similar conditions were used to estimate the rate of apoptotic cell death by using Cell Death Elisa Plus assay (Boehringer Mannheim Canada, Laval, Quebec), which quantitates nucleosomes released into the cytoplasm of dying cells. In this case cells were collected from multiple spheroid-containing wells, lysed and incubated with a mixture of biotinylated antihistone antibody and peroxidase-conjugated anti-DNA antibody, both of which bind to histon-DNA complexes and initiate color reaction in the presence of the ABTS 2,2′-Azino-di[3-ethylbenzthiazolin-sulfonat] substrate (Boehringer Mannheim Canada, Laval, Quebec). The OD reading at 405 nm was corrected according to negative control and expressed as an enrichment factor, according to the manufacturer's instruction. In spheroid culture the intensity of this reaction peaked between 24 and 72 hours, with some variation between the cell lines and under the influence of culture conditions (e.g., confluence before plating could also influence the results). For H-ras-transformed cells and monolayer cultures the reaction only becomes, positive at later time points (after day 4) as a function of cell density and depletion of the media.

Tumorigenicity Assay

Single-cell suspensions of all cell lines indicated were prepared from monolayer cultures and 2–5 x 10⁶ cells were injected subcutaneously into BALB/c nu/nu mice (Taconic, Germantown, NY). Tumors were measured weekly and their volume calculated according to the standard formula: (a² x b)/2, where a is the width and b is the length of the horizontal tumor perimeter. The animals were sacrificed before any visible signs of discomfort were present according to the guidelines of the Animal Care Commitee at Sunnybrook Health Science Centre.

Detection of VEGF by Northern Analysis

Total cellular RNA was extracted by using Trizol reagent according to the manufacturer's instructions (Gibco BRL). The samples were resolved on 1% agarose gel, and the equal loading was verified by staining with ethidium bromide (EtBr). RNA was then transferred to Hybond N(+) membrane (Amersham Canada, Oakville, Ontario) which was hybridized with VEGF, cDNA, or GAPDH cDNA probes, as described previously [10]. All cell lines expressing VEGF mRNA were also positive for VEGF protein, which was secreted into their conditioned medium and detected by antimouse VEGF antibody (for ras-3, ras-4, AR1.10, AR2.10 cells) and antihuman VEGF antibody (18V1 and 18V7) with Quantakine Elisa Kits (R&D Systems Inc, Minneapolis, MN).

Data Analysis

All tests were repeated at least 2 to 3 times with similar results. The numerical data were presented as average values, with error bars representing standard deviations.

Results

Derivation of Spontaneous Anoikis-Resistant Variants

IEC-18 cells were originally established from embryonic rat intestinal (crypt) epithelium. They are known to be highly sensitive to anoikis [6]. Thus, the cells do not form colonies in soft agar [17] and undergo massive apoptosis when forced to grow in a tumor-like fashion as large (1 mm), multicellular spheroids [6] on nonadhesive (agarose-coated) dishes. During this death process, a proportion of the cells initially form a tight aggregate while remaining cells disintegrate outside the spheroid, forming a halo of debris (Figure 1B). Because we have previously observed some degree of heterogeneity between different IEC-18 derived clones with respect to their sensitivity to cell death in spheroid culture, we reasoned that cryptic subpopulations that are relatively resistant to anoikis might exist, and that such cells may be selected (and/or further induced) by long-term growth as spheroids. To test this hypothesis, IEC-18 cells were subjected to repeated cycles of selection for survival and growth in spheroid culture (Figure 1A). After each 4- to 7-day incubation period on agarose-coated dishes, the remnants of IEC-18 spheroids were collected from several wells, plated in monolayer culture, and left to recover for several weeks. The few clones that survived were expanded and plated in another round of spheroid culture. After 10 cycles of such selection we observed that the resulting cellular variants obtained were capable of forming larger spheroids (Figure 1B) composed of cells that indeed displayed a reduced rate of cell death, albeit to a lesser extent than, for example, sublines of IEC-18 overexpressing a mutant H-ras oncogene (Figure 1C). Two such anoikis-resistant variants, designated AR1.10 and AR2.10, each derived in an independent series of selections, were found to be stable in that their survival properties in spheroid culture were retained even after several weeks of prior growth as standard monolayer cultures.

Figure 1.

Derivation of anoikis-resistant variants of IEC-18 cells. A. Experimental protocol (see text). B. Morphology of spheroids composed of IEC-18, AR1.10, AR2.10 or ras-3 cells after 48 hours in 3-dimensional culture. (Original magnification x 100) C. Cumulative survival curves (MTS assay) of anoikis resistant variants of IEC-18 cells derived by spheroid selection protocol (AR1.10, AR2.10) or transformation with mutant H-ras oncogene (ras-4).

Selection for Anoikis Resistance in Vitro Cosegregates with a Tumorigenic Phenotype in Vivo

Our prior experience with the parental IEC-18 cells have shown that these cells are absolutely nontumorigenic after subcutaneous injection into nude mice, regardless of cell number in the inoculum (1–5 x 10⁶) or the duration of the follow up time (up to 12 months). Even the presence of Matrigel in the inoculum, which has been shown to significantly increase tumorigenic competence of immortalized fibroblasts [18], was unable to render the IEC-18 cells tumorigenic (unpublished observations). In contrast, as we have shown previously, expression of the mutant H-ras oncogene had a pleiotropic transforming effect on IEC-18 cells manifested by exuberant tumorigenic properties (compare respective tumor takes and growth rates in Figure 2A) that were attributed to the combination of effects this oncogene has on the ability of the cells to survive, proliferate, and induce an effective angiogenic response under 3-dimensional growth conditions in vivo [6,10,19]. However, the relative contribution of each of these changes has not been formally examined. In this regard, K-ras-dependent upregulation of VEGF and associated angiogenic capacity appear to be necessary, but not sufficient, for expression of a tumorigenic phenotype by human colorectal cancer cells [20]. Furthermore, the inability of immortalized epithelial cells to form tumors in vivo in some cases may be overcome (at least to a small degree) solely by increasing their angiogenic competence, e.g., by transfection-mediated overexpression of VEGF [21]. Such a manipulation was clearly ineffective when applied to IEC-18 cells. Thus, transfected IEC-18 cells overexpressing human VEGF₁₂₁ (18V7) at levels comparable to those observed in the aforementioned H-ras transfectants (Figure 2B) completely failed to form tumors, even after injection of as many as 5 x 10⁶ cells (Figure 2A). This is consistent with the notion that such VEGF transfectants would in theory remain susceptible to anoikis and thus die regardless of and prior to generation of new blood vessels. Indeed, growth of the VEGF overexpressing 18V7 IEC-18 cells in spheroid culture resulted in a similar degree of apoptosis to the parental IEC-18 cell line as measured by nucleosomal release to the cytoplasm (apoptotic enrichment factor) and low cumulative survival capacity (MTS) assay (Figure 3).

Figure 2.

Tumor-forming capacity of IEC-18 variant cell lines. IEC-18—parental cells; 18V1 and 18V7—IEC-18 cells transfected with the human VEGF₁₂₁; AR1.10 and AR2.10—anoikis-resistant variants selected for survival in spheroid culture; ras-3 and ras-4—IEC-18 cells transfected with mutant H-ras oncogene. A. Tumor growth kinetics in nude mice (AR1.10, AR2.10, ras-3, ras-4—2 x 10⁶ per inoculum; remaining cells—5 x 10⁶ per inoculum). Three independent series of injections were carried out with both AR1.10 and AR2.10 cell lines as well as appropriate IEC-18 controls. In each case the results were highly reproducible in that no tumors resulted from IEC-18 injections while the variants were consistently tumorigenic. Mice were monitored for at least 4 months, and even up to 12 months (e.g., the IEC-18 injections). The data pooled from two such experiments are presented. B. Expression of hVEGF₁₂₁ by transfected IEC-18 cell lines, total RNA Northern blot.

Figure 3.

Survival properties of IEC-18 derived tumorigenic variants in spheroid culture. Cumulative cell survival after 3 days in spheroid culture (MTS assay). (Inset) Rate of nucleosomal release (apoptotic enrichment factor) as measured after 48 hours in spheroid culture by Cell Death Elisa assay (see text).

Surprisingly, selection for anoikis resistance alone resulted in acquisition of a tumorigenic phenotype by both AR1.10 and AR2.10 cell lines (see Figure 2). Injection of 2 x 10⁶ cells of each variant resulted in formation of slow-growing tumors in 9 of 10 and 5 of 8 mice, respectively. Despite their slow growth, the tumors eventually reached considerable sizes, e.g., 500 to 2000 mm³, within 4 to 5 months. Cell lines reisolated from some of these tumors (AR2T1 and AR2T2) were even more resistant to anoikis in spheroid culture and morphologically indistinguishable from their parental AR2.10 cell line (Figure 3 and data not shown).

Variants Selected for Resistance to Anoikis Constitutively Upregulate VEGF

The observation that injection of AR1.10 and AR2.10 cells in nude mice leads to formation of solid tumors that expand beyond 1 to 2 mm in diameter suggested that, unlike the parental IEC-18 cells, the anoikis-resistant variants are able to mount an angiogenic response in vivo. This suggests the possibility of a functional link between (resistance to) anoikis and (induction of) angiogenesis. To evaluate this possibility, we decided to examine the anoikis-resistant variants for their expression of VEGF. We did so for 3 reasons. First, parental IEC-18 cells are known to be VEGF negative [10]. Second, VEGF is a potent proangiogenic regulator [22]. Third, high levels of VEGF are induced in other transformed and tumorigenic sublines of IEC-18 cells [10]. Indeed, testing for VEGF expression revealed a several-fold constitutive upregulation of the transcript in both anoikis-resistant cell lines tested (Figure 4). This up-regulation was not as pronounced as in the case of mutant ras-transformed IEC-18 cells (ras-3). Nevertheless, the levels of VEGF corresponded well with the respective degrees of both tumorigenicity and anoikis resistance in both variants of IEC-18 cells selected in spheroid culture (see Figure 2 and Figure 3).

Figure 4.

Upregulation of VEGF mRNA expression in anoikis-resistant cell lines selected in spheroid culture. IEC-18—parental cells; AR1.10 and AR2.10—spontaneously derived anoikis-resistant variants of IEC-18; ras-4—IEC-18 cells transfected with mutant H-ras.

Discussion

Two findings are presented in this article. First, an unbiased selection of immortalized epithelial cells for resistance to anoikis in vitro confers a tumorigenic (albeit relatively low-grade) phenotype in vivo, and second, that anoikis resistance and tumor angiogenesis may be functionally coupled and thereby coselected during tumor progression. The first finding could shed new light on the significance of anchorage-independent growth. Thus, the ability to grow as colonies in soft agar remains perhaps the most consistent in vitro correlate of the in vivo tumorigenic phenotype [1,2,23]. By growing cells as large multicellular aggregates on a thin layer of agarose, rather than as relatively inaccessible single cells or very small clonal colonies within soft agar, we were able to select for and recover variant subpopulations having the competence to grow as anchorage-independent multicellular spheroids. To our knowledge this is the first demonstration that intentional selection for ability to grow as anchorage independent entities not only correlates with tumorigenicity but can actually directly contribute to this phenotype.

The second finding, namely, a possible link between anoikis resistance and tumor angiogenesis, is implied by the first finding because progressive tumorigenic growth in vivo cannot proceed in the absence of induction of angiogenesis [24]. In this regard, the ability of the anoikis-resistant variants to produce spontaneously VEGF could obviously contribute to induction of the angiogenic response in vivo, although induction of other angiogenesis effectors as well as loss of endogenous inhibitors could also be involved [25]. These possibilities are currently being studied in more detail.

Our results may be pertinent to the factors that govern the angiogenic switch in developing tumors and to the question of when this switch actually begins to take place. Although there may be instances of a chronic avascular phase responsible for keeping tumors microscopic and dormant, evidence is accumulating that the angiogenic switch can occur at very early stages of tumor progression, i.e., during the premalignant phases of tumor development in some types of cancer [26]. For example, in 3 different transgenic mouse models of tumorigenesis involving formation of pancreatic carcinomas, squamous cell carcinomas, and fibrosarcomas, the angiogenic switch was found to occur in the various precursor lesions before overt tumor formation [26]. Similarly, there is evidence that in human cancers in which the early stage precursor lesions can be identified, angiogenesis can be detected as early as in the horizontal growth phase of primary melanomas [27], in cervical dysplasias [28], and in similar lesions associated with the mammary duct [26]. In these situations, the ability of transformed cells to grow and survive as a multicellular mass rather than as a sheet or monolayer of cells resting on a basement membrane implies some degree of resistance to anoikis, and this, in turn (as suggested by our results), may contribute to ability of the surviving cells to trigger angiogenesis. Clearly, other factors could contribute to this switch as well, e.g., hypoxia [29] and/or additional genetic alterations [25]. This possibility is consistent with the low-grade tumorigenic phenotype of the anoikis-resistant sublines of IEC-18 cells.

Whereas the molecular mechanisms of tumor angiogenesis, and specifically the role of VEGF, are becoming increasingly better defined, it is not clear how the angiogenic phenotype is actually selected for. Such a phenotype per se would be expected to provide a relatively weak selective advantage because new blood vessels may be shared by both angiogenically competent and angiogenically incompetent tumor cell populations [30]. In this regard, it is important to note that genetic lesions implicated in triggering tumor angiogenesis, such as activating mutations of the ras proto-oncogene or inactivating mutations of p53, are also known to render tumor cells resistant to apoptosis-inducing stimuli [6,31]. Hence, it is possible that strong selection pressures favoring the latter property facilitate dominance of tumor cell clones which are at the same time angiogenically competent. The pleiotropic phenotypes of AR1.10 and AR2.10 cell lines seem to suggest such a possibility in the context of anoikis. A corollary to this discussion might be that at the later stages of tumor progression at least some of the genetic changes involved (e.g., p53 mutations in colon cancer) may not only trigger a more aggressive proangiogenic phenotype, but also simultaneously reduce cellular susceptibility to undergo apoptosis under microenvironmental stress conditions (such as hypoxia) [31] and in doing so, double their impact on tumor growth by effectively lowering the relative dependency of tumors cells on the blood supply [32].

The molecular basis of the anoikis-resistant-like phenotype in the AR-IEC-18 sublines is unknown. It is possible that this phenotype is reflective of some degree of cryptic spontaneous transformation that occurred during the long period that IEC-18 cells were maintained in monolayer tissue culture. A considerable number of candidates for transforming genes that could confer such a phenotype have already been identified by previous gene transfection studies. Mutant H- and K-ras have been shown to abrogate anoikis in IEC-18 cells [6,33], as have v-src (our unpublished observation), integrin-linked kinase (ILK) [34] and BCL-2 [6]. With the exception of BCL-2 (our unpublished observation), transfection of these transforming oncogenes was found to result in the simultaneous expression of a tumorigenic and angiogenic phenotype.

Although the molecular pathways responsible for simultaneous induction of anoikis resistance and upregulation of VEGF remain unknown, we believe that a valuable lesson can be drawn from the studies on human counterparts of the IEC-18 system that are currently being developed. This is because the method of selection in 3-dimensional culture used in our experimentation is fairly objective and not dependent on the preconceptions usually associated with transfection of a specific transforming gene into a recipient nontumorigenic cell line. This strategy also has a potential for gene expression cloning and comparative gene expression analysis, both of which can be used for unbiased identification of genes involved in early stages of progression of colorectal cancer, breast cancer, and other epithelial human malignancies derived from cell types naturally susceptible to anoikis. It can be speculated that in colorectal cancer such candidate genes could include the mutant APC tumor suppressor gene and the K-ras oncogene, because both of these genetic changes are frequently found at relatively early stages of disease progression [35]; moreover, both have been linked to regulation of cell survival [6,36].

Finally, growth of the anoikis resistant IEC-18 variant sublines as tumors in vivo further increased their ability to survive in the 3-dimensional culture (Figure 3). This is indicative of the cumulative and quantitative nature of anoikis resistance in which various secondary changes inflicted on the cells during their tumorigenic growth in vivo may have contributed to increasing survival capacity. Indeed, such secondary changes may account for the acute acceleration of tumor growth that was noticeable after first 3 to 4 months after injection of the cells. Nevertheless, despite the large sizes of tumors from which the AR2T1 and AR2T2 cell lines were isolated, their resistance to anoikis was still considerably lower than that of H-ras-transfected cells. This difference may reflect a distinct nature of the molecular defects expressed by both AR2.10 derivatives.

In summary, two critical aspects of the transformed phenotype, namely resistance to anoikis and angiogenic competence, may be functionally coupled, suggesting that selection of cells for competence to grow 3-dimensionally favors certain types of pleiotropic molecular defects rather than simply cell survival in general. Therapeutic targeting of such pleiotropic defects could be highly effective as it could simultaneously suppress or obliterate multiple essential components of the tumorigenic phenotype, such as tumor angiogenesis and the apoptosis (i.e. anoikis) resistant phenotype of tumor cells.

Abbreviations

VEGF

vascular-endothelial growth factor

ARPCD

adhesion regulated programmed cell death

APC

adenomatous polyposis coli

ILK

integuin linked kinase

GAPDH

glutaraldehyde phosphate dehydrogenase

MTS

(3-4,5-dimethylthiazol-z-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H tetrazolium, inner salt