p53 and Angiogenesis
The TP53 gene is the most mutated gene in human cancer and as a consequence has been one of the most extensively studied genes in the human genome. Over half of all human cancers carry direct mutations of the p53 coding region. In addition to sporadic mutations in human cancer, inherited mutations in TP53 cause a genetic predisposition to cancer called Li–Fraumeni syndrome. Individuals with Li–Fraumeni exhibit early onset of a wide variety of cancers including soft-tissue sarcoma, leukemia, osteosarcoma, and tumors of the breast and brain (Li et al., 1988). The TP53 gene is not essential for development but seems to have evolved a primary function to prevent neoplasia in multicellular organisms. The p53 tumor suppressor protein has the structure of a classical transcription factor possessing a central domain with sequence-specific DNA binding activity and an N-terminal acidic region required for transcriptional regulation of target genes.
KeywordsVascular Endothelial Growth Factor Down Syndrome Vascular Endothelial Growth Factor Expression Angiogenesis Inhibitor Maspin Expression
Cast of Characters
The TP53 gene is the most mutated gene in human cancer and as a consequence has been one of the most extensively studied genes in the human genome. Over half of all human cancers carry direct mutations of the p53 coding region. In addition to sporadic mutations in human cancer, inherited mutations in TP53 cause a genetic predisposition to cancer called Li–Fraumeni syndrome. Individuals with Li–Fraumeni exhibit early onset of a wide variety of cancers including soft-tissue sarcoma, leukemia, osteosarcoma, and tumors of the breast and brain (Li et al., 1988). The TP53 gene is not essential for development but seems to have evolved a primary function to prevent neoplasia in multicellular organisms. The p53 tumor suppressor protein has the structure of a classical transcription factor possessing a central domain with sequence-specific DNA binding activity and an N-terminal acidic region required for transcriptional regulation of target genes. The majority of mutations observed in human cancer fall within the DNA binding region of p53 and hence disrupts its ability to modify gene expression (Soussi et al., 2006). The major tumor suppressive properties of p53 are derived from increasing expression of hundreds of target genes that inhibit cell cycle progression and promote apoptosis. In addition, p53 is also thought to regulate target genes that can affect the tumor microenvironment to inhibit angiogenesis and metastasis.
Introduction: Hardwiring the Angiogenic Switch
Mammalian cells have intensive requirements for oxygen and nutrients and are provided with a constant supply of both via the capillary network of the vascular system. In general, cells cannot be located further than approximately 100 μm from a capillary, which represents the diffusion limit of oxygen through tissues. As a result, an inherent property of rapidly proliferating tissues is the formation of new blood vessels – a process termed angiogenesis. During angiogenesis, cells comprising the vasculature (endothelial cells) are stimulated to divide and migrate to regions of low oxygen (hypoxia) requiring vascularization. Part of the angiogenic response also involves the proteolytic breakdown of the extracellular matrix (ECM) to allow for expansion and growth of existing blood vessels. Widespread angiogenesis occurs during the normal physiological processes of embryogenesis and at specific periods in adulthood such as wound healing and cycling of the female reproductive system. The inappropriate induction of angiogenesis is associated with several pathologies including psoriasis, macular degeneration, inflammation, and atherosclerosis. Aberrant angiogenesis is also associated with the progression of cancer, where it plays a critical role in the growth, invasion, and metastases of solid tumors.
The process of angiogenesis is promoted by the production of growth factors that stimulate endothelial cells to divide and sprout into new blood vessels. This angiogenic potential is countered by the production of endogenous factors that actively inhibit angiogenesis. The resulting balance of pro- and anti-angiogenic factors determines whether a given tissue or tumor cell mass undergoes angiogenesis. The early stages of tumor development are thought to be characterized by an “angiogenic switch” in which the output of pro-angiogenic factors exceeds that of anti-angiogenic factors, resulting in the tumor becoming highly vascularized and aggressive. Prior to the angiogenic switch, a small occult tumor can lie dormant for many years before transitioning to become a vascularized, rapidly growing tumor.
Because new blood vessel formation is an inherent property of rapidly dividing tissues, genes that promote cell division are often hardwired to concomitantly stimulate angiogenesis. For example, some of the most potent growth-promoting genes, such as the oncogenes ras (Rak et al. 2000), src (Mukhopadhyay, Tsiokas and Sukhatme 1995a; Mukhopadhyay et al. 1995b)], myc (Pelengaris et al. 1999), and fos (Saez et al. 1995), are also powerful stimulators of angiogenesis. Conversely, the same logic applies to tumor suppressor proteins, which are capable of limiting cellular division and at the same time preventing angiogenesis. The Rb tumor suppressor, for example, is a key regulator of cell cycle progression and has been shown to inhibit angiogenesis through several mechanisms (Gabellini, Del Bufalo and Zupi 2006). As discussed in this chapter, p53 – a crucial cell cycle regulator and major tumor suppressor protein in humans – also plays an important role in inhibiting angiogenesis and is key to maintaining the angiogenic switch in an “off” state.
p53: The Guardian of the Genome
The tumor suppressor protein p53 is a key regulator of the cellular response to genotoxic damage, and thus plays a pivotal role in preventing cancer formation. Once DNA damage has been incurred, p53 can elicit several different responses to either correct the error(s) or destroy the damaged cell. First, p53 can induce G1 cell cycle arrest, which stops the cell from dividing and allows time to repair the damage before the DNA is replicated. Second, p53 can activate DNA repair proteins to drive the repair of damaged DNA. Third, as a last resort, p53 can induce damaged cells to undergo programmed cell death (apoptosis), thereby eliminating damaged – and potentially dangerous – cells at risk for neoplastic transformation. Due to its critical importance in maintaining genetic stability, p53 has been called the “gatekeeper” or “guardian” of the genome.
Given the crucial role of p53 in maintaining genomic stability, it is perhaps not surprising to find that the gene encoding p53 is the most commonly altered gene in human cancers, being mutated or deleted in half of all tumors. Moreover, many cancers involve cellular or viral oncogenes that target and inactivate p53. For example, the simian virus 40 (SV40) large T-antigen (Mietz et al. 1992; Jiang et al. 1993), the adenoviral E1A, E1B, and E4orf6 proteins (Yew and Berk 1992; Dobner et al. 1996; Steegenga et al. 1996; Somasundaram and El-Deiry 1997), and the E6 oncoprotein from human papilomavirus (Band et al. 1993) all inhibit p53. Loss of functional p53 creates an environment that is permissive for genome instability, and allows cells with DNA damage (i.e., mutations and chromosomal aberrations) to continue replicating, which, left unchecked, can contribute to tumorigenesis.
The biological effects of p53 are predominantly carried out by its ability to transcriptionally regulate – either through activation or repression – the expression of specific target genes (although p53 can also perform functions that are independent of its transcriptional activity, some examples of which are provided in this chapter). To date, more than 150 p53 target genes have been identified; these genes act at essentially every step involved in carcinogenesis, providing an explanation for how p53 is capable of exerting such extensive restraints on tumor formation. In general, the p53 target genes that have been most well characterized are those involved in cell cycle arrest, such as p21 (recently renamed CDKN1A), and those that induce apoptosis, such as BAX, PUMA (or BBC3), and NOXA (or PMAIP1).
p53 is a tetrameric protein that contains four functional domains and possesses the hallmarks of a classical transcription factor. The N-terminus of the protein contains a prototypical acidic transactivation domain, which interacts with components of the basal transcription machinery and promotes the transcription of genes harboring p53-binding elements in their promoters. The central core domain is involved in sequence-specific DNA binding and is the location of the majority of oncogenic p53 mutations. The oligomerization domain contains nuclear localization signals and is involved in tetramerization of the protein. Lastly, the basic C-terminal domain is a negative regulatory domain that can inhibit DNA binding by the core domain.
The activity of p53 is modulated by the coordinated interplay of covalent modifications within the N- and C-terminal domains and a suite of interacting protein partners, which function to keep p53 levels and activity under extremely tight control. In normal cells (i.e., those that are unstressed or undamaged), p53 is maintained at a very low level and in a relatively inactive form, held in check by interaction with its primary negative regulator, MDM2. MDM2 binds the N-terminal domain of p53 and restricts its function in three ways. First, MDM2 translocates p53 from the nucleus to the cytoplasm, thus preventing p53 from accessing DNA. Second, binding of MDM2 to p53 results in the concealment of its transcriptional activation domain, thereby suppressing p53-mediated transactivation. Third, MDM2, which is an E3 ubiquitin ligase, promotes the ubiquitination of p53 and targets it for degradation by the 26S proteasome. The continual degradation of p53 results in an extremely short half-life for the protein, in the range of mere minutes.
In response to a variety of cellular stress stimuli – such as hypoxia, DNA damage (induced by either UV, IR, or chemical agents), X-ray irradiation, or oncogene activation (for example, ras, myc, or E2F) – p53 undergoes substantial post-translational modification and becomes functionally active. The critical event leading to p53 activation is phosphorylation of its N-terminal domain, which inhibits MDM2 binding, resulting in the rapid accumulation of p53 in the nucleus. Post-translational modifications in the C-terminal domain, including not only phosphorylation but also acetylation, sumoylation, and methylation, have diverse affects on p53 stability and function. For example, acetylation of C-terminal residues has been shown to protect p53 from ubiquitination, to potentiate its interaction with other transcription factors, and to induce a conformational change in the protein that exposes the DNA binding domain, allowing it to activate or repress target genes (Appella and Anderson 2001). Different genotoxic stresses activate p53 through distinct pathways and result in post-translational modification of distinct subsets of residues, but the net result of these post-translational modifications is an increase in not only the stability of the protein but also its biological activity.
A Role for p53 in Inhibiting Angiogenesis
In addition to its well-known functions in regulating cell-autonomous effects (i.e., cell cycle arrest, DNA repair, and apoptosis), the tumor suppressive role of p53 is also mediated through more complex host–tumor interactions, in particular angiogenesis. Several lines of evidence support the notion that p53 limits tumor vascularization. First, clinical studies involving a variety of cancers have demonstrated that tumors carrying p53 mutations are more highly vascularized than tumors harboring wild-type p53. For example, prostate tumors expressing mutated p53 have significantly greater microvessel density (MVD), a semi-quantitative measure of tumor vascularization, than tumors expressing wild-type p53 (Yu et al. 1997; Takahashi et al. 1998). Similar correlations have been observed between p53 status and MVD in colon cancer (Kang et al. 1997; Takahashi et al. 1998; Faviana et al. 2002), head and neck tumors (Gasparini et al. 1993) and breast cancer (Gasparini et al. 1994). Of particular clinical interest is that even early stage node-negative breast cancers harboring p53 mutations appear to have higher MVD than tumors with wild-type p53 and correlate with poor prognosis (Gasparini et al. 1994). Normally, node-negative breast cancers have a favorable prognosis, but the increased angiogenic potential provided by early loss of p53 in these cancers may be sufficient to render them much more aggressive.
In addition to the clinical evidence, experimental studies in human cell lines and mouse models have also shown connections between p53 and angiogenesis. Initial clues came from studies showing that p53 was able to perform tumor suppressor functions independent of its anti-proliferative and pro-apoptotic effects and could result in avascular, dormant tumors in vivo (Holmgren, Jackson and Arbiser 1998). Further evidence came from reports showing that overexpression of wild-type p53 could inhibit differentiation of human umbilical vein endothelial cells (HUVECs) into capillary-like structures in vitro and neovascularization in vivo (Riccioni et al. 1998). Finally, additional experiments using mouse models have shown that p53-deficient cell lines show a less pronounced and slower response to anti-angiogenic therapy compared with p53 wild-type cell lines, indicating anti-angiogenic therapy is sensitive to p53 status (Yu et al. 2002).
The molecular basis by which p53 interacts with and inhibits the regulatory pathways of angiogenesis have recently begun to be elucidated. These studies have defined three basic mechanisms by which p53 inhibits angiogenesis: (1) inhibition of hypoxia-sensing systems, (2) downregulation of pro-angiogenesis genes, and (3) upregulation of anti-angiogenesis pathways. The overall combination of these effects dramatically dampens the angiogenic output of tumor cells and tilts the tumor microenvironment in the host substantially toward angiogenesis suppression.
Inhibition of Hypoxia-Sensing Systems
The p53 tumor suppressor inhibits the HIF-1 pathway by directly binding to HIF-1α and targeting the protein for degradation (Ravi et al. 2000) (see Fig. 9.1). Because HIF-1 is a master regulator of the hypoxia-induced response, inhibiting HIF-1 will have major effects on tumor angiogenesis. Notably, the ability of p53 to inhibit the HIF-1 system is mediated by its physical interaction with HIF-1α and does not require its transcriptional activity.
The ability of p53 to inhibit HIF-1 activity does not occur under normal physiological conditions (Rempe et al. 2007). Rather, it most likely occurs only in the extreme hypoxic environment of a tumor or when accompanied by other forms of genotoxic stress (i.e., conditions under which p53 is activated). Evidence in support of this model has shown that inhibition of HIF-1 by p53 only occurs when p53 becomes stabilized in tumor cells by stimuli such as DNA damage (Kaluzova et al. 2004) or acidosis and nutrient deprivation that occur within tumors (Pan et al. 2004). Thus, one of the roles of p53 in tumor suppression may be to act as a final checkpoint in allowing angiogenesis to progress, much as it does for cell division.
Downregulation of Pro-angiogenic Factors
p53-Regulated genes implicated in angiogenesis
VEGF (vascular endothelial growth factor)
FGF2/bFGF (basic fibroblast growth factor)
FGFBP1/bFGF-BP (basic fibroblast growth factor-binding protein)
BAI1 (brain-specific angiogenesis inhibitor 1)
EPHA2 (ephrin receptor A2)
COL18A1 (α1 collagen 18)
COL4A1 (α1 collagen 4)
P4HA2 (α [II] 4-prolyl hydroxylase)
p53 inhibits VEGF expression during hypoxia by binding the transcription factor Sp1 and inhibiting its ability to bind the VEGF promoter and activate VEGF transcription (Pal, Datta and Mukhopadhyay 2001). As noted above, p53 also indirectly modulates VEGF expression by inhibiting another transcriptional activator of VEGF, HIF-1. Thus, p53 employs redundant mechanisms to inhibit one of the most potent pro-angiogenic factors known.
Mutations in p53 have been associated with increased VEGF expression in a number of human cancers including lung (Fontanini et al. 1997), bladder (Crew et al. 1997), and colorectal (Takahashi et al. 1998) carcinomas. The correlation between p53 status and VEGF expression appears to have important prognostic value, as breast cancer prognosis can be predicted with greater accuracy using VEGF expression and p53 status as prognostic markers than either VEGF or p53 status alone (Gasparini et al. 1997; Obermair et al. 1997; Linderholm et al. 2000; Linderholm et al. 2001). Finally, the ability of wild-type p53 to inhibit angiogenesis has been linked to suppression of VEGF in human leiomyosarcoma and synovial sarcoma (Zhang et al. 2000a).
COX-2 (also known as PTGS2) is an enzyme that converts arachidonic acid to prostaglandin H2, an intermediate that is subsequently converted to numerous other molecules known as prostanoids, which are key players in inflammation. Interestingly, the prostanoids produced by COX-2 can also stimulate the expression of pro-angiogenic factors (Tsujii et al. 1998; Williams et al. 2000). Thus, by directly repressing the COX-2 gene, p53 may inhibit a significant pathway of prostanoid-mediated angiogenesis. p53-mediated repression of COX-2 occurs by a mechanism whereby p53 inhibits TATA box-binding protein (TBP) from binding to the COX-2 promoter (Subbaramaiah et al. 1999). Although the precise mechanism is not yet clear, previous studies have shown that p53 suppresses a number of promoters that contain TATA elements by directly binding to TBP and interfering with its ability to stably bind the TATA box (Seto et al. 1992; Ragimov et al. 1993).
COX-2 is upregulated in many types of human cancers (Masferrer et al. 2000), consistent with an important role for this enzyme in promoting tumor angiogenesis. In some cases, COX-2 overexpression has been associated with altered p53 status (for example, Erkinheimo et al., 2004). Notably, inhibition of COX-2 has been shown to suppress tumor formation in various animal models of carcinogenesis (Oshima et al. 1996; Chulada et al. 2000). These observations have generated significant interest in the use of COX-2 inhibitors as potential cancer therapeutics.
FGF2 and FGF-BP
FGF2 (also known as basic FGF or bFGF) is a potent pro-angiogenic cytokine present in the basement membranes and subendothelial ECM of blood vessels. FGF2 is held in an inactive form in the ECM through binding to heparan sulfates and proteoglycans present in the ECM and is liberated during angiogenesis from these ECM reservoirs. Release of FGF2 from the ECM requires the action of a secreted protein called FGF-BP (also called bFGF-BP). p53 represses expression of FGF2 through direct repression of the FGF2 core promoter (Ueba et al. 1994), although the precise mechanism is not yet clear as the FGF2 promoter does not contain a TATA box. Moreover, p53 directly represses expression of the FGF-BP gene (Sherif et al. 2001). Thus, similar to regulation of VEGF, p53 uses redundant approaches to suppress a potent pro-angiogenic factor.
In addition to directly repressing transcription of pro-angiogenic genes, p53 may also inhibit production of pro-angiogenic factors through post-transcriptional mechanisms. In particular, a proteomics-based approach has identified a number of secreted factors induced or repressed by p53 (Khwaja et al. 2006). Among the repressed pro-angiogenic factors identified were known direct targets of p53, including VEGF and transforming growth factor-beta (TGF-β), as well as previously unidentified targets CYR61, FGF4, and the cytokine interleukin-8. However, p53 did not appear to modulate the mRNA levels of these genes, suggesting that p53 may increase the levels of these proteins by post-translational mechanisms such as enhancing their stability or secretion. Such secondary effects caused by p53 activation may also be significant in mediating changes in the tumor microenvironment and underscores the notion that not all p53-dependent effects are transcriptionally based.
Transcriptional Activation of Anti-Angiogenic Factors
p53 has been shown to upregulate a number of genes that inhibit angiogenesis (see Table 9.1). Unlike transcriptional repression by p53, which can occur by a variety of mechanisms, activation by p53 is strictly dependent on binding of p53 to a cognate DNA sequence motif in the promoter of its target gene. Anti-angiogenic factors upregulated by p53 are, in most cases, factors that are secreted into the ECM. The biological importance of these p53-induced factors in limiting tumor growth likely depends on both tumor type and location. As discussed below, several anti-angiogenic factors that are stimulated in response to p53 also require proteolytic cleavage in order to release a biologically active angiogenesis inhibitor.
Thrombospondin-1 (TSP-1 or THBS1)
One of the first p53 target genes identified was TSP-1; interestingly, it was also the first naturally occurring angiogenesis inhibitor to be discovered (Good et al. 1990; Tolsma et al. 1993; Dameron et al. 1994; Weinstat-Saslow et al. 1994). TSP-1 expression is reciprocally regulated by tumor suppressor genes and oncogenes: for instance, TSP-1 expression is upregulated by the p53 and PTEN tumor suppressors, but downregulated by a number of oncogenes, including v-src (Slack and Bornstein 1994), c-jun (Dejong et al. 1999), myc (Tikhonenko, Black and Linial 1996; Janz et al. 2000), and ras (Rak et al. 2000; Watnick et al. 2003). TSP-1 is a large, 450-kDa glycoprotein that is secreted into the ECM. The anti-angiogenic activity of TSP-1 has been localized to two of its three so-called thrombospondin type 1 repeats (TSRs), a motif that is present in over 70 human proteins. TSP-1 has been shown to potently inhibit angiogenesis through a number of different mechanisms. For example, TSP-1 negatively regulates endothelial cell proliferation and migration both in vitro and in vivo (Hsu et al. 1996; Jimenez et al. 2000; Nor et al. 2000). Many of these inhibitory effects are mediated by direct binding of TSP-1 to endothelial cells via interaction between the TSRs and the cell surface receptor, CD36 (Dawson et al. 1997; Jimenez et al. 2000). TSP-1 also directly activates the multifunctional cytokine TGF-β1, a secreted factor that plays a role in a variety of biological functions, including inhibition of angiogenesis. TGF-β1 is secreted in a latent form that is activated by TSP-1 in the ECM (Crawford et al. 1998). Notably, TGF-β1 has been shown to have potent tumor suppressive properties in several types of human cancers (Derynck, Akhurst and Balmain 2001).
TSP-1 expression has been shown to correlate with p53 status in a variety of cancers, including prostate cancer (Kwak et al. 2002), advanced epithelial ovarian carcinoma (Alvarez et al. 2001), and melanoma (Grant et al. 1998). However, in other cancers, such a correlation has not been found (Kawahara et al. 1998; Tokunaga et al. 1998; Grossfeld et al. 2002), suggesting that regulation of TSP-1 by p53 may be cell-type or tissue-type specific. In some cases, loss of p53 and the resulting decrease in TSP-1 has been shown to correlate with increased tumor angiogenesis (Grant et al. 1998; Alvarez et al. 2001). Fibroblasts from patients with Li–Fraumeni syndrome, a cancer predisposition disease, show a loss of p53 function that correlates with a reduction in TSP-1 protein expression and a switch from an inhibitory to stimulatory angiogenic phenotype (Stellmach et al. 1996; Volpert, Dameron and Bouck 1997). These results strongly suggest that loss of p53 function contributes, at least in part, to the development of an angiogenic phenotype by decreasing TSP-1 expression.
Brain-Specific Angiogenesis Inhibitor 1 (BAI1)
Expressed almost exclusively in the brain, BAI1 was originally identified in a screen for p53-target genes in glioblastoma cells (Tokino et al. 1994), and subsequent studies verified that expression of BAI1 could be induced directly by wild-type p53 (Nishimori et al. 1997). The BAI1 gene product is a large, seven-pass transmembrane protein, with extended extracellular and cytoplasmic regions, that belongs to the adhesion-type family of G-protein-coupled receptors. The N-terminal extracellular region of BAI1 contains five TSRs, which bear homology to the TSRs that are found in TSP-1. BAI1 is cleaved at a conserved G-protein-coupled receptor proteolytic cleavage site within the extracellular domain to generate a soluble 120-kDa fragment containing the five TSRs. This fragment has been shown to inhibit endothelial cell migration and proliferation in vitro and to inhibit angiogenesis in vivo in a mouse model; because of these anti-angiogenic properties, the N-terminal fragment was dubbed vasculostatin (Kaur et al. 2005). The fragment also suppresses the growth of glioma tumor xenografts in mice. BAI1 was the first transmembrane protein identified that harbored a releasable proteolytic fragment with anti-angiogenic and anti-neoplastic properties. Interestingly, an earlier study had identified a secreted activity in glioblastoma cells that could potently inhibited angiogenesis and was also stimulated by p53 (Van Meir et al. 1994). This factor was called glioma-derived angiogenesis inhibitory factor (GD-AIF), but thus far, the identity of the protein has not been determined. It is tempting to speculate that this activity may be vasculostatin.
Another p53-inducible gene that has been implicated in regulation of angiogenesis is the EPHA2 gene (Dohn, Jiang and Chen 2001; Brantley et al. 2002). EPHA2 is a member of a family of transmembrane receptor tyrosine kinases, called the ephrin receptors, which, together with their ligands (ephrins), are thought to play important roles in angiogenesis (Dodelet and Pasquale 2000; Pasquale 2005). Like most ephrin receptors, EPHA2 signaling is decidedly complex and has been shown to have either an oncogenic or a tumor suppressor effect depending on physiological context. In cancer cells, the EPHA2 gene is typically overexpressed and functions as a potent oncogene (Walker-Daniels et al. 1999; Easty and Bennett 2000; Ogawa et al. 2000; Zelinski et al. 2001; Miyazaki et al. 2003). In normal cells, however, EPHA2 is expressed at low levels and acts as a negative regulator of cell growth (Coffman et al. 2003; Kinch and Carles-Kinch 2003). These different activities appear to depend on the ability of EPHA2 to bind its ligand, ephrin-A1, which is anchored to the membrane of adjacent cells; in malignant cells, unstable cell–cell contacts prevent EPHA2 from binding ephrin-A1. Adenoviral-mediated delivery of ephrin-A1 has been shown to decrease the tumorigenic potential of EPHA2-overexpressing breast cancer cells in vivo (Noblitt et al. 2004). Interestingly, both the EPHA2 receptor and ephrin-A1 are upregulated in response to p53, suggesting that modulation of signaling through the ephrin pathway may represent another mechanism by which p53 disrupts angiogenesis (Dohn et al. 2001; Brantley et al. 2002).
The SERPINB5 gene encodes a 42-kDa protein, commonly called maspin, that belongs to the serine protease inhibitor (serpin) superfamily. Serpins play a number of important biological functions, including regulating cell adhesion and differentiation. Maspin was initially identified as a candidate tumor suppressor gene involved in breast cancer based on its reduced expression in mammary tumor cell lines compared to normal mammary epithelial cells, and subsequent studies revealed that expression of maspin in breast tumor cells could inhibit tumor cell invasion and metastasis (Zou et al. 1994). Maspin is also downregulated in several other types of cancers, and clinical studies have associated loss of maspin expression with the progression and poor prognosis (Khalkhali-Ellis 2006). The ability of maspin to inhibit tumor cell invasion and metastasis appears to derive, at least in part, from its potential to inhibit angiogenesis. Maspin inhibits endothelial cell migration in vitro and blocks neovascularization in an animal model (Zhang et al. 2000b). p53 induces maspin expression by binding the p53-response element present in the promoter, indicating maspin is a direct p53 target gene (Zou et al. 2000). Interestingly, several studies have found correlations between maspin expression, p53 status, and MVD in colon cancers (Song et al. 2002).
A relatively recently identified class of potent, endogenous angiogenesis inhibitors are those that are derived from proteolytic fragments of certain types of collagen, an abundant insoluble fibrous protein present in the ECM and connective tissue. Most anti-angiogenic collagens are constitutively expressed in vascular basement membranes; these include type IV collagens, the main constituent of basement membranes, as well as type XVIII collagen (Kalluri 2003). These collagens can be proteolytically processed by matrix metalloproteases, serine proteases, or cysteine proteases to produce anti-angiogenic peptides such as endostatin, which is derived from α1 collagen 18, and tumstatin, canstatin, and arresten, which are derived, respectively, from α3, α2, and α1 collagen 4. These peptides exert their effects by interacting with specific receptors on the endothelial cell surface and inhibiting angiogenesis by, for example, reducing endothelial cell proliferation or migration, or increasing endothelial cell apoptosis. Collagen-derived anti-angiogenic peptides have generated significant therapeutic interest as angiogenesis inhibitors for the treatment of cancer, but recent evidence has begun to suggest that these molecules are also part of the body’s naturally occurring tumor suppressor mechanisms. An understanding of these mechanisms first requires an introduction to the collagen biosynthesis pathway.
The collagen family of proteins is characterized by the presence of so-called collagen repeats, which comprise several copies of the amino acid sequence glycine-X-Y, in which X is often a proline residue and Y is frequently a 4-hydroxyproline residue. To create a mature collagen molecule, the collagen repeats from three separate collagen α chains wind around one another to form a rigid triple helix structure. Formation of this triple helix requires hydrogen bonding between the 4-hydroxyproline residues; without these modified amino acid residues, collagen molecules are unstable and do not form.
Proline hydroxylation is the rate-limiting step in collagen biosynthesis. The two enzymes that post-translationally modify prolines within the collagen repeats are called α[I] and α[II] prolyl 4-hydroxylase (hereafter referred to as α(I)PH and α(II)PH). The α(I)PH isoenzyme is ubiquitously expressed and is the predominant collagen prolyl hydroxlase activity in connective tissues. By contrast, the α(II)PH isoform has a much more restricted expression pattern and is mostly expressed in chondrocytes and capillary endothelial cells (Nissi et al. 2001). The prolyl 4-hydroxylases that modify collagens are catalytically very similar to the PHDs involved in HIF-1α hydroxylation mentioned above in that they require molecular oxygen, as well as ascorbic acid, Fe2+, and α-ketoglutarate, as cofactor for catalytic activity. However, the collagen prolyl 4-hydroxylases differ from PHDs in their subcellular localization: prolyl 4-hydroxylases are localized to the endoplasmic reticulum where collagens are assembled, whereas PHDs reside primarily in the cytoplasm. Loss of collagen prolyl 4-hydroxylase enzymatic activity, which can occur when its essential cofactor ascorbic acid (vitamin C) is lacking in the diet, causes scurvy, highlighting the importance of prolyl hydroxylation in collagen biosynthesis.
Collagen-derived anti-angiogenic factors
α1 collagen 4
α2 collagen 4
α3 collagen 4
α6 collagen 4
α1 collagen 8
α1 collagen 15
α1 collagen 18
As yet unnamed
As stated above, type IV collagens are the most abundant component of vascular basement membranes. Notably, of the six type IV collagens, four have been reported to possess domains that are anti-angiogenic (see Table 9.2). The presence of anti-angiogenic collagens in the basement membrane is thought to maintain endothelial cells in a quiescent (non-dividing) state (Kalluri 2003). In order for angiogenesis to occur, endothelial or cancer cells must proteolytically degrade the vascular basement membrane by secreting matrix metalloproteases or other ECM-degrading enzymes. However, this process would also liberate vast amounts of collagen-derived angiogenesis inhibitors, and unless tumor cells under these conditions express sufficient pro-angiogenic factors to overcome the release of angiogenesis inhibitors, neovascularization will not occur.
It is interesting to note that small molecule inhibitors of matrix metalloproteases were once regarded as ideal drug targets to prevent metastasis. Unfortunately, such inhibitors are proved unsuccessful for treating cancer and, in some tumor types, even resulted in accelerated tumor growth during clinical trials (Coussens, Fingleton and Matrisian 2002; Fingleton 2006). One possible explanation for the ineffectiveness of these inhibitors is that by inhibiting matrix metalloproteases, they also inhibit the production of collagen-derived anti-angiogenic peptides, and thus may actually promote angiogenesis rather than inhibiting the process.
Clinical implications of p53-mediated Upregulation of Endogenous Angiogenesis Inhibitors
Collagen-Derived Angiogenesis Inhibitors May Contribute to the Body’s Natural Tumor Suppressor Mechanisms
Although activation of p53 can lead to the production of collagen-derived angiogenesis inhibitors, whether or not these factors are effective in limiting tumor growth in vivo remains to be determined. Several lines of evidence suggest that increasing biosynthesis of endogenous angiogenesis inhibitors could potentially have dramatic impact on tumor incidence and growth. Perhaps the most intriguing piece of evidence has come from an interesting observation made in individuals with trisomy 21, also known as Down syndrome. In particular, certain common pediatric cancers such as neuroblastoma, which typically represents approximately 30% of childhood solid tumors, are only very rarely observed in individuals with Down syndrome. Likewise, in older individuals, prevalent cancers, such as breast cancer, are completely absent in Down’s patients. The extremely low incidence of solid tumors in Down’s individuals has led to the proposal that chromosome 21 harbors a potent tumor suppressor activity. Indeed, studies using mouse models have identified two candidate tumor suppressor genes, Dscr1 (Down syndrome candidate region 1) (Minami et al. 2004) and Ets2 (Sussan et al. 2008), present on chromosome 21.
Interestingly, the gene encoding α1 collagen 18, COL18A1, is also located on chromosome 21, and it has been proposed that increased dosage of its C-terminal anti-angiogenic fragment, endostatin, could help explain the reduced solid tumor incidence in individuals with Down syndrome (Folkman and Kalluri 2004; Sund et al. 2005). Consistent with this proposal, serum levels of endostatin are approximately 30% higher in Down’s individuals compared to the general population (Zorick et al. 2001). Experiments in mice have demonstrated that an increase in endostatin levels as little as 1.7-fold results in significantly slower rates of tumor growth (Sund et al. 2005), suggesting that even a modest increase in the dosage of a collagen-derived anti-angiogenic peptide could lead to dramatically reduced tumor growth rates. Conversely, mice deficient in either the Col4a3 or Col18a1 gene exhibit increased rates of tumor growth (Sund et al. 2005). Collectively, these observations suggest that collagen-derived angiogenesis inhibitors may have physiological relevance in limiting tumor formation and growth.
Endogenous Angiogenesis Inhibitors May Mediate Long-Range Host–Tumor Interactions
The p53-mediated production of secreted collagen-derived angiogenesis inhibitors may also be relevant to abscopal effects that occur following surgical removal of a tumor (see Fig. 9.3). It has been reported that primary tumors can inhibit the growth of a distant metastasis through inhibition of angiogenesis (Prehn 1991; O’Reilly et al. 1994; Sckell et al. 1998). In these situations, surgical removal of the primary tumor results in dramatically increased metastatic growth. However, when the primary tumor is irradiated, angiogenesis at a distal site is inhibited and growth of the metastases is limited (Hartford et al. 2000). Interestingly, in this same study, irradiation of the primary tumor was shown to lead to an increase in serum endostatin levels, consistent with the notion that p53-dependent production of collagen-derived angiogenesis inhibitors plays a role in this abscopal effect.
A Potential Role for Endogenous Angiogenesis Inhibitors in Promoting Tumor Dormancy
The p53-mediated production and secretion of anti-angiogenic factors may also help explain how small tumors are maintained in a dormant state for long periods of time. For instance, almost all individuals are thought to carry small microscopic growths of transformed cells – termed in situ carcinomas – in the thyroid, and yet thyroid cancer is a relatively rare type of malignancy (Harach, Franssila and Wasenius 1985; Black and Welch 1993). Similarly, approximately 40% of women between the ages of 40 and 50 harbor small malignant mammary cell foci; however, only 1% of the female population is diagnosed with breast cancer (Nielsen et al. 1987; Black and Welch 1993). It has been proposed that production of anti-angiogenic factors by in situ carcinomas limits their growth and prevents the majority of these tumors from becoming life threatening (Folkman and Kalluri 2004).
An important question in the field of cancer biology is: What are the molecular mechanisms that drive small benign growths to become highly vascularized, rapidly growing cancers? The goals of future research in this area are to understand the molecular events that control the angiogenic switch and how small in situ carcinomas can be maintained in dormant states for years or even decades. Once a clear understanding of the process is gained, it may then be possible to devise therapies that revert large tumors back into a dormant state. Moreover, current cancer diagnoses generally occur after the tumor has reached a dangerously large size. As better biomarkers are discovered that allow cancer to be detected at earlier stages, administration of anti-angiogenic therapy may be an effective method of maintaining tumors in a small, asymptomatic state.
p53-Induced Angiogenesis Inhibitors as Potential Cancer Therapeutics
In the early 1970s, Judah Folkman proposed that targeting angiogenesis in tumor tissues could be a therapeutic approach to preventing tumor progression. To date, more than 30 angiogenesis inhibitors are in clinical trials, representing one of the most promising avenues of cancer drug discovery. In particular, endogenous p53-induced angiogenesis inhibitors possess several important properties for therapeutic applications: they are generally non-toxic, they are secreted and function extracellularly, and often the biologically active portion of these molecules is a relatively small peptide. These ideal properties increase the possibility that endogenous p53-induced anti-angiogenic factors may make their way into the clinic as therapeutic treatments for cancer. The first endogenous angiogenesis inhibitor to enter clinical trials was endostatin. Although proven safe to use, recombinant human endostatin had only minimal efficacy when tested in clinical trials for advanced neuroendocrine tumors (Herbst et al. 2002; Kulke et al. 2006). Despite setbacks, there remains interest in further clinical trials for endostatin and other endogenous angiogenesis inhibitors such as tumstatin. One limitation of the endostatin trials was that the therapy was conducted as a monotherapy. However, successful trials with other angiogenesis inhibitors have generally been conducted in combination with conventional cancer therapeutics and may be a more feasible approach. In support of this notion, a modified form of endostatin called “Endostar” has been approved, in combination with chemotherapy, for treatment of non-small cell lung cancer in China (Jia and Kling 2006).
The limited success of endostatin has also prompted speculation that a more successful therapeutic approach may involve targeting multiple anti-angiogenic pathways. In this regard, one of the most promising angiogenesis inhibitors to be developed is Avastin (also known as bevacizumab), a monoclonal antibody that targets and inhibits the pro-angiogenic factor VEGF. Avastin was approved in 2004 in the United States for the treatment of metastatic colon cancer, and is now a frontline therapy for several types of malignancies. However, results of clinical trials with Avastin have not been overwhelming (Yang et al. 2003). One possibility yet to be addressed is whether agents that block pro-angiogenic factors, such as Avastin, can synergize with endogenous anti-angiogenic factors such as endostatin or tumstatin. Cocktail therapies containing a variety of anti-angiogenesis inhibitors targeting several pathways may provide significantly broader anti-cancer effects and dramatically improve the efficacy of current anti-angiogenesis approaches.
An additional strategy that may have potential in future therapies is to mobilize endogenous anti-angiogenic factors from their storage pools in the ECM by activating wild-type p53. The feasibility of this approach stems from the recent discovery of a family of small molecule MDM2 antagonists called Nutlins, which inhibit the p53-MDM2 interaction and thereby lead to activation of p53. One of these family members, Nutlin-3, has been shown to induce the expression of p53-regulated genes and to exhibit potent anti-proliferative activity in cells with functional p53. Nutlin-3 also inhibits the growth of human tumor xenografts in nude mice (Vassilev et al. 2004). Currently, there is interest in utilizing molecules such as Nutlin-3 as sensitizers for cytotoxic cancer therapies such as chemotherapy or radiation. It may be worthwhile to also test if these molecules also sensitize tumors to anti-angiogenic therapy in vivo.
The Role of p53 Status in Anti-angiogenic Therapies
It is well established that loss of p53 – which occurs in more than half of all human tumors – can diminish the therapeutic benefits of conventional chemotherapies such as radiation and other DNA damaging agents by reducing the cellular apoptosis response. With the recent development of angiogenesis inhibitors for cancer treatment, one of the most pressing clinical issues is whether the p53 status of tumors similarly influences the efficacy of drugs targeting tumor angiogenesis. Indeed, a number of studies have indicated that loss of p53 may reduce tumor responsiveness to anti-angiogenic treatments. In one study, for example, mice bearing tumors derived from p53–/– human colorectal cancer cells were more resistant than mice bearing isogenic p53+/+ tumors to an anti-angiogenic therapy involving an antibody, DC101, targeting the VEGF receptor (Yu et al. 2002). This resistance could be due, at least in part, to one or more of the p53-dependent mechanisms discussed above.
There are several additional reasons why tumors carrying p53 mutations may be somewhat refractory to anti-angiogenic therapy. Hypoxia-induced apoptosis, for example, appears to be at least partially dependent on p53 (Graeber et al. 1994), and when grown under hypoxic conditions, cells harboring p53 mutations rapidly overtake wild-type p53 counterparts, demonstrating the selective advantage that p53 null cells have under hypoxic conditions (Graeber et al. 1996). However, even if loss of p53 diminishes the efficacy of anti-angiogenic therapies, this does not necessarily reduce the potential of these drugs for treating cancer. For example, even though the p53-negative tumors described above became less sensitive to the DC101 antibody, it is important to note that these tumors still responded to treatment (Yu et al. 2002).
Inhibition of angiogenesis is now widely recognized as an important component of the p53 tumor suppressor pathway. p53 has been shown to inhibit angiogenesis by interfering with central regulators of hypoxia that mediate angiogenesis, inhibiting production of pro-angiogenic factors, and directly increasing the production of endogenous angiogenesis inhibitors. The combination of these effects allows p53 to efficiently shut down the angiogenic potential of cancer cells. Inactivation of p53 during tumorigenesis reverses these effects and provides a potent stimulus for tumor angiogenesis; as a result, tumors carrying p53 mutations are more vascularized, often more aggressive and frequently correlate with poor prognosis for treatment. Thus, the loss of functional p53 during tumorigenesis likely represents an essential step in the switch to an angiogenic phenotype that is displayed by tumors. p53-induced angiogenesis inhibitors represent novel therapies for the treatment of cancer, although clinical data suggest that maximal therapeutic benefit may require these inhibitors to be used in conjunction with other therapies. Future directions of research should attempt to find optimal combinations of conventional chemotherapies and anti-angiogenic therapies.
- Alvarez, A. A., Axelrod, J. R., Whitaker, R. S., Isner, P. D., Bentley, R. C., Dodge, R. K., and Rodriguez, G. C (2001) Thrombospondin-1 expression in epithelial ovarian carcinoma: Association with p53 status, tumor angiogenesis, and survival in platinum-treated patients. Gynecol. Oncol. 82:273–278.PubMedCrossRefGoogle Scholar
- Appella, E. and Anderson, C. W (2001) Post-translational modifications and activation of p53 by genotoxic stresses. Eur. J. Biochem. 268:2764–2772.Google Scholar
- Carmeliet, P., Dor, Y., Herbert, J. M., Fukumura, D., Brusselmans, K., Dewerchin, M., Neeman, M., Bono, F., Abramovitch, R., Maxwell, P., Koch, C. J., Ratcliffe, P., Moons, L., Jain, R. K., Collen, D., and Keshert, E (1998) Role of HIF-1alpha in hypoxia-mediated apoptosis, cell proliferation and tumour angiogenesis. Nature 394:485–490.PubMedCrossRefGoogle Scholar
- Erkinheimo, T. L., Lassus, H., Finne, P., van Rees, B. P., Leminen, A., Ylikorkala, O., Haglund, C., Butzow, R., and Ristimaki, A (2004) Elevated cyclooxygenase-2 expression is associated with altered expression of p53 and SMAD4, amplification of HER-2/neu, and poor outcome in serous ovarian carcinoma. Clin. Cancer Res. 10:538–545.PubMedCrossRefGoogle Scholar
- Faviana, P., Boldrini, L., Spisni, R., Berti, P., Galleri, D., Biondi, R., Camacci, T., Materazzi, G., Pingitore, R., Miccoli, P., and Fontanini, G (2002) Neoangiogenesis in colon cancer: Correlation between vascular density, vascular endothelial growth factor (VEGF) and p53 protein expression. Oncol Rep. 9:617–620.PubMedGoogle Scholar
- Fontanini, G., Vignati, S., Lucchi, M., Mussi, A., Calcinai, A., Boldrini, L., Chine, S., Silvestri, V., Angeletti, C. A., Basolo, F., and Bevilacqua, G (1997) Neoangiogenesis and p53 protein in lung cancer: Their prognostic role and their relation with vascular endothelial growth factor (VEGF) expression. Br. J. Cancer 75:1295–1301.PubMedCrossRefGoogle Scholar
- Gasparini, G., Weidner, N., Bevilacqua, P., Maluta, S., Dalla Palma, P., Caffo, O., Barbareschi, M., Boracchi, P., Marubini, E., and Pozza, F (1994) Tumor microvessel density, p53 expression, tumor size, and peritumoral lymphatic vessel invasion are relevant prognostic markers in node-negative breast carcinoma. J. Clin. Oncol. 12:454–466.PubMedGoogle Scholar
- Gasparini, G., Toi, M., Gion, M., Verderio, P., Dittadi, R., Hanatani, M., Matsubara, I., Vinante, O., Bonoldi, E., Boracchi, P., Gatti, C., Suzuki, H., and Tominaga, T (1997) Prognostic significance of vascular endothelial growth factor protein in node-negative breast carcinoma. J. Natl. Cancer Inst. 89:139–147.PubMedCrossRefGoogle Scholar
- Giuriato, S., Ryeom, S., Fan, A. C., Bachireddy, P., Lynch, R. C., Rioth, M. J., van Riggelen, J., Kopelman, A. M., Passegue, E., Tang, F., Folkman, J., and Felsher, D. W (2006) Sustained regression of tumors upon MYC inactivation requires p53 or thrombospondin-1 to reverse the angiogenic switch. Proc. Natl. Acad. Sci. USA 103:16266–16271.PubMedCrossRefGoogle Scholar
- Good, D. J., Polverini, P. J., Rastinejad, F., Le Beau, M. M., Lemons, R. S., Frazier, W. A., and Bouck, N. P (1990) A tumor suppressor-dependent inhibitor of angiogenesis is immunologically and functionally indistinguishable from a fragment of thrombospondin. Proc. Natl. Acad. Sci. USA 87:6624–6628.PubMedCrossRefGoogle Scholar
- Grossfeld, G. D., Carroll, P. R., Lindeman, N., Meng, M., Groshen, S., Feng, A. C., Hawes, D., and Cote, R. J (2002) Thrombospondin-1 expression in patients with pathologic stage T3 prostate cancer undergoing radical prostatectomy: Association with p53 alterations, tumor angiogenesis, and tumor progression. Urology 59:97–102.PubMedCrossRefGoogle Scholar
- Herbst, R. S., Hess, K. R., Tran, H. T., Tseng, J. E., Mullani, N. A., Charnsangavej, C., Madden, T., Davis, D. W., McConkey, D. J., O’Reilly, M. S., Ellis, L. M., Pluda, J., Hong, W. K., and Abbruzzese, J. L (2002) Phase I study of recombinant human endostatin in patients with advanced solid tumors. J. Clin. Oncol. 20:3792–3803.PubMedCrossRefGoogle Scholar
- Iyer, N. V., Kotch, L. E., Agani, F., Leung, S. W., Laughner, E., Wenger, R. H., Gassmann, M., Gearhart, J. D., Lawler, A. M., Yu, A. Y., and Semenza, G. L (1998) Cellular and developmental control of O2 homeostasis by hypoxia-inducible factor 1 alpha. Genes Dev. 12:149–162.PubMedCrossRefGoogle Scholar
- Jimenez, B., Volpert, O. V., Crawford, S. E., Febbraio, M., Silverstein, R. L., and Bouck, N (2000) Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1. Nat. Med. 6:41–48.Google Scholar
- Kang, S. M., Maeda, K., Onoda, N., Chung, Y. S., Nakata, B., Nishiguchi, Y., and Sowa, M (1997) Combined analysis of p53 and vascular endothelial growth factor expression in colorectal carcinoma for determination of tumor vascularity and liver metastasis. Int. J. Cancer 74:502–507.PubMedCrossRefGoogle Scholar
- Koumenis, C., Alarcon, R., Hammond, E., Sutphin, P., Hoffman, W., Murphy, M., Derr, J., Taya, Y., Lowe, S. W., Kastan, M., and Giaccia, A (2001) Regulation of p53 by hypoxia: Dissociation of transcriptional repression and apoptosis from p53-dependent transactivation. Mol. Cell. Biol. 21:1297–1310.PubMedCrossRefGoogle Scholar
- Kulke, M. H., Bergsland, E. K., Ryan, D. P., Enzinger, P. C., Lynch, T. J., Zhu, A. X., Meyerhardt, J. A., Heymach, J. V., Fogler, W. E., Sidor, C., Michelini, A., Kinsella, K., Venook, A. P., and Fuchs, C. S (2006) Phase II study of recombinant human endostatin in patients with advanced neuroendocrine tumors. J. Clin. Oncol. 24:3555–3561.PubMedCrossRefGoogle Scholar
- Minami, T., Horiuchi, K., Miura, M., Abid, M. R., Takabe, W., Noguchi, N., Kohro, T., Ge, X., Aburatani, H., Hamakubo, T., Kodama, T., and Aird, W. C (2004) Vascular endothelial growth factor- and thrombin-induced termination factor, Down syndrome critical region-1, attenuates endothelial cell proliferation and angiogenesis. J. Biol. Chem. 279:50537–50554.PubMedCrossRefGoogle Scholar
- Nishimori, H., Shiratsuchi, T., Urano, T., Kimura, Y., Kiyono, K., Tatsumi, K., Yoshida, S., Ono, M., Kuwano, M., Nakamura, Y., and Tokino, T (1997) A novel brain-specific p53-target gene, BAI1, containing thrombospondin type 1 repeats inhibits experimental angiogenesis. Oncogene 15:2145–2150.PubMedCrossRefGoogle Scholar
- Noblitt, L. W., Bangari, D. S., Shukla, S., Knapp, D. W., Mohammed, S., Kinch, M. S., and Mittal, S. K (2004) Decreased tumorigenic potential of EphA2-overexpressing breast cancer cells following treatment with adenoviral vectors that express EphrinA1. Cancer Gene Ther. 11:757–766.PubMedCrossRefGoogle Scholar
- Obermair, A., Kucera, E., Mayerhofer, K., Speiser, P., Seifert, M., Czerwenka, K., Kaider, A., Leodolter, S., Kainz, C., and Zeillinger, R (1997) Vascular endothelial growth factor (VEGF) in human breast cancer: Correlation with disease-free survival. Int. J. Cancer 74:455–458.PubMedCrossRefGoogle Scholar
- Rak, J., Mitsuhashi, Y., Sheehan, C., Tamir, A., Viloria-Petit, A., Filmus, J., Mansour, S. J., Ahn, N. G., and Kerbel, R. S (2000) Oncogenes and tumor angiogenesis: Differential modes of vascular endothelial growth factor up-regulation in ras-transformed epithelial cells and fibroblasts. Cancer Res. 60:490–498.PubMedGoogle Scholar
- Sund, M., Hamano, Y., Sugimoto, H., Sudhakar, A., Soubasakos, M., Yerramalla, U., Benjamin, L. E., Lawler, J., Kieran, M., Shah, A., and Kalluri, R (2005) Function of endogenous inhibitors of angiogenesis as endothelium-specific tumor suppressors. Proc. Natl. Acad. Sci. USA 102:2934–2939.PubMedCrossRefGoogle Scholar
- Tokunaga, T., Nakamura, M., Oshika, Y., Tsuchida, T., Kazuno, M., Fukushima, Y., Kawai, K., Abe, Y., Kijima, H., Yamazaki, H., Tamaoki, N., and Ueyama, Y (1998) Alterations in tumour suppressor gene p53 correlate with inhibition of thrombospondin-1 gene expression in colon cancer cells. Virchows Arch. 433:415–418.PubMedCrossRefGoogle Scholar
- Ueba, T., Nosaka, T., Takahashi, J. A., Shibata, F., Florkiewicz, R. Z., Vogelstein, B., Oda, Y., Kikuchi, H., and Hatanaka, M (1994) Transcriptional regulation of basic fibroblast growth factor gene by p53 in human glioblastoma and hepatocellular carcinoma cells. Proc. Natl. Acad. Sci. USA 91:9009–9013.PubMedCrossRefGoogle Scholar
- Wei, C. L., Wu, Q., Vega, V. B., Chiu, K. P., Ng, P., Zhang, T., Shahab, A., Yong, H. C., Fu, Y., Weng, Z., Liu, J., Zhao, X. D., Chew, J. L., Lee, Y. L., Kuznetsov, V. A., Sung, W. K., Miller, L. D., Lim, B., Liu, E. T., Yu, Q., Ng, H. H., and Ruan, Y (2006) A global map of p53 transcription-factor binding sites in the human genome. Cell 124:207–219.PubMedCrossRefGoogle Scholar
- Weinstat-Saslow, D. L., Zabrenetzky, V. S., VanHoutte, K., Frazier, W. A., Roberts, D. D., and Steeg, P. S (1994) Transfection of thrombospondin 1 complementary DNA into a human breast carcinoma cell line reduces primary tumor growth, metastatic potential, and angiogenesis. Cancer Res. 54:6504–6511.PubMedGoogle Scholar
- Yang, J. C., Haworth, L., Sherry, R. M., Hwu, P., Schwartzentruber, D. J., Topalian, S. L., Steinberg, S. M., Chen, H. X., and Rosenberg, S. A (2003) A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic renal cancer. N. Engl. J. Med. 349:427–434.PubMedCrossRefGoogle Scholar