Transcript
Page 1: Hypoxic stress proteins: Survival of the fittest

H y p o x i c Stress Proteins: Survival of t h e F i t t e s t AmatoJ. Giaccia The importance of tumor microenvironment in malig- nant progression has been largely ignored. Tumor cells protect themselves from changes in the microenviron- ment due to a decrease in nutrients and oxygen by reducing macromolecular synthesis and inducing genes that will promote angiogenesis and tissue remodeling, This ability of transformed cells to survive fluctuations in oxygen tensions is clinically important, as tumors with high hypoxic fractions respond poorly to many forms of cancer therapy. While it is commonly accepted that the decrease in molecular oxygen in hypoxic cells makes them more refractory to killing by agents such as ionizing radiation which use oxygen radical formation, the cessation of division and loss of apoptotic (cell suicide) potential in hypoxic cells are also important in

their resistance to killing by radiotherapy and chemother- apy. In this chapter, we will discuss hypoxia-induced stress proteins in regards to three clinically relevant end points: inhibition of cell proliferation, induction of apop- tosis, and regulation of genes modulating angiogenesis. Although these three end points may seem unrelated, in fact, they are intimately linked with each other in the cellular response to hypoxia and malignant progression. One of the goals of this review is to inform both clinician and scientist of these interrelationships and discuss how hypoxia selects for tumors that are clonal expan- sions of cells that have lost their apoptotic ability and have switched to a proangiogenic phenotype. Copyright �9 1996 by W.B, Saunders Company

Inhibition of Cell Proliferation by Hypoxia

Insights From In Vitro Studies

I t has long been known that cells exposed to low oxygen tensions (hypoxia) stop growing) -6 This

inhibition of cell proliferation by hypoxia is not the result of a general effect on the entire cell-cycle, for example due to a lack of energy. Instead, hypoxia specifically inhibits replicon initiation in early S- phase. 7,a Because the two most critical events in the cell-cycle are S-phase where chromosomes are repli- cated and M-phase where duplicated chromosomes are segregated, it is not that surprising that hypoxia prevents DNA replicon initiation. However, it is surprising that even stringent hypoxic conditions (0.02% oxygen) do not impede cells from entering mitosis by the activation of a G~ checkpoint, a common feature of cells exposed to ionizing radia- tion. 9,1~ The cell-cycle response of cells exposed to hypoxia is to exhibit an immediate inhibition of replicon initiation in early S-phase and an accumula- tion of Gi-phase cells that cannot enter S-phase. However, S-phase replicons that initiated before exposure to hypoxia will continue to elongate, com-

From the Department of Radiation Oncology, Division ~ Radiation Biology, Stanford University School of Medicine, Stanford, CA:

Supported in part by Grant No. CA64489j%m the National Cancer Institute and ACSjunior faculty research award.

Address reprint requests to AmatoJ. Giaccia, PILD, CBRL, Rm GK220, Division of Radiation Biology, Department of Radiation Oncology, Stanford University School of Medicine, Stanford, CA 94305-5468.

Copyright �9 1996by W..B. Saunders Company 1053-4296/96/0601-0006505.00/0

plete chromosome replication, enter mitosis, and arrest somewhere in the subsequent cell-cycle be- tween late GI and early S-phase (Fig 1).

Little is known about why cells accumulate in the Gl/S-phase of the cell cycle when exposed to hypoxia, except that this accumulation is different than growth arrest induced by medium depletion (starvation) and DNA damaging agents such as ionizing radiation, ll One distinctive feature of growth arrest induced by hypoxia is that replicon initiation is rapidly (within minutes) resumed when cells are reoxygenated, 7 indicating that all of the essential machinery for replicon initiation is present, but cannot be used. This finding is to be contrasted with cells that require many hours to reengage in cell-cycle activity after release from medium depletion, indicating that these cells do not possess all the essential components for DNA replication and must wait until they accumu- late the essential precursors to begin DNA synthesis. The induction of a G j/S-phase arrest by hypoxia has been hypothesized to protect cells from proceeding into S-phase and duplicating their genetic material under adverse conditions that could lead to genomic instability or errors in chromosome duplication. 2,6,12 Interestingly, the reversibility of the inhibition of replicon initiation by hypoxia is lost as the oxygen tension falls below 0.02%. 13

What proteins are needed for replicon initiation? Most of our understanding about eukaryotic DNA replication originates from in vitro studies on simian virus 40 (SV40) replication. I4-I6 Initiation of DNA replication occurs at specifically defined sequences called origins of replication. In the SV40 system, the replicon origin possesses two binding sites for the

46 Seminars in Radiation Oncology, Vol 6, No 1 (January), 1996:pp 46-58

Page 2: Hypoxic stress proteins: Survival of the fittest

Hypoxia "s Role in Cell Proliferation 47

t ~Low 02 / Inhibition of Replicon initiation

Figure 1. Cell-cycle response of cells exposed to low oxygen. The cell-cycle response of cells exposed to low oxygen conditions is to exhibit an immediate inhibition of replicon initiation in early S-phase and an accumulation of Gl-phase cells that cannot enter S-phase. However, S- phase replicons that initiated betbre exposure to hypoxia will continue to elongate, complete chromosome replica- tion, enter mitosis, and arrest somewhere in the subse- quent cell-cycle between late Gffearly S-phase, but will not initiate new replicons.

SV40 encoded protein large T antigen. Once large T antigen binds to these sites, it catalyzes an adenosine triphosphate (ATP) driven DNA unwinding reaction. Following the DNA unwinding reaction, virally en- coded large T antigen interacts with the host cell's single-stranded DNA binding protein, replication protein A (RPA). RPA is composed of 70, 34, and 14 kDa molecular weight subunits.17-19 The 70-kDa pro- tein is functionally conserved from bacteria to man and has attracted accessory subunits of 34 and 14 kDa in eukaryotic cells. Although all of the functions of these accessory subunits have not yet been identi- fied, antibodies against the 34-kDa protein will in- hibit SV40 DNA replication, indicating that it plays an essential role in DNA replication. 18 The p34 subunit is phosphorylated in a cell-cycle dependent manner and is hyperphosphorylated (a shift of 2 kDa in molecular weight) after cellular exposure to ioniz- ing radiation 2~ and ultraviolet light. 2j The signifi- cance of RPA hyperphosphorylation by stress induc- ing agents is still under investigation, but recent studies seem to indicate that the hyperphosphoryla- tion of the p34 subunit by hypoxia inhibits RPA's single-strand DNA binding activity (C. Koumenis, S. Greene, M. Tsai, R. Fischel, and AJ. Giaccia, unpub- lished observations). Therefore, as RPA binding to single-stranded DNA origins is necessary for replicon initiation, hypoxia could act to inhibit RPA binding activity by inducing its hyperphosphorylation (Fig 2). This relationship raises the intriguing and still specu-

lative possibility that RPA hyperphosphorylation in- hibits replicon initiation and signals for early-S-phase arrest under hypoxic conditions. At present, the identity of the kinase (s) that induce hyperphosphory- lation of RPA and the pathway that modulates this hypoxia inducible kinase activity are unknown, al- though several candidate kinases have been identi- fied by in vitro studies? 2

RPA is not the only protein involved in replicon initiation that could be a target for hypoxia. Once large T antigen has catalyzed the unwinding of the

I Ill ~ III I

F i g u r e 2. Hyperphosphorylation of replication proteinA (RPA) and cell-cycle arrest by hypoxia. (A) Under normal replication conditions RPA is a high-affinity DNA single- strand binding protein and is found phospho171ated on its 34-kDa subunit. (B) When cells are exposed to low oxygen conditions, the 34 kDa subunit of RPA is found to be hyperphosphoITlated. (C) The increase in RPA's hyperphos- phorylation correlates with a decrease in its ability to bind single-stranded DNA in vitro.

Page 3: Hypoxic stress proteins: Survival of the fittest

48 AmatoJ. Giaccia

double helix, DNA polymerase a/primase synthe- sizes primers that will permit the initiation of leading strand DNA synthesis and for initiation of each Okazaki fragment during lagging strand synthe- sis. 23,24 Once primers have been synthesized, replica- tion factor C (RFC, not to be confused with RPA above) displaces pol a/primase and binds with prolif- erating cell nuclear antigen (PCNA) to act as a "brace-clamp" complex to recruit DNA polymerase 3. 24,25 At this point, initiation of DNA synthesis is complete and elongation of nascent DNA strands proceeds. Recent studies by several groups have shown that PCNA will bind to two different inhibi- tors: the cyclin-cdk inhibitor p2126 and the growth arrest and DNA damage inducible gene GADD45. 27 Although both p21 and GADD45 proteins will bind with high-affinity to PCNA, only GADD45 mRNA has been found to be inducible by low oxygen condi- tions. 28 The binding of PCNA with an inhibitor, such as GADD45, prevents it from binding to RFC and creating the necessary brace-clamp structure needed to interact with DNA polymerase, thereby resulting in an inhibition of DNA synthesis. To implicate GADD45 in the inhibition of replicon initiation by hypoxia, future studies will be required to determine if GADD45 protein levels are induced by hypoxia to levels that will physiologically inhibit PCNA activity.

Cell-cycle arrest induced by low oxygen conditions is independent of the activity of the p53 tumor suppressor gene (Fig 3). Previous work in our labora-

G~/S Ratio

AG1522 2184B RC.neo.1RC.IOHL-60 PC-3 DU-145 NIH 3T3

I II II 1 Wild-Type p53 WT pS3 Mutant p53

+ E6

Figure 3. The influence of p53 on cell-cycle arrest by hypoxia. Comparison of the G~ to S-phase ratio of cells under aerobic conditions or after a 14-hour exposure to hypoxia. Cells are grouped according to their p53 geno- type: wild-type, mutant, or wild-type expressing HPV-E6 gene. In all human cells examined to date, the GI/S-phase ratio increases under hypoxia, indicative of a cell-cycle arrest. Cell-cycle distributions were determined as previ- ously described by Graeber et al. 11 N, Untreated cells; m, hypoxia-treated cells.

tory '1 indicated that, like ionizing radiation, low oxygen conditions caused the nuclear accumulation of p53 protein, increased p53's DNA binding, and transactivation activity in cells that are wild-type for p53, but not in cells that possess mutant p53. Unlike ionizing radiation, the accumulation of cells in Gl- phase by hypoxia was not solely dependent on wild- type p53 function at low oxygen conditions (0.02% oxygen). These results suggest that there are mul- tiple pathways le~ding to cell-cycle growth arrest, and that the signal transduction pathway for growth arrest induced by hypoxia is at least partially distinct from the signal transduction pathway for growth arrest induced by ionizing radiation.

Detection of Growth Arrest in Experimental Mouse Tumors

The kinetics of tumor cell labeling in situ after the administration of thymidine analogs, such as Budr or Iudr, which can be detected by histochemical meth- ods, have clearly shown that tumor cells closely bordering the lumen of a blood vessel are actively proliferating. 29 Cells that are greater than 100 to 150 ~m from the blood vessel lumen do not actively incorporate thymidine analogs and are not cycling. This recent data using immunohistochemical meth- ods to detect proliferating cells supported the oxygen diffusion model by Tannock, which proposed that chronically hypoxic cells in a tumor stopped cycling and were arrested in GI/S phase? ~ Instead ofimmu- nohistochemistry, cell proliferation in Tannock's stud- ies was assessed by the incorporation of [3HI thymi- dine into DNA and subsequent autoradiography. As even further proof that hy-poxic cells in vivo arrest in late Gl/early S-phase, Siemann and Keng injected Hoechst 33342 into mice bearing KHT sarcomas, dissociated the tumors and used flow cytometric analysis to examine the cell-cycle profiles of hypoxic cells and oxic cells based on the amount of Hoechst dye uptake.31 The power of this technique rests in the ability to directly compare the cell-cycle profiles of the most "dim" chronically hypoxic cells with the "brightest" oxic cells. Their results demonstrated that the most dim hypoxic cells had the smallest percentage of cells in S-phase, but the largest in G1 compared with bright oxic cells. As cells incorporated more Hoechst, which presumably reflected an in- creased oxygen uptake, their cell-cycle profile showed more cells in S-phase and a corresponding decrease in cells in Gi phase. Therefore, by three different but related approaches, we can conclude that tumor cells in vivo also exhibit the same biochemical events that

Page 4: Hypoxic stress proteins: Survival of the fittest

Hypoxia's Role in Cell Proliferation 49

lead to inhibition of replicon initiation in early S-phase.

What is the clinical significance of these results? Cell-cycle checkpoints are a mechanism the cell uses to maintain its genomic integrity. Because cells in S-phase of the cell-cycle are the most sensitive to killing by hypoxia, the early S-phase checkpoint induced by hypoxia, in part, serves to protect the cell from the deleterious effects of entering S-phase under adverse environmental conditions. Therefore, "forcing" hypoxic cells to enter S-phase and initiate DNA replication should lead to celt death. In vitro studies have shown that the supplemental addition of deoxycytidine and uridine will partially reverse the inhibition of replicon initiation by low oxygen condi- tions. 13 However, the use of exogenously supplied deoxynucleosides to force cells into S-phase is not a practical approach in vivo. A second approach de- scribed by Koong et aP 2 was to induce proliferative activity of hypoxic cells by treating them with mito- gen stimulating factors, such as phorbol esters that will activate a family of kinases known as the mitogen activated protein kinases (MAP kinases) or extracel- lular signal regulated protein kinases (ERKs). These kinases respond to diverse extracellular signals, espe- cially growth factors and cytokines, and increase cell proliferation by activating the appropriate transcrip- tional regulatory factors. 33-35 Koong et al showed that chronically hypoxic cells were extremely sensitive to treatment by agents, such as phorbol esters and thapsigargin, which induced proliferation (Fig 4). As these studies were performed with mitogenic agents that can also act as tumor promotors, the identifica- tion of small molecules or growth-factor receptor activating-antibodies, which could be used instead of these tumor promoting agents, would be required for any application of such an approach in the clinic.

Hypoxia-Induced Apoptos is Is Enhanced by t h e Act iv i ty of t h e p53 Tumor Suppressor Gene

Role of p53 in Modulating Apoptosis in Minimally Transformed Cells

The induction of p53 by low oxygen conditions has suggested a potential role for hypoxia in tumorigen- esis. If one looks at the well-studied model of colo- rectal tumorigenesis, it is relatively clear that p53 mutations occur late in the transformation process. 36 Therefore, at some time during the neoplastic pro- cess, a colorectal tumor will possess both homozygous wild-type and mutant p53 cells. 36 Eventually, the

tumor becomes more or less a clonal expansion of the mutant p53 tumor genotype. Thus, the question has been raised as to what physiological stimulus in a tumor selects for the growth advantage of mutant p53 tumor cells over wild-type p53 tumor cells. 37,38 Implicit in this statement is that some agent in a tumor is able to cause the activation of wild-type p53 that results in the growth restriction or elimination of wild-type p53 tumor cells. Under the same condi- tions, mutant p53 cells do not respond to this agent and continue growing. Although ionizing radiation is a strong inducer of wild-type p53, it is not a physiologi- cal stress that would be normally found in a develop- ing tumor. 39 The most well-documented physiologi- cal stress, which could cause restrictive growth and elimination of wild-type, but not mutant p53 cells in a tumor, is hypoxia.

The supporting evidence for this hypothesis is as follows. Areas of low oxygen do not occur in a tumor until it contains regions further than 150 Ixm from the nearest blood vessel. 4~ Therefore, the presence of hypoxia correlates temporally with the timing of mutations in p53, a late event in several tumor types. Secondly, hypoxia, as we discussed above, is a potent inducer of wild-type p53 activity. II Thirdly, a strong relationship has been found between wild-type p53 activity and stress-induced apoptosis. 41,42 In one model system, genetically matched cell lines differing only in their p53 status were transfected with the E 1A and Ha-ras oncogenes to increase cell proliferation and then exposed to serum deprivation or chemothera- peutic agents. Cell lines homozygous ( + / + ) or heterozygous ( + / - ) for wild-type p53 died an apop- totic death when exposed to these agents both in vitro and in vivo. 41,43 In contrast, homozygous-deleted ( - / - ) p53 cells survived these cellular insults and only a small fraction of the cells died an apoptotic death. Interestingly, clones that were derived from the wild-type populations that survived these above described stresses were now found to be homozygous for mutant/inactive p53.43'Hypoxia fits the necessary criteria for a selective stress, as it induces apoptosis in a p53-dependent manner and the magnitude of this selection would be large enough to enhance the probability for mutant p53 tumor cells to become the predominant genotype in a tumor. 44 Furthermore, developing tumors derived from oncogenically trans- formed cells that possess wild-type p53 contain sub- stantially more apoptosis than tumors lacking wild- type p53. 44 Apoptosis in tumors that contained wild-type p53 was found distal to blood vessels and

Page 5: Hypoxic stress proteins: Survival of the fittest

50 AmatoJ. Giaccia

Hypoxia Alone

C

Hypoxia + Thapsigargin

Hypoxia + PDA Hypoxia + TPA

F i g u r e 4. Increased cytotoxicity of human tumor A549 cells treated with hypoxia and protein kinase C activating agents as described by Koong et al. 32 (A) Colony forming ability ofA549 cells exposed to 19 hours of hypoxia (Hypoxia Alone); (B) cells exposed to the combination of 19 hours of hypoxia and 0.5 hours to 2 Ixmol/L of thapsigargin, an inhibitor of the endoplasmic reticulum Ca2+-ATPase (Hypoxia + Thapsigargin); (C) cells exposed to 19 hours of hypoxia and 0.5 hours to the phorbol ester 4c~-phorbol-12-13-didecanoate (PDA), a phorbol ester that does not bind nor activate protein kinase C (Hypoxia + PDA); (D) cells exposed to 19 hours of hypoxia and 0.5 hours to 1 Ixmol/L of 12-O-tetradecanoylphorbol-13- acetate (TPA), a phorbol ester that binds and strongly activates PKC (Hypoxia + TPA).

co-localized with hypoxic regions 44 (Fig 5), as de- tected by binding of a hypoxic specific marker EF5. 45 Apoptotic cells in tumors lacking wild-type p53 showed no strong relationship with their distance from a blood vessel, and apoptosis did not colocalize with the hypoxic marker EF544 (Fig 5). These results have

important therapeutic implications. Because the same genes that regulate hypoxia-induced apoptosis also regulate radiation and chemotherapy-induced apop- tosis, 41,46,47 hypoxia-mediated selection of cells with diminished apoptotic potential could also explain the poor response of solid tumors to therapy. 43

Page 6: Hypoxic stress proteins: Survival of the fittest

51 Hypoxia's Role in Cell Proliferation

B

Figure 5. Relationship between hypoxic and apoptotic regions in tumors differing in their p53 status. Tumors were derived from mouse embryonic fibroblasts that were transformed with adenovirus E1A and Ha-ras oncogenes. In situ detection of fragmented DNA by TUN-EL staining (fluorescein-green) of a p53 +/+ tumor (A) and a p53 - / - tumor (C). Hypoxic regions were detected by EF5 staining (orange-red) of the same p53 +/+ tumor (B) and p53 - / - tumor (D) sections. (A) and (C) show both TUNEL and EF-5 staining by using a dual-band pass fluorescence set for direct comparison of apoptotic and hypoxic regions.

Roles of Human PapiUomavirus E6 and E7 in Hypoxia-Induced Apoptosis

Genetic analysis of the human papillomavirus (HPV) genomes of high-risk types have identified two genes, E6 and E7, as being solely sufficient to cause immor- talization of keratinocytes. 48,49 The oncogenic poten- tial of the E6 and E7 genes has been attributed in large part to their interactions with p535~ and Rb 51 proteins, respectively. Although the induction of apoptosis by hypoxia is strongly enhanced by wild- type p53 activity, it is presently unknown whether hypoxia can induce apoptosis in cells expressing the combination of E6 and E7 from high-risk types of HPV. Evidence that supports the functionally impor- tant role both proteins play in apoptosis was reported by two different groups studying retina and lens development. 52,53 In one series of experiments, trans- genic mice were generated in which HPV E6, E7 or both genes were targeted to the lens. Transgenic mice that only expressed HPV E7 in the lens exhib- ited inhibition of lens differentiation, unregulated

cell proliferation, and apoptosis. In mice that ex- pressed both E6 and E7, apoptosis was suppressed and lens tumors developed. 53 In a second series of experiments, when E7 was targeted to developing photoreceptors in the retina, mice that possessed wild-type p53 would undergo apoptosis in the retina, whereas mice lacking wild-type p53 activity would develop retinoblastoma. 52 Because in many ways E7 is functionally equivalent to the adenovirus E 1A gene in priming cells for apoptosis, its expression should make minimally transformed cells prone to apoptosis when exposed to low oxygen conditions. The un- known question is whether minimally transformed cells expressing both E6 and E7 will die by apoptosis, cell-cycle arrest, exhibit genomic instability, or a combination of all three when subjected to low oxygen conditions. We favor the hypothesis that HPV cells early in the development of cervical carci- noma will die by apoptosis if hypoxia is able to uncouple wild-type p53 from E6-mediated degrada- tion as evidenced by Graeber e t al.ll If this hypothesis

Page 7: Hypoxic stress proteins: Survival of the fittest

52 AmatoJ. Giaccia

proves to be correct, it may explain why the presence of hypoxia correlates with a poor outcome of cervical cancers independent of treatment modality. 54,55 As the majority of cexvical carcinomas possess a high- risk type HPV and approximately half of these tumors possess oxygen poor regions, hypoxia could select for oncogenically transformed cervical epithe- lial cells that have lost their apoptotic ability and respond poorly to therapy.

Induction of Apoptosis by p53 Independent Pathways

Although wild-type p53 is strongly associated with certain types of apoptosis induced by oncogenic overexpression, there is mounting evidence for the existence of both p53-dependent and independent pathways. 56 If p53 is a proximal step in apoptosis, then it may be possible to induce apoptosis by activating a more distal step in the apoptotic path- way. For example, androgen withdrawal by castra- tion induces apoptotic cell death of prostatic glandu- lar cells in mice that are either homozygous wild-type or homozygous null for p53. 57 Furthermore, the induction of apoptosis is not an "all or none" phenom- enon. If one examines the difference in apoptosis induced by hypoxia between cells expressing wild- type p53 and those not expressing wild-type p53, one finds that wild-type p53 cells are extremely sensitive to induction ofapoptosis, but that a small percentage of mutant p53 cells do undergo apoptosis. This latter finding seems to be particularly true for transformed human cells. 58,59 These results indicate that stress activated apoptotic pathways in oncogenically trans- formed cells can be activated by other modulators in addition to wild-type p53 or different thresholds of damage are needed to induce apoptosis independent of p53. A second family of genes represented by bcl-2 or bcl-x proto-oncogenes have been shown to inhibit both p53-dependent and independent apoptosis. Mechanistically, bcl-2 protection from apoptosis was thought to involve the scavenging of free radicals, but this is certainly not the case under hypoxic condi- tions.60, 61

R e g u l a t i o n of A n g i o g e n i c G r o w t h Factors by Low O x y g e n C o n d i t i o n s

Tumor Angiogenesis, the Net Difference Between Negative and Positive Regulators

What is the relationship between hypoxia and angio- genesis? The answer to this question lies in under- standing what induces a tumor to become angio-

genic. Human tumors are not highly angiogenic when they initially develop and, furthermore, can remain without vascularization for months to years. 62 However, for the tumor to grow and become malig- nant or metastatic, neovascularization is required. Therefore, a very likely hypothesis is that tumor regions poor in oxygen possess cells that secrete angiogenic mitogens to overcome their lack of oxy- gen. According to the current model, hypoxia would only be a potent' inducer of angiogenesis when negative growth regulators such as thrombospondin 1 (TSP1) have decreased in expression. 63 Therefore, understanding how the tumor microenvironment modulates intracellular signaling pathways involved in angiogenic growth factor expression would give us new targets in cancer therapy. As in the case with the loss of tumor suppressor genes, the ability to reacti- vate or modulate angiogenic suppressors in tumor cells may present us with a specific mechanism to inhibit tumor cell growth.

Negative Regulators of Angiogenesis The balance between positive regulators and nega- tive regulators of angiogenesis is stringently con- trolled in normal tissues. However, in neoplastic tissue, this fine balance is tipped in the direction of neovascularization. A net increase in angiogenic growth signals is achieved by attenuation or loss of angiogenic suppressors such as TSP-1. In normal untransformed fibroblasts, TSP-1 is constitutively expressed at a high level. However, tumors derived from transformed cells possess a significantly re- duced amount ofTSP-1 compared with their untrans- formed counterparts. Cells derived from Li-Frau- meni patients serve as a good model system to investigate the relationship between TSP1 activity and wild-type p53 activity. These individuals who are born heterozygous for wild-type p53 possess an in- creased risk of losing their remaining wild-type allele and developing a tumor. Using fibroblasts derived from Li-Fraumeni patients, Dameron et a164 showed that late passage fibroblasts from these individuals who lost their remaining wild-type p53 allele exhib- ited a marked decrease in TSP-1 levels. If they restored wild-type p53 activity to cells that lost their endogenous p53 activity, they could restore TSP-1 expression to similar levels found in early passage cells. More recently, it has been reported that TSP-1 possesses two p53 regulatory elements, suggesting that it can be modulated by wild-type p53 activity. In untransformed cells, little TSP-1 would be needed, as these cells are under growth factor regulation. In

Page 8: Hypoxic stress proteins: Survival of the fittest

Hypoxia's Role in Cell Proliferation 53

contrast to untransformed fibroblasts, elevated TSP- 1 levels in minimally transformed cells, stimulated by increased levels of wild-type p5365 would help control tumor growth. Therefore, early events in oncogenic transformation seem to poise the cell to resist the switch to the angiogenesis phenotype by activating p53 to enhance the expression of anti-angiogenic agents.

Let us speculate on how hypoxia may effect malignant progression through the selection of cells that have lost wild-type p53 activity. Hypoxia acti- vates intracellular signaling pathways needed for the transcriptional induction of positive effectors of anglo- genesis, but at the same time, it also increases the activity of anti-angiogenic regulators such as TSP- 1, presumably through the transcriptional activity of p53. To allow minimally transformed cells to com- pletely make the switch to the angiogenic phenotype, hypoxia exerts a selective pressure for the clonal outgrowth of minimally transformed cells that have lost their apoptotic ability. As a consequence of this clonal outgrowth, the expression of negative angio- genic regulators such as TSP-1 will be attenuated or lost as wild-type p53 expression is lost. Therefore, the net result of all these interactions will be the selec- tion for transformed cells that have reduced apop- totic potential and are primed for the switch to the angiogenic phenotype (Fig 6).

Positive Regulators of Angiogenesis Compared with our dirth of knowledge about nega- tive regulators of angiogenesis, the identity and mechanism of action of angiogenesis activators is voluminous. Positive angiogenic activators can be divided into two groups--growth factors and cyto- kines. Two of the most potent angiogenic polypep- tides are fibroblast growth factor (FGF) and vascular endothelial cell growth factor (VEGF). 66 Both the acidic and basic forms of fibroblast growth factor stimulate proliferation of cells that embryonically originated from mesoderm and neuroectoderm, 67,6a promote endothelial cell migration, 6a and modulate

Normal

Selection for cells [ a" ptohrtetdu;edentlal] Hyperplasia ~ Pre-Mallgnant ~ Malignant Growth ~ Metastasis

Growth l t . y . o x , a ~ l~;'re;;:~o~;'J::l

Figure 6. Schematic of how hypoxia enhances malig- nant progression.

protease and plasminogen expression. 69,7~ In contrast to FGF, VEGF is an endothelial cell specific growth factor. 71,72 Both FGF and VEGF activate proteases that are needed to degrade the basement membrane to allow for the formation of new blood vessels. 7~

Hypoxia increases the transcriptional activity of VEGF, 74'75 FGF, 76 and platelet derived growth factor PDGF] 7,7a Recent studies have brought forth evi- dence that the induction ofVEGF gene expression by hypoxia is through a nuclear protein designated hypoxia inducible factor-1 (HIF-1), 7a-al This factor was originally identifed in the transcriptional regula- tion of the erythoprotein (EPO) gene by hypoxia, but has since been found in many cell types that do not express EPO. a2 In addition, numerous genes encod- ing glycolytic enzymes such as phosphoglycerate kinase 1 (PGK1)possess HIF-1 regulatory sequences and are transcriptionally regulated by HIF-1 under hypoxic conditions, a~ Taken together, these results indicate that HIF-1 is a unique hypoxia inducible transcription factor that is ubiquitous in many differ- ent cell types and regulates a wide variety of genes involved in cellular homeostasis.

The signal transduction pathway involved in HIF- 1/VEGF induction by hypoxia is modulated by mem- brane associated kinases such as the src and ras proto-oncogenes. If kinase defective alleles of either src or ra t are overexpressed in cells before exposure to hypoxia, VEGF induction is inhibited a3 (N.M. Mazure, E.Y. Chen, P. Yeh, N. Mivechi, K.R. Laderoute, and AJ. Giaccia, manuscript in prepara- tion). Because the growth factor receptor poison suramin, which is presently in clinical trials, is also able to inhibit VEGF induction by hypoxia, growth factor receptors must also play a role in signaling for VEGF induction and present a realistic target for novel therapeutic strategies. Figure 7 summarizes the possible sequential activation of kinases in the signaling pathway for VEGF induction by hypoxia. Clearly, understanding the signal transduction path- way inw)lved in VEGF, FGF, and PDGF induction would present new molecular targets for cancer therapy.

Working cooperatively with growth factors, cyto- kines also aid in tumor angiogenesis in two impor- tant ways. In one way, they work as paracrine effectors. For example, tumor necrosis factor (TNF)~ is induced by hypoxia in some cell types that are resident in a tumor such as mononuclear cells. 84,a5 This localized TNF~ production will then induce endothelial cells to produce basic FGF. FGF will then stimulate mitogenesis and migration of more endo-

Page 9: Hypoxic stress proteins: Survival of the fittest

54 AmatoJ. Giaccia

HYPOXIA

Growth Factor Receptors/Src

Ras

Pl3- Kinase

Diacylglycerol

PKC

?

HIF-1/VEGF

Suramin, Genistein

N17 Ras (dorninente negative)

Wortrnannin

Calphostin C

Figure 7. Signal transduction pathway for the induction of HIF- 1/VEGF by hypoxia.

thelial cells until neovascularization takes place. The second way cytokines such as transforming growth factor (TGF)[3 function to enhance angiogenesis is by modulating the cellular production of extracellular matrix proteins, a6

Several of the cytokines mentioned above have been shown to be induced by hypoxia. In contrast to the regulation of growth factors and glycolytic en- zymes by HIF-1, cytokine regulation by hypoxia has not extensively been studied. However, as most of the cytokines implicated in angiogenesis possess regulatory sequences to which the transcription fac- tor NF-KB binds, 87 and NF-KB is directly activated under hypoxic conditions, 88,a9 it seems highly prob- able that transcriptional regulation of cytokine gene expression by hypoxia is, in part, through NF-KB regulatory elements.

How Can Our Knowledge of the Genes Involved in Angiogenesis be Therapeutically Exploited?

One approach to this problem is to inhibit the intracellular signaling pathway involved in the activa- tion of angiogenic growth factors or cytokines. One common feature to both VEGF induction by HIF-1 and cytokine induction by NF-KB is the requirement of ras signaling. 89,9~ In normal cells this signaling is transitory, but in tumor cells that possess mutated forms of ras, this activity is constitutively "turned on. ''91 Genetically, ras activity can be inhibited by dominant negative mutant alleles ofras that prevent transmission of extra cellular signals through ra~ to downstream effectors. 92,93 Although, the use of domi-

nant mutant alleles of ras would be impractical therapeutically, recent success has been made in inhibiting rm- activity in cell culture systems by blocking its ability to associate with the cell mem- brane, ras requires posttranslational modification in the form of isoprenylation by the enzyme farnesyl- protein transferase (FPTase) to associate with the cell membrane. 94,95 Several studies using different assays have shown that the present group of FPTase inhibitors are selective for farnesylated proteins and inhibit transformed cell proliferation? 6 Without ques- tion, ras represents an excellent molecular target for cancer therapy. However, the ability to selectively, and with minimal toxicity, inhibit ras activity in a tumor awaits animal testing.

A second approach that would clearly have the largest impact on cancer therapy would be to restore wild-type p53 activity. Although this idea at one time would have seemed unrealistic, several lines of evi- dence suggest that it may be possible to achieve. In tumors, p53 mutations do not occur randomly and are clustered within the central core region of the protein. 97,98 Several different groups have shown that specific p53 monoclonal antibodies or second site suppressor mutations can restore p53's DNA binding activity to some mutant forms of the protein. 99,1~176 Although both of these approaches would not be clinically useful, they do indicate that in some cases wild-type p53 activity can be restored to mutant forms of p53.

One possible therapeutic approach that is now being actively pursued is to develop a small organic molecule that would be able to tether mutant p53 to its binding sequence in a gene. Implicit in this approach is that p53's transcriptional enhancing is required for it to act as a tumor suppressor gene. However, p53's role as a transcriptional activator may not be necessary to induce apoptosis.101,J0~ There- fore, the heroic task of finding a small molecule that can restore p53's transcriptional activity could be complemented by hunting for small molecules that can restore p53's role in apoptosis.

C o n c l u s i o n s

Our appreciation for how the tumor microenviron- ment influences malignant progression is only begin- ning. Recent studies have provided us with impor- tant insights into how regions of low oxygen in a tumor can select for the clonal expansion of a transformed cell that has lost its ability to kill itself when faced with inadequate growth conditions. Al-

Page 10: Hypoxic stress proteins: Survival of the fittest

Hypoxia's Role in Cell Proliferation 55

though hypoxia can act as a selective pressure in

exper imenta l systems both in vi tro and in vivo, it still

remains to be d e t e r m i n e d whe the r hypoxia selects

for apoptot ic deficient t u m o r cells in h u m a n neopla-

sias. This ques t ion will be best addressed in neo-

plasms that progress th rough pathologically distinct

s tages dur ing thei r deve lopment , where one can

associate the loss o fapoptos i s in hypoxic regions wi th

the acquisi t ion of genet ic a l terat ions that will de-

crease apoptot ic potent ia l such as p53 muta t ion or

bcl-2 overexpression.

Al though the inhibi t ion of cell prol i ferat ion by

hypoxia is th rough a p53 independen t pathway,

wild-type p53 ac tMty is increased by cellular expo-

sure to hypoxia. In the cellular response to hypoxia,

p53 plays several impor t an t roles. One , it signals for

the induct ion o f apoptosis when oxygen tension falls

below 0.2% oxygen. Two, it may transcript ionally

increase the expression of ant i -angiogenic factors

such as TSP- 1. Wi th both of these critical regula tory

roles, the possibility o f res tor ing wild-type p53 func-

tion would be o f p a r a m o u n t impor tance in reversing

the cellular phenotype acquired by its loss.

In conclusion, the old d o g m a tha t regions of low

oxygen tension in a t u m o r would m a k e cells refrac-

tory to killing by rad io therapy and some chemothe ra -

peut ic agents requires a m e n d m e n t . Viable hypoxic

cells that are refractory to cell killing may be a

by-product of a selective process for cells that have

lost the i r endogenous sensitivity (apoptosis) to low

oxygen conditions and at the same t ime have adopted

a pro-angiogenic phenotype. Because the same genes

that modu la te hypoxia- induced apoptosis also modu-

late radio therapy and chemotherapy- induced apopto-

sis, hypoxia-media ted selection of cells wi th dimin-

ished apoptot ic potent ia l could also explain the

resistance of some tumors to these same therapies.

A c k n o w l e d g m e n t The author thanks Luis Alvarez, Eunice Chen, Thomas Graeber, Susannah Green, Albert Koong, Constantinos Koumenis, Nath- alie Mazure, and Mitchell Tsal for their contributions to the preparation of the manuscript, and Tess Gonzales for typing the manuscript.

References

1. BedfordJS, MitchellJB: The effect of hypoxia on the growth and radiation response of mammalian cells in culture. BrJ Radio147:687-696, 1976

2. Amellem O, Penersen EO: Cell inactivation and cell cycle inhibition as induced by extreme hypoxia: The possible role

of cell cycle arrest as a protection against hypoxia-induced lethal damage. Cell Prolif 24:127-141, 1991

3. Pallavicini MG, Laland ME, Miller RG, et al: Cell-cycle distribution of chronically hypoxic cells and determination of the clonogenic potential of cells accumulated in G2-M phases after irradiation of a solid tumor. Cancer Res 39:1891-1897, 1979

4. Pettersen EO, Lindmo T: Inhibition of cell-cycle progression by acute treatment with various degrees of hypoxia: Modifi- cations induced by low concentrations of misonidazole pre- sent during hypoxia. BrJ Cancer 48:809-817, 1983

5. LudlowJW, Howell RL, Smith HC: Hypoxic stress induces reversible hypophosphorylation of pRB and reduction in cyclin A abundance independent of cell cycle progression. Oncogene 8:331-339, 1993

6. Amellem O, Penersen EO: Cell cycle progression in human cells following reoxygenation after extreme hypoxia: Conse- quences concerning initiation of DNA synt~aesis. Cell Prolif 26:25-35, 1993

7. Riedinger HJ, Gekeler V, Probst H: Reversible shutdown of replicon initiation by transient hypoxia in Ehrlich ascites cells. Dependence of initiation on short-lived protein. EurJ Biochem 210:389-398, 1992

8. Gekeler V, Epple J, Kleymann G, et al:.Selective and synchronous activation of early S-phase replicons of Ehrlich ascites cells. Mol Gell Bio113:5020-5033, 1993

9. Maity A, McKenna WG, Muschel RJ: The molecular basis for cell cycle delays following ionizing radiation: A review. Radiother Oncol 31:1-13, 1994

10. Weinert TA, Kiser GL, Hartwell LH: Mitotic checkpoint genes in budding yeast and the dependence of mitosis on DNA replication and repair. Genes Dev 8:652-665, 1994

11. Graeber TG, PetersonJF, Tsai M, et al: Hypoxia induces the accumulation of p53 protein, but the activation ofa Gl-phase checkpoint by low oxygen conditions is independent of p53 status. Mol Cell Biol 14:6264-6277, 1994

12. Spiro IJ, Rice GC, Durand RE, et al: Cell killing, radiosensiti- zation and cell cycle redistribution induced by chronic hy- poxia. IntJ Radiat Oncol Biol Phys 10:1275-1280, 1984

13. Loftier M: Restimulation of cell cycle progression by hypoxic cells with deoxynucleosides requires ppm oxygen tension. Exp Cell Res 169:255-261, 1987

14. LiJJ, Kelly TJ: Simian virus 40 DNA replication in vitro. Proc Natl Acad Sci USA 81:6973-6977, 1984

15. Weinberg DH, Collins KL, Simancek P, et al: Reconstitution of simian virus 40 DNA replication with purified proteins. Proc Natl Acad Sci USA 87:8692-8696, 1990

16. Stillman B: Smart machines at the DNA replication fork. Cell 78:725-728, 1994

17. Wold MS, Kelly T: Purification and characterization of replication protein A, a cellular protein required for in vitro replication of simian virus 40 DNA. Proc Natl Acad Sci USA 85:2523-2527, 1988

18. Kenny MK, Schlegel U, Furneaux H, et al: The role of human single-stranded DNA binding-protein and its indi- vidual subunits in simian virus 40 DNA replication. J Biol Chem 265:7693-7700, 1990

19. Fairman MP, Stillman B: Cellular factors required for multiple stages of SV40 DNA replication in vitro. EMBOJ 7:1211-1218, 1988

20. Liu VF, Weaver DT: The ionizing radiation-induced replica- tion protein A phosphorylation response differs between

Page 11: Hypoxic stress proteins: Survival of the fittest

56 ArnaloJ. Giaccia

ataxia telangiectasia and normal human cells. Mol Cell Biol 13:7222-7231, 1993

21. Carty MP, Zernik-Kobak M, McGrath S, et al: UV light- induced DNA synthesis arrest in HeLa cells is associated with changes in phosphorylation of human single-stranded DNA- binding protein. EMBOJ 13:2114-2123, I994

22. Pan Z-Q, Amin AA, Gibbs E, et at: Phosphorylation of the p34 subunit of human singlc-stranded-DNA-binding protein in cyclin cyclin A-activated Gl extracts is catalyzed by cdk-cyclin A complex and DNA-dependent protein kinase. Proc Natl Acad Sci USA 91:8343-8347, 1994

23. Nethanel T, Kaufmann G: Two DNA polymerases may be required for synthesis of the lagging DNA strand of simian virus 40.J Virology 64:5912-5918, 1990

24. Tsurimoto T, Melendy T, Stillman B: Sequential initiation of lagging and leading strand synthesis by two different polymer- ase complexes at the SV40 DNA replication origin. Nature 346:534-539, 1990

25. Lee SH, KwongAD, Pan Z-Q, et al: Studies on the activator 1 protein complex, an accessory factor for proliferating cell nuclear antigen-dependent DNA polymerase delta. J Biol Chem 266:594-602, 1991

26. Waga S, Hannon GJ, Beach D, et al: The p21 inhibitor of cyclin-dependent kinases controls DNA replication by interac- tion with PCNA. Nature 369:574-578, 1994

27. Smith ML, Chen I, Zhan Q, et al: Interaction of the p53-rcgulated protein Gadd45 with proliferating cell nuclear antigen. Science 266:1376-1380, 1994

28. Price BD, Calderwood SK: Gadd45 and Gadd153 messenger I/2NA levels are increased during hypoxia and after exposure of cells to agents which elevate the levels of the glucose- regulated proteins. Cancer Res 52:3814-3817, 1992

29. Rodriguez R, Ritter MA, Fowler JF, et al: Kinetics of cell labeling and thymidine replacement after continuous infu- sion of halogenated pyrimidines in vivo. Int J Radiat Oncol Biol Plays 29:105-1 t 3, t 994

30. Tannock IF: The relation between cell proliferation and the vascular system in a transplanted mouse mammary tumour. BrJ Cancer 22:258-273, 1968

3 h Siemann DW, Keng PC: Characterization of radiation resis- tant hypoxic cell subpopulations in KHT sarcomas. (ii) Cell sorting. BrJ Cancer 58:296-300, 1988

32. Koong AC, Chen EY, Kim CY, et al: Activators of protein kinase C selectively mediate cellular cytotoxicity to hypoxic cells, and not aerobic cells. Int J Radiat Oncol Biol Phys 29:259-265, 1994

33. Cobb MH, HeplerJE, ChengM, et ah The mitogen-activated protein kinases, ERKI and ERK2. Semin Cancer Biol 5:261- 268, 1994

34. WestwickJK, Cox AD, Der CJ, et al: Oncogenic Ras activates c-Jun via a separate pathway from the activation of extracel- lular signal-regulated kinases. Proc Natl Acad Sci USA 91:6030-6034, 1994

35. Gille H, Kortenjann M, Thomae O, et al: ERK phosphoryla- tion potentiates Elk-l-mediated ternary complex formation and transactivation. EMBOJ 14:951-962, 1995

36. Baker SJ, Preisenger AC, MilburnJessupJ, eta]: p53 gene mutations occur in combination with 17p allelic deletions as late events in colorectal cancer. Cancer Res 50:7717-7722, 1990

37. PietenpolJA, Vogelstein B: No room at the p53 inn. Nature 365:17-I8, 1993

38. Vogelstein B, Kinzler KW: p53 function and dysfunction. Cell 70:523-526, 1992

39. Nelson WG, Kastan M: DNA strand breaks: The DNA template alterations that trigger p53-dependent DNA dam- age response pathways. Mol Cell Biol 14:1815-1823, 1994

40. Moulder JE, Rockwell S: Tumor hypoxia: Its impact on cancer therapy. Cancer Metastasis Rev 5:313-341, 1987

41. Lowe SW, Ruley HE,Jacks T, et al: p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 74:957- 967, 1993

42. Lowe SW, Jacks T, Housman DE, et al: Abrogation of oncogene-associate~l apoptosis allows transformation of p53- deficient cells. Proc Natl Acad Sci USA 91:2026-2030, 1994

43. Lowe SW, Bodis S, McClatchey A, et ah p53 status and the efficacy of cancer therapy in vivo. Science 266:807-810, 1994

44. Graeber TG, Osmanian C,Jacks T, et al: Hypoxia, apoptosis and p53 in tumor development. Nature (submitted)

45. Lord EM, Harwell L, Koch CJ: Detection of hypoxic cells by monoclonal antibody recognizing 2-nitroimidazole adducts. Cancer Res 53:5721-5726, 1993

46. Lotem J, Sachs L: Regulation by bcl-2, c-myc, and p53 of susceptibility to induction of apoptosis by heat shock and cancer chemotherapy compounds in differentiation-compe- tent and defective myeloid leukemic cells. Cell Growth Differ 4:41-47, 1993

47. Kitada S, Takayama S, De Riel K, et al: Reversal of chemoresistance of lymphoma cells by antisense-mediated reduction of bcl-2 gene expression. Antisense Res Dev 4:71- 79, 1994

48. Munger KW, Phelps WC, Bubb V, ct al: The E6 and E7 genes of the human papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratin- ocytes.J Viro163:4417-4421, 1989

49. Hawley NP, Vousden KH, Hubbert LDR, et al: HPV16 E6 and E7 proteins cooperate to immortalize human foreskin keratinocytes. EMBOJ 8:3905-3910, 1992

50. Werness BA, Levine AJ, HoMey PM: Association of human papillomavirus types 16 and 18 proteins with p53. Science 248:76-79, 1990

51. Chellapan S, Krauss VB, Kroger B, et ah Adenovirus EIA, simian virus 40 tumor antigen, and human papillomavirus E7 protein share the capacity to disrupt the interaction between transcription factor E2F and the retinoblastoma gene product. Proc Nat1Acad Sci USA 89:4549-4553, 1992

52. Howes KA, Ransom N, Papermaster DS, et al: Apoptosis or retinoblastoma: Alternative fates of photoreceptors express- ing the HPV-16 E7 gene in the presence or absence of p53. Genes Dev 8:1300-1310, 1994

53. Pan H, Griep AE: Altered cell cycle regulation in the lens of HPV-16 E6 or E7 transgenic mice: Implications for tumor suppressor gene function in development. Genes Dev 8:1285- 1299, 1994

54. Hockel M, Knoop C, Schlenger K, et al: Intracmural pO2 predicts survival in advanced cancer of the uterine cervix. Radiother Onco126:45-50, 1993

55. Hockel M, Schlenger K, Knoop C, et al: Oxygenation of carcinomas of the uterine cervix: Evaluation by computer- ized 02 tension measurements. Cancer Res 51:6098-6102, 1991

56. White E: Death-defying acts: A meeting review on apoptosis. Genes Dev 7:2277-2284, 1993

57. Berges RR, Furuya Y, Remington L, et al: Cell proliferation,

Page 12: Hypoxic stress proteins: Survival of the fittest

Hypoxia's Role in Cell Proliferation 57

DNA repair, and p53 function are not required for pro- grammed death of prostatic glandular cells induced by androgen ablation. Proc Natl Acad Sci USA 90:8910-8914, 1993

58. Yao K-S, Clayton M, O'Dwycr PJ: Apoptosis in human adenocarcinoma HT29 cells induced by exposure to hypoxia. JNat l Cancer Inst 87:117-122, 1995

59. Muschel RJ, Bernhard EJ, Garza L, et al: Induction of apoptosis at different tensions: Evidence that oxygen radicals do not mediate apoptotic signaling. Cancer Res 55:995-998, 1995

60. Shimizu S, Eguchi Y, Kosaka I-I, et al: Prevention of hypoxia- induced cell death by Bcl-2 and Bcl-xL. Nature 374:811-813, 1995

61. Jacobson MD, RaffMC: Programmed cell death and Bcl-2 protection in very low oxygen. Nature 374:814-8 l 6, 1995

62. FolkmanJ: What is the evidence that tumors arc angiogenic? J Natl Cancer Inst 82:4-6, 1990

63. Rastinejad F, Polverini PJ, Bouck NP: Regulation of the activity of a new inhibitor of angiogenesis by a cancer suppressor gene. Cell 56:345-355, 1989

64. Dameron KM, Volpert OV, Tainsky MA, et al: Control of angiogenesis in fibroblasts by p53 regulation of Thrombospon- din- 1. Science 265:1582-1584, 1994

65. Lowe SW, Ruley HE: Stabilization of the p53 tumor suppres- sor is induced by adenovirus 5 E IA and accompanies apopto- sis. Genes Dev 7:535-545, 1993

66. Klagsbrun M, D'Amore PA: Regulators of angiogenesis. Annu Rev Physio153:217-239, 199 l

67. Ledoux D, Gannoun-Zaki L, Barritanlt D: Interactions of FGFs with target cells. Prog Growth Factor Res 4:107-120, 1992

68. Montesano R, Vassalli JD, Baird A, et al: Basic fibroblast growth factor induces angiogenesis in vitro. Proc Natl Acad Sci USA 83:7297-7301, 1986

69. Presta M, Moscatelli D, Joseph-Silverstein J, et al: Purifica- tion from a human hepatoma celt line of a basic fibroblast growth factor-like molecule that stimulates capillmT endothe- lial cell plasminogen activator production, DNA synthesis, and migration. Mol Cell Biol 6:4060-4066, 1986

70. Pepper MS, VassalliJD, Orci L, et al: Proteolytic balance and capillary morphogenesis in vitro. EXS 61:137-145, 1992

7l. Gospodarowicz D, Abraham JA, Schilling J: Isolation and characterization of a vascular endothelial cell mitogen pro- duced by pituitary-derived follicular stellate cells. Proc Natl Acad Sci USA 86:7311-7315, 1989

72. Ferrara N, Henzel WJ: Pituitary lbllicular cells secrete a novel heparin-binding growth factor specific for vascular endothelial cells. BBRC 161:851-858, 1989

73. Pepper MS, Ferrara N, Orci L, et al: Vascular endothelial growth factor (VEGF) induces plasminogen activators and plasminogen activator inhibitor-1 in microvascular endothe- lial cells. BBRC 181:902-906, 1991

74. Shweiki D, Itin A, Softer D, et al: Vascular endothelial growth factor induced by hypoxia may mediate hypoxia- initiated angiogenesis. Nature 359:843-845, 1992

75. Plate KH, Brier G, Welch HA, et al: Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo. Nature 359:845-848, 1992

76. Kuwabara K, Ogawa S, Matsumoto M, et al: Hypoxia- mediated induction of acidic/basic fibroblast growth factor and platelet-derived growth factor in mononuclear phago-

cytes stimulates growth of hypoxic endothelial cells. Proc Natl Acad Sci USA 92:4606-4610, 1995

77. Kourembanas S, Hannan RL, Failer DV: Oxygen tension regulates the expression of the platelet-derived growth lactor-B chain in human endothelial cells. J Clin Invest 86:670-674, I990

78. Semenza GL, Koury ST, Nejfelt MK, et al: Cell-type-specific and hypoxia-inducible expression of the human erythropoi- etin gene in transgenic mice. Proc Natl Aead Sci USA 88:8725-8729, 1991

79. Semenza GL, Wang GL: A nuclear factor induced by hypoxia via de novo protein synthesis binds to the hmnan erythropoi- etin gene enhancer at a site required for transcriptional activation. Mol Cell Biol 12:5447-5454, 1992

80. Semenza GL, Roth PH, Fang HM, et al: Transcriptional regulation of genes encoding glycolytic enzymes by hypoxia- inducible factor 1.J Biol Chem 269:23757-23763, 1994

81. Wang GL, Scmenza GL: Purification and characterization of hypoxia-indudble factor 1.J Biol Chem 270:1230-1237, 1995

82. Wang GL, Semenza GL: General involvement of hypoxia- inducible factor I in transcriptional response to hypoxia. Proc Natl Acad Sci USA 90:4304-4308, 1993

83. Mukhopadhyay D, Tsiokas L, Zhou X-M, et al: Hypoxic induction of" human vascular endothelial growth factor expres- sion through c-Src activation. Nature 375:577-581, 1995

84. Ghezzi P, Dinarello CA, Bianchi M, et al: Hypoxia increases production of interleukin-1 and tumor necrosis factor by human mononuclear cells. Cytokine 3:189-194, 1991

85. Scannell G, Waxman K, Kaml GJ, et al: Hypoxia induces a human macrophage cell line to release tumor necrosis factor-alpha and its soluble receptors in vitro. J Surg Res 54:281-285, 1993

86. Kuzuya M, KinsellaJL: Reorganization of endothelial cord- like structures on basement membrane complex (Matrigel): Involvement of transforming growth factor beta 1. J Cell Physiol 161:267-276, 1994

87. Baeurele PA: The inducible transcription activator NF-kB: Regulation by distinct protein subunits. Biochim Biophys Acta 1072:63-80, 1991

88. Koong AC, Chen EY, Giaccia AJ: Hypoxia causes the activation of nuclear factor-kB through the phosphorylation of IKBa on tyrosine residues. Cancer Res 54: I425-I430, 1994

89. KoongAC, Chen EY, Mivechi NF, et ah Hypoxic activation of nuclear factor-kB is mediated by a Ras and Raf signaling pathway and does not involve MAP kinasc (ERKI or ERK2). Cancer Res 54:5273-5279, 1994

90. Devary Y, Rosette C, DiDonatoJA, et al: Nf-kB activation by ultraviolet light not dependent on a nuclear signal. Science 261:1442-I445, 1993

91. Low,/DR, Willumsen BM: Fuhction and regulation of ras. Ann Rev Biochem 62:851-891, 1993

92. Stacey DW, Feig LA, Gibbs JB: Dominant inhibitory ras mutants selectively inhibit the activity of either cellular or oncogenic ras. Mol Cell Biol 11:4053-4064, 1991

93. Quilliam LA, Kato K, Rabun KM, et al: Identification of residues for Ras(17N) Growth-inhibitory phenotype and for ras interaction with guanine nucleotide exchange factors. Mol Ccll Bio114:l 113-112 i, 1994

94. HancockJF, Magee AI, ChildsJE, et al: All ras proteins are polyisoprenylated but only some are palmitoylated. Cell 57:1167-1177, I989

95. HancockJF, Cadwallader K, Paterson H, et al: A CAAX or a

Page 13: Hypoxic stress proteins: Survival of the fittest

58 AmatoJ. Giaccia

CAAL motif and a second signal are sufficient for plasma membrane targeting of ras proteins. EMBOJ 10:4033-4039, 1991

96. Gibbs JB, OliffA, Kohl NE: Farnesyltransferase inhibitors: Ras research yields a potential cancer therapeutic. Cell 77:175-178, 1994

97. Hollstein M, Rice K, Greenblatt MS, et al: Database of p53 gene somatic mutations in human tumors and cell lines. Nucleic Acids Res 22:3551-3555, 1994

98. Greenblatt MS, Bennett WP, Hollstein M, et al: Mutations in the p53 tumor suppressor gene: Clues to cancer etiology and molecular pathogenesis. Cancer Res 54:4855-4878, 1994

99. Hupp TR, Meek DW, Midgley CA, et al: Activation of the

cryptic DNA binding function of mutant forms of p53. Nucleic Acids Res 21:3167-3174, 1993

100, Halazonetis TD, Kandil AN: Conformational shifts propa- gate the oligermization domain of p53 to its tetramerLx DNA binding domain and restore DNA binding to select p53 mutants. EMBOJ 12:5057-5064, 1993

101. Wagner AJ, Kokontis JM, Hay N: Myc-mediated apoptosis requires wild-type p53 in a manner independent of cell-cycle arrest and the ability of p53 to induce p21 waf-1/cipl. Genes Dev 8:2817-2830, 1994

102. Caelles C, Helmberg A, Karin M: p53-dependent apoptosis in the absence of' transcriptional activation of p53 target genes. Nature 370:220-223, 1994


Top Related