Akt mediates an angiogenic switch in transformed keratinocytes Carmen Segrelles, Sergio Ruiz, Mirentxu Santos, Jesus Martı ´nez-Palacio, M.Fernanda Lara and Jesus M.Paramio 1 Department of Cell and Molecular Biology, CIEMAT, Av. Complutense 22, E-28040 Madrid, Spain 1 To whom correspondence should be addressed Email: [email protected]Akt signaling is involved in tumorigenesis via a number of different mechanisms that result in increased proliferation and decreased apoptosis. Previous data have demonstrated that Akt-mediated signaling is functionally involved in keratinocyte transformation. This work investigates the involvement of angiogenesis as a mediator of tumorigenesis in Akt-transformed keratinocytes. Tumors produced by subcutaneous injection of the latter showed increased angiogenic profiles associated with increased vascular endothelial growth factor (VEGF) protein levels. However, in contrast to v-ras Ha -transformed keratinocytes, VEGF mRNA levels were not increased. The induction of VEGF protein by Akt is associated with increased phosphoryla- tion and thus activation of p70S6K and eIF4E-binding protein 1, leading to increased VEGF translation. In addi- tion, we observed increased metaloproteinases 2 and 9 expression, but not thrombospondin 1, in tumors derived from Akt-transformed keratinocytes. Collectively, these results demonstrate that Akt is an important mediator of angiogenesis in malignant keratinocytes through a post- transcriptional mechanism. Introduction The mouse skin carcinogenesis model has provided an impor- tant instrumental framework for understanding many of the concepts currently applied to human neoplasia (reviewed in refs 1,2). The importance of Akt-dependent signaling in mouse skin tumorigenesis has recently been demonstrated as Akt activity increases in parallel with the process of tumor progression and precedes that of MAPK/ERK (3). In addition, over-expression of Akt, which leads to increased Akt activity, exacerbates the tumorigenic behavior of murine keratinocytes (increased pro- liferation, decreased apoptosis and impaired differentiation) (3). In agreement, specific ablation of PTEN tumor suppressor gene in the epidermis leads to the formation of spontaneous epider- mal tumors and increased sensitivity to chemical mouse skin carcinogenesis (4). Further, in experiments with transgenic mice, the ectopic expression of keratin K10—which inhibits Akt activation (5,6)—also results in dramatic inhibition of tumor development (5,7). These observations are in agreement with the current view that Akt functions in tumorigenesis in association with increased proliferation and survival of the transformed cells (8,9). Besides alterations in proliferation and apoptosis, the induc- tion of angiogenesis is essential in tumor growth since the generation of new vessels allows rapid tumor expansion and increases the likelihood of metastatic events. The acquisition of the angiogenic phenotype during tumorigenesis, the so- called angiogenic switch (10), is thought to be induced by a change in the balance of positive and negative regulators of endothelial cell growth (10). Among these, vascular endothelial growth factor (VEGF) is thought to be one of the major angiogenesis factors in malignant tumor growth (11–13). In the mouse skin carcinogenesis model, angiogenesis is an early event. The development of papillomas is preceded by a burst of angiogenesis (14). In addition, the activation of Ha-ras, the major critical event in tumor initiation in this system, plays a major role in the tumor angiogenic response inducing VEGF expression (15,16). The importance of VEGF in mouse skin carcinogenesis is demonstrated by accelerated tumor development in transgenic mice expressing VEGF (15), and by the rescue of tumor growth inhibition caused by func- tional EGFR abrogation promoted by VEGF expression (17). In addition to VEGF, other important factors such matrix metaloproteinases 2 (MMP2) and 9 (MMP9), thrombospon- dins 1 and 2 (TSP1, TSP2) have also recently emerged as important regulators of tumoral angiogenesis in mouse skin carcinogenesis (18–24). There is evidence that suggests the PI3K/PTEN/Akt pathway may be involved in tumor angiogenesis (reviewed in ref. 25). This appears to proceed mainly through the regulation of VEGF expression (26–28) and TSP1 (29). However, the precise involvement of the different elements of the pathway and the molecular mechanisms resulting in the angiogenic response are not well understood. Given the reported essential role of Akt in mouse keratinocyte transformation (3), the aim of the present study was to investigate whether Akt signaling also regulated tumor angiogenesis in this system. The results show that Akt modulates the angiogenic profile and VEGF up-regulation by a post-transcriptional mechanism associated with increased p70S6K and eIF4E-binding protein 1 (4E-BP1) phosphorylation. Further, Akt-induced tumors also show increased MMP2 and MMP9 expression, but no alteration in TSP1. This provides evidence that Akt plays a central role in the establishment of stromal changes leading to skin tumoral growth. Materials and methods Cell culture, transfection and in vivo tumorigenic assays Mouse PB keratinocytes (30) and Akt-transfected derivatives (as 40–80 pooled clones) were grown and subcutaneously injected into nu/nu mice as reported Abbreviations: HIF1a, hypoxia-inducible factor 1; MMP2, metaloproteinase 2; MMP9, metaloproteinase 9; TSP1, thrombospondin 1; TSP2 thrombo- spondin 2; VEGF, vascular endothelial growth factor. Carcinogenesis vol.25 no.7 # Oxford University Press 2004; all rights reserved. 1137 Carcinogenesis vol.25 no.7 pp.1137–1147, 2004 DOI: 10.1093/carcin/bgh132 by guest on June 2, 2013 http://carcin.oxfordjournals.org/ Downloaded from
Transcript
Akt mediates an angiogenic switch in transformed keratinocytes
Carmen Segrelles, Sergio Ruiz, Mirentxu Santos, Jes�usMartõÂnez-Palacio, M.Fernanda Lara and Jes�usM.Paramio1
Department of Cell and Molecular Biology, CIEMAT, Av. Complutense 22,E-28040 Madrid, Spain
Akt signaling is involved in tumorigenesis via a number ofdifferent mechanisms that result in increased proliferationand decreased apoptosis. Previous data have demonstratedthat Akt-mediated signaling is functionally involved inkeratinocyte transformation. This work investigates theinvolvement of angiogenesis as a mediator of tumorigenesisin Akt-transformed keratinocytes. Tumors produced bysubcutaneous injection of the latter showed increasedangiogenic profiles associated with increased vascularendothelial growth factor (VEGF) protein levels. However,in contrast to v-rasHa-transformed keratinocytes, VEGFmRNA levels were not increased. The induction of VEGFprotein by Akt is associated with increased phosphoryla-tion and thus activation of p70S6K and eIF4E-bindingprotein 1, leading to increased VEGF translation. In addi-tion, we observed increased metaloproteinases 2 and 9expression, but not thrombospondin 1, in tumors derivedfrom Akt-transformed keratinocytes. Collectively, theseresults demonstrate that Akt is an important mediator ofangiogenesis in malignant keratinocytes through a post-transcriptional mechanism.
Introduction
The mouse skin carcinogenesis model has provided an impor-tant instrumental framework for understanding many of theconcepts currently applied to human neoplasia (reviewed in refs1,2). The importance of Akt-dependent signaling in mouse skintumorigenesis has recently been demonstrated as Akt activityincreases in parallel with the process of tumor progression andprecedes that of MAPK/ERK (3). In addition, over-expressionof Akt, which leads to increased Akt activity, exacerbates thetumorigenic behavior of murine keratinocytes (increased pro-liferation, decreased apoptosis and impaired differentiation) (3).In agreement, specific ablation of PTEN tumor suppressor genein the epidermis leads to the formation of spontaneous epider-mal tumors and increased sensitivity to chemical mouse skincarcinogenesis (4). Further, in experiments with transgenicmice, the ectopic expression of keratin K10Ðwhich inhibits
Akt activation (5,6)Ðalso results in dramatic inhibition oftumor development (5,7). These observations are in agreementwith the current view that Akt functions in tumorigenesis inassociation with increased proliferation and survival of thetransformed cells (8,9).
Besides alterations in proliferation and apoptosis, the induc-tion of angiogenesis is essential in tumor growth since thegeneration of new vessels allows rapid tumor expansion andincreases the likelihood of metastatic events. The acquisitionof the angiogenic phenotype during tumorigenesis, the so-called angiogenic switch (10), is thought to be induced by achange in the balance of positive and negative regulators ofendothelial cell growth (10). Among these, vascular endothelialgrowth factor (VEGF) is thought to be one of the majorangiogenesis factors in malignant tumor growth (11±13).
In the mouse skin carcinogenesis model, angiogenesis is anearly event. The development of papillomas is preceded by aburst of angiogenesis (14). In addition, the activation ofHa-ras, the major critical event in tumor initiation in thissystem, plays a major role in the tumor angiogenic responseinducing VEGF expression (15,16). The importance of VEGFin mouse skin carcinogenesis is demonstrated by acceleratedtumor development in transgenic mice expressing VEGF (15),and by the rescue of tumor growth inhibition caused by func-tional EGFR abrogation promoted by VEGF expression (17).In addition to VEGF, other important factors such matrixmetaloproteinases 2 (MMP2) and 9 (MMP9), thrombospon-dins 1 and 2 (TSP1, TSP2) have also recently emerged asimportant regulators of tumoral angiogenesis in mouse skincarcinogenesis (18±24).
There is evidence that suggests the PI3K/PTEN/Akt pathwaymay be involved in tumor angiogenesis (reviewed in ref. 25).This appears to proceed mainly through the regulation ofVEGF expression (26±28) and TSP1 (29). However, theprecise involvement of the different elements of the pathwayand the molecular mechanisms resulting in the angiogenicresponse are not well understood. Given the reported essentialrole of Akt in mouse keratinocyte transformation (3), the aimof the present study was to investigate whether Akt signalingalso regulated tumor angiogenesis in this system.
The results show that Akt modulates the angiogenic profileand VEGF up-regulation by a post-transcriptional mechanismassociated with increased p70S6K and eIF4E-binding protein1 (4E-BP1) phosphorylation. Further, Akt-induced tumors alsoshow increased MMP2 and MMP9 expression, but no alterationin TSP1. This provides evidence that Akt plays a central role inthe establishment of stromal changes leading to skin tumoralgrowth.
Materials and methods
Cell culture, transfection and in vivo tumorigenic assays
Mouse PB keratinocytes (30) and Akt-transfected derivatives (as 40±80 pooledclones) were grown and subcutaneously injected into nu/nu mice as reported
previously (3). In some cases, prior to subcutaneous injection, the cells wereeither mock- or retrovirus-infected as described previously (31). Helper-freeretrovirus coding for v-rasHa was obtained from C2(ras) cells (32). Tumorgrowth was monitored with an external caliper for up to 7 weeks, measure-ments being made every 2 days [volume was calculated as p(4/3)� (width/2)2 �(length/2)]. One hour prior to death, mice were subcutaneously injected withBrdU (0.1 mg/g weight in 0.9% NaCl) to monitor proliferation. Tumorsamples were excised and processed for histopathology, RNA or proteinanalysis. Similarly sized pieces from different tumors with similarstroma-tumor content, as determined by histopathology, and obtained at thesame time points upon subcutaneous injection, were used in biochemicalanalyses. Proliferation and apoptosis measurements in the tumor sampleswere performed by double immunofluorescence using anti K5 (to detecttumor cells) and either anti BrdU mAb or TUNEL (Roche, Mannhein,Germany) essentially as described (3).
For luciferase assays plasmid pSV-Renilla was obtained from Promega andpVGEF-Luc KpnI (33) from ATCC. PB cells were transfected using Superfect(Qiagen) with pSV-Renilla, pVGEF-Luc KpnI and empty vector (pcDNA3) orplasmid coding from wt Akt or Ha-ras (Val12). Lysates were prepared 36 hafter transfection and analyzed with the Dual Luciferase Reporter Assay system(Promega, Madison, WI). Relative luciferase expression was determined as theratio of firefly to Renilla luciferase activity. Transfections were performed intriplicate, and the mean and standard error were calculated for each condition.PD98059, Wortmanin and rapamycin were purchased from Sigma (St Louis,MO) and used at 15 mM, 10 and 1 nM, respectively. These drugs were added tothe transfected cells 24 h after transfection and the cultures were incubated fora further 24 h. At this time lysates were prepared and analyzed as above.Fluorescence-activated cell sorter (FACS) analysis was performed withmethanol-fixed cells. DNA content was estimated with propidium iodide,and the cell cycle profile was analyzed by using multicycle software.
Immunohistochemical analysis of blood vessels and microvessel counting
Frozen tumor samples embedded in OCT (Tissue Tek, Sakura, Zoeterwoude,The Netherlands) and sectioned (6 mm thick) were fixed in acetone at ÿ20�Cfor 5 min and the blood vessels visualized using rat anti-mouse CD31(Pharmingen, diluted 1:30). To monitor vessel maturation, sections were alsostained with smooth muscle a-actin using a mouse monoclonal (Sigma diluted1/400). Microvessel counting was performed in the five areas of greatestvascularization in each tumor sample and expressed as an average. The areacovered by the vessels and the mean vessel area was determined (using Micro-image 4 software, Olympus, Silver Spring, MD) in the same fields screened forvessel counts after digitizing the images. For each type of tumor (either controlor Akt), vessel counts and vessel-covered areas were recorded as the means offour to six individual tumor samples.
Northern blotting
Total RNA from different tumors or from pooled cultured clones (20±40different clones each) was isolated by the guanidine isothiocyanate±phenol±chloroform extraction and probed by northern blotting (15 mg/lane) for VEGFexpression using mouse-specific VEGF probes as described previously (15).Equal loading was confirmed by hybridization with a 7S RNA probe. Quanti-fication was performed using a Phosphorimager and Quantity One software(Bio-Rad, Hercules, CA).
Immunohistochemistry
The detection of CD � 31, a-actin, MMP-2, MMP-9, VEGF and TSP1 wasperformed in frozen sections in parallel with antibodies reacting with mousekeratins (either rabbit polyclonal against K5 or mouse monoclonal against K14),essentially as described previously (3) using specific antibodies (see below).Secondary antibodies were purchased from Jackson Immunoresearch andused as described (3,5,6,34). Observations were made with a Zeiss Axiophotphotomicroscope equipped with epifluorescence illumination and the corres-pondent filters to avoid cross channel contamination. At least four tumors ofeach type were analyzed. Controls omitting primary antibodies, or after the pre-incubation of the antibodies with the immunizing peptide (when available),were routinely performed.
Protein extracts and western blotting
Protein extracts were obtained from different tumors and used in westernblotting (3). The following antibodies were used: Akt1 and 4E-BP1 (SantaCruz Biotechnology), hypoxia-inducible factor 1 (HIF1a) (35,36) (Transduc-tion Labs), phosphorylated Akt (Ser473; Cell Signaling, Beverly, MA), phos-phorylated p70S6Kinase [Thr389 (37); Thr421/Thr424 (38); Cell Signaling],phosphorylated 4E-BP1 [Ser65; Cell Signaling (39)], MMP-2 (39), MMP-9(41), VEGF (42) and TSP1 (43) (Neomarkers, Fremont, CA). WestPicoSignal(Pierce, Rockford, IL) was used to detect the bands according to the manufac-turer's recommendations. Quantification was performed using Quantity Onesoftware (Bio-Rad).
Results and discussion
Akt induces tumorigenesis and tumor angiogenesis inkeratinocytes
The PB keratinocyte cell line was obtained from a chemicallyinduced mouse skin papilloma (30). However, it does not dis-play increased EGFR expression nor mutations in Ha-ras gene(17,30). As a consequence, it is poorly tumorigenic in xeno-graft experiments, displaying long latency, reduced growth rateand the tumors obtained show highly differentiated phenotypes[Figure 1A±D; see also (3,17,30)]. These characteristics, aspotentially initiated cells, make this cell line a very useful modelto analyze changes in possible increased tumorigenic proper-ties upon a limited number of experimental alterations. In thisregard, increased expression of Akt, leading to increased kinaseactivity [see below and (3,44)], promoted a dramatic enhance-ment in tumorigenic behavior, with reduced latency, increasedgrowth rate and less differentiated phenotypes (Figure 1A, B,C0 and D; see also ref. 3). Analysis of these tumors clearlydemonstrated that Akt expression leads to increased prolifera-tion and reduced apoptosis (Figure 1E) in agreement with ourprevious data (3). However, the increased proportion of cells inS phase upon Akt transfection in vitro (Figure 1F) does notseem to account for the growth rate observed in the tumorsin vivo (Figure 1B) and suggests the existence of othermechanisms that support tumor development.
The induction of angiogenesis is essential in tumor growthsince the generation of new vessels allows rapid tumor expan-sion providing the environment necessary to allow the unre-strained growth of tumor cells and to prevent necrosis andcorrelates with aggressiveness. The `angiogenic switch' fromvascular quiescence to up-regulation of angiogenesis has beenobserved in the early stages of skin carcinogenesis (14). Ourprevious data demonstrating increased Akt activation in paral-lel with the process of tumor progression (3) could suggestthat such early Akt activation might also parallel angiogenesis.In agreement, we observed a reddish appearance in Akt-induced tumors, compared with a pale aspect of control tumors(Figure 2A).
To confirm the possible changes in angiogenesis we analyzedthe tumor-associated vascularization. Frozen sections werestained for the endothelial junction molecule CD31 (14). Incontrol tumor samples, the vessels appeared clear and lacunar(Figure 2B, arrows), similar to those observed in papillomasand highly differentiated squamous carcinomas and suggestiveof immaturity and poor functionality (14). On the contrary,Akt-induced tumors showed an increased number of bloodvessels, which appeared also to be narrower than those ofcontrols (Figure 2B0) and similar to those observed in poorlydifferentiated carcinomas, which have been associated with amore mature and functional stage (14). To verify this sugges-tion, double labeling against keratin (to label tumor cells) andsmooth muscle a actin was performed. This is a well-knownmarker of mature vessels (45,46). The results (Figure 2C and C0)confirm that the tumors generated by Akt-transfected kerati-nocytes show a higher grade of blood vessel maturation com-pared with controls. To further substantiate these observationsin a quantitative manner, computer-assisted morphometricimage analysis was performed in the CD31 stained sections.This revealed that both vessel density and the relative areaoccupied by tumor blood vessels were increased in Akt-derived tumors compared with control samples (Figure 3Aand B). On the other hand, the mean size of the vessels was
C.Segrelles et al.
1138
by guest on June 2, 2013http://carcin.oxfordjournals.org/
Fig. 1. Increased tumorigenic potential of Akt-transfected keratinocytes. Pooled clones (20±40) of PB keratinocytes after transfection with pcDNA3 (opensquares) or wt Akt (closed squares) were injected subcutaneously into nude mice. The time of appearance (A) and the mean volume of tumors (B) weresubsequently monitored. (C and D) Histological appearance of tumors from control keratinocytes showing well-differentiated morphology (C), whereas Akttransfection leads to poorly differentiated (C0 and D) morphologies. (E) Increased proliferation and decreased apoptosis in tumors generated by Akt-transfected PBkeratinocytes. The proliferation of tumors (closed bars) was analyzed by the ability of the cells to incorporate BrdU, whereas apoptosis induction was measuredby TUNEL labeling (open bars). At least six fields per tumor and five tumors of each type were scored for each analysis. Data are shown as mean � SD.(F) Percentage of Akt- and pcDNA-transfected pooled clones in S-phase in vitro analyzed by FACS after being stained with propidium iodide.
Akt induces angiogenesis in keratinocytes
1139
by guest on June 2, 2013http://carcin.oxfordjournals.org/
larger in control than in Akt-induced tumors (Figure 3C).Collectively, these data show that Akt induces an angiogenicswitch in keratinocytes.
Akt-induced angiogenesis is not mediated by increasedVEGF mRNA
VEGF mediates the essential angiogenic response required formouse skin tumor progression (15±17). VEGF expression can
be regulated at the transcriptional and mRNA stability levels,leading to increased VEGF mRNA steady state (47,48). Toinvestigate if Akt can drive VEGF mRNA expression, mRNAsextracted from several control and Akt-derived tumors obtainedat the same time and containing similar tumor±stroma ratiowere used in northern blot analysis. No significant differenceswere seen among the different samples (Figure 4A and A0). Tosubstantiate this observation, the VEGF mRNA levels werealso analyzed in pooled clones from control or Akt-transfectedPB cells prior to the generation of subcutaneous tumors (Figure 4Band B0). Again, we did not observe increased expression ofVEGF mRNA by Akt transfection. Finally, as a positive controland given the reported induction of VEGF mRNA by activatedHa-ras (15) we monitored VEGF expression in tumors uponv-rasHa expression. Empty vector and Akt-transfected PBkeratinocytes were transduced with a v-rasHa-coding retro-virus. The mock or v-rasHa infected cells were subsequentlyused in subcutaneous injection experiments and VEGF mRNAexpression was then analyzed in four types of tumor (control,
Fig. 2. Increased angiogenic aspect of Akt-induced tumors. (A) Appearanceof tumors upon subcutaneous injection of pcDNA3 (right flank) orAkt-transfected keratinocytes (left flank). Note the reddish appearance ofthe Akt-induced tumors. (B and B0) Representative CD31 immunostaining(green) of tumors from control (B) and Akt-transfected keratinocytes(B0) showing increased number of more mature vessels in tumors fromAkt-transfected PB cells. (C and C0) Detection of smooth musclea-actin (green) in tumors from control (C) and Akt-transfected keratinocytes.Note that only small narrow vessels displayed positive reactivityin the control tumors. Keratin K5 (red in B, B0, C and C0) was usedto counterstain tumoral cells. Arrows in (B) denote lacunarblood vessels commonly observed in control tumors.Bars � 100 mm.
Fig. 3. Histomorphometric analysis of blood vessels of different tumors.Computer-assisted analysis of blood vessels following CD31immunostaining showing increased numbers of vessels (A) covering majorareas (B) and their narrow and smaller size (C) in Akt-derived tumorscompared with control samples. At least five tumors of each type wereanalyzed using four to six fields on each. Data are shown as mean � SD.
C.Segrelles et al.
1140
by guest on June 2, 2013http://carcin.oxfordjournals.org/
Fig. 4. Akt does not induce VEGF mRNA. Total mRNA from different tumor samples (A and C) or pooled cell clones prior to subcutaneous injection(B) was probed for the expression of VEGF by northern blotting. To control loading, a 7S probe was used in all the cases. Note that no increase in VEGFexpression was observed in Akt-derived samples, whereas v-rasHa infection promoted the increased VEGF mRNA expression both in control and Akt-derivedtumors (C). (A0±C0) Semi-quantitative analysis obtained by densitometry of the corresponding northern blot data. (D) Luciferase activity from the VEGF promoterin PB cells co-transfected with pcDNA3, Akt or Ha-ras (Val12). Note that only Ha-ras (Val12) expression leads to significant increase in VEGF promoter activity.(D) Luciferase activity from the VEGF promoter in PB cells co-transfected with Ha-ras (Val12) and treated for 24 h with the stated inhibitors. Data in(D and D0) come from three independent experiments and are shown as mean � SD.
Akt induces angiogenesis in keratinocytes
1141
by guest on June 2, 2013http://carcin.oxfordjournals.org/
control plus v-rasHa, Akt and Akt plus v-rasHa). The obtainedresults (Figure 4C and C0) confirmed the induction of VEGFmRNA expression mediated by ras, in agreement with theprevious reported data in mouse keratinocytes in vivo andin vitro (15). On the other hand, such increases were similarin all cases, irrespective of whether the cells were emptyvector- or Akt-transfected keratinocytes (Figure 4C). Collec-tively, these data indicated that Akt is not involved in VEGFmRNA expression in PB keratinocytes.
To further support this observation we monitored the activityof the VEGF promoter (33) in control or Akt-transfected PBkeratinocytes. To this we co-transfected Akt, ras or emptyvector (pcDNA3) together with a reporter plasmid coding forfirefly luciferase under the control of mouse VEGF promoter(33) and the activity of this latter was analyzed. The obtaineddata confirmed that ras, but not Akt induces the transcriptionof VEGF gene in PB keratinocytes (Figure 4D). The presentobservations disagree with the previously reported involvementof the PI3K/Akt pathway in VEGF mRNA expression (28,49).This could be due to a cell type-specific effect similar to thosereported previously in fibroblast and some epithelial cell lines(50), or to a major effect of other alternative pathways. In thisregard it has been shown that VEGF mRNA is induced by TPAtreatment in mouse skin (15) by the p38-dependent pathwayin breast cancer cells (51), and through both MAPK/ERK andPI3K/Akt in human squamous cell carcinoma cells (52).Although at present it cannot be said which of these are re-sponsible for VEGF mRNA up-regulation, the observation thatHa-ras induces VEGF mRNA in parental or Akt-transfectedkeratinocytes (Figure 4) strongly suggests that other ras-dependent pathways (53,54) different to that of Akt activationare involved. In this regard, experiments using drugs thatspecifically block particular signaling pathways indicate thatinhibition of ERK and to a lesser extent PI-3K or mTOR, leadsto a substantial reduction in the transcription of VEGF gene inPB keratinocytes (Figure 4D0), thus suggesting that in ourexperimental settings multiple Ha-ras-dependent pathwaysconverge to increase the VEGF transcription.
Akt induces VEGF, MMP2 and MMP9 protein expression
The expression of VEGF is regulated at many levels by dis-parate stimuli. In addition to mRNA steady state levels, VEGFregulation can also occur at the level of VEGF translation(55±58). Consequently, we monitored the expression of VEGFprotein in parallel with Akt and active-phosphorylated-Akt(P-Ser 473) in extracts from control- and Akt-derived tumors.In agreement with previous results (3,44), the tumors derivedfrom Akt-transfected keratinocytes also showed increasedactive Akt (Figure 5A and A0). In addition, the expression ofVEGF protein in tumors derived from Akt-transfected kerati-nocytes was increased (Figure 5A), while no increase inHIF1a protein expression was seen (Figure 5A). Given thatthe transcriptional activation of VEGF is mediated mainlyby this transcription factor, the absence of increased HIF1aprotein levels might correspond to the lack of induction ofVEGF mRNA observed (Figure 4). The transcriptional activa-tion of VEGF is mediated mainly by the transcription factorHIF1a. The present data, which show no increase inHIF1a protein (Figure 5A and A0), are therefore in agreementwith the lack of change in VEGF mRNA in cells and tumors(Figure 4).
The finding that Akt regulates the translation of VEGFprompted an investigation of the possible mechanism
underlying this effect. It has been shown that the VEGF con-tains an unusually long and structured 50 untranslated region(UTR) (59). VEGF mRNA translation can take place by aninternal ribosome entry process (56,57,60±62). In addition,VEGF mRNA translation can be regulated by the activity ofthe mRNA cap-binding protein eIF-4E, the rate-limiting mem-ber of the eIF-4F translation initiation complex (58,63,64). Bybinding the cap structure, eIF-4E recruits mRNAs to the eIF-4F complex to enable ribosome loading. mRNAs harboringlengthy, highly structured 50 UTRs, is suppressed except wheneIF-4E is engaged with the eIF-4F complex. In addition, thiscomplex is further regulated by translational repressors, the4E-BPs. It has been reported that Akt can mediate the phos-phorylation of 4E-BP1 disrupting its binding to eIF-4E (65).Furthermore, 4E-BP1 is co-ordinately regulated with p70S6Kin many circumstances, and the p70S6 kinase is also activatedby an Akt-dependent pathway (66,67). We consequently hypo-thesized that Akt might regulate VEGF expression throughincreased translation. The phosphorylation of 4E-BP1 andp70S6 kinase was thus assessed in different tumors derivedfrom control and Akt-transfected keratinocytes. A markedincrease in 4E-BP1 (on Ser 65) and p70S6K (on Thr389)phosphorylation was evident in tumors derived from Akt-transfected keratinocytes compared with those derived fromcontrol keratinocytes (Figure 5B and B0). On the other hand,similar levels of total 4E-BP1 and p70S6K proteins wereobserved, as well as phosphorylation of p70S6K on Thr421/424 (which is dependent on the raf/MAPK cascade), irrespec-tive of Akt activity (Figure 5B and B0).
At present we may not discern, whether eIF4 or p70S6K-dependent mechanism is responsible for the induced VEGFtranslation. In this regard, the presence of a prototype for theterminal oligopyrimidine (`TOP'), whose expression is con-trolled by the activity of p70S6K, in the VEGF mRNA has notbeen demonstrated, thus suggesting a major involvement of4E-BP1 phosphorylation. However, the fact that p70S6K ismore sensitive to down-regulation by glucocorticoids than is4E-BP1 (66,67), and glucocorticoids can repress skin tumordevelopment by preventing Akt activation (68,69) may sug-gest a preponderant role of p70S6K. These aspects will bestudied in the near future.
On the other hand, both p706SK and 4E-BP1 phosphorylationare rapamycin sensitive (65,70±74), thus suggesting the invol-vement of mTOR in the increased VEGF protein translation.In agreement, preliminary experiments indicate that rapamycintreatment decreases tumorigenic behavior of Akt and Ha-rastransduced keratinocytes (Segrelles et al., unpublished results).Although the potential involvement of anti-angiogenic processhas not been yet tested, these data indicate that mTOR actsdownstream of Akt and Ha-ras in keratinocyte transformation.Collectively, these data indicate that the increase in VEGF canbe attributed to the ability of Akt to regulate the mTOR signal-ing pathway, which may account for increased VEGF transla-tion. On the other hand, HIF1a expression has also beendescribed to be under the control of the PI3K/PTEN/Akt/mTOR pathway (75±77). Given the increased Akt activity inthe tumors (Figure 5, see also refs 3,44), the lack of increasedlevels of HIF1a might be due to hypoxic conditions irrespectiveof tumor origin. Indeed, it has been reported that hypoxia-dependent induction of HIF1a is moderately affected by inhibi-tors of this pathway (36,75,78,79).
The process of tumor angiogenesis is not only dependent onVEGF production but rather on the balance of positive and
C.Segrelles et al.
1142
by guest on June 2, 2013http://carcin.oxfordjournals.org/
Fig. 5. Akt induces VEGF, MMP2 and MMP9 proteins. (A) Protein extracts from different tumors obtained with Akt-transfected cells and from control cells[including well-differentiated (w-d) and poorly differentiated (p-d)] were probed by western blotting with the quoted antibodies. Note that increased Akt proteinleads to increased active Akt levels and also to increased VEGF (A), MMP2 and MMP9 (C) protein levels. Of note also is the increased phosphorylation ofp70S6K and 4E-BP1 protein promoted by this increase in Akt activity (B). No significant alterations in the protein levels of HIF1a (A) and TSP-1 (C) wereobserved. Blots were reprobed with an antibody against tubulin as a loading control. (A0±C0) Semi-quantitative analysis obtained by densitometry of thecorresponding western blot data.
Akt induces angiogenesis in keratinocytes
1143
by guest on June 2, 2013http://carcin.oxfordjournals.org/
negative regulators of endothelial cell growth (10). Amongthese factors, and besides VEGF, MMP2 and MMP9, are well-recognized inducers of tumor angiogenesis (19,20,80±82),whereas TSP1 is considered an inhibitor (20±23,83). Theirexpression was therefore monitored in different tumorsderived from control or Akt-transfected keratinocytes. Theresults indicate that Akt induced the expression of MMP2and MMP9 (Figure 5C and C0) in agreement with previousdata (18,84±86). On the other hand we found that TSP-1 levels
were similar in the different samples (Figure 5C and C0)indicating that TSP-1 levels are not modulated by Akt. Theseresults seemed to be at variance with the reported activity ofPTEN to induce TSP1 expression (29). However, it is worthnoting that PTEN may influence other PI3K-dependent path-ways besides the activation of Akt. Therefore, our data wouldindicate that TSP-1 expression depends on these alternatepathways. In this regard, it has recently been demonstratedthat the activation of Rho mediated by increased PI3K activity
Fig. 6. Immunohistochemical detection of VEGF, MMP2 and TSP1 in tumors. Frozen sections of tumors obtained after subcutaneous injection of parental (A±D),Akt-transfected (A0±D0), and control keratinocytes upon v-rasHa infection (A00±D00), were stained for VEGF (green in A, A0 and A00), MMP2 (green in B, B0and B00), MMP9 (green in C, C0 and C00) and TSP-1 (green in D, D0 and D00). Together with K14 or K5 (in red) to denote the tumoral cells. Note that the Akt andv-rasHa infection led to increased VEGF, MMP2 and MMP9 protein expression, whereas only v-rasHa infection decreased TSP1 expression. In addition, MMP2and MMP9 immunoreactivity is predominant at the tumor boundary (B0, B00, C0 and C00). Bar � 50 mm.
C.Segrelles et al.
1144
by guest on June 2, 2013http://carcin.oxfordjournals.org/
is responsible for the down-regulation of TSP-1 mediated byras in mammary and kidney cells (87).
Finally, to further substantiate these observations, the expres-sion of these molecules in parallel with VEGF was studiedimmunohistochemically in tumors derived from subcutaneousinjections of Akt-transfected and control PB keratinocytes.In agreement with the western data (Figure 5), Akt led toincreased VEGF, MMP2 and MMP9 expression (Figure 6A0,B0 and C0) compared with control samples (Figure 6A, B and C).Interestingly, the expression of MMP2 and MMP9 was foundmainly in the tumor periphery and few tumor cells, either Aktor v-rasHa (Figure 6B0, B00, C0 and C00, respectively) also displayimmunoreactivity against these molecules, thus suggesting thattumor cells are not the primary source of these MMP. On theother hand, no changes in TSP1 expression were observedbetween control and Akt-derived tumors (Figure 5D and D0),whereas v-rasHa transduction led to decreased expression(Figure 5D00).
Collectively the results presented here demonstrate that Aktis an important mediator of angiogenesis in keratinocytes,leading to an increased number of blood vessels with a moremature appearance. This seemed to proceed through apost-transcriptional process due to the activation of mTORsignaling leading to increased VEGF levels. These allow theneovascularization of the tumors allowing the recruitment ofother cell types that then may secrete other molecules such asMMP2 and MMP9 (83,88) relevant for the angiogenic switch.In addition, present data imply that drugs that reduce Akt orAkt kinase activity, or which inhibit Akt-mediated translationincrease, could be of crucial clinical interest to modulateangiogenic profile in human cancers.
Acknowledgements
We thank F.Larcher for providing VEGF probes and critical reading of themanuscript, the animal facility personnel of the CIEMAT for the excellent careof the animals, and I.de los Santos and P.Hern�andez for their work involving thehistological preparations. This work was partially supported by grant SAF2002-01037 from the MCYT and 08.1/0054/2001.1 from the CAM (to J.M.P.).
References
1.Slaga,T.J., Budunova,I.V., Gimenez-Conti,I.B. and Aldaz,C.M. (1996)The mouse skin carcinogenesis model. J. Invest. Dermatol. Symp. Proc., 1,151±156.
2.Yuspa,S.H., Dlugosz,A.A., Denning,M.F. and Glick,A.B. (1996)Multistage carcinogenesis in the skin. J. Invest. Dermatol. Symp. Proc.,1, 147±150.
3.Segrelles,C., Ruiz,S., Perez,P. et al. (2002) Functional roles of Aktsignaling in mouse skin tumorigenesis. Oncogene, 21, 53±64.
4.Suzuki,A., Itami,S., Ohishi,M. et al. (2003) Keratinocyte-specific Ptendeficiency results in epidermal hyperplasia, accelerated hair folliclemorphogenesis and tumor formation. Cancer Res., 63, 674±681.
5.Santos,M., Paramio,J.M., Bravo,A., Ramirez,A. and Jorcano,J.L. (2002)The expression of keratin k10 in the basal layer of the epidermis inhibitscell proliferation and prevents skin tumorigenesis. J. Biol. Chem., 277,19122±19130.
6.Paramio,J.M., Segrelles,C., Ruiz,S. and Jorcano,J.L. (2001) Inhibition ofprotein kinase B (PKB) and PKCzeta mediates keratin K10-induced cellcycle arrest. Mol. Cell. Biol., 21, 7449±7459.
7.Santos,M., Ballestin,C., Garcia-Martin,R. and Jorcano,J.L. (1997) Delaysin malignant tumor development in transgenic mice by forced epidermalkeratin 10 expression in mouse skin carcinomas. Mol. Carcinog., 20, 3±9.
8.Nicholson,K.M. and Anderson,N.G. (2002) The protein kinase B/Aktsignalling pathway in human malignancy. Cell Signal., 14, 381±395.
9.Brazil,D.P. and Hemmings,B.A. (2001) Ten years of protein kinase Bsignalling: a hard Akt to follow. Trends Biochem. Sci., 26, 657±664.
10.Hanahan,D. and Folkman,J. (1996) Patterns and emerging mechanisms ofthe angiogenic switch during tumorigenesis. Cell, 86, 353±364.
11.Conti,C.J. (2002) Vascular endothelial growth factor: regulation in themouse skin carcinogenesis model and use in antiangiogenesis cancertherapy. Oncologist, 7, 4±11.
12.Ferrara,N. (2002) VEGF and the quest for tumour angiogenesis factors.Nat. Rev. Cancer, 2, 795±803.
13.Harmey,J.H. and Bouchier-Hayes,D. (2002) Vascular endothelial growthfactor (VEGF), a survival factor for tumour cells: implications for anti-angiogenic therapy. Bioessays, 24, 280±283.
14.Bolontrade,M.F., Stern,M.C., Binder,R.L., Zenklusen,J.C., Gimenez-Conti,I.B.andConti,C.J. (1998) Angiogenesis is an earlyevent in thedevelop-ment of chemically induced skin tumors. Carcinogenesis, 19, 2107±2113.
15.Larcher,F., Robles,A.I., Duran,H., Murillas,R., Quintanilla,M., Cano,A.,Conti,C.J. and Jorcano,J.L. (1996) Up-regulation of vascular endothelialgrowth factor/vascular permeability factor in mouse skin carcinogenesiscorrelates with malignant progression state and activated H-ras expressionlevels. Cancer Res., 56, 5391±5396.
16.Larcher,F., Murillas,R., Bolontrade,M., Conti,C.J. and Jorcano,J.L. (1998)VEGF/VPF overexpression in skin of transgenic mice induces angiogen-esis, vascular hyperpermeability and accelerated tumor development.Oncogene, 17, 303±311.
17.Casanova,M.L., Larcher,F., Casanova,B., Murillas,R., Fernandez-Acenero,M.J., Villanueva,C., Martinez-Palacio,J., Ullrich,A., Conti,C.J.and Jorcano,J.L. (2002) A critical role for ras-mediated, epidermal growthfactor receptor-dependent angiogenesis in mouse skin carcinogenesis.Cancer Res., 62, 3402±3407.
18.Park,B.K., Zeng,X. and Glazer,R.I. (2001) Akt1 induces extracellularmatrix invasion and matrix metalloproteinase-2 activity in mousemammary epithelial cells. Cancer Res., 61, 7647±7653.
19.Papathoma,A.S., Zoumpourlis,V., Balmain,A. and Pintzas,A. (2001) Roleof matrix metalloproteinase-9 in progression of mouse skin carcinogen-esis. Mol. Carcinog., 31, 74±82.
20.Bergers,G., Brekken,R., McMahon,G. et al. (2000) Matrix metalloproteinase-9triggers the angiogenic switch during carcinogenesis. Nat. Cell. Biol.,2, 737±744.
21.Bleuel,K., Popp,S., Fusenig,N.E., Stanbridge,E.J. and Boukamp,P. (1999)Tumor suppression in human skin carcinoma cells by chromosome 15transfer or thrombospondin-1 overexpression through halted tumorvascularization. Proc. Natl Acad. Sci. USA, 96, 2065±2070.
22.Hawighorst,T., Oura,H., Streit,M., Janes,L., Nguyen,L., Brown,L.F.,Oliver,G., Jackson,D.G. and Detmar,M. (2002) Thrombospondin-1selectively inhibits early-stage carcinogenesis and angiogenesis but nottumor lymphangiogenesis and lymphatic metastasis in transgenic mice.Oncogene, 21, 7945±7956.
23.Streit,M., Velasco,P., Brown,L.F., Skobe,M., Richard,L., Riccardi,L.,Lawler,J. and Detmar,M. (1999) Overexpression of thrombospondin-1decreases angiogenesis and inhibits the growth of human cutaneoussquamous cell carcinomas. Am. J. Pathol., 155, 441±452.
24.Hawighorst,T., Velasco,P., Streit,M., Hong,Y.K., Kyriakides,T.R.,Brown,L.F., Bornstein,P. and Detmar,M. (2001) Thrombospondin-2 playsa protective role in multistep carcinogenesis: a novel host anti-tumordefense mechanism. EMBO J., 20, 2631±2640.
25.Shiojima,I. and Walsh,K. (2002) Role of Akt signaling in vascularhomeostasis and angiogenesis. Circ. Res., 90, 1243±1250.
26.Zundel,W., Schindler,C., Haas-Kogan,D. et al. (2000) Loss of PTENfacilitates HIF-1-mediated gene expression. Genes Dev., 14, 391±396.
27.Huang,J. and Kontos,C.D. (2002) PTEN modulates vascular endothelialgrowth factor-mediated signaling and angiogenic effects. J. Biol. Chem.,277, 10760±10766.
28. Jiang,B.H., Zheng,J.Z., Aoki,M. and Vogt,P.K. (2000) Phosphatidylinositol3-kinase signaling mediates angiogenesis and expression of vascularendothelial growth factor in endothelial cells. Proc. Natl Acad. Sci. USA,97, 1749±1753.
30.Yuspa,S.H., Morgan,D., Lichti,U., Spangler,E.F., Michael,D., Kilkenny,A.and Hennings,H. (1986) Cultivation and characterization of cells derivedfrom mouse skin papillomas induced by an initiation-promotion protocol.Carcinogenesis, 7, 949±958.
31.Paramio,J.M., Segrelles,C., Ruiz,S., Martin-Caballero,J., Page,A.,Martinez,J., Serrano,M. and Jorcano,J.L. (2001) The ink4a/arf tumorsuppressors cooperate with p21cip1/waf in the processes of mouseepidermal differentiation, senescence and carcinogenesis. J. Biol. Chem.,276, 44203±44211.
32.Roop,D.R., Lowy,D.R., Tambourin,P.E., Strickland,J., Harper,J.R.,Balaschak,M., Spangler,E.F. and Yuspa,S.H. (1986) An activated
Akt induces angiogenesis in keratinocytes
1145
by guest on June 2, 2013http://carcin.oxfordjournals.org/
34.Santos,M., Perez,P., Segrelles,C., Ruiz,S., Jorcano,J.L. and Paramio,J.M.(2003) Impaired NF-kappa B activation and increased production of tumornecrosis factor alpha in transgenic mice expressing keratin K10 in thebasal layer of the epidermis. J. Biol. Chem., 278, 13422±13430.
35.Srinivas,V., Leshchinsky,I., Sang,N., King,M.P., Minchenko,A. andCaro,J. (2001) Oxygen sensing and HIF-1 activation does not require anactive mitochondrial respiratory chain electron-transfer pathway. J. Biol.Chem., 276, 21995±21998.
36.Arsham,A.M., Plas,D.R., Thompson,C.B. and Simon,M.C. (2002)Phosphatidylinositol 3-kinase/Akt signaling is neither required for hypoxicstabilization of HIF-1 alpha nor sufficient for HIF-1-dependent target genetranscription. J. Biol. Chem., 277, 15162±15170.
37.Weng,Q.P., Kozlowski,M., Belham,C., Zhang,A., Comb,M.J. andAvruch,J. (1998) Regulation of the p70 S6 kinase by phosphorylation invivo. Analysis using site-specific anti-phosphopeptide antibodies. J. Biol.Chem., 273, 16621±16629.
38.Pullen,N. and Thomas,G. (1997) The modular phosphorylation andactivation of p70s6k. FEBS Lett., 410, 78±82.
39.Aoki,M., Blazek,E. and Vogt,P.K. (2001) A role of the kinase mTOR incellular transformation induced by the oncoproteins P3k and Akt. Proc.Natl Acad. Sci. USA, 98, 136±141.
40.Margulies,I.M., Hoyhtya,M., Evans,C., Stracke,M.L., Liotta,L.A. andStetler-Stevenson,W.G. (1992) Urinary type IV collagenase: elevatedlevels are associated with bladder transitional cell carcinoma. CancerEpidemiol. Biomarkers Prev., 1, 467±474.
41.Nikkari,S.T., Hoyhtya,M., Isola,J. and Nikkari,T. (1996) Macrophagescontain 92-kd gelatinase (MMP-9) at the site of degenerated internalelastic lamina in temporal arteritis. Am. J. Pathol., 149, 1427±1433.
42.Boocock,C.A., Charnock-Jones,D.S., Sharkey,A.M., McLaren,J.,Barker,P.J., Wright,K.A., Twentyman,P.R. and Smith,S.K. (1995)Expression of vascular endothelial growth factor and its receptors fltand KDR in ovarian carcinoma. J. Natl Cancer Inst., 87, 506±516.
43.Dixit,V.M., Galvin,N.J., O'Rourke,K.M. and Frazier,W.A. (1986)Monoclonal antibodies that recognize calcium-dependent structures ofhuman thrombospondin. Characterization and mapping of their epitopes.J. Biol. Chem., 261, 1962±1968.
44.Leis,H., Segrelles,C., Ruiz,S., Santos,M. and Paramio,J.M. (2002)Expression, localization and activity of glycogen synthase kinase 3betaduring mouse skin tumorigenesis. Mol. Carcinog., 35, 180±185.
45.Huang,J., Frischer,J.S., Serur,A. et al. (2003) Regression of establishedtumors and metastases by potent vascular endothelial growth factorblockade. Proc. Natl Acad. Sci. USA, 100, 7785±7790.
46.Hungerford,J.E., Owens,G.K., Argraves,W.S. and Little,C.D. (1996)Development of the aortic vessel wall as defined by vascular smoothmuscle and extracellular matrix markers. Dev. Biol., 178, 375±392.
47.von Marschall,Z., Cramer,T., Hocker,M., Finkenzeller,G.,Wiedenmann,B. and Rosewicz,S. (2001) Dual mechanism of vascularendothelial growth factor upregulation by hypoxia in human hepatocel-lular carcinoma. Gut, 48, 87±96.
48.Nabors,L.B., Gillespie,G.Y., Harkins,L. and King,P.H. (2001) HuR, aRNA stability factor, is expressed in malignant brain tumors and binds toadenine- and uridine-rich elements within the 30 untranslated regions ofcytokine and angiogenic factor mRNAs. Cancer Res., 61, 2154±2161.
49.Mazure,N.M., Chen,E.Y., Laderoute,K.R. and Giaccia,A.J. (1997)Induction of vascular endothelial growth factor by hypoxia is modulatedby a phosphatidylinositol 3-kinase/Akt signaling pathway in Ha-ras-transformed cells through a hypoxia inducible factor-1 transcriptionalelement. Blood, 90, 3322±3331.
50.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 tumorangiogenesis: differential modes of vascular endothelial growth factor up-regulation in ras-transformed epithelial cells and fibroblasts. Cancer Res.,60, 490±498.
51.Xiong,S., Grijalva,R., Zhang,L., Nguyen,N.T., Pisters,P.W., Pollock,R.E.and Yu,D. (2001) Up-regulation of vascular endothelial growth factorin breast cancer cells by the heregulin-beta1-activated p38 signal-ing pathway enhances endothelial cell migration. Cancer Res., 61,1727±1732.
52.Dong,G., Chen,Z., Li,Z.Y., Yeh,N.T., Bancroft,C.C. and Van Waes,C.(2001) Hepatocyte growth factor/scatter factor-induced activation of MEK
and PI3K signal pathways contributes to expression of proangiogeniccytokines interleukin-8 and vascular endothelial growth factor in head andneck squamous cell carcinoma. Cancer Res., 61, 5911±5918.
54.Vojtek,A.B. and Der,C.J. (1998) Increasing complexity of the Rassignaling pathway. J. Biol. Chem., 273, 19925±19928.
55.Kevil,C.G., De Benedetti,A., Payne,D.K., Coe,L.L., Laroux,F.S. andAlexander,J.S. (1996) Translational regulation of vascular permeabilityfactor by eukaryotic initiation factor 4E: implications for tumorangiogenesis. Int. J. Cancer, 65, 785±790.
56.Akiri,G., Nahari,D., Finkelstein,Y., Le,S.Y., Elroy-Stein,O. and Levi,B.Z.(1998) Regulation of vascular endothelial growth factor (VEGF)expression is mediated by internal initiation of translation and alternativeinitiation of transcription. Oncogene, 17, 227±236.
57.Stein,I., Itin,A., Einat,P., Skaliter,R., Grossman,Z. and Keshet,E. (1998)Translation of vascular endothelial growth factor mRNA by internalribosome entry: implications for translation under hypoxia. Mol. Cell.Biol., 18, 3112±3119.
58.Chung,J., Bachelder,R.E., Lipscomb,E.A., Shaw,L.M. and Mercurio,A.M.(2002) Integrin (alpha 6 beta 4) regulation of eIF-4E activity and VEGFtranslation: a survival mechanism for carcinoma cells. J. Cell. Biol., 158,165±174.
59.Shima,D.T., Kuroki,M., Deutsch,U., Ng,Y.S., Adamis,A.P. andD'Amore,P.A. (1996) The mouse gene for vascular endothelial growthfactor. Genomic structure, definition of the transcriptional unit andcharacterization of transcriptional and post-transcriptional regulatorysequences. J. Biol. Chem., 271, 3877±3883.
63.Scott,P.A., Smith,K., Poulsom,R., De Benedetti,A., Bicknell,R. andHarris,A.L. (1998) Differential expression of vascular endothelial growthfactor mRNA vs protein isoform expression in human breast cancer andrelationship to eIF-4E. Br. J. Cancer, 77, 2120±2128.
64.Graff,J.R. and Zimmer,S.G. (2003) Translational control and metastaticprogression: enhanced activity of the mRNA cap-binding protein eIF-4Eselectively enhances translation of metastasis-related mRNAs. Clin. Exp.Metastasis, 20, 265±273.
65.Gingras,A.C., Kennedy,S.G., O'Leary,M.A., Sonenberg,N. and Hay,N.(1998) 4E-BP1, a repressor of mRNA translation, is phosphorylatedand inactivated by the Akt (PKB) signaling pathway. Genes Dev., 12,502±513.
66.Goncharova,E.A., Goncharov,D.A., Eszterhas,A. et al. (2002) Tuberinregulates p70 S6 kinase activation and ribosomal protein S6phosphorylation. A role for the TSC2 tumor suppressor gene inpulmonary lymphangioleiomyomatosis (LAM). J. Biol. Chem., 277,30958±30967.
67.Shi,Y., Hsu,J.H., Hu,L., Gera,J. and Lichtenstein,A. (2002) Signalpathways involved in activation of p70S6K and phosphorylation of4E-BP1 following exposure of multiple myeloma tumor cells tointerleukin-6. J. Biol. Chem., 277, 15712±15720.
68.Leis,H., Page,A., Ramirez,A., Bravo,A., Segrelles,C., Paramio,J.,Barettino,D., Jorcano,J.L. and Perez,P. (2004) Glucocorticoid receptorcounteracts tumorigenic activity of Akt in skin through interference withthe phosphatidylinositol 3-kinase signaling pathway. Mol. Endocrinol., 18,303±311.
69.Budunova,I.V., Kowalczyk,D., Perez,P., Yao,Y.J., Jorcano,J.L. andSlaga,T.J. (2003) Glucocorticoid receptor functions as a potent suppressorof mouse skin carcinogenesis. Oncogene, 22, 3279±3287.
70.Neshat,M.S., Mellinghoff,I.K., Tran,C., Stiles,B., Thomas,G., Petersen,R.,Frost,P., Gibbons,J.J., Wu,H. and Sawyers,C.L. (2001) Enhancedsensitivity of PTEN-deficient tumors to inhibition of FRAP/mTOR.Proc. Natl Acad. Sci. USA, 98, 10314±10319.
71.Podsypanina,K., Lee,R.T., Politis,C. et al. (2001) An inhibitor of mTORreduces neoplasia and normalizes p70/S6 kinase activity in Pten�/± mice.Proc. Natl Acad. Sci. USA, 98, 10320±10325.
C.Segrelles et al.
1146
by guest on June 2, 2013http://carcin.oxfordjournals.org/
72.Sekulic,A., Hudson,C.C., Homme,J.L., Yin,P., Otterness,D.M.,Karnitz,L.M. and Abraham,R.T. (2000) A direct linkage between thephosphoinositide 3-kinase-AKT signaling pathway and the mammaliantarget of rapamycin in mitogen-stimulated and transformed cells. CancerRes., 60, 3504±3513.
73.Raught,B. and Gingras,A.C. (1999) eIF4E activity is regulated at multiplelevels. Int. J. Biochem. Cell Biol., 31, 43±57.
74.Nomura,M., He,Z., Koyama,I., Ma,W.Y., Miyamoto,K. and Dong,Z.(2003) Involvement of the Akt/mTOR pathway on EGF-induced celltransformation. Mol. Carcinog., 38, 25±32.
75.Zhong,H., Chiles,K., Feldser,D., Laughner,E., Hanrahan,C.,Georgescu,M.M., Simons,J.W. and Semenza,G.L. (2000) Modulation ofhypoxia-inducible factor 1alpha expression by the epidermal growthfactor/phosphatidylinositol 3-kinase/PTEN/AKT/FRAP pathway in humanprostate cancer cells: implications for tumor angiogenesis and therapeutics.Cancer Res., 60, 1541±1545.
77.Treins,C., Giorgetti-Peraldi,S., Murdaca,J., Semenza,G.L. and VanObberghen,E. (2002) Insulin stimulates hypoxia-inducible factor 1 througha phosphatidylinositol 3-kinase/target of rapamycin-dependent signalingpathway. J. Biol. Chem., 277, 27975±27981.
78.Blancher,C., Moore,J.W., Robertson,N. and Harris,A.L. (2001) Effects ofras and von Hippel-Lindau (VHL) gene mutations on hypoxia-induciblefactor (HIF)-1alpha, HIF-2alpha and vascular endothelial growth factorexpression and their regulation by the phosphatidylinositol 30-kinase/Aktsignaling pathway. Cancer Res., 61, 7349±7355.
79.Alvarez-Tejado,M., Alfranca,A., Aragones,J., Vara,A., Landazuri,M.O.and del Peso,L. (2002) Lack of evidence for the involvement of the phos-phoinositide 3-kinase/Akt pathway in the activation of hypoxia-induciblefactors by low oxygen tension. J. Biol. Chem., 277, 13508±13517.
80.Coussens,L.M., Tinkle,C.L., Hanahan,D. and Werb,Z. (2000) MMP-9supplied by bone marrow-derived cells contributes to skin carcinogenesis.Cell, 103, 481±490.
81.Silletti,S., Kessler,T., Goldberg,J., Boger,D.L. and Cheresh,D.A. (2001)Disruption of matrix metalloproteinase 2 binding to integrin alpha vbeta 3by an organic molecule inhibits angiogenesis and tumor growth in vivo.Proc. Natl Acad. Sci. USA, 98, 119±124.
82.Pfeifer,A., Kessler,T., Silletti,S., Cheresh,D.A. and Verma,I.M. (2000)Suppression of angiogenesis by lentiviral delivery of PEX, a noncatalyticfragment of matrix metalloproteinase 2. Proc. Natl Acad. Sci. USA, 97,12227±12232.
83.Coussens,L.M., Raymond,W.W., Bergers,G., Laig-Webster,M.,Behrendtsen,O., Werb,Z., Caughey,G.H. and Hanahan,D. (1999)Inflammatory mast cells up-regulate angiogenesis during squamousepithelial carcinogenesis. Genes Dev., 13, 1382±1397.
84.Kubiatowski,T., Jang,T., Lachyankar,M.B., Salmonsen,R., Nabi,R.R.,Quesenberry,P.J., Litofsky,N.S., Ross,A.H. and Recht,L.D. (2001)Association of increased phosphatidylinositol 3-kinase signaling withincreased invasiveness and gelatinase activity in malignant gliomas.J. Neurosurg., 95, 480±488.
85.Kim,D., Kim,S., Koh,H., Yoon,S.O., Chung,A.S., Cho,K.S. and Chung,J.(2001) Akt/PKB promotes cancer cell invasion via increased motility andmetalloproteinase production. FASEB J., 15, 1953±1962.
86.Thant,A.A., Nawa,A., Kikkawa,F., Ichigotani,Y., Zhang,Y., Sein,T.T.,Amin,A.R. and Hamaguchi,M. (2000) Fibronectin activates matrixmetalloproteinase-9 secretion via the MEK1-MAPK and the PI3K-Aktpathways in ovarian cancer cells. Clin. Exp. Metastasis, 18, 423±428.
87.Watnick,R.S., Cheng,Y.N., Rangarajan,A., Ince,T.A. and Weinberg,R.A.(2003) Ras modulates Myc activity to repress thrombospondin-1expression and increase tumor angiogenesis. Cancer Cell, 3, 219±231.
88.Vu,T.H., Shipley,J.M., Bergers,G., Berger,J.E., Helms,J.A., Hanahan,D.,Shapiro,S.D., Senior,R.M. and Werb,Z. (1998) MMP-9/gelatinase B is akey regulator of growth plate angiogenesis and apoptosis of hypertrophicchondrocytes. Cell, 93, 411±422.
Received November 6, 2003; revised February 5, 2004;accepted February 20, 2004
Akt induces angiogenesis in keratinocytes
1147
by guest on June 2, 2013http://carcin.oxfordjournals.org/
