The mdm2 proto-oncogene sensitizes human medullary thyroid carcinoma cells to ionizing radiation

The mdm2 proto-oncogene sensitizes human medullary thyroid carcinomacells to ionizing radiation

Tatiana Dilla1, JesuÂ s Romero2, Pilar Sanstisteban*,1 and Juan A Velasco3,4

1Instituto de Investigaciones BiomeÂdicas Alberto Sols, Consejo Superior de Investigaciones CientõÂ®cas, Madrid, Spain;2Departamento de OncologõÂa RadioteraÂpica, Hospital Puerta de Hierro, Madrid, Spain; 3Molecular Oncology Program, CentroNacional de Investigaciones OncoloÂgicas Carlos III, Madrid, Spain

We have analysed the radiation response of a humanmedullary thyroid carcinoma cell line (MTT), character-ized by the absence of a functional p53 protein, and theconsequences of MDM2 overexpression in this process.We show that the product of the mdm2 proto-oncogene isable to sensitize MTT cells to ionizing radiation. Afterradiation treatment, MTT cells display histogramsconsistent with a G2M arrest. MTT cells expressingMDM2 (MTT-mdm2) are unable to respond to DNAdamage with G2M arrest, and display a high percentageof apoptosis. MTT-mdm2 cells show high levels of E2F-1protein, suggesting that the induction of apoptosisobserved upon MDM2 overexpression could be depen-dent on E2F-1. This observation is further supportedwith assays showing that E2F-1 binding to speci®c DNAsequences is enhanced in MTT-mdm2 cells. Likewise,transactivation of reporter constructs exclusively depen-dent on E2F-1 is also elevated after transfection withMDM2. This e�ect can be reverted by transienttransfection with p19ARF. To link the expression ofE2F-1 with the induction of apoptosis, we generatedclonal cell lines overexpressing E2F-1. Transfection withE2F-1 results in a low number of outgrowing colonieswith reduced proliferation rates, indicating that E2F-1 isdeleterious for cell growth. This negative regulationcorrelates with an increase in the percentage of the cellpopulation with DNA content below 2N, suggesting thatE2F-1 promotes apoptosis. Finally, overexpression ofE2F-1 sensitizes MTT cells to radiation exposure. Weconclude that the e�ects observed by MDM2 over-expression could be mediated by E2F-1.Oncogene (2002) 21, 2376 ± 2386. DOI: 10.1038/sj/onc/1205307

Keywords: radiation; MDM2; E2F-1; medullary thyr-oid carcinoma

Introduction

Cellular response to ionizing radiation is a complexmechanism involving multiple pathways. Over thelast years, extensive research has determined howdi�erent oncogenes and tumor suppressor genesmodify the intrinsic sensitivity of neoplastic cells toradiation treatment. A positive correlation betweenthe expression of oncogenes and the radiationresponse of various malignant cell lines has beenestablished, and it is now widely accepted that someoncogenes such as ras, raf, myc or trk play a role inmodulating the sensitivity to ionizing radiation(Sawey et al., 1987; Kasid et al., 1989; Chang etal., 1987; Pirollo et al., 1993). The establishment ofthis association may have clinical implications insupporting rational use of radiotherapy for di�erenttumor types.

Medullary carcinoma of the thyroid (MTC) is anuncommon malignancy derived from the C cells ofthe thyroid gland and accounts for up to 10% of allthyroid neoplasias. Most of the cases of MTC aresporadic, although a signi®cant percentage (around20%) are genetically determined and are associatedwith inherited syndromes known as Multiple Endo-crine Neoplasia (MEN) 2A, MEN 2B and FamilialMedullary Thyroid Carcinoma (FMTC) (Giu�ridaand Gharib, 1998; Nelkin et al., 1989, Vecchio andSantoro, 2000). Clinical management of thyroidcarcinoma relies mainly on surgery, although externalradiation treatment, alone or in combination withchemotherapeutic agents, is also applied in somecases of metastatic MTC (Heshmati et al., 1997). Theanalysis of the genetic markers related to thyroidcarcinogenesis could help to improve the managementof this disease (Puxeddu and Fagin, 2001) Severalhuman and rat MTC cell lines have been establishedand analysed both genetically and phenotypically(Carlomagno et al., 1995; Velasco et al., 1997).Among them, the MTT cell line displays most ofthe MTC markers including expression of calcitonin,somatostatin and somatostatin receptors (Medina etal., 1999). These cells, highly tumorigenic wheninjected into nude mice, carry a major rearrangementof the TP53 locus (Velasco et al., 1997).

MTT cells are also devoid of MDM2 (Dilla et al.,2000), the product of the murine double-minute

Oncogene (2002) 21, 2376 ± 2386ã 2002 Nature Publishing Group All rights reserved 0950 ± 9232/02 $25.00

www.nature.com/onc

*Correspondence: P Santisteban, Instituto de InvestigacionesBiomedicas Alberto Sols, Arturo Duperier 4, 28029 Madrid, Spain;E-mail: [email protected] address: Lilly Research Laboratories, Avda. de la Industria30, 28018 Alcobendas, Madrid, SpainReceived 15 August 2001; revised 2 January 2002; accepted 8January 2002

oncogene, initially identi®ed in spontaneously trans-formed murine ®broblasts (Cachilly-Snyder et al.,1987). The mdm2 oncogene possesses transformingactivity and it is activated by overexpression,ampli®cation or enhanced translation in di�erenttypes of human tumors, mainly in sarcomas (Olineret al., 1992). At low incidence, ampli®cation ofMDM2 has also been detected in gliomas, breastcancer and non-small cell lung carcinomas (Momandand Zambetti, 1997; Zhang and Wang, 2000). MDM2plays a pivotal role in cell proliferation throughfunctional interactions with the tumor suppressor geneproduct p53 (Momand et al., 1992). This interactionresults in the inactivation of p53 function by amechanism involving protein degradation (Haupt etal., 1997; Kubbutat et al., 1997). Besides, MDM2 alsointeracts with other proteins, such as retinoblastoma,E2F-1, the TATA box binding protein and theribosomal protein L5 (Lozano and Montes de OcaLuna, 1998).

Overexpression of MDM2 is generally linked withan advantage in cell proliferation and predispositionto tumorigenesis. In vitro, MDM2 is able totransform di�erent cell lines in combination withother oncogenes (Finlay, 1993). Likewise, transgenesencoding the MDM2 protein promote tumorigenesisin di�erent tissue environments (Ganguli et al.,2000). However, some unexpected properties ofMDM2 have been also described. When over-expressed in NIH3T3 cells, MDM2 is able to arrestthe cell cycle (Brown et al., 1998). In humanmedullary carcinoma cells, our laboratory hasrecently shown that MDM2 promotes apoptosisthrough a mechanism involving a downregulationof Bcl2 (Dilla et al., 2000). Results from tissuespeci®c transgenic mice have also shown thatMDM2 not always predispose to tumorigenesis(Alkhalaf et al., 1999), supporting the notion thatthe biological response to MDM2 overexpression iscell type speci®c.

The participation of MDM2 in response to DNAdamage has been investigated. Particularly, MDM2protein levels are elevated upon ionizing radiationexposure (Chen et al., 1994). This response seems torequire the presence of a functional p53 protein, andoccurs in parallel with other p53 downstreame�ectors such as p21Cip1/WAF1 and Bax (El-Deiry etal., 1993; Zhan et al., 1994). The biologicalconsequences of this MDM2 upregulation remainunclear. In the presence of p53, MDM2 could beresponsible for targeting p53 for degradation. How-ever, MDM2 may have additional functions in theabsence of p53, situation commonly found in humantumors. In this paper we have analyzed the responseof the human medullary thyroid carcinoma cell lineMTT to ionizing radiation, and the e�ect of MDM2overexpression in this process. The results show thatMDM2 induces a signi®cant sensitization of MTTcells to ionizing radiation. Further assays allow us toimplicate the E2F-1 transcription factor in thissensitization process.

Results

MDM2 sensitizes human medullary thyroid carcinomacells (MTT) to ionizing radiation

The e�ect of MDM2 on cell proliferation of the humanmedullary thyroid carcinoma cell line MTT has beenpreviously investigated. The isolation of clones thatconstitutively expressed MDM2 allowed us to demon-strate that this oncoprotein exerts a negative e�ect oncell growth by promoting apoptosis (Dilla et al., 2000).Here in this study we have analysed the response ofthese human MTC cells to ionizing radiation, and thee�ect of MDM2 in this response. For that purpose,mock-transfected (pC-MTT) and two independentclones of MDM2-transfected MTT cells (MTT-mdm2-c1 and -c4) were seeded at low density andsubjected to increasing doses of g-radiation (0 ± 10 Gy).Then, surviving cells were allowed to grow asindividual clones. Visual examination of the cells underthe microscope revealed that after low doses ofradiation (4 2 Gy), MTT cells still maintained atypical ®broblast-like morphology (Velasco et al.,1997), whereas colonies from MTT-mdm2 cells showedsigns of cellular fusion, condensation of nuclei andeventual nuclear disintegration (data not shown). Toquantify the e�ects of the radiation treatment, colonieswere stained and scored by clonogenic assays. Datawere analysed and the survival curves corresponding tothe linear quadratic model represented (Figure 1a). Thesurvival fraction at 2 Gy (SF2) was signi®cantly lowerin MDM2 expressing cells compared with the parentalcells (0.88 for MTT, 0.59 and 0.61 for MTT-mdm2clones c1 and c4, respectively), indicating that MDM2sensitizes MTT cells to g-radiation.

To assure that constitutively transfected cells expressMDM2, nuclear extracts from control transfectionswith empty vector and the two clones described abovewere isolated, resolved by electrophoresis and immu-noblotted with an speci®c human MDM2 monoclonalantibody. A unique polipeptide was detected at theexpected molecular mass was observed in the twotransfectants (Figure 1b).

Cell cycle distribution in MTT and MTT-mdm2 cellsafter radiation

To further analyse the radiation e�ect on cellproliferation, we determined cell cycle distribution ofthe parental (MTT) and transfected cells (MTT-mdm2clone c1 and c4) after radiation exposure (Table 1).Cultures were treated with 5 Gy radiation dose andDNA content was then analysed by ¯ow cytometry(Figure 2). As previously described, cell cycle distribu-tion of the parental cell line showed histogramscorresponding with actively dividing cells (59.21%G0±G1, 17.29% S, 21.46% G2-M). Only 2% of thecell population presented DNA content below 2N.Radiation treatment of MTT cells resulted in a G2-Marrest of the cell population. After 24 h, we observed asigni®cant reduction in the G1 fraction (59.21 to

Oncogene

Radiosensitization of human medullary thyroid cells by MDM2T Dilla et al

2377

39.27%), concomitant with an increase in the G2-Mphase (21.46 to 33.16%). This G2-M accumulationincreased reaching the maximum percentage 48 h aftertreatment (52.04%). Cell cycle distribution of MTT-mdm2 cells after radiation was also analysed. Non-irradiated controls in both clones showed a 35% cellfraction below diploid status, in keeping with pre-viously described data showing MDM2-mediatedinduction of apoptosis (Dilla et al., 2000). Twenty-fourhours after treatment, we found no signi®cant changesin the histograms. Unlike MTT cells, MDM2 expres-sing clones failed to accumulate in G2-M, and

continued showing a high proportion of hypodiploidcells. After 48 h, this percentage of apoptotic cellsincreased signi®cantly, reaching almost half of the cellpopulation (46.18 and 39.95% for c1 and c4,respectively). To demonstrate that the e�ect is a directconsequence of the radiation treatment and not due tothe evolution of MTT-mdm2 cells in culture, weanalysed cell cycle distribution of non-irradiated cellsafter 24 and 48 h in culture. Hypodiploid valuesranged again around 35% and cell cycle distributionwas undistinguishable to that obtained at the beginningof the experiment (data not shown).

To unambiguously demonstrate that hypodiploidcells detected by ¯ow cytometry correspond to cellsundergoing apoptosis, MTT and MTT-mdm2-c1 cellswere collected and labeled with annexin V (Table 2).We found only background levels of ¯uorescence incontrol and irradiated MTT cells (around 1%),con®rming that radiation does not induce apoptosisin the parental cells. Consistent with cell cyclehistograms, we observed a signi®cant percentage ofnon-irradiated MTT-mdm2 cells positive for annexinlabeling. This population of annexin positive cellsslightly increased 24 h after radiation, reaching max-imum value at 48 h, when di�erences with non-irradiated controls were signi®cant (31.3 vs 18.2). Fromthese results, we conclude that radiation enhancesapoptotic e�ects mediated by MDM2 in MTT cells.

MDM2 up-regulates E2F-1 protein levels

We investigated molecular mechanisms that couldexplain the phenotypic characteristics of MTT-mdm2cells, including growth retardation, induction ofapoptosis and sensitization to g-radiation. We searchedfor changes in regulatory proteins that upon eitherinduction or repression mediated by MDM2 could beresponsible for the promotion of apoptosis in MTTcells. Since these cells are naturally devoid of p53, wenarrowed down the analysis to proteins that couldinduce apoptosis through p53-independent pathways.

The product of the E2F-1 gene, initially identi®ed asa growth promoting transcription factor, has beenrecently described as an apoptosis inducer in di�erenttumor cell lines (Phillips et al., 1999; Hunt et al., 1997;Liu et al., 1999; Dong et al., 1999). Moreover, E2F-1induction of apoptosis does not require a functionalp53, at least in human breast and ovarian carcinomacells (Hunt et al., 1997). We tested whether in MTTcells, the e�ect observed upon MDM2 overexpressioncould be mediated by E2F-1. Figure 3 shows Westernanalysis using polyclonal antibodies for the E2F-1protein in MTT and MTT-mdm2 cell extracts (clone c1and c4). Results demonstrate that MDM2 upregulatescellular levels of E2F-1. Quanti®cation of the signalfrom three independent experiments revealed that basallevels of the E2F-1 transcription factor are increasedmore that threefold in both clones overexpressingMDM2.

To support any functional implication to theupregulation of E2F-1 detected by immunoblotting,

Figure 1 Overexpression of MDM2 sensitizes MTT cells toionizing radiation. (a) Dose response curves for the humanmedullary carcinoma cell line MTT, and MTT cells transfectedwith MDM2. Lineal quadratic ®ts are shown and representsurvival fraction vs radiation dose of three independentexperiments. (b) MDM2 protein levels in the parental andtransfected clones. Protein extracts from MTT cells, and MTTcells either transfected with an eukaryotic expression vector(pCDNA3) or a MDM2 expression vector (MTT-mdm2) wereresolved by SDS±PAGE electrophoresis and probed with aspeci®c MDM2 antibody. Membranes were stripped and probedwith an actin antibody to assess equal loading of the samples


2378

Oncogene

we tested the ability of E2F-1 to bind a speci®c DNAsequence in gel shift assays. As shown in Figure 4a,basal levels of E2F-1 binding were observed whenMTT cell extracts were incubated with an oligonucleo-tide containing a consensus E2F-1 recognition site.This signal corresponds to the endogenous levels ofE2F-1 present in MTT cells. The protein-DNAcomplex was clearly augmented in two independentclones (c1 and c4) transfected with MDM2, indicatingthat the elevation of E2F-1 protein levels detected byWestern blot in MTT-mdm2 cells correlates with agreater capacity of E2F-1 to bind DNA. In fact, theintensity of the signal was similar to that obtainedwhen an expression vector for E2F-1 was transfected inMTT cells. The speci®city of the DNA/protein complexwas con®rmed by competition assays. Extracts fromMTT cells were incubated with an excess of unlabeledE2F-1 binding oligonucleotide, or an unrelatedsequence prior to the addition of the labeled probe.Unlabeled E2F-1 binding oligonucleotide, but not theunrelated one, competed with the radiactive probe.

E2F-1 binding to speci®c DNA sequences promotestranscription of E2F-1-regulated target genes. Toascertain whether the increased DNA-protein complexresults in any e�ect in gene transcription, we carriedout transient transfection assays using expressionvector for MDM2, in combination with reporterconstructs in which expression of luciferase is drivenby three tandem sequences of E2F-1 consensus bindingsites (p36E2F-Luc). A reporter constructs carryingmutated E2F-1 sequences were used as negative control(p36E2Fmut-Luc). Results are represented in Figure4b. When p36E2F-Luc was transfected alone, trans-activation from the endogenous E2F-1 was detected inMTT cell extracts, whereas p36E2Fmut-Luc yieldedvery low values of luciferase. Transactivation ofp36E2F-Luc was clearly increased in MTT cells whencotransfected with pC-mdm2, con®rming the capacityof MDM2 to promote E2F-1 dependent transcription.As expected, no activity was observed when p36E2F-mut-Luc was used instead of p36E2F-Luc.

E2F-1 protein levels and transactivation after exposure ofMTT and MTT-mdm2 cells to ionizing radiation

We then analysed the e�ects of radiation exposure onE2F-1 protein levels and transactivation in MTT and

Table 1 Cell cycle distribution of parental and MDM2-expressing cells following radiation treatment

Radiation dose (Gy) Incubation after radiation (h) Sub G0 ±G1 G0 ±G1 S G2-M

MTT±pCDNA3 0 ± 2.04+0.71 59.21+5.44 17.29+2.08 21.46+3.155 24 2.98+0.87 39.27+6.48 22.59+2.85 33.16+3.875 48 3.34+1.36 25.69+5.97 18.93+3.12 52.04+5.81

MTT±mdm2 (cl) 0 ± 34.97+4.42 42.23+5.07 10.71+2.75 12.09+2.775 24 31.58+4.63 41.16+6.15 14.70+2.62 12.56+3.595 48 46.18+6.77 39.31+5.52 8.06+3.03 6.45+2.64

MTT±mdm2 (c4) 0 ± 32.45+5.48 46.28+5.89 9.38+4.56 11.89+3.055 24 35.81+7.12 46.80+6.14 7.94+3.19 9.45+4.045 48 39.95+8.07 45.28+5.27 7.15+4.96 7.62+5.23

Cells were collected, ®xed and analysed by ¯ow cytometry after propidium iodide staining. Experiments were performed in triplicate, and data isrepresented as mean+s.d.

Figure 2 Cell cycle distribution of asynchronous MTT andMTT-mdm2 clones following radiation treatment (5 Gy). Atdi�erent intervals after radiation exposure (0, 24 and 48 h), cellswere ®xed with ethanol, stained with propidium iodide andanalysed by ¯ow cytometry. Histograms from the parental (MTT,left panels) and one representative MDM2 transfected clone(MTT-mdm2-cl, right panels) are represented. Limits for thedi�erent phases of the cell cycle are indicated (a) subG0-G1; (b)G0-G1; (c) S (d) G2-M

Table 2 Apoptosis in MTT and MTT-mdm2 cells after radiationexposure

Incubation after Percentage of annexin V positive cellsradiation (h) MTT MTT±mdm2 cells (cl)

0 1.50+0.43 18.2+4.3524 1.29+0.29 20.3+5.8348 2.15+0.63 31.3+6.92

Cells were collected, and incubated with FITC-annexin V asdescribed in Materials and methods. Annexin V positive cells weredetected by ¯ow cytometry. Experiments were performed in triplicate,and data is represented as mean+s.d.

Oncogene


2379

MTT-mdm2 cells. Cultures were subjected to radiationtreatment (5 Gy) and extracts were collected forWestern blot after 48 h. Basal levels of E2F-1 inMTT cells were not modi®ed by radiation treatment(Figure 5a). As previously observed, E2F-1 proteinlevels in MTT-mdm2 were clearly elevated respect tocontrol cells. Similar levels of E2F-1 protein werefound in MTT-mdm2 cells exposed to radiationtreatment.

To determine whether exposure to radiation a�ectsE2F-1-dependent gene transcription, transactivationassays were performed in MTT and MTT-mdm2 cells.Twenty-four hours after transfection with p36E2F-Luc, cells were subjected to a 5 Gy dose of radiation,and 48 h later extracts were assayed for luciferaseactivity to measure E2F-1 dependent transactivation.Data was normalized for transfection e�ciency bymeasuring CAT activity driven by pRSV-CAT (Figure5b). Compared with basal transcription detected inMTT cells, radiation slightly increased E2F-1 activity,although this e�ect was not signi®cant. Consistent with

previous observations (Figure 4), E2F-1 activity wasaugmented in MTT-mdm2 cells. Radiation exposurefurther increased E2F-1 dependent transactivation.Results demonstrated that radiation enhances theability of MDM2 to up-regulate E2F-1 dependentactivity, not by increasing protein levels, but bymodulating E2F-1 transactivation capacity.

p19ARF reverts the ability of MDM2 to transactivateE2F-1

We have previously shown that MDM2 induction ofapoptosis can be partially reverted by transienttransfection with p19ARF (Dilla et al., 2000). Therefore,if E2F-1 transactivation was dependent on MDM2, itshould be reverted by p19ARF when co-expressed withMDM2. Also, p19ARF should be capable to decreasethe constitutively activated E2F-1 transcription inMTT-mdm2 cells. To test this hypothesis, MTT cellswere transfected with p36E2F-Luc and pC-mdm2 inthe absence or presence of p19ARF. To assure that bothMDM2 and p19ARF proteins are e�ciently expressed inthe transient transfection experiments, we checkedprotein levels in Western analysis using extracts fromMTT and MTT cells transfected with either pC-mdm2or pC-p19ARF. As shown in Figure 6b, p19ARF andMDM2 protein levels were clearly elevated upontransfection with the corresponding plasmids.

Results from the transfection experiments showedthat pC-mdm2 increased activity from p36E2F-Luc bythreefold. This activation was blocked by p19ARF,which reverted luciferase values to almost basal levels.In the absence of MDM2, p19ARF had little e�ect onE2F-1 transcription (Figure 6a, left). In MTT-mdm2,we measured the ability of p19ARF to modulateconstitutive activity of E2F-1. As shown in Figure 6a(right), p19ARF inhibited E2F-1-dependent transactiva-tion. From these results, we conclude that E2F-1transcriptional activity is regulated by MDM2 througha p19ARF controlled mechanism.

E2F-1 exerts a negative effect on MTT cell growth bypromoting apoptosis

The results presented above demonstrated that E2F-1protein levels are upregulated in MTT cells transfectedwith MDM2. As mentioned, several reports indicatedthat E2F-1 inhibits cell growth by promoting apoptosisin particular tumor cell lines (Oswald et al., 1994;Phillips et al., 1999; Hunt et al., 1997; Liu et al., 1999).From both observations we reasoned that in MTTcells, elevation of E2F-1 protein levels could beresponsible for the phenotypic characteristics observedin MTT-mdm2 cells, including growth retardation,apoptosis and sensitization to radiation. If this is thecase, we should be able to mimic MDM2 e�ect inMTT cells by directly upregulating E2F-1. To test thishypothesis, a mammalian expression vector carryingthe full-length human E2F-1 cDNA was transfectedinto MTT cells. Selection (G-418) was maintained for 4weeks. After that period, the e�ect of E2F-1 on MTT

Figure 3 E2F-1 transcription factor is upregulated in MTT cellsoverexpressing MDM2. Protein extracts from the parental (MTT)and MDM2 transfected cells (clones c1 and c4) were resolved bySDS±PAGE electrophoresis, probed with an E2F-1 polyclonalantibody and revealed by chemiluminiscence. After exposure, themembrane was stripped and probed with an actin antibody toassess equal loading of the samples. Quanti®cation of threeindependent experiments is shown in the lower panel, where anarbitrary value of one was assigned to the average intensity ofE2F-1 in MTT cells. SD is also shown for each bar


2380

Oncogene

growth was determined by the ability of the transfectedcells to grow and form single colonies. Representativeassays are shown in Figure 7a. Cells were stained withcrystal violet to determine colony formation in thepresence or absence of E2F-1. Compared with thecontrol transfections, expression of E2F-1 signi®cantlydecreased colony formation. Colony outgrowth inhibi-tion was comparable to that found after transfectionwith MDM2. The results indicate that E2F-1 exerts anegative e�ect on MTT cell growth.

To quantify growth retardation induced by E2F-1,we generated stable cell lines overexpressing E2F-1.As shown in Figure 7b, this cells constituvelyoverexpressed E2F-1. We then expanded and seededpC-MTT and MTT-E2F-1 to determine the prolif-eration rate of both cell lines for 4 consecutive days.After 2 days in culture, the ability of E2F-1 toinhibit cell growth was already evident, reaching50% cell growth inhibition after 4 days (Figure 7c).We then tested whether, as hypothesized, E2F-1could be promoting apoptosis of MTT cells. Wedetermined sub G0 ±G1 population in cell cyclehistograms from asynchronous MTT and MTT-E2F-1 cells by ¯ow cytometry. As shown in Figure7d, overexpression of E2F-1 signi®cantly elevatedthe proportion of subG0 ±G1 nuclei to valuessimilar to those previously found after MDM2overexpression. From all these results, we concludethat E2F-1 transcriptional activity is responsible forMDM2 induction of apoptosis.

E2F-1 sensitizes MTT cells to ionizing radiation

Finally, we analysed whether overexpression of E2F-1was able to modify the intrinsic radiation response ofMTT cells. Cells were seeded at low density andsubjected to increasing doses of g-radiation. Out-growing colonies were counted to determine the e�ecton cell survival. Results from the linear quadratic ®tsare represented in Figure 8. Analysis of the datarevealed signi®cant di�erences in the radiation survivalcurves. SF2 values determined for MTT-E2F-1 cellswere lower than those found in the controls (0.58 vs0.88).

Discussion

Intrinsic cellular response to ionizing radiation can bemodi®ed by ectopic oncogene expression, suggestingthat oncogenes play a role in the radiation response ofmammalian cells both in vivo and in vitro. Here wereport that the product of the mdm2 oncogene is ableto increase the radiation sensitivity of human tumorcells derived form medullary thyroid carcinoma.

After radiation treatment, MTT cells are accumu-lated in the G2M phase of the cell cycle. This responseto DNA damage is commonly found in many tumorcell lines regardless of the status of the p53 protein(Hwang and Muschel, 1998). Normal cells, as well astumor cells carrying wild type p53, respond to

Figure 4 Binding and transactivation assays of E2F-1 to speci®c DNA consensus sequences. (a) Gel retardation assays of nuclearextracts of MTT cells, MTT-mdm2 (clones C1 and C4) and MTT cells transfected with E2F-1 were performed by incubation of aradiolabeled E2F-1 consensus oligonucleotide. The arrow indicates the position of the protein/DNA complex. For competitionexperiments, cell extracts from MTT cells were preincubated with an excess of unlabeled E2F-1 consensus oligonucleotide (related)or and oligonucleotide with a random sequence (unrelated) (b) E2F-1 dependent transactivation in MTT is enhanced by MDM2.Equal number of cells (26106) were transfected with either a wild-type or mutated E2F-1 reporter constructs (p36E2F-Luc andp36E2Fmut-Luc), alone or in combination with an expression plasmid for MDM2 (pC-mdm2). In all cases, transfection e�ciencywas corrected with the activity of chloranfenicol acetyltransferase, driven by the pRSV±CAT construct. Values are represented asLuc/CAT activity, in arbitrary units. The mean+s.d. of three independent experiments are shown

Oncogene


2381

radiation with a G1 arrest, which allow the repairmachinery to act before DNA is duplicated in the Sphase (Hwang and Muschel, 1998; Illiakis, 1997). Infact, we have observed that when p53 is introduced inthese MTT cells, response to DNA damage is partiallyrestored, cells accumulated in G1 after radiation andsurvival is increased (Velasco and Santisteban, unpub-lished observations). On the other hand, here we showthat the introduction of MDM2 causes MTT cells tofail G2M arrest and instead, induces cell death byapoptosis. Results also show that MDM2 overexpres-sion leads to sensitization of MTT cells to radiation-induced apoptosis. This observation might explainclonogenic inhibition detected in the survival curvesafter radiation exposure.

We have explored molecular mechanisms that couldlink the expression of MDM2 with the induction of

apoptosis in MTT cells. In previous reports, we havedemonstrated that this process is associated with adown-regulation of the antiapoptotic protein bcl-2 andinvolves the upregulation of caspase 2 (Dilla et al.,2000). Experiments described in Figures 3 and 4 allowus to conclude that MDM2 overexpression leads to anincrease in E2F-1 protein levels, which in turn resultsin a greater capacity to bind speci®c DNA sequences ingel shift assays. Moreover, results obtained fromtransient transfections ruled out the possibility thatoverexpression of E2F-1 in MTT-mdm2 cells was just acasual phenomenon that occurred during the selectionprocess of the cells. As shown in Figure 4b, E2F-1-dependent transactivation is detected in MTT extracts72 h after transfection with MDM2. From all theseresults, we hypothesized that the e�ect induced byMDM2 could be in part mediated by E2F-1. Elevationof E2F-1 upon MDM2 overexpression could be eitherdue to a transcriptional mechanism or stabilization ofthe protein. Recently, the latter mechanism has beendescribed in pathways leading to apoptosis upon DNAdamage. Moreover, speci®c phosphorylation sites with-in the E2F-1 sequences have been identi®ed asresponsible for accumulation of the E2F-1 protein(Lin et al., 2001). Whether this particular mechanismfor E2F-1 accumulation occurs in MTT cells uponMDM2 overexpression remains to be investigated.

It is described that this transcription factor is crucialfor a precise cell regulation of the cell cycle (Ohtani,1999). Other studies have attributed oncogenic proper-ties to E2F-1, although more recent reports have alsodemonstrated that E2F-1 is able to promote apoptosiswhen overexpressed (Phillips et al., 1999). For instance,adenoviral transfer of E2F-1 produced generalizedapoptosis in di�erent tumor types, including breast,ovarian, head and neck carcinoma and melanoma cells(Hunt et al., 1997; Liu et al., 1999; Dong et al., 1999).Therefore, whereas the normal function of E2F-1 incell cycle progression controlling transition to S phaseis well described, the consequences of its overexpres-sion may lead to di�erent responses, favoring trans-formation or apoptosis depending on the cell type(Johnson, 2000).

The interplay between MDM2 and E2F-1 has beenanalysed recently in several reports. Initially, it wasshown that MDM2 promotes E2F-1 transactivationthough a mechanism involving a physical contact ofboth proteins (Martin et al., 1995). The observationwas made in SAOS cells, characterized by the absenceof a functional p53, suggesting that the e�ect isindependent of p53. Our results are in the context ofthis observation since (i) MDM2 leads to an incrementin E2F-1 transactivation and (ii) MTT cells are alsodevoid of p53. Other reports, although con®rm thephysical interaction between the two proteins in thesame cell line, propose a di�erent biological outcomedescribing a reduction of E2F-1 dependent activitymediated by MDM2 (Loughran and La Thange, 2000).

Radiation exposure neither changes E2F-1 proteinlevels nor transactivation activity in MTT cells. Thisresults is in keeping with the knowledge that upregulation

Figure 5 E2F-1 protein levels and activity in MTT and MTT-mdm2 cells exposed to radiation. (a) protein samples from eithernon-irradiated cells (0) or cultures exposed to 5 Gy of radiation(5) were collected after 48 h and separated by SDS±PAGE.Immunoblotting was performed using E2F-1 antibodies. (b) E2F-1-dependent transactivation in MTT and MTT-mdm2 cells afterradiation. MTT and MTT-mdm2 cells were transfected with thereporter vector p36E2F-Luc. After 24 h, plates were either non-treated or exposed to a 5 Gy dose of radiation and 48 h later,E2F-1 dependent transactivation was measured as the levels ofluciferase activity. The results were corrected for transfectione�ciency using a RSV±CAT expression plasmid. Results wereperformed by triplicate and are represented as the mean+s.d.


2382

Oncogene

of E2F-1 in response to DNA damage requires thepresence of a functional p53 (Blattner et al., 1999).However, it is of special interest the observation that inMTT-mdm2 cells, althoughE2F-1 levels are notmodi®edby radiation, transactivation from a E2F-1 speci®cconstruct was signi®cantly elevated. This fact maycontribute to the increased amount of apoptotic cellsdetected after radiation exposure.

Previous observations from our laboratory indicatedthat MDM2 induction of apoptosis is partially revertedby p19ARF (Dilla et al., 2000). The product from thistumor suppressor gene has been shown inhibit MDM2function (Zhang et al., 1998). We then tested whetherE2F-1 activity could be blocked by transient expressionof p19ARF. We show that in MTT-mdm2 cells, p19ARF

overexpression reverts E2F-1 dependent transactivationto values similar to those obtained in the parental cellline. These results are also reproduced in the parentalMTT cell line by co-transfection of MDM2 andp19ARF. The results indicate that E2F-1 transactivationactivity is directly dependent on MDM2 and mediatesMDM2-induction of apoptosis in MTT cells.

Our data also support the notion that overexpression ofE2F-1 induces sensitization of MTT cells to radiation.Either through MDM2 overexpression, or when directlytransfected in MTT cells, resulting clones exhibit greatersensitivity to radiation than parental or empty vector-transfected cells. This link between E2F-1 overexpressionand radiation sensitization is in keeping with previousreports describing sensitizing properties of E2F-1 todi�erent DNA damage (Banerjee et al., 1998). Forinstance, intrinsic overexpression of E2F-1 is found inAtaxia telangectasia cells (Varghese and Jung, 1998),which are hypersensitive to radiation. Also, ectopicexpression of E2F-1 in ®brosarcoma cells results in

increased cytotoxicity induced by radiation (Pruschy etal., 1999). As we observe in our study, E2F-1 sensitizationto radiation is independent of p53, since both reports havein common the use of cell lines devoid of wild-type p53. Itis not de®ned yet whether lack of p53 is a requirement forE2F-1 dependent apoptosis.

In summary, our results indicate that MDM2,presumably through elevation of E2F-1 protein levels,is responsible for a radiosensitization of MTT cells toionizing radiation. Screening of tumor samples and celllines in which MDM2 is overexpressed, either byincreased translation or gene ampli®cation, will con-tribute to determine whether radiosensitizing propertiesof MDM2 are restricted to particular tumor types.

Materials and methods

Cell culture

Medullary thyroid carcinoma cells MTT were derived fromthe human TT cell line (Leong et al., 1981). They carry asevere rearrangement of the TP53 locus that leads to theabsence of either p53-speci®c transcripts or protein. Thesecells were routinely maintained in RPMI1640 mediumsupplemented with 10% FBS, 2 mM glutamine, 100 mg/mlsodium piruvate, 100 U/ml penicillin and 100 mg/ml strepto-mycin. MTT cells overexpressing MDM2 and E2F-1 weremaintained in the same medium with a 200 mg/ml concentra-tion of geneticin (G-418).

Plasmids and transfections

Generation of MTT cells lines overexpressing MDM2 hasbeen described previously (Dilla et al., 2000). A mammalianexpression vector for E2F-1 (pC-E2F-1) was generated by

Figure 6 p19ARF reverts MDM2 induction of E2F-1 transactivation in MTT and MTT-mdm2 cells. (a) Transactivation assayswere performed using expression vectors for MDM2 and/or p19ARF, along with reporter plasmids for E2F-1. Transfection e�ciencywas normalized with a RSV±CAT plasmid, and data is represented as Luc/CAT activity in arbitrary units. Experiments wereperformed in triplicate, and data represented as the mean+s.d. (b) p19 and MDM2 are e�ciently expressed in the transfectionexperiments. Cell extracts from control transfections and either pC-p19ARF or pC-MDM2 cells were subjected to SDS±PAGE andblotted with speci®c antibodies for p19 or MDM2. Blots were then stripped and probed with and actin antibody for loading control

Oncogene


2383

cloning of the human full-length E2F-1 cDNA (Helin et al.,1992) into the BamHI site of the pCDNA expression vector.Transfection of both pCDNA and pC-E2F-1 (2 mg each)were performed using Lipofectin reagent (Life Technologies,Gaithersburg MD, USA). An expression vector for MDM2was also used as control. Transfected cells were selected forthree weeks with G418 (400 mg/ml). At that point, repre-sentative plates were stained with crystal violet to determinethe number of outgrowing colonies from each experimentalgroup. Other plates were used to generate established celllines overexpressing E2F-1. For that purpose, isolatedcolonies were trypsinized individually and expanded. Fortransactivation assays, transient transfections with wild-typeor mutated E2F-1 reporter constructs (Lam and Watson,1993), with or without expression vectors for MDM2 werealso carried out in MTT cell cultures. Seventy-two hours aftertransfection, cells extracts were prepared and chlorampheni-col acetyltransferase (CAT) and luciferase (Luc) weredetermined as previously described (Velasco et al., 1998).For these assays, an expression vector for p19ARF was alsoused (Kelle et al., 1995)

Radiation survival curves

Cells in exponentially growing cultures were collected bytrypsinization, washed twice in PBS and plated at di�erentdensities in 25 cm2 tissue culture ¯asks. Radiation treatments

were performed with an AECL Theratron 80 Clinical 60Coirradiator at a dose of 2 Gy/min. After treatment, colonieswere allowed to grow for 12 days, ®xed in 70% methanol andthen stained with 0.5% crystal violet in 70% methanol.Colonies containing at least 50 cells were counted todetermine cell survival. The surviving fraction was calculatedas the ratio of the plating e�ciency or irradiated cells to thatof control cells. Data from triplicate experiments weregrouped for each dose and ®tted to the linear-quadraticmodel, using the least square method (Albright, 1987). Therewas a good ®tting of the experimental values with acorrelation coe�cient (r2) over 0.9 in all cases. The survivalfraction at 2 Gy (SF2) was calculated to compare radiationsensitivity.

Growth, cell cycle distribution and detection of apotosis

Growth curves were performed in 6 cm plates, seeding 56104

cells/plate. Cell number was determined in triplicate experi-ments every 24 h for 4 consecutive days. Trypan blue stainingwas used to exclude non-viable cells. Cell cycle distributionwas determined from asynchronous cultures by FACSanalysis. Cells were trypsinized, washed in PBS and ®xed in70% ethanol. After centrifugation, cells were resuspended ina solution containing 10 mg/ml propidium iodide and 5 mg/mlRNase and analysed in a ¯ow cytometer (Becton Dickinson,San Jose CA, USA). At least 10 000 events were collected for

Figure 7 Phenotypic analysis of MTT cells transfected with E2F-1. (a) MDM2 and E2F-1 reduce colony outgrowth of MTT cellsafter transfection. Cells were transfected with either control vector (pCDNA3), MDM2 (pC-mdm2) or E2F-1 (pC-E2F-1).Neomycin resistant colonies were allowed to grow for 3 weeks and then stained with crystal violet to quantify colony number. Platesfrom representative experiments are shown. (b) Growth curves of MTT cells transfected with control vector pCDNA3 (dark circles)or an expression vector for E2F-1 (open circles). Cells were seeded at 56104 cells/plate, and the number of viable cells from threeindependent experiments was determined for 4 consecutive days. Mean and s.d. values are represented. (c) Immunodetection ofE2F1 in constitutively transfected clones. Protein extracts were separated by SDS±PAGE and probed with a E2F-1 antibody. Afterstripping, the same blot was incubated with an actin antibody for loading control. (d) E2F-1 promotes apoptosis of MTT cells. Cellswere ®xed with ethanol, stained with propidium iodide and analysed by ¯ow cytometry to determine the percentage of the subG0±G1 fractions (DNA content below 2N). Results represent the mean and s.d. from three independent determinations


2384

Oncogene

each histogram. Data from triplicate experiments wereanalysed to calculate the percentage of each cell cycle phaseusing the CellQuest software (Becton Dickinson).Apoptosis was monitored by Fluorescein (FITC)-conju-

gated Annexin V labeling (Pharmingen, San Diego CA,USA). Cells were trypsinized, washed with PBS andresuspended in 0.1 M HEPES/NaOH, pH 7.4, 1.4 M NaCl,25 mM CaCl2 containing Annexin V-FITC. After 15 minincubation at room temperature, cells were analysed by ¯owcytometry.

Western analysis

After radiation treatment, cultures were washed withphosphate bu�ered saline and disrupted with lysing bu�er

(10 mM sodium phosphate pH 7.5, 100 mM NaCl, 1% TritonX-100, 0.5% deoxycholate, 1 mM phenylmethylsulfonyl¯uor-ide, 15 mg/ml aprotinin). Equal amounts of protein lysates(20 mg) were subjected to 12.5% polyacrylamide gel electro-phoresis and then transferred to nitrocellulose membranes.After blocking membranes with 1% low fat milk in Trisbu�ered saline containing 0.05% Tween-20, immunodetectionwas performed using appropriate dilutions of the primaryantibodies. MDM2, E2F-1 and actin antibodies werepurchased from Santa Cruz Biotechnology (Santa Cruz,CA, USA). Anti-p19ARF was from AbCam (Cambridge, UK)After incubation with the appropiate secondary antibody,membranes were revealed with western blotting luminolreagent (Santa Cruz Biotechnology). Quantitation of thesignals was performed by densitometric analysis andrepresented as arbitrary optical density units.

Electrophoretic mobility shift assays

Nuclear extracts were prepared as described (Andrews andFaller, 1991) and incubated with the oligonucleotide 5'-AGGCTTGGCGGGAAAAAGAACG-3', which contains aconsensus binding site for the E2F-1 transcription factor(Oswal et al., 1994). Conditions for in vitro binding were40 mM HEPES pH 7.9, 200 mM KCl, 0.5 mM DTT, 0.2 mM

EDTA, 5% Ficoll and 3 mg/ml poly (dI-dC). After 30 min ofincubation at room temperature, DNA-protein complexeswere separated from free 32P-radiolabeled oligonucleotide ona 5% polyacrylamide gel in 0.56TBE bu�er (45 mM Tris,45 mM boric acid and 0.5 mM EDTA, pH 8.0). Thespeci®city of the DNA-protein complexes were con®rmedby competition experiments with either unlabeled E2F-1oligonucleotide or an unrelated DNA sequence. For thiscompetition experiments, unlabeled oligonucleotides werepreincubated 30 min before the addition of the radioactiveprobe.

AcknowledgmentsWe thank Dr Vicente Notario from Georgetown UniversityMedical Center for critical reading of the manuscript.Tatiana Dilla is recipient of a fellow from the Fondo deInvestigaciones Sanitarias. The work is supported bygrants from the Comunidad de Madrid (CAM 08.1/0025/1997), DireccioÂ n General de Ciencia y TecnologõÂ a (BMC-2001-2087) and FundacioÂ n Salud 2000 (Spain).

References

Albright N. (1987). Radiat. Res., 112, 331 ± 340.Alkhalaf M, Ganguli G, Messaddeq N, Le Meur M and

Wasylyk B. (1999). Oncogene, 18, 1419 ± 1434.Andrews NC and Faller DV. (1991). Nucleic Acid Res., 19,

2499.Banerjee D, Schnieders B, Fu JZ, Adhikari D, Zhao SC and

Bertino JR. (1998). Cancer Res., 58, 4292 ± 4296.Blattner C, Sparks A and Lane D. (1999). Mol. Cell. Biol.,

19, 3704 ± 3713.Brown DR, Thomas CA and Deb SP. (1998). EMBO J., 17,

2513 ± 2525.Cachilly-Snyder L, Yang-Feng T, Francke U and George

DL. (1987). Somatic Cell. Mol. Genet., 13, 235 ± 244.Carlomagno F, Salvatore D, Santoro M, de Franciscis V,

Quadro L, Panariello L, Colantuoni V and Fusco A.(1995). Biochem. Biophys. Res. Commun., 207, 1022 ± 1028.

Chang EH, Pirollo KF, Zou ZQ, Cheung HY, Lawler EL,Garner R, White E, Bernstein WB, Fraumeni JW andBlattner WA. (1987). Science, 237, 1036 ± 1038.

Chen CY, Oliner JD, Zhan Q, Fornace AJ, Vogelstein B andKastan MB. (1994). Proc. Natl. Acad. Sci. USA, 91,2684 ± 2688.

Dilla T, Velasco JA, Medina DL, Gonzalez-Palacios JF andSantisteban P. (2000). Endocrinology, 141, 420 ± 429.

Dong YB, Yang HL, Jane M, Liu TJ, Stilwell A, Atienza Cand McMasters KM. (1999). Cancer, 86, 2021 ± 2033.

El-Deiry WS, Tokino T, Velculescu VE, Levy DB, ParsonsR, Trent JM, Lin D, Mercer WE, Kinzler KW andVogelstein B. (1993). Cell, 75, 817 ± 825.

Finlay CA. (1993). Mol. Cell. Biol., 13, 301 ± 306.Ganguli G, Abecassis J andWasylyk B. (2000). EMBO J., 19,

5135 ± 5147.

Figure 8 E2F-1 sensitizes MTT cells to ionizing radiation. Doseresponse curves for MTT and MTT±E2F-1 cells are shown.Linear quadratic ®ts represent survival fraction vs. radiation doseof three independent experiments

Oncogene


2385

Giu�rida D and Gharib H. (1998). Ann. Oncol., 9, 695 ± 701.Haupt Y, Maya R, Kazaz A and Oren M. (1997). Nature,

387, 296 ± 299.Helin K, Lees JA, Vidal M, Dyson N, Harlow E and Fattaey

A. (1992). Cell, 24, 337 ± 350.Heshmati HM, Gharib H, Van Heerden JA and Sizemore

GW. (1997). Am. J. Med., 103, 60 ± 69.Hunt KK, Deng J, Liu TJ, Wilson-Heiner M, Swisher SG,

Clayman G and Hung MC. (1997). Cancer Res., 57, 4722 ±4726.

Hwang A and Muschel RJ. (1998). Radiat. Res., 150, S52 ±S59.

Illiakis G. (1997). Seminars Oncol., 24, 602 ± 615.Johnson DG. (2000). Mol. Carcinog., 27, 151 ± 157.Kasid U, Pfei�er A, Brennan T, Beckett M, Weichselbaum

RR, Dritschilo A and Mark GE. (1989). Science, 243,1354 ± 1356.

Kelle DE, Zyndi F, Ashmun RA and Sherr JC. (1995). Cell,83, 993 ± 1000.

Kubbutat MHG, Jones SN and Vousden KH. (1997).Nature, 387, 299 ± 303.

Lam EW and Watson RJ. (1993). EMBO J., 12, 2705 ± 2713.Leong SS, Horoszewicz JS, Shimaoka K, Friedman M,

Kawinski E, Song MJ, Zeigel R, Chu TM, Bayling SB andMirand EA. (1981). Advances in thyroid neoplasia.Andreoli M, Monaco F, Robbins J (eds). Rome, FieldEducational Italia: pp. 95 ± 108.

Lin WC, Lin FT and Nevins JR. (2001). Genes Dev., 15,1833 ± 1844.

Liu TJ, Wang M, Breau LR, Henderson Y, El Naggar AK,Steck KD, Siccard MW and Clayman GL. (1999). CancerGene Ther., 6, 163 ± 171.

Loughran O and La Thange NB. (2000). Mol. Cell. Biol., 20,2186 ± 2197.

Lozano G and Montes de Oca Luna R. (1998). Biochim.Biophys. Acta, 1377, M55 ±M59.

Martin K, Trouche D, Hagemeler C, Sorensen TS, LaThange NB and Kouzarides T. (1995). Nature, 375, 691 ±694.

Medina DL, Velasco JA and Santisteban P. (1999).Endocrinology, 140, 87 ± 95.

Momand J and Zambetti GP. (1997). J. Cell. Biochem., 64,343 ± 352.

Momand J, Zambetti GP, Olson DC, George D and LevineAJ. (1992). Cell, 69, 1237 ± 1245.

Nelkin BD, De Bustos A, Mabry M and Baylin SB. (1989). J.Amer. Med. Ass., 261, 3130 ± 3135.

Ohtani K. (1999). Front. Biosci., 4, D793 ±D804.Oliner JD, Kinzler KW, Meltzer PS, George DL and

Vogelstein B. (1992). Nature, 358, 80 ± 83.Oswald F, Lovec H, Moroy T and Lipp M. (1994).Oncogene,

9, 2029 ± 2036.Phillips AC, Ernst MK, Bates S, Rice NR and Vousden KH.

(1999). Mol. Cell, 4, 771 ± 781.Pirollo KF, Tong YA, Villegas Z, Chen Y and Chang EH.

(1993). Radiat. Res., 135, 234 ± 243.Puxeddu E and Fagin JA. (2001). Endocrinol Metab Clin

North Am., 30, 493 ± 513.Pruschy M, Wirbelauer C, Glanzmann C, Bodis S and Krek

W. (1999). Cell Growth Di�er., 10, 141 ± 146.Sawey MJ, Hoot TA, Burns FJ and Garte SJ. (1987). Mol.

Cell. Biol., 7, 932 ± 935.Varghese S and Jung M. (1998). Arch. Pharma. Res., 21,

640 ± 644.Vecchio G and Santoro M. (2000). Clin. Chem. Lab. Med.,

38, 113 ± 116.Velasco JA, Acebron A, Zannini M, Martin-Perez J, Di

Lauro R and Santisteban P. (1998). Endocrinology, 139,2796 ± 2802.

Velasco JA, Medina DL, Romero J, Mato ME andSantisteban P. (1997). Int. J. Cancer, 73, 449 ± 455.

Zhan Q, Fan S, Bae I, Guillouf C, Liebermann DA,O'Connor PM and Fornace Jr AJ. (1994). Oncogene, 9,3743 ± 3751.

Zhang R and Wang H. (2000). Current Pharmac. Design, 6,393 ± 416.

Zhang Y, Xiong Y and Yarbrough G. (1998). Cell, 92, 725 ±734.


2386

Oncogene

Documents

The mdm2 proto-oncogene sensitizes human medullary thyroid carcinoma cells to ionizing radiation