Overexpression of mouse Mdm2 induces developmental phenotypes in Drosophila

(2002) 21, 2413 ± 2417 ã 2002 Nature Publishing Group All rights reserved 0950 ± 9232/02 $25.00 www.nature.com/onc SHORT REPORTS Overexpression of mouse Mdm2 induces developmental phenotypes in Drosophila Adriana Folberg-Blum 1, Amir Sapir 2, Ben-Zion Shilo 2 and Moshe Oren*,1 1 Department of Molecular Cell Biology, Weizmann Institute of Science, Rehovot, 76100, Israel; 2 Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, 76100, Israel The Mdm2 proto-oncogene is ampli ed and overexpressed in a variety of tumors. One of the major functions of Mdm2 described to date is its ability to modulate the levels and activity of the tumor suppressor protein p53. Mdm2 binds to the N-terminus of p53 and, through its action as an E3 ubiquitin ligase, targets p53 for rapid proteasomal degradation. Mdm2 can also bind to other cellular proteins such as hnumb, E2F1, Rb and Akt; however, the biological signi cance of these interactions is less clear. To gain insight into the function of Mdm2 in vivo, we have generated a transgenic Drosophila strain bearing the mouse Mdm2 gene. Ectopic expression of Mdm2, using the UAS/ GAL4 system, causes eye and wing phenotypes in the y. Analysis of wing imaginal discs from third instar larvae showed that expression of Mdm2 induces apoptosis. Crosses did not reveal genetic interactions between Mdm2 and the Drosophila homolog of E2F, Numb and Akt. These transgenic ies may provide a unique experimental model for exploring the molecular interactions of Mdm2 in a developmental context. (2002) 21, 2413 ± 2417. DOI: 10.1038/sj/ onc/1205305 Keywords: Mdm2; apoptosis; Drosophila The mouse Mdm2 gene, originally cloned from a spontaneously transformed derivative of mouse 3T3 cells (Fakharzadeh et al., 1991), is frequently found ampli ed and overexpressed in a variety of tumors (for reviews on Mdm2 see Juven-Gershon and Oren, 1999; Lohrum and Vousden, 2000; Momand et al., 2000; Caspari, 2000). The oncogenicity of Mdm2 has been attributed mainly to its ability to bind the p53 tumor suppressor protein and to inhibit its transcriptional activity (Momand et al., 1992; Oliner et al., 1993). Mdm2 functions as an E3 ubiquitin ligase (Honda et al., 1997), that binds to p53 and targets it to proteasome-mediated degradation (Kubbutat et al., 1997; Haupt et al., 1997). *Correspondence: M Oren, E-mail: moshe.oren@weizmann.ac.il Received 19 July 2001; revised 2 January 2002; accepted 8 January 2002 Several lines of evidence suggest that Mdm2 may regulate normal and abnormal growth not only by inhibiting p53 function, but through p53-independent mechanisms as well. For example, Mdm2 has been shown to bind E2F1 and enhance E2F-mediated transcriptional activity (Martin et al., 1995). Additionally, Mdm2 can interact with prb and relieve at least some of its inhibitory functions (Xiao et al., 1995; Sun et al., 1998). Hence, Mdm2 may enhance E2F transcriptional activity either through direct binding or by releasing E2F from repression by prb. Recently, it has been shown that Mdm2 can interact with the transcription factor Sp1 and inhibit its DNA-binding activity (Johnson-Pais et al., 2001). Interestingly, prb can counteract this inhibition (Johnson-Pais et al., 2001). Other proteins that have been shown to interact with Mdm2 in a p53-independent manner are the developmental regulator hnumb (Juven-Gershon et al., 1998) and Akt/PKB, a kinase with anti-apoptotic properties (Mayo and Donner, 2001; Gottlieb et al., 2002). A homology search of the y database did not reveal any sequence with signi cant similarity to mammalian Mdm2 (data not shown), suggesting that a true Mdm2 homolog does not exist in Drosophila. Furthermore, while Drosophila contains a p53 homolog (dp53), the residues that are involved in the mammalian p53- Mdm2 interaction are not conserved in dp53 (Ollmann et al., 2000; Brodsky et al., 2000; Jin et al., 2000). The presumptive inability of mammalian Mdm2 to interact with the y p53 makes Drosophila an interesting system for studying p53-independent interactions of Mdm2 with its partners in vivo. For this purpose, we generated a Drosophila y line bearing an UAS-Mdm2 transgene. We report here that the overexpression of mammalian Mdm2 in the y causes marked developmental phenotypes. At least in the wing, this phenotype can be attributed to the induction of apoptosis. The mouse Mdm2 cdna was cloned into the puas-t vector (Brand and Perrimon, 1993) and the resulting UAS-Mdm2 construct injected into Drosophila embryos. Eleven independent transgenic Drosophila lines were generated. To visualize the expression of Mdm2 protein in Drosophila, one UAS-Mdm2 line was crossed with the en-gal4 line (engrailed enhancer), which is segmentally expressed in the embryo. As shown in Figure 1, expression of transgenic mouse Mdm2 was detected in

2414 Figure 1 Detection of UAS-Mdm2 expression in the embryo. Flies carrying UAS-Mdm2 were generated by P-element transformation. The EcoRI fragment of the mouse Mdm2 cdna (subclined from plasmid X2; Barak et al., 1994) was ligated into the puast vector (Brand and Perrimon, 1993) linearized with EcoRI. The resulting plasmid (UAS-Mdm2) was injected into embryos of the Drosophila line yw, using standard procedures. This allowed screening for UAS-Mdm2 lines by eye color. Expression of UAS-Mdm2 in the embryo was driven by en- GAL4. Detection of Mdm2 expression was done using standard immunohistochemistry staining procedures for Drosophila embryos, utilizing the 4B2 anti-mdm2 antibody (Chen et al., 1993) a pattern corresponding to the normal expression pattern of the engrailed gene product. At that stage, there was no noticeable e ect of Mdm2 expression on the stripe pattern, suggesting that Mdm2 overexpression does not impair early embryonic development. However, adult Mdm2 expressing ies were never obtained from this cross (data not shown), implying that the mouse Mdm2 did interfere at a later developmental stage. All eleven UAS-Mdm2 lines were next crossed with Drosophila strains expressing GAL4 under the control of either MS1096 or the glass multimer reporter (GMR) (Hay et al., 1994). The MS1096 enhancer is expressed in the wing imaginal disc, while GMR is expressed posterior to the morphogenetic furrow in the eye imaginal disc (Chang et al., 1994). Both crosses gave rise to Mdm2-induced phenotypes. In each case, the phenotypes varied in severity, depending on the insertion line used (Figures 2 and 4). In the wing, the consequence of Mdm2 overexpression was either blistered or gnarled wings (Figure 2b,c, respectively; Figure 2d shows a larger magni cation of the gnarled wing in Figure 2c). In the eye, we observed either rough or small eyes (Figure 4b,e,c,f, respectively). The di erent severity can usually be explained by insertion site-speci c e ects on the transgene expression level (Spradling and Rubin, 1983). Accordingly, one representative weak UAS-Mdm2 (Mdm2 w ) and one representative strong UAS-Mdm2 (Mdm2 s ) line were selected for further characterization. In mammals, Mdm2 interacts with proteins that are key regulators of apoptosis and cell proliferation. Therefore, we tested whether either of these processes was a ected in Mdm2-expressing larvae. Apoptosis was monitored in wing imaginal discs of third instar larvae through the use of TUNEL staining. In wing imaginal discs of control ies, there were only very few TUNEL positive cells (Figure 3a). In contrast, we observed a dramatic increase in the number of TUNEL-positive Figure 2 Overexpression of mouse Mdm2 in the wing imaginal disc leads to aberrant wing development. (a) Wild type wing. (b ± d) Wings derived from Mdm2 transgenic ies. Expression of UAS-Mdm2 in the wing imaginal disc was driven by MS1096- GAL4. The UAS-Mdm2 lines presented either a weaker, blistered phenotype (b) or a stronger, gnarled phenotype (c ± d). The photographs in a±c are shown at the same magni cation, whereas (d) represents a higher magni cation of (c) cells in Mdm2-expressing larvae (Figure 3b). The induction of extensive apoptosis was con rmed by staining the wing imaginal discs with the CM1 antibody (Figure 3c), which detects activated forms of caspase-3 and caspase-3-related proteases that appear in apoptotic cells (Srinivasan et al., 1998). Hence, the extent of apoptosis in UAS-Mdm2 expressing wing imaginal discs is signi cantly higher than in wild type discs. It thus seems plausible that the phenotype observed in adult wings is a result of ectopic apoptosis induced by Mdm2 in third instar larvae. Unexpectedly, double staining for Mdm2 and CM1 reactivity (Figure 3d) revealed that the imaginal disc areas displaying the most intense CM1 staining appeared negative for Mdm2 (arrow). In contrast, expression of GFP under the same UAS revealed positive staining in the corresponding areas (Figure 3e). This suggests that at the time when caspases become fully activated following Mdm2 induction, the Mdm2 protein is degraded. It is of note that, in mammalian cells, the endogenous Mdm2 protein is indeed cleaved by caspases during apoptosis (Chen et al., 1997; Erhardt et al., 1997; Pochampally et al., 1998); most probably, the same happens also in the insect wing imaginal disc cells. In contrast to the dramatic induction of apoptosis, staining for proliferation with an anti-phosphohistone antibody (Upstate Biotechnology Inc.) revealed no increase in the number of proliferating cells in wing imaginal discs overexpressing Mdm2 (data not shown). Mdm2 was also expressed in the compound eye under the control of GMR. This resulted in either rough or small eyes (Figure 4). In the rough eyes (Figure 4b,e), one can observe disorganization of the bristles, while some ommatidia are fused. In the small eyes (Figure 4c,f), there was complete disorganization

2415 Figure 3 Expression of UAS-Mdm2 triggers apoptosis in wing imaginal discs. (a) Wild type wing imaginal disc stained for TUNEL. TUNEL staining was performed using the In situ Cell Death Detection kit (Roche, Cat No 1 684 817). A few scattered positively stained nuclei can be observed (examples indicated by arrowheads), re ecting the ongoing normal apoptosis in the developing y larva. (b) TUNEL staining of a wing imaginal disc expressing UAS-Mdm2 driven by MS1096-GAL4. The staining reveals a high rate of apoptosis in the transgenic disc. (c) Staining of a transgenic wing disc with the CM1 antibody. CM1 (IDUN Pharmaceuticals) recognizes activated, processed caspase-3, present in apoptotic cells (Srinivasan et al., 1998). Staining was performed according to standard immunohistochemistry procedures employed for imaginal discs. Note the extensive accumulation of active caspase in a large region of the disc. No distinct cell boundaries can be discerned within the most intensely stained area, suggesting that cells within this area may have disintegrated, releasing excessive amounts of activated caspase into the resultant cavity. (d) Double staining of a transgenic wing disc for activated caspase (CM1, red) and Mdm2 (4B2, green). 4B2 staining was visualized with the aid of FITC-conjugated donkey anti-mouse immunoglobulins (diluted 1 : 1000), whereas CM1 staining was visualized with Cy3-conjugated donkey anti-rabbit immunoglobulins (diluted 1 : 1000). (e) Expression of GPF in the wing imaginal disc was driven by MS1096-GAL4. GFP was visualized by direct inspection under a uorescent microscope of the bristles, whereas ommatidia were hardly visible at all. Apoptosis assays performed on the eye imaginal disc of third instar larvae failed to reveal an increase in the number of apoptotic cells within the Mdm2- expressing domains (data not shown), unlike what was observed in the wing. No increase in proliferation was evident either (data not shown). To further investigate the mechanism underlying the phenotypic e ects of Mdm2 in the Drosophila compound eye, we generated a recombinant chromosome 2 carrying both the Mdm2 s transgene and GMR- GAL4 (the resultant ies are hereafter referred to as the Mdm2;GMR/CyO line). This line is convenient for the study of potential genetic interactions between Mdm2 and Drosophila genes. We initially crossed Mdm2;GMR/CyO with a line containing the antiapoptotic baculovirus p35 gene (UAS-p35; Hay et al., 1994). This cross did not rescue the Mdm2-induced eye phenotype. This result is consistent with the lack of evidence for apoptosis in the Mdm2-overexpressing eye disc. However, it still remains possible that the eye phenotype is due to apoptosis by a p35-independent pathway, occurring at a later stage during development (e.g. pupation) (Meier et al., 2000). As Mdm2 has been shown to interact with E2F1, hnumb and Akt in cultured mammalian cells, possible genetic interactions between Mdm2 and these proteins in the Drosophila eye were evaluated by setting up

2416 Figure 4 Overexpression of mouse Mdm2 a ects the development of the Drosophila compound eye. (a, d) Wild type Canton-S adult eye. Note the highly regular array of ommatidia and bristles. All panels show scanning electron micrographs at magni cations of 6200 (a ± c), or 61200 (d ± f). Expression of UAS-Mdm2 in the eye, driven by GMR-GAL4, resulted in either a rough (b, e) ora small (c, f) eye phenotype pertinent genetic crosses. Mdm2;GMR/CyO ies were crossed with lines Akt1 04226 and E2F 07172, that are mutants for Dakt and de2f, respectively, as well as with lines UAS-Dakt, GMR-E2F and UAS-Numb, that overexpress the corresponding proteins. Neither of these crosses was able to rescue the eye phenotype (data not shown). It remains possible that genetic interactions do take place between Mdm2 and either de2f, Dakt or Numb, but manipulation of either of these proteins alone is insu cient to achieve a major change in the phenotype. It is interesting to note that, in an experimental system where Mdm2 was overexpressed in mouse mammary gland, Mdm2 was found to cause S phase deregulation independently of either p53 or E2F1 (Reinke et al., 1999). Mdm2;GMR/CyO ies were also crossed with a line overexpressing dominant-negative dp53 (UAS- DNdp53) (Ollmann et al., 2000). Even though the two proteins are not expected to interact directly, it is still plausible that Mdm2 exerts inhibitory e ects on eye development in the y through a p53-dependent stress pathway. However, as in the other crosses, DNdp53 failed to rescue the Mdm2-induced eye phenotype. The failure of DNp53 to counteract the Mdm2 phenotype is not surprising, since p53 appears to only induce apoptosis and not cell cycle arrest in Drosophila eyes, and furthermore it only does so in response to radiation (Ollmann et al., 2000; Brodsky et al., 2000; Jin et al., 2000). A variety of earlier studies have addressed the e ect of overexpressed Mdm2 on mammalian cells, in culture as well as in vivo. Of particular interest, it was found that excess Mdm2 can exert a growth inhibitory e ect in some, but not all, cultured mammalian cells (Brown et al., 1998). The relevance of those observations to our ndings is presently unknown, but it is tempting to speculate that some of the underlying inhibitory mechanisms may be shared between these two very di erent types of experimental systems. In vivo studies, based primarily on tissue speci c Mdm2 overexpression in transgenic mouse models, revealed that excess Mdm2 can disrupt normal tissue di erentiation and cell cycle control, often leading to aberrant cell proliferation and increased tumorigenicity (Alkhalaf et al., 1999; Ganguli et al., 2000; Jones et al., 1998; Lundgren et al., 1997; Reinke et al., 1999). Of note, at least some of these e ects were p53-independent, reinforcing the notion that Mdm2 has additional molecular targets, at least when expressed in high amounts. The experiments described in the present study demonstrate that mouse Mdm2 is able to interfere with normal Drosophila development, leading to conspicuous phenotypic aberrations. Hence, Mdm2 can impinge on pathways that operate during y development, strongly suggesting that it is able to interact with one or more regulatory y proteins. The ner analysis of the mechanisms triggered by excess Mdm2 in the y may provide new insights into understanding Mdm2 function. Acknowledgments WethankHSteller,ESchejter,TVolkandLGlazerfor stimulating discussions and valuable advice. We are grateful to Idun Pharmaceuticals for the CM1 antibody, and to Y Jan, A Manoukian, W Du, H Steller, Exelixis Inc., and the Bloomington Stock Center for generously providing y strains. We would like to thank the members of the Shilo and Oren labs for their support and advice throughout this work. This work was supported in part by grant RO1 CA 40099 from the National Cancer Institute and by Yad Abraham Center for Cancer Diagnosis and Therapy. A Folberg-Blum was recipient of a Clore Post-Doctoral Fellowship from the Weizmann Institute of Science throughout this work.

References Alkhalaf M, Ganguli G, Messaddeq N, Le Meur M and Wasylyk B. (1999)., 18, 1419 ± 1434. Barak Y, Gottlieb E, Juven-Gershon T and Oren M. (1994). Genes Dev., 8, 1739 ± 1749. Brand AH and Perrimon N. (1993). Development, 118, 401 ± 415. Brodsky MH, Nordstrom W, Tsang G, Kwan E, Rubin GM and Abrams JM. (2000). Cell, 101, 103 ± 113. Brown DR, Thomas CA and Deb SP. (1998). EMBO J., 17, 2513 ± 2525. Chen L, Marechal V, Moreau J, Levine AJ and Chen J. (1997). J. Biol. Chem., 272, 22966 ± 22973. Caspari T. (2000). Curr. Biol., 10, R315 ± 317. Chang HC, Karim FD, O'Neill EM, Rebay I, Solomon NM, Therrien M, Wassarman DA, Wol T and Rubin GM. (1994). Cold Spring Harb. Symp. Quant. Biol., 59, 147 ± 153. Chen J, Marechal V and Levine AJ. (1993). Mol. Cell. Biol., 13, 4107 ± 4114. Erhardt P, Tomaselli KJ and Cooper GM. (1997). J. Biol. Chem., 272, 15049 ± 15052. Fakharzadeh SS, Trusko SP and George DL. (1991). EMBO J., 10, 1565 ± 1569. Ganguli G, Abecassis J and Wasylyk B. (2000). EMBO J., 19, 5135 ± 5147. Gottlieb TM, Leal JFM, Seger R, Taya Y and Oren M. (2002)., (in press). Haupt Y, Maya R, Kazaz A and Oren M. (1997). Nature, 387, 296 ± 299. Hay BA, Wol T and Rubin GM. (1994). Development, 120, 2121 ± 2129. Honda R, Tanaka H and Yasuda H. (1997). FEBS Lett., 420, 25 ± 27. Jin S, Martinek S, Joo WS, Wortman JR, Mirkovic N, Sali A, Yandell MD, Pavletich NP, Young MW and Levine AJ. (2000). Proc. Natl. Acad. Sci. USA, 97, 7301 ± 7306. Johnson-Pais T, Degnin C and Thayer MJ. (2001). Proc. Natl. Acad. Sci. USA, 98, 2211 ± 2216. Jones SN, Hancock AR, Vogel H, Donehower LA and Bradley A. (1998). Proc. Natl. Acad. Sci. USA, 95, 15608 ± 15612. Juven-Gershon T and Oren M. (1999). Mol. Med., 5, 71 ± 83. Juven-Gershon T, Shifman O, Unger T, Elkeles A, Haupt Y and Oren M. (1998). Mol. Cell. Biol., 18, 3974 ± 3982. Kubbutat MH, Jones SN and Vousden KH. (1997). Nature, 387, 299 ± 303. Lohrum MA and Vousden KH. (2000). Trends Cell Biol., 10, 197 ± 202. Lundgren K, Montes de Oca Luna R, McNeill YB, Emerick EP, Spencer B, Bar eld CR, Lozano G, Rosenberg MP and Finlay CA. (1997). Genes Dev., 11, 714 ± 725. Martin K, Trouche D, Hagemeier C, Sorensen TS, La Thangue NB and Kouzarides T. (1995). Nature, 375, 691 ± 694. Mayo LD and Donner DB. (2001). Proc. Natl. Acad. Sci. USA, 98, 11598 ± 11603. Meier P, Silke J, Leevers SJ and Evan GI. (2000). EMBO J., 19, 598 ± 611. Momand J, Wu HH and Dasgupta G. (2000). Gene, 242, 15 ± 29. Momand J, Zambetti GP, Olson DC, George D and Levine AJ. (1992). Cell, 69, 1237 ± 1245. Oliner JD, Pietenpol JA, Thiagalingam S, Gyuris J, Kinzler KW and Vogelstein B. (1993). Nature, 362, 857 ± 860. OllmannM,YoungLM,DiComoCJ,KarimF,BelvinM, Robertson S, Whittaker K, Demsky M, Fisher WW, Buchman A, Duyk G, Friedman L, Prives C and Kopczynski C. (2000). Cell, 101, 91 ± 101. Pochampally R, Fodera B, Chen L, Shao W, Levine EA and Chen J. (1998)., 17, 2629 ± 2636. Reinke V, Bortner DM, Amelse LL, Lundgren K, Rosenberg MP, Finlay CA and Lozano G. (1999). Cell Growth Di er., 10, 147 ± 154. Spradling AC and Rubin GM. (1983). Cell, 34, 47 ± 57. Srinivasan A, Roth KA, Sayers RO, Shindler KS, Wong AM, Fritz LC and Tomaselli KJ. (1998). Cell Death Di er., 5, 1004 ± 1016. Sun P, Dong P, Dai K, Hannon GJ and Beach D. (1998). Science, 282, 2270 ± 2272. Xiao ZX, Chen J, Levine AJ, Modjtahedi N, Xing J, Sellers WR and Livingston DM. (1995). Nature, 375, 694 ± 698. 2417