Oncogene (2008) 27, 2345–2350 & 2008 Nature Publishing Group All rights reserved 0950-9232/08 $30.00 www.nature.com/onc REVIEW FOXO animal models reveal a variety of diverse roles for FOXO factors

KC Arden

Department of Medicine, Ludwig Institute for Cancer Research, University of California San Diego School of Medicine, La Jolla, CA, USA

The Foxo subfamily of FOX transcription factors plays a organismal development, function of the immune and variety of roles in a broad assortment of diverse nervous systems, and metabolism. It has also allowed physiological processes including cellular differentiation, models ofhuman disease to be generated. These models tumor suppression, metabolism, cell cycle arrest, cell have been instrumental in determining the mechanism of death and protection from stress.Animal models have action ofnew therapies fordisease and have allowed proved to be invaluable tools in furthering our under- preclinical testing ofnew therapeutic approaches. standing of the role of particular in complex Multiple animal models have made unique contribu- organismal processes.Multiple animal models in diverse tions and invaluably aided our understanding ofthe species, including , Drosophila. various roles ofthe FOXO familyoftranscription melanogaster and the , exist for the factors. Foxo family of transcription factors. Foxo genes are highly conserved throughout the evolution and each of these model systems has provided valuable insight into the roles of Foxo factors.Many roles are conserved among The FOXO subfamily of FOX genes the different model organisms.Several Foxo-related animal model systems are reviewed here along with the The FOXO family of transcription factors regulates a knowledge gleaned to date from each model system. wide variety ofcellular and organismal processes both Oncogene (2008) 27, 2345–2350; doi:10.1038/onc.2008.27 during development and in the adult. The forkhead superfamily of transcription factors shares a highly Keywords: FOXO; FKHR; daf-16; dFoxo; animal conserved 100 amino-acid DNA binding or FOX models (Forkhead box) domain. The three predominant members ofthe mammalian Foxo subfamily ofthe Fox transcription factors (Foxo1, Foxo3 and Foxo4) were identified through the cloning ofchromosomal trans- location break points associated with cancer (reviewed Animal models by Arden, 2006). All three Foxo factors are similarly regulated through phosphorylation by a variety of The recent awarding ofthe 2007 Noble Prize in kinases, including AKT/ kinase B (PKB), serum- Physiology or Medicine to Mario R Capecchi and and glucocorticoid-inducible kinase, casein kinase 1, Oliver Smithies and Sir Martin J Evans for dual-specificity tyrosine-phosphorylated and -regulated targeting in mice (Vogel, 2007) only serves to emphasize kinase 1A and anti-I-kappa-B kinase b, typically the importance ofanimal models in understanding the resulting in Foxo inactivation by active nuclear export role ofparticular genes in processes such as embryonic to the cytoplasm. In contrast, in response to stress development, adult physiology, disease and aging. stimuli, Foxo can be phosphorylated by the mitogen- Organisms used as animal models share the distinctive activated protein kinase family member, jun N-terminal characteristic that it is possible to manipulate the kinase (JNK), resulting in the opposite effect, the genome and modify specific genes. It is through the movement ofFoxo fromthe cytoplasm to the nucleus. raising and careful observation of these genetically The AKT and JNK pathways appear to operate in modified animals that clues are provided about the opposition by regulating subcellular localization of normal function of the gene or genes in question. These Foxo factors using environmental cues promoting signal types ofexperiments have provided valuable insight into transduction from different pathways (reviewed in Greer and Brunet, 2005). The Foxos have overlapping patterns Correspondence: Dr KC Arden, Department ofMedicine, Ludwig both during development and in the adult. In addition, Institute for Cancer Research, University of California San Diego School ofMedicine, 9500 Gilman Drive, No. 0660, La Jolla, CA Foxo1, Foxo3 and Foxo4 bind to the same DNA target 92093-0660 USA. sequence and are capable ofregulating the same target E-mail: [email protected] genes. A fourth Foxo family member, FOXO6, whose Foxo function in complex organisms is revealed in model systems KC Arden 2346 expression is restricted to the brain, has been identified, regulatory pathway, so elegantly dissected in worms, but is more distantly related and is not regulated by may be conserved in higher level organisms that set subcellular localization (Jacobs et al., 2003). Foxo1, the stage for studies in higher organisms. Studies on Foxo3 and Foxo4 have been found to play important DAF-16 in C. elegans did not end with the discovery of roles in variety of processes including cellular differ- this regulatory mechanism. Further studies have entiation, tumor suppression, metabolism, cell cycle revealed that DAF-16 regulation is complex and arrest, cell death and protection from stress. These roles impacted by several additional pathways important to have been discovered through manipulation ofthe genes the biological processes of C. elegans including sensory encoding these factors in cell-based systems and in perception, germ line signaling and JNK signaling animals including Caenorhabditis elegans, Drosophila (reviewed by Mukhopadhyay and Tissenbaum, 2007). melanogaster and laboratory mice.

Drosophila melanogaster Caenorhabditis elegans The fruit fly is another manip- The round worm C. elegans is the first animal in which a ulable animal model system. D. melanogaster has been FOXO ortholog was modified and the resulting used for genetic experiments for over a century. One of phenotype was characterized. This organism offers the reasons it continues to be a valuable model animal is several advantages for the study of gene function: it that much is known about how to optimally handle the has a short life cycle; every animal has a specific number flies. Similar to C. elegans, these are small animals with a ofcells; and much is known about how these cells relatively short life cycle of about 2 weeks. Mutant flies contribute to the adult organism. It also has the with defects in thousands of genes are readily available. distinction ofbeing the first multicellular animal to In fact, the name ‘forkhead’ comes from observations of have its genome sequenced. flies with mutations in the ‘forkhead’ gene. These flies Dauer larva formation in C. elegans is controlled by developed ectopic head structures giving the appearance environmental cues. The word ‘dauer’ comes from the oftwo heads. There are established and well-understood German meaning ‘endurance’ or ‘duration.’ C. elegans methods for manipulating the fly genome and the entire has the ability to halt development under conditions of genome has now been sequenced. starvation or overcrowding and to enter a dauer larval The amino-acid sequence ofthe D. melanogaster stage. This is a reversible stage ofdevelopment, in which homolog of daf-16, dFOXO, was highly conserved when the worms develop impermeable cuticles, their metabolic compared to DAF-16 (Kramer et al., 2003). It was rate slows down dramatically and they do not feed for up discovered that overexpression of dFOXO leads to a to 3 months. Ifnutrients become more abundant and the reduction in the size ofthe animal due to decreases in environment is less crowed, the worms sense the change, cell size and cell number. This phenotype was very exit the dauer stage and complete the developmental similar to starvation, suggesting that dFOXO plays a process to become a mature adult. The FOXO ortholog, role in nutrient sensing, similar to that observed in daf-16, was one ofa number ofgenes identified as a C. elegans (Kramer et al., 2003). Interestingly, over- mediator ofdauer formation by studying the effects of expression of dFOXO targeted to specific tissues in the gene mutation in C. elegans (Riddle et al., 1981). Over a fly results in a reduction in organ size by reducing cell decade later, the connection between DAF-16’s role in number but does not seem to have an effect on cell size dauer formation and its role in C. elegans life span was (Puig et al., 2003). Flies lacking dFOXO appear discovered (Kenyon et al., 1993). Daf-16 influences life physically normal but show increased sensitivity span in the worm by mediating insulin-like metabolic to oxidative stress (Ju¨ nger et al., 2003). Expression signaling and confers as well as resistance to of dFOXO in the adult pericerebral at body of stress (Murakami and Johnson, 1996; Lin et al., 1997; D. melanogaster has been shown to decrease mortality Ogg et al., 1997). Mutations in daf-16 prevent dauer and increase life span significantly (Giannakou et al., formation under the environmental conditions that 2004; Hwangbo et al., 2004). dFOXO has also been normally promote dauer formation. Through additional shown to play a role in the control ofinsulin experimentation it was determined that the insulin-like (Insr). Starvation ofwild-type flies upregulates Insr pathway in worms included two C. elegans AKT-PKB expression. However, flies deficient for dFOXO cannot homologs that function primarily to inactivate Daf-16 upregulate Insr in response to starvation, suggesting that providing the first evidence that a dFOXO-mediated feedback of Insr transcriptional could be a downstream target ofAKT-PKB (Paradis and activation is key to the feedback mechanism of Ruvkun, 1998). A search ofthe DAF-16 amino-acid controlling the regulation ofinsulin signaling (Puig sequence revealed three putative AKT/PKB consensus and Tjian, 2005). sites offering further proof that DAF-16 was regulated It should be noted that information related to the by the worm insulin-like signaling pathway resulting in function of Foxo can be gained from genetic alteration the inactivation ofDAF-16 when the insulin-like path- of other genes that have an effect on Foxo. Although way is activated. It was these important observations in this is not the focus of this article, one such example in C. elegans and the speculation that aspects ofthe Drosophila is the manipulation ofthe insulin/insulin

Oncogene Foxo function in complex organisms is revealed in model systems KC Arden 2347 growth factor-like signaling pathway using an inducible cytoplasm thereby interfering with Foxo1 inhibition of expression system (Giannakou et al., 2007). When this Pdx1 expression and promoting b-cell proliferation. pathway is altered, dFOXO levels also change. This Foxo1 haploinsufficiency has also been shown to confer effect lessens with age, suggesting that some process protection from diet-induced diabetes in mice. These exists that prevents or lessens the physiological response mice have reduced levels ofcytokines known to be to changes in insulin/insulin growth factor-like signaling associated with insulin resistance (serum tumor necrosis pathway as the organism ages. factor-a and resistin) (Nakae et al., 2003). In a study ofhepatic stellate cells (HSCs) and liver fibrosis, inactivation ofFoxo1 promotes proliferation Mouse and transdifferentiation of HSCs resulting in the increase in susceptibility of Foxo1 þ / mice to experi- Although C. elegans and D. melanogaster have many mentally induced liver fibrosis (Adachi et al., 2007). characteristics that make them useful model systems, Hyperinsulinemia inactivates Foxo1 in HSCs, resulting many useful discoveries are still to be made in these in HSC activation, proliferation and transdifferentia- nonmammalian systems. However, to be useful and tion, and may contribute to the fibrosis observed in relevant to the human condition, these discoveries must nonalcoholic fatty liver disease. be validated in the context ofthe more complex All ofthe experiments in mice detailed above had to mammalian system. The laboratory mouse (Mus mus- be conducted using the Foxo1 þ / mouse, because the culus) has become the premier mammalian model system Foxo1-null (Foxo/) state results in an embryonic lethal for genetic research. Much is known about the mouse phenotype. However, transgenic mice have been created genome, and there are clear genetic similarities to the that overexpress wild-type Foxo. Transgeneic mice that . Humans and mice are also physiologi- express Foxo1 targeted to exhibit a cally similar. Parallels can be drawn between the human reduction in skeletal muscle mass, glycemic control and and mouse immune, nervous and cardiovascular systems expression ofslow twitch/red muscle fiber (Kamei et al., to name just a few. Mice also develop some of the same 2004). The creation oftransgenic mice expressing a conditions seen in humans such as cancer, hypertension, mutated Foxo1 cDNA, bearing a single amino-acid diabetes and osteoporosis. There exist well-known substitution replacing the phosphorylatable Ser253 with methods to manipulate the mouse genome. The mouse the nonphosphorylatable amino-acid alanine, results in is fairly small, which is important when calculating a constitutively nuclear/constitutively active form of experimental costs and, relative to other mammalian Foxo1. The constitutively active form of Foxo1, when species, mice have a short life span and reproduce fairly expressed from a transgene containing a promoter that quickly. Foxo genes have been genetically modified in a targeted expression to the liver and pancreatic b cells, number ofmouse models. These models have been resulted in mice that developed glucose intolerance, instrumental in furthering our understanding of the b-cell failure and diabetes in an age-dependent multiple roles occupied by the various Foxo transcrip- manner, lending further support to a role for Foxo1 as tion factors. an important downstream regulator ofthe insulin pathway The results ofexperiments conducted in C. elegans and suggesting that Foxo1 could be a useful therapeutic and D. melanogaster model systems suggested that Foxo target in the treatment ofdiabetes (Nakae et al., 2002). factors might play a role in signaling downstream from As mentioned above, Foxo1-null embryos die during the insulin receptor in mammals. The first experiments embryonic development. Death takes place on embryo- using a genetically modified mouse model for Foxo used nic day 10.5 as a consequence ofincomplete vascular a mouse deficient for one allele of Foxo1 (Foxo1 þ /) development. Foxo1/ embryos are approximately 50% (Nakae et al., 2002). Breeding these mice with insulin- the size oftheir Foxo1 þ / þ littermates. Cardiac looping resistant diabetic mice, caused by a mutation in Insr, of Foxo1/ embryos is retarded and the pericardium is resulted in restored insulin sensitivity and a rescue ofthe distended compared to Foxo1 þ / þ and Foxo1 þ /. The diabetic phenotype. This result was achieved through dorsal aorta in Foxo1/ embryos appears thin and downregulation ofglucogenetic gene expression in the disorganized, and the intersomitic vessels are also liver and coordinate upregulation ofinsulin-sensitizing irregularly developed. The head vasculature of Foxo1/ genes. This initial study suggested that Foxo1 has a role embryos also appears to lack properly formed as a negative regulator ofinsulin sensitivity. The role of branches ofthe internal carotid artery. In addition, pancreatic b cells is integral to diabetes. Type 1 diabetes developed vasculature was not present in Foxo1/ yolk is characterized by the complete absence ofinsulin- sacs. Foxo1 expression was found in a variety of producing b cells, whereas type II diabetes is associated embryonic vessels, supporting a crucial role ofthis with a deficiency ofinsulin-producing b cells. Foxo1 transcription factor in vascular formation (Hosaka haploinsufficiency reversed the characteristic b-cell fail- et al., 2004). In vitro differentiation of embryonic stem ure observed in mice lacking Insr substrate-2 (Irs2/)by cells demonstrated that endothelial cells derived from partial restoration ofthe ability ofthe residual b cells to Foxo1-deficient embryonic stem cells in the absence of proliferate and increased expression of Pdx1 (Kitamura exogenous vascular endothelial growth factor showed a et al., 2002). These data suggest that Foxo1 is an markedly different morphological response compared important regulator of b-cell proliferation. Insulin with wild-type endothelial cells, suggesting that Foxo1 is promotes cellular relocalization ofFoxo1 to the important in endothelial cell response to vascular

Oncogene Foxo function in complex organisms is revealed in model systems KC Arden 2348 endothelial growth factor (Furuyama et al., 2004). Thus, When Cre-mediated gene disruption ofall three Foxo in the developing embryo, Foxo1 plays a critical role in genes was induced at 4–5 weeks ofage, the animals establishing normal vasculature. developed lymphoblastic thymic lymphomas between 19 Both Foxo3a- and Foxo4-null mice are viable and and 30 weeks ofage. These tumors possessed the ability grossly indistinguishable from their littermate controls, to spread to additional tissues (spleen, liver and lymph indicating dispensability ofthese two members ofthe nodes) (Paik et al., 2007). The Foxo1-, Foxo3-, Foxo4- Foxo transcription factor family for normal vascular null mice also developed hemangiomas first appearing in development. Foxo3-null female mice showed age- the uterus at 6–8 weeks ofage. The hemangiomas dependent infertility and had abnormal ovarian folli- eventully progressed to massive fatal tumors found in cular development (Castrillon et al., 2003; Hosaka et al., skeletal muscle, abdominal wall, liver, adrenal glands, 2004). The Foxo3/ female mice exhibit widespread, bone marrow, omentum, lymph nodes and skin. Lethal almost uniform follicular activation leading to oocyte angiosarcomas were observed in 9% ofthe animals. degeneration, resulting in the early total depletion of Functional redundancy of Foxo genes was apparent, as functional ovarian follicles by 12 weeks of age and loss ofall three Foxo alleles was required for the tumor infertility. Foxo3 appears to function normally to phenotype. Disruption ofany combination oftwo ofthe suppress follicular activation and help maintain a three Foxo genes resulted in a more moderate pheno- reserve resting follicle pool. The increased levels of type. Thus, the functional redundancy of these closely serum follicle-stimulating hormone and luteinizing related transcription factors was confirmed. hormone levels in Foxo3/ female mice suggest that In light ofthe widespread tissue distribution of Foxo this phenotype may be very similar to accelerated expression in mammals, the restricted tumor phenotype follicular initiation and ultimate follicle depletion was unexpected. In humans, individual FOXO gene observed in human premature ovarian failure, a disruption by translocation has been common cause ofinfertility and premature aging in associated with and the solid tumor, alveolar women (Castrillon et al., 2003). rhabdomyosarcoma, whereas misregulation of FOXO Foxo3 is the predominant Foxo member expressed in genes has been associated with a variety oftumor types peripheral lymphoid organs. Histological examination including prostate, stomach, brain and breast. As the of different tissues in Foxo3-deficient mice revealed that FOXO DNA-binding domain is disrupted by the Foxo3 deficiency leads to lymphoproliferation, wide- chromosomal translocations, thereby eliminating regu- spread organ inflammation (Lin et al., 2004). Compared lation ofits target genes, it has been proposed that to their wild-type counterparts, Foxo3/ hyperactivated Foxos may function as tumor suppressors. Activation of helper T cells proliferated more vigorously and pro- Foxo factors by mutation of the phosphorylation sites, duced more Th1 and Th2 cytokines. Normally, Foxo3 thereby restricting the localization ofthe Foxo inhibits NF-kB activation. Therefore, in the absence of to the nucleus, results in cell-cycle arrest or cell death Foxo3, activation ofNF- kB promotes T-cell hyperac- as would be expected ofa tumor suppressor gene. tivity. In a role seemingly unrelated to ovarian function, Furthermore, all three predominant Foxo factors are Foxo3 regulates helper T-cell activation by inhibiting directly regulated by AKT and a variety oftumor types NF-kB activity, suggesting a role for the Foxo proteins are associated with mutations affecting the PI3K/AKT in the inhibition ofinflammation. pathway. It was surprising therefore that, despite the In contrast, to the Foxo1- and Foxo3-deficient mice, absence ofFoxo proteins in several cell and tissue types Foxo4-null mice are grossly normal in appearance, and where Foxos are normally expressed, the tumor pheno- histological analysis ofover 30 tissues from9-week-old type was restricted to endothelial-derived cells. Perhaps Foxo4/ mice compared to their normal counterparts more interesting, not all tissues containing endothelial revealed no consistent histological differences (Hosaka cells were affected. For example, lungs and kidneys were et al., 2004). The important conclusion from these unaffected, suggesting that the phenotype is not only cell individual gene disruption experiments is that the lineage restricted but also organ-specific. Foxo-deficient physiological roles of Foxo factors are functionally liver-derived endothelial cells show an increased capa- diverse in mammals. city for proliferation and survival. These characteristics The experiments described above relied on the were absent in Foxo-deficient endothelial cells derived constitutive individual disruption ofeach ofthe three from lung tissue. Microarray analysis revealed non- predominant Foxo genes. Individual gene disruption overlapping lists ofputative direct targets when RNA resulting in differing phenotypes suggests that some from liver-derived Foxo-deficient endothelial cells was degree offunctional diversification exists during devel- compared to Foxo-deficient thymocytes further reinfor- opment and in the adult animal (Castrillon et al., 2003; cing the concept ofFoxo cell- and tissue-type specificity. Hosaka et al., 2004). However, as the three genes show Oxidative stress is implicated in the etiology ofmany overlapping patterns ofexpression, there remains still human diseases and Foxos have been implicated as the possibility offunctional redundancy. The develop- important mediators ofcellular response to oxidative ment ofa conditional Foxo1/, Foxo3/ and Foxo4/ stress (Kops et al., 2002). Using the same system ofCre- mouse model and the observation ofthe distinct mediated disruption ofall three Foxo genes but focusing phenotype conferred by disruption of all three genes on the hematopoietic system revealed a significant simultaneously addressed the issue offunctional redun- decrease in the long-term hematopoietic dancy (Paik et al., 2007; Tothova et al., 2007). population and the common lymphoid progenitor

Oncogene Foxo function in complex organisms is revealed in model systems KC Arden 2349 compartment (Tothova et al., 2007). The Foxo-deficient in gene products required for the detoxification of bone marrow was found to be defective in reactive oxygen species. colony-formation activity in vitro and in its ability to support long-term bone marrow repopulation in vivo. A significant increase in the number ofHSCs exiting Conclusion G0/G1 and entering S/G2/M was observed, suggesting that Foxos are important in maintaining the quiescent The animal models discussed above illustrate the value state ofHSCs and preserving their replicative ofstudying conserved pathways in a variety ofmodel and renewal capacity. This is reminiscent ofthe role systems. The shared effects of Foxo gene deficiency are ofFoxo3 in maintaining a resting pool ofovarian significant. Consistent across all model systems was the follicles. Foxo-deficient HSCs also exhibit increased role ofFoxo in the insulin-signaling pathway. In levels ofapoptosis furthercontributing to the overall addition, Foxo appears to be important in the main- decrease in cell number. Similar to the tumor phenotype tenance ofquiescent cells. This role applies to a variety described above, deficiency in any one or two ofsetting fromlong-term hematopoietic stem Foxo genes in combination does not produce the full cell population, HSCs and ovarian follicles in mammals cell cycle and phenotype corroborating to the quiescent stage ofthe dauer larval stage in the functional redundancy of Foxo proteins. A C. elegans. The animal model systems developed marked increase in reactive oxygen species levels in and studies to date have revealed much about the multiple HSC isolated from the Foxo-deficient mice was also roles ofFoxo factors at the complex organismal level; observed. Gene expression analysis revealed a decrease however, many questions remain to be explored. in expression ofa subset ofgenes associated Unraveling the multifaceted aspects of FOXO regula- with cellular response to oxidative stress, suggesting tion will provide important insights ofall the processes that the increase in reactive oxygen species detected in involving Foxos including development, immunity, the Foxo-deficient HSCs is due to a deficiency neurology, physiology, metabolism and aging.

References

Adachi M, Osawa Y, Uchinami H, Kitamura T, Accili D, Brenner D. skeletal muscle mass, down-regulated Type I (slow twitch/red (2007). The forkhead transcription factor FoxO1 regulates prolif- muscle) fiber genes, and impaired glycemic control. J Biol Chem 279: eration and transdifferentiation of hepatic stellate cells. Gastro- 41114–41123. enterology 132: 1434–1446. Kenyon C, Chang J, Gensch E, Rudner A, Tabtiang R. (1993). A C. Arden KC. (2006). Multiple roles ofFOXO transcription factorsin elegans mutant that lives twice as long as wild type. Nature 366: mammalian cells point to multiple roles in cancer. Exp Gerontol 41: 461–464. 709–717. Kitamura T, Nakae J, Kitamura Y, Kido Y, Biggs III WH, Wright CV Castrillon DH, Miao L, Kollipara R, Horner JW, DePinho RA. et al. (2002). The forkhead transcription factor Foxo1 links insulin (2003). Suppression ofovarian follicleactivation in mice by the signaling to Pdx1 regulation ofpancreatic beta cell growth. J Clin transcription factor Foxo3a. Science 301: 215–218. Invest 110: 1839–1847. Furuyama T, Kitayama K, Shimoda Y, Ogawa M, Sone K, Yoshida- Kops GJ, Dansen TB, Polderman PE, Saarloos I, Wirtz KW, Araki K et al. (2004). Abnormal angiogenesis in Foxo1 (Fkhr)- Coffer PJ et al. (2002). Forkhead transcription factor deficient mice. J Biol Chem 279: 34741–34749. FOXO3a protects quiescent cells from oxidative stress. Nature Giannakou ME, Goss M, Jacobson J, Vinti G, Leevers SJ, Partridge 419: 316–321. L. (2007). Dynamics ofthe action ofdFOXO on adult mortality in Kramer JM, Davidge JT, Lockyer JM, Staveley BE. (2003). Expres- Drosophila. Aging Cell 6: 429–438. sion of Drosophila FOXO regulates growth and can phenocopy Giannakou ME, Goss M, Ju¨ nger MA, Hafen E, Leevers SJ, Partridge starvation. BMC Dev Biol 3:5. L. (2004). Long-lived Drosophila with overexpressed dFOXO in Lin K, Dorman JB, Rodan A, Kenyon C. (1997). daf-16: An HNF-3/ adult fat body. Science 305: 361. forkhead family member that can function to double the life-span of Greer EL, Brunet A. (2005). FOXO transcription factors at the Caenorhabditis elegans. Science 278: 1319–1322. interface between and tumor suppression. Oncogene 24: Lin L, Hron JD, Peng SL. (2004). Regulation ofNF-kappaB, Th 7410–7425. activation, and autoinflammation by the forkhead transcription Hosaka T, Biggs W III H, Tieu D, Boyer AD, Varki NM, Cavenee factor Foxo3a. Immunity 21: 203–213. WK et al. (2004). Disruption of forkhead transcription factor Ogg S, Paradis S, Gottlieb S, Patterson GI, Lee L, Tissenbaum HA (FOXO) family members in mice reveals their functional diversifica- et al. (1997). The Fork head transcription factor DAF-16 transduces tion. Proc Natl Acad Sci USA 101: 2975–2980. insulin-like metabolic and longevity signals in C. elegans. Nature Hwangbo DS, Gershman B, Tu MP, Palmer M, Tatar M. (2004). 389: 994–999. Drosophila dFOXO controls lifespan and regulates insulin signalling Mukhopadhyay A, Tissenbaum HA. (2007). Reproduction and in brain and fat body. Nature 429: 562–566. longevity: secrets revealed by C. elegans. Trends Cell Biol 17: 65–71. Jacobs FM, van der Heide LP, Wijchers PJ, Burbach JP, Hoekman Murakami S, Johnson TE. (1996). A genetic pathway conferring life MF, Smidt MP. (2003). FoxO6, a novel member ofthe FoxO class extension and resistance to UV stress in Caenorhabditis elegans. oftranscription factorswith distinct shuttling dynamics. J Biol Genetics 143: 1207–1218. Chem 278: 35959–35967. Nakae J, Biggs III WH, Kitamura T, Cavenee WK, Wright CV, Arden Ju¨ nger MA, Rintelen F, Stocker H, Wasserman JD, Ve´ gh M, KC et al. (2002). Regulation ofinsulin action and pancreatic beta- Radimerski T et al. (2003). The Drosophila forkhead transcription cell function by mutated alleles of the gene encoding forkhead factor FOXO mediates the reduction in cell number associated with transcription factor Foxo1. Nat Genet 32: 245–253. reduced insulin signaling. J Biol 2: 20. Nakae J, Kitamura T, Kitamura Y, Biggs III WH, Arden KC, Accili Kamei Y, Miura S, Suzuki M, Kai Y, Mizukami J, Taniguchi T et al. D. (2003). The forkhead transcription factor Foxo1 regulates (2004). Skeletal muscle FOXO1 (FKHR) transgenic mice have less adipocyte differentiation. Dev Cell 4: 119–129.

Oncogene Foxo function in complex organisms is revealed in model systems KC Arden 2350 Paik J-H, Kollipara R, Chu G, Ji H, Xiao Y, Ding Z et al. (2007). Puig O, Tjian R. (2005). Transcriptional feedback control of insulin FoxOs are lineage-restricted redundant tumor suppressors and receptor by dFOXO/FOXO1. Genes Dev 19: 2435–2446. critical regulators ofendothelial cell homeostasis. Cell 128: 309–323. Riddle DL, Swanson MM, Albert PS. (1981). Interacting genes in Paradis S, Ruvkun G. (1998). Caenorhabditis elegans Akt/PKB dauer larva formation. Nature 290: 268–271. transduces insulin receptor-like signals from AGE-1 PI3 kinase to Tothova Z, Kollipara R, Huntly BJ, Lee BH, Castrillon DH, Cullen the DAF-16 transcription factor. Genes Dev 12: 2488–2498. DE et al. (2007). Foxo are critical mediators ofhematopoietic stem Puig O, Marr MT, RuhfML, Tjian R. (2003). Control ofcell number cell resistance to physiologic oxidative stress. Cell 128: 325–339. by Drosophila FOXO: downstream and feedback regulation of the Vogel G. (2007). Nobel Prizes. A knockout award in medicine. Science insulin receptor pathway. Genes Dev 17: 2006–2020. 318: 178–179.

Oncogene