Oncogene (2011) 30, 1784–1797 & 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc ORIGINAL ARTICLE Divergent functions for eIF4E and S6 kinase by sonic hedgehog mitogenic signaling in the developing cerebellum

LA Mainwaring1,2 and AM Kenney1,2

1Biochemistry, Cell, and Molecular Biology Program, Weill Medical College of Cornell University, New York, NY, USA and 2Department of Cancer Biology and Genetics, Memorial Sloan-Kettering Cancer Center, New York, NY, USA

Cerebellar development entails rapid peri-natal prolifera- growth-related messenger RNAs (mRNAs), enabling tion of cerebellar granule neuron precursors (CGNPs), cell cycle progression and cell size increase (Hay and proposed cells-of-origin for certain medulloblastomas. Sonenberg, 2004; Ma and Blenis, 2009). Studies in cell CGNPs require insulin-like growth factor (IGF) for lines have established that activated mTOR phosphor- survival and sonic hedgehog (Shh)—implicated in medul- ylates its substrates, 4E-binding proteins (4E-BPs) and loblastoma—for proliferation. The IGF-responsive kinase ribosomal protein (rp) S6 kinase (S6K), in parallel to mammalian target of rapamycin (mTOR) drives prolif- promote simultaneous (a) release of eukaryotic initia- eration-associated protein synthesis. We asked whether tion factor 4E (eIF4E) from inhibition and (b) activation Shh signaling regulates mTOR targets to promote CGNP of rpS6, thus enhancing activity of downstream effectors proliferation despite constitutive IGF signaling under to promote synthesis of proteins required for cell cycle proliferative and differentiation-promoting conditions. progression (Gingras et al., 2001). However, in the case Surprisingly, Shh promoted eukaryotic initiation factor of primary cells and in vivo systems, it is unclear how the 4E (eIF4E) expression, but inhibited S6 kinase (S6K). mRNA translation machinery responds to environmen- In vivo, S6K activity specifically marked the CGNP tal cues regulating proliferation or differentiation, an population transitioning from proliferation-competent to energetically expensive process that frequently entails post-mitotic. Indeed, eIF4E was required for CGNP cell type-specific morphological adaptations. This issue proliferation, while S6K activation drove cell cycle exit. is particularly highlighted in the central nervous system, Protein phosphatase 2A (PP2A) inhibition rescued S6K wherein cell numbers are rigorously controlled and mature activity. Moreover, Shh upregulated the PP2A B56c neurons achieve a volume that is much larger than that of subunit, which targets S6K for inactivation and was dividing progenitor cells, because of extension of axonal required for CGNP proliferation. These findings reveal and dendritic processes or myelin sheath production as is unique developmental functions for eIF4E and S6 kinase the case in glial cells (Altman and Bayer, 1997). wherein their activity is specifically uncoupled by mito- The developing mammalian cerebellum is an ideal genic Shh signaling. system for investigating mTOR pathway regulation in Oncogene (2011) 30, 1784–1797; doi:10.1038/onc.2010.564; proliferating neural progenitors. The cerebellum is published online 21 February 2011 characterized by a small number of cell types, which undergo well-characterized developmental programs to Keywords: sonic hedgehog; eIF4E; S6 kinase; achieve the final stereotypical architecture of the mature proliferation cerebellum, enabling this brain structure to process cues that regulate coordination, movement and aspects of motor learning (Altman and Bayer, 1997). Moreover, aberrant proliferation of cerebellar neural precursors is Introduction implicated in medulloblastoma, the most common solid malignancy of childhood (Wechsler-Reya and Scott, The kinase mammalian target of rapamycin (mTOR) 2001). Thus, studies of cerebellar development yield integrates extracellular and intracellular signals to insight into normal and tumorigenic central nervous regulate cell growth in mammalian cells. Growth system growth control mechanisms. factors, nutrients and adenosine triphosphate levels Cerebellar granule neuron precursors (CGNPs) un- converge upon mTOR to propagate signals that dergo a burst of rapid, post-natal expansion in the promote translation of proliferation, metabolism and external granule layer (EGL) on the dorsal surface of the cerebellum (Hatten, 1998). This expansion phase lasts B2 weeks in mice and 12 months in human infants. When CGNPs exit the cell cycle, they migrate internally Correspondence: Dr AM Kenney, Department of Neurosurgery, and differentiate into interneurons, where they function Vanderbilt University, Nashville, TN 37232-2308, USA. E-mail: [email protected] as signal integrators between mossy fibers and Purkinje Received 14 June 2010; revised 2 September 2010; accepted 29 September neurons (Altman and Bayer, 1997). In contrast to its 2010; published online 21 February 2011 roles in neural cell-fate specification or differentiation eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1785 elsewhere in the nervous system (Marti et al., 1995; Results Dutton et al., 1999a, b), sonic hedgehog (Shh) serves as a mitogen for CGNPs (Dahmane and Ruiz-i-Altaba, Proliferating CGNPs differentially regulate mTOR 1999; Wallace, 1999; Wechsler-Reya and Scott, 1999). substrates Shh binds to its receptor Patched (Ptc), lifting inhibition To address the role of mTOR signaling in cerebellar of the signaling pathway activated by the transmem- development, we used a well-characterized primary brane protein smoothened (Smo) (Ho and Scott, 2002), culture system of cerebellar granule neurons wherein and resulting in activation of Smo target genes, isolated CGNPs from post-natal day (PN) 5 mice are including Ptc itself (Hepker et al., 1997). Downstream maintained under serum-free conditions, in the presence effectors of Smo that regulate CGNP proliferation of IGF which promotes their survival and cooperates include Gli transcription factors, N-myc, insulin recep- with Shh to maintain proliferation (Dudek et al., 1997; tor substrate 1 (IRS1), YAP1 and the microRNA cluster Kenney et al., 2004). Continued proliferation is pro- miR17/92 (Dahmane et al., 2001; Kenney et al., 2003; moted by addition of Shh protein (Shh-treated) to the Oliver et al., 2003; Parathath et al., 2008; Fernandez culture medium; cell cycle exit and differentiation into et al., 2009; Northcott et al., 2009; Uziel et al., 2009). mature neurons occurs in the absence of Shh (vehicle- In addition to requiring Shh for proliferation, CGNPs treated) (Wechsler-Reya and Scott, 1999; Kenney and also require signaling by insulin-like growth factor Rowitch, 2000). When we examined levels and activity (IGF) for cell survival (Dudek et al., 1997; D’Mello of mTOR pathway components in vehicle- and Shh- et al., 1997). Notably, aberrant activation of the Shh and treated CGNPs by western blot analysis and densito- IGF signaling pathways is implicated in medulloblas- metry, we found that proliferating (Shh-treated) CGNPs toma (Hahn et al., 2000; Wechsler-Reya and Scott, 2001; show upregulation of proteins consistent with a need for Hartmann et al., 2005). IGF positively regulates the increased cap-dependent mRNA translation, specifically mTOR pathway (Corradetti and Guan, 2006), and mTOR itself, the mTOR partner Raptor, eIF4G and CGNPs are continuously exposed to IGF in vivo and other eIF4F components (Figure 1a and Supplementary in vitro (Ye et al., 1996; D’Mello et al., 1997). Previous Figure S1A). We also observed that Shh treatment reports demonstrated that in addition to providing pro- promoted increases in total eIF4E and 4EBP2 levels, as survival signals, IGF signaling cooperates with the well as 4EBP2 phosphorylation, suggesting that the mitogenic Shh pathway to promote stabilization of negative regulation 4EBP2 has on eIF4E is relieved. N-myc, a Shh target required for CGNP proliferation Surprisingly, Shh suppressed S6K activity as deter- and upregulated in Shh-associated medulloblastoma mined by rpS6 phosphorylation (Figure 1a, Supplemen- (Pomeroy et al., 2002; Kenney et al., 2004; Northcott tary Figure S1B). When we compared the levels of et al., 2009). However, CGNPs only proliferate in phosphorylated protein with total protein we found that response to Shh (Wechsler-Reya and Scott, 1999) despite in both vehicle- and Shh-treated CGNPs 4EBP2 is constitutive IGF pathway activation. This conundrum phosphorlyated; however, the amount of total rpS6 that suggests that IGF-independent mechanisms exist down- is phosphorylated at either serine residue in Shh-treated stream of Shh to promote preferential activity of mTOR CGNPs is significantly decreased (Figure 1b). Previous substrates during the CGNP proliferation phase. reports from studies conducted largely in cell lines have To test this hypothesis, we examined how Shh indicated that mTOR phosphorylates both 4E-BPs and signaling affects expression and activity of mTOR S6K (von Manteuffel et al., 1997; Fingar et al., 2004) in pathway components in CGNPs. Surprisingly, we parallel, but our results suggest that mechanisms exist to observed differential regulation of eIF4E and S6K preferentially specify phosphorylation of these mTOR activity, two of the best-characterized mTOR effectors, substrates in primary neural precursors responding to in response to Shh in proliferation-competent CGNPs. mitogenic Shh signaling. Our findings contrast with models that have been based To further demonstrate that 4EBP2 phosphorylation on serum-stimulated cell line studies, which have combined with Shh-driven increases in total eIF4E levels indicated that eIF4E and S6K function cooperatively results in active eIF4E we performed immunoprecipita- to promote proliferation (von Manteuffel et al., 1996; tion experiments to pull down eIF4E with eIF4G, the Burnett et al., 1998; Fingar et al., 2004; Corradetti and scaffold protein required for formation of the transla- Guan, 2006). In primary CGNP cultures we correlated tion initiation complex. eIF4E only binds to eIF4G eIF4E expression with Shh-induced proliferation, when it has been released from the negative inhibition whereas S6K activity was associated with CGNP cell by 4EBPs (Pause et al., 1994; Gingras et al., 1999). As cycle exit; indeed phosphorylation of the S6K substrate shown in Supplementary Figure S1C eIF4E–eIF4G rpS6 marked the population of cells transitioning from complexes form in both vehicle- and Shh-treated the proliferative to the post-mitotic state in the neonatal CGNPs, but at much higher levels in the presence of cerebellum in vivo. The positive effects of Shh on eIF4E Shh. were at the transcriptional and protein upregulation To further investigate a role for eIF4E downstream of level, whereas the suppression of S6K activity involved Shh in CGNP proliferation, we stained primary CGNP protein phosphatase 2A (PP2A) activity, specifically cultures with antibodies against eIF4E and phospho- through Shh-mediated upregulation of the B56g sub- rpS6. Consistent with our western blot results, eIF4E unit, which targets PP2A to dephosphorylate and protein levels significantly increased in Shh-treated inactivate S6K. CGNPs (Figures 1c, d and k) while phospho-rpS6

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1786

Figure 1 Shh-treated proliferating CGNPs show increased levels of translation initiation components and suppressed S6K activity. (a) Western blot analysis of protein isolated from vehicle- or Shh-treated CGNP primary cultures. Shh treatment stimulates mTOR activity as demonstrated by eIF4G and 4EBP2 hyperphosphorylation, but blocks S6K activity measured by rpS6 phosphorylation. Levels of the mTOR effector eIF4E increase in Shh-treated CGNPs. (b) Quantification of western blots by densitometry. Phosphorylated protein levels were compared with total protein levels after normalization to b-Tubulin n ¼ 3(**Po0.002, ***Po0.0005). (c–i) Immunofluorescence analysis of vehicle- and Shh-treated CGNPs plated on coverslips. eIF4E increases in Shh-treated CGNPs (d compared with c) while phospho-S235/236- rpS6 accumulates in vehicle-treated CGNPs (e compared with f). Total rpS6 levels do not change in proliferating or differentiated cells (g, h). Furthermore phospho-S235/236-rpS6 is expressed in vehicle-treated CGNPs marked by the differentiation marker, MEF2D (i arrowheads), whereas eIF4E protein co-localizes with PCNA-positive cells (j arrowheads) in Shh-treated CGNPs. (k) Quantification of immunostaining in c–h, n ¼ 3(***Po0.0001, **Po0.001). (l) Percentage of cells expressing eIF4E only, PCNA only or both eIF4E and PCNA in Shh-treated CGNPs. (m) Percentage of cells expressing MEF2D only, phospho-rps6 only or both MEF2D and phospho-rpS6 in vehicle-treated CGNPs.

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1787 decreased (Figures 1e, f and k). As a control we also demonstrates that eIF4E correlates with CGNP pro- stained for total rpS6 in vehicle- and Shh-treated liferation (Figure1l). Furthermore, the majority phos- CGNPs and found no significant differences (Figures pho-rpS6 positive cells (red) also co-label with MEF2D 1g, h and k). We also analyzed proliferating cell nuclear (green) protein (Figures 1i and m), indicating an antigen (PCNA) in CGNPs as an indicator of prolifera- association between S6K activity and CGNP cell cycle tion, and levels of MEF2D, a transcription factor exit. induced during CGNP differentiation (Fogarty et al., To determine whether our in vitro observations 2007), to address whether eIF4E expression is linked to linking eIF4E with proliferation in CGNPs recapitu- proliferation and S6K activity with differentiation on a lated the in vivo protein accumulation pattern, we per cell basis. As shown in Figure 1j, PCNA-positive carried out immunohistochemistry for eIF4E in the cells (green) also express eIF4E (red). Quantification of mouse cerebellum at PN 7 (Hematoxylin and eosin- the number of single and double positive cells further stained sagittal section is shown in Figure 2a). As shown

Figure 2 Evaluation of eIF4E and phospho-S235/236 rpS6 localization in vivo.(a) Hematoxylin and eosin staining of wild type (wt) PN7 sagittal cerebellar sections. eIF4E is elevated in CGNPs located in the EGL (b). eIF4E (green) also co-localizes with the proliferation marker PCNA (red). (c) Higher magnification of area indicated in C further illustrates that cells in the EGL label with both eIF4E which is cytoplasmic and nuclear PCNA, arrows (c0). eIF4E is not present in p27 (d) or NeuN (e) positive cells, which indicate post-mitotic or differentiated neurons, respectively. GFAP staining marks astrocyte cell bodies and processes that extend into the EGL. (f) Confocal microscopy on area indicated in F demonstrates that eIF4E-positive cells are on a separate plane from the astrocyte processes (arrows) (f0). (g) H&E staining on inner folia of wt PN7 sections to define area used to analyze phospho-S235/236- rpS6 protein, an indicator of S6K activity. (h) Cells with increased S6K activity are restricted to specific regions within EGL, indicated by arrows. Phospho-S235/236-rpS6 does not co-localize with p27 (i), NeuN (j, j0), or the astrocyte marker GFAP. (k) Confocal microscopy on area defined in k confirms that astrocyte processes run along the cell membrane of phospho-S235/236-S6 positive cells in the EGL (arrows) (k0) Scale bars ¼ 100 mm.

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1788 in Figure 2b, eIF4E was strongly expressed in the EGLa, mTOR is not required for their maintenance. Interest- the region wherein CGNP expansion takes place, as well ingly, 24 h of rapamycin treatment did not impair as Purkinje neurons. Cells bearing eIF4E largely over- CGNP proliferation as determined by quantification of lapped with PCNA-positive cells (Figures 2c and c0)in PCNA-positive cells. Decreases in CGNP proliferation the EGLa (the mitotic region of the EGL). In contrast, were only observed after 48 h of rapamycin treatment as shown in Figures 2d and e, eIF4E was excluded from (Figure 3d), by which time rapamycin may also be cells of the EGLb (the post-mitotic region of the EGL) impacting upon mTOR:Rictor and impairing Akt that express the cyclin-dependent kinase inhibitor p27 or activity through inhibition of S473 phosphorylation NeuN, markers of CGNP cell cycle exit. eIF4E (Sarbassov et al., 2006). These results suggest that the immunostaining did not reflect astrocytic expression, increases in eIF4E are likely to be mediated through a as astrocyte processes stained positive for glial fibrillary Shh-dependent mechanism, independent of mTOR. acidic protein (Figures 2f and f0), but were not eIF4E positive. We also analyzed phospho-S6 levels in vivo.We CGNPs require eIF4E for proliferation observed a unique pattern of rpS6 phosphorylation in We next wished to determine whether manipulation of the EGL wherein phospho-rpS6 positive cells marked a eIF4E levels affects CGNP proliferation. We first specific cell population bordering the EGLa and EGLb carried out loss-of-function analysis using lentiviruses (Figure 2h), where CGNPs transition from proliferating carrying short hairpin RNA (shRNA) sequences targeting to post mitotic. Additionally, Purkinje neurons and cells eIF4E. We found that three out of the five shRNA of the IGL were also positive. Interestingly, phospho- lentiviruses tested in a murine cell line were effective in rpS6 staining does not co-localize with previously knocking down eIF4E (Supplementary Figure S2), and described markers of post-mitotic (p27, Figure 2i) or then used those lentiviruses as a pool to infect CGNPs. differentiated granule cells (NeuN, Figures 2j and j0)in Effects on eIF4E were specific as determined by the EGL, indicating that these phospho-prS6-positive comparison of eIF4E levels in CGNPs infected with cells have not completed cell cycle exit. The absence control (GFP-targeting) lentiviruses and by assessment of phospho-rpS6 staining in glial fibrillary acidic of proteins not expected to change by eIF4E knock- protein-positive cells suggests that in the EGL S6K down. Although the infection efficiency was only B20% activity may have a novel CGNP-specific role in in Shh-treated CGNPs (Supplementary Figure S2), we mediating the transition from cycling to a non-cycling observed decreased levels of cyclin D2, eIF4G, IRS1 in state (Figures 2k and k0). eIF4E-knocked down CGNPs. We previously reported that IRS1 is required for Shh-mediated proliferation (Parathath et al., 2008), thus these results hint at Shh effects on eIF4E do not require mTOR activity reduced CGNP proliferation in the absence of eIF4E. It has been reported that Shh signaling induces rapid Interestingly, we also observed a recovery of rpS6 upregulation of cell proliferation-related mRNAs phosphorylation in these cells (Figure 4a), indicating through either direct or indirect mechanisms (Katoh rescue of S6K activity, known to de-stabilize IRS1 in and Katoh, 2009). To test whether eIF4E is a transcrip- CGNPs (Parathath et al., 2008). In contrast, phosphory- tional target of Shh signaling, we isolated RNA from lation of 4E-BP2 was not affected by eIF4E loss, in vehicle and Shh-treated CGNPs at different time points keeping with our observation (Figure 1a) that 4E-BP2 and then measured eIF4E transcripts by quantitative and S6K phosphorylation are independently regulated PCR. As shown in Figure 3a, after 24 h of Shh treatment in Shh-treated CGNPs. These results indicate that an a significant two-fold induction in eIF4E transcripts was inverse relationship exists between eIF4E expression and observed (P ¼ 0.0005). The delay and levels of eIF4E S6K activity, in that when eIF4E is present, S6K activity induction compared with Gli1 (data not shown) suggest is suppressed, but when eIF4E is lost and subsequently that Shh upregulates eIF4E through an indirect CGNP proliferation decreases, S6K activity returns. mechanism. The induction of eIF4E was prevented Reduced cyclin D2 and IRS1 levels in eIF4E-knocked when Shh-treated CGNPs were incubated in the down CGNPs suggests that loss of eIF4E decreased presence of the smoothened antagonist, SANT-2 proliferation of these cells (Figure 4a). To quantitatively (Figure 3a). To determine whether eIF4E upregulation determine the effect of eIF4E loss on proliferation in is independent of mTOR activity, we incubated Shh- Shh-treated CGNPs, we measured BrdU incorporation treated CGNPs with the mTORC1 inhibitor rapamycin in control-infected and eIF4E shRNA-infected cells for increasing periods of time. We then isolated RNA or following a 2-h BrdU pulse (Figures 4b–d). eIF4E protein and looked at effects on eIF4E and prolifera- knockdown reduced BrdU incorporation by 50% tion. As shown in Figure 3a, rapamycin did not compared with Shh-treated control-infected CGNPs, significantly reduce eIF4E expression. Exposure to as determined by quantification of immunofluorescent rapamycin for increasing periods of time further staining (Po0.02) (Figure 4e); however, eIF4E loss did suppressed rpS6 phosphorylation and blocked 4E-BP2 not compromise CGNP survival, as we saw no increase phosphorylation, indicating that rapamycin is inhibiting in cleaved caspase-3-positive cells (data not shown). mTORC1 kinase activity in CGNPs (Figure 3b). Rapa- These results suggest a requirement for eIF4E in CGNP mycin did not decrease eIF4E or cyclin D2 protein levels proliferation, as well as indicating a role for S6K activity at 24 h (Figure 3b), indicating that rapamycin-sensitive as CGNPs leave the cell cycle.

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1789

Rapa 10nm (hr): --6 3 12 24 Shh (48 hr):- + ++ + + eIF4E Rapa (10nM): Cyclin D2 Shh (48 hrs): PS235/236 rpS6 PS235/236 rpS6

PS240/244 rpS6 PS240/244 rpS6

rpS6 rpS6 PT37/404EBP2 β-Tubulin 4EBP2

β-Tubulin % PCNA+ cells

Figure 3 Shh induction of eIF4E does not depend on rapamycin-sensitive mTOR activity. (a) Quantitative PCR analysis performed on CGNPs shows that eIF4E expression increases over periods of prolonged Shh incubation suggesting an indirect upregulation of eIF4E mRNA n ¼ 3 (***P ¼ 0.0005). Incubation with mTORC1 inhibitor rapamycin (10 nM) for 12 or 24 h did not effect eIF4E induction, however, the smoothened antagonist, SANT-2 (100 nM), prevented Shh-mediated eIF4E induction (**P ¼ 0.0012), (b) western blot analysis from CGNPs treated with the mTORC1 inhibitor, rapamycin (10 nM), for increasing periods of time. Suppression of mTORC1 activity, confirmed by 4EBP2 and rps6 phosphorylation, did not alter eIF4E or cyclin D2 levels. (c) Western blot analysis from vehicle- and Shh-treated CGNPs treated with rapamycin (10 nM) for 12 h to confirm suppression of mTOR activity in untreated CGNPs based on rpS6 phosphorylation. (d) Quantification of proliferation differences based on PCNA staining in vehicle- and Shh-treated CGNPs after addition of rapamycin for either 24 or 48 h n ¼ 4 (**Po0.001). eIF4E overexpression is sufficient to drive cultures. As demonstrated by western blot analysis, Shh-independent CGNP proliferation vehicle-treated b-actin-Eif4e CGNPs have increased As we observed that eIF4E is upregulated in the eIF4E levels compared with vehicle-treated WT but presence of Shh and is required for CGNP proliferation, similar to Shh-treated WT and b-actin-Eif4e CGNPs we wished to determine how eIF4E overexpression (Figure 4f). Furthermore, the increase in eIF4E is affects CGNP proliferation. To this end we used a sufficient to drive CGNP proliferation independent of recently described in vivo mouse model, in which eIF4E Shh as evidenced by cyclin D2 levels in vehicle-treated is ubiquitously overexpressed under control of the b-actin-Eif4e CGNPs. This ability of b-actin-Eif4e b-actin promoter (b-actin-Eif4e) (Ruggero et al., 2004). CGNPs to proliferate in the absence of Shh was As adults, these mice develop lymphomas, angiosarco- confirmed by quantification of BrdU incorporated cells mas, adenocarcinomas and hepatomas, but they do not (Figures 4g and h), which demonstrates that increased develop medulloblastomas, indicating that eIF4E over- eIF4E results in twice as many cycling cells compared expression alone is not sufficient to transform CGNPs with WT (Po0.0003). Interestingly, though increased in vivo. We isolated CGNPs from b-actin-Eif4e mice or eIF4E does not cooperate with Shh to promote WT age-matched controls mice and established primary proliferation, as the number of BrdU-positive cells in

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1790 eIF4E shRNAs: -- + GFP shRNA: -+ - 48 hr Shh: -+ + BrdU/DAPI BrdU/DAPI BrdU/DAPI

eIF4E

PS235/236-rpS6

S240/244 P -rpS6 Veh Shh Shh+eIF4E shRNAs

rpS6 * eIF4G 60

IRS1 40 Cyclin D2

PT37/464EBP2 20 % BrdU + cells

4EBP2 0 β-tubulin untreated Shh Shh+elF4EshRNAs

WT β-actin Eif4e Vehicle-treated Shh-treated 24hr Shh: - +- + BrdU/DAPI BrdU/DAPI eIF4E

Cyclin D2 WT

β-tubulin

80 WT β-actin Eif4e BrdU/DAPI BrdU/DAPI 60

40 β-actin Eif4e

% BrdU+ cells 20

0 vehicle-treated Shh Figure 4 eIF4E expression is required for the full mitogenic response of CGNPs to Shh. (a) Western blot analysis from CGNPs infected with a pool of lentiviruses containing sequences targeting eIF4E (see also Supplementary Figure S1). Loss of eIF4E expression results in decreased proliferation, measured by cyclin D2, as well as recovery of S6K activity. (b–d) Immunohistochemical analysis of CGNPs infected with shRNA containing lentiviruses. Effects on proliferation quantified by measuring BrdU incorporation in untreated (b), Shh-treated, GFP shRNA infected (c) and Shh-treated eIF4E shRNAs infected (d) CGNPs demonstrating a 50% reduction in cycling cells n ¼ 3(*Po0.02); (e) when eIF4E was knocked down. (f) Western blot analysis from b-actin-Eif4e or WT CGNPs demonstrating that increased eIF4E levels are sufficient to promote Shh-independent proliferation. Vehicle-treated b-actin- Eif4e CGNPs have increased cyclin D2 levels compared with WT. (g) Immunohistochemical analysis of proliferation by BrdU incorporation (green) in vehicle and Shh-treated WT or b-actin-Eif4e CGNPs. (h) Differences in proliferation quantified by comparing number of BrdU incorporated cells in vehicle-treated WT and b-actin-Eif4e CGNPs or Shh-treated WT and b-actin-Eif4e CGNPs. n ¼ 3 (***Po0.0003).

Shh-treated WT and b-actin-Eif4e CGNPs were similar. asked how CGNPs responded to differentiation cues These results suggest that increased eIF4E levels alone with respect to S6K activity. We incubated Shh-treated can promote CGNP proliferation independent of Shh, CGNPs with bFGF, which has been reported to induce however, this increase is insufficient to cause trans- CGNP differentiation (Fogarty et al., 2007). Under formation as b-actin-Eif4e CGNPs still maintain cell conditions of bFGF treatment, a dramatic upregulation intrinsic cues to exit the cell cycle and initiate of phosphorylated rpS6 occurred (Figure 5a). Western differentiation programs. blot analysis further supports the conclusion that FGF treatment caused cell cycle exit as indicated by decreased cyclin D2 and N-myc levels (Figure 5a). We also S6K drives CGNP cell cycle exit evaluated activity of other kinases, and consistent with On the basis of our initial western blot and immuno- our previous studies (Kenney and Rowitch, 2000; fluorescence observations, S6K activity seems to be Parathath et al., 2008) we observe no effects of Shh on induced at a point when CGNPs are poised to exit the the activity of these kinases (Figure 5a). Additionally, cell cycle, and pro-proliferative eIF4E activity antag- incubating Shh-treated CGNPs with BMP2, which has onizes S6K activity. To further investigate the connec- been shown to induce neuronal differentiation through tion between S6K and cell cycle exit in CGNPs, we Smad5-mediated mechanisms (Rios et al., 2004),

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1791 48 hr Shh:- + ++ + + 20 ng/mL bFGF (hr): -- 36 12 24 Infection: -GFP HA-S6K PS235/236rpS6 48 hr Shh: -+ + HA PS240/244rpS6

rpS6 S6 Kinase

eIF4E PT389S6 Kinase

Cyclin D2 PS235/236rpS6

Nmyc PS240/244rpS6 PT202Y204Erk rpS6 Erk

IRS1 PS473-Akt

Akt Cyclin D2

4EBP2 β-tubulin

β-tubulin

BrdU/DAPI BrdU/DAPI BrdU/DAPI

Veh Shh+GFP Shh+S6K

*** 50

40

30

20 % Brdu + cells 10

0 untreated Shh Shh+HA-S6K Figure 5 Increased S6K activity in proliferating CGNPs promotes cell cycle exit. (a) Western blot analysis of CGNPs treated with bFGF, which promotes CGNP cell cycle exit. S6K activity increases while the proliferation markers cyclin D2 and N-myc decrease. (b) S6K overexpression decreases CGNP proliferation, as determined by cyclin D2 levels. (c–e) Immunohistochemical analysis of CGNPs infected with S6 K-expressing retroviruses. Untreated- (c) Shh-treated infected with control GFP viruses (d) or Shh-treated infected with S6K viruses (e) were incubated for 48 h and then analyzed for proliferation by BrdU incorporation. (f) Proliferation of CGNPs was quantified by BrdU incorporation. Increasing S6K activity inhibited CGNP proliferation by an average of 20%. n ¼ 3 (***Po0.003). increased rpS6 phosphorylation while decreasing CGNP lend further support to our previous report where proliferation (data not shown). we demonstrated that CGNPs stabilize IRS1 through Because S6K activity correlated with CGNP differ- S6K suppression to promote proliferation (Parathath entiation signals, we next asked whether S6K over- et al., 2008). When we measured BrdU incorporation in expression is sufficient to force CGNPs to exit the cell Shh-treated S6K-infected CGNPs (Figures 5c–e), we cycle. We infected primary CGNP cultures with retro- observed a significant decrease in the number of BrdU- viruses expressing S6K and assayed proliferation by positive cells as compared with control (GFP-infected) carrying out western blot analysis and measuring BrdU cells (Po0.003) (Figure 5f). The reduction of proliferat- incorporation. As shown in Figure 5b, overexpression of ing CGNPs in S6K-infected cultures cannot be attrib- S6K resulted in increased rpS6 phosphorylation and uted to an increase in apoptosis as determined by reduced IRS1 and cyclin D2 protein levels. These data quantification of cleaved caspase-3-positive cells (data

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1792 not shown). The inhibitory effects on proliferation as a proliferation, perhaps by inhibiting S6K. We next result of increased S6K activity further support the wished to ask whether Shh-mediated B56g induction hypothesis that in CGNPs, the downstream mTOR might lie downstream of eIF4E. Indeed, when we effectors, eIF4E and S6K, carry out distinct roles in knocked down eIF4E with lentiviruses containing regulation of progenitor cell expansion and cell cycle shRNAs targeting eIF4E we observed a decrease in exit, respectively. B56g protein (Figure 6e). Taken together with our previous knockdown experiments where we observed rescue of S6K activity, these data support our hypo- CGNPs regulate S6K activity through activation thesis that PP2A–B56g mediates S6K suppression of a specific PP2A complex during CGNP proliferation; when CGNPs are cued to Our data demonstrate that Shh treatment results in exit the cell cycle, the PP2A-B56g complex is disrupted, increased levels of eIF4E protein in CGNPs, which can allowing for S6K activity, destabilization of IRS1 and be explained by increased eIF4E mRNA expression; completion of cell cycle exit. however, Shh caused decreased S6K activity, but did not It has been reported that the Drosophila ortholog of reduce total S6K or rpS6 levels. We therefore sought to this PP2A subunit mediates Drosphila S6K de-phos- determine the mechanism by which S6K activity is phorylation (Bielinski and Mumby, 2007), and that suppressed in proliferating CGNPs. We have shown that knocking down the mammalian ortholog increases S6K treatment with okadaic acid, a potent inhibitor of PP2A phosphorylation in HeLa cells (Hahn et al., 2010). Thus, function, promotes S6K activity in Shh-treated CGNPs we next asked whether S6K activity increases when the (Parathath et al., 2008), and we have also identified a B56g subunit is knocked down in proliferating CGNPs. positive relationship between Shh mitogenic signaling Lentiviral-mediated knockdown of the B56g subunit in and PP2A activity in that PP2A de-phosphorylates proliferating CGNPs resulted in increased S6K and rpS6 N-myc, thereby increasing its stability (Sjostrom et al., phosphorylation but no differences in 4EBP2 phosphor- 2005). Previous studies have suggested that mTOR may ylation, indicating that B56g loss specifically relieves the bind to and inhibit PP2A (Hartley and Cooper, 2002). inhibition of S6K (Figure 6f). Moreover, B56g knock- However, these observations do not explain why 4E-BP down resulted in decreased cyclin D2 and IRS1 levels, as phosphorylation is not suppressed by Shh in CGNPs, well as reduced proliferation in Shh-treated CGNPs as but that of S6K is. indicated by BrdU incorporation into cycling cells To further investigate the involvement of PP2A in (Po0.03) (Figures 6f and g). To address the specificity suppressing S6K activity we infected proliferating of the B56g subunit in mediating the suppression of S6K CGNPs with retroviruses carrying small-T antigen, activity we knocked down another subunit, B56e and which is known to antagonize PP2A (Chen et al., evaluated rpS6 phosphorylation. Loss of this subunit 2004). As shown in Figure 6a, overexpression of small-T did not rescue S6K activity and had little effect on antigen restored S6K activity and reduced levels of CGNP proliferation (Supplementary Figure 3). Taken proliferation mediators IRS1 and cyclin D2, suggesting together these data strongly suggest that Shh-mediated that PP2A may have a role in downstream mTOR induction of the PP2A–B56g complex downstream of signaling. We next measured the number of cells that eIF4E drives S6K de-phosphorylation and by extension incorporated BrdU in control or small-T infected maintains CGNP proliferation competency. CGNPs. As shown in Figure 6b, small-T infection resulted in a greater than 50% decrease in the number of BrdU-positive cells, further supporting our hypothesis Discussion that PP2A activity has a role in CGNP proliferation. As a trimeric protein complex, PP2A contains a In this study, we have investigated how the Shh catalytic subunit (C), a scaffold subunit (A), and a signaling pathway affects regulators of protein transla- regulatory ‘B’ subunit, which directs PP2A to its specific tion typically lying downstream of mTOR. Previously, substrates (Virshup and Shenolikar, 2009). We speculate we showed that Shh suppresses S6K activity in order to that Shh signaling could regulate a ‘B’ subunit targeting stabilize IRS1, through an unknown mechanism. In this S6K as a way to specifically de-phosphorylate S6K. We study, we show that Shh signaling modulates individual therefore examined expression of specific PP2A ‘B’ mTOR effectors separately in order to maintain a subunits in Shh-treated CGNPs. Using our primary proliferation-competent state. We identified a novel culture system we found that the protein levels of the mode of mTOR effector regulation wherein Shh induces B56g regulatory subunit were upregulated in Shh- eIF4E in CGNPs, which we show is required for their treated CGNPs (Figure 6c), whereas the B56e subunit proliferation, while simultaneously suppressing S6K levels, which have been previously demonstrated to activity via promoting its PP2A-mediated dephosphor- positively modulate Shh signaling during eye develop- ylation (Figure 7). Conversely, induction of S6K activity ment, remained constant (Rorick et al., 2007). Addi- through ectopic overexpression or PP2A inhibition tionally, treatment with the smoothened inhibitor results in CGNP cell cycle exit. cyclopamine (Berman et al., 2002) not only rescued Previous work performed in cell lines demonstrates S6K activity, but also reduced B56g protein levels that mTOR inhibits 4EBPs and activates S6K in order (Figure 6d and Supplementary Figure S3), further to promote protein synthesis (von Manteuffel et al., supporting a role for PP2A–B56g complex in promoting 1996; Fingar et al., 2004). However, immortalized cell

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1793

Infection: - GFP Small-T 48 hr Shh: - + + PS235/236rpS6 50 *** CGNP culture: 1 2 S240/244 P rpS6 40 48 hr Shh: - + -+

γ rpS6 30 PP2A B56 3 20 IRS1 PP2A B56ε % BrdU+ cells 10 Cyclin D2 0 β-tubulin untreated Shh+GFP Shh+Small-T GFP

β-tubulin

Shh (48 hr):- ++++ + Cyclop (hr): - - 3 6 12 24 PP2A B56γ3 GFP eIF4E Infection: - shRNA shRNA PS235/236rpS6 48 hr Shh: - + +

PS240/244rpS6 PP2A B56γ3

rpS6 β-tubulin Cyclin D2

β-tubulin

GFP B56γ Infection: - shRNA shRNA 48 hr Shh: - + + PP2A B56γ3

PP2A B56ε 60 *** PS235/236rpS6

PS240/244rpS6 40 rpS6

T389 P S6K 20 % BrdU+ cells S6K

IRS1 0 untreated Shh Shh+B56γ Cyclin D2 shRNA

PT37/46 4EBP2

4EBP2

β-tubulin

Figure 6 (a) Western blot analysis of lysates from CGNPs infected with retroviruses expressing small-T antigen. Small-T rescues rpS6 phosphorylation while decreasing proliferation, measured by cyclin D2 levels. (b) Effects on proliferation after control or small-T infection measured by BrdU incorporation, n ¼ 3 (***Po0.0008). (c) Evaluation of PP2A regulatory subunit protein levels in vehicle and Shh-treated CGNPs by western blot analysis reveals that Shh specifically upregulates the B56g3 subunit while B56e levels remain constant. (d) Western blot analysis demonstrating that treatment with cyclopamine (1 mg/ml) for increasing periods of time reduces levels of the PP2A B56g3 subunit, as well as rescues S6K activity. (e) eIF4E knockdown in Shh-treated CGNPs reduces B56g subunit levels. (f) Knockdown of the PP2A B56g3 subunit in Shh-treated CGNPs restores rpS6 phosphorylation, but does not affect 4EBP2 phosphorylation. Loss of this PP2A B subunit also results in reduction of IRS1 and cyclin D2 levels (see also Supplementary Figure S3). Loss of PP2A B56g3 subunit reduces proliferation by an average of 18% as determined by BrdU incorporation quantification (g); n ¼ 3 (**Po0.03). lines do not recapitulate the normal life cycle of most promote cell cycle progression; when cued to exit the cell types, especially those cells that dramatically alter cell cycle, CGNPs utilize S6K functions. However, we their morphology on differentiation. Using primary note that as ectopic S6K expression or B56g shRNA- cultures established from post-natal mouse cerebella mediated inhibition of S6K dephosphorylation did not allows us to more accurately determine how CGNPs completely eliminate CGNP proliferation, other down- regulate cellular processes, as these cultures have been stream effectors of smoothened, such as Gli, are likely to proven to mimic in vivo conditions. We propose that Shh have essential roles in achieving the complete mitogenic signaling in CGNPs preferentially activates eIF4E to response of CGNPs to Shh (Figure 7).

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1794 rate-limiting component of the initiation complex, eIF4E Shh binds to the cap structure on mRNAs and helps to recruit other initiation factors such as the scaffolding protein, eIF4G and an RNA helicase, eIF4A, to the mRNA. The cap structure consists of a 50 methyl-7- guanosine linked to the initial nucleotide of the mRNA Ptch Smo molecule and this cap functions to target the message for translation, as well as protect it from enzymatic S6K IRS 1 degradation. Assembly of the eIF4F complex on cellular mRNAs initiates the complex process of protein translation, which includes mRNA unwinding, initiation codon scanning as well as ribosome assembly. Response to growth factor signals regulates activity of eIF4E, but generally the overall levels of eIF4E remain limited in C cells and are kept inactive by translation inhibitor A proteins; therefore, capped mRNAs must compete with B56γ eIF4E Gli1, Gli2 N-myc the limited amount of eIF4E available in order to be translated. We find that during the rapid expansion phase of cerebellum development, Shh signaling induces expres- M sion and promotes increased protein levels of eIF4E, as G1 G2 well as other components of the translation initiation S machinery, including eIF4G. The eIF4E not only co- localizes with proliferation markers in vitro, but is also Figure 7 Model depicting how Shh promotes CGNP proliferation expressed in proliferating precursor cells of the EGL through combined effects on Gli1/2 and the mTOR pathway in vivo. The eIF4E promoter is known to contain myc- components eIF4E and S6K. Shh binding to Ptc leads to Smo binding sites and increased levels of cmyc have been activation, resulting in upregulation of Gli and N-myc. eIF4E mRNA and protein are upregulated, up-stream of the PP2A-B56g shown to induce eIF4E expression (Rosenwald et al., subunit, which directs PP2A activity towards S6K. Inactivation 1993). Shh signaling induces N-myc, which promotes (dephosphorylation) of S6K results in IRS1 stabilization, which expression of G1 cyclins and may provide a mechanism can also drive N-myc expression, thus creating a pro-proliferative for the induction of eIF4E (Ciemerych et al., 2002; feed-forward loop in CGNPs. Kenney et al., 2003; Oliver et al., 2003). On the basis of lines of evidence presented in this Although it is thought that rpS6 is an important study, we propose that Shh signaling in CGNPs regulator of proliferation-associated mRNAs containing preferentially activates eIF4E while simultaneously a highly structured 50-untranslated region, our findings inhibiting S6K activity through the activity of in neural precursors suggest that this downstream effect PP2A–b56g, thereby promoting cell cycle progression of S6K is not required for proliferation. Indeed, it has at least in part through stabilization of IRS1, which we been shown that mice expressing knocked-in mutant have shown drives expression of N-myc and cyclin D2 non-phosphorylatable rpS6 show increased prolifera- (Parathath et al., 2008). Although it has been shown tion, reduced size and enhanced protein synthesis in that ectopic eIF4E expression in cell lines can impair certain cell types (Ruvinsky et al., 2005), further S6K activity (Khaleghpour et al., 1999), the mechanism supporting alternate roles for S6K activity. However, a underlying this phenomenon was not determined, normal biological setting wherein rpS6 activity is and a relevant biological context has not previously reduced in developing cells has not been shown. In this been identified. Thus, our results demonstrate a study, we suggest that CGNPs exemplify a develop- novel mechanism wherein eIF4E and S6K are differen- mental paradigm for shifting the balance of mTOR tially regulated in proliferating cerebellar neural effectors between eIF4E and rpS6, as they proliferate precursors by Shh, to ultimately promote proliferation with a very short cell cycle (Mares et al., 1970), and these in primary neural precursor cells. Further, they cells are very small, consisting mostly of nucleus. During suggest that eIF4E inactivation may be an effective differentiation, they extend neural processes, resulting in way to reduce proliferation in CGNP-derived medullo- an increased cell size, where a role for S6K may become blastomas. more important (Altman and Bayer, 1997; Shima et al., 1998; Bateman and McNeill, 2004). Furthermore other kinases known to target rpS6, such as p90RSK (Roux Materials and methods et al., 2007), may have a role in activating rpS6 phosphorylation to promote cap-dependent translation Mice while S6K mediated rpS6 phosphorylation leads to Harvest of cerebellar neural precursors and cerebella prepara- CGNP cell cycle exit. tion for histological analysis from murine neonates were Our data indicate an important role for eIF4E carried out in compliance with the Memorial-Sloan Kettering in Shh-mediated CGNP proliferation. A critical and Institutional animal care and use committee guidelines.

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1795 Cerebellar granule precursor cell cultures Immunohistochemistry and immunofluorescence Cerebella were isolated from PN day 5 Swiss-Webster (SW) Paraffin-embedded sections (5 mM) were de-waxed, rehydrated mice and primary cultures were prepared as described (Parathath and fixed in ice-cold acetone for 20 min, then boiled in 0.01 M. et al., 2008). Recombinant Shh (R&D Systems, Minneapolis, MN, citric acid for 15 min for antigen retrieval. Sections were USA, concentration 3 mg/ml) was used to maintain proliferation as blocked with 10% normal goat serum (Sigma) with 0.1% indicated. Where applicable, cyclopamine (1 mg/ml, kind gift of Triton X-100/phosphate-buffered saline. Primary antibodies Dale Gardner, USDA), rapamycin (10 nM,Sigma,StLouis,MO, included eIF4E, phospho-S6 (235/236), glial fibrillary acidic USA) or bFGF (20 ng/ml, Peprotech, Rocky Hill, NJ, USA) were protein, NeuN (Cell Signaling), MEF2D, p27 (BD-Pharmingen, added for indicated times after 24 h of Shh treatment. Franklin Lakes, NJ, USA), BrdU (Becton Dickinson, Franklin Lakes, NJ, USA) and PCNA (Calbiochem, Gibbstown, NJ, USA). After washing in phosphate-buffered saline, slides were Protein preparation and western blotting incubated with either goat anti-rabbit or goat anti-mouse fluo- Adherent and floating cells were collected, washed once with rescently tagged secondary antibodies (Invitrogen). Sections phosphate-buffered saline, then resuspended in lysis buffer and were mounted using Vectashield mounting media with DAPI processed as previously described (Kenney and Rowitch, 2000). (Vector Laboratories, Burlingame, CA, USA). For immuno- For each sample 30 mg were separated on 8 or 10% sodium histochemical DAB detection an automated staining processor dodecyl sulfate–polyacrylamide gels and then transferred to was used (Discovery, Ventana Medical Systems, Inc., Tucson, activated polyvinylidene difluoride membranes (Millipore, AZ, USA). Billerica, MA, USA). Western blotting was carried out accord- ing to standard protocols. Primary antibodies included eIF4E, phospho-eIF4E, eIF4G, phospho-eIF4G, 4EBP2, phospho- Quantitative PCR 4EBP2, rpS6, phospho-rpS6 (235/236 or 240/244), mTOR, RNA from CGNPs was isolated using the TRIZOL (Invitro- S6K, phospho-S6K (Cell Signaling, Danvers, MA, USA), Raptor, gen) reagent and resuspended in 35 ml of DEPC-treated Rictor (Bethyl Laboratories, Montgomery, TX, USA), cyclin D2, water. cDNA was generated with SuperScript First-Strand N-myc (Santa Cruz Biotechnology, Santa Cruz, CA, USA), GFP Synthesis System for reverse transcriptase–PCR (Invitrogen). (Invitrogen, Carlsbad, CA, USA), HA (Chemicon, Billerica, MA, TaqMan Gene Expression Arrays (Applied Biosystems, USA), PP2A B56g (Abcam,Cambridge,MA,USA),PP2AB56e Carlsbad, CA, USA) using TaqMan custom designed (Novus Biologicals, Littleton, CO, USA) and b-tubulin (Sigma). MGB probes for eIF4E (Mm01317468_m1) and b-actin Horseradish–peroxidase-conjugated secondary antibodies were (Mm01191484_m1) were performed in triplicate on an ABI anti-mouse (Jackson Labs, Bar Harbor, ME, USA) or anti-rabbit 7000 Sequence Detection System. Data were analyzed with (Pierce, Rockford, IL, USA). Blots were developed using ABI GeneAmp SDS software (Applied Biosystems). The Amersham (Amersham, ECL kits, Piscataway, NJ, USA). average threshold cycle (CT) was determined to quantify Chemiluminescence was detected by exposing membranes to transcript levels, normalized against b-actin and the results Kodak biomax film for various intervals to obtain a non- reported as fold changes. saturated image. Image capturing Immunostaining performed on cultured CGNPs or tissue Immunoprecipitation sections was visualized using a Leica DM5000B microscope A measure of 1 mg of protein extract was used from either and images were captured with Leica FW400 software (Solms, vehicle-treated or Shh-treated CGNP cultures. A total of 10 mg Germany). For quantification of BrdU incorporation into of antibody were incubated with Protein A-Sepharose beads primary cells, TIFF images of four random fields were taken (Invitrogen) for 2 h. Protein extracts were pre-cleared with for each experimental group using the 10X objective. The Protein A-Sepharose beads for 2 h, and then incubated with percentage of BrdU-positive cells over the total number the antibody plus Protein A-Sepharose solution overnight at of cells, as determined by DAPI staining, was calculated 4 1C. The precipitate was washed four times and proteins were using Image Pro Plus software (MediaCybernetics, Bethesda, eluted with 0.2 M glycine. Antibodies used for immunopreci- MD, USA). pitation were eIF4E (Santa Cruz Biotechnologies), eIF4G (Cell Signaling) and mouse and rabbit IgG (Upstate Statistics Biotechnologies, Billerica, MA, USA). Statistical analysis was performed using one-way ANOVA followed by a two-tailed t-test for comparisons between Virus production and infection groups. All results are given as mean±s.e.m. All in vitro Lentiviruses were produced as previously described (Parathath experiments were performed at least three times with separate et al., 2008). Briefly, 293 EBNA (Invitrogen) packaging cells were litters to confirm reproducibility and consistency. co-transfected with lentiviral constructs (Mission shRNA, Sigma) expressing shRNAs targeting eIF4E, PP2R5c, PP2R5e or GFP, delta 8.9, and vesicular stomatitis virus G glycoprotein (vsvg) Conflict of interest plasmids, using Fugene 6 transfection reagent (Roche). The media was changed at 12 h after transfection and supernatants (10 ml) The authors declare no conflict of interest. were harvested every 24 h for 72 h and kept at 4 1C, then pooled, filtered through 0.45 mm syringe filters, concentrated by centrifuga- tion, and stored at 4 1C until use. CGNPs were infected with 50-ml concentrated viral supernatant and maintained in N2 media with Acknowledgements Shh. For retroviral production HA-S6K, provided by John Blenis (Harvard), and small-T, provided by William Hahn (Harvard) We thank John Blenis (Harvard Medical School) for providing were cloned into pWZL IRES-GFP vector and co-transfected with the HA-tagged S6 kinase plasmid. These studies were gagpol and vsvg into 293E packaging cells. Collections and supported by grants to AMK from the NINDS infections were carried out similar to lentivirus procedures. (R01NS061070) and the Handler Foundation.

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1796 References

Altman J, Bayer SA. (1997). Development of the cerebellar system in Hartley D, Cooper GM. (2002). Role of mTOR in the degradation of relation toits eveolution, structure, and functions. CRC Press: Boca IRS-1: regulation of PP2A activity. J Cell Biochem 85: 304–314. Raton, FLA. Hartmann W, Koch A, Brune H, Waha A, Schuller U, Dani I et al. Bateman JM, McNeill H. (2004). Temporal control of differentiation (2005). Insulin-like growth factor II is involved in the proliferation by the insulin receptor/tor pathway in Drosophila. Cell 119: 87–96. control of medulloblastoma and its cerebellar precursor cells. Am J Berman DM, Karhadkar SS, Hallahan AR, Pritchard JI, Eberhart Pathol 166: 1153–1162. CG, Watkins DN et al. (2002). Medulloblastoma growth inhibition Hatten ME, Gao W-Q, Morrison ME, Mason CA. (1998). The by hedgehog pathway blockade. Science 297: 1559–1561. cerebellum: purification and co-culture of identified cell popula- Bielinski VA, Mumby MC. (2007). Functional analysis of the PP2A tions. In: Banker GaG K (ed). Culturing Nerve Cells, 2nd edn. subfamily of protein phosphatases in regulating Drosophila S6 MIT Press: Cambridge, MA, pp 419–459. kinase. Exp Cell Res 313: 3117–3126. Hay N, Sonenberg N. (2004). Upstream and downstream of mTOR. Burnett PE, Barrow RK, Cohen NA, Snyder SH, Sabatini DM. (1998). Genes Dev 18: 1926–1945. RAFT1 phosphorylation of the translational regulators p70 S6 Hepker J, Wang QT, Motzny CK, Holmgren R, Orenic TV. (1997). kinase and 4E-BP1. Proc Natl Acad Sci USA 95: 1432–1437. Drosophila cubitus interruptus forms a negative feedback loop with Chen W, Possemato R, Campbell KT, Plattner CA, Pallas DC, Hahn patched and regulates expression of Hedgehog target genes. WC. (2004). Identification of specific PP2A complexes involved in Development 124: 549–558. human cell transformation. Cancer Cell 5: 127–136. Ho KS, Scott MP. (2002). Sonic hedgehog in the nervous system: Ciemerych MA, Kenney AM, Sicinska E, Kalaszczynska I, Bronson functions, modifications and mechanisms. Curr Opin Neurobiol 12: RT, Rowitch DH et al. (2002). Development of mice expressing a 57–63. single D-type cyclin. Genes Dev 16: 3277–3289. Katoh Y, Katoh M. (2009). Hedgehog target genes: mechanisms of Corradetti MN, Guan KL. (2006). Upstream of the mammalian target carcinogenesis induced by aberrant hedgehog signaling activation. of rapamycin: do all roads pass through mTOR? Oncogene 25: Curr Mol Med 9: 873–886. 6347–6360. Kenney AM, Cole MD, Rowitch DH. (2003). Nmyc upregulation by D’Mello SR, Borodezt K, Soltoff SP. (1997). Insulin-like growth factor sonic hedgehog signaling promotes proliferation in developing and potassium depolarization maintain neuronal survival by distinct cerebellar granule neuron precursors. Development 130: 15–28. pathways: possible involvement of PI 3-kinase in IGF-1 signaling. Kenney AM, Rowitch DH. (2000). Sonic hedgehog promotes G(1) J Neurosci 17: 1548–1560. cyclin expression and sustained cell cycle progression in mammalian Dahmane N, Ruiz-i-Altaba A. (1999). Sonic hedgehog regulates neuronal precursors [In Process Citation]. Mol Cell Biol 20: the growth and patterning of the cerebellum. Development 126: 9055–9067. 3089–3100. Kenney AM, Widlund HR, Rowitch DH. (2004). Hedgehog and PI-3 Dahmane N, Sanchez P, Gitton Y, Palma V, Sun T, Beyna M et al. kinase signaling converge on Nmyc1 to promote cell cycle progression (2001). The Sonic Hedgehog-Gli pathway regulates dorsal brain in cerebellar neuronal precursors. Development 131: 217–228. growth and tumorigenesis. Development 128: 5201–5212. Khaleghpour K, Pyronnet S, Gingras AC, Sonenberg N. (1999). Dudek H, Datta SR, Franke TF, Birnbaum MJ, Yao R, Cooper GM Translational homeostasis: eukaryotic translation initiation factor et al. (1997). Regulation of neuronal survival by the serine-threonine 4E control of 4E-binding protein 1 and p70 S6 kinase activities. Mol protein kinase Akt. Science 275: 661–665. Cell Biol 19: 4302–4310. Dutton R, Yamada T, Turnley A, Bartlett PF, Murphy M. (1999a). Ma XM, Blenis J. (2009). Molecular mechanisms of mTOR-mediated Regulation of spinal motoneuron differentiation by the combined translational control. Nat Rev Mol Cell Biol 10: 307–318. action of Sonic hedgehog and neurotrophin 3. Clin Exp Pharmacol Mares V, Lodin Z, Srajer J. (1970). The cellular kinetics of the Physiol 26: 746–748. developing mouse cerebellum. I. The generation cycle, growth Dutton R, Yamada T, Turnley A, Bartlett PF, Murphy M. (1999b). fraction and rate of proliferation of the external granular layer. Sonic hedgehog promotes neuronal differentiation of murine spinal Brain Res 23: 323–342. cord precursors and collaborates with neurotrophin 3 to induce Marti E, Bumcrot DA, Takada R, McMahon AP. (1995). Require- Islet-1. J Neurosci 19: 2601–2608. ment of 19K form of Sonic hedgehog for induction of distinct Fernandez LA, Northcott PA, Dalton J, Fraga C, Ellison D, Angers S ventral cell types in CNS explants. Nature 375: 322–325. et al. (2009). YAP1 is amplified and up-regulated in hedgehog- Northcott PA, Fernandez LA, Hagan JP, Ellison DW, Grajkowska W, associated medulloblastomas and mediates Sonic hedgehog-driven Gillespie Y et al. (2009). The miR-17/92 polycistron is up-regulated neural precursor proliferation. Genes Dev 23: 2729–2741. in sonic hedgehog-driven medulloblastomas and induced by N-myc Fingar DC, Richardson CJ, Tee AR, Cheatham L, Tsou C, Blenis J. in sonic hedgehog-treated cerebellar neural precursors. Cancer Res (2004). mTOR controls cell cycle progression through its cell growth 69: 3249–3255. effectors S6K1 and 4E-BP1/eukaryotic translation initiation factor Oliver TG, Grasfeder LL, Carroll AL, Kaiser C, Gillingham CL, Lin 4E. Mol Cell Biol 24: 200–216. SM et al. (2003). Transcriptional profiling of the Sonic hedgehog Fogarty MP, Emmenegger BA, Grasfeder LL, Oliver TG, Wechsler- response: a critical role for N-myc in proliferation of neuronal Reya RJ. (2007). Fibroblast growth factor blocks Sonic hedgehog precursors. Proc Natl Acad Sci USA 100: 7331–7336. signaling in neuronal precursors and tumor cells. Proc Natl Acad Sci Parathath SR, Mainwaring LA, Fernandez LA, Campbell DO, USA 104: 2973–2978. Kenney AM. (2008). Insulin receptor substrate 1 is an effector of Gingras AC, Raught B, Sonenberg N. (1999). eIF4 initiation factors: sonic hedgehog mitogenic signaling in cerebellar neural precursors. effectors of mRNA recruitment to ribosomes and regulators of Development 135: 3291–3300. translation. Annu Rev Biochem 68: 913–963. Pause A, Belsham GJ, Gingras AC, Donze O, Lin TA, Lawrence Jr JC Gingras AC, Raught B, Sonenberg N. (2001). Regulation of et al. (1994). Insulin-dependent stimulation of protein synthesis translation initiation by FRAP/mTOR. Genes Dev 15: 807–826. by phosphorylation of a regulator of 50-cap function. Nature 371: Hahn H, Wojnowski L, Specht K, Kappler R, Calzada-Wack J, Potter 762–767. D et al. (2000). Patched target Igf2 is indispensable for the Pomeroy SL, Tamayo P, Gaasenbeek M, Sturla LM, Angelo M, formation of medulloblastoma and rhabdomyosarcoma. J Biol McLaughlin ME et al. (2002). Prediction of central nervous system Chem 275: 28341–28344. embryonal tumour outcome based on gene expression. Nature 415: Hahn K, Miranda M, Francis VA, Vendrell J, Zorzano A, Teleman 436–442. AA. (2010). PP2A regulatory subunit PP2A-B0 counteracts S6K Rios I, Alvarez-Rodriguez R, Marti E, Pons S. (2004). Bmp2 phosphorylation. Cell Metab 11: 438–444. antagonizes sonic hedgehog-mediated proliferation of cerebellar

Oncogene eIF4E and S6K in dividing cerebellar precursors LA Mainwaring and AM Kenney 1797 granule neurones through Smad5 signalling. Development 131: phosphorylation and turnover in neural precursors. Dev Cell 9: 3159–3168. 327–338. Rorick AM, Mei W, Liette NL, Phiel C, El-Hodiri HM, Yang J. Uziel T, Karginov FV, Xie S, Parker JS, Wang YD, Gajjar A et al. (2007). PP2A:B56epsilon is required for eye induction and eye field (2009). The miR-17B92 cluster collaborates with the Sonic Hedge- separation. Dev Biol 302: 477–493. hog pathway in medulloblastoma. Proc Natl Acad Sci USA 106: Rosenwald IB, Rhoads DB, Callanan LD, Isselbacher KJ, Schmidt 2812–2817. EV. (1993). Increased expression of eukaryotic translation initiation Virshup DM, Shenolikar S. (2009). From promiscuity to precision: factors eIF-4E and eIF-2 alpha in response to growth induction by protein phosphatases get a makeover. Mol Cell 33: 537–545. c-myc. Proc Natl Acad Sci USA 90: 6175–6178. von Manteuffel SR, Dennis PB, Pullen N, Gingras AC, Sonenberg N, Roux PP, Shahbazian D, Vu H, Holz MK, Cohen MS, Taunton J et al. Thomas G. (1997). The insulin-induced signalling pathway leading (2007). RAS/ERK signaling promotes site-specific ribosomal to S6 and initiation factor 4E binding protein 1 phosphorylation protein S6 phosphorylation via RSK and stimulates cap-dependent bifurcates at a rapamycin-sensitive point immediately upstream of translation. J Biol Chem 282: 14056–14064. p70s6k. Mol Cell Biol 17: 5426–5436. Ruggero D, Montanaro L, Ma L, Xu W, Londei P, Cordon-Cardo C von Manteuffel SR, Gingras AC, Ming XF, Sonenberg N, Thomas G. et al. (2004). The translation factor eIF-4E promotes tumor (1996). 4E-BP1 phosphorylation is mediated by the FRAP-p70s6k formation and cooperates with c-Myc in lymphomagenesis. Nat pathway and is independent of mitogen-activated protein kinase. Med 10: 484–486. Proc Natl Acad Sci USA 93: 4076–4080. Ruvinsky I, Sharon N, Lerer T, Cohen H, Stolovich-Rain M, Nir T Wallace VA. (1999). Purkinje-cell-derived Sonic hedgehog regulates et al. (2005). Ribosomal protein S6 phosphorylation is a determi- granule neuron precursor cell proliferation in the developing mouse nant of cell size and glucose homeostasis. Genes Dev 19: 2199–2211. cerebellum. Curr Biol 9: 445–448. Sarbassov DD, Ali SM, Sengupta S, Sheen JH, Hsu PP, Bagley AF Wechsler-Reya RJ, Scott MP. (1999). Control of neuronal precursor et al. (2006). Prolonged rapamycin treatment inhibits mTORC2 proliferation in the cerebellum by Sonic Hedgehog [see comments]. assembly and Akt/PKB. Mol Cell 22: 159–168. Neuron 22: 103–114. Shima H, Pende M, Chen Y, Fumagalli S, Thomas G, Kozma SC. Wechsler-Reya R, Scott MP. (2001). The developmental biology of (1998). Disruption of the p70(s6k)/p85(s6k) gene reveals a small brain tumors. Annu Rev Neurosci 24: 385–428. mouse phenotype and a new functional S6 kinase. EMBO J 17: Ye P, Xing Y, Dai Z, D’Ercole AJ. (1996). In vivo actions of insulin- 6649–6659. like growth factor-I (IGF-I) on cerebellum development in Sjostrom SK, Finn G, Hahn WC, Rowitch DH, Kenney AM. (2005). transgenic mice: evidence that IGF-I increases proliferation of The cdk1 complex plays a prime role in regulating N-myc granule cell progenitors. Brain Res Dev Brain Res 95: 44–54.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene