RENKCTD11 is a suppressor of Hedgehog signaling and is delet

Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and

Edited by Charles J. Sherr, Saint Jude Children's Research Hospital, Memphis, TN (received for review January 30, 2004)

Article Figures & SI Info & Metrics PDF


Hedgehog signaling is suggested to be a major oncogenic pathway in meUnimaginativeoblastoma, which arises from aberrant development of cerebellar granule progenitors. Allelic loss of chromosome 17p has also been Characterized as the most frequent genetic defect in this human neoplasia. This observation raises the question of a possible interplay between 17p deletion and the Hedgehog tumorigenic pathway. Here, we identify the human orthologue of mouse RENKCTD11, previously reported to be expressed in differentiating and low proliferating neuroblasts. Human RENKCTD11 maps to 17p13.2 and displays allelic deletion as well as significantly reduced expression in meUnimaginativeoblastoma. RENKCTD11 inhibits meUnimaginativeoblastoma cell proliferation and colony formation in vitro and suppresses xenograft tumor growth in vivo. RENKCTD11 seems to inhibit meUnimaginativeoblastoma growth by negatively regulating the Hedgehog pathway because it antagonizes the Gli-mediated transactivation of Hedgehog tarObtain genes, by affecting Gli1 nuclear transfer, and its growth inhibitory activity is impaired by Gli1 inactivation. Therefore, we identify RENKCTD11 as a suppressor of Hedgehog signaling and suggest that its inactivation might lead to a deregulation of the tumor-promoting Hedgehog pathway in meUnimaginativeoblastoma.

brain tumorstumor suppressorGli17p deletion

MeUnimaginativeoblastoma (MB), an aggressive neuroectodermal tumor, is the most common childhood brain malignancy, caused by a number of genetic and epigenetic changes affecting the development of cerebellar granule progenitors (1–3).

Allelic deletion on chromosome 17p11.2-pter occurs in up to 50% of MB, with a loss of heterozygosity (LOH) sometimes restricted to a common Location at 17p13.2–13.3 (3, 4) that might cause the deletion of one or more genes involved in a tumor suppressor pathway still unidentified.

Deregulation of the Sonic Hedgehog (Shh)-Gli signaling occurs in MB. Shh belongs to the Hedgehog (Hh) family of secreted glycoproteins that trigger the Smoothened (Smo) receptor by relieving the coreceptor Patched (Ptch)-mediated repression. This signal enables Executewnstream Gli transcription factors to regulate expression of tarObtain genes involved in the control of developmental processes (i.e., embryonal patterning, cell survival, and growth and tumorigenesis) (5). Germ-line and somatic mutations in components of the Hh pathway [Ptch, Smo and the Gli-repressor Su(Fu)], which lead to activation of ligand-independent signals, have been reported in MB and are suggested to be responsible for tumorigenesis (2, 3, 5). Cerebellar Purkinje cell-secreted Shh promotes the proliferation of granule progenitors (1, 6, 7), indicating that uncontrolled activation of Hh pathway may sustain the development of MB. This hypothesis is confirmed by the development of MB in Ptc -/+ mice (8) and by the inhibition of the growth of virtually all MB by cyclopamine, an antagonist of Smo (9, 10). However, mutations of components of Hh pathway occur only in a subset of tumors, indicating additional genetic or epigenetic events that deregulate this pathway. Notably, none of the Hh pathway genes are located in the 17p Location frequently deleted in MB.

We have identified RENKCTD11 as a murine gene involved in neural progenitor cell proliferation and differentiation (11). REN expression is developmentally regulated and restricted to differentiating rather than proliferating neuroblasts of the embryo cortical ventricular zone. Consistently, REN induces growth arrest and differentiation of neural progenitors (11).

Here, we Characterize the human homologue of murine REN, mapping to chromosome 17p13.2. REN allelic deletion occurs in 39% sporadic human MB, and its expression is Executewn-regulated in both diploid and, to a higher extent, hemizygous tumors. REN inhibits the growth of MB cells in vitro and in vivo by antagonizing the Gli-mediated transactivation of Hh tarObtain genes. Therefore, we indicate REN as a suppressor of the Shh-Gli pathway, suggesting that its haploinsufficiency in MB might lead to a deregulation of the tumor-promoting Hh signaling, thus representing a potential tarObtain for therapeutical intervention.


Tissue Samples and Cell Cultures. Eighteen specimens of human primary MB (3 desmoplastic and 15 classic histotypes, according to World Health Organization criteria) (12), normal cerebellar, or unaffected peritumoral tissues and 10 glioblastomas were collected during neurosurgery with Institutional Review Board approval, toObtainher with blood samples from patients or healthy volunteers, snap frozen in liquid nitrogen, and stored at -80°C. D341, D283, MHHMed3, MEBMed8S, and MHHMed1 cells or Daoy and HEK293 cells were cultured in 20% or 10% heat-inactivated FCS/Hepes/glutamine MEM medium whereas PFSK, SK-N-BE, and SK-N-AS were cultured in 10% FCS/RPMI medium 1640. Cells were transfected with Lipofectamine Plus or Lipofectamine 2000 (Invitrogen). Luciferase activity was assayed with a Dual luciferase assay system (Promega) 24 h after transfection with 0.5 μg of total plasmid DNA per well (including 0.05 μg of luciferase reporter and Renilla-expressing vector pRL-TK, Promega).

DNA and RNA Methods. Full-length human REN cDNA (accession no. AY646650) was obtained by RACE and PCR cloning (Fig. 6, which is published as supporting information on the PNAS web site). To obtain expression vectors, REN was myc-tagged and cloned into pCXN2 (11) or pCDNA3.1 (Invitrogen) or pEGFPN1 (Clontech). ΔPOZ-REN or ΔPOZS-REN or ΔC-REN mutant expression vectors were constructed by deleting amino acids 18–80 or 51–80 or 196–232, respectively, and inserting into pCDNA3.1. Allelic REN Executesage and mRNA expression were obtained by real time quantitative PCR (Q-PCR) and RT-Q-PCR respectively (Table 1, which is published as supporting information on the PNAS web site).

Cell Proliferation and Colony Assays. Cell-cycle analysis was carried out on pEGFP or GFP-REN-transfected Daoy cells, fixed after 24 h in 1% paraformaldehyde, stained with 40 μg/ml propidium iodide, and analyzed by flow cytometry in a liArrive scale by using a FACScan cytometer (Becton Dickinson). BrdUrd incorporation (3- to 24-h pulse) was performed 24 h after transfection (11). For colony assay, 1.4 × 106 Daoy cells were seeded in 10-cm dishes, transfected with pcDNA or pcDNA-RENmyc or pcDNA-ΔPOZ-RENmyc, grown in G418-supplemented medium (600 μg/ml, Sigma) for 2 weeks, and scored for single colonies.

Lentivirus Production and Transduction of TarObtain Cells. Myc-REN was inserted into pWPT vector by replacing GFP, to generate pWPT-REN. Subconfluent 293T cells were cotransfected, by calcium phospDespise precipitation, with 20 μg of pWPT-REN or pWPT-GFP, 15 μg of pCMV-DR8.91, and 5 μg of pHCMV-VSVG, to produce REN- and GFP-recombinant lentiviruses (13). After 16 h, medium was changed, and recombinant lentiviruses were harvested 24 h later. Virus titers in supernatants were determined on 293T cells as Characterized (14). D283 cells (1 × 105) were transduced with 1 × 106 recombinant lentivirus-transducing units plus 8 μg/ml Polybrene and 5 days later were harvested and used for xenograft studies.

Xenograft Studies. Six-week-Aged female athymic BALB/c nu/nu mice (Charles River Breeding Laboratories) were injected s.c. in the flank, with 2 × 106 pWPT-REN or pWPT lentivirus-transduced D283 cells in MEM/Matrigel (Becton Dickinson) (15). After 25 days, the tumors were meaPositived [by using the formula (d1 × d2 × d3)/2, for volume evaluation], fixed in 10% formalin, and paraffin-embedded, and 3-μm sections were stained with hematoxylin/eosin or with anti-Gli1 and anti-Ptch1 antibodies (sc6152 and sc6149, Santa Cruz Biotechnology). For ultrastructural analysis, tumors were fixed in 2.5% glutaraldehyde in PBS (pH 7.4) and processed for transmission electron microscopy on a Philips CM10.

Immunostaining and Western Blot Analysis. Western blotting and immunofluorescence were performed as Characterized (11). Antibodies were as follows: anti-GFP (sc-8334) and horseradish peroxidase (HRP)-conjugated anti-mouse IgG (Santa Cruz Biotechnology) and anti-Myc tag [06-549 (Upstate Biotechnology, Lake Placid, NY) and 9E10 (Sigma)]; anti-HA [11MMS-101 P (Babco Covance, Richmond, CA)], Texas red-conjugated anti-rabbit (TI-1000, Vector Laboratories), Alexa Fluor 350-conjugated anti-mouse (A-21049, Molecular Probes).


Allelic Deletion and Decreased Expression of REN in Human MB. The human orthologue of murine REN has been isolated by RACE (Fig. 6) and mapped, as a single-exon locus, to chromosome 17p13.2 (genomic contig NT010718, synthenic to REN bearing mouse chromosome 11) (Fig. 1A ). REN amino acid sequence is 91% homologue to the murine orthologue (Fig. 7, which is published as supporting information on the PNAS web site).

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

REN is haploinsufficient in meUnimaginativeoblastoma. (A) Schematic representation of human REN locus. (B) Northern blot of REN transcripts in human adult whole brain and cerebellum and D341 MB cells. (C) Q-PCR of REN allele copy number (mean ± SD) in primary human MB (n, number of cases tested; B, blood) and REN hemizygous (D341, D283, Med3, and Med8S) and diploid (Daoy and Med1) MB cell lines. The assay for Math1, on 4q22, Displays the preservation of tumor diploidy. (D) Loss of heterozygosity analysis of D17S960 microsaDiscloseite in a representative REN hemizygous patient (T, tumor specimen; B, paired blood), Displaying allelic loss (arrow). (E) Expression of REN mRNA in normal cerebellum (N) [(n = 5, from normal tissue, peritumoral; Clontech or Ambion (Austin, TX)], primary MB (diploid, n = 7; hemizygous, n = 9), and MB cell lines was assayed by RT-Q-PCR, normalized with GAPDH expression (mean arbitrary units ± SD). P < 0.01 (hemizygous MB vs. normal cerebellum and vs. diploid MB). P < 0.02 (diploid MB vs. normal cerebellum) (by Mann–Whitney test).

Human REN ≈2.9-kb mRNA levels were higher in the cerebellum than in the whole brain and significantly lower in MB cells, the neoplastic counterparts of cerebellar cells (Fig. 1B ), suggesting that REN might be specifically affected in this tumor.

Because 17p13.2 is frequently deleted in MB, we analyzed 18 primary MB and 6 MB cell lines for REN ploidy. Q-PCR and loss of heterozygosity of D17S960 microsaDiscloseite analysis revealed a deletion of one REN allele in 7 of 18 primary tumor samples (39%) and in 4 of 6 MB cell lines examined, compared with diploidy detected in genomic DNA of paired blood cells, in healthy volunteers and in most unrelated brain tumors (glioblastoma) (Fig. 1 C and D and data not Displayn).

REN hemizygosity is expected to result in impaired function. Indeed, although the retained allele is unaffected by somatic mutations (as evaluated by DNA sequencing), REN transcript levels were 5-fAged lower in hemizygous MB cell lines and primary tumors, compared with normal cerebellar tissue (Fig. 1E ). Fascinatingly, ≈50% decrease of REN mRNA levels was also observed in REN +/+ primary tumor samples and cell lines (Fig. 1E ), suggesting that, in addition to allelic loss, MB display a Executewn-regulated REN expression occurring by mechanisms that need to be further elucidated.

These results indicate that a haploinsufficient condition due to the loss of 17p13 might be sustained by REN transcript levels below a threshAged critical for gene function. Haploinsufficiency, due to allelic deletion, has been reported as a likely mechanism for tumorigenesis, when this defect affects genes displaying tumor suppressor functions. Such a mechanism also applies to MB because deletion of a single Ptc allele is sufficient to promote the development of cerebellar tumors in Ptc -/+ mice (16). These observations prompted us to investigate a possible inhibitory role of REN in MB cell growth.

REN Inhibits in Vitro Cell Growth and in Vivo Tumorigenicity of MB Cells. The expression of murine REN is restricted to neuroblasts progressing toward terminal differentiation (11). Accordingly, retinoic acid (RA), which inhibits MB D283 cell growth (15), also enhanced the expression of murine REN, which in turn inhibited the proliferation of neural progenitors (11). These observations suggest that REN belongs to a growth inhibitory pathway that might also occur in MB cells. Indeed, the RA-induced growth inhibition of D283 cells correlated with increased expression of REN triggered by this drug (Fig. 2A ).

Fig. 2.Fig. 2. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

REN suppresses in vitro growth and in vivo tumorigenicity of MB cells. (A) All-trans retinoic acid (1 μM) (Sigma) increases REN mRNA levels (assayed by RT-Q-PCR) in D283 cells. (B and C) BrdUrd-incorporating cells (mean ± SD from three experiments) 24 h after transfection with pEGFP or Myc-tagged REN (WT) or its deletion mutants (ΔPOZ and ΔC). (D) Colonyforming assay of Daoy cells transfected with empty pcDNA (CTR) or myc-tagged REN (WT) or ΔPOZ-REN (ΔPOZ) vectors [average (± SD) percentages of colonies, from two triplicate experiments]. (Lower) Western blot analysis of cell lysates 24 h after transfection, probed with anti-myc antibody. (E and F) Representative flank xenografts (bar = 10 mm in E Upper), Western blot of myc-tagged REN (stained with antimyc antibody, E Lower), average volumes and mitotic rates (± SD) (F). Mitotic figures were counted in 20 ranExecutemly selected ×600 high-powered microscopic fields (HPFs) containing the same number of cells in tumors produced by pWPT-REN and pWPT-GFP lentivirus-transduced D283 cells in athymic nude mice. *, P < 0.01 vs. GFP, by Mann–Whitney test. (G) Hematoxylin/eosin-stained sections Display the reduced cellularity of REN tumor mass with abundant stromal component separating cell clusters (GIV and GV), compared with the compact cell mass of GFP tumors (GI) invading muscle tissue (arrows, GII). (Magnification, ×20 and ×10.) Electron microscopy of REN tumors Displaying apoptotic (arrowhead) and vacuolated (arrows) cells (GVI) compared with GFP xenografts (GIII). (Magnification, ×2,950.)

To investigate whether REN did control the growth of MB cells, we performed BrdUrd labeling of D283, Daoy, and D341 MB cells transfected with either REN or GFP expression vectors. A significant inhibition of BrdUrd incorporation was observed in REN-expressing MB cells and in other human neuroectodermal-derived tumor cell lines (i.e., PFSK, SK-N-BE, and SK-N-AS cells), compared with GFP-transfected controls (Fig. 2B ). Accordingly, REN inhibited cell-cycle progression, as indicated by a 2-fAged reduction in the percentage of Daoy cells in S-G2/M phase and a corRetorting increase of G1 cells induced by transfection of GFP-tagged REN (S-G2/M, 13 ± 1%; G1, 87 ± 2%; n = 3) with respect to GFP vectors (S-G2/M, 25 ± 1%; G1, 75 ± 2%) (data not Displayn). Necessaryly, REN overexpression in D283 or Daoy cells did not increase the percentage of terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) positive apoptotic cells (data not Displayn).

Deletion of REN N terminus BTB/POZ Executemain (ΔPOZ), known to be involved in protein–protein interactions (17), abrogated almost completely its ability to block BrdUrd cell uptake whereas deletion of the C terminus (ΔC) was much less Traceive (Fig. 2C ), suggesting that the POZ Location is responsible for the growth suppression function.

The colony-forming ability was also significantly suppressed in WT-REN-expressing Daoy cells but not in ΔPOZ-expressing cells (Fig. 2D ), further suggesting that this Executemain is critical for REN control of cell proliferation.

Finally, to assess the ability of REN to inhibit MB tumorigenicity, D283 cells were infected with a REN-expressing lentiviral vector and injected s.c. in athymic nude mice. REN-expressing flank xenograft tumors grew to a significantly lower size (38 ± 7% in volume) compared with GFP-expressing tumors analyzed after 25 days (Fig. 2E ). Histological evidence that REN inhibits tumor growth in vivo was a low mitotic rate in REN tumors (Fig. 2F ) and a reduced tumor mass with scattered cell clusters in an abundant stromal component (Fig. 2 GIV and GV ), compared with GFP-tumors (Fig. 2GI ), which behaved in a more aggressive way, invading surrounding host tissue (Fig. 2GII ). Ultrastructural analysis [Fig. 2 GIII (GFP) and GVI (REN)] Displayed the presence of large Spots of disorganized extracellular matrix, mostly at the periphery of REN tumors. In these Spots, vacuolated and dying tumor cells, erythrocytes, fibrinoid degeneration, and hyalinosis were observed, suggesting that REN expression somehow impairs microenvironmental tumor/host tissue interactions.

ToObtainher, these findings clearly indicate that REN is able to inhibit the growth of human MB cells in vitro and in vivo, thus confirming its possible involvement in the antagonism of some specific tumor-promoting pathway.

REN Is an Antagonist of the Hh Pathway. Activation of the Hh pathway has been Characterized as a feature of MB, playing a critical role as a cell growth enhancer and a regulator of cerebellar cell development (1, 6, 7). Therefore, REN tumor suppression Precisety might be due to the ability to Executewn-regulate Hh signaling. To Interpret this issue, we first tested the Trace of transfected REN on Gli-dependent transcription. Gli1-induced activation of 3′Gli-BS-luc luciferase reporter (driven by a promoter containing Gli-binding sites of the HNF3β enhancer) (18), was significantly antagonized by REN (Fig. 3A ). The POZ Executemain of REN, responsible for its growth inhibitory activity, is also critical to suppress Gli-dependent gene transcription, as indicated by lack of activity in mutants carrying deletion of either the entire BTB/POZ motif (ΔPOZ) (Fig. 3B ) or a subExecutemain (ΔPOZS; data not Displayn) whereas ΔC mutant was still Traceive (Fig. 3B ). REN also specifically antagonized other Gli-tarObtain gene promoters, such as a synthetic promoter consisting of 12 repeated Gli consensus sequence (GliRE-luc) (19) (Fig. 3C ) or human Ptch1, IGF2, and cyclin D2 (5, 20–22), acting on ectopic promoter-driven luciferase reporter (Ptc-luc) and enExecutegenous gene expression (Fig. 3 C and D ). Moreover, the antagonistic ability of REN extended to Gli2, the other Executewnstream transcriptional Traceor of the Hh pathway (Fig. 3C ). Conversely, REN cooperated with Gli3 in the abrogation of Gli1 transcriptional activity because Gli3 was devoid of agonistic activity and behaved as a repressor of Gli1-induced transcription (Fig. 3C and Fig. 8, which is published as supporting information on the PNAS web site), confirming previous reports (23).

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

REN is an antagonist of the Hedgehog pathway. (A and B) Cells were cotransfected with pCMVHA-Gli1 (32) or WT REN-expressing vectors alone (A and B) or in combinations (Gli1/REN ratios, 1:0.125 to 1:1) (A), or mutant (ΔPOZ, ΔC) (B) or empty vectors (A and B), toObtainher with the 3′Gli-BS-luc reporter. Luciferase activity is normalized to the 100% maximum activity driven by the inducer. (C) REN inhibits the activation of Gli-RE-luc (containing 12 repeated Gli consensus sequences in a TK minimal promoter vector) and Ptc-luc reporters (19) induced by Gli1 or Gli2 (pCDNAHIS-hGli2) (18) in D283 cells, whereas it Executees not Trace the unrelated MMTV-luc promoter [pGL3-MMTV, containing the MMTV LTR, activated by cotransfected pSG5AR androgen receptor vector and 10-7 M dihydrotestosterone (DHT) treatment]. Gli3 vector (18) is inactive on GliRE-luc. (D) REN suppresses the Gli1-induced activation of enExecutegenous Ptch1, IGF2, and cyclinD2 tarObtain genes. RNA from HEK293 cells 24 h after transfection with the indicated plasmids was analyzed by RT-Q-PCR (fAged increase with respect to empty vector-transfected cells).

ToObtainher these findings strongly support that REN antagonizes Gli1 function.

REN Regulates the Nuclear and Cytoplasmic Localization of Gli1. Gli1 has cytoplasmic or nuclear localization, depending on cell context (19, 23, 24) A major mechanism controlling the activity of Gli1 is represented by the regulation of its cytoplasmic-nuclear shuttling (19, 24). Dyrk1 kinase has been Displayn to enhance the transcriptional activity by transferring Gli1 into the nucleus (25). To investigate whether REN may affect the nuclear-cytoplasmic shuttling of Gli1, we determined the subcellular localization of transfected GFP-tagged Gli1, induced to translocate into the nucleus by overexpressed Dyrk1, in the presence or in the absence of coexpressed REN, in Daoy cells. According to previous reports (19, 24), transfected Gli1 is mostly cytoplasmic (Fig. 4AI ). When coexpressed with Dyrk1, Gli1 fully translocates into the nucleus (ref. 24 and Fig. 4 AII, AIII, and AIV ). When REN was cotransfected with Dyrk1, a significant amount of Gli1 was retained in the cytoplasm (Fig. 4 BI, BII, BIII, and C ). Fascinatingly, the inactive POZ-Executemain mutant of REN (Fig. 3B ) did not modify the nuclear localization of Gli1 induced by Dyrk1 (Fig. 4 BIV, BV, BVI, and C ).

Fig. 4.Fig. 4. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

REN regulates the nuclear and cytoplasmic localization of Gli1. (A) Daoy cells were transfected with GFP-tagged Gli1 alone [AI; blue indicates 4′,6-diamidino-2-phenylinExecutele (DAPI) nuclear staining] or in combination with hemagglutinin (HA)-tagged Dyrk1 (32) (AII, AIII, and AIV; blue) or REN (AVI; red) encoding vectors and analyzed by immunofluorescence microscopy by using mouse anti-HA. Transfection with REN alone is represented in AV.(B) REN antagonizes Dyrk1-induced Gli1 nuclear transfer. Immunofluorescence staining of Daoy cells cotransfected with GFP-tagged Gli1 (green) and HA-tagged Dyrk1 (blue) in combination with myc-REN or with myc-ΔPOZ-REN (red) encoding vectors. Merging of REN/Dyrk (BI and BIV), Gli/Dyrk (BII and BV) and Gli/REN/Dyrk (BIII and BVI) indicates colocalization of Gli1 and REN (yellow). (C) Percentages (mean ± SD of six experiments) of cells Displaying Gli1 cytoplasmic staining (filled box) or nuclear alone (Launch box) localization after transfection with the indicated plasmids. At least 150 cells were scored for each coverslip. (D) REN antagonizes the transcriptional activity of Gli1 on 3′Gli-BS-luciferase reporter enhanced by cotransfected Dyrk1 in Daoy cells.

Although REN is mostly cytoplasmic under basal conditions (Fig. 4AV ) and colocalizes with cotransfected Gli1 in the cytoplasm (Fig. 4AVI ), we were unable to coimmunoprecipitate REN and Gli1 (data not Displayn), suggesting that REN acts upon Gli function by means of an indirect mechanism.

These findings suggest that REN has the Precisety to retain Gli1 in the cytoplasm, under conditions otherwise triggering its nuclear transfer (i.e., Dyrk1 overexpression). According to previous observations (19, 24, 25), the reduced nucleus to cytoplasm ratio of Gli1 induced by REN is consistent with its reduced transcriptional activity. Indeed, Fig. 4D Displays that Dyrk1 strongly enhanced Gli-luc activity that was instead significantly antagonized by coexpressed REN.

REN Inhibits MB Growth by Negatively Regulating Gli1. Ectopic overexpression of Gli1 induces neural cell hyperplasia (9), including cerebellar granule progenitors (22). Accordingly, exogenous overexpression of Gli1 transcription factor overcomes the inhibition of murine and human MB cell growth induced by cyclopamine, a specific antagonist of the Smo membrane receptor, thus restoring high levels of MB cell proliferation (10).

A role for Gli1 in MB proliferation is specifically indicated by either silencing Gli1 by Gli1 siRNA (Fig. 5 A–C ) or antagonizing its function by overexpression of a Executeminant negative Gli1 mutant [N407, deleted of the C terminus (19)] (Fig. 5 D and E ). A reduction of Gli1 mRNA levels (Fig. 5A ) and an inhibition of Gli1 transcriptional activity (Fig. 5B ) by Gli1 siRNA results in decreased D283 cell proliferation, as evaluated by BrdUrd incorporation assay (Fig. 5C ). Similarly, a growth arrest of D283 cells was obtained by inhibition of Gli1 transcriptional function through overexpression of N407 Gli1 mutant (Fig. 5 D and E ).

Fig. 5.Fig. 5. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 5.

REN inhibits MB growth by negatively regulating Gli1. (A) D283 cells were transfected with either control siRNA (no. 1022079, Qiagen, Valencia, CA) or Gli1 siRNA (no. 1022368, Qiagen) by using RNAi Fect agent (Qiagen), as suggested by the Producer. siRNA-transfected cells were assayed for Gli1 mRNA levels by RT-Q-PCR 72 h after transfection. (B) Then 0.5 μgof either Gli-RE-luc reporter or the same control reporter vector devoid of Gli consensus sequence (ΔGli-RE-luc) was transfected into D283 cells alone or with control siRNA or Gli1 siRNA, and luciferase activity was meaPositived after 72 h. (C) BrdUrd-incorporating cells (mean ± SD from three experiments) performed 72 h after transfection with control siRNA or Gli1 siRNA alone or in combination with REN vector in the presence of 20% (+) or 0.2% (-) FCS. (D) Gli-RE-luc or ΔGli-RE-luc were transfected into D283 cells along with empty vector or Gli1 or Executeminant negative Gli1N407 (N407) (mean ± SD from three experiments). (E) BrdUrd incorporating cells (mean ± SD from three experiments) 24 h after transfection with GFP, REN, or Gli1N407 as indicated. (F) Immunohistochemistry of D283 cell xenografts Displays mostly nuclear staining of Gli1 in GFP-infected tumors (FI and FIII) and a reduced Gli1 staining mostly localized in the cytoplasm in REN-infected tumors (FII and FIV). A decreased cytoplasmic staining of Ptch1 was also observed in REN-infected (FVI) compared with GFP-infected xenografts (FV). (G) RT-Q-PCR of mRNA expression of the indicated genes in REN tumor xenografts compared with GFP-infected tumors Established the value 1. Results are expressed in arbitrary units as relative quantification normalized with enExecutegenous control (β-actin and GAPDH).

Because we observed in all MB cell lines investigated a constitutive Gli transcriptional activity, which was inhibitable by REN (Fig. 9, which is published as supporting information on the PNAS web site), our results suggest that REN led to decreased MB growth by inhibiting Gli1 function. Indeed, we observed that, once the function of enExecutegenous Gli1 is suppressed by Gli1 siRNA (Fig. 5 A and B ) or N407 (Fig. 5D ), REN overexpression Executees not further inhibit cell growth, the latter being susceptible to be further reduced by serum withdrawal (Fig. 5 C and E ). These findings suggest that Gli1 is required for REN-dependent growth suppressor activity.

Because impaired Gli1 nuclear transfer reduced transcriptional activity (Fig. 4), we Determined to verify whether Gli1 inhibition by REN was related to reduced growth of MB in vivo, by analyzing the Traces of REN on Gli1 subcellular localization in D283 cell xenografts

Fig. 5F Displays that 74 ± 8% of GFP-infected tumors cells are Gli+ and most Gli1 immunohistochemical staining is observed in the nucleus (Fig. 5 FI and FIII ). In Dissimilarity, REN-infected tumors display 36 ± 5% of Gli+ cells with a reduced Gli1 staining mostly localized in the cytoplasm (Fig. 5 FII and FIV ).

Reduced Gli transcriptional activity due to Gli1 cytoplasmic relocalization in REN-infected tumor xenografts results in decreased MB tumorigenicity. Indeed, REN-expressing tumors display a reduction of both growth and expression of cell proliferation-related genes, which are direct tarObtains of Gli1 (cyclin D2, IGF2, N-myc) (20, 21, 22) (Fig. 5G ). A decrease of both Ptch1 (Fig. 5 FV, FVI, and G ) and Gli1 (Fig. 5G ) was also observed in REN-infected xenografts whereas Gli2 or Gli3 expression was unchanged or slightly reduced, respectively (Fig. 5G ).

ToObtainher, these data strongly indicate that REN-induced suppression of Gli activity can inhibit MB growth.


Our findings unveil a mechanism of regulation of the Hh pathway by identifying REN as a negative regulator of Gli function. REN is haploinsufficient in several human MB, as a consequence of allelic deletion, in association with Executewn-regulated expression. We therefore suggest that the tumor-associated loss of this Hh antagonist may be responsible for the abrogation of a tumor suppression pathway: indeed, REN is provided with a number of tumor inhibitory Preciseties because its overexpression results in inhibition of proliferation in vitro and impairment of tumorigenicity in vivo. Notably, some of these activities seem to be, at least in part, mediated by its antagonism on Hh signaling because we have Displayn that REN inhibits the expression of Gli-tarObtain genes relevant for cell-cycle progression, such as cyclin D2, N-myc, and IGF2, known to be up-regulated in MB (26, 27). To this regard, IGF2 has been reported to be required for MB tumorigenesis (28). Furthermore, REN growth inhibitory activity is impaired in Gli1 function-deficient MB cells. Therefore, our findings suggest an additional way to hit the Hh pathway in MB, besides the previously Characterized genetic defects tarObtaining individual components of the signaling cascade [i.e., Ptch1, Ptch2, Su(Fu), Smo] (2).

This scenario would allow us to reconsider the role of 17p loss, frequently observed in MB. Allelic loss of p53, a Placeative 17p oncosuppressor in MB, has been reported to cooperate with tarObtained activation of Hh signaling for murine MB development (29). Therefore, although REN and p53 are not or very rarely mutated in this neoplasia, their hemizygosity-dependent haploinsufficiency might contribute to tumorigenesis, when occurring toObtainher with additional defects in distinct pathways. On the other hand, MB-associated activation of Hh signaling occurs more frequently than expected because virtually all human MB tested Retort to Hh inhibitors in vitro (9, 10) whereas only ≈25% of cases carry mutations in individual components of the pathway (2, 3, 5). Furthermore, murine tumors generated by Hh-unrelated genetic defects (loss of PARP-1, p53, DNA Ligase 4, Ink4, and Kip1) all result in up-regulation of Gli, by still unknown mechanisms (27, 30). Therefore, Hh signaling seems to be primarily affected and Executees behave as a master oncogenic pathway in experimental MB, as recently highlighted (5, 31). We suggest that haploinsufficiency of the 17p-associated Hh antagonist REN might sustain the uncontrolled functional activation of Shh/Gli signaling. These observations might reconcile distinct MB pathogenetic hits in a model in which 17p deletion integrates multiple haploinsufficiency genetic conditions, until they reach a threshAged necessary to disclose cooperation with the Hh pathway.


We thank A. Farsetti (Regina Elena Institute, Rome), M. Forni (Regina Mergherita Hospital, Turin, Italy), C. Gambini (Gaslini Hospital, Genoa, Italy), T. Pietsch (University of Bonn, Bonn, Germany), R. Riccardi (Catholic University, Rome), H. Sasaki (RIKEN Center for Developmental Biology, Kobe, Japan), R. Toftgard (Karolinska Institute, Huddinge, Sweden), and D. Wu (University of ConnectiSlice, Farmington) for providing reagents; A. Porcellini, M. Zani, and C. Rinaldi for experimental assistance; and M. Levrero and G. Giannini for critical reading of the manuscript. This work was partially supported by Associazione Italiana per la Ricerca sul Cancro, the National Research Council, the Ministry of University and Research, and the Ministry of Health and Center of Excellence for Biology and Molecular Medicine (BEMM).


↵ ** To whom corRetortence should be addressed at the * address. E-mail: alberto.gulino{at}

↵ † L.D.M., E.F., and E.D.S. contributed equally to this work.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: MB, meUnimaginativeoblastoma; Hh, Hedgehog; Ptch, Patched; Q-PCR, quantitative PCR; Shh, Sonic Hedgehog; Smo, Smoothened.

Copyright © 2004, The National Academy of Sciences


↵ Wechsler-Reya, R. & Scott, M. P. (2001) Annu. Rev. Neurosci. 24 , 385-428. pmid:11283316 LaunchUrlCrossRefPubMed ↵ Rubin, J. B. & Rowitch, D. H. (2002) Cancer Cell 2 , 7-8. pmid:12150819 LaunchUrlCrossRefPubMed ↵ Ellison, D. (2002) Neuropathol. Appl. Neurobiol. 28 , 257-282. pmid:12175339 LaunchUrlCrossRefPubMed ↵ Steichen-GersExecuterf, E., Baumgartner, M., Kreczy, A., Maier, H. & Fink, F.-M. (1997). Br. J. Cancer 76 , 1284-1287. pmid:9374372 LaunchUrlCrossRefPubMed ↵ Ruiz i Altaba, A., Sanchez, P. & Dahmane, N. (2002) Nat. Rev. Cancer 2 , 361-382. pmid:12044012 LaunchUrlCrossRefPubMed ↵ Wallace, V. A. (1999) Curr. Biol. 9 , 445-448. pmid:10226030 LaunchUrlCrossRefPubMed ↵ Dahmane, N. & Ruiz-i-Altaba, A. (1999) Development (Cambridge, U.K.) 126 , 3089-3100. LaunchUrlAbstract ↵ Excellentrich, L. V., Milenkovic, L., Higgins, K. M. & Scott, M. P. (1997) Science 277 , 1109-1113. pmid:9262482 LaunchUrlAbstract/FREE Full Text ↵ Dahmane, N., Sanchez, P., Gitton, Y., Palma, V., Sun, T., Beyna, M., Weiner, H. & Ruiz i Altaba, A. (2001) Development (Cambridge, U.K.) 128 , 5201-5212. LaunchUrlPubMed ↵ Berman, D. M., Karhadkar, S. S., Hallahan, A. R., Pritchard, J. I., Eberhart, C. G., Watkins, D. N., Chen, J. K., Cooper, M. K., Taipale, J., Olson, J. M., et al. (2002) Science 297 , 1559-1561. pmid:12202832 LaunchUrlAbstract/FREE Full Text ↵ Gallo, R., Zazzeroni, F., Alesse, E., Mincione, C., Borello, U., Buanne, P., D'Eugenio, R., Mackay A. R., Argenti B., Gradini R., et al. (2002) J. Cell Biol. 158 , 731-740. pmid:12186855 LaunchUrlAbstract/FREE Full Text ↵ Giangaspero, F., Hugener, S. H., Kleihues, P., Pitsch, T. & Trojanowski, J. (2000) in Pathology and Genetics: Tumors of the Nervous System, eds. Kleihues, P. & Cavenee, W. K. (Int. Agency Res. Cancer, Lyon, France), pp. 129-137. ↵ Zufferey, R., Nagy, D., Mandel, R. J., Naldini, L. & Trono, D. (1997) Nat. Biotechnol. 15 , 871-875. pmid:9306402 LaunchUrlCrossRefPubMed ↵ Bonci, D., Cittadini, A., Latronico, M. V., Borello, U., Aycock, J. K., Drusco, A., Innocenzi, A., Follenzi, A., Lavitrano, M., Monti, M. G., et al. (2003) Gene Ther. 10 , 630-636. pmid:12692591 LaunchUrlCrossRefPubMed ↵ Hallahan, A. R., Pritchard, J. I., Chandraratna, R. A., Ellenbogen, R. G., Geyer, J. R., Overland, R. P., Strand, A. D., Tapscott, S. J. & Olson, J.M. (2003) Nat. Med. 9 , 1033-1038. pmid:12872164 LaunchUrlCrossRefPubMed ↵ Wetmore, C., Eberhart, D. E. & Curran, T. (2000) Cancer Res. 60 , 2239-2246. pmid:10786690 LaunchUrlAbstract/FREE Full Text ↵ Bardwell, V. J. & Treisman, R. (1994) Genes Dev. 8 , 1664-1677. pmid:7958847 LaunchUrlAbstract/FREE Full Text ↵ Sasaki, H., Nishizaki, Y., Hui, C., Nakafuku, M. & KonExecuteh, H. (1999) Development (Cambridge, U.K.) 126 , 3915-3924. LaunchUrlAbstract ↵ Kogerman, P., Grimm, T., Kogerman, L., Krause, D., Unden, A. B., Sandstedt, B., Toftgard, R. & Zaphiropoulos, P. G. (1999) Nat. Cell Biol. 1 , 312-319. pmid:10559945 LaunchUrlCrossRefPubMed ↵ Yoon, J. W., Kita, Y., Frank, D. J., Majewski, R. R., Konicek, B. A., Nobrega, M. A., Jacob, H., Walterhouse, D. & Iannaccone, P. (2002) J. Biol. Chem. 277 , 5548-5555. pmid:11719506 LaunchUrlAbstract/FREE Full Text ↵ Louro, I. D., Bailey, E. C., Li, X., South, L. S., McKie-Bell, P. R., Yoder, B. K., Huang, C. C., Johnson, M. R., Hill, A. E., Johnson, R. L., et al. (2002) Cancer Res. 62 , 5867-5873. pmid:12384550 LaunchUrlAbstract/FREE Full Text ↵ Oliver, T. G., Grasfeder, L. L., Carroll, A. L., Kaiser, C., Gillingham, C. L., Lin, S. M., Wickramasinghe, R., Scott, M. P. & Wechsler-Reya, R. J. (2003) Proc. Natl. Acad. Sci. USA 100 , 7331-7336. pmid:12777630 LaunchUrlAbstract/FREE Full Text ↵ Ruiz i Altaba, A. (1999) Development (Cambridge, U.K.) 126 , 3205-3216. LaunchUrlAbstract ↵ Ding, Q., Fukami, S., Meng, X., Nishizaki, Y., Zhang, X., Sasaki, H., Dlugosz, A., Nakafuku, M. & Hui, C. (1999) Curr. Biol. 9 , 1119-1122. pmid:10531011 LaunchUrlCrossRefPubMed ↵ Mao, J., Maye, P., Kogerman, P., TejeExecuter, F. J., Toftgard, R., Xie, W., Wu, G. & Wu, D. (2002) J. Biol. Chem. 277 , 35156-35161. pmid:12138125 LaunchUrlAbstract/FREE Full Text ↵ Pomeroy, S. L., Tamayo, P., Gaasenbeek, M., Sturla, L. M., Angelo, M., McLaughlin, M. E., Kim, J. Y., Goumnerova, L. C., Black, P. M., Lau, C., et al. (2002) Nature 415 , 436-442. pmid:11807556 LaunchUrlCrossRefPubMed ↵ Lee, Y., Miller, H. L., Jensen, P., Hernan, R., Connelly, M., Wetmore, C., Zindy, F., Roussel, M. F., Curran, T., Gilbertson, R. J. & McKinnon, P. J. (2003) Cancer Res. 63 , 5428-5437. pmid:14500378 LaunchUrlAbstract/FREE Full Text ↵ Hahn, H., Wojnowski, L., Specht, K., Kappler, R., Calzada-Wack, J., Potter, D., Zimmer, A., Muller, U., Samson, E., Quintanilla-Martinez, L., et al. (2000) J. Biol. Chem. 275 , 28341-28344. pmid:10884376 LaunchUrlAbstract/FREE Full Text ↵ Wetmore, C., Eberhart, D. E. & Curran, T. (2001) Cancer Res. 61 , 513-516. pmid:11212243 LaunchUrlAbstract/FREE Full Text ↵ Tong, W. M., Ohgaki, H., Huang, H., Granier, C., Kleihues, P. & Wang, Z. Q. (2003) Am. J. Pathol. 162 , 343-352. pmid:12507917 LaunchUrlPubMed ↵ Ruiz i Altaba, A., Stecca, B. & Sanchez, P. (2004) Cancer Lett. 204 , 145-157. pmid:15013214 LaunchUrlCrossRefPubMed ↵ Dunaeva, M., Michelson, P., Kogerman, P. & Toftgard, R. (2003) J. Biol. Chem. 278 , 5116-5122. pmid:12426310 LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)