The SWI/SNF Complex Protein Snr1 Is a Tumor Suppressor in ... · way to understand the mechanism how SMARCB1 suppresses MRTs during early childhood. Materials and Methods Drosophila

Molecular and Cellular Pathobiology

The SWI/SNF Complex Protein Snr1 Is a TumorSuppressor in Drosophila Imaginal TissuesGengqiang Xie1, Hanqing Chen2, Dongyu Jia1, Zhiqiang Shu1,William Hunt Palmer1,Yi-Chun Huang1, Xiankun Zeng3, Steven X. Hou3, Renjie Jiao2,4, and Wu-Min Deng1

Abstract

Components of the SWI/SNF chromatin-remodeling complexare among the most frequently mutated genes in various humancancers, yet only SMARCB1/hSNF5, a core member of the SWI/SNF complex, is mutated in malignant rhabdoid tumors (MRT).HowSMARCB1/hSNF5 functions differently fromothermembersof the SWI/SNF complex remains unclear. Here, we useDrosophilaimaginal epithelial tissues to demonstrate that Snr1, the con-served homolog of human SMARCB1/hSNF5, prevents tumori-genesis by maintaining normal endosomal trafficking-mediatedsignaling cascades. Removal of Snr1 resulted in neoplastic tumor-igenic overgrowth in imaginal epithelial tissues, whereas deple-

tion of any other members of the SWI/SNF complex did notinduce similar phenotypes. Unlike other components of the SWI/SNF complex that were detected only in the nucleus, Snr1 wasobserved in both the nucleus and the cytoplasm. Aberrant regu-lation of multiple signaling pathways, including Notch, JNK, andJAK/STAT, was responsible for tumor progression upon snr1-depletion. Our results suggest that the cytoplasmic Snr1 mayplay a tumor suppressive role in Drosophila imaginal tissues,offering a foundation for understanding the pivotal role ofSMARCB1/hSNF5 in suppressing MRT during early childhood.Cancer Res; 77(4); 862–73. �2017 AACR.

IntroductionThe mammalian SWI/SNF complex, also termed the Brahma

(Brm or Brg1) complex, regulates cellular processes such as celldifferentiation and cell cycling. Numerous studies in mammalshave shown that several subunits of this complex play a tumor-suppressor role in different tissues (1, 2). The Drosophila SWI/SNF complex regulates cell proliferation and differentiation in asimilar manner to its mammalian counterpart. Two recentstudies reported that the SWI/SNF complex also functions asa tumor suppressor in Drosophila neural stem cells (3, 4). Forexample, loss of components of the SWI/SNF complex resultedin aberrant dedifferentiation and proliferation of Drosophilaneuroblasts.

Although several components of the SWI/SNF complex, suchas SMARCA4/Brg1 (Brm in Drosophila), SMARCC1/BAF155(Mor in Drosophila), SMARCB1 (also known as hSNF5/INI1),and ARID1 (Osa in Droosphila), have been found mutated indifferent human cancers, SMARCB1 is distinct from others.Cancers with mutations of other components of the SWI/SNF

complex always have a number of more mutations in therecurrent samples, whereas cancers with loss of SMARCB1essentially have no additional mutations (5). Mutation ofSMARCB1 was first found in malignant rhabdoid tumors(MRT) that are very aggressive and highly lethal pediatrictumors (6, 7). No mutations in other subunits of the SWI/SNFcomplex have been found to be related to the MRTs, suggestinga potentially unique role of SMARCB1 in tumorigenesis. Com-ponents of the SWI/SNF complex in Drosophila, such as Brm,Mor and Osa, whereas not Snr1 (homolog of humanSMARCB1), have been reported to be essential for the repres-sion of Wingless target genes (8). Another study by Zraly andcolleagues (9) has indicated that Snr1 is dispensable in sometissues where Brm complex activities are necessary. These stud-ies suggest that Snr1 functions differently from the othercomponents of the Brm complex in certain situations.

Drosophila has been used as a model for studying the mechan-isms of tumorigenesis and for screening for antitumor drugs (10,11). In addition to the conservation of genes between Drosophilamelanogaster and humans, Drosophila tumors share many similar-ities to human tumors (12). Therefore, knowledge gained fromstudies of Drosophila tumors would help to understand humancounterparts.

In this study, we report that Snr1, which is distinct from othercomponents of the SWI/SNF complex, plays a unique role insuppressing tumor growth in the imaginal epithelial tissues.Loss of snr1 leads to neoplastic tumorigenic overgrowth, whichis characterized by the disruption of cell polarity, failure ofdifferentiation, and upregulation of invasion markers. Furtherexperiments demonstrate that multiple signaling pathways,including Notch, JNK, and JAK/STAT, are deregulated in snr1loss-of-function-induced tumorigenic discs. Deregulation ofthe pathways is caused by defects of the endosomal traffickingpathway, which may be attributed to the cytoplasmic functionof Snr1 that is different from that of other components of the

1Department of Biological Science, Florida State University, Tallahassee, Florida.2StateKey LaboratoryofBrain andCognitive Science, InstituteofBiophysics, theChinese Academy of Sciences, Beijing, China. 3The Basic Research Laboratory,National Cancer Institute, National Institutes of Health, Frederick, Maryland.4Sino-French Hoffmann Institute, Guangzhou Medical University, Guangzhou,China.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Authors: Wu-Min Deng, Florida State University, 319 StadiumDrive, King Life Science Building, Tallahassee, FL 32306-4295. Phone: 850-645-1501; Fax: 850-645-8774; E-mail: [email protected]; and Renjie Jiao, ChineseAcademy of Sciences, Beijing 100101, China. E-mail, [email protected]

doi: 10.1158/0008-5472.CAN-16-0963

�2016 American Association for Cancer Research.

CancerResearch

Cancer Res; 77(4) February 15, 2017862

on August 19, 2020. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst December 6, 2016; DOI: 10.1158/0008-5472.CAN-16-0963

http://cancerres.aacrjournals.org/

SWI/SNF complex. Our results have provided an alternativeway to understand the mechanism how SMARCB1 suppressesMRTs during early childhood.

Materials and MethodsDrosophila strains and clonal analysis

Drosophila stocks used in this study were shown in the Supple-mentary Data. The snr119-21-mutant allele was generated by theCRISPR technique and procedures are essentially as previouslydescribed (13, 14). The target sequence is 50-GGGCAGGGGATT-GAGGTCCAGG-30, in which "AGG" (in bold) is the PAM site and"GGTCC" (in italic) is a cutting site for Sau96I, designed foridentification of the mutagenesis. The forward primer sequencefor producing the guiding RNA (gRNA) is: 50-TAATACGACTC-ACTATAGGGCAGGGGATTGAGGTCCGTTTTAGAGCTAGAAAT-AGC-30. The concentration of the gRNA used for injection was50 ng/mL. Two independent F0 lines with mutations in the targetsequence of snr1 were obtained.

GenerationofUAS-HA-Snr1FL,UAS-HA-Snr1DNES, andUAS-HA-Snr1DC transgenic flies

The cDNA clone of snr1 from DGRC was used as a template toamply Snr1FL, Snr1DNES (nuclear export sequence, amino acids248–260, was deleted), and Snr1DC (the 66 C-terminal aminoacids were deleted). Only the coding sequences were amplified,and subsequently sequenced and subcloned into the pUASpvector by the Kpn I and Xba I restriction sites. An HA-tag wasincluded in the N-terminal of each construct. After standard Pelement-mediated random germline transformation, severalindependent lines of transgenic flies of each genotype wereobtained.

Anti-Brm antibody generationPolyclonal rabbit antisera were raised against 6�His fusion

proteins containing amino acids 501–775 of the Brm protein. Toproduce the 6�His fusion protein, an 825 bp fragment of brmwasamplified by PCR using primers 50-GTAGAATTCGCTGAAC-GAAAGCGTCGCCA-30 and 50-GTAGCGGCCGCTCATTCCTT-GAGCGTACCATTAAC-30 (restriction sites underlined). Theamplified fragments were cloned into the EcoRI and NotI sitesof PET-28a (þ) (Novagen). 6�His-Brm fusion protein wasexpressed in E. coli BL21 (DE3) pLysS, purified onNi-NTA agarosecolumns (Qiagen) and used to immunize rabbits using standardprotocol.

Immunocytochemistry and image captureAntibody staining was performed as previously described (15).

Bromodeoxyuridine (BrdUrd) labeling was carried out as previ-ously described (16) and endocytosis assay in live wing disccells was performed by following the protocol kindly providedby Dr. David Bilder (University of California, Berkeley, Berke-ley, CA; ref. 17). Rhodamine-conjugated and Alexa Fluor 647-conjugated phalloidin (Molecular Probe, 1:100) were used tostain F-actin. Nuclei were co-stained with DAPI (1:1,000, Invi-trogen) for 10 minutes at room temperature.

Images were captured on a Zeiss LSM-510 confocal micro-scope in the Biological Science Imaging Resource Facility atFlorida State University and Ziess LSM-800 confocal micro-scope in our laboratory. Figures were processed and arranged inAdobe Illustrator.

ResultsSnr1 functions as tumor suppressor in Drosophilaimaginal discs

Mutations in SMARCB1 are responsible for human MRT for-mation in children (6, 7), so to determine whether Snr1, theconservedDrosophila homolog of SMARCB1, has a similar tumor-suppressor role during development, we examined the functionofsnr1 in the imaginal disc, an epithelial tissue that has been widelyused for modeling cancer in Drosophila (18, 19). Animals withhomozygous snr1 mutation die at early larval stages (9). Tocircumvent early larval lethality, we used the Mosaic Analysiswith a Repressible Cell Marker (MARCM) technique (20) togenerate snr1mosaic clones with a null allele (snr1R3; ref. 9). Thearea of snr1R3 mutant clones generated by the MARCM system inthe wing imaginal disc was much smaller than that of the mockclones 72 hours after clonal induction (Supplementary Fig. S1Dand S1E). The mutant cells underwent apoptosis, as indicated bythe expression of an apoptotic marker, cleaved Drosophila Dcp-1,and were basally extruded (Supplementary Fig. S1B0 and S1E0).Mutant clones of another snr1 allele (snr119-21) showed similarphenotypes to those of snr1R3 (Supplementary Fig. S1C,C0 andS1F,F0). We further tested three independent RNAi lines targetingdifferent coding regions of snr1. The efficiency of these RNAi lineswas confirmed by immunostaining of Snr1 in the wing disc whereSnr1 was barely detected in snr1-RNAi cells (Fig. 1D-D0 0). Con-sistently, wing-imaginal-disc cells with snr1 knockdown inducedby the flip-out Gal4 further confirmed strong apoptosis and basalextrusion phenotypes (Fig. 1B-C0 0 0).

Imaginal disc cells in the wing pouch area bearing a singlemutation in tumor-suppressor genes, such as scribble (scrib), disclarge (dlg), or lethal giant larvae (lgl), have previously been reportedto be outcompeted by their wild-type neighbors (21–23). Pre-vention of cell competition by expression of oncogenic Ras(RasV12) in thesemutant cells resulted in neoplastic tumor growth(21, 24). We reasoned that the cell lethality phenotype of snr1mosaic discs could also be caused by cell competition, because inthe Minute background, the area of snr1 mutant clones wassignificantly bigger than that observed in wild-type background(not shown). The cell death phenotype caused by cell competitionwas generally more pronounced in the wing pouch area (25, 26).Although themajority of snr1 knockdown cells in the wing poucharea underwent apoptosis (yellow circle, Fig. 1C0), we noticed anovergrowth phenotype in the notum region of the wing disc(outlined by red circles in Fig. 1B and C, compared with thatin Fig. 1A). This overgrowth phenotype was detected in both snr1-depleted cells and their neighboring wild-type cells, as indicatedby BrdUrd and PH3 labeling (Fig. 1E–H). Similar to mutations ofept and vps25, removal of snr1 caused mainly nonautonomousovergrowth (arrows in Fig. 1F andH). Toprevent the cell death,wecoexpressed the antiapoptotic protein p35 with flip-out Gal4induced snr1 RNAi. Four days after RNAi induction, both wingand eye imaginal discs showed highly aggressive overgrowth; thesize of either the wing or eye-antennal disc was at least two timeslarger than that of the wild-type disc (Fig. 2F and G). When snr1-RNAi and p35 were ectopically expressed under the control of atissue-specific Gal4 driver, ptc-Gal4 or en-Gal4, some of theanimals stayed at larval stages for up to ten days, and their wingdiscs showed massive overgrowth as well as disrupted anterior–posterior compartment borders (Supplementary Fig. S2A–S2D).It is unlikely that this tumorigenic overgrowth is caused by

Different Tumor Suppressor Role of Snr1 from SWI/SNF Complex

www.aacrjournals.org Cancer Res; 77(4) February 15, 2017 863




compensatory proliferation induced by "undead" cells (27–29),because coexpression of proapoptotic gene reaper (rpr) and p35 ina wild-type background did not result in a similar overgrownphenotype (Supplementary Fig. S3B, compared with Supplemen-tary Fig. 3A). Together, these results suggest that Drosophila snr1plays a tumor-suppressor role in imaginal discs.

Depletion of snr1 leads to neoplastic tumorigenic overgrowthin Drosophila imaginal tissues

Drosophila epithelial tumors are normally either hyperplasticor neoplastic, with the latter showing defects in epithelialstructure, differentiation, and tissue architecture (18). To deter-mine the characteristics of snr1-LOF induced tumors, we firstexamined whether the apical-basal polarity was intact by stain-ing the mosaic tissue with an apical marker atypical proteinkinase C (aPKC), the basal lateral marker Dlg, and markers foradhesion junctions DE-Cadherin (DE-Cad) and Armadillo(Arm), as well as the actin cytoskeleton protein F-actin. Insnr1-depleted tumor cells, we found that columnar epitheliawere no longer maintained (Fig. 2A–B0). The overall levels ofthese markers in snr1-depleted cells were significantly increasedcompared with those in the wild-type neighbors, and subcel-lular localizations of the markers were disrupted (Fig. 2A–B0

and Supplementary Fig. S4D–S4F). Instead of its typicalapical localization in the wild-type cells, aPKC was localizedto all submembranes (Supplementary Fig. S4D), whereasArm and DE-Cad were not restricted to adhesion junctionsin snr1-knockdown cells as in the wild-type cells (Fig. 2A,A0,and Supplementary Fig. S4E). F-actin and Dlg, on the otherhand, were evenly distributed in all submembranes of snr1-depleted cells (Fig. 2B,B0, and Supplementary Fig. S4F). Fromsuch aberrant localizations, we can infer that the apical-basalpolarity is disrupted cell-autonomously in snr1-depleted epi-thelial cells.

Next, we examinedwhether cell differentiation is altered in snr1LOF cells. In the wild-type eye disc, differentiated photoreceptorcells normally had Elav expression posterior to the morphoge-netic furrow (MF) marked by accumulation of Ci (arrow in Fig.2C), whereas snr1 LOF cells did not express this photoreceptormarker (Fig. 2D), indicating that terminal cell differentiationfailed to occur in the absence of Snr1. This cell differentiationdefect in snr1-depleted cells is probably not caused by loss of MFmovement because the MF is anterior to snr1-depleted cells thatlack Elav expression (arrow in Fig. 2D). In addition, we detectedincreased cell proliferation in snr1 knockdown cells, as revealed byincreased BrdUrd incorporation, which labels proliferating cells

A

B

C

D H

A’

B’

C’

D’

B”

C”

D”

B’”

C’”

F

G

A” A”’ E

Figure 1.

Knockdown of snr1 causes cell death in the wing pouch area and overgrowth in the notum region. A-A0 0 0,Wild-type wing disc as control expressing flip-out clones.B–C0 0 0,Wing disc-bearing snr1-RNAi mosaic cells induced by the flip-outGal4 with apical view (B-B0 0 0) and basal view (C-C0 0 0); clonesweremarked by the expressionof GFP. D-D0 0, The Snr1 protein level is greatly reduced in snr1-depleted cells marked by the expression of GFP. E–H, Knockdown of snr1 causes bothautonomous and nonautonomous overgrowths. E and F, The flip-out control and snr1RNAi wing discs stained with PH3; the notum of control disc in E outlinedby dashed line has few PH3 labeling, and snrRNAi mosaic disc notum (F) has much more PH3 signals that are more prominent in neighboring wild-type area(arrow in F, compared with snr1RNAi area outlined by dashed line). G and H, The flip-out control and snr1RNAi wing discs labeled with BrdUrd; much moreBrdUrd signals were observed in the snr1RNAi clone (dashed line) and neighboring wild-type cells (arrow).

Xie et al.

Cancer Res; 77(4) February 15, 2017 Cancer Research864




(Supplementary Fig. S4A and S4B), and increased mitotic activitylabeled by mitotic marker phospho-histone H3 (PH3; Supple-mentary Fig. S4C).

Another feature of neoplastic fly tumors is upregulated expres-sionof the pro-invasion factorMatrixMetalloproteinase-1 (Mmp-1; ref. 30). This marker was strongly expressed in snr1-depletedcells (Fig. 2E,E0). This upregulated Mmp-1 phenotype was alsoobserved in wing discs carrying snr1-mutant clones in a lesscompetitive background wherein their neighboring twin spotswere eliminated due to mutation of STAT92E (STAT92E06346;Supplementary Fig. S5B0, compared with Supplementary Fig.S5A0). Cell polarity was also disrupted in wing disc cells of snr1mutant clones, as indicated by F-actin staining (arrowhead inSupplementary Fig. S5D0, compared with Supplementary Fig.S5C0). Together, the upregulation of proinvasion factors, and theloss of polarity demonstrate that snr1–LOF-induced tumors areneoplastic tumors in the imaginal tissues.

Disruption of other components of the SWI/SNFremodeling complex fails to cause tumorigenic overgrowthin imaginal tissues

Although not related to MRTs, several other members of theSWI/SNF complex have been found to be mutated in othertypes of cancer (2). To determine whether disruption of othermembers of the Drosophila SWI/SNF remodeling complexwould cause tumorigenesis in the imaginal disc, we knockeddown each component by use of the same flip-out Gal4 driver.Similar to the removal of snr1, cells of brm or osa knockdown inimaginal epithelial tissues failed to survive and showed appar-ent cell death phenotype with basal extrusion in the wingpouch region (Supplementary Fig. S6), implicating the SWI/SNF complex as required for cell survival. To block apoptosis,we introduced p35 into the knockdown cells. Unlike knock-down of snr1, removal of other components of the SWI/SNFcomplex appeared to cause no obvious overgrowth in theimaginal discs (Fig. 3B and C, compared with Fig. 3A); thosemosaic clones still kept their intact cell polarity and differen-tiated properly (not shown). These results suggest that thetumor-suppressor role Snr1 plays in Drosophila imaginal discsis probably independent from other components of the SWI/SNF remodeling complex.

The subcellular localization of Snr1 is different from that ofother components of the SWI/SNF complex

To understand why Snr1 has a unique role in suppressingtumor growth, we examined its subcellular localization in thewing disc cells. The SWI/SNF chromatin-remodeling complex isexpected to be localized in the nucleus, and indeed, Brm andOsa,two conserved members of this complex, were detected in thenucleus of thewingdisc cells (Fig. 3E andG). In contrast, althoughendogenous Snr1 wasmainly detected in the nucleus of disc cells,its expression was also detected in the cytoplasm of the cells (Fig.3E). In addition, we examined the subcellular localization ofSnr1, Brm and Osa in salivary gland cells. Consistent with theresults in disc cells, Brm and Osa were exclusively localized in thenucleus of salivary gland cells (Supplementary Fig. S7B and S7C),whereas Snr1 was detected in both the nucleus and the cytoplasm(Supplementary Fig. S7A). To further explore the cytoplasmiclocalization of Snr1, we generated HA-tagged Snr1 full-lengthprotein (HA-Snr1FL), and induced its expression in both the disccells and salivary gland cells. In concordance with endogenousSnr1 protein, HA-Snr1FL localized to both the nucleus and thecytoplasm in these two type of cells (Fig. 7B and SupplementaryFig. S7D), which suggests a potential cytoplasmic role of Snr1,

A

A’

B

B’C

D

E

E’

F G

Figure 2.

Loss of snr1 function in Drosophila imaginal disc induces neoplastic tumors. A,B, and E, Wing discs with coexpression of snr1-RNAi and p35 induced by theflip-out-Gal4 driver, stained with Arm (A), phalloidin to label F-actin (B), andMmp1 (E). A0 , B0, and E0, Transverse sections of the white lines in A, B, E. Cand D, Eye discs with expression of p35 (C) or snr1-RNAi and p35 (D) labeledby Elav and Ci, and the MF indicated by arrows. Clones were marked byexpression of GFP (A–E0). F and G, Wing (F) and eye-antennal discs (G) withp35 expression as control (left side of F and G) and snr1-RNAi and p35 (rightside of F and G) stained with DAPI.






besides its nuclear function in Drosophila. This cytoplasmic local-ization of Snr1 is consistent with the report that SMARCB1 canshuttle between nucleus and cytoplasm upon HIV-1 infection orunder normal growth conditions (31–33).

Snr1 is required for endosomal traffickingSMARCB1 has been suggested to regulate endocytosis by inter-

acting with Dynamin-2 (Shibire in Drosophila; ref. 31). Becausedisruption of components involved in endosomal trafficking hasbeen shown to cause tumorigenesis inDrosophila (17, 34, 35), we

sought to determine whether cytoplasmic localization of Snr1 isrelated to a role in endosomal trafficking of membrane proteins.To this end, we first examined the expression of two endosomaltrafficking components,Hepatocyte growth factor–regulated tyro-sine kinase substrate (Hrs) and the syntaxin Avalanche (Avl). Hrsis required for sorting of ubiquitinatedmembrane proteins to lateendosomes, whereas Avl is localized to early endosomes (36).Both Hrs and Avl were cell-autonomously enriched in snr1-depleted wing disc cells (Fig. 4A and B and Supplementary Fig.S8B and S8G), though not in brm-, osa-, or bap180-knockdown

A

D

E

F

G

B C

Figure 3.

Differences between Snr1 and other components ofDrosophila SWI/SNF complex. A–C, Wing discs withmosaic clonal expression of snr1-RNAi and p35 (A), brm-RNAi and p35 (B), or osa-RNAi and p35 (C) under the flip-out-Gal4 inducer. D, Snr1 protein was not detected inwing disc cells with snr1R3mutant clones (GFP negative),demonstrating the specificity of anti-Snr1 antibody. E–G,Wild-type wing disc cells with clonal expression ofhistone-RFP (hRFP, red) by flip-out Gal4 wereimmunolabeled with antibodies against Snr1 (E), Brm (F),or Osa (G); hRFP was used to label cell nuclei.

Xie et al.





A

B

C

DE

F

G

H

Figure 4.

Snr1 is required for endocytic trafficking signaling. A–C,Wing discs harboring snr1-depleted cells labeled byexpression of GFP were stainedwith Hrs (A), Avl (B), andNICD (C). D and F, Transverse sections of the whitelines in C and E. Wing disc with snr1R3 mutant clonesmarked by the absence of GFP was stained with Dl (E). Gand H, Endocytic assay with anti-Notch ECD (NECD)to label Notch at the apical surface of live wing discs.






cells (Supplementary Fig. S8C–S8E, S8H–S8J). The enrichment ofHrs and Avl was previously shown in mutant cells of Vps25, aneoplastic tumor suppressor, which is a component of the ESCRT(endosomal sorting complex required for transport) machinerythat sorts membrane proteins into multivesicular bodies duringendocytic trafficking to the lysosome (17, 36). Next, we examinedthe subcellular localization of transmembrane proteins: Delta(Dl), the ligand of the Notch signaling pathway, and its receptor,Notch, in snr1 mosaic imaginal wing discs. The distributions ofboth Dl and Notch proteins are highly dependent on membranetrafficking, and they were strongly upregulated in snr1-depletedcells as compared with neighboring wild-type cells (Fig. 4C–F).Again, accumulation of Dl or Notch was not found in brm- or osa-knockdown cells in the disc (Supplementary Fig. S9A, S9B, S9D,and S9E). The efficiency of brm or osa RNAi lines was confirmedby the undetectable or greatly reduced protein levels in theposterior compartment where brm or osa RNAi was induced(Supplementary Fig. S9C and S9F). We further performed anendocytosis assay with anti-Notch ECD to label Notch at theapical surface of live wing disc cells and confirmed the traffickingdefect in snr1-depleted tumor cells (Fig. 4G and H). We thereforespeculate that the cytoplasmic function of Snr1, which separatesSnr1 from other SWI/SNF complex components, contributes tomembrane trafficking.

Multiple signaling pathways are deregulated insnr1-depleted tumors

Endosomal trafficking is important for processing transmem-brane ligands and receptors of many signaling pathway cas-cades (37, 38); therefore, disruption of the trafficking pathwayusually leads to disregulation of signaling pathways (37, 39).Multiple signaling pathways, such as Notch, JAK/STAT and JNK,are found to be involved in tumorigenesis in Drosophila (17, 35,40–43). We have shown that Notch and Dl proteins accumu-lated in snr1 mutant or knockdown cells (Fig. 4C–F). In thesesnr1 LOF tissues, Notch signaling activity was strongly upregu-lated as monitored by its direct reporters, E(spl)-CD2 and E(spl)-m7-lacZ (Fig. 5E and Supplementary Fig. S10B). In con-trast, Notch signaling was not affected in the wing disc upondepletion of brm or osa (Supplementary Fig. S10C and S10D).The overgrowth phenotype caused by snr1 depletion was par-tially suppressed by expression of Notch RNAi (Fig. 5H, com-pared with 5F, and Fig. 5K). These results suggest that Notchsignaling is autonomously upregulated in snr1-LOF cells andaugments the tumor phenotype caused by snr1 depletion inDrosophila imaginal tissues.

The activation of JAK/STAT signaling has been reported to beessential for tumorigenesis in vps25, Tsg101, and Polycomb groupgenesmutant discs (17, 35, 41). In wild-type wing discs, Upd, theligand for the JAK/STAT pathway, was barely detected. However,in wing discs with snr1-RNAimosaic clones, Updwas upregulated(not shown). Induction of an upd-lacZ transgene was detected inmany snr1 LOF clones (Fig. 5A), suggesting increased Upd expres-sion is predominantly due to increased transcription. Further-more, JAK/STAT signaling activity, assessed by the 10xSTA-T92E>GFP (STAT-GFP) reporter (44),was robustly hyperactivatedin snr1-depleted tissues (Fig. 5C and Supplementary Fig. S11B),whereas STAT-GFP was expressed at low levels in wild-type wingdiscs (Fig. 5B and Supplementary Fig. S11A) orwing discs carryingbrm- or osa-RNAi mosaic clones (Supplementary Fig. S11C andS11D). Genetic interaction experiment suggested that increased

JAK-STAT activity contributes to tumorigenic overgrowth uponsnr1 loss (Fig. 5G, compared with 5F, and Fig. 5K).

The JNK signaling pathway is upregulated in many flytumors. We have shown that expression of Mmp1, which isinduced by JNK signaling, was upregulated in snr1–LOF-induced tumors (Fig. 2E,E0). To further analyze the activationof JNK signaling in snr1-LOF discs, we examined the expressionof two direct reporters, puc-lacZ and TRE-GFP (45). Consis-tently, the labeling of both reporters was significantly increasedcell-autonomously in snr1-depleted cells (Fig. 5D and Supple-mentary Fig. S12A), whereas not in brm- or osa-depleted cells(Supplementary Fig. S12B and S12C). To determine whetherJNK signaling was indeed involved in the tumorigenesis of snr1LOF mosaic discs, we reduced JNK activity by expressing eithera UAS-puc or UAS-bskDN construct. As expected, expression ofeither construct reduced Mmp1 levels in snr1 LOF clones (Fig.5I and J, compared with 5F); it partially decreased the tissue sizeof snr1-depleted mosaic discs (Fig. 5K), thus implying that JNKactivation was at least partially responsible for the tumorigenicphenotype caused by snr1 LOF.

RNA-seq analyses of snr1-RNAi–induced tumors confirmaberrant regulation of multiple signaling pathways

To understand the global change of gene expression in snr1-depletion–induced tumors, we carried out RNA-seqencing(RNA-seq) analyses to compare the expression profile betweencontrol and snr1-depleted tumorigenic wing imaginal discs.We focused on changes common to neoplasm by sequencingcDNA libraries generated from two independent snr1 knock-down tumors that phenocopy each other (Fig. 6A–C). Analysisrevealed 393 genes misregulated by at least two-fold increases/decreases in both mutant tissues (FDR <5%), with 336 upre-gulated and 57 downregulated (Fig. 6D and E). Differentiallyexpressed genes included several previously identified neoplas-tic effectors, such as the proinvasion factor Mmp-1 (30), whichhas been discussed previously, and the pupation regulatorinsulin-like peptide 8 (Ilp8; Fig. 6F; refs. 46, 47). The tran-scriptome dataset therefore accurately captures the expressionprofile of neoplastic tissues, and contains genes that promotetumorigenesis upon snr1 loss. However, depletion of mmp-1 orilp8 by RNAi was unable to suppress tumor growth caused bysnr1 knockdown (Supplementary Fig. S13), similar to the casein avl-RNAi tumors (47).

Consistent with the findings that multiple signaling path-ways, including Notch, JAK-STAT, and JNK were deregulated,RNA-seq analysis confirmed that expression of at least onetarget of each pathway was upregulated (Table 1, Supplemen-tary file S1). We previously showed that levels of Notch and Dlproteins were greatly upregulated in snr1-depleted cells; how-ever, we did not detect significant increases of their mRNAlevels in RNA-seq transcriptome analysis, which confirms thataccumulation of Notch and Dl is due to endocytic defects insnr1 LOF cells, irrelevant to the transcription regulation activityof the SWI/SNF chromatin-remodeling complex. In contrast,the mRNA level of the Notch signaling target E(spl)-m3 wasincreased nearly 2-folds in wing discs with snr1-depleted clones(Table 1). In addition, expression of multiple Notch targetssuch as chinmo, fruitless, abrupt, ken, and CG3835, as identifiedin Notch-induced hyperplastic wing discs by Djiane and hiscolleagues (48), was also significantly upregulated in snr1-depleted wing discs (Table 1, Supplementary File S1).

Xie et al.





Regarding JAK-STAT signaling, in accord with upd-lacZ upregu-lation (Fig. 5A), the level of upd-mRNA was greatly increased(7.41-folds, Table 1) in wing discs bearing snr1-depletion–induced tumors. In addition, transcriptional levels of the othertwo JAK-STAT pathway ligands, Upd2 and Upd3, were signif-icantly increased as well (Table 1). For the JNK signalingpathway, mRNA levels of its targets puc and mmp-1 increased2.9-fold and 8.5-fold, respectively (Table 1). Therefore, ourRNA-seq analyses of snr1 RNAi wing disc tumors providedvaluable and reliable information worth further exploration.

The cytoplasmic function of Snr1 contributes totumorigenesis in the disc

Because Snr1 localizes to both the cytoplasm and thenucleus, we wondered which disrupted functions of Snr1contributed to the tumorigenic phenotype in imaginal discs:nuclear, cytoplasmic, or both? A conserved nuclear exportsequence (NES) motif was found in the Drosophila Snr1 pro-tein, similar to its human counterpart, SMARCB1 (31–33). Wegenerated HA-tagged Snr1 protein with NES deletion (HA-Snr1DNES) under the UASp promoter (Fig. 7A). Immunostain-ing in both wing disc cells and salivary gland cells showed HA-Snr1DNES in the nucleus (Fig. 7C and Supplementary Fig. S7E),whereas the control HA-tagged wild-type Snr1 (HA-Snr1FL)was detected in both the nucleus and the cytoplasm in both

type of cells (Fig. 7B and Supplementary Fig. S7D) resemblingthat of endogenous Snr1 protein (Fig. 3E and SupplementaryFig. S7A).

As shown above, snr1-depleted wing imaginal discs pre-sented tumorigenic overgrowth (Fig. 2 and Supplementary Fig.S2A–S2D), we therefore tested whether expression of NES-deleted Snr1 could suppress the tumorigenic phenotype.Expectedly, expression of HA-Snr1FL was able to fully rescuethe tumorigenic phenotype in the snr1-RNAi þ p35 wing discs(Fig. 7F, compared with Fig. 7E, and Fig. 7I), whereas expres-sion of HA-Snr1DNES, of which the cytoplasmic function of Snr1is compromised, failed to suppress overgrowth (Fig. 7H and I).A truncated human SNF5 protein lacking the 66 C-terminalamino acids (hSNF5DC) is accumulated in the cytoplasm (32).This 66-amino-acid sequence is highly conserved betweenDrosophila Snr1 and human SNF5, we therefore generatedan HA-tagged Snr1 truncation mimicking this C-terminal dele-tion (HA-Snr1DC) under the UASp promoter. Similar tohSNF5DC, HA-Snr1DC is mainly localized in the cytoplasm ofwing disc cells and salivary gland cells (Fig. 7D and Supple-mentary Fig. S7F). This construct partially suppressed the snr1-depleted tumor phenotype (Fig. 7G, compared with Fig. 7E,and Fig. 7I). These data suggest that the cytoplasmic function ofSnr1 is required for its tumor-suppressing role in Drosophilaimaginal disc.

A F

G

I J

HB

C

D

E

K

Win

g d

isc

area

(mm

2 )

400,000

300,000

200,000

100,000

0

snr1

RNAi +p35

snr1

RNAi +p35

+snr1

FL

snr1

RNAi +p35

+STA

TRNAi

snr1

RNAi +p35

+Notc

hRNAi

snr1

RNAi +p35

+puc

snr1

RNAi +p35

+bsk

DN

GFPMmp1DAPI

Figure 5.

Notch, JAK-STAT, and JNK signalingare upregulated in snr1-RNAi–inducedwing disc tumor cells. A, D, and E,Close-up images of tumorous wingdiscs coexpressing snr1-RNAi and p35driven by the flip-out Gal4, each withtransgene upd-lacZ (A), puc-lacZ (D),or E(spl)CD2 (E). B, The flip-out Gal4-driven p35 control wing discexpressing transgene 10 X STAT92E-GFP (STAT-GFP). C, Tumorous wingdisc caused by depletion of snr1 andexpression of p35 under the control ofthe flip-out Gal4 labeled by theexpression of RFP expressed STAT-GFP; arrows indicate cell-autonomousupregulation of STAT-GFP in snr1-RNAi tumors and asterisks show non-autonomous upregulation of STAT-GFP in their neighboring wild-typecells. Note B and C had the sameimaging settings for STAT-GFP. F–K,Tumorigenic overgrowth of snr1-depleted wing disc can be partiallysuppressed by downregulation ofJAK-STAT (G), Notch (H), or JNK(I and J) signaling. Upregulatedexpression of JNK reporter Mmp1 insnr1-RNAi tumors (F) was onlysuppressed by downregulation ofJNK (I and J), but not bydownregulation of JAK-STAT (G),Notch (H). Statistical data of F–Jare shown in K.






DiscussionIn Drosophila neuroblasts, the SWI/SNF chromatin remodeling

complex functions alone, or cooperatively with another chroma-tin remodeling factor, HDAC3, to suppress dedifferentiation ofneural stem cells, suggesting its tumor suppressor role in neuralstem cells (3, 4). In this scenario, there is no difference amongcomponents of the SWI/SNF complex. In Drosophila imaginalepithelial tissues, however, only knockdown of snr1 resulted intumorigenic overgrowth, unlike depletion of other subunits of theSWI/SNF complex (Fig. 3A–C), indicating a unique and SWI/SNF-independent role of Snr1 in suppressing tumor progression inDrosophila imaginal epithelial tissues. Indeed, depletion of snr1caused endosomal trafficking defects wherein transmembraneproteins and components of the endosomal trafficking pathwayswere accumulated in snr1-RNAi tumors (Fig. 4), whereas RNAitargeting against other SWI/SNF components did not result insuch visible trafficking defects (Supplementary Figs. S8 and S9).

As a core component of the SWI/SNF chromatin-remodelingcomplex, only SMARCB1 is found mutated in childhood MRTs(5). An intrinsic distinction of SMARCB1 from other SWI/SNFsubunits is that it can shuttle between the nucleus and thecytoplasm via its nuclear export sequence (31–33). A recentstudy by Alfonso-Perez and his colleagues demonstratedthat SMARCB1 directly interacts with the GTPase Dynamin-2in the cytoplasm (31). Depletion of SMARCB1, but not othercomponents of the SWI/SNF complex, destabilizes Dynamin-2and impairs Dynamin-2-dependent endocytosis. Similar to

SMARCB1, we found Snr1 was localized in both the nucleusand the cytoplasm of wing disc cells and salivary gland cells(Fig. 3E and Supplementary Fig. S7A), whereas other DrosophilaSWI/SNF components, Brm and Osa, were only detected in thenucleus in these two type of cells (Fig. 3F and G and Supple-mentary Fig. S7B and S7C), which separates Snr1 from the SWI/SNF complex. Our novel findings that its cytoplasmic functionis required for its tumor-suppression function provide a mech-anistic explanation for snr1's involvement in cell proliferationregulation and early childhood cancers.

Numerous studies in Drosophila have demonstrated thatmutations of components of the trafficking signaling pathway,such as Avl, Vps25, and Tsg101, lead to tumorigenic over-growth. In these tumors, multiple signaling pathways are dis-regulated. The snr1-depletion-induced tumors also showdefects in Notch, JNK, and JAK-STAT signaling. Though in Avl,Vps25, and Tsg101 tumors, blocking Notch or JAK-STAT acti-vation successfully blocked tumor growth, disruption of anyindividual pathways in snr1-tumors only mildly alleviated thetumorous phenotype, suggesting a more complicated involve-ment of signaling pathways in these tumors. In the mammaliansystem, the involvement of SMARCB1 in endosomal traffickingis related to its Dynamin interaction. This interaction may havebeen conserved through evolution, as knockdown of shibire, theDrosophila homolog of dynamin, also resulted in tumorousgrowth in the imaginal discs (not shown). It is tempting tohypothesize that the cytoplasmic function of Snr1 is the maindriving force for its tumor-suppressor role.

10

8

6

4

2

0

–2

–4

Upregulated Downregulated

117 336 155 35 57 58

Mmp1 Ilp8

ImpE2

snr1TRiP

snr1IR22645

D E F

A B C

Figure 6.

Comparison of RNA-seq analyses offlip-out control wild-type wing discsand snr1-depleted tumor discs. A–C,Wing discs from wandering instarlarvae expressing only p35 (A), or withsnr1-RNAi: snr1TRiP in B and snr1IR12645

in C, stained with DAPI; clonal cellsmarked by expression of GFP.D and E,Overlap of genes upregulated (D) ordownregulated (E) with at least 2-foldchange in snr1TRiP and snr1IR12645

wing imaginal discs. F, Genespreviously implicated in neoplasticcharacteristics are differentiallyexpressed in snr1-depleted tumordiscs.

Xie et al.





The difference between childhood cancer and adult cancer maylie in themutation rates. In childhood tumors,mutant cells donothave enough time to accumulate significant numbers of muta-tions; therefore, genes responsible for such diseases must bemultifunctional. Snr1/SMARCB1 might be a perfect example inthis regard; its functional roles in both chromatin remodeling andendocytic trafficking fit into its function of growth regulation.

The tumor-suppressor role of Snr1 appears to be tissuespecific; in follicle cells of developing egg chambers, anothertype of epithelial cell, the removal of snr1 through mutation orRNAi did not cause tumorigenesis, even in the presence ofp35. In humans, SMARCB1-associated tumors are observed insoft tissues. Therefore, the involvement of this gene maydepend on its specific role in a tissue, or if there is a redun-

dancy in certain tissues where loss of snr1 might be compen-sated by a different gene. In Drosophila neural stem cells,several components of the SWI/SNF complex are required forpreventing tumorigenesis through proper temporal patterningand self-renewal control (4). These components include corecomponents of the SWI/SNF complex, Brm, Mor, and Snr1,and BAP-specific subunit Osa, though not any subunits of thePBAP complex, underlining the functional specificity of theSWI/SNF complex. Therefore, our study on Snr1's cytoplasmicrole in Drosophila imaginal discs provides insight into under-standing the occurrence and progression of human MRTs.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

A

I

B

C

D

E

G H

F400,000

300,000

200,000

100,000

0

Win

g d

isc

area

(mm

2 )

snr1

RNAi +p35p35

snr1

RNAi +p35

+snr1

FL

snr1

RNAi +p35

+snr1

DC

snr1

RNAi +p35

+snr1

DNES

NS

NS

HA-Snr1FL

HA-Snr1DC

HA-Snr1DNES

flip

-ou

t-G

al4>

snr1

RN

Ai ,U

AS

-p35

Figure 7.

Construction of Snr1-truncatedversions and their subcellularlocalization and rescue effects ontumorous wing discs caused by snr1depletion. A, HA-tagged differentforms of Snr1 under theUASppromoter.B–D, Close-up images of imaginal wingdiscs expressing HA-Snr1FL (B), HA-Snr1DNES (C), or HA-Snr1DC (D) werestained with anti-HA antibody; thetransgenes were induced by the flip-out Gal4 driver and cells expressing thetransgenesweremarked byhRFP (red).E–H, Tumorigenic overgrowthphenotype of snr1-RNAi wing disc (E) iscompletely rescued by the expressionof UASp-HA-Snr1FL (F), partially by theexpression of UASp-HA-Snr1DC (G), butnot by the expression of UASp-HA-Snr1DNES (H). I, Statistical data showingwing disc size difference in E–H. NS,nonsignificant.






Authors' ContributionsConception and design: G. Xie, W.-M. DengDevelopment of methodology: G. Xie, X. Zeng, S.X. Hou, R. JiaoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): G. Xie, H. Chen, D. Jia, Z. Shu, Y.-C. Huang, R. JiaoAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): G. Xie, W.H. Palmer, Y.-C. HuangWriting, review, and/or revision of themanuscript:G. Xie, Z. Shu,W.-M. DengAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): G. Xie, Z. Shu, R. JiaoStudy supervision: G. Xie, W.-M. Deng

AcknowledgmentsWe thank Trudi Schupbach, the Bloomington Drosophila Stock Center,

the Vienna Drosophila RNAi Center, the National Institute of Genetics,and the TRiP at Harvard Medical School for fly stocks; Andrew Ding-wall, Hugo J. Bellen, and the Developmental Studies Hybridoma Bankfor antibodies; David Bilder for the endocytosis assay protocol; LexiangYu for technical assistance; Thomas J. Fellers from the Biological

Science Imaging Resource of Florida State University (FSU) for technicalsupport; the Translational Science Laboratory of the FSU College ofMedicine for running the Illumina HiSeq machine; Jen Kennedy andGabriel Calvin for critical reading and helpful comments on the article;and members of the Deng and Jiao laboratories for feedback andsuggestions.

Grant SupportW.-M. Deng is supported by the NIH (grant number R01GM072562) and

the National Science Foundation (grant number IOS-1052333). R. Jiao issupported by grants from the National Natural Science Foundation of China(81470846, 31271573, and 31529004) and the Chinese Academy ofSciences (XDA04020413-02).

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received April 8, 2016; revised November 2, 2016; accepted November 18,2016; published OnlineFirst December 6, 2016.

References1. Roberts CW, Biegel JA. The role of SMARCB1/INI1 in development of

rhabdoid tumor. Cancer Biol Ther 2009;8:412–6.2. Reisman D, Glaros S, Thompson EA. The SWI/SNF complex and cancer.

Oncogene 2009;28:1653–68.3. KoeCT, Li S, Rossi F,Wong JJ,Wang Y, Zhang Z, et al. The Brm-HDAC3-Erm

repressor complex suppresses dedifferentiation in Drosophila type IIneuroblast lineages. Elife 2014;3:e01906.

4. Eroglu E, Burkard TR, Jiang Y, Saini N, Homem CC, Reichert H, et al. SWI/SNF complex prevents lineage reversion and induces temporal patterningin neural stem cells. Cell 2014;156:1259–73.

5. Lee RS, Stewart C, Carter SL, Ambrogio L, Cibulskis K, Sougnez C, et al. Aremarkably simple genome underlies highly malignant pediatric rhabdoidcancers. J Clin Invest 2012;122:2983–8.

6. Biegel JA, Zhou JY, Rorke LB, Stenstrom C, Wainwright LM, Fogelgren B.Germ-line and acquired mutations of INI1 in atypical teratoid and rhab-doid tumors. Cancer Res 1999;59:74–9.

7. Versteege I, Sevenet N, Lange J, Rousseau-Merck MF, Ambros P, Handgre-tinger R, et al. Truncatingmutations of hSNF5/INI1 in aggressive paediatriccancer. Nature 1998;394:203–6.

8. Collins RT, Treisman JE. Osa-containing Brahma chromatin remodelingcomplexes are required for the repression of wingless target genes. GenesDev 2000;14:3140–52.

9. Zraly CB, Marenda DR, Nanchal R, Cavalli G, Muchardt C, Dingwall AK.SNR1 is an essential subunit in a subset of Drosophila brm complexes,targeting specific functions during development. Dev Biol 2003;253:291–308.

10. Miles WO, Dyson NJ, Walker JA. Modeling tumor invasion and metastasisin Drosophila. Dis Model Mech 2011;4:753–61.

11. Gladstone M, Su TT. Chemical genetics and drug screening in Drosophilacancer models. J Genet Genomics 2011;38:497–504.

12. Gateff E. Malignant neoplasms of genetic origin in Drosophila melano-gaster. Science 1978;200:1448–59.

13. Yu Z, Chen H, Liu J, Zhang H, Yan Y, Zhu N, et al. Variousapplications of TALEN- and CRISPR/Cas9-mediated homologousrecombination to modify the Drosophila genome. Biol Open 2014;3:271–80.

14. Yu Z, Ren M, Wang Z, Zhang B, Rong YS, Jiao R, et al. Highly efficientgenome modifications mediated by CRISPR/Cas9 in Drosophila. Genetics2013;195:289–91.

15. Xie G, Yu Z, Jia D, Jiao R, Deng WM. E(y)1/TAF9 mediates thetranscriptional output of Notch signaling in Drosophila. J Cell Sci2014;127:3830–9.

16. Sun JJ, Deng WM. Notch-dependent downregulation of the home-odomain gene cut is required for the mitotic cycle/endocycle switchand cell differentiation in Drosophila follicle cells. Development2005;132:4299–308.

17. Vaccari T, Bilder D. The Drosophila tumor suppressor vps25 preventsnonautonomous overproliferation by regulating notch trafficking. DevCell 2005;9:687–98.

18. Hariharan IK, Bilder D. Regulation of imaginal disc growth by tumor-suppressor genes in Drosophila. Annu Rev Genet 2006;40:335–61.

19. Pastor-Pareja JC, Xu T. Dissecting social cell biology and tumors usingDrosophila genetics. Annu Rev Genet 2013;47:51–74.

20. Lee T, Luo L. Mosaic analysis with a repressible cell marker forstudies of gene function in neuronal morphogenesis. Neuron 1999;22:451–61.

21. Brumby AM, Richardson HE. scribble mutants cooperate with oncogenicRas or Notch to cause neoplastic overgrowth in Drosophila. EMBO J2003;22:5769–79.

22. Tamori Y, Bialucha CU, Tian AG, Kajita M, Huang YC, Norman M, et al.Involvement of Lgl and Mahjong/VprBP in cell competition. PLoS Biol2010;8:e1000422.

23. Igaki T, Pastor-Pareja JC, Aonuma H, Miura M, Xu T. Intrinsic tumorsuppression and epithelial maintenance by endocytic activation ofEiger/TNF signaling in Drosophila. Dev Cell 2009;16:458–65.

24. Pagliarini RA, Xu T. A genetic screen in Drosophila for metastatic behavior.Science 2003;302:1227–31.

25. Lolo FN, Casas-Tinto S, Moreno E. Cell competition time line: winners killlosers, which are extruded and engulfed by hemocytes. Cell Rep 2012;2:526–39.

26. de Beco S, Ziosi M, Johnston LA. New Frontiers in cell competition. DevDyn 2012;241:831–841.

27. Perez-Garijo A, Martin FA, Morata G. Caspase inhibition during apoptosiscauses abnormal signalling and developmental aberrations in Drosophila.Development 2004;131:5591–8.

28. RyooHD,Gorenc T, StellerH.Apoptotic cells can induce compensatory cellproliferation through the JNK and the Wingless signaling pathways. DevCell 2004;7:491–501.

29. Huh JR, Guo M, Hay BA. Compensatory proliferation induced bycell death in the Drosophila wing disc requires activity of the apicalcell death caspase Dronc in a nonapoptotic role. Curr Biol 2004;14:1262–6.

30. Uhlirova M, Bohmann D. JNK- and Fos-regulated Mmp1 expressioncooperates with Ras to induce invasive tumors in Drosophila. EMBO J2006;25:5294–304.

31. Alfonso-Perez T, Dominguez-Sanchez MS, Garcia-Dominguez M,Reyes JC. Cytoplasmic interaction of the tumour suppressor proteinhSNF5 with dynamin-2 controls endocytosis. Oncogene 2014;33:3064–74.

32. Craig E, Zhang ZK, Davies KP, Kalpana GV. A masked NES in INI1/hSNF5mediates hCRM1-dependent nuclear export: implications for tumorigen-esis. EMBO J 2002;21:31–42.

Xie et al.





33. Turelli P, Doucas V, Craig E, Mangeat B, Klages N, Evans R, et al. Cyto-plasmic recruitment of INI1 and PML on incoming HIV preintegrationcomplexes: interference with early steps of viral replication. Mol Cell2001;7:1245–54.

34. Thompson BJ, Mathieu J, Sung HH, Loeser E, Rorth P, Cohen SM. Tumorsuppressor properties of the ESCRT-II complex component Vps25 inDrosophila. Dev Cell 2005;9:711–20.

35. Moberg KH, Schelble S, Burdick SK, Hariharan IK. Mutations inerupted, the Drosophila ortholog of mammalian tumor susceptibilitygene 101, elicit non-cell-autonomous overgrowth. Dev Cell 2005;9:699–710.

36. Lu H, Bilder D. Endocytic control of epithelial polarity and proliferation inDrosophila. Nat Cell Biol 2005;7:1232–9.

37. Fischer JA, Eun SH, Doolan BT. Endocytosis, endosome trafficking, and theregulation of Drosophila development. Annu Rev Cell Dev Biol 2006;22:181–206.

38. Murphy JE, Padilla BE, Hasdemir B, Cottrell GS, Bunnett NW. Endosomes:a legitimate platform for the signaling train. Proc Natl Acad Sci U S A2009;106:17615–22.

39. Fortini ME, Bilder D. Endocytic regulation of Notch signaling. Curr OpinGenet Dev 2009;19:323–8.

40. Martinez AM, Schuettengruber B, Sakr S, Janic A, Gonzalez C, Cavalli G.Polyhomeotic has a tumor suppressor activity mediated by repression ofNotch signaling. Nat Genet 2009;41:1076–82.

41. Classen AK, Bunker BD, Harvey KF, Vaccari T, Bilder D. A tumor suppressoractivity of Drosophila Polycomb genes mediated by JAK-STAT signaling.Nat Genet 2009;41:1150–5.

42. Ohsawa S, Sato Y, Enomoto M, Nakamura M, Betsumiya A, IgakiT. Mitochondrial defect drives non-autonomous tumour progres-sion through Hippo signalling in Drosophila. Nature 2012;490:547–51.

43. Enomoto M, Igaki T. Src controls tumorigenesis via JNK-dependentregulation of the Hippo pathway in Drosophila. EMBO Rep 2012;14:65–72.

44. Bach EA, Ekas LA, Ayala-Camargo A, Flaherty MS, Lee H, Perrimon N, et al.GFP reporters detect the activation of the Drosophila JAK/STAT pathway invivo. Gene Expr Patterns 2007;7:323–31.

45. Chatterjee N, Bohmann D. A versatile PhiC31 based reporter system formeasuring AP-1 and Nrf2 signaling in Drosophila and in tissue culture.PLoS ONE 2012;7:e34063.

46. Garelli A, Gontijo AM, Miguela V, Caparros E, Dominguez M. Imaginaldiscs secrete insulin-like peptide 8 to mediate plasticity of growth andmaturation. Science 2012;336:579–82.

47. Colombani J, AndersenDS, Leopold P. Secreted peptide Dilp8 coordinatesDrosophila tissue growth with developmental timing. Science 2012;336:582–5.

48. Djiane A, Krejci A, Bernard F, Fexova S, Millen K, Bray SJ. Dissecting themechanisms of Notch induced hyperplasia. EMBO J 2013;32:60–71.






2017;77:862-873. Published OnlineFirst December 6, 2016.Cancer Res Gengqiang Xie, Hanqing Chen, Dongyu Jia, et al.

Imaginal TissuesDrosophilaThe SWI/SNF Complex Protein Snr1 Is a Tumor Suppressor in

Updated version

10.1158/0008-5472.CAN-16-0963doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2016/12/06/0008-5472.CAN-16-0963.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/77/4/862.full#ref-list-1

This article cites 48 articles, 17 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/77/4/862.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/77/4/862To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-16-0963

http://cancerres.aacrjournals.org/content/suppl/2016/12/06/0008-5472.CAN-16-0963.DC1

http://cancerres.aacrjournals.org/content/77/4/862.full#ref-list-1

http://cancerres.aacrjournals.org/content/77/4/862.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/77/4/862


The SWI/SNF Complex Protein Snr1 Is a Tumor Suppressor in ... · way to understand the mechanism how SMARCB1 suppresses MRTs during early childhood. Materials and Methods Drosophila

Documents