dedd interacts with pi3kc3 to activate autophagy and ... · silencing dedd expression in...

Molecular and Cellular Pathobiology

DEDD Interacts with PI3KC3 to Activate Autophagy andAttenuate Epithelial–Mesenchymal Transition in HumanBreast Cancer

Qi Lv, Wei Wang, Jianfei Xue, Fang Hua, Rong Mu, Heng Lin, Jun Yan, Xiaoxi Lv,Xiaoguang Chen, and Zhuo-Wei Hu

AbstractEpithelial-to-mesenchymal transition (EMT), a crucial developmental program, contributes to cancer invasion

and metastasis. In this study, we show that death-effector domain-containing DNA-binding protein (DEDD)attenuates EMTand acts as an endogenous suppressor of tumor growth andmetastasis.We found that expressionlevels of DEDD were conversely correlated with poor prognosis in patients with breast and colon cancer. Bothin vitro and in vivo, overexpression of DEDD attenuated the invasive phenotype of highlymetastatic cells, whereassilencing of DEDD promoted the invasion of nonmetastatic cells. Via direct interaction with the class III PI-3-kinase (PI3KC3)/Beclin1, DEDD activated autophagy and induced the degradation of Snail and Twist, twomasterregulators of EMT. The DEDD–PI3KC3 interaction led to stabilization of PI3KC3, which further contributed toautophagy and the degradation of Snail and Twist. Together, our findings highlight a novel mechanism in whichthe intracellular signaling protein DEDD functions as an endogenous tumor suppressor. DEDD expressionthereforemay represent a prognosticmarker and potential therapeutic target for the prevention and treatment ofcancer metastasis. Cancer Res; 72(13); 3238–50. �2012 AACR.

IntroductionMetastasis is the spread of cancer cells to other parts of the

body and metastatic tumor is primarily responsible for mostcancer deaths (1). Despite the mechanism of tumor metastasisremains most enigmatic in the field of cancer biology, newconcepts and mechanistic insights into the process of tumormetastasis have emerged from recent advances (2, 3). Forinstance, the epithelial-to-mesenchymal transition (EMT), anessential cell-biological program during embryonic develop-ment, plays a critical role in the regulation of tumor progres-sion and metastasis (4–6). In contrast to the temporally andspatially precise regulation of EMT during development, EMTin tumor progression and metastasis may be droved by auton-omous oncogenic activation or inactivation of signaling mole-cules with or without additional stimulation (7). Moreover,recent studies show that the activation of EMT can promotenoncancer stem cells (CSC) to enter into a CSCs-like state: cell

obtaining self-renewal and tumor-initiating ability (8–10).Hence, once undergoing EMT, cancer cells acquiring invasiveproperties can detach from primary tumor tissue and enter thesurrounding stroma and circulation, creating a favorable tissuemicroenvironment for the cancer cell colonization in distanttissue (11, 12).

The death-effector domain-containingDNA-binding protein(DEDD) is a member of family of death effector domain-containing proteins that comprise 7 proteins including FADD,c-FLIP, Caspase-8/10, and DEDD1/2 (13). DEDD plays animportant role in CD95-mediated apoptosis in response tothe activation of TNF-a through activating caspase-3 in cyto-plasm or caspase-6 in nucleus as a scaffold protein (14). Recentstudies found that DEDD can suppress the activity of Cdk1/cyclin B1 complexes and keep S6K1 activity, suggesting thatDEDD participates in the regulation of cell cycle and inhibitscell mitosis (15–17). Although other members of DED-contain-ing proteins such as FADD have been implicated in a diversearray of diseases such asHuntington's disease, Type II diabetes,autoimmune disease, arthritis, and cancer (18), there is nodirect evidence to show whether DEDD involves in the regu-lation of tumor development, growth, invasion, andmetastasis.Using a yeast 2-hybrid screening system, we have recentlyidentified a number of molecules, including DEDD and TRB3,physically interacting with TGF-b1 signal molecule SMAD3 toattenuate or enhance TGF-b1/Smad3-mediated cancer cellinvasion or EMT (19, 20).

In this study, we therefore hypothesized that DEDD mightplay a role in the regulation of tumor metastasis throughregulation of EMT. We found that the expression levels of

Authors' Affiliation: Molecular Immunology and Cancer PharmacologyGroups, State Key Laboratory of Bioactive Substance and Function ofNatural Medicines, Institute of Materia Medica, Chinese Academy ofMedical Sciences & Peking Union Medical College, Beijing, PR China

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

Corresponding Author: Zhuo-Wei Hu, Institute of Materia Medica, Chi-nese Academy of Medical Sciences & Peking Union Medical College,Beijing 100050, PR China. Phone: 86-10-8316-5034; Fax: 86-10-8316-5034; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-11-3832

�2012 American Association for Cancer Research.

CancerResearch

Cancer Res; 72(13) July 1, 20123238

on January 14, 2020. © 2012 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst June 19, 2012; DOI: 10.1158/0008-5472.CAN-11-3832

http://cancerres.aacrjournals.org/

DEDD correlate conversely with the poor prognosis of breastand colon cancers. Having discovered the functional asso-ciation of DEDD expression with cancer progression andmetastasis, we study the potential mechanism responsiblefor the DEDD-mediated antimetastatic effect. Our studiessuggest that DEDD can function as an endogenous suppres-sor of tumor growth and metastasis through regulating thestability of critical autophagy modulators to maintain autop-hagy activation and promote the degradation of master EMTinducers.

Materials and MethodsCell cultureBreast cancer and other cancer cell lines were purchased

from the American Type Culture Collection (ATCC) and main-tained as described (21–24). 293ET cells were obtained fromthe Cell Culture Center of Peking Union Medical College(PUMC, China). All of the cell lines were authenticated by thePUMC using the AmpF/STR Identifiler Kit (Applied Biosys-tems) and the ATCC fingerprint database. The last cell authen-tication was tested at December 2011.

PlasmidsFlag-DEDD was a gift from Dr. Marcus E. Peter, University

of Chicago, USA (25). HA-FADD was a gift from Dr. Sug HyungLee, The Catholic University of Korea, South Korea (26). Theconstructs pcDNA6-HA-PI3KC3, pCMV6-Entry-Flag-Beclin1,pGEX-4T-1-DEDD, pCMV2-Flag-Snail/Twist, and pGL6-E-cad-herin-lucwere generated byPCRand confirmedby sequencing.

DEDD-silencing cellsThree shRNAs were designed on the basis of the DEDD

sequence (NM_032998) identified with siRNA Target Finder(Ambion). Stable cell lines were selected hygromycin. A singleclone was amplified by a dilution cloning technique.

DEDD-overexpressing cellsHumanDEDD cDNAwas cloned into the pCDNA 3.1 expres-

sion vector (Invitrogen) and transfected into MDA-MB-231cells. Stable cell lines were selected by G418. A single clone wasamplified by a dilution cloning technique.

Tissue microarray slides and immunohistochemistryHuman breast or colon cancer tissue microarray slides (US

Biomax) were detected as previously described (27).

Chemoinvasion assayAssays were conducted following previous report (28) as in

Supplementary Methods.

Soft agar assaySoft agar assay were carried out as previously described

(29).

Analysis of metastasisFemale BALB/C-nu mice were injected with MCF-7 cells or

MDA-MB-231 cells according to a previous study (30). Nine

weeks after injection, all mice were killed and the number ofsurfacemetastases per lungwas determined under a dissectingmicroscope.

Analysis of tumor growthFemale BALB/C-nu mice were injected with MCF-7 cells or

MDA-MB-231 cells into the 4th mammary fat pad from flank.Tumor growth wasmonitored externally using vernier calipersfor 48 to 52 days. More details are provided in SupplementaryMethods.

ResultsDEDD expression negatively correlates with poorprognosis of breast and colon cancers

To determine the potential role of DEDD in the regulation oftumor growth and metastasis, relationship between theexpression levels of DEDD and the metastatic grade of cancerswas determined using cancer tissue microarrays with knownclinical follow-up records, containing 60 cases ductal breastcarcinoma specimens (Fig. 1A, top) and 60 cases colon ade-nocarcinoma specimens (Fig. 1B, top) were analyzed. Tissueswere scored and classified (low or high DEDD staining) on thebasis of the intensity of DEDD labeling and percentage ofDEDD-positive tumor cells. Among these specimens, Kaplan–Meier survival analysis of ductal breast carcinoma specimensand colon adenocarcinoma specimens revealed a correlationbetween the lower DEDD expression and reduced overallsurvival times (P < 0.05; Fig. 1A and B, bottom). In addition,the expression of DEDD was examined in 6 breast epithelialcancer cell lines, including 3 nonmetastatic cancer cell linesand 3 metastatic cancer cell lines (Fig. 1C). The expression ofDEDD could be detected in metastatic and nonmetastaticcancer cell lines but the metastatic cancer cells expressedmuch lower level of DEDD than the nonmetastatic cancer cellsdid. The identical results were also observed in several cancercell lines derived from liver, colon, pancreas, and prostatecancers (Fig. 1D). These data suggest that the expression levelof DEDD negatively correlates with the metastatic phenotypesof tumor and may be used as a prognosis marker for thesecancers.

DEDD inhibits metastatic phenotypes of cancer cellsWe investigated whether the expression level of DEDD

regulated invasive phenotypes of breast cancer cells byexpressing short hairpin RNAs (shRNA) to knock downDEDD expression in nonmetastatic MCF-7 cells or bystably overexpressing DEDD in metastatic MDA-MB-231cells. The shRNAs from 3 different DEDD sequences(shRNA1, shRNA2, or shRNA3) expressed in MCF-7 cellsreduced DEDD expression by 50%, 90%, and 84%, respec-tively (Supplementary Fig. S1A), compared with the cellsexpressing a control shRNA whose sequence did notmatch any known human gene. Overexpressing DEDD inMDA-MB-231 by transfecting pCDNA3.1-DEDD signifi-cantly increased the expression of DEDD compared withthe cells transfected with pCDNA3.1-control vector (Sup-plementary Fig. S1B).

DEDD-Activated Autophagy against Metastasis

www.aacrjournals.org Cancer Res; 72(13) July 1, 2012 3239




Silencing DEDD expression in nonmetastatic MCF-7 cellsresulted in the acquisition of invasive ability as indicated byTranswell invasion assay (Fig. 2A). Consistently, depletion ofDEDD stimulated colony formation of these cells in soft agar,indicating that silencing DEDD increases anchorage-depen-dent growth (Fig. 2B). Moreover, knocking down of DEDD

increased the proliferation (Supplementary Fig. S2A) andinhibited the TNF-a–induced apoptosis in MCF-7 cells (Sup-plementary Fig. S2B). Major differences in cell morphology hadbeen observed between cells expressing control shRNA andcells expressing DEDD shRNAs grown onMatrigel. The formerexhibited a cobble-stone-like morphology, the border of cells

Figure 1. The expression levels of DEDD conversely correlate with the poor prognosis of breast and colon cancers. A and B, DEDD levels in breast (A)and colon (B) tumor tissues (top; scale bars, 40 mm) and Kaplan–Meier plots (bottom) of overall survival of 60 patients with ductal breast carcinomasand colon adenocarcinomas stratified by DEDD expression levels. Tissues were scored as low or high DEDD staining for all tumor cells. C and D,DEDD levels were detected by immunoblotting (IB) in the nonmetastatic or metastatic breast cancer cell lines (C) and other tissue-derived cancercell lines (D). Data are representative of 3 independent assays with identical results.

Lv et al.

Cancer Res; 72(13) July 1, 2012 Cancer Research3240




andMatrigel was sharp; although the latter exhibited a cobble-stone-like morphology too, the border of cells showed a spike-like structure, data suggesting that MCF-7 cells silencingDEDD acquire a metastatic phenotype (Fig. 2C, left). BecauseDEDD shRNA2 wasmore efficient to silence DEDD expression,it was chosen for the subsequent studies. In contrast, over-expressing DEDD significantly reduced the invasive capacity ofMDA-MB-231 cells as indicated in Transwell invasion assay(Fig. 2A), and prevented colony formation of these cells in soft

agar (Fig. 2B). Overexpressing DEDD also decreased prolifer-ation but promoted apoptosis of MDA-MB-231 cells (Supple-mentary Fig. S2A and S2B). Moreover, cells expressing emptyvector exhibited a spindle-like fibroblastic morphology butDEDD-overexpressing cells had a cobble-stone-like morphol-ogy as showed by 3-dimensional Matrigel culture (Fig. 2C,right). Interestingly, prolonged growth time or increasedcell number did not enhance the invasive capacity ofMCF-7-control shRNA cells or MDA-MB-231-DEDD cells

Figure 2. Ectopic expression ofDEDD inhibits but depletion of DEDDpromotes tumor cell migration,invasion, and the anchorage-independent growth. A, the invasiveproperties of DEDD shRNA MCF-7and pcDNA3.1-DEDD MDA-MB-231cells togetherwith their controlswereanalyzed by Transwell invasionassay (magnification, �200). Dataare presented as mean � SEM of 3independent assays. B,representative photographsindicating soft agar colony formationwere taken at 20 days after culture ofthe indicated cells (top) and summaryof 3 independent assays (bottom) arepresented as mean � SEM. C, Dataare phase-contrast micrographs ofMCF-7 (left, magnification, �400)and MDA-MB-231 cells (right,magnification, �100) cultured onMatrigel. ��, P < 0.01.






(Supplementary Fig. S2C and S2D), indicating that the regu-latory effect of DEDD expression level on the invasive capacityis not a simple growth-dependent phenomenon.

DEDD inhibits tumor metastasis and growthWe next evaluated the in vivo effects of DEDD-changing

expression on tumor metastasis and growth. Control shRNAMCF-7 and DEDD shRNA MCF-7 cells were injected into thelateral tail vein of 6-week-old athymic mice. Nine weeks afterinjection, injection of mice with DEDD shRNAMCF-7 cells butnot with control shRNA MCF-7 cells resulted in the formation

of large metastatic nodules in the lungs. In contrast, thenumber and volume of metastatic nodules in the lung of miceinjected with MDA-MB-231-DEDD cells were significantlyreduced as compared with that injected with MDA-MB-231-control vector cells (Fig. 3A and B). These results indicate thatexpression level of DEDD correlates inversely with the meta-static ability of breast cancer cells.

To determine the effect of DEDD expression level on tumorgrowth, the DEDD-silencing or DEDD-overexpressing breastcancer cells were implanted under the 4th mammary fat padsof athymic nude mice. Comparing with mice implanted with

Figure 3. Cancer cellsoverexpressing DEDD loss butcancer cells silencing DEDDacquires the capacity of tumorgrowth and metastasis. A and, B,MDA-MB-231 cells (A) or MCF-7cells (B) were i.v. injected into malenude mice. Representative lungsand hematoxylin and eosin stainingof lung sections were presented toshow metastases in lungs (scalebar, 100 mm). The metastaticnodules are indicated with yellowarrows. The number of metastaticnodules is presented as mean �SEM (n¼ 6). C, MDA-MB-231 cellswere injected into the fat pads ofnude mice. Photographs ofrepresentativemice and tumors areshown, along with H&E staining oftumor (T) and normal (N) breasttissue (left, scale bar, 100 mm).Each data point of the growth curveis mean � SEM of primary tumors(right). D,MCF-7 cellswere injectedinto the fat pads of nude mice.Photographs of representativetumors (left) and the growth curve(right) are shown. Data arepresented as mean � SEM(n ¼ 6); ��, P < 0.01.

Lv et al.





MDA-MB-231-control vector cells, which formed large tumorswithin 52 days (Fig. 3C), themice implantedwith theMDA-MB-231-DEDD cells showed a significant reduction in the tumorgrowth, even no tumors were observed in some mice. Also,silencing DEDD significantly promoted the tumor growth inmice implanted with MCF-7 cancer cells containing DEDDshRNA compared with mice implanted with MCF-7 cancercells containing control shRNA (Fig. 3D). These results indicatethat the expression level of DEDD correlates negatively withthe tumor growth of breast cancer.

DEDD inhibits EMT by downregulating the expression ofSnail and TwistEMT, a process by which tumor-associated epithelial cells

obtain mesenchymal features and result in a reduced cell–cellcontact and increased motility, has a critical role in metastasis

(10, 11). To assess the potential roles of DEDD in the regulationof EMT, we examined the expression of EMT-associated mar-kers in the presence or absence of DEDD expression. We foundthatDEDDshRNAMCF-7 cells showed spindle-likefibroblasticmorphology, one of the main characteristics of EMT (data noshown). The expression of epithelial marker E-cadherin wassignificantly reduced in these cells (Fig. 4A and B, left). Incontrast, the expression of mesenchymal markers Vimentinanda-SMAwas dramatically upregulated in these cells (Fig. 4AandB, left). However, overexpression ofDEDD inMDA-MB-231cells led to an increase in the expression of epithelial marker E-cadherin and a decrease in the expression of mesenchymalmarkers vimentin and a-SMA (Fig. 4A and B, right). To verifythe universal effects of DEDD on EMT, DEDD was knock-downed in a nonmetastatic colon cancer cell line SW480 byDEDD shRNA.We found that the expression of E-cadherin was

Figure 4. Overexpressing DEDDinhibits but silencing DEDDpromotes EMT. A, expression ofE-cadherin and vimentin weredetected by immunofluorescencestaining (scale bar, 35 mm). B, theexpression of epithelial ormesenchymal markers was detectedby Western blotting (IB) in MCF-7cells and MDA-MB-231 cells. C, left,silencing DEDD increased theexpression of Snail and Twist inMCF-7 cells; overexpression ofDEDD reduced the expression ofSnail and Twist inMDA-MB-231 cells(right). D, the mRNA levels of Snailand Twist were detected byquantitative PCR (qPCR). Data arepresented as mean � SEM of 3independent assays. E, luciferaseactivity of E-cadherin (cad) reportergene was detected in MCF-7 cellsand MDA-MB-231 cells. Data arepresented as the mean � SEM of 3independent assays; �, P < 0.05.






significantly reduced but the expression of Vimentin wasincreased in SW480 DEDD shRNA cells (Supplementary Fig.S3A). Taken together, our studies indicate that DEDD is anegative EMT modulator.

A number of factors, particularly transcriptional factorsSnail and Twist, play critical role in the regulation of EMT(31–34). We examined the expression of Snail and Twist in thepresence or absence of DEDD. The expression of Snail andTwist was very low in the control shRNAMCF-7 cells. SilencingDEDD in MCF-7 cells significantly upregulated the expressionof Snail and Twist (Fig. 4C). In contrast, overexpression ofDEDDsignificantly decreased the expression of Snail andTwistin MDA-MB-231 cells (Fig. 4C). However, DEDD–knockingdown or DEDD-overexpressing cancer cells did not changethe amounts of Snail and Twist mRNAs compared with theircontrols as indicated by quantitative PCR analysis (Fig. 4D).Because Snail and Twist negatively regulate the expression ofepithelial marker E-cadherin, the upregulated Snail and Twistexpression in DEDD-silencing cells resulted in a suppressivepromoter activity of E-cadherin as indicated by luciferasereporter assays (Fig. 4E). In contrast, the downregulatedexpression of Snail and Twist in DEDD-overexpressing cellsresulted in an enhanced promoter activity of E-cadherin incompared with the control vector cells (Fig. 4E).

DEDD induces the degradation of Snail and Twist byactivating the autophagy–lysosome degradation system

Because the expression level of DEDD did not significantlychange the expression of Snail/Twist mRNAs (Fig. 4D), weexamined the potential effect of DEDD-changing expression onthe stability of Snail and Twist. The pCMV2-Flag-snail orpCMV2-Flag-twsit expression vector was transiently trans-fected into the MCF-7-control shRNA or MCF-7-DEDD shRNAcells. pEGFP-N1 vector expressing EGFP protein was used ascontrol to normalize the experimental variation betweendifferent time points. We found that the half-life of Flag-Snailwas 5.7 hours in MCF-7-control shRNA cells but the degrada-tion of Flag-Snail was restrained and the half-life was signif-icantly prolonged to 10 hours in theMCF-7-DEDD shRNA cells(Fig. 5A). Similarly, depletion of DEDD inhibited the degrada-tion of Twist and the half-life of Twist was extended from 6.2hours in MCF-7-control shRNA cells to 9.7 hours in MCF-7-DEDD shRNA cells (Supplementary Fig. S4A).

Previous work indicated the degradation of Snail and Twistdepends on the ubiquitin-proteasome degradation pathway(35, 36), we thus examined if the inhibition of proteasomechanged theDEDD-mediated degradation of Snail/Twist. Flag-Snail was transfected into MCF-7-control shRNA or MCF-7-DEDD shRNA cells. The cells were then treated with proteinsynthesis inhibitor cycloheximide and proteasome inhibitorMG132 to block de novo synthesis and the ubiquitin-protea-some degradation of Snail. We found that the Flag-Snail'sdegradation in MCF-7-control shRNA cells was much rapidthan that in MCF-7-DEDD shRNA cells (Fig. 5B). The identicalresult was also observed in MCF-7 cells transfected with Flag-Twist (Supplementary Fig. S4B). However, treatment of cellswith proteosome inhibitor MG132 did not change Snail/Twistdegradation. These data indicate that the DEDD-mediated

degradation of Snail/Twist does not depend on the ubiqui-tin-proteasome system.

Autophagy is another intracellular degradation system forthe majority of proteins and some organelles (37). To verifywhether autophagy-lysosome system participated in the reg-ulation of DEDD-mediated degradation of Snail/Twist, GFP-LC3 and Flag-Snail or Flag-Twist were cotransfected intoMCF-7 cells, and the cellular localization of LC3 and Snail/Twist wasobserved by immunofluorescence. We found that LC3 wascolocalized with Snail/Twist in the presence or absence pro-tein synthesis inhibitor cycloheximide (Fig. 5C; SupplementaryFig. S4C). Furthermore, inhibiting autophagy by 3-MA couldsignificantly attenuate the degradation of Snail/Twist becausea similar degradation rate was observed in MCF-7-controlshRNA andMCF-7-DEDD shRNA cells (Fig. 5D; SupplementaryFig. S4D). Thus, activation of the autophagy-lysosome systemparticipates in the degradation of Snail/Twist, and this processis DEDD dependent.

DEDD activates autophagy modulating the expression ofPI3KC3

Because we had found that DEDD-mediated degradation ofSnail/Twist, we examined if altering the expression level ofDEDD directly activate autophagic signal molecules. We foundthat silencing or overexpressing DEDD did not change theexpression of autophagic signal molecules such as Akt, p-Akt,p-mTOR in MCF-7 or MDA-MB-231 cells (Fig. 6A). However,the expression of class III PI-3-kinase (PI3KC3) was suppressedin MCF-7-DEDD shRNA cells compared with MCF-7-controlshRNA cells. In contrast, the expression of PI3KC3 wasincreased in DEDD-overexpressing MDA-MB-231 cells com-pared with the cells with the empty vector. Also, Beclin1expression was upregulated in DEDD-overexpressing MDA-MB-231 cells (Fig. 6A).

LC3 and p62/SQSTM are 2 indicators reflecting the effectualautophagy flux (38). We found that alternating DEDD expres-sion did not change the expression level of LC3 I/II. However,silencing DEDD in MCF-7 cells significantly increased theexpression and aggregation of p62 but overexpressing DEDDin MDA-MB-231 cells reduced the expression and aggregationof p62 (Fig. 6A and B). These results suggest that the enhancedexpression of DEDD induces but the reduced expression ofDEDD attenuates autophagy activation.

PI3KC3 and Beclin1 are 2 critical components of the PI3KC3core complex, which plays a central role in the activation ofautophagy (39, 40). To examine the effect of PI3KC3 andBeclin1 on the DEDD-mediated EMT, different concentrationsof PI3KC3-siRNA pools or Beclin1-siRNA pools were trans-fected into MCF-7-control shRNA cells. The different amountsof pCDNA6-HA-PI3KC3 or pCMV6-Flag-Beclin1 expressingvector were also transfected into MCF-7-DEDD shRNA cells.We found that silencing of PI3KC3 or Beclin1 could increasethe expression of Snail and Twist, downregulate the expressionof E-cadherin, and inhibit the degradation of p62 in a concen-tration-dependent manner (Fig. 6C andD, left). In contrast, theoverexpression of PI3KC3 or Beclin1 could decrease theexpression of Snail/Twist, increase the expression of E-cad-herin, and the degradation of p62 in a concentration-

Lv et al.





Figure 5. Depletion of DEDD stabilizes Snail through attenuating the autophagy–lysosome degradation system. A and B, MCF-7 cells werecotransfected with pCMV2-Flag-Snail and pEGFP-N1 vector. Cells were treated with 20 mmol/L of cycloheximide (CHX; A) or CHX plus 10 mmol/L ofMG132 (B) for the indicated times. Flag-Snail and EGFP were detected by immunoblotting (IB) with indicated antibodies. C, LC3-GFP (green) andFlag-snail (red) were cotransfected into the MCF-7 cells containing control shRNA and incubated for 36 hours. The cells were treated with 20 mmol/L ofCHX for the indicated times. The cell nuclei were stained with DAPI (scale bar, 35 mm). The arrows indicate the colocalization of LC3 and Snail. D,cells were handled as described in A and treated with 20 mmol/L of CHX and 10 mmol/L of 3-methyladenine (3-MA) for the indicated times. Cell extractswere prepared and immunoblotting was carried out as described in A. DAPI, 40,6-diamidino-2-phenylindole.






dependent manner (Fig. 6C and D, right). These data indicatethat PI3KC3 and Beclin1 participate in the DEDD-mediatedregulation of EMT by promoting the autophagy-associateddegradation of Snail and Twist.

DEDD stabilizes PI3KC3 by interacting with PI3KC3/Beclin1

The function or activity of PI3KC3/Beclin1 complex canbe regulated by several signal molecules such as UVRAG,Bcl-2, and Rubicon (39, 41). We wondered if DEDD inter-

acted directly with the complex to regulate autophagy. Wefound that DEDD interacted physically with PI3KC3 andBeclin1, as showed by co-precipitated of Flag-DEDD withendogenous PI3KC3 or Beclin1 (Fig. 7A), which was con-firmed by glutathione S-transferase pull-down assay (Fig.7B), and by immunoprecipitation of endogenous DEDD withPI3KC3 or Beclin1 (Fig. 7C). To determine whether theinteraction of DEDD with PI3KC3 and Beclin1 have effecton the formation of core complex, the interaction of PI3KC3with Beclin1 was determined by co-immunoprecipitation in

Figure 6. DEDD regulatesautophagy through modulating theexpression of PI3KC3. A, theexpression of autophagy-associated proteins was detectedby immunoblotting with indicatedantibodies. B, the expressionof p62 was detected byimmunofluorescence staining. Thegreen signal indicates the stainingof corresponding protein and theblue signal represents the nuclearDNA staining by DAPI (scale bar,35 mm). C and D, left, MCF-7-control shRNA cells weretransfected with the indicatedconcentrations of PI3KC3-siRNA(C) or Beclin1-siRNA (D); right,MCF-7-DEDD shRNA cells weretransfected with the indicatedconcentrations of pCDNA6-HA-PI3KC3 (C) or pCMV2-Flag-Beclin1 (D). Protein lysates wereanalyzed by immunoblotting withthe indicated antibodies. DAPI,40,6-diamidino-2-phenylindole.

Lv et al.





MDA-MB-231 cells containing empty or DEDD-expressingvector. Overexpression of DEDD significantly enhanced theinteraction between PI3KC3 and Beclin1 (Fig. 7D). Wefurther found that silencing DEDD promoted the degrada-tion of PI3KC3 and the half-life of PI3KC3 was reduced from9.2 to 3.3 hours (Fig. 7E). A similar result was also observedin SW480 control shRNA and DEDD shRNA cells (Supple-mentary Fig. S3B). However, silencing DEDD did not signif-icantly change the half-life of Beclin1 (data not shown).Taken together, DEDD supports the stability of PI3KC3through interacting with PI3KC3/Beclin1.

DiscussionOur recent work indicates that DEDD attenuates TGF-b1/

Smad3 signaling through interacting with transcription factorSmad3 to suppress the TGF-b1/Smad3-mediaed invasion incancer cells (19). In this study, we have showed that the ex-pression level of DEDD in breast and colon cancers correlatesconversely with the poor prognosis of breast and colon cancers:the lower is DEDD expression in beast and colon cancers, thehigher is the cancer-induced death rate because the expressionlevel of DEDD decides the metastatic phenotypes of the cancercells, including migration, invasion, growth, and metastasis

Figure 7. DEDD stabilizes PI3KC3through a physical interaction withPI3KC3/Beclin1. A–C, DEDDinteracts with PI3KC3 and Beclin1was confirmed by semi-endogenousco-immunoprecipitation (A),glutathione S-transferase (GST) pull-down (B), and endogenous co-immunoprecipitation (C). D, whole-cell extracts of MDA-MB-231 cellswere immunoprecipitaed (IP) andblotted (IB) with indicated antibodies.E, MCF-7 cells were transfected withpCDNA6-HA-PI3KC3 and treatedwith 20 mmol/L of cycloheximide(CHX) for the indicated times. Cellextracts were prepared and detectedby Western blotting with indicatedantibodies. F, schematic diagramillustrates the role of DEDD in theregulation of metastatic phenotypes.HA, hemagglutinin; MW, molecularweight.






(1, 5). Indeed, in metastatic MAD-MB-231 cells losing DEDDexpression, ectopic expression of DEDD stabilizes and activatesthe PI3KC3 through physically interacting with this protein,leading to the autophagy-lysosome dependent degradation ofSnail and Twist, 2 master inducers of EMT, to attenuate theEMT process in these cancer cells. In contrast, in nonmeta-static MCF-7 cells normally expressing DEDD, silencing DEDDexpression promotes the degradation and inactivation of thePI3KC3, leading to an attenuation of autophagy activity and theautophagy-lysosome–dependent degradation pathway. Hence,accumulation and activation of Snail and Twist in these cancercells promote the activation of the EMT process. Indeed,restoring DEDD expression in metastatic MAD-MB-231 breastcells results in a loss of metastatic phenotype whereas knock-down of DEDD expression in the nonmetastatic MCF-7 breastcells results in the acquisition of the most central traits of theCSCs: invasive/metastatic and tumor-initiating ability (1–3, 7).Thus, our studies indicate thatDEDD is critically involved in thenegative regulation of cancer progression and metastasisthrough maintaining autophagy activity to suppress the EMT(summarized in Fig. 7F).

Metastasis is a complex andmultistep process, which has tobe divided into 2 phases, namely, physical translocation of atumor cell from the primary tumor to a distant tissue to seedand colonization of disseminated tumor cell in the tissue (4–6).Recent researches indicate that EMT is not only involved in the2-phase events in cancer to promote metastasis, but alsoparticipates in the regulation of tumor progression (42). Themolecular and cellular mechanisms for the regulation of EMThave been emerged from these studies and, particularly, manypositive inducers and signal pathways for the induction of theEMT have been identified (5, 42). For example, a mechanisticlinkage between the Snail/Twist-induced EMT and CSC-likebehaviors has been indicated for the understanding of prome-tastatic role of this EMT inducer because these cancer cellson the EMT process acquire CSC-like traits (7). Our currentstudies provide evidence to show that the alteration in theexpression levels of DEDD regulates the EMT and results inloss or acquisition of metastatic phenotypes, thereby deter-mining the process of tumor progression and metastasis.Interestingly, a very recent study byMori and colleagues reportthat DEDD deficiency results in a defect in uterine decidua-lization and infertile because of reducing Akt expression,indicating that DEDD is indispensable for the establishmentof an adequate uterine environment to support early pregnan-cy in mice (43). This finding is consistent and support to thefact that DEDD is involved in the regulation of embryosdevelopmental programs including EMT (19). However, itneeds to point out, previous work indicated that DEDD canenter nucleus to inhibit rDNA transcription (25), regulate cellcycle, and inhibit cell mitosis to reduce cell and body size viamodulation of rRNA synthesis (17), attenuate protein synthe-sis, and promote apoptosis (44, 45). These growth-impedingor apoptosis-promoting roles of DEDD may also contributeto the DEDD-induced antimetastatic role. Taken together,these studies support a notion that tumor metastatic abilitymay be an innate property shared by the bulk of cells presentearly in a developing tumor mass (1). Hence, further study is

required to decipher the precise role and mechanism forantimetastatic roles of DEDD.

Our studies indicate that the alterations in the expressionlevel of DEDD regulate the EMT through DEDD interactingwith the PI3KC3/Beclin1 complex, which plays central role inthe initiation and maturation of autophagosomes (39, 40). TheDEDD's effect on the stability of PI3KC3/Beclin1 complexseems to be achieved through DEDD interacting with PI3KC3and stabilizes this protein. Indeed, we find that DEDD caninteract with PI3KC3 or Beclin1, respectively. Ectopic expres-sion of DEDD results in an increase in the expression of PI3KC3and Beclin1 in MDA-MB-231 cancer cells, whereas knockdownof DEDD results in a decrease in the expression of PI3KC3 andBeclin1 in MCF-7 cancer cells. However, silencing DEDDexpression results in a marked decrease in the degradationhalf-life of PI3KC3 but not Beclin1 (Fig. 7), indicating that theDEDD–PI3KC3 interaction plays a crucial role in maintainingthe stability of the PI3KC3/Beclin1 complex. Thus, alterationsin the expression of PI3KC3 result in either activation orinactivation of autophagy-dependent degradation of Snail/Twist and other positive EMT markers. Despite we do notknow the precise mechanism how DEDD interacts withPI3KC3 to maintain the stability of this protein and why lossof DEDD expression promotes the degradation of PI3KC3,recent work indicate that DEDD induces a diversity of cellbiologic activity through interactingwith several keymoleculesto attenuate or support the stability of these proteins in thecells (19, 46). Thus, a key question to be addressed is: whatprogram is responsible for shutdown or turn-on of DEDDexpression under the physiologic or pathologic situations.

Although targeting the activity of E-cadherin repressorssuch as Snail/Twist to attenuate EMT may be promisingtherapeutic strategy against tumor metastasis and progres-sion, these EMT inducers are transcription factors and difficultto access or target (5). Our discovery, DEDD interacting withPI3KC3/Beclin1 to promote the autophagy-dependent degra-dation of Snail and Twist, may provide mechanistic insightsinto developing novel therapeutic strategy against tumormetastasis andprogression. Currently, awide range of selectivePI3K inhibitors has been tested in preclinical studies and somehave entered clinical trials in oncology. Unveiling the Vps34structure (47) and discovery of selective PI3KC3 inhibitor (48)may shed the new light on the perspective anticancer targets.However, because of the complexity of PI3K signaling path-ways, developing an effective anticancer therapy may bedifficult. Particularly, the potential therapeutic effects of phar-macologic Vps34 modulation may well be context-dependent,and thus there could perhaps be a need for biomarkers forpatient selection. For this perception, our current finding mayprovide mechanistic insights into design of chemical tools andpotential drugs to activate or antagonize this key moleculeinvolved in vesicular trafficking and autophagy.

In summary, our current studies show that DEDD is anendogenous suppressor of tumor growth and metastasis inhuman breast cancer and the expression level of DEDDdecidesthe phenotypes of cancer cells in migration, invasion, growth,metastasis, and maybe resistance to the chemotherapybecause it negatively regulates the EMT. As tumor metastasis

Lv et al.





is a devastative disease, our findings suggest that the expres-sion level of DEDD can be detected and used as the prognosticmarker of cancers and therapeutic target for the preventionand treatment of cancer progression and metastasis. There-fore, additional insight into the regulatory role of DEDD–PI3KC3 interaction in autophagy and EMT may ultimatelyserve as an entry point to translate modifiers of these inter-actions into clinical endpoints.

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

Authors' ContributionsConception and design: Q. Lv, Z.-W. HuDevelopment of methodology: Q. Lv, Z.-W. HuAcquisition of data (provided animals, acquired and managedpatients, provided facilities, etc.): Q. Lv, W. Wang, J. Xue, F. Hua, R. Mu,H. Lin, J. Yan, X. Lv

Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Q. Lv, J. Xue, Z.-W. HuWriting, review, and/or revision of the manuscript: Q. Lv, F. Hua, X. Chen,Z.-W. HuAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): Q. LvStudy supervision: Z.-W. Hu

Grant SupportThis work was supported by grants from National Major Basic Research

Programof China (973: #2006CB503808), National Natural Science Foundation ofChina (81030056, 30973557, 81101595), and Creation of Major New Drugs(2009ZX09301-003-13, 2009ZX09301-003-9-1).

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicate thisfact.

Received November 28, 2011; revised March 16, 2012; accepted April 6, 2012;published OnlineFirst June 19, 2012.

References1. Chaffer CL, Weinberg RA. A perspective on cancer cell metastasis.

Science 2011;331:1559–64.2. Coghlin C, Murray GI. Current and emerging concepts in tumour

metastasis. J Pathol 2010;222:1–15.3. Klein CA. Cancer: the metastasis cascade. Science 2008;321:

1785–87.4. Ledford H. Cancer theory faces doubts. Nature 2011;472:273.5. Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial-mesenchy-

mal transitions in development and disease. Cell 2009;139:871–90.

6. Reiman JM,KnutsonKL, RadiskyDC. Immune promotion of epithelial–mesenchymal transition and generation of breast cancer stem cells.Cancer Res 2010;70:3005–8.

7. Scheel C, Eaton EN, Li SH, Chaffer CL, Reinhardt F, Kah KJ, et al.Paracrine and autocrine signals induce and maintain mesenchymaland stem cell states in the breast. Cell 2011;145:926–40.

8. Chaffer CL, Brueckmann I, Scheel C, Kaestli AJ, Wiggins PA, Rodri-gues LO, et al. Normal and neoplastic nonstem cells can spontane-ously convert to a stem-like state. Proc Natl Acad Sci U S A2011;108:7950–5.

9. Chaffer CL, Weinberg RA. Cancer cell of origin: spotlight on luminalprogenitors. Cell Stem Cell 2010;7:271–2.

10. Gupta PB, Chaffer CL, Weinberg RA. Cancer stem cells: mirage orreality? Nat Med 2009;15:1010–2.

11. Singh A, Settleman J. EMT, cancer stem cells and drug resistance:an emerging axis of evil in the war on cancer. Oncogene 2010;29:4741–51.

12. Steeg PS. Tumor metastasis: mechanistic insights and clinical chal-lenges. Nat Med 2006;12:895–904.

13. Tibbetts MD, Zheng L, LenardoMJ. The death effector domain proteinfamily: regulators of cellular homeostasis. Nat Immunol 2003;4:404–9.

14. Valmiki MG, Ramos JW. Death effector domain-containing proteins.Cell Mol Life Sci 2009;66:814–30.

15. Kurabe N, Arai S, Nishijima A, Kubota N, Suizu F, Mori M, et al. Thedeath effector domain-containing DEDD supports S6K1 activity viapreventing Cdk1-dependent inhibitory phosphorylation. J Biol Chem2009;284:5050–5.

16. Miyazaki T, Arai S. Two distinct controls of mitotic cdk1/cyclin B1activity requisite for cell growth prior to cell division. Cell Cycle2007;6:1419–25.

17. Arai S, Miyake K, Voit R, Nemoto S, Wakeland EK, Grummt I, et al.Death-effector domain-containing protein DEDD is an inhibitorof mitotic Cdk1/cyclin B1. Proc Natl Acad Sci U S A 2007;104:2289–94.

18. Barnhart BC, Lee JC, Alappat EC, PeterME. Thedeath effector domainprotein family. Oncogene 2003;22:8634–44.

19. Xue JF, Hua F, Lv Q, Lin H, Wang ZY, Yan J, et al. DEDD negativelyregulates transforming growth factor-beta1 signaling by interactingwith Smad3. FEBS Lett 2010;584:3028–34.

20. Hua F, Mu R, Liu J, Xue J, Wang Z, Lin H, et al. TRB3 interacts withSMAD3 promoting tumor cell migration and invasion. J Cell Sci2011;124:3235–46.

21. Neve RM, Chin K, Fridlyand J, Yeh J, Baehner FL, Fevr T, et al. Acollectionof breast cancer cell lines for the studyof functionally distinctcancer subtypes. Cancer Cell 2006;10:515–27.

22. Lee TK, Poon RT, Yuen AP, Ling MT, Kwok WK, Wang XH, et al. Twistoverexpression correlates with hepatocellular carcinoma metastasisthrough induction of epithelial-mesenchymal transition. Clin CancerRes 2006;12:5369–76.

23. Buck E, Eyzaguirre A, Barr S, Thompson S, Sennello R, Young D, et al.Loss of homotypic cell adhesion by epithelial-mesenchymal transitionor mutation limits sensitivity to epidermal growth factor receptorinhibition. Mol Cancer Ther 2007;6:532–41.

24. Beach S, TangH, Park S, Dhillon AS, Keller ET, KolchW, et al. Snail is arepressor of RKIP transcription in metastatic prostate cancer cells.Oncogene 2007;27:2243–48.

25. Stegh AH, Schickling O, Ehret A, Scaffidi C, Peterh€ansel C, HofmannTG, et al. DEDD, a novel death effector domain-containing protein,targeted to the nucleolus. EMBO J 1998;17:5974–86.

26. Shin MS, Kim HS, Lee SH, Lee JW, Song YH, Kim YS, et al. Alterationsof Fas-pathway genes associated with nodal metastasis in non-smallcell lung cancer. Oncogene 2002;21:4129–36.

27. Liu YY, Cai WF, Yang HZ, Cui B, Chen ZR, Liu HZ, et al. BacillusCalmette-Guvrin and TLR4 agonist prevent cardiovascular hypertro-phy and fibrosis by regulating immune microenvironment. J Immunol2008;180:7349–57.

28. Yang HZ, Cui B, Liu HZ, Mi S, Yan J, Yan HM, et al. Blocking TLR2activity attenuates pulmonarymetastases of tumor. PLoSOne 2009;4:e6520.

29. Han HJ, Russo J, Kohwi Y, Kohwi-Shigematsu T. SATB1 repro-grammes gene expression to promote breast tumour growth andmetastasis. Nature 2008;452:187–93.

30. Yan J,WangZY, YangHZ, LiuHZ,Mi S, LvXX, et al. Timing is critical foran effective anti-metastatic immunotherapy: the decisive role of IFNg/STAT1-mediated activation of autophagy. PLoS One 2011;6:e24705.

31. Foubert E, De Craene B, Berx G. Key signalling nodes in mammarygland development and cancer. The Snail1-Twist1 conspiracy inmalignant breast cancer progression. BreastCancerRes 2010;12:206.

32. Tran DD, Corsa CA, Biswas H, Aft R, Longmore GD. Temporal andspatial cooperation of Snail1 and Twist1 during epithelial-mesenchy-mal transition predicts for human breast cancer recurrence. MolCancer Res 2011;9:1644–57.






33. Peinado H, Olmeda D, Cano A. Snail, Zeb and bHLH factors in tumourprogression: an alliance against the epithelial phenotype? Nat RevCancer 2007;7:415–28.

34. Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, Come C,et al. Twist, a master regulator of morphogenesis, plays an essentialrole in tumor metastasis. Cell 2004;117:927–39.

35. Zhou BP, Deng J, Xia W, Xu J, Li YM, Gunduz M, et al. Dualregulation of Snail by GSK-3[beta]-mediated phosphorylation incontrol of epithelial-mesenchymal transition. Nat Cell Biol 2004;6:931–40.

36. Demontis S, Rigo C, Piccinin S, Mizzau M, Sonego M, Fabris M, et al.Twist is substrate for caspase cleavage and proteasome-mediateddegradation. Cell Death Differ 2005;13:335–45.

37. Yang Z, Klionsky DJ. An overview of the molecular mechanism ofautophagy. Curr Top Microbiol Immunol 2009;335:1–32.

38. Roy S, Debnath J. Autophagy and tumorigenesis. Semin Immuno-pathol 2010;32:383–96.

39. Funderburk SF,WangQJ, Yue Z. TheBeclin 1-VPS34 complex—at thecrossroads of autophagy and beyond. Trends Cell Biol 2010;20:355–62.

40. Backer JM. The regulation and function of class III PI3Ks: novel rolesfor Vps34. Biochem J 2008;410:1–17.

41. Zhong Y, Wang QJ, Li X, Yan Y, Backer JM, Chait BT, et al. Distinctregulation of autophagic activity by Atg14L and Rubicon associated

with Beclin1-phosphatidylinositol-3-kinase complex. Nat Cell Biol2009;11:468–76.

42. Iwatsuki M, Mimori K, Yokobori T, Ishi H, Beppu T, Nakamori S, et al.Epithelial-mesenchymal transition in cancer development and its clin-ical significance. Cancer Sci 2010;101:293–9.

43. MoriM, KitazumeM,OseR, Kurokawa J, KogaK,OsugaY, et al. Deatheffector domain-containing protein (DEDD) is required for uterinedecidualization during early pregnancy in mice. J Clin Invest 2011;121:318–27.

44. Lee JC,WangGX, Schickling O, PeterME. Fusing DEDDwith ubiquitinchanges its intracellular localization and apoptotic potential. Apopto-sis 2005;10:1483–95.

45. Alcivar A, Hu S, Tang J, Yang X. DEDD and DEDD2 associate withcaspase-8/10 and signal cell death. Oncogene 2003;22:291–7.

46. KurabeN,MoriM,KurokawaJ, Taniguchi K,AoyamaH,AtsudaK, et al.The death effector domain-containing DEDD forms a complex with Aktand Hsp90, and supports their stability. Biochem Biophys Res Com-mun 2010;391:1708–13.

47. Miller S, TavshanjianB,OleksyA,PerisicO,HousemanBT,ShokatKM,et al. Shaping development of autophagy inhibitors with the structureof the lipid kinase Vps34. Science 2010;327:1638–42.

48. Liu J, Xia H, KimM, Xu L, Li Y, Zhang L, et al. Beclin1 controls the levelsof p53by regulating the deubiquitination activity of USP10 andUSP13.Cell 2011;147:223–34.

Lv et al.





2012;72:3238-3250. Published OnlineFirst June 19, 2012.Cancer Res Qi Lv, Wei Wang, Jianfei Xue, et al.

Mesenchymal Transition in Human Breast Cancer−Epithelial DEDD Interacts with PI3KC3 to Activate Autophagy and Attenuate

Updated version

10.1158/0008-5472.CAN-11-3832doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2012/06/27/0008-5472.CAN-11-3832.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/72/13/3238.full#ref-list-1

This article cites 48 articles, 14 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/72/13/3238.full#related-urls

This article has been cited by 8 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/72/13/3238To 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-11-3832

http://cancerres.aacrjournals.org/content/suppl/2012/06/27/0008-5472.CAN-11-3832.DC1

http://cancerres.aacrjournals.org/content/72/13/3238.full#ref-list-1

http://cancerres.aacrjournals.org/content/72/13/3238.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/72/13/3238


dedd interacts with pi3kc3 to activate autophagy and ... · silencing dedd expression in...

Documents