acute myelogenous leukemia–derived smad4 mutations target the protein to ubiquitin-proteasome...

HUMANMUTATION 27(9), 897^905,2006

RESEARCH ARTICLE

Acute Myelogenous Leukemia–Derived SMAD4Mutations Target the Protein to Ubiquitin-Proteasome Degradation

Lei Yang,1,2 Ning Wang,1 Yi Tang,2 Xu Cao,2 and Mei Wan2�

1School of Medicine, Shihezi University, Shihezi, Xinjiang, People’s Republic of China; 2Department of Pathology, School of Medicine,University of Alabama at Birmingham, Birmingham, Alabama

Communicated by Haig H. Kazazian

Disruption of transforming growth factor-b (TGFB1/TGF-b) signaling contributes to the formation of humanhematological malignancies. Smad4, a tumor suppressor, functions as an essential intracellular signal transducerof the TGF-b signaling pathway. Recent studies have demonstrated that some tumor-derived mutations ofSmad4 are associated with protein instability; however, the precise mechanism by which mutated Smad4proteins undergo rapid degradation remains to be elucidated. A missense mutation of the SMAD4 gene in theMad homology 1 (MH1) domain (c.305C4T, Pro102Leu) and one frameshift mutation resulting intermination in the Mad homology 2 (MH2) domain (c.1447_1448insAATA, D483–552) have been identifiedin acute myelogenous leukemia. It is not known whether protein instability of these SMAD4 mutants is one ofthe contributors to TGF-b signaling disruption in acute myelogenous leukemia. Here we report that these twoacute myelogenous leukemia–derived SMAD4 mutants are degraded rapidly when compared to their wild-typecounterpart. We have demonstrated that both mutated proteins exhibit enhanced polyubiquitination (orpolyubiquitylation) and proteasomal degradation. Importantly, we found that b-transducin-repeat-containingprotein 1 (b-TrCP1), an F-box protein in the ubiquitin E3 ligase Skp1-Cullin-F-box protein (SCF) complex,directly interacts with and acts as a critical determinant for degradation of both mutated SMAD4 proteins.In addition, small interference RNA (siRNA)-triggered endogenous b-TrCP1 suppression increased the proteinexpression level of both overexpressed SMAD4 mutants and endogenous mutated SMAD4 protein in acutemyelogenous leukemia cells. These data suggest that mutated SMAD4 proteins undergo rapid degradation inacute myelogenous leukemia cells via SCFb-TrCP1 E3 ligase-mediated protein ubiquitination (or ubiquitylation)and subsequent proteasomal degradation. Hum Mutat 27(9), 897–905, 2006. Published 2006 Wiley-Liss, Inc.y

KEY WORDS: SMAD4 gene mutation; acute myelogenous leukemia; protein instability; ubiquitin-proteasome pathway;TGFB1; TGF-b; b-TrCP1

INTRODUCTION

Transforming growth factor-b (HUGO gene symbol, TGFB1;also TGF-b; MIM] 190180) signaling plays a central role in theregulation of a broad range of cellular responses, including cellproliferation, recognition, differentiation, apoptosis, and specifica-tion of developmental fate [Blobe et al., 2000; Derynck andZhang, 2003]. Loss of TGF-b signaling has been implicated inmalignant transformation of various tissues [Massague, 1998; Yueand Mulder, 2001]. The Smad family proteins mediate TGF-bsignaling from the cell membrane to the nucleus and are,therefore, critical components of the TGF-b signaling pathway.SMAD4 (MIM] 600993), characterized as a key downstreamdeterminant in TGF-b signaling [Massague, 1998], is a transcrip-tional comodulator capable of integrating cellular responses tomultiple signaling cascades. SMAD4, also known as deleted inpancreatic carcinoma locus 4 (DPC4), was originally isolated fromhuman chromosome 18q21.1 as a tumor suppressor gene forpancreatic cancer [Hahn et al., 1996]. Mutations in SMAD4/DPC4 are frequently found in tumors, including pancreaticadenocarcinomas and colorectal carcinoma [Hahn et al., 1996;Schutte et al., 1996], suggesting a pivotal role for SMAD4 in TGF-

b functional loss in tumorigenesis. In fact, defects in SMAD4play a significant role in the malignant progression of tumors.Tumors lacking functional SMAD4 tend to be more invasive andangiogenic and, consequently, are more likely to form metastaticlesions [Sunamura et al., 2002].

TGF-b is one of the most potent endogenous negativeregulators of hematopoiesis. Recent studies have demonstratedthat TGF-b plays an important role in hematopoiesis by regulating

Published online 24 July 2006 in Wiley InterScience (www.interscience.wiley.com).

DOI10.1002/humu.20387

Received 22 June 2005; accepted revised manuscript 27January 2006.

Grant sponsor: National Institutes of Health (NIH); Grant number:CA112942-01; DK53757;CA101955-01.

LeiYang and NingWang contributed equally to this work.

�Correspondence to: Mei Wan, MD., Ph.D., Department of Patho-logy, School of Medicine, University of Alabama at Birmingham,Birmingham, AL 35294. E-mail: [email protected]

yThis article is a US Government work, and, as such, is in the publicdomain in the United States of America.

PUBLISHED 2006 WILEY-LISS, INC.

the proliferation and differentiation of hematopoietic cells[Fortunel et al., 2000; Kim and Letterio, 2003]. Autocrineproduction of TGF-b by hematopoietic stem cells acts to maintaintheir quiescence and to protect hematopoietic stem cells fromagents that selectively kill cycling cells [Grzegorzewski et al.,1994]. Inactivation of genes involved in the TGF-b signaltransduction pathway may serve as the mechanism whereby somehematopoietic progenitors escape from quiescence and cell-cyclinginhibition. Abnormalities in the expression of TGF-b receptorshave been described in the proliferation syndromes associatedwith both myeloid and lymphoid leukemia [DeCoteau et al., 1997;Lagneaux et al., 1997; Rooke et al., 1999]. Inhibition of TGF-bsignaling by an oncogene is involved in the formation of humanhematological malignancies [Kurokawa et al., 1998], suggestingthat the TGF-b signaling pathway acts as a tumor suppressor inhematopoietic cells.

Recently, it has been demonstrated that one missense mutationin the Mad homolog 1 (MH1) domain (c.305C4T, Pro102Leu)and one frameshift mutation resulting in termination of the Madhomolog 2 (MH2) domain (c.1447_1448insAATA, D483–552) inthe SMAD4 gene are associated with the pathogenesis of acutemyelogenous leukemia, indicating that disruption of the TGF-bsignaling pathway could lead to acute myelogenous leukemia [Imaiet al., 2001]. The mechanisms underlying SMAD4 inactivationcaused by mutations in neoplastic cells are not fully understood.Recently, mutations in SMAD4 have been shown to targetproteins for rapid degradation via the ubiquitin-proteasomepathway [Xu and Attisano, 2000; Maurice et al., 2001; Morenet al., 2003], indicating that protein instability of SMAD4 maycontribute to the loss in cellular responsiveness to TGF-b signalingin tumors. Although it has been shown that the SMAD4mutations identified in acute myelogenous leukemia lead to itsfunctional inactivation [Imai et al., 2001], it is possible that thesetwo mutations actually cause SMAD4 protein instability. Sincecellular responses are highly sensitive to the expression level ofSmad protein [Shimizu and Gurdon, 1999], studies of proteindegradation of mutated SMAD4 should be instructive for gaininga better understanding of the role of SMAD4 in human acutemyelogeneous leukemia.

Ubiquitin-dependent degradation by the 26S proteasome hasemerged as a central mechanism governing protein turnover. Apolyubiquitin chain is built onto either one or multiple lysineresidues of a substrate to target it for capture and degradation bythe 26S proteasome [DeSalle and Pagano, 2001; Pickart, 2001].The polyubiquitination (or polyubiquitylation) reaction requiresthe coordination of three classes of different enzymes: E1, E2, andE3. As a key component of the ubiquitination (or ubiquitylation)pathway, the ubiquitin ligases (E3 ligases) control both thespecificity and timing of substrate ubiquitination. Several classes ofE3 ligases have been identified, among which are the Cullin-basedE3 ligases. The SCF (Skp1-Cul1-F-box) complex is thus farthe most well characterized Cullin-based ligase [Deshaies, 1999;Jackson and Eldridge, 2002]. The variable F-box protein serves asthe substrate recognition subunit. The b-transducin-repeat-containing protein 1 (b-TrCP1; MIM] 603482) members of theFbw subfamily of F-box proteins is known to recognize thephosphorylated DSG(X)21nS motif within its substrates, whichincludes many important cellular regulatory proteins [Maniatis,1999; Jin et al., 2003; Fuchs et al., 2004]. Skp2 (MIM] 601436) isanother F-box protein of the SCF ubiquitin E3 ligase thatrecognizes and binds to many protein substrates [Nakayama andNakayama, 2005]. Therefore, SCF E3 ligases ubiquitinate andcontrol the stability of specific protein substrates. The E3 ligases

also play a pivotal role in the regulation of cell division and varioussignal transduction pathways that are essential for many aspectsof tumorigenesis.

We have previously demonstrated that SCFb-TrCP1, a ubiquitin(E3) ligase, is a critical determinant for degradation of SMAD4protein [Wan et al., 2004], and that SCFb-TrCP1 is one of thecontributors to rapid protein degradation of pancreatic tumor–derived SMAD4 mutants [Wan et al., 2005]. A recent studydemonstrated that SCFSkp2 is also an E3 ligase that mediatesincreased ubiquitination and accelerated proteolysis of cancer-derived SMAD4 mutants [Liang et al., 2004]. In this study, weexamined the protein stability of SMAD4 mutants found in acutemyelogenous leukemia. We found that SMAD4 mutants exhibit asignificant decrease in protein stability. Importantly, the F-boxprotein b-TrCP1 in SCF E3 ligase interacts with SMAD4 andexhibits a higher affinity for acute myelogenous leukemia–derivedSMAD4 mutants. These results suggest that the SCFb-TrCP1 E3ligase complex mediates ubiquitination of SMAD4 mutants. Usingsmall interference RNA (siRNA)-induced F-box protein b-TRCP1gene silencing, the protein steady state level of SMAD4 was shownto be elevated in acute myelogenous leukemia cells.

MATERIALSANDMETHODSConstructs

Wild-type SMAD4 expression plasmid was amplified by PCRusing pRK5-SMAD4 as the template and then subcloned into theEcoRI and SalI sites of the pCMV5B vector with a Flag tag at theamino terminus. Flag-tagged point mutated (P102L) and deleted(D483–552) SMAD4 expression plasmid were constructed usingsite-directed mutagenesis with the Quick Change kit (Stratagene,La Jolla, CA) according to the manufacturer’s protocol. TheSMAD4 gene sequence was compared with a reference cDNAsequence (GenBank accession: NM_005359.3). The mutationnumbering was based on cDNA sequence (c., cDNA). The firstbase of the ATG of initiation Met codon was reported asnucleotide 11.

Green fluorescent protein (GFP) siRNA and b-TrCP1 siRNAplasmids were generated using the BS/U6 vector [Sui et al., 2002].Briefly, a 22-nt oligonucleotide (oligo 1) corresponding tonucleotides 106–127 of GFP or nucleotides 453–474 of thehuman b-TRCP1 coding region was first inserted into the BS/U6vector digested with ApaI (blunted) and HindIII. The invertedmotif that contains the 6-nt spacer and five Ts (oligo 2) was thensubcloned into the HindIII and EcoRI sites of the intermediateplasmid to generate BS/U6/GFP and BS/U6/ b-TrCP1. For cloninginto retrovectors, the U6 promoter region plus the siRNA cassettewas digested with XbaI and cloned into the XbaI site of a retrovirusvector DU3.

Antibodies

Anti-human SMAD4 monoclonal antibody, anti-human b-TrCPpolyclonal antibody, and antiubiquitin polyclonal antibody werepurchased from Santa Cruz Biotechnology (Santa Cruz, CA).Anti-Flag and anti-HA monoclonal antibodies were purchasedfrom CPR, Inc. (Denver, PA). Anti-Myc and anti-b-actinmonoclonal antibodies were purchased from Sigma-Aldrich(St. Louis, MO).

Cell Culture andTransfection

293T, COS-1 cells were obtained and cultured according toprotocols from the American Tissue Cell Culture Collection(www.atcc.org). A monocytoid leukemia cell line CTV-1 cells were

898 HUMANMUTATION 27(9), 897^905,2006

Human Mutation DOI 10.1002/humu

maintained in RPMI 1640 with 10% FCS. 293T cells weretransfected using Tfx-20 Reagent (Promega, Madison, WI)according to the manufacturer’s instruction. For transienttransfection, CTV-1 cells were transfected by the electroporationmethod [Schakowski et al., 2004] using an electroporationapparatus (GENE PULSER II; Bio-Rad Laboratories, Hercules,CA). Briefly, 5 � 106 cells were suspended in 500 ml completemedium, mixed with 30mg of plasmid in a 4-mm electroporationcuvette, and incubated on ice for 10 min. After electroporationwith a single pulse, the cells were transferred into completemedium at a density of 1� 106 cells per ml.

Immunoprecipitation andWestern Blotting Analysis

Cells were lysed in buffer A (150 mM NaCl, 1% Triton X-100,0.5% deoxycholic acid, 50 mM Tris buffer pH7.5, 1 mM phenyl-methylsulfonyl fluoride, 10 mg/ml aprotinin, and 10mg/ml leupep-tin). For immunoprecipitation, lysed cells were incubated withdifferent antibodies as indicated in Figures 3A, 3B, and 3C, andincubated at 41C for 3 hr. Protein G plus agarose beads(Amersham, Piscataway, NJ) were added to the antigen-antibodymixture and incubated overnight. Immunocomplexes were washedthoroughly with buffer A containing 0.1% SDS and then separatedon a 12% SDS-polyacrylamide gel and blotted to nitrocellulose.All blots were developed by the enhanced chemiluminescenceECL technique (Amersham, Piscataway, NJ).

InVivo Ubiquitination Assays

At 40 hr after transfection, cell lysates were immunoprecipitatedusing antibody against Flag, boiled in SDS, and then reprecipitatedprior to immunoblotting. To detect ubiquitination of precipitatedSMAD4, Western blot analysis was performed using anti-HAantibody.

Immuno£uorescence

CTV-1 cells were cotransfected with GFP plasmid (2 mg) andb-TrCP1 RNAi plasmid (20mg) or GFP plasmid (2mg) with BS/U6vector (20mg). After 48 hr, cells were harvested and reseeded onpoly-L-lysine (Sigma, St. Louis, MO) pretreated glass coverslips.Cells were fixed in 3.7% formaldehyde and permeabilized with0.1% Triton X-100. After preblocking, the cells were treated withanti-SMAD4 or anti-b-TrCP1 antibodies and then incubated withTexas-red-conjugated donkey anti-mouse immunoglobulin G (IgG)or Texas-red-conjugated donkey anti-rabbit IgG (Jackson Immu-noResearch Laboratories, Inc., West Grove, PA) as secondaryantibodies. Images were observed with a fluorescence microscope(Olympus IX70; Tokyo, Japan), and photos were taken using thesame optical parameters to ensure comparable luminosity. Fivephotographs from each slides were randomly taken at 400�magnification, cells with green light (GFP transfecting showing thetransfection efficiency) or significant elevated red light (SMAD4level increases) were counted in each photograph. The results wereexpressed as percentage of GFP cells per 100 cells (transfectionefficiency) or SMAD4 elevated cells per 100 GFP cells. Data werecompiled and analyzed by Microsoft Excel (www.microsoft.com).

For virus infection, CTV-1 cells were infected with viruscontaining DU3/U6-GFP (si-GFP) or DU3/U6-b-TrCP1 (si-b-TrCP1) and were fixed, stained with either b-TrCP1 or SMAD4antibodies and then incubated with Texas-red-conjugated donkeyanti-mouse IgG as described above. Nuclei were counterstainedwith To-Pro-3 iodide (blue). Fluorescence density in the slidesstained with SMAD4 was analyzed and quantitated using softwareIPlab (version 3.0). Florescence intensity of the whole slide or the

mean density of 10 represented cells on each slide was measuredand was presented as histogram or relative intensity with standarddeviation, respectively.

RESULTSInstability of Mutated SMAD4 Proteins DerivedFrom Acute Myelogenous Leukemia

The stability of two mutated SMAD4 proteins derived fromacute myelogenous leukemia was examined. Flag-tagged SMAD4expression constructs harboring SMAD4 mutations identified inacute myelogenous leukemia were generated, among which one isa point mutation in the MH1 domain (P102L) and the other is adeletion in the MH2 domain (D483–552) in SMAD4. We firstexamined the expression levels of mutated SMAD4 protein in293T cells by transient transfection with the same amount ofSMAD4 wild-type (WT), P102L or D483–552 plasmid. As shownin Figure 1A, P102L exhibits a much lower level of SMAD4protein expression when compared to WT SMAD4 protein (Fig.1A; lane 2), and D(483–552) almost completely lacks SMAD4protein expression (Fig. 1A; lane 3).

The observed low protein expression level of SMAD4 mutantssuggests that the two mutations might lead to an increasedturnover of protein. To investigate this possibility, COS-1 cellswere transfected with WT SMAD4 and the two mutated

FIGURE 1. Instability of mutated SMAD4 proteins derived fromacute myelogeneous leukemia. A:The expression levels of acutemyelogeneous leukemia^derived SMAD4 mutants. 293T cellswere transfected with equal amounts of Flag-tagged wild typeandmutated SMAD4 expression plasmids as indicated. Extractswere assayed by Western blotting with antibodies speci¢c forFlag and b-actin. B: The steady-state levels of acute myeloge-neous leukemia^derived SMAD4mutants.293Tcellswere trans-fectedwith Flag-taggedwild type andmutated SMAD4 plasmidsas indicated. After 40 hr, cycloheximide (40 mg/ml) was added tothe culture to prevent any further SMAD4 synthesis.Whole cellextracts were prepared at di¡erent time points as indicated andassayedbyWesternblottingwith ananti-Flagantibody recogniz-ing both the full-length and truncated SMAD4 proteins. C:Theintensity of the bands was quantitated by phosphorimaging andplotted relative to the amount present at time 0.This is one of therepresentative data from three repeats of the same experiments,fromwhichwe got quite similar results. Note:Themutation num-beringwas basedoncDNA sequence (c., cDNA).The ¢rst baseofthe ATG of initiation Met codon was reported as nucleotide 11(GenBank, Ref SeqNM_005359.3).

HUMANMUTATION 27(9), 897^905,2006 899


SMAD4s. After 40 hr, cycloheximide (80 mg/mL) was added toprevent further protein synthesis. Whole cell extracts wereprepared from the cells at different times after cycloheximideaddition, and SMAD4 protein was visualized by Western blotting.WT SMAD4 was very stable and its expression level remainedrelatively constant at 12 hr after the addition of cycloheximide(Fig. 1B). The level of SMAD4 P102L decreased much faster thanwild-type SMAD4 (Fig. 1B). The level of SMAD4 D(483–552)rapidly decreased after cycloheximide addition, and was undetect-able at 12 hr. The half-life of wild-type SMAD4 was 412 hr,P102L was �5 hr, and D(483–552) was �3 hr (Fig. 1C). Thisexperiment has been repeated three times, and the degradationrate of SMAD4 proteins exhibit very similar pattern each time.The results indicate that rapid protein degradation is the maincontributor leading to protein loss in acute myelogenousleukemia–derived SMAD4 mutations.

SMAD4 Mutants Undergo Protein DegradationThrough the Ubiquitin-Proteasome Pathway

To determine whether the two mutated SMAD4 proteins aretargeted for degradation via the ubiquitin-proteasome pathway,we examined the expression of SMAD4 mutants using MG-132,a highly specific proteasome inhibitor [Lee and Goldberg, 1998].293Tcells were transfected with different amounts of each plasmid(Flag-tagged WT SMAD4 800 ng/dish, P102L 800 ng/dish, andD(483–552) 6mg /dish), following which MG-132 was added.MG132 significantly elevated the SMAD4 protein expression levelin both point-mutated and deleted SMAD4 transfected cells(Fig. 2A; lanes 4 and 6), but not in WT SMAD4 transfected cells(Fig. 2A; lane 2). This result indicates that mutated SMAD4undergoes rapid degradation via the 26S proteasome.

We also examined whether the enhanced turnover observed inSMAD4 mutants is mediated through ubiquitination. Individualmutation or WT SMAD4 expression plasmids were cotransfectedwith or without ubiquitin expression plasmid in 293T cells. Asshown in Figure 2B, WT SMAD4 exhibits very little ubiquitina-tion with ubiquitin overexpression (Fig. 2B; lane 2 vs. lane 1).Stronger ladders of high molecular weight, ubiquitin-conjugatedSMAD4 products were observed in both point-mutated anddeleted SMAD4 transfected cells (Fig. 2B; lanes 4 and 6). Theseresults suggest that SMAD4 harboring acute myelogeneousleukemia–derived mutations are degraded more rapidly throughthe ubiquitin-proteasome pathway when compared to their wild-type counterparts.

Instability of SMAD4 Mutants Mediated by SCFb-TrCP1

Previously, we found that the SCFb-TrCP1 E3 ligase complex isresponsible for SMAD4 degradation. The F-box protein b-TrCP1in this complex associates with SMAD4, and SCFb-TrCP1 over-expressing cells display increased ubiquitination and degradationof SMAD4 [Wan et al., 2004]. The decreased protein stability ofacute myelogenous leukemia–derived SMAD4 mutants raises thequestion of whether mutated SMAD4 protein interacts with b-TrCP1. To address this question, 293T cells were individuallytransfected with either empty vector, WT SMAD4 or mutatedSMAD4 plasmids, and immunoprecipitation assays were per-formed as described above. Both WT and mutated SMAD4 forman interaction complex with b-TrCP1 (Fig. 3A and B; first panel).The results indicate that b-TrCP1 is an F-box protein thatrecognizes both SMAD4 P102L and D(483–552) protein. Skp2was previously reported as another F-box protein that specificallybinds to WT SMAD4 and very strongly to some cancer-derived

SMAD4 mutants [Liang et al., 2004]. We then examined whetherthese two acute myelogenous leukemia–derived SMAD4 mutantsalso bind to Skp2. 293T cells were individually cotransfected witheither empty vector or mutated SMAD4 plasmids and HA-Skp2,and immunoprecipitation assays were performed. Neither of themutated SMAD4 proteins interacts with Skp2 (Fig. 3C). SMAD4R100T, previously being demonstrated has strong interaction withSkp2 [Liang et al., 2004], interacts with Skp2 in our systemserving as a positive control. These results indicate that SCFSkp2 isnot an E3 ligase that mediates the ubiquitination and degradationof these two mutated SMAD4 proteins.

We then investigated whether SCFb-TrCP1 would enhanceubiquitination of the mutated SMAD4 proteins. Individualmutation or WT SMAD4 expression plasmids were cotransfectedwith or without CUL1, ROC1, and b-TRCP1 expression plasmidin 293T cells. As shown in Figure 4A, point-mutated and deletedSMAD4 protein exhibit polyubiquitination in the presence ofubiquitin. SCFb-TrCP1 significantly enhances protein ubiquitinationof these two proteins. Since we have previously demonstrated thatthe use of vector-based siRNA is successful in inhibiting the

FIGURE 2. SMAD4 mutants undergo protein degradationthrough theubiquitin-proteasomepathway. A: Proteasome inhi-bitor elevates the protein level of mutated SMAD4. 293T cellswere transfected with di¡erent amount of wild-type andmutatedSMAD4 plasmids as follows: wild-type 800 ng/dish; P102L800 ng/dish; and D(483^552) 6 mg/dish). MG-132 (20 mM), aninhibitor of the 26S proteasome,was added 4 hr before cell lysis.Total cell extracts were analyzed by immunoblotting using anti-bodies against Flag and b-actin. B: Mutated SMAD4 exhibitshigher ubiquitination thanwild-type SMAD4 protein.293Tcellswere transfected with the indicated plasmids, and MG132(20 mM) was added 4 hr before cell lysis. Cell lysates were immu-noprecipitated with Flag antiserum, and then reprecipitatedprior to immunoblottingwith antiubiquitin.

900 HUMANMUTATION 27(9), 897^905,2006


expression of b-TrCP1 and enhancing the expression of endogen-ous SMAD4 [Wan et al., 2004], we sought to determine whetherRNAi can increase the stability of acute myelogenous leukemia–-derived SMAD4 mutants. WT SMAD4 and mutated SMAD4were cotransfected with either BS/U6/b-TrCP1 (si-b-TrCP1) orBS/U6 (empty vector control) plasmids in 293T cells. Si-b-TrCP1increased the expression level of mutated SMAD4 (Fig. 4B; lanes4 and 6), suggesting that b-TrCP1 is a key factor in mediatingthe instability of acute myelogenous leukemia–derived SMAD4mutants.

b-TRCP1Gene Silencing Elevated SMAD4Protein Level in Acute Myelogeneous Leukemia Cells

CTV-1 acute myelogenous leukemia cells, which harbor a MH2deletion (D483–552), provide a useful tool to investigate theeffects of b-TrCP1 on SMAD4 mutant degradation. One-tenth theamount of pCDNA3 GFP was cotransfected with BS/U6/b-TrCP1(si-b-TrCP1) or BS/U6 (empty vector control), and immuno-fluorescence assays were performed. Since GFP and si-b-TrCP1were cotransfected with a ratio 1:10, the cells showing greenfluorescence (GFP-transfected) should also express si-b-TrCP1.Figure 5A demonstrates that cells expressing GFP exhibit muchless expression of b-TrCP1 in si-b-TrCP1-transfected cells (Fig. 5A;second line showing in red), whereas GFP had no significant effect

on empty vector transfected cells (Fig. 5A; first line showing inred). These results indicate that si-b-TrCP1 successfully sup-pressed b-TrCP1 expression. Importantly, cells expressing GFPexhibited significant elevation of SMAD4 protein expression (Fig.5A; fourth line showing in red) compared with that in non-GFPexpressed cells or cells expressing siRNA empty vector (Fig. 5A;3rd line showing in red). We also quantitated the SMAD4elevated cells in sib-TrCP1–positive cells. The transfectionefficiency is about 15.671.1%, which indicates the percentageof sib-TrCP1–positive cells, among which about 89.074.2% of thecells have elevated SMAD4 staining. However, only 3.670.7% ofthe empty vector control-transfected cells has elevated SMAD4staining. This result indicates that b-TrCP1 suppression in cellsresults in a concomitant increase in SMAD4 expression.

The efficiency of transient transfection of siRNA into CTV-1cells is relatively low and a Western blot of proteins harvested fromthese cells could barely show the differences between siRNA-treated and untreated CTV-1 cells. We therefore generatedretrovirus constructs that contain DU/U6/GFP (si-GFP, irrelevantsiRNA control) and DU/U6/b-TrCP (si-b-TrCP1). High infectionefficiencies (495%) were achieved by detecting the percentage of

FIGURE 3. AssociationofSMAD4mutantswith b-TrCP1.A:293Tcellswere transfectedwith the indicatedplasmids. Immunopreci-pitation assayswere performed using anti-b-TrCP1antibody andthe immunocomplex was detected by Western blotting usinganti-Flag antibody.The expression levels of Flag proteins and b-TrCP1incellswere also detected, as indicated in the lower panel.B: 293Tcells were transfected with the indicated plasmids. Im-munoprecipitation assays were performed using anti-Flag anti-body and the immunocomplex was detected byWestern blottingusing anti-b-TrCP1 antibody.The expression levels of Flag pro-teins and b-TrCP1 in cells were also detected, as indicated in thelower panel. C: 293T cells were transfected with the indicatedplasmids. Immunoprecipitation assays were performed usinganti-Flag antibody and the immunocomplex was detected byWesternblottingusinganti-Skp2 antibody.Theexpression levelsof Flag proteins and Skp2 in cells were also detected, as indi-cated in the lower panel.

FIGURE 4. InstabilityofSMAD4mutantsmediatedby b-TrCP1.A:SCFb-TrCP1 increases ubiquitination of mutated SMAD4 protein.293T cells were transfected with the indicated plasmids, andMG132 (20 mM) was added 4 hr before cell lysis.Cell lysateswereimmunoprecipitatedwithFlagantiserumand thenreprecipitatedprior to immunoblotting with antiubiquitin. B: si-b-TrCP1 ele-vated protein level of acute myelogenous leukemia^derivedSMAD4 mutants. BS/U6 (empty vector control) or BS/U6/b-TrCP1 (si-b-TrCP1) plasmid, as indicated, were cotransfectedinto 293T cells with wild-type or individual SMAD4 mutants.Cells were harvested 3 days later and the extracts were assayedbyWestern blotting with antibodies speci¢c for Flag (upper pa-nel) or b-actin (lower panel).

HUMAN MUTATION 27(9), 897^905,2006 901


GFP-positive cells in flow cytometry (data not shown). si-b-TrCP1significantly decreases b-TrCP1 level as shown in red (Fig. 5B; thirdline). Conversely, the SMAD4 expression level is elevated in si-b-TrCP1 infected CTV-1 cells also shown in red (Fig. 5B; fourth line).To further compare the expression level of SMAD4 in siRNAcontrol and sib-TrCP1-infected cells, fluorescence density in thesetwo slides was analyzed and quantitated using software IPlab(version 3.0). Both the florescence intensity of whole slide (Fig.5C) and the mean density of 10 represented cells (Fig. 5D) in sib-TrCP1-infected slides are significantly higher than the siRNAcontrol slides. In addition, Western blot was also conducted,and similar results were obtained (Fig. 5E). Taken together, theseresults demonstrate that b-TrCP1 is a critical factor in regulatingmutated SMAD4 protein stability in acute myelogeneousleukemia cells.

DISCUSSION

Loss of SMAD4 is associated with a poor prognosis in humancancers [Sunamura et al., 2002]. Recently, several lines of

evidence have shown that SMAD4 mutations identified in humancancer patients are rapidly degraded via the ubiquitin-proteasomepathway [Xu and Attisano, 2000; Maurice et al., 2001; Morenet al., 2003], indicating that protein instability of SMAD4/DPC4contributes significantly to a loss in cellular responsiveness toTGF-b in tumorigenesis. We provide evidence here that acutemyelogenous leukemia–derived SMAD4 mutations target SMAD4to the ubiquitin-proteasome pathway, indicating a possible role forSMAD4 in acute myelogeneous leukemia progression. Importantly,our data provides new insight into the mechanism by whichSMAD4 point mutations undergo rapid degradation in acutemyelogenous leukemia cells. We found a strong associationbetween SCFb-TrCP1 and the point mutated SMAD4 protein.Consequently, ubiquitination of point-mutated SMAD4 is sig-nificantly higher than that of wild-type SMAD4.

b-TrCP regulates the turnover of several key cell regulatoryproteins, including IkB [Yaron et al., 1998], b-catenin [Fuchset al., 1999], ATF4 [Lassot et al., 2001], Emi1 [Guardavaccaroet al. 2003], p100 nuclear factor (NF)-kB B2 [Fong and Sun,2002], NF-kB p105 [Lang et al., 2003], Dlg [Mantovani and

FIGURE 5. Suppressionof endogenous b-TrCP1inhibits the degradation of SMAD4 in acutemyelogenous leukemia cells. A: Loweredb-TrCP1 expression in CTV-1 cells by si-b-TrCP1 transient overexpression. Empty BS/U6 (empty vector control) or BS/U6/b-TrCP1(si-b-TrCP1) plasmids were cotransfected with 1:10 amount of pCDNA3-GFP plasmid into CTV-1 cells.Three days later, cells were¢xed, incubated with anti-b-TrCP1or anti-SMAD4 antibody and then incubated withTexas-red-conjugated donkey anti-mouse IgG.Nuclei were counter-stained withTo-Pro-3 iodide (blue). Images were observed by £uorescence microscopy (Olympus IX70;Tokyo,Japan). B:CTV-1cellswere infectedwith virus containing DU3/U6-GFP (si-GFP) or DU3/U6-b-TrCP1 (si-b-TrCP1) andwere stainedwith antibodies as described above. Images were observed with a Leica SP1 confocal microscope. C,D: Fluorescence density in theslides stainedwith SMAD4 was analyzed and quantitated using software IPlab (version 3.0). Florescence intensity of thewhole slide(C) or the mean density of10 represented cells on each slide (D) was measured and was presented as histogram or relative intensitywith standard deviation, respectively. E:CTV-1cells were infected with virus containing DU3/U6-GFP (si-GFP) or DU3/U6-b-TrCP1(si-b-TrCP1).Three days later, cells were harvested, and aWestern blot assay was performedwith anti-b-TrCP, anti-SMAD4, or anti-b-actin antibody as indicated.

902 HUMAN MUTATION 27(9), 897^905,2006


Banks, 2003], IFNAR1 Cdc25A [Busino et al., 2003], Wee1[Watanabe et al., 2004], prolactin receptor [Li et al., 2004], andSnail [Zhou et al., 2004] proteins. Therefore, SCFb-TrCP E3 ligasesubiquitinate and regulate the stability of specific protein substratesand play a pivotal role in the regulation of cell division and varioussignal transduction pathways that are essential for many aspects oftumorigenesis. In fact, many lines of evidence indicate thataberrant upregulation of b-TrCP is often found in cancer cell linesand primary tumors [Fuchs et al., 2004; Muerkoster et al., 2005],and that blocking b-TrCP activities represents a potentiallyeffective means for cancer treatment [Tang et al., 2005]. Ourpresent work elucidates the role of b-TrCP in mediating rapidturnover of acute myelogeneous leukemia–derived SMAD4mutants. This is the first demonstration that b-TrCP activityis correlated with hematopoietic malignancy. Since SMAD4 isa key downstream effector of TGF-b signaling pathway, a furtherexamination of the role of b-TrCP1 in leukemia will provideimportant new insight into the pathogenesis of this malignantdisease.

SCFb-TrCP1 may not be the only E3 ligase that mediates rapidprotein turnover in SMAD4 mutants. Skp2, another F-box proteinin the SCF E3 ligase family, preferentially binds to several cancer-derived SMAD4 mutants and accelerates their protein ubiquitina-tion and degradation [Liang et al., 2004]. However, in this study,we could not demonstrate Skp2 binding to these two acutemyelogenous leukemia–derived SMAD4 mutations. Both Skp2and b-TrCP1 are F-box proteins in the same SCF E3 ligase family,and both strongly bind to and increase protein turnover in somecancer-derived SMAD4 mutants. However, Skp2 and b-TrCP1seem to differentially regulate protein stability in SMAD4mutants. It will, therefore, be interesting to further investigatewhy and how Skp2 and b-TrCP1 recognize and bind to differentSMAD4 mutants.

The detailed mechanism by which b-TrCP1 more easilydegrades these mutated SMAD4 proteins remains to be deter-mined. It has been demonstrated that a SCF complex containingb-TrCP as the F-box protein (SCFb-TrCP) recognizes phosphory-lated substrates and mediates their protein degradation.One possibility is that some point-mutated SMAD4 might bephosphorylated more easily by certain protein-associated kinases.If so, the mutated protein would be much easier for b-TrCP1to recognize. SMAD4 is a multifunctional protein that containsa DNA binding domain MH1, a protein-interaction domain MH2,and a linker region [Massague, 1998]. A mutational hotspotwithin the MH2 domain corresponding to codons 330 to 370was termed the mutation cluster region (MCR) in a recent study[Iacobuzio-Donahue et al., 2004]. The immunohistochemicalstudies indicated that the majority of missense mutationsinactivate SMAD4 by protein degradation, whereas carcinomaswith missense mutations within the MCR retain SMAD4 proteinstability. It will be interesting to examine whether other mutationswithin MCR will affect their recognition by the F-box proteinb-TrCP1.

Many substrates of the SCF E3 ligases share a common feature:phosphorylation as a prerequisite for recognition by the ligase.Therefore, it is possible that SMAD4 mutants are degraded fasterdue to protein phosphorylation by a certain kinase, allowing foreasier recognition by F box proteins. JNK/p38 kinase has beenreported to induce serine phosphorylation and proteasomaldegradation of the pancreatic cancer–derived SMAD4 R100Tmutant [Liang et al., 2004]. It is also possible that these two acutemyelogenous leukemia–derived SMAD4 mutants form consensusphosphorylation sites for JNK or other kinases or change their

protein confirmations, thus increasing the possibility for interac-tion with b-TrCP1.

Considering the cytoplasmic position of the known b-TrCP andSkp2 substrates [Maniatis, 1999; Foster, 2003], we predict thatSMAD4 protein degradation occurs mainly in the cytosol. The factthat SMAD4 D(483–552) has lost its ability to translocate into thenucleus and persists in the cytoplasm, regardless of TGF-bstimulation [Imai et al., 2001], may be one of the reasons forthe more rapid degradation of this mutated SMAD4 protein. Iftrue, this may also explain why the binding of F-box proteinb-TrCP1 with SMAD4 D(483–552) is not stronger and that thedegradation rate is much faster than wild-type SMAD4. However,since both b-TrCP and Skp2 are also localized in the nucleus, wecannot rule out the possibility that b-TrCP1 recognizes SMAD4 orSMAD4 mutants in the nucleus, but the complexes are rapidlyexported to the cytoplasm where protein degradation occurs.

Functional inactivation caused by these SMAD4 mutations,especially deletions in the MH2 domain, cannot be excluded.Some SMAD4 mutations lose the ability to form homo- orheterotrimeric Smad complexes [Shi et al., 1997], while othersexhibit increased autoinhibition of the Smads by stabilizing theintramolecular interactions between the MH1 and MH2 domains[Hata et al., 1997]. In contrast, Xu and Attisano [2000] reportedthat some Smad mutations do not interfere with most of thefunctions of these tumor suppressor proteins, but insteadinactivate the proteins by inducing their targeting for theubiquitin-proteasome pathway. The SMAD4 D(483–552) identi-fied in acute myelogeneous leukemia has been reported to lose itsnuclear translocation ability, and thus disrupt TGF-b signaling[Imai et al., 2001]. However, the protein rapid degradation mayalso be a primary contributor for its functional loss because cellularresponses are highly sensitive to the levels of Smad proteins[Shimizu and Gurdon, 1999]. Therefore, we suggest that reductionin the steady-state levels of these Smad proteins, which isinsufficient to mediate TGF-b signaling, may be the primary defectin acute myelogenous leukemia. SMAD4 is a key signalingmolecule in TGF-b signaling and, therefore, is an attractivetherapeutic target in oncology because of its strong cancersuppression activity. Our study and other publications indicatethat it is becoming apparent that the instability of SMAD4proteins is a general phenomenon in cancer cells harboring pointmutations in SMAD4. Therefore, further characterization of thestructure and function of the SMAD4/SCFb-TrCP1 interactionmight lead to the discovery of new targets for anticancer drugdevelopment.

ACKNOWLEDGMENTS

We thank Drs. Yoichi Imai and Matsuo for the CTV-1 cell line.We thank Dr. Yang Shi for the BS/U6 plasmid, Dr. J. Wade Harperfor the Myc-tagged b-TrCP1 expression plasmid, and Drs. ZafarNawaz and Bert O’Malley for the HA-tagged ubiquitin expressionplasmid. This work was supported by National Institutes of Healthgrants CA112942-01 (to M.W.), DK53757 (to X.C.), andCA101955-01 (to S.M.V).

REFERENCES

Blobe GC, Schiemann WP, Lodish HF. 2000. Role of transforminggrowth factor beta in human disease. N Engl J Med 342:1350–1358.

Busino L, Donzelli M, Chiesa M, Guardavaccaro D, Ganoth D,Dorrello NV, Hershko A, Pagano M, Draetta GF. 2003.

HUMANMUTATION 27(9), 897^905,2006 903


Degradation of Cdc25A by beta-TrCP during S phase and inresponse to DNA damage. Nature 426:87–91.

DeCoteau JF, Knaus PI, Yankelev H, Reis MD, Lowsky R, LodishHF, Kadin ME. 1997. Loss of functional cell surface transforminggrowth factor beta (TGF-beta) type 1 receptor correlateswith insensitivity to TGF-b in chronic lymphocytic leukemia.Proc Natl Acd Sci USA 94:5877–5881.

Derynck R, Zhang YE. 2003. Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature425:577–584.

DeSalle LM, Pagano M. 2001. Regulation of the G1 to S transitionby the ubiquitin pathway. FEBS Lett 490:179–189.

Deshaies RJ. 1999. SCF and Cullin/Ring H2-based ubiquitinligases. Annu Rev Cell Dev Biol 15:435–467.

Fong A, Sun SC. 2002. Genetic evidence for the essentialrole of beta-transducin repeat-containing protein in the indu-cible processing of NF-kB2/p100. J Biol Chem 277:22111–22114.

Fortunel NO, Hatzfeld A, Hatzfeld JA. 2000. Transforming growthfactor-beta: pleiotropic role in the regulation of hematopoiesis.Blood 96:2022–2036.

Foster JS, Fernando RI, Ishida N, Nakayama KI, Wimalasena J.2003. Estrogens down-regulate p27Kip1 in breast cancer cellsthrough Skp2 and through nuclear export mediated by the ERKpathway. J Biol Chem 278:41355–41366.

Fuchs SY, Chen A, Xiong Y, Pan ZQ, Ronai Z. 1999. HOS, ahuman homolog of Slimb, forms an SCF complex with Skp1 andCullin1 and targets the phosphorylation-dependent degradationof IkB and beta-catenin. Oncogene 18:2039–2046.

Fuchs SY, Spiegelman VS, Kumar KG. 2004. The many faces ofbeta-TrCP E3 ubiquitin ligases: reflections in the magic mirror ofcancer. Oncogene. 23:2028–2036.

Grzegorzewski KF, Ruscetti FW, Usui N, Damia G, Longo DL,Carlino JA, Keller JR, Wiltrout RH. 1994. Recombinanttransforming growth factor beta1 and beta 2 protect mice fromacutely lethal doses of 5-fluorouracil and doxorubicin. J ExpMed 180:1047–1057.

Guardavaccaro D, Kudo Y, Boulaire J, Barchi M, Busino L, DonzelliM, Margottin-Goguet F, Jackson PK, Yamasaki L, Pagano M.2003. Control of meiotic and mitotic progression by the F boxprotein b-Trcp1 in vivo. Dev Cell 4:799–812.

Hahn SA, Schutte M, Hoque AT, Moskaluk CA, da Costa LT,Rozenblum E, Weinstein CL, Fischer A, Yeo CJ, Hruban RH,Kern SE. 1996. DPC4, a candidate tumor suppressor gene athuman chromosome 18q21.1. Science 271:350–353.

Hata A, Lo RS, Wotton D, Lagna G, Massague J. 1997. Mutationsincreasing autoinhibition inactivate tumour suppressors Smad2and Smad4. Nature 388:82–87.

Iacobuzio-Donahue CA, Song J, Parmiagiani G, Yeo CJ, HrubanRH, Kern SE. 2004. Missense mutations of MADH4: char-acterization of the mutational hot spot and functionalconsequences in human tumors. Clin Cancer Res 10:1597–1604.

Imai Y, Kurokawa M, Izutsu K, Hangaishi A, Maki K, Ogawa S,Chiba S, Mitani K, Hirai H. 2001. Mutations of the Smad4 genein acute myelogeneous leukemia and their functional implica-tions in leukemogenesis. Oncogene 20:88–96.

Jackson PK, Eldridge AG. 2002. The SCF ubiquitin ligase: anextended look. Mol Cell 9:923–925.

Jin J, Shirogane T, Xu L, Nalepa G, Qin J, Elledge SJ,Harper JW. 2003. SCFb-TRCP links Chk1 signaling todegradation of the Cdc25A protein phosphatase. Genes Dev17:3062–3074.

Kim S, Letterio J. 2003. Transforming growth factor-beta signalingin normal and malignant hematopoiesis. Leukemia 17:1731–1737.

Kurokawa M, Mitani K, Irie K, Matsuyama T, Yakahashi T, ChibaS, Yazaki Y, Matsumoto K, Hirai H. 1998. The oncoproteinEvi-1 represses TGF-beta signalling by inhibiting Smad3. Nature394:92–96.

Lagneaux L, Delforge A, Bron D, Massy M, Bernier M, StryckmansP. 1997. Heterogenous response of B lymphocytes to transform-ing growth factor-beta in B-cell chronic lymphocytic leukaemia:correlation with the expression of TGF-beta receptors. Br JHaematol 97:612–620.

Lang V, Janzen J, Fischer GZ, Soneji Y, Beinke S, Salmeron A,Allen H, Hay RT, Ben Neriah Y, Ley SC. 2003. bTrCP-mediatedproteolysis of NF-kB1 p105 requires phosphorylation of p105serines 927 and 932. Mol Cell Biol 23:402–413.

Lassot I, Segeral E, Berlioz-Torrent C, Durand H, Groussin L, HaiT, Benarous R, Margottin-Goguet F. 2001. ATF4 degradationrelies on a phosphorylation-dependent interaction with theSCF(b-TrCP) ubiquitin ligase. Mol Cell Biol 21:2192–2202.

Lee DH, Goldberg AL. 1998. Proteasome inhibitors: valuable newtools for cell biologists. Trends Cell Biol 8:397–403.

Li Y, Suresh Kumar KG, Tang W, Spiegelman VS, Fuchs SY. 2004.Negative regulation of prolactin receptor stability and signalingmediated by SCFb-TrCP E3 ubiquitin ligase. Mol Cell Biol 24:4038–4048.

Liang M, Liang YY, Wrighton K, Ungermannova D, Wang XP,Brunicardi FC, Liu X, Feng XH, Lin X. 2004. Ubiquitinationand proteolysis of cancer-derived Smad4 mutants by SCFSkp2.Mol Cell Biol. 24:7524–7537.

Maniatis T. 1999. A ubiquitin ligase complex essential for the NF-kappaB, Wnt/Wingless, and Hedgehog signaling pathways.Genes Dev 13:505–510.

Mantovani F, Banks L. 2003. Regulation of the discs large tumorsuppressor by a phosphorylation-dependent interaction with theb-TrCP ubiquitin ligase receptor. J Biol Chem 278:42477–42486.

Massague J. 1998. TGF-beta signal transduction. Annu RevBiochem 67:753–791.

Maurice D, Pierreux CE, Howell M, Wilentz RE, Owen MJ, HillCS. 2001. Loss of Smad4 function in pancreatic tumors:C-terminal truncation leads to decreased stability. J Biol Chem276:43175–43181.

Moren A, Hellman U, Inada Y, Imamura T, Heldin CH, MoustakasA. 2003. Differential ubiquitination defines the functionalstatus of the tumor suppressor Smad4. J Biol Chem 278:33571–33582.

Muerkoster S, Arlt A, Sipos B, Witt M, Grossmann M, Kloppel G,Kalthoff H, Folsch UR, Schafer H. 2005. Increased expression ofthe E3-ubiquitin ligase receptor subunit betaTRCP1 relates toconstitutive nuclear factor-kappaB activation and chemoresis-tance in pancreatic carcinoma cells. Cancer Res 65:1316–1324.

Nakayama KI, Nakayama K. 2005. Regulation of the cell cycle bySCF-type ubiquitin ligases. Semin Cell Dev Biol 16:323–333.

Pickart CM. 2001. Mechanisms underlying ubiquitination. AnnuRev Biochem 70:503–533.

Rooke HM, Vitas MR, Crosier PS, Crosier KE. 1999. The TGF-beta type II receptor in chronic myeloid leukemia: analysis ofmicrosatellite regions and gene expression. Leukemia 13:535–541.

Schakowski F, Buttgereit P, Mazur M, Marten A, Schottker B,Gorschluter M, Schmidt-Wolf IG. 2004. Novel non-viral methodfor transfection of primary leukemia cells and cell lines. GenetVaccines Ther 2:1.

904 HUMANMUTATION 27(9), 897^905,2006


Schutte M, Hruban RH, Hedrick L, Cho KR, Nadasdy GM,Weinstein CL, Bova GS, Isaacs WB, Cairns P, Nawroz H,Sidransky D, Casero RA Jr, Meltzer PS, Hahn SA, Kern SE.1996. DPC4 gene in various tumor types. Cancer Res 56:2527–2530.

Shi Y, Hata A, Lo RS, Massague J, Pavletich NP. 1997. A structuralbasis for mutational inactivation of the tumour suppressorSmad4. Nature 388:87–93.

Shimizu K, Gurdon JB. 1999. A quantitative analysis of signaltransduction from activin receptor to nucleus and its relevanceto morphogen gradient interpretation. Proc Natl Acad Sci USA9:6791–6796.

Sui G, Soohoo C, Affare B, Gay F, Shi Y, Forrester WC, Shi Y.2002. A DNA vector-based RNAi technology to suppress geneexpression in mammalian cells. Proc Natl Acad Sci USA99:5515–5520.

Sunamura M, Yatsuoka T, Motoi F, Duda DG, Kimura M,Abe T, Yokoyama T, Inoue H, Oonuma M, Takeda K, MatsunoS. 2002. Gene therapy for pancreatic cancer based on geneticcharacterization of the disease. J Hepatobiliary Pancreat Surg9:32–38.

Tang W, Li Y, Yu D, Thomas-Tikhonenko A, Spiegelman VS,Fuchs SY. 2005. Targeting beta-transducin repeat-containingprotein E3 ubiquitin ligase augments the effects of antitumordrugs on breast cancer cells. Cancer Res 65:1904–1908.

Wan M, Tang Y, Tytler EM, Lu C, Jin B, Vickers SM, Shi X, Cao X.2004. Smad4 protein stability is regulated by ubiquitin ligaseSCF beta-TrCP1. J Biol Chem 279:14484–14487.

Wan M, Huang J, Jhala NC, Tytler EM, Yang L, Vickers SM, TangY, Lu C, Wang N, Cao X. 2005. SCFb-TrCP1 controls Smad4protein stability in pancreatic cancer cells. Am J Pathol166:379–392.

Watanabe N, Arai H, Nishihara Y, Taniguchi M, Watanabe N,Hunter T, Osada H. 2004. M-phase kinases induce phospho-dependent ubiquitination of somatic Wee1 by SCFb-TrCP. ProcNatl Acad Sci USA 101:4419–4424.

Xu J, Attisano L. 2000. Mutations in the tumor suppressors Smad2and Smad4 inactivate transforming growth factor beta signalingby targeting Smads to the ubiquitin-proteasome pathway. ProcNatl Acad Sci USA 97:4820–4825.

Yaron A, Hatzubai A, Davis M, Lavon I, Amit S, Manning AM,Andersen JS, Mann M, Mercurio F, Ben Neriah Y. 1998.Identification of the receptor component of the IkBa-ubiquitinligase. Nature 396:590–594.

Yue J, Mulder KM. 2001. Transforming growth factor-beta signaltransduction in epithelial cells. Pharmacol Ther 91:1–34.

Zhou BP, Deng J, Xia W, Xu J, Li YM, Gunduz M, Hung MC. 2004.Dual regulation of Snail by GSK-3b-mediated phosphorylationin control of epithelial-mesenchymal transition. Nat Cell Biol6:931–940.

HUMAN MUTATION 27(9), 897^905,2006 905


acute myelogenous leukemia–derived smad4 mutations target the protein to ubiquitin-proteasome...

Documents