Epidermal Growth Factor Receptor Turnover Complex ...

Bcr Interacts with Components of the Endosomal Sorting Complex Required for Transport-I and Is Required for Epidermal Growth Factor Receptor Turnover Oyenike O. Olabisi, Gwendolyn M. Mahon, Elena V. Kostenko, et al. Cancer Res 2006;66:6250-6257.

Updated version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/66/12/6250

Cited Articles

This article cites by 49 articles, 27 of which you can access for free at: http://cancerres.aacrjournals.org/content/66/12/6250.full.html#ref-list-1

Citing articles

This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/66/12/6250.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on June 3, 2013. © 2006 American Association for Cancer Research.

Research Article

Bcr Interacts with Components of the Endosomal Sorting Complex Required for Transport-I and Is Required for Epidermal Growth Factor Receptor Turnover Oyenike O. Olabisi, Gwendolyn M. Mahon, Elena V. Kostenko, Zhuoming Liu, Harvey L. Ozer, and Ian P. Whitehead Department of Microbiology and Molecular Genetics and University Hospital Cancer Center, New Jersey Medical School, University of Medicine and Dentistry of New Jersey, Newark, New Jersey

Abstract Virtually all patients with chronic myelogenous leukemia (CML) express an aberrant protein (p210 Bcr-Abl) that contains NH2-terminal sequences from Bcr fused to COOHterminal sequences from Abl. In a yeast two-hybrid screen, we have identified TSG101 as a binding partner for Bcr. Because TSG101 is a subunit of the mammalian endosomal sorting complex required for transport (ESCRT), which regulates protein sorting during endosomal trafficking, this association suggests that Bcr may have a related cellular function. The docking site for TSG101 has been mapped to the COOH terminus of Bcr, indicating that this interaction may be disrupted in CML. Overexpression studies with full-length TSG101 and Bcr reveal that this interaction can be recapitulated in mammalian cells. The association can also be observed between natively expressed proteins in a panel of hematopoietic and nonhematopoietic cell lines, where a second subunit of the ESCRT complex, vacuolar sorting protein 28 (Vps28), was also found to interact with Bcr. Both Bcr and TSG101 exhibit a punctate cytoplasmic distribution and seem to colocalize in HeLa cells, which would be consistent with an in vivo association. Bacterially purified Bcr and TSG101 also bind, suggesting that the interaction is direct and is not dependent on ubiquitination. Disruption of the endosomal pathway with an ATPase-defective Vps4 mutant results in the cellular redistribution of Bcr, and suppression of Bcr in HeLa cells by small interfering RNA impairs epidermal growth factor receptor turnover. Taken together, these observations suggest that Bcr is a component of the mammalian ESCRT complexes and plays an important role in cellular trafficking of growth factor receptors. (Cancer Res 2006; 66(12): 6250-7)

Introduction Chronic myelogenous leukemia (CML) is a malignant hematopoietic stem cell disorder that occurs at a frequency of 1 to 1.5 per 100,000 people (1). The cytologic hallmark of CML is a translocation [t(9;22)(q34;q11)], termed the Philadelphia chromosome, that occurs in >90% of all patients (2). In virtually all cases, translocation breakpoints fall within introns 13 or 14 of the BCR gene, and the first intron of the ABL gene. This produces a 210 kDa Requests for reprints: Ian P. Whitehead, Department of Microbiology and Molecular Genetics, New Jersey Medical School, University of Medicine and Dentistry of New Jersey, 225 Warren Street, Newark, NJ 07103. Phone: 973-972-4483, ext. 25215; Fax: 973-972-8981; E-mail: [email protected]. I2006 American Association for Cancer Research. doi:10.1158/0008-5472.CAN-06-0536

Cancer Res 2006; 66: (12). June 15, 2006

in-frame fusion protein (p210 Bcr-Abl) that contains NH2-terminal sequences of Bcr fused to COOH-terminal sequences of Abl. Elevated tyrosine kinase activity residing within Abl is considered to be a critical event in the induction of CML. Therapeutic agents that specifically target this activity are effective in the treatment of CML (3), and loss of this activity is invariably associated with diminished transforming activity (4). Sequences that reside within the Bcr component of p210 Bcr-Abl are also required for transformation. Autophosphorylated Tyr177 binds to growth factor receptor binding protein 2, which complexes with SOS to activate Ras (5). Loss of this residue results in a dramatic loss in the transforming activity. The oligomerizing property of the NH2terminal domain of Bcr also plays an essential role in the activation of Bcr-Abl tyrosine kinase activity (6). Removal of the domain abrogates transforming activity, which can be at least partially restored by the addition of unrelated sequences that possess oligomerizing ability (7). Finally, the NH2 terminus of Bcr binds to the SH2 domain of Abl in a phosphotyrosine-independent manner, constitutively activating the tyrosine kinase (8). The major product of the BCR gene is a 160 kDa protein that exists primarily as a soluble 600 kDa tetramer (1). Oligomerization is mediated by an NH2-terminal coiled-coil domain located within residues 1 to 71. Also contained within Bcr are an NH2-terminal (residues 1-414) serine/threonine kinase domain, a centrally located Rho-specific guanine nucleotide exchange factor (RhoGEF) domain, and COOH-terminal C2 and Rac-specific GTPase-activating protein (RacGAP) domains. RhoGEF domains catalyze the exchange of GDP for GTP on Rho family GTPases, thus converting them into their biologically active state (9). Although the C2 domain of Bcr has not been characterized, they have been shown to bind phospholipids in a calcium-dependent manner in other signaling molecules (10). GAP domains, including the one in Bcr, stimulate the intrinsic rate of hydrolysis of GTPases, thus converting them to the inactive GDP-bound form (11). The native function of Bcr remains largely unknown. The protein is broadly expressed and is predominantly associated with cytoplasmic membranes (12). Sporadic reports have also provided evidence for nuclear localization, which is regulated in a cell cycle– dependent manner (12–14). There are no orthologues of Bcr in yeast, Caenorhabditis elegans, or Drosophila, and homozygous null mice are viable, exhibiting mild neurologic disorders (15). The simultaneous disruption of both Bcr and the closely related Abr leads to more severe abnormalities in postnatal cerebellar development, which is consistent with the high expression of both proteins that is observed in neural tissue (16, 17). Interestingly, the neutrophils of bcr / mice show unusually high levels of reactive oxygen species when exposed to Gram-negative endotoxins, which would be consistent with chronic activation of Rac2 (18). This

6250

www.aacrjournals.org


Bcr Interacts with the ESCRT-I Complex

suggests that the COOH-terminal RacGAP domain of Bcr may play an important role in the regulation of the oxidative burst in professional phagocytes. The function of Bcr in other cell types is unknown. TSG101 was originally identified as a mammalian tumor suppressor gene through its insertional inactivation in NIH 3T3 mouse fibroblasts (19). An analysis of the protein structure of TSG101 revealed several identifiable domains. These include an NH2-terminal noncanonical E2 ubiquitin–conjugating enzyme domain (UEV; residues 23-160); centrally located proline-rich (residues 139-216) and coiled-coil (residues 242-312) domains; and a unique domain in the COOH-terminal region (residues 346-391) that has been implicated in protein stability (20). Like Bcr, TSG101 is a ubiquitously expressed protein whose subcellular localization is regulated in a cell cycle–dependent manner (21). TSG101 was initially identified by its ability to transform mouse fibroblasts when its expression is suppressed (19). Somewhat paradoxically, an analysis of tsg101 / mice revealed that the protein is essential for cell growth and embryonic development (22, 23). Subsequent to its identification, TSG101 was implicated in a variety of cellular processes, including transcriptional activation (24), modulation of the MDM2/p53 feedback loop (25), and cell cycle progression (26). More recently, it has been firmly established that TSG101 plays a central role in the sorting of endosomal cargo. In both yeast and mammalian cells, TSG101 is a component of a 350 kDa soluble protein complex [designated endosomal sorting complex required for transport-1 (ESCRT-I)], which recognizes ubiquitinated cargo on the endosomal membrane and mediates their internalization during formation of the multivesicular body (27, 28). TSG101 is thought to bind to ubiquitinated proteins through its UEV domain and facilitate their distribution into the multivesicular body. Mutations in Vps23p, the yeast homologue of TSG101, lead to defective transport of membrane proteins into yeast vacuoles, which are the yeast counterparts of the mammalian lysosome (27, 28). In this current study, we have identified TSG101 as a binding partner for Bcr. This interaction occurs in mammalian cells, is direct, and is not mediated by ubiquitination. An association between natively expressed Bcr and Vps28 has also been detected, which strongly suggest that Bcr is a bona fide component of the mammalian ESCRT-I complex. Bcr and TSG101 show a similar subcellular distribution in HeLa cells, and targeted suppression of both is associated with defective trafficking, suggesting a similar role in the endosomal pathway. Because TSG101 binds to a region of Bcr that is lost in p210 Bcr-Abl, this raises the possibility that a Bcr-mediated defect in trafficking may occur in the context of CML.

Materials and Methods Molecular constructs. The pGBT9 and pGAD(GH) yeast expression vectors are as described by the manufacturer (Clontech, Mountain View, CA). pGBT9-bcr (1-1,271), pGBT9-bcr(871-1,271), pGBT9-bcr(911-1,271), pGBT9-bcr(1,037-1,271), and pGBT9-bcr(1,165-1,271) contain cDNAs that encode the indicated amino acids of the human Bcr protein fused in frame to the GAL4 DNA-binding domain. pGAD-tsg101(1-390), pGAD-tsg101(223390), pGAD-tsg101(1-345), and pGAD-tsg101(223-345) contain cDNAs that encode the indicated amino acids of the human TSG101 protein fused in frame to the GAL4 DNA activation domain. The pAX142 mammalian expression vector has been described previously (29). pAX142-bcr(1-1,271) encodes the full-length human Bcr protein fused at the NH2-terminus to a hemagglutinin (HA) epitope tag (14). pAX142-tsg101(1-390) encodes the full-length TSG101 protein fused at the NH2 terminus to a myc-epitope tag. GFP-hVps4 and GFP-hVps4(KQ) contain wild-type and dominant inhibitory


versions of human Vps4 fused to green fluorescent protein (GFP; ref. 30: kindly provided by P. Woodman, Faculty of Life Sciences, University of Manchester, Manchester, United Kingdom). pGEX5X-1-bcr(871-1271) contains a cDNA that encodes the indicated amino acids of Bcr fused at the NH2-terminus to glutathione S-transferase (GST; ref. 14). pET21bHis6-tsg101(223-390) contains a cDNA that encodes the indicated amino acids of TSG101 fused at the COOH terminus to a polyhistidine tag. All PCR products that were used in the assembly of these constructs were sequenced in their entirety. Detailed descriptions of the cloning strategies that were used to generate each construct are available upon request. Yeast two-hybrid screening. A Bcr cDNA fragment encoding the fulllength protein (residues 1-1,271) was subcloned in frame into pGBT9 [pGBT9-bcr(1-1,271)] containing the GAL4 DNA-binding domain. A human HeLa MATCHMAKER cDNA library (S3; Clontech) was screened with the PJ69 yeast strain according to the instructions of the manufacturer (Clontech). Double selection by HIS3 and lacZ followed by retransformation of yeast with recovered cDNAs identified two independent TSG101 clones (residues 1-390 and 223-390) as binding partners for Bcr. Cell culture and transfection. 293T, HeLa, and ts20 3T3 cells (31) were maintained in DMEM (high glucose) supplemented with 10% fetal bovine serum (FBS; Sigma, St. Louis, MO). ts20 cells were maintained and passaged at the permissive temperature (35jC). K-562 and HL-60 cells (American Type Tissue Culture Collection, Manassas, VA) were maintained in Iscove’s modified medium supplemented with 4 mmol/L L-glutamine, 1.5 g/L sodium bicarbonate, and 20% FBS (Sigma, St. Louis, MO). Transient transfection of HeLa and 293T cells was by LipofectAMINE 2000 according to the instructions of the manufacturer (Invitrogen, Carlsbad, CA). Protein expression. Protein expression in transiently transfected 293T cells, or untransfected HeLa, 293T, K-562, ts20, and HL-60 cells was determined by Western blot analysis as described previously (32). Protein was visualized with enhanced chemiluminescence reagents (Amersham Pharmacia Biotech, Piscataway, NJ) or by Odyssey IR imaging. Coimmunoprecipitations. Cells were harvested in GST fluorescence in situ hybridization (FISH) buffer [50 mmol/L Tris-HCl (pH 7.4), 2 mmol/L MgCl2, 100 mmol/L NaCl, 10% glycerol, 1% NP40] supplemented with protease inhibitor cocktails (Calbiochem, San Diego, CA). Lysates were precleared with 0.25 Ag agarose-conjugated normal IgG (Santa Cruz Biotechnology, Santa Cruz, CA) by rotation for 1 hour at 4jC and precipitated with 10 Ag agarose-conjugated monoclonal antibody by rotation at 4jC overnight. The immunoprecipitate was washed thrice with cold lysis buffer and then resuspended in 2 SDS loading dye. In vitro affinity precipitations. GST-Bcr(871-1,271) was expressed in BL21(DE3) cells by induction with 0.1 mmol/L isopropyl-L-thio-h-Dgalactopyranoside (IPTG) and was purified with glutathione Sepharose 4B beads according to the instructions of the manufacturer (Amersham Pharmacia Biotech). His6-TSG101(223-390) was expressed in BL21(DE3) cells by induction with 0.4 mmol/L IPTG and was purified with Talon Metal Affinity Resin according to the instructions of the manufacturer (Clontech). Expression was confirmed by Coomassie staining and by Western blot using an anti-GST and anti-His antibody, respectively. Approximately 20 Ag glutathione Sepharose–linked GST or GST-Bcr(871-1,271) was incubated with f20 Ag bacterially purified His6-TSG101(223-390) with rotation for 1 hour at 4jC. Beads were then washed briefly thrice in GST FISH buffer, pelleted at 500 rpm for 5 minutes, separated by SDS-PAGE, and transferred to polyvinylidene difluoride membrane (Amersham). Bound TSG101 was detected by Western blot using an anti-TSG101 monoclonal antibody (C2, Santa Cruz Biotechnology) and then confirmed using an anti-histidinetag monoclonal antibody (H3, Santa Cruz Biotechnology). Immunostaining. Immunostaining of HeLa cells was done as described previously (33). Briefly, at 24 hours posttransfection, cells were trypsinized and plated on 0.02% poly-L-lysine–coated glass coverslips. At 48 hours posttransfection, cells were fixed with 3.7% formaldehyde in PBS for 20 minutes, and then permeabilized and blocked for 1 hour in 0.1% Triton X-100, 3% bovine serum albumin (BSA) in PBS. Coverslips were then incubated in a humidity chamber with the appropriate monoclonal antibody for 1 hour in 0.1% Triton X-100, 0.1% BSA in PBS, washed in PBS, and then incubated with either Alexa Fluor 568 (red) or 488

6251

Cancer Res 2006; 66: (12). June 15, 2006


Cancer Research (green)–conjugated secondary antibody (Molecular Probes, Carlsbad, CA) in the same buffer for 1 hour in the dark. Cells were viewed with a Zeiss Axiovert 200M microscope fitted with an ApoTome Imaging system. Image stacks in the axial direction were acquired and all images shown are from a representative axial plane. Epidermal growth factor receptor degradation assay. HeLa cells were transiently transfected by SilentFect reagent (Bio-Rad, Hercules, CA) with 20 nmol/L mixed small interfering RNA (siRNA) oligonucleotides targeted at either TSG101 or Bcr (Santa Cruz Biotechnology). Scrambled siRNAs were also included as negative controls. After 24 hours, cells were serum starved for an additional 12 hours and then 150 ng/mL epidermal growth factor (EGF; E9644, Sigma) was added to the medium for 90 minutes. Lysates were then collected and examined by Western blot for expression of the EGF receptor (EGFR; 1005, Santa Cruz Biotechnology), Bcr (N20, Santa Cruz Biotechnology), and TSG101 (C2, Santa Cruz Biotechnology).

Results p160 Bcr interacts with TSG101 in yeast. To identify new binding partners for p160 Bcr, a yeast two-hybrid screen was done.

Figure 2. p160 Bcr interacts with TSG101 and Vps28 in mammalian cells. IP, antibody used in immunoprecipitations; IB, antibody used in a Western blot to detect an interaction. A, 293T cells were cotransfected with the indicated combinations of plasmids [vector, pAX142; Bcr, pAX142-bcr (1-1271); TSG101, pAX142-tsg101 (1-390)]. Lysates were collected at 48 hours and examined by Western blot for expression of TSG101 (middle ) or Bcr (bottom ). Immunoprecipitations were then done with an anti-Myc antibody to detect an interaction (top ). B and C, endogenous Bcr and TSG101 interact in hematopoietic (HL60, K562) and nonhematopoietic (HeLa, 293T) cells. B, lysates were collected from K-562, HL-60, HeLa, and 293T cells and then immunoprecipitations were done by using either an anti-TSG101 antibody or an irrelevant anti-HA antibody. Interactions were shown by Western blot using an anti-Bcr antibody that recognizes the amino terminus of p160 Bcr (top ). C, specificity of the interaction between endogenous Bcr and TSG101 in HeLa and 293T cells was confirmed using a blocking peptide for the Bcr antibody (N20-P, Santa Cruz Biotechnology). D, p160 Bcr interacts with endogenous Vps28 in HeLa and 293T cells. Lysates were collected and then immunoprecipitations were done with an anti-Vps28 antibody. An interaction was shown by Western blot using an anti-Bcr antibody (N20, Santa Cruz Biotechnology). The specificity of the binding was again confirmed with the Bcr N20 blocking peptide.

Figure 1. Mammalian TSG101 and p160 Bcr interact in yeast. A, mapping of the TSG101 binding site within p160 Bcr. Top (top line ), domain structure of the full-length Bcr protein (O, oligomerization domain; kinase, serine/threonine kinase domain; DH/PH, RhoGEF domain; C2, calcium-binding domain); bottom lines, regions of the protein included in predicted translation products of the various cDNA derivatives. Bottom, yeast colonies that grew on plates lacking leucine and tryptophan (Leu/Trp ) were examined for growth on histidine-deficient plates (Leu/Trp/His). Growth in the absence of leucine indicates the presence of pGAD containing residues 1 to 390 of TSG101, whereas growth in the absence of tryptophan indicates the presence of pGBT9 containing the indicated derivatives of Bcr. Interactions between proteins are shown by the ability to activate the his3 reporter gene. B, full-length Bcr binds to the carboxyl-terminal region of TSG101. Top (top line ), domain structure of the full-length TSG101 protein (PP, proline-rich domain; CC, coiled-coil domain; SB, steadiness box). Bottom, interactions between full-length p160 Bcr and the indicated derivatives of TSG101, determined as described above.

Cancer Res 2006; 66: (12). June 15, 2006

Multiple independent copies of four structurally unrelated clones were identified that interacted with the full-length p160 Bcr bait. One of these clones encodes TSG101, the mammalian orthologue of yeast Vps23 (Fig. 1A). Using an extended panel of Bcr baits, we were able to map the TSG101 docking site to a COOH-terminal region of Bcr (residues 1,037-1,165) that is lost in p210 Bcr-Abl. Yeast twohybrid analysis was also used to map the Bcr binding site in TSG101 to the COOH-terminal 45 amino acids (Fig. 1B; residues 345-390). p160 Bcr interacts with TSG101 in mammalian cells. Next, we wished to determine whether the association between TSG101 and Bcr can be recapitulated in mammalian cells. To address this possibility, we coexpressed Myc epitope–tagged full-length TSG101 and HA-tagged full-length Bcr in 293T cells and then immunoprecipitated TSG101 (Fig. 2A) with an anti-Myc monoclonal antibody (9E10, Santa Cruz Biotechnology). Immunoprecipitates were then separated by Western blot and probed for the presence of Bcr using an anti-HA antibody (HAY11, Santa Cruz Biotechnology). Using this approach, we were readily able to detect an association between

6252




the two full-length proteins. To confirm that binding can be detected in a more physiologically relevant context, the association between natively expressed proteins was also examined. Initially, we examined the interaction in a panel of hematopoietic (K562 and HL60) and nonhematopoietic (293T and HeLa) cell lines (Fig. 2B). K-562 is a human myeloid cell line that was isolated from a patient with CML and, thus, is also positive for p210 Bcr-Abl. HL-60 cells are a leukemic cell line of myeloid origin that does not contain p210 Bcr-Abl. In all cell lines, we were able to coimmunoprecipitate Bcr with an antibody that detects endogenous TSG101 (Fig. 2B, top) but not with an irrelevant antibody (Fig. 2B, bottom). The specificity of the interaction was then confirmed in 293T and HeLa cells using a blocking peptide for the Bcr antibody (N20-P, Fig. 2C). As predicted from our mapping studies, the TSG101 antibody did not coimmunoprecipitate p210 Bcr-Abl from the K562 cells (not shown). TSG101 interacts with Vps28. TSG101 is a component of mammalian ESCRT-I, a protein complex that is found on the limiting membrane of the multivesicular body (34). In addition to TSG101, this complex also contains Vps28 and Vps37. To determine whether the TSG101-Bcr interaction is occurring in the context of ESCRT-I, we determined whether we can also coimmunoprecipitate endogenous Bcr with a Vps28 antibody (N12, Santa Cruz Biotechnology). Using an antibody specific for Bcr (N20, Santa Cruz Biotechnology), we were readily able to detect the association between endogenous Vps28 and Bcr in HeLa and 293T cells (Fig. 2D). The specificity of the interaction was also confirmed using the blocking peptide for the Bcr antibody (N20-P, Santa Cruz Biotechnology). The interaction between TSG101 and Bcr is direct and ubiquitin independent. Because TSG101 is thought to bind to its targets in a ubiquitin-dependent manner, we wondered whether the association between Bcr and TSG101 is mediated by ubiquitination (28, 35). To test this possibility, we used two approaches. First, we purified His6-tagged TSG101 and GST-tagged Bcr in bacteria and did an in vitro binding assay. An interaction was observed between His6-TSG101(223-390) and GST-Bcr(871-1,271) but not with GST alone, indicating that the interaction is direct (Fig. 3A). Because ubiquitination does not occur in prokaryotes, this result would also suggest that the interaction does not require

Figure 3. The interaction between TSG101 and Bcr is direct and ubiquitin independent. A, His6-TSG101(223-390) interacts with GST-Bcr(871-1,271) in an in vitro affinity precipitation assay. Bacterially purified His6-TSG101(223-390) was incubated with bacterially expressed, Sepharose-linked GST or GST-Bcr(871-1,271). The beads were then washed and the interaction was shown by Western blot using an anti-TSG101 antibody. B, Bcr and TSG101 interact in mammalian cells in the absence of ubiquitination. ts20 3T3cells were cultured at the permissive (35jC) and restrictive (39jC) temperatures for 18 hours and then lysates were collected and examined by coimmunoprecipitation for an interaction between endogenous TSG101 and Bcr. The defect in ubiquitination was shown by the accumulation of cyclin D2 at the restrictive temperature. RhoB expression was also monitored as a negative control.


the ubiquitination of either protein. To confirm that ubiquitination does not mediate the association between full-length Bcr and TSG101 in mammalian cells, we took advantage of a cell line that contains a temperature-sensitive allele of the mammalian E1 (ts20 3T3; ref. 31). We have shown in previous studies that ubiquitination does not occur in these cells when they are cultured at the restrictive temperature (31). For this analysis, cells were either left at the permissive temperature or shifted to the restrictive temperature for 18 hours and then lysates were collected. The defect in the ubiquitination pathway was shown by the accumulation of cyclin D2 at the restrictive temperature (Fig. 3B; ref. 31). Because we were able to coimmunoprecipitate endogenous Bcr and TSG101 at both the permissive and restrictive temperatures, we conclude that the interaction does not require ubiquitination of either protein in mammalian cells. TSG101 and Bcr have similar patterns of subcellular distribution. Although binding partners for Bcr have been described previously, they generally do not exhibit a pattern of subcellular distribution that is consistent with the association. To further confirm the association between TSG101 and Bcr, we did dual staining in HeLa cells that express high endogenous levels of both proteins. In accordance with previous reports, TSG101 exhibited a granular cytoplasmic staining that would be consistent with an association with endosomal vesicles (36). Bcr exhibited an almost identical staining pattern, although an additional pool of immunoreactivity was identified in the nucleus. The specificity of the Bcr staining was confirmed using two independent antibodies (not shown). The nuclear localization was observed with both antibodies and is consistent with several previous reports, including one from our own laboratory (12, 14). When we merged the images of TSG101 and Bcr, we observed that much, but not all, of the cytoplasmic immunoreactivity was colocalized. When we did a similar colocalization experiment with Bcr and a marker for the late endosome (LAMP-I, H5G11, Santa Cruz Biotechnology), convergence of signal was not obvious. TSG101 was included in this analysis for comparison and also did not colocalize with LAMP-I. This suggests that the association between TSG101 and Bcr may be occurring at a site other than the surface of the late endosomal membrane (Fig. 4). A Vps4 mutant that blocks endosomal trafficking interferes with the subcellular distribution of Bcr. To determine whether Bcr can cycle onto the endosomal membrane, we used a dominantinhibitory mutant of Vps4 [Vps4(KQ), ref. 31]. Vps4 is a member of a family of ATPases that have been implicated in multiple cellular processes, including endosomal/vacuolar sorting (37). In both yeast and mammalian cells, Vps4 is a class E protein that is required for the disassembly of the ESCRT complexes at the completion of the sorting reaction (38–40). The dominant inhibitory mutant lacks ATPase activity and is thought to accumulate on the limiting membranes of the late endosome (30, 41). Importantly, other class E molecules, such as TSG101, have been shown to accumulate on the same membrane compartments in the presence of the mutant (36). For our analysis, HeLa cells were transiently transfected with Bcr or TSG101, and either GFP-Vps4(KQ) or GFP-Vps4(WT) (see Fig. 5). GFP alone was also included in the analysis for comparison. There seemed to be no obvious changes in overall cellular morphology in response to expression of either GFP-Vps4(KQ) or GFP-Vps4(WT) (Fig. 5). As described previously, GFP-Vps4(WT) exhibited a diffuse staining throughout the cell, whereas GFPVps4(KQ) accumulated in clustered cytoplasmic foci (30). In the presence of the mutant, both TSG101 and Bcr also accumulated in

6253

Cancer Res 2006; 66: (12). June 15, 2006


Cancer Research

Figure 4. TSG101 and Bcr exhibit a similar subcellular distribution. HeLa cells were examined by indirect immunofluorescence for the cellular distribution of endogenous TSG101, Bcr, and LAMP-1. Cells were viewed with a Zeiss Axiovert 200M microscope fitted with an ApoTome Imaging system. Image stacks in the axial direction were acquired and all images shown are from a representative axial plane. Cells shown are typical of >50 cells that were examined for the expression of the indicated combination of proteins.

strongly staining cytoplasmic foci that were not apparent in the presence of GFP alone or wild-type Vps4. When we merged the images, a limited, but reproducible convergence of signal between hVps4(KQ) and either Bcr (Fig. 5A) or TSG01 (Fig. 5B) was observed. The level of convergence that is shown is representative and is observed in 75% of all cells examined. Although these observations suggest that both Bcr and TSG101 may be accumulating on intracellular membranes in response to Vps4 expression, they both seem to have a distribution that is, for the most part, discrete from the Vps4 mutant. Bcr is required for EGFR degradation. The interaction of Bcr with components of the ESCRT-I complex suggests that it may function in the sorting machinery of the endosome. It has been previously shown that components of ESCRT-I mediate a sorting pathway that regulates the internalization and lysosome-mediated degradation of growth factor receptors (27, 35, 42, 43). For example, interference with TSG101 in HeLa cells blocks EGFmediated receptor turnover and promotes increased receptor recycling to the plasma membrane (27). To determine whether receptor turnover also requires Bcr function, we suppressed Bcr expression in HeLa cells and then monitored the degradation of the receptor in response to EGF stimulation (Fig. 6). TSG101 was included as a positive control because it has been shown previously that siRNA-mediated suppression of TSG101 in these cells causes the accumulation of the receptor at the plasma membrane (27). Consistent with previous observations, we observed an accumulation of the receptor in EGF-stimulated cells in which TSG101 expression was suppressed by siRNA. A similar accumulation of the receptor was observed when Bcr expression was suppressed. No accumulation was noted when cells were treated with scrambled control siRNAs. This observation implies overlapping functions for Bcr and TSG101 in HeLa cells, and suggests that each may be required for endosome-mediated receptor internalization.

Cancer Res 2006; 66: (12). June 15, 2006

Discussion Previous studies that describe putative binding partners for Bcr have provided relatively little information about its cytoplasmic function. The NH2-terminus of Bcr interacts with and phosphorylates at least five isoforms of the 14-3-3 proteins (44). Because different members of the 14-3-3 family have been implicated in a variety of cytoplasmic activities, including apoptosis, signal transduction, trafficking, and secretion, the meaning of these interactions remains unclear. Bcr contains a COOH-terminal PDZbinding domain through which it is reported to interact with several cytoplasmic proteins, including AF-6, PDZK1, and Mint3 (45, 46). AF-6 has been colocalized with tight junctions and adhesion junctions and is thought to mediate the interaction between the plasma membrane and the actin cytoskeleton (46). PDZ-K1 is also a plasma membrane–associated protein that is predominantly found in association with apical membrane proteins in polarized epithelia (45). Although these associations are consistent with the presence of Rho-modulating domains within Bcr, most Bcr immunoreactivity is found on intracellular membranes and none is readily observed on the plasma membrane. This suggests that these associations are either transient or highly cell specific. Interestingly, the binding partner for Bcr that has the most similar cellular distribution is Mint3 (45). Mint3 is found primarily in the Golgi compartment where it has been implicated in protein processing and vesicular trafficking in the distal secretory pathway. Mint3 has a punctate cytoplasmic staining that partially overlaps with Bcr and it has been previously proposed that Bcr may have a role in cellular trafficking (45). In the current study, we have identified two structurally unrelated subunits of the endosomal sorting machinery as binding partners for Bcr. Bcr interacts with TSG101 and Vps28 in HeLa and 293T cells, and Bcr and TSG101 exhibit a similar subcellular distribution in these cell types. The association of Bcr with both TSG101 and Vps28 suggests a very specific role for Bcr in the

6254




endosomal pathway. In eukaryotic cells, transmembrane proteins can be removed from the cell surface by endocytosis, followed by lysosome-mediated degradation. On the limiting membrane of the multivesicular body, proteins that are being transported in this pathway are sorted into vesicles that first bud into the lumen of the compartment and then are delivered to the lysosome (34). Although the mechanism by which cargo is sorted as it passes through the multivesicular body is poorly understood, there is evidence to suggest that monoubiquitination is used as the flag

that targets proteins into this pathway (28). The formation of the multivesicular body and the sorting of ubiquitinated proteins is controlled by a large collection of proteins that was originally identified in yeast as the class E Vps (reviewed in ref. 47). Because both TSG101 and Vps28 are class E Vps proteins in yeast and mammalian cells, we propose that Bcr may also be a member of this class. Vps proteins are organized into three discrete complexes, termed ESCRT-I, ESCRT-II, and ESCRT-III, that are recruited from the

Figure 5. A dominant inhibitory mutant of Vps4 alters the cellular distribution of Bcr and TSG101. HeLa cells were transiently transfected with GFP alone or GFP fused to wild-type or dominant-inhibitory Vps4. Cells were examined at 48 hours by indirect immunofluorescence for the cellular distribution of Vps4 and Bcr (A ), or Vps4 and TSG101 (B). Cells were viewed with a Zeiss Axiovert 200M microscope fitted with an ApoTome Imaging system. Image stacks in the axial direction were acquired and all images shown are from a representative axial plane. Cells shown are typical of >50 cells that were examined for the expression of the indicated combination of proteins.


6255

Cancer Res 2006; 66: (12). June 15, 2006


Cancer Research

Figure 6. Bcr is required for ligand-mediated turnover of the EGFR. A, siRNAs were used to transiently suppress the expression Bcr or TSG101 in HeLa cells. Cells were then treated for 90 minutes with 150 ng/mL EGF and then lysates were collected and examined for expression of EGFR (anti-EGFR), Bcr (anti-Bcr), and TSG101 (anti-TSG101). B, relative levels of receptor expression were determined using an Odyssey IR imager and expressed as integrated intensity units. Average of three independent experiments.

cytoplasm to act sequentially at the surface of the multivesicular body (28, 39, 48). ESCRT-I is thought to recognize ubiquitinated proteins and trigger the assembly and activation of ESCRT-II. ESCRT-II, in turn, is required for the recruitment of ESCRT-III whose role is to concentrate the multivesicular body cargo for internalization and remove the ubquitination tag. Once sorting is completed, Vps4 binds to ESCRT-III and disassembles the complex (37, 38). The Vps proteins are highly conserved from yeast to mammals, and mutations in the mammalian orthologues exhibit phenotypes consistent with defects in endosomal trafficking. For most components of the ESCRT complexes, one or more mammalian counterparts have now been identified. Because the mammalian counterpart of the ESCRT-I complex contains both TSG101 and Vps28 (reviewed in ref. 34), our binding data suggests that Bcr is either a component or a regulator of this complex. Whether Bcr also binds directly to Vps28 in this complex is unclear. Because it has been shown previously that TSG101 interacts with mammalian Vps28 in HeLa cells, our immunoprecipitations with the Vps28 antibody may be detecting Bcr indirectly through its association with TSG101 (36). To resolve this issue, in vitro studies are currently under way to examine the ability of Bcr to directly interact with the various components of the mammalian ESCRT complexes. However, regardless of whether the interaction with Vps28 is direct, the existence of a complex containing Bcr, TSG101, and Vps28 in mammalian cells strongly suggests that Bcr is a bona fide component of the mammalian endosomal sorting machinery. Support for a role for Bcr in endosomal sorting also comes from our functional studies. In mammalian cells, the ESCRT-I complex is required for the down-regulation of growth factor receptors (27, 35, 42, 43). The complex is recruited from the cytoplasm to the

References 1. Maru Y. Molecular biology of chronic myeloid leukemia. Int J Hematol 2001;73:308–22. 2. Nowell PC, Hungerford DA. The etiology of leukemia: some comments on current studies. Semin Hematol 1966;3:114–21.

Cancer Res 2006; 66: (12). June 15, 2006

multivesicular body in response to growth factor stimulation where it identifies and routes the internalized receptors to the lysosome. TSG101 binds directly to ubiquitinated proteins and thus may serve to recognize the targeted receptors (28, 35). Because the interaction between TSG101 and Bcr does not require ubiquitination, it is unlikely that Bcr is simply a preferred endosomal cargo in this pathway that is being recognized by TSG101. Once the receptor internalization is complete, the ESCRT complexes are disassembled from the limiting membrane of the multivesicular body in response to the action of the Vps4 ATPase. Interference of TSG101 through siRNA or Vps28 by neutralizing antibodies causes an accumulation of the EGFR on the cell surface in response to treatment with EGF (35, 49). Similarly, we have observed that siRNAs directed against both TSG101 and Bcr cause equivalent accumulations of EGFR on the cell surface. Thus, the association of Bcr with a complex that includes TSG101 and Vps28 seems to have functional relevance with regard to the efficient turnover of growth factor receptors. Despite the fact that our binding studies suggest that Bcr interacts with ESCRT complexes, we do not observe a good convergence of signal between either TSG101 or Bcr, and a marker for the late endosomal membrane (LAMP-I). Because ESCRT complexes cycle between the cytoplasm and the limiting membrane of the late endosome, we assume that the majority of the TSG101-Bcr complexes in HeLa cells are in the soluble, resting state. Our observation that the dominant-inhibitory Vps4 mutant causes both TSG101 and Bcr to accumulate on densely staining cytoplasmic foci suggests that at least some of these complexes are being cycled onto membranes and can be trapped by the action of the mutant before membrane dissociation. Our observation that Bcr interacts with subunits of the ESCRT complexes raises the possibility that endosomal trafficking may be impaired in CML. Although the interaction between Bcr and TSG101 is still observed in K562 cells that express p210 Bcr-Abl, we have also observed that incomplete suppression of either Bcr or TSG101 by siRNA is sufficient to impair growth factor receptor turnover in HeLa cells. Thus, the reduced dosage of Bcr that occurs in Philadelphia chromosome–positive cells may in fact be sufficient to cause impairment in trafficking. Alternatively, p210 Bcr-Abl may exhibit a dominant-inhibitory effect with respect to the normal endosomal function of Bcr. Although the TSG101 docking site is lost in p210 Bcr-Abl, the NH2 terminus of Bcr may interact with additional components of the sorting complex, such as Vps28. Thus, p210 Bcr-Abl may be recruited to the endosome and interfere with the normal function of Bcr. Both of these possibilities are currently under investigation.

Acknowledgments Received 2/10/2006; revised 3/29/2006; accepted 4/12/2006. Grant support: NIH grants R01 CA097066 (I. Whitehead), RO1 AG04821 (H. Ozer), F31 CA117049 (O. Olabisi), and F32 CA113049 (E. Kostenko). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

3. O’Dwyer ME, Druker BJ. Chronic myelogenous leukaemia—new therapeutic principles. J Intern Med 2001; 250:3–9. 4. Lugo TG, Pendergast AM, Muller AJ, Witte ON. Tyrosine kinase activity and transformation potency of bcr-abl oncogene products. Science 1990;247:1079–82. 5. Pendergast AM, Quilliam LA, Cripe LD, et al. BCR-

6256

ABL-induced oncogenesis is mediated by direct interaction with the SH2 domain of the GRB-2 adaptor protein. Cell 1993;75:175–85. 6. McWhirter JR, Galasso DL, Wang JY. A coiled-coil oligomerization domain of Bcr is essential for the transforming function of Bcr-Abl oncoproteins. Mol Cell Biol 1993;13:7587–95.



Bcr Interacts with the ESCRT-I Complex 7. Maru Y, Afar DE, Witte ON, Shibuya M. The dimerization property of glutathione S -transferase partially reactivates Bcr-Abl lacking the oligomerization domain. J Biol Chem 1996;271:15353–7. 8. Muller AJ, Young JC, Pendergast AM, et al. BCR first exon sequences specifically activate the BCR/ABL tyrosine kinase oncogene of Philadelphia chromosome-positive human leukemias. Mol Cell Biol 1991;11: 1785–92. 9. Whitehead IP, Campbell S, Rossman KL, Der CJ. Dbl family proteins. Biochim Biophys Acta 1997;1332:F1–23. 10. Nalefski EA, Falke JJ. The C2 domain calcium-binding motif: structural and functional diversity. Protein Sci 1996;5:2375–90. 11. Chuang TH, Xu X, Kaartinen V, Heisterkamp N, Groffen J, Bokoch GM. Abr and Bcr are multifunctional regulators of the Rho GTP-binding protein family. Proc Natl Acad Sci U S A 1995;92:10282–6. 12. Wetzler M, Talpaz M, Van Etten RA, Hirsh-Ginsberg C, Beran M, Kurzrock R. Subcellular localization of Bcr, Abl, and Bcr-Abl proteins in normal and leukemic cells and correlation of expression with myeloid differentiation. J Clin Invest 1993;92:1925–39. 13. Laurent E, Talpaz M, Wetzler M, Kurzrock R. Cytoplasmic and nuclear localization of the 130 and 160 kDa Bcr proteins. Leukemia 2000;14:1892–7. 14. Mahon GM, Wang Y, Korus M, et al. The c-Myc oncoprotein interacts with Bcr. Curr Biol 2003;13: 437–41. 15. Voncken JW, Baram TZ, Gonzales-Gomez II, et al. Abnormal stress response and increased fighting behavior in mice lacking the bcr gene product. Int J Mol Med 1998;2:577–83. 16. Kaartinen V, Nagy A, Gonzalez-Gomez I, Groffen J, Heisterkamp N. Vestibular dysgenesis in mice lacking Abr and Bcr Cdc42/RacGAPs. Dev Dyn 2002;223:517–25. 17. Kaartinen V, Gonzalez-Gomez I, Voncken JW, et al. Abnormal function of astroglia lacking Abr and Bcr RacGAPs. Development 2001;128:4217–27. 18. Voncken JW, van Schaick H, Kaartinen V, et al. Increased neutrophil respiratory burst in bcr-null mutants. Cell 1995;80:719–28. 19. Li L, Cohen SN. Tsg101: a novel tumor susceptibility gene isolated by controlled homozygous functional knockout of allelic loci in mammalian cells. Cell 1996; 85:319–29. 20. Feng GH, Lih CJ, Cohen SN. TSG101 protein steadystate level is regulated posttranslationally by an evolutionarily conserved COOH-terminal sequence. Cancer Res 2000;60:1736–41. 21. Xie W, Li L, Cohen SN. Cell cycle-dependent subcellular localization of the TSG101 protein and mitotic and nuclear abnormalities associated with


TSG101 deficiency. Proc Natl Acad Sci U S A 1998;95: 1595–600. 22. Ruland J, Sirard C, Elia A, et al. p53 accumulation, defective cell proliferation, and early embryonic lethality in mice lacking tsg101. Proc Natl Acad Sci U S A 2001;98: 1859–64. 23. Wagner KU, Krempler A, Qi Y, et al. Tsg101 is essential for cell growth, proliferation, and cell survival of embryonic and adult tissues. Mol Cell Biol 2003;23: 150–62. 24. Watanabe M, Yanagi Y, Masuhiro Y, et al. A putative tumor suppressor, TSG101, acts as a transcriptional suppressor through its coiled-coil domain. Biochem Biophys Res Commun 1998;245:900–5. 25. Li L, Liao J, Ruland J, Mak TW, Cohen SN. A TSG101/ MDM2 regulatory loop modulates MDM2 degradation and MDM2/p53 feedback control. Proc Natl Acad Sci U S A 2001;98:1619–24. 26. Zhong Q, Chen Y, Jones D, Lee WH. Perturbation of TSG101 protein affects cell cycle progression. Cancer Res 1998;58:2699–702. 27. Babst M, Odorizzi G, Estepa EJ, Emr SD. Mammalian tumor susceptibility gene 101 (TSG101) and the yeast homologue, Vps23p, both function in late endosomal trafficking. Traffic 2000;1:248–58. 28. Katzmann DJ, Babst M, Emr SD. Ubiquitin-dependent sorting into the multivesicular body pathway requires the function of a conserved endosomal protein sorting complex, ESCRT-I. Cell 2001;106:145–55. 29. Whitehead I, Kirk H, Tognon C, Trigo-Gonzalez G, Kay R. Expression cloning of lfc, a novel oncogene with structural similarities to guanine nucleotide exchange factors and to the regulatory region of protein kinase C. J Biol Chem 1995;270:18388–95. 30. Bishop N, Woodman P. ATPase-defective mammalian VPS4 localizes to aberrant endosomes and impairs cholesterol trafficking. Mol Biol Cell 2000;11:227–39. 31. Chowdary DR, Dermody JJ, Jha KK, Ozer HL. Accumulation of p53 in a mutant cell line defective in the ubiquitin pathway. Mol Cell Biol 1994;14:1997–2003. 32. Whitehead IP, Lambert QT, Glaven JA, et al. Dependence of Dbl and Dbs transformation on MEK and NF-nB activation. Mol Cell Biol 1999;19:7759–70. 33. Rossman KL, Cheng L, Mahon GM, et al. Multifunctional roles for the PH domain of Dbs in regulating Rho GTPase activation. J Biol Chem 2003;278:18393–400. 34. Babst M. A protein’s final ESCRT. Traffic 2005;6:2–9. 35. Bishop N, Horman A, Woodman P. Mammalian class E vps proteins recognize ubiquitin and act in the removal of endosomal protein-ubiquitin conjugates. J Cell Biol 2002;157:91–101. 36. Bishop N, Woodman P. TSG101/mammalian VPS23 and mammalian VPS28 interact directly and are

6257

recruited to VPS4-induced endosomes. J Biol Chem 2001;276:11735–42. 37. Babst M, Sato TK, Banta LM, Emr SD. Endosomal transport function in yeast requires a novel AAA-type ATPase, Vps4p. EMBO J 1997;16:1820–31. 38. Babst M, Wendland B, Estepa EJ, Emr SD. The Vps4p AAA ATPase regulates membrane association of a Vps protein complex required for normal endosome function. EMBO J 1998;17:2982–93. 39. Babst M, Katzmann DJ, Snyder WB, Wendland B, Emr SD. Endosome-associated complex, ESCRT-II, recruits transport machinery for protein sorting at the multivesicular body. Dev Cell 2002;3:283–9. 40. Odorizzi G, Katzmann DJ, Babst M, Audhya A, Emr SD. Bro1 is an endosome-associated protein that functions in the MVB pathway in Saccharomyces cerevisiae . J Cell Sci 2003;116:1893–903. 41. Yoshimori T, Yamagata F, Yamamoto A, et al. The mouse SKD1, a homologue of yeast Vps4p, is required for normal endosomal trafficking and morphology in mammalian cells. Mol Biol Cell 2000;11:747–63. 42. Bache KG, Slagsvold T, Cabezas A, Rosendal KR, Raiborg C, Stenmark H. The growth-regulatory protein HCRP1/hVps37A is a subunit of mammalian ESCRT-I and mediates receptor down-regulation. Mol Biol Cell 2004;15:4337–46. 43. Lu Q, Hope LW, Brasch M, Reinhard C, Cohen SN. TSG101 interaction with HRS mediates endosomal trafficking and receptor down-regulation. Proc Natl Acad Sci U S A 2003;100:7626–31. 44. Reuther GW, Fu H, Cripe LD, Collier RJ, Pendergast AM. Association of the protein kinases c-Bcr and BcrAbl with proteins of the 14-3-3 family. Science 1994;266: 129–33. 45. Malmberg EK, Andersson CX, Gentzsch M, et al. Bcr (breakpoint cluster region) protein binds to PDZdomains of scaffold protein PDZK1 and vesicle coat protein Mint3. J Cell Sci 2004;117:5535–41. 46. Radziwill G, Erdmann RA, Margelisch U, Moelling K. The Bcr kinase downregulates Ras signaling by phosphorylating AF-6 and binding to its PDZ domain. Mol Cell Biol 2003;23:4663–72. 47. Katzmann DJ, Odorizzi G, Emr SD. Receptor downregulation and multivesicular-body sorting. Nat Rev Mol Cell Biol 2002;3:893–905. 48. Babst M, Katzmann DJ, Estepa-Sabal EJ, Meerloo T, Emr SD. Escrt-III: an endosome-associated heterooligomeric protein complex required for mvb sorting. Dev Cell 2002;3:271–82. 49. Doyotte A, Russell MR, Hopkins CR, Woodman PG. Depletion of TSG101 forms a mammalian ‘‘Class E’’ compartment: a multicisternal early endosome with multiple sorting defects. J Cell Sci 2005;118:3003–17.

Cancer Res 2006; 66: (12). June 15, 2006
