Skip to main content
NCBI home page
Neoplasia (New York, N.Y.) logoLink to Neoplasia (New York, N.Y.)
. 2011 Jul;13(7):570–578. doi: 10.1593/neo.11632

Regulation of EGFR Protein Stability by the HECT-type Ubiquitin Ligase SMURF21,2

Dipankar Ray *, Aarif Ahsan *,3, Abigail Helman *,3, Guoan Chen †, Ashok Hegde *, Susmita Ramanand Gurjar *, Lili Zhao ‡, Hiroaki Kiyokawa §, David G Beer †, Theodore S Lawrence *, Mukesh K Nyati *
PMCID: PMC3132843  PMID: 21750651

Abstract

Epidermal growth factor receptor (EGFR) is overexpressed in a variety of epithelial tumors and is considered to be an important therapeutic target. Although gene amplification is responsible for EGFR overexpression in certain human malignancies including lung and head and neck cancers, additional molecular mechanisms are likely. Here, we report a novel interaction of EGFR with an HECT-type ubiquitin ligase SMURF2, which can ubiquitinate, but stabilize EGFR by protecting it from c-Cbl-mediated degradation. Conversely, small interfering RNA (siRNA)-mediated knockdown of SMURF2 destabilized EGFR, induced an autophagic response and reduced the clonogenic survival of EGFR-expressing cancer cell lines, with minimal effects on EGFR-negative cancer cells, normal fibroblasts, and normal epithelial cells. UMSCC74B head and neck squamous cancer cells, which form aggressive tumors in nudemice, significantly lost in vivo tumor-forming ability on siRNA-mediated SMURF2 knockdown. Gene expressionmicroarray data from 443 lung adenocarcinoma patients, and tissue microarray data from 67 such patients, showed a strong correlation of expression between EGFR and SMURF2 at the messenger RNA and protein levels, respectively. Our findings suggest that SMURF2-mediated protective ubiquitination of EGFR may be responsible for EGFR overexpression in certain tumors and support targeting SMURF2-EGFR interaction as a novel therapeutic approach in treating EGFR-addicted tumors.

Introduction

Epidermal growth factor receptor (EGFR) overexpression is a common occurrence in human malignancies of epithelial origin including lung and head and neck cancers and has been correlated with poor prognosis [1–3], and small molecules directed at EGFR kinase activity have provided definite but limited success [4–6]. We have shown that therapies causing EGFR degradation, as opposed to simple inhibition of its kinase activity, are far more potent both in killing cancer cells and in sensitizing tumor cells to chemotherapy than simply inhibiting EGFR kinase activity [7–9]. We, therefore, hypothesized that, instead of simply inhibiting EGFR kinase function, degrading EGFR may improve the clinical outcome of already-existing strategies to control tumor cell growth. Thus, it is important to better understand the molecular regulators involved in maintaining EGFR protein stability.

In many cases, protein stability of important oncogenes or tumor suppressor genes is controlled by multiple ubiquitin ligases. For example, the stability of the checkpoint regulator and oncogene, CDC25A, is maintained by at least two different ubiquitin ligases, SCFβ-TrCP and APCCdh1 [10–12]. Similarly, in the case of insulin-like growth factor receptor I (IGF-IR), Mdm2, Nedd4, and c-Cbl act as ubiquitin ligases [13,14]. The RING-type ubiquitin ligase, Cbl, is the major ligase that catalyzes EGFR ubiquitination [15]; however, because EGFR is an important regulator of various cellular functions, we hypothesized that there may be multiple ubiquitin ligases responsible for controlling its stability.

One such promising ligase to investigate in this regard is the HECT-type ubiquitin ligase, Smad ubiquitination regulatory factor 2 (SMURF2) [16,17]. Like other HECT-type ubiquitin ligases, SMURF2 has the ability to ubiquitinate and degrade some proteins, e.g., the TGF-β receptor I [18] and associated SMAD proteins [19,20], but also has a unique ability to protect other substrates, e.g., spindle assembly checkpoint protein, MAD2 [21] and NEDD9/HEF1 [22]. The ubiquitin ligase activity of SMURF2 is dependent on the presence of a cofactor, SMAD7, which helps to recruit the E2 enzyme, UbcH7 to the HECT domain of SMURF2 [23]. It has been reported that Smad7 overexpression can accelerate the oncogenic Ras-mediated tumor progression in a mouse squamous cell carcinoma model not only by inhibiting TGF-β signaling but also by upregulating EGFR activity [24]. SMURF2 and EGFR are known to be overexpressed in various squamous cell carcinomas including esophagus [25–27], and Smad7 overexpression activates EGFR signaling [24]. We have also identified probable SMURF2-interacting motifs in EGFR, and SMURF2 has the unique ability either to protect or to degrade key cellular proteins. Therefore, we hypothesized that EGFR may be a novel substrate for SMURF2. Indeed, we found that SMURF2 is capable of directly binding to EGFR; that knocking down SMURF2 led to ligand-dependent EGFR ubiquitination and degradation, induced autophagic response, and reduced clonogenic cell survival; and that SMURF2 targeting affected the tumor-forming ability of UMSCC74B cells in nude mice. When we discovered that SMURF2 could have these EGFR-directed effects in vitro and in vivo, we decided to evaluate whether the presence of SMURF2 and EGFR was correlated in patient specimens.

Materials and Methods

Reagents

A monoclonal antibody against human SMURF2 has been described previously [18]. Rabbit polyclonal SMURF2 antibody was purchased from Upstate Biotechnology (Lake Placid, NY) and c-Cbl, glyceraldehyde-3-phosphate dehydrogenase, AKT, phospho-AKT (Ser473), STAT3, and phospho-STAT3 (Tyr705) antibodies were purchased from Cell Signaling Technology (Danvers, MA). EGFR (sc-03), TGF-β receptor I (sc-398), and ubiquitin (P4D1) antibodies were acquired from Santa Cruz Biotechnology (Santa Cruz, CA). Sepharose conjugated with FLAG (M2) antibody, protease inhibitor cocktail, EGF, and glutathione agarose beads were obtained from Sigma (St Louis, MO). Anti-chicken cytokeratin 8 antibody was purchased from Novus Biologicals (Littleton, CO). Tetra-His antibody was purchased from Qiagen (Chatsworth, CA). Alexa 488 antimouse IgG, Alexa 488 antirabbit, Alexa 594 antirabbit, and Alexa 555 antichicken IgG were purchased from Invitrogen (Carlsbad, CA). SMURF2 small interfering RNA(siRNA) used in this study has been described previously [21] and was purchased from Thermo Fisher Scientific (Lafayette, CO).

Cell Cultures

Chinese hamster ovarian (CHO), SW620, A549, Panc1, BxPC3, MiaPaca2, NIH3T3, and Het1A cell lines were purchased from the American Type Culture Collection (Manassas, VA). The human head and neck squamous cell carcinoma (HNSCC) cell lines UMSCC1 and UMSCC74B were kindly provided by Dr Thomas E. Carey (University of Michigan, Ann Arbor, MI). NCI-H1975 cell line was a gift from Dr Jeffrey Engelman (Massachusetts General Hospital, Boston, MA). Stably transfected HeLa-LC3B-GFP-expressing cells were maintained in puromycin selection medium. All cell lines were grown in either RPMI 1640 or Dulbecco modified Eagle medium supplemented with 10% fetal calf serum (Sigma). CHO cells were transiently transfected with the constructs using Lipofectamine (Invitrogen) according to the instructions of the manufacturer. siRNA transfections were performed using Lipofectamine RNAi-max (Invitrogen) following the manufacturer's protocol.

Immunofluorescence Microscopy

Immunostaining was performed as previously described [21]. In brief, fixed cells were washed with PBS, permeabilized with 0.2% Triton X-100 on ice for 5 minutes, and stained with primary antibodies at 4°C for 16 hours. Washed slides were then incubated with fluorescence-conjugated secondary antibodies at 4°C for 1 hour, stained with DAPI (Invitrogen), and photographed using an Olympus DP70 camera fitted to an Olympus 1X-71 microscope (Olympus America, Inc, Melville, NY). Images were captured at either x20 or x60 magnifications and processed in Adobe PhotoShop CS4 by removing the unsharp mask.

Protein Analyses

For immunoblot analysis or immunoprecipitation, cells were lysed by sonication in lysis buffer as described previously [28]. Immunoprecipitation of EGFR was performed as previously described [29], and immunoprecipitation for FLAG-tagged SMURF2 was performed using Affi-FLAG(M2) beads as previously described [30].

GST Pull-down Assay

GST-SMURF2 was purified from bacteria as previously described [30], and 1 µg of fusion protein still attached to agarose beads was equilibrated in 0.5x Superdex buffer (1x Superdex buffer: 25 mM HEPES, pH 7.5, 12.5 mM MgCl2, 10 µM ZnSO4, 150 mM KCl, 20% glycerol, 0.1% Nonidet P-40, and 1 mM EDTA) for 30 minutes at 4°C and then washed three times with 0.5x Superdex buffer. Beads were then incubated with either about 200 or 600 ng of purified His-tagged EGFR protein (Creative BioMart, New York, NY) was then added to the washed beads and incubated overnight at 4°C. The beads were washed three times using 0.5 Superdex buffer and boiled in Laemmli buffer, and the bound SMURF2-EGFR complex was immunodetected after immunoblot analysis with SMURF2 and Tetra-His (for EGFR) antibodies.

Reverse Transcription-Polymerase Chain Reaction

Semiquantitative reverse transcription (RT)-polymerase chain reaction (PCR) was performed as described previously [31]. In brief, total RNA was isolated using Qiagen RNeasy mini kit (Qiagen, Inc, Valencia, CA) from cells according to the manufacturer's protocol. Total RNA (1 µg) was subjected to RT reaction with the use of the SuperScript first-strand synthesis system (Invitrogen). After the RT reaction, RNase H was added to remove the RNA template from the reaction mixture. Subsequently, PCR was performed in a total volume of 25 µl with 1 µl of the RT product. The primers used for human SMURF2 messenger RNA (mRNA) were 5′ tagccctggcagacctctta 3′ and 5′ aatacacctggccttgttgc 3′ for amplification of a 218-bp product and for human EGFR mRNA were 5′ cagcgctaccttgtcattca 3′ and 5′ tgcactcagagagctcagga 3′ for a 195-bp product. The primers used for coamplification (238 bp) of the control glyceraldehyde-3-phosphate dehydrogenase mRNA were 5′ gagtcaacggatttggtcgt 3′ and 5′ ttgattttggagggatctcg 3′. The reaction was performed in an Eppendorf PCR machine, at 94°C for 30 seconds, followed by 30 cycles of 94°C for 30 seconds, 54°C for 45 seconds, and 72°C for 1 minute. Amplified DNAs were analyzed by agarose gel electrophoresis, and signals in ethidium bromide-stained gels were quantified using the EDAS-290 imaging system (Kodak, Rochester, NY).

Clonogenic Cell Survival Assay

Clonogenic assays were performed using standard techniques [32]. The fraction surviving for each treatment was normalized to the survival of the control cells. The effects of SMURF2 siRNA-induced clonogenic death were calculated by comparing the survival fraction with that of the control siRNA-treated group.

In Vivo Tumor Growth Studies

Mice were handled according to the established procedures of the University of Michigan's Laboratory Animals Maintenance Manual. UMSCC74B cells were either left untreated or treated with either control or SMURF2 siRNA as previously described. To generate tumor xenografts, 50,000 UMSCC74B cells from each of the three subgroups (untreated, control siRNA treated, or SMURF2 siRNA treated) were transplanted subcutaneously at four different locations into athymic nude Foxn1nu mice (Harlan Laboratories, Indianapolis, IN). For the untreated and control siRNA-treated groups, three animals were injected in each group, whereas in SMURF2 siRNA group, we used four mice. Thus, in case of untreated and control siRNA-treated groups, we have followed the tumor growth of 12 individual tumors, whereas we followed 16 tumors in the SMURF2 siRNA-treated group. Tumor latency was calculated based on the post-injection day and when a palpable tumor was first detected. Animals were killed when the largest tumor diameter reached to the protocol approved size of 2 cm. The length and width of the tumors were measured every other day, and the tumor volume was calculated using the following equation: 0.5 x length x width2 [33]. Relative tumor volume was calculated by normalizing to the tumor size when first detected.

Tissue Microarray Immunofluorescence Staining and Evaluation by Automated Quantitative Analysis

The lung adenocarcinoma tissue microarray (TMA) used in the study was constructed as previously described [34]. The TMA slides were deparaffinized in xylene and rehydrated using serial ethanol dilutions, and antigen site unmasking was performed by immersing slides in 100-nM citrate buffer for 20 minutes at high pressure and temperature inside a pressure cooker. Slides were then washed in TBS, blocked for 1 hour, and incubated in primary antibody at 4°C overnight. Slides were then washed again in TBS, incubated in secondary antibody for 1 hour at room temperature, rewashed, and prepared with a coverglass after a drop of ProLong Gold antifade reagent with 4′, 6-diamidino-2-phenylindole (Molecular Probes, Eugene, OR) was added to each sample. Fluorescence images were acquired using an Olympus DP70 camera fitted in an Olympus 1X-71 microscope for representative pictures, and automated quantitative analysis (AQUA) was used for automated image acquisition and analysis as previously described [35], supported by University of Michigan Comprehensive Cancer Center (UMSCC) tissue core.

Statistics

Pearson correlation was used to measure the degree of association between two markers. Pearson correlation coefficient (r) was calculated, together with the P value (null hypothesis is that r is in fact zero). Results are presented as mean ± SEM of at least three experiments. A significance level threshold of P < .05 was used.

Results

SMURF2 Directly Binds, Ubiquitinates, and Protects EGFR from c-Cbl-Mediated Degradation

We began by examining if there were physical and molecular interactions between EGFR and SMURF2. To investigate such interactions, we used two different experimental systems: 1) CHO cells to overexpress EGFR and/or SMURF2 (because CHO does not express any EGFR) and 2) UMSCC1 head and neck squamous cancer cell lines that overexpress EGFR and SMURF2 and depend on EGFR for their survival. To determine whether SMURF2 can interact with EGFR, we performed immunoprecipitation using EGFR-specific antibodies from cell lysates isolated from either CHO cells overexpressing EGFR and/or SMURF2 or UMSCC1 cells. In both systems, we detected interaction between EGFR and SMURF2 (Figure 1A). Similarly, immunoprecipitation studies using FLAG antibodies to pull down overexpressed FLAG-tagged SMURF2 also detected EGFR-SMURF2 immunocomplex (Figure 1B). Because immunoprecipitation under nondenaturing condition does not rule out a direct or indirect interaction between two proteins, we used a classic GST pull-down assay using bacterially purified GST-SMURF2 and His-tagged EGFR purified from HEK-293 cells. As shown in Figure 1C, we detected a direct physical interaction between EGFR and SMURF2.

Figure 1.

Figure 1

Open in a new tab

SMURF2 interacts with, ubiquitinates and protects EGFR from c-Cbl-mediated down-regulation. (A) EGFR was immunoprecipitated (IP) from CHO cells cotransfected with EGFR and SMURF2 (left panel) or from UMSCC1 cells (right panel) and immunoblotted (IB) with indicated antibodies. NRS indicates normal rabbit serum. MW marker indicated on the side. (B) Immunoprecipitation was performed using Affi-FLAG Sepharose beads to pull-down FLAG-tagged SMURF2 from CHO cells cotransfected with EGFR and SMURF2, and immunoblot analysis was performed using indicated antibodies. (C) GST-SMURF2 agarose beads were incubated with purified Histagged EGFR protein, and a GST pull-down assay was performed as described in Materials and Methods. Immunoblot analyses were performed using indicated antibodies. (D) EGFR was either expressed alone or coexpressed either with WT or ligase dead (CA) mutant of SMURF2 in CHO cells. Direct IBs as well as EGFR IP followed by Ub IB were performed on cell lysates 24 hours after transfection. (E) EGFR was co expressed either with c-Cbl or SMURF2 in various combinations as indicated. Six hours after transfection, medium was replaced with 10% FBS containing complete medium, and 24 hours after transfection, cells were either left untreated (-) or treated (+) with 10 ng/ml EGF for 6 hours. Immunoblot analyses were performed using indicated antibodies.

To determine whether EGFR is ubiquitinated by SMURF2, we overexpressed EGFR in CHO cells in the presence or absence of either catalytically active wild-type (WT) or a catalytically dead C716A (CA) mutant of SMURF2. EGFR was immunoprecipitated using specific antibody, and immunoblot analysis was performed using antiubiquitin antibody. As shown in Figure 1D (lower panel), WT SMURF2 significantly increased the ubiquitination of EGFR, whereas SMURF2 (CA) mutant had no effect, indicating that SMURF2 is an ubiquitin ligase for EGFR. Interestingly, coexpression of EGFR and WT SMURF2 significantly enhanced the steady-state level of EGFR, whereas SMURF2 (CA) mutant had minimal effects (Figure 1D, upper panel). These data indicate that SMURF2 can interact with EGFR leading to its ubiquitination; however, unlike c-Cbl-mediated EGFR ubiquitination, SMURF2-dependent ubiquitination is protective for EGFR. Because of its autoubiquitinating ability, SMURF2 (CA) is more highly expressed than SMURF2 (WT) [36]. We believe that the increased presence of SMURF2 (CA) (Figure 1D) is responsible for a modest alteration of EGFR steady-state levels, probably by counteracting c-Cbl binding.

Next we addressed the question of why SMURF2 (WT) overexpression increased EGFR steady-state levels. One hypothesis was that SMURF2-mediated ubiquitination may be counteracting Cbl's effect on EGFR, the only RING-type ubiquitin ligase reported to ubiquitinate and downregulate EGFR. To test this hypothesis, we cotransfected CHO cells with EGFR and c-Cbl either in the presence or in the absence of SMURF2 (WT) construct (Figure 1E). We found that SMURF2 overexpression abrogated c-Cbl-mediated ligand-induced down-regulation of EGFR. Addition of excess EGF, which also causes receptor internalization and degradation, was also significantly inhibited by SMURF2 (WT) overexpression. These experiments demonstrated a novel cooperative interaction between EGFR and SMURF2, which may be important for oncogenic EGFR overexpression in human cancers.

siRNA-Mediated Knockdown of SMURF2 Decreased EGFR Protein Levels with Enhanced Ubiquitination and Leads to Reduced Clonogenic Survival of EGFR-Dependent Cancer Cells

To better understand the physiological importance of the interactions reported above, we performed siRNA-mediated acute knockdown of SMURF2 in various tumor cell lines including lung (A549 and NCI-H1975), head and neck (UMSCC1, UMSCC74B), and pancreatic (Panc-1) cells and also in immortalized normal esophageal squamous epithelial cells (Het-1A). In all the cell lines tested, SMURF2 siRNA caused significant loss of endogenous EGFR protein levels within 48 hours of siRNA transfection (Figure 2A). However, SMURF2 siRNA had no effects on EGFR mRNA levels as exemplified in UMSCC1 cells (Figure 2B). To show the specificity of SMURF2 siRNA-mediated down-regulation of EGFR, the membrane was reprobed with a TGF-β receptor I antibody, which remained essentially unaltered wherever expressed (Figure 2A). These data indicate that SMURF2 plays a critical and potentially selective role in maintaining EGFR protein stability.

Figure 2.

Figure 2

Open in a new tab

siRNA-mediated SMURF2 down-regulation decreased EGFR protein levels and clonogenic cancer cell survival. (A) Different cancer (lung [A549, NCI-H1975], head and neck [UMSCC1 and 74B], and pancreatic [Panc1]) and a normal epithelial cell line (Het1A) were transfected either with 50 nM of control (C) or SMURF2 (S) siRNA. Cell lysates were prepared 48 hours after transfection, and immunoblot analyses were performed using different antibodies as indicated. (B) Total RNAs were isolated from UMSCC1 cells treated with either C or S siRNA and RT-PCR was performed as described previously [31]. (C) Different cancer (A549, NCI-H1975, UMSCC1, UMSCC74B, BxPC3, MiaPaca2, Panc1, and SW620) and normal cell lines (Het1A, NIH3T3, and CHO) were transfected with siRNA (C or S). Twenty-four hours after transfection, cells were trypsinized and replated to determine the clonogenic survival efficiency. Survival efficiency for untreated group was normalized to 1 to determine the survival fraction for the control and SMURF2 siRNA group for each cell line and presented as mean ± SEM from three independent experiments.

Because EGFR down-regulation affects the downstream survival pathways, we examined the colony-forming efficiency of different cell lines on SMURF2 siRNA and control siRNA treatments. For this study, besides all these cell lines, we also used two additional pancreatic cell lines (BxPC3 and Miapaca2), two EGFR-null cell lines (SW620 [colorectal] and CHO), and the normal mouse fibroblasts (NIH3T3). Cells were either left untreated or transfected with either control or SMURF2 siRNA as mentioned. At 24 hours after transfection, cells were plated for clonogenic survival assays (Figure 2D). SMURF2 siRNA had minimal effects on clonogenic survival of EGFR-null SW620 and CHO cells and also on normal epithelial cells (Het1A) and fibroblasts (NIH3T3). However, all the EGFR-positive cancer cells, irrespective of gefitinib resistance (T790M) mutation (NCI-H1975), were sensitive to SMURF2 siRNA to varying extents. Among them, UMSCC74B and NCI-H1975 cells were the two most affected lines with approximately 98% reduction in clonogenic survival efficiency (Figure 2D). These data indicate that SMURF2 siRNA had a great effect on EGFR-dependent cancer cell survival than on normal cells, and thus SMURF2 knockdown might have therapeutic potential. We have also looked at the EGFR downstream signaling events, particularly, AKT and STAT3 phosphorylation, which also showed significant down-regulation in SMURF2 siRNA-treated samples (Figure W1).

SMURF2 siRNA Treatment Induced Autophagic Cell Death in Cancer Cells

As loss of EGFR is known to induce autophagy [37], we wanted to explore whether SMURF2 siRNA treatment was inducing an autophagic response, which may be responsible for reduction in clonogenic cancer cell survival. Light chain 3 (LC3) is a well-established marker for autophagy. During autophagy, the unmodified LC3-I form is converted to the LC3-II form through lipidation, allowing LC3-II to be associated with autophagosomes, appearing as punctuate spots under microscope on immunostaining and running as a faster migrating band by immunoblot analysis [38]. To test our hypothesis, we have used HeLa cells stably expressing GFP-LC3, a well-studied model for autophagy [39]. HeLa, an EGFR-positive cell line, when treated with SMURF2 siRNA showed significant reduction in the EGFR steady-state level compared with control siRNA (Figure 3B). Like other EGFR-positive cancer cell lines (as shown in Figure 2C), the clonogenic survival was significantly reduced (0.95 ± 0.04 vs 0.49 ± 0.03) (Figure 3C) in SMURF2 siRNA-treated HeLa-GFP-LC3 cells. Furthermore, SMURF2 siRNA-transfected cells showed increased GFP-LC3 punctuate pattern 48 hours after transfection (Figure 3A), and the same sample showed an increase in the LC3-II form on immunoblot analysis (Figure 3B). These data suggest that autophagy could be one of the mechanisms for the loss of clonogenic survival in cancer cell lines on SMURF2 siRNA treatment.

Figure 3.

Figure 3

Open in a new tab

SMURF2 knockdown induced autophagic response in tumor cells. (A) HeLa-LC3-GFP cells were transfected with either control or SMURF2 siRNA. Forty-eight hours after transfection, cells were observed by fluorescence microscopy and photographed. SMURF2 siRNA-treated cells showed increased punctate fluorescence. Bar, 50 µm. (B) Protein lysates were isolated from the samples and immunoblotted for indicated antibodies. Relative quantification was performed using ImageJ software setting the control siRNA value as 1. (C) HeLa cells were transfected with siRNA (C or S). Twenty-four hours after transfection, cells were trypsinized and replated to determine the clonogenic survival efficiency. Survival efficiency for untreated group was normalized to 1 to determine the survival fraction for the control and SMURF2 siRNA group and presented as mean ± SEM from three independent experiments.

Smurf2 Knockdown Abrogated Tumor-Forming Ability of UMSCC74B Cells in Nude Mice

To explore the therapeutic potential of SMURF2 siRNA, we assessed the in vivo effects of siRNA on UMSCC74B cancer cell line. This cell line was chosen for two reasons: 1) according to our clonogenic survival data (Figure 2C), SMURF2 siRNA had a robust effect in this cell line; and 2) UMSCC74B cells form aggressive tumors in nude mice within 4 weeks of subcutaneous injection. We found that SMURF2 siRNA-treated UMSCC74B cells, rarely formed tumors within the 75-day observation period, and those that did form had a latency of more than 54 days (Figure 4). In contrast, nude mice carrying untreated UMSCC74B cells or cells pretreated with control siRNA formed aggressive tumors with a median time-to-tumor detection were more than 29 days and more than 26 days, respectively (Figure 4). However, tumor growth rates were comparable between the three groups once a palpable tumor was detected (Figure 4, upper panel). On the basis of the log-rank tests, there was significant difference in time-to-tumor initiation between control siRNA and SMURF2 siRNA-treated groups (P = .001); however, there was no significant difference on time-to-tumor initiation between control siRNA and untreated group (P = .17). The difference was marginally significant between SMURF2 siRNA-treated and untreated groups (P = .09). From these studies, we concluded that SMURF2 siRNA has therapeutic potential, which may be mediated through EGFR down-regulation.

Figure 4.

Figure 4

Open in a new tab

siRNA-mediated SMURF2 down-regulation reduced tumor-forming potential of an HNSCC cell line in nude mice. Upper panel: Relative tumor volume was calculated for each tumor by normalizing the first time detected tumor volume as 1. Tumor growth kinetics was presented by plotting the relative tumor volume with respect to the days after injection. Lower panel: Total number of tumors detected in each group and the median time-to-tumor detection in days.

Correlation of Expression between EGFR and SMURF2 in Lung Adenocarcinoma Patients at the mRNA and Protein Levels

To better understand the potential clinical importance of the interaction between EGFR and SMURF2, we analyzed gene expression microarray data from 443 lung adenocarcinoma patients [34]. Among all the different HECT family members, SMURF2, which belongs to the Nedd4 subfamily, showed the strongest correlation of expression (r = 0.42, n = 443, P < .001) with EGFR mRNA (Figure 5A and Table 1). To determine the correlation between EGFR and SMURF2 expression at the protein level, we performed coimmunofluorescence staining on a TMA isolated from 67 lung adenocarcinoma patients. We used cytokeratin 8 (CK8) as an epithelial cell marker to specify tumor cells, which helped quantifying the tumor cell-specific EGFR and SMURF2 protein levels. On the basis of the AQUA, EGFR and SMURF2 showed a strong correlation (r = 0.711, n = 67, P < .0001) at the protein level (Figure 5B). Representative fluorescence micrographs of a patient's lung tumor showing tumor-specific EGFR and SMURF2 colocalization along with CK8 staining is shown in Figure 5C. Interestingly, based on the same gene expression microarray data, among the various families of ubiquitin ligases (e.g., RING and HECT), EGFR mRNA levels was most positively correlated with SMURF2 (data not shown).

Figure 5.

Figure 5

Open in a new tab

Significant correlation of expression between EGFR and SMURF2 in lung adenocarcinoma and H&N cancer patients. (A) Plots of EGFR and SMURF2 matched expression from a gene expression microarray showing strong correlation (r = 0.42, n = 443, P < .001). (B) TMAs obtained from lung adenocarcinoma patients were stained with SMURF2, EGFR, CK8, and DAPI. Immunofluorescence intensities were quantified using AQUA for individual antibodies. To quantify tumor-specific EGFR and SMURF2 staining, CK8 staining was used as a reference. Such analysis showed very strong correlation (r = 0.71, n = 67, P < .0001) between EGFR and SMURF2. (C) Representative immunofluorescence pictures of TMA from lung and head and neck cancer patients stained with anti-EGFR, SMURF2, and CK8 antibodies. Leftmost panels show the merged images of EGFR and SMURF2 staining showing colocalization (orange). Bars, 100 µm.

Table 1.

Among All HECT Ubiquitin Ligases, SMURF2 Is Best Correlated with EGFR at the mRNA Level in Lung Adenocarcinoma Patients.

HECT Ligases r* to EGFR
HERC family
HERC6 0.2841
HERC5 0.1207
HERC3 0.0664
HERC4 0.0311
HERC1 0.0102
HERC2 -0.0072
NEDD4 family SMURF2
0.4186
SMURF1 0.2655
WWP2 0.2079
WWP1 0.0918
NEDD4 0.0585
ITCH -0.0514
NEDD4L -0.057
HECW1 -0.1091
Other HECTs
UBE3C 0.0559
KIAA0317 0.0543
HUWE1 0.0188
TRIP12 0.0047
UBE3B -0.1147
Open in a new tab

The Pearson correlation analysis was performed for 19 HECT ubiquitin ligases of three different subfamilies (HERC family, Nedd4 family, and other HECTs) to EGFR from gene expression data set of lung adenocarcinoma patients as described previously [34] and tabulated. Among all the tested family members, SMURF2 showed the best positive correlation (r = 0.42, n = 443, P < .001) to EGFR.

*

Pearson correlation (n = 443, P < .001).

Discussion

In this study, we have identified EGFR as a novel substrate for SMURF2, which ubiquitinates but stabilizes EGFR; conversely, EGFR undergoes EGF-mediated rapid turnover in the absence of SMURF2. Such observations have both basic as well as significant clinical relevance: (a) it identifies a molecular regulator, which may be critical in tumor cell-specific EGFR overexpression, a common occurrence in a variety of epithelial tumors; and (b) it also identifies SMURF2 as a novel therapeutic target, down-regulation of which can degrade EGFR protein leading to reduced clonogenic survival of EGFR-addicted cancer cells and inhibition of tumor initiation in a xenograft model. A major advantage of this approach is that it does not target receptor kinase activity, so that tumor response is independent of potential drug-resistant EGFR (T790M) mutations. These findings motivate a search for an agent that would disrupt EGFR-SMURF2 binding, degrade EGFR, and cause cancer cell death.

Our study shows that EGFR joins a growing group of receptors that can undergo ubiquitination by HECT-type ubiquitin ligases. In this study, we have demonstrated a direct physical interaction between EGFR and SMURF2, although the domain(s) involved in this interaction remains to be identified. Besides, it is interesting to study the subcellular localization where EGFR-SMURF2 interaction is taking place. Because SMURF2 localization is highly dynamic (nuclear as well as cytosolic) and as previously reported SMURF2 interacts with TGF-β receptor in distinct endosomal compartments [40], we hypothesize that SMURF2 may be interacting mainly with the endocytosed EGFR pool and thus plays a critical role in receptor recycling/degradation, an area of research slightly underinvestigated.

We also found that SMURF2 can inhibit c-Cbl-mediated EGFR down-regulation, but the molecular mechanism is not yet certain. One plausible hypothesis is the differential ubiquitination pattern exerted by the two different families of ubiquitin ligases. c-Cbl-mediated ubiquitination of EGFR leads to its lysosomal degradation [41]. Because K63-linked polyubiquitin chain formation facilitates lysosomal degradation of proteins [42,43] and c-Cbl was shown to catalyze such ubiquitin linkages on EGFR [44,45], we are hypothesizing that SMURF2 may be involved in attaching non-K63-linked polyubiquitin chains, thereby counteracting c-Cbl activity. Furthermore, different deubiquitinating enzymes [46,47] may be a critical in balancing the c-Cbl and SMURF2 activity, thus tightly regulate the EGFR protein stability, alteration of which during oncogenesis leads to EGFR overexpression. For future studies, the involvement of EGFR in transcriptional/posttranscriptional regulation of SMURF2 also remains an interesting question.

In conclusion, we have shown that EGFR is a novel substrate of SMURF2, an interaction found to be critical for the receptor protein stability in an ubiquitination-dependent manner. We also demonstrated that the loss of SMURF2 caused rapid EGFR degradation irrespective of the presence or absence of a drug-resistant mutation in EGFR and decreased cancer cell survival and inhibition of tumor initiation in nude mice. The presence of a strong correlation between the two oncogenes in lung cancer patients suggested that our findings are clinically relevant. These findings motivate the investigation of the therapeutic efficacy of SMURF2 knockdown in treating EGFR-addicted cancers more effectively either as an individual therapy or in combination with already existing chemotherapy and/or radiotherapy.

Supplementary Material

Supplementary Figures and Tables
neo1307_0570SD1.pdf (66.2KB, pdf)

Acknowledgments

The authors thank Yi Sun for discussion and also providing HeLa-LC3B-GFP-expressing cells. SMURF2-FLAG construct and mouse monoclonal SMURF2 antibody were kind gifts from Dr Gerald Thomsen (Stony Brook University, NY). The authors thank Thomas Dafydd for automated image acquisition and analysis using AQUA, members of the Kiyokawa and Nyati laboratories for technical assistance, Mary Davis for assistance in article preparation, and Steven Kronenberg for graphic assistance.

Footnotes

1

This work was supported by grants R01CA131290 and P50 CADE97248 (to M.K.N.) and University of Michigan Cancer Center support grant 5 P30 CA46592.

2

This article refers to supplementary material, which is designated by Figure W1 and is available online at www.neoplasia.com.

References

  • 1.Zwick E, Bange J, Ullrich A. Receptor tyrosine kinase signalling as a target for cancer intervention strategies. Endocr Relat Cancer. 2001;8:161–173. doi: 10.1677/erc.0.0080161. [DOI] [PubMed] [Google Scholar]
  • 2.Grandis JR, Tweardy DJ. Elevated levels of transforming growth factor alpha and epidermal growth factor receptor messenger RNA are early markers of carcinogenesis in head and neck cancer. Cancer Res. 1993;53:3579–3584. [PubMed] [Google Scholar]
  • 3.Hirsch FR, Varella-Garcia M, Bunn PA, Jr, Di Maria MV, Veve R, Bremmes RM, Baron AE, Zeng C, Franklin WA. Epidermal growth factor receptor in non-small-cell lung carcinomas: correlation between gene copy number and protein expression and impact on prognosis. J Clin Oncol. 2003;21:3798–3807. doi: 10.1200/JCO.2003.11.069. [DOI] [PubMed] [Google Scholar]
  • 4.Soulieres D, Senzer NN, Vokes EE, Hidalgo M, Agarwala SS, Siu LL. Multicenter phase II study of erlotinib, an oral epidermal growth factor receptor tyrosine kinase inhibitor, in patients with recurrent or metastatic squamous cell cancer of the head and neck. J Clin Oncol. 2004;22:77–85. doi: 10.1200/JCO.2004.06.075. [DOI] [PubMed] [Google Scholar]
  • 5.Ciardiello F, Tortora G. EGFR antagonists in cancer treatment. N Engl J Med. 2008;358:1160–1174. doi: 10.1056/NEJMra0707704. [DOI] [PubMed] [Google Scholar]
  • 6.Harari PM. Epidermal growth factor receptor inhibition strategies in oncology. Endocr Relat Cancer. 2004;11:689–708. doi: 10.1677/erc.1.00600. [DOI] [PubMed] [Google Scholar]
  • 7.Ahsan A, Hiniker SM, Ramanand SG, Nyati S, Hegde A, Helman A, Menawat R, Bhojani MS, Lawrence TS, Nyati MK. Role of epidermal growth factor receptor degradation in cisplatin-induced cytotoxicity in head and neck cancer. Cancer Res. 2010;70:2862–2869. doi: 10.1158/0008-5472.CAN-09-4294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ahsan A, Hiniker SM, Davis MA, Lawrence TS, Nyati MK. Role of cell cycle in epidermal growth factor receptor inhibitor-mediated radiosensitization. Cancer Res. 2009;69:5108–5114. doi: 10.1158/0008-5472.CAN-09-0466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Chun PY, Feng FY, Scheurer AM, Davis MA, Lawrence TS, Nyati MK. Synergistic effects of gemcitabine and gefitinib in the treatment of head and neck carcinoma. Cancer Res. 2006;66:981–988. doi: 10.1158/0008-5472.CAN-05-2665. [DOI] [PubMed] [Google Scholar]
  • 10.Busino L, Chiesa M, Draetta GF, Donzelli M. Cdc25A phosphatase: combinatorial phosphorylation, ubiquitylation and proteolysis. Oncogene. 2004;23:2050–2056. doi: 10.1038/sj.onc.1207394. [DOI] [PubMed] [Google Scholar]
  • 11.Busino L, Donzelli M, Chiesa M, Guardavaccaro D, Ganoth D, Dorrello NV, Hershko A, Pagano M, Draetta GF. Degradation of Cdc25A by β-TrCP during S phase and in response to DNA damage. Nature. 2003;426:87–91. doi: 10.1038/nature02082. [DOI] [PubMed] [Google Scholar]
  • 12.Donzelli M, Squatrito M, Ganoth D, Hershko A, Pagano M, Draetta GF. Dual mode of degradation of Cdc25 A phosphatase. EMBO J. 2002;21:4875–4884. doi: 10.1093/emboj/cdf491. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Sehat B, Andersson S, Girnita L, Larsson O. Identification of c-Cbl as a new ligase for insulin-like growth factor-I receptor with distinct roles from Mdm2 in receptor ubiquitination and endocytosis. Cancer Res. 2008;68:5669–5677. doi: 10.1158/0008-5472.CAN-07-6364. [DOI] [PubMed] [Google Scholar]
  • 14.Monami G, Emiliozzi V, Morrione A. Grb10/Nedd4-mediated multi-ubiquitination of the insulin-like growth factor receptor regulates receptor internalization. J Cell Physiol. 2008;216:426–437. doi: 10.1002/jcp.21405. [DOI] [PubMed] [Google Scholar]
  • 15.Levkowitz G, Waterman H, Zamir E, Kam Z, Oved S, Langdon WY, Beguinot L, Geiger B, Yarden Y. c-Cbl/Sli-1 regulates endocytic sorting and ubiquitination of the epidermal growth factor receptor. Genes Dev. 1998;12:3663–3674. doi: 10.1101/gad.12.23.3663. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Inoue Y, Imamura T. Regulation of TGF-β family signaling by E3 ubiquitin ligases. Cancer Sci. 2008;99:2107–2112. doi: 10.1111/j.1349-7006.2008.00925.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Rotin D, Kumar S. Physiological functions of the HECT family of ubiquitin ligases. Nat Rev Mol Cell Biol. 2009;10:398–409. doi: 10.1038/nrm2690. [DOI] [PubMed] [Google Scholar]
  • 18.Kavsak P, Rasmussen RK, Causing CG, Bonni S, Zhu H, Thomsen GH, Wrana JL. Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGF beta receptor for degradation. Mol Cell. 2000;6:1365–1375. doi: 10.1016/s1097-2765(00)00134-9. [DOI] [PubMed] [Google Scholar]
  • 19.Izzi L, Attisano L. Regulation of the TGFβ signalling pathway by ubiquitin-mediated degradation. Oncogene. 2004;23:2071–2078. doi: 10.1038/sj.onc.1207412. [DOI] [PubMed] [Google Scholar]
  • 20.Zhang Y, Chang C, Gehling DJ, Hemmati-Brivanlou A, Derynck R. Regulation of Smad degradation and activity by Smurf2, an E3 ubiquitin ligase. Proc Natl Acad Sci USA. 2001;98:974–979. doi: 10.1073/pnas.98.3.974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Osmundson EC, Ray D, Moore FE, Gao Q, Thomsen GH, Kiyokawa H. The HECT E3 ligase Smurf2 is required for Mad2-dependent spindle assembly checkpoint. J Cell Biol. 2008;183:267–277. doi: 10.1083/jcb.200801049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Moore FE, Osmundson EC, Koblinski J, Pugacheva E, Golemis EA, Ray D, Kiyokawa H. The WW-HECT protein Smurf2 interacts with the docking protein NEDD9/HEF1 for Aurora A activation. Cell Div. 2010;5:22. doi: 10.1186/1747-1028-5-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Ogunjimi AA, Briant DJ, Pece-Barbara N, Le Roy C, Di Guglielmo GM, Kavsak P, Rasmussen RK, Seet BT, Sicheri F, Wrana JL. Regulation of Smurf2 ubiquitin ligase activity by anchoring the E2 to the HECT domain. Mol Cell. 2005;19:297–308. doi: 10.1016/j.molcel.2005.06.028. [DOI] [PubMed] [Google Scholar]
  • 24.Liu X, Lee J, Cooley M, Bhogte E, Hartley S, Glick A. Smad7 but not Smad6 cooperates with oncogenic ras to cause malignant conversion in a mouse model for squamous cell carcinoma. Cancer Res. 2003;63:7760–7768. [PubMed] [Google Scholar]
  • 25.Cassell A, Grandis JR. Investigational EGFR-targeted therapy in head and neck squamous cell carcinoma. Expert Opin Investig Drugs. 2010;19:709–722. doi: 10.1517/13543781003769844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Hanawa M, Suzuki S, Dobashi Y, Yamane T, Kono K, Enomoto N, Ooi A. EGFR protein overexpression and gene amplification in squamous cell carcinomas of the esophagus. Int J Cancer. 2006;118:1173–1180. doi: 10.1002/ijc.21454. [DOI] [PubMed] [Google Scholar]
  • 27.Fukuchi M, Fukai Y, Masuda N, Miyazaki T, Nakajima M, Sohda M, Manda R, Tsukada K, Kato H, Kuwano H. High-level expression of the Smad ubiquitin ligase Smurf2 correlates with poor prognosis in patients with esophageal squamous cell carcinoma. Cancer Res. 2002;62:7162–7165. [PubMed] [Google Scholar]
  • 28.Tsutsui T, Hesabi B, Moons DS, Pandolfi PP, Hansel KS, Koff A, Kiyokawa H. Targeted disruption of CDK4 delays cell cycle entry with enhanced p27(Kip1) activity. Mol Cell Biol. 1999;19:7011–7019. doi: 10.1128/mcb.19.10.7011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Feng FY, Varambally S, Tomlins SA, Chun PY, Lopez CA, Li X, Davis MA, Chinnaiyan AM, Lawrence TS, Nyati MK. Role of epidermal growth factor receptor degradation in gemcitabine-mediated cytotoxicity. Oncogene. 2007;26:3431–3439. doi: 10.1038/sj.onc.1210129. [DOI] [PubMed] [Google Scholar]
  • 30.Zou X, Tsutsui T, Ray D, Blomquist JF, Ichijo H, Ucker DS, Kiyokawa H. The cell cycle-regulatory CDC25A phosphatase inhibits apoptosis signal-regulating kinase 1. Mol Cell Biol. 2001;21:4818–4828. doi: 10.1128/MCB.21.14.4818-4828.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Ray D, Terao Y, Nimbalkar D, Chu LH, Donzelli M, Tsutsui T, Zou X, Ghosh AK, Varga J, Draetta GF, et al. Transforming growth factor beta facilitates β-TrCP-mediated degradation of Cdc25A in a Smad3-dependent manner. Mol Cell Biol. 2005;25:3338–3347. doi: 10.1128/MCB.25.8.3338-3347.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Nyati MK, Maheshwari D, Hanasoge S, Sreekumar A, Rynkiewicz SD, Chinnaiyan AM, Leopold WR, Ethier SP, Lawrence TS. Radio-sensitization by pan ErbB inhibitor CI-1033 in vitro and in vivo. Clin Cancer Res. 2004;10:691–700. doi: 10.1158/1078-0432.ccr-1041-03. [DOI] [PubMed] [Google Scholar]
  • 33.Hahnfeldt P, Panigrahy D, Folkman J, Hlatky L. Tumor development under angiogenic signaling: a dynamical theory of tumor growth, treatment response, and postvascular dormancy. Cancer Res. 1999;59:4770–4775. [PubMed] [Google Scholar]
  • 34.Shedden K, Taylor JM, Enkemann SA, Tsao MS, Yeatman TJ, Gerald WL, Eschrich S, Jurisica I, Giordano TJ, Misek DE, et al. Gene expression-based survival prediction in lung adenocarcinoma: a multi-site, blinded validation study. Nat Med. 2008;14:822–827. doi: 10.1038/nm.1790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Camp RL, Chung GG, Rimm DL. Automated subcellular localization and quantification of protein expression in tissue microarrays. Nat Med. 2002;8:1323–1327. doi: 10.1038/nm791. [DOI] [PubMed] [Google Scholar]
  • 36.Wiesner S, Ogunjimi AA, Wang HR, Rotin D, Sicheri F, Wrana JL, Forman-Kay JD. Autoinhibition of the HECT-type ubiquitin ligase Smurf2 through its C2 domain. Cell. 2007;130:651–662. doi: 10.1016/j.cell.2007.06.050. [DOI] [PubMed] [Google Scholar]
  • 37.Weihua Z, Tsan R, Huang WC, Wu Q, Chiu CH, Fidler IJ, Hung MC. Survival of cancer cells is maintained by EGFR independent of its kinase activity. Cancer Cell. 2008;13:385–393. doi: 10.1016/j.ccr.2008.03.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Tanida I, Ueno T, Kominami E. LC3 and autophagy. Methods Mol Biol. 2008;445:77–88. doi: 10.1007/978-1-59745-157-4_4. [DOI] [PubMed] [Google Scholar]
  • 39.Kimura S, Fujita N, Noda T, Yoshimori T. Monitoring autophagy in mammalian cultured cells through the dynamics of LC3. Methods Enzymol. 2009;452:1–12. doi: 10.1016/S0076-6879(08)03601-X. [DOI] [PubMed] [Google Scholar]
  • 40.Di Guglielmo GM, Le Roy C, Goodfellow AF, Wrana JL. Distinct endocytic pathways regulate TGF-β receptor signalling and turnover. Nat Cell Biol. 2003;5:410–421. doi: 10.1038/ncb975. [DOI] [PubMed] [Google Scholar]
  • 41.Sorkin A, Goh LK. Endocytosis and intracellular trafficking of ErbBs. Exp Cell Res. 2009;315:683–696. doi: 10.1016/j.yexcr.2008.07.029. [DOI] [PubMed] [Google Scholar]
  • 42.Lauwers E, Jacob C, Andre B. K63-linked ubiquitin chains as a specific signal for protein sorting into the multivesicular body pathway. J Cell Biol. 2009;185:493–502. doi: 10.1083/jcb.200810114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Barriere H, Nemes C, Du K, Lukacs GL. Plasticity of polyubiquitin recognition as lysosomal targeting signals by the endosomal sorting machinery. Mol Biol Cell. 2007;18:3952–3965. doi: 10.1091/mbc.E07-07-0678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Huang F, Kirkpatrick D, Jiang X, Gygi S, Sorkin A. Differential regulation of EGF receptor internalization and degradation by multiubiquitination within the kinase domain. Mol Cell. 2006;21:737–748. doi: 10.1016/j.molcel.2006.02.018. [DOI] [PubMed] [Google Scholar]
  • 45.Argenzio E, Bange T, Oldrini B, Bianchi F, Peesari R, Mari S, Di Fiore PP, Mann M, Polo S. Proteomic snapshot of the EGF-induced ubiquitin network. Mol Syst Biol. 2011;7:462. doi: 10.1038/msb.2010.118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Alwan HA, van Leeuwen JE. UBPY-mediated epidermal growth factor receptor (EGFR) de-ubiquitination promotes EGFR degradation. J Biol Chem. 2007;282:1658–1669. doi: 10.1074/jbc.M604711200. [DOI] [PubMed] [Google Scholar]
  • 47.McCullough J, Clague MJ, Urbe S. AMSH is an endosome-associated ubiquitin isopeptidase. J Cell Biol. 2004;166:487–492. doi: 10.1083/jcb.200401141. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figures and Tables
neo1307_0570SD1.pdf (66.2KB, pdf)

Articles from Neoplasia (New York, N.Y.) are provided here courtesy of Neoplasia Press

RESOURCES