Background
The ubiquitin-proteasome system (UPS) plays a prominent role in cell regulatory mechanism to maintain homeostasis. Evidently, every part of the cell is under the control of the ubiquitin-proteasome regulatory system. The peptide products of this proteasome system provides additional outcome by functioning as a critical factor in deciding whether a cell will be recognized as infected or tumorigenic by the immune system and destroyed, or recognized as self and spared [
1]. In many cases, stability of several important scaffold proteins is controlled by ubiquitination, a predominant posttranslational modification central to eukaryotic regulation [
2,
3]. E3s carry out the key rate-limiting step in ubiquitin-mediated proteolysis. Multiple E3 ubiquitin ligases can targets same protein and vice versa, corroborating a highly complex and dynamic regulation [
4]. Hence, the dynamic role of cellular ubiquitination can be further confirmed by identifying new targets for a given ubiquitin ligase and, conversely, new ubiquitin ligases for a given target [
4].
Smurf2 is a C2-WW-HECT E3 ubiquitin ligase that has the ability to ubiquitinate and degrade some proteins, e.g., the TGF-β receptor I and associated Smad proteins involved in TGF- β/BMP signaling, but also have a unique ability to protect other substrates, e.g., spindle assembly checkpoint protein, MAD2 and NEDD9/HEF1 [
5]. It contains WW domains, which directly bind to a PP
XY motif (also known as PY motif) in its targets which is further stabilized by the PY tail, a six-amino acid stretch immediately carboxylterminal to the PP
XY motif, although additional interactions exist [
6]. Smurfs are frequently deregulated in human cancers and are critical regulators of multiple oncogenic factors. However, the specific role of Smurf2 in tumorigenesis has been highly obscure, with numerous studies showing that it is either pro-tumorigenic or, conversely, anti-tumorigenic in a context-dependent manner. Smurf2 has involved in diverse signal pathways and cellular processes by targeting broad spectrum of proteins. Recent studies emphasize the growing appreciation that the pleiotropic effects of Smurf2 might be explained by the ability of Smurf2 to target potential molecular substrates in a ligase dependent and independent manner. Thus, expanding the family of Smurf2 targets may provide further insights into regulatory role of Smurf E3 ubiquitin ligases in the cellular environment [
5].
CNK (Connector enhancer of kinase suppressor of ras) is a putative multi-adaptor scaffold protein required in multiple receptor tyrosine kinase pathways particularly in RAS-RAF/ mitogen–activated protein kinase (MAPK) cascade which is involved in conveying several proliferative and differentiative signals to the nucleus [
7]. CNKSR2 (Connector enhancer of kinase suppressor of ras 2) also named as Membrane-Associated GUanylate kinase Interacting protein (MAGUIN) is a mammalian isoform of CNK, and is the closest homolog to
Drosophila melanogaster and
C. elegans CNK, thus likely represents the orthologous member. The neuronal isoform CNK2 has an essential function in nerve growth factor-induced, sustained stimulation of ERK leading to neuronal differentiation [
8]. But CNK2 is not required for epidermal growth factor-dependent, transient stimulation of ERK occurring during cell proliferation. CNK’s multidomain architecture suggests that it has the ability to bind and bring together different molecules as previously shown for several other multidomain molecules [
8]. In particular, consistent with other known potential substrates of Smurf2, we have also identified a probable Smurf2-interacting motif in CNKSR2, a ‘SPPPPY’ motif at 702–707 sequence region that shows a strong PY motif match with Smurf2. Indeed, we observed that knockdown of Smurf2 downregulated the expression of CNKSR2 and reduced the proliferative potential of human breast cancer cells [
9]. Hence we hypothesized that CNKSR2 may be a novel substrate for Smurf2 E3 ubiquitin ligase which seems to perform a crucial role in regulating the stability of CNKSR2. However, whether additional ubiquitin ligases for CNKSR2 exist and how CNKSR2 is regulated by various ubiquitin ligases are not clearly defined.
Methods
Plasmids, cell lines and culture conditions
pCMV5B-Flag-Smurf2 and pCMV5B-Flag-Smurf2 C716A plasmids have been described [
10] and were purchased from Addgene, Cambridge, MA, USA. Human embryonic kidney 293 (HEK293) cells were purchased from National Centre for Cell Science (NCCS, Pune, India). Immortalized normal human breast epithelial cell line MCF10A, and the human breast cancer MCF-7 and MDA-MB-231 cells were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). The pre-malignant and in situ grade cell lines-MCF10AT and MCFDCIS were a kind gift from Dr. Suresh Kumar Rayala (Indian Institute of Technology-Madras, India). MCF10A and MCF10AT cells were cultured in Dulbecco’s modified eagles medium (DMEM)/Nutrient F12-Ham (1:1) supplemented with 10% fetal bovine serum (FBS), 20 ng/ml Epidermal Growth Factor (EGF), 100 ng/ml cholera toxin, 0.01 mg/ml Insulin-Transferrin-Selenium (ITS), 500 ng/ml hydrocortisone, and 1 ng/ml Fibroblast Growth Factor (FGF). MCFDCIS, MCF-7 and MDA-MB-231 cells were grown at 37 °C with 5% CO
2 in Dulbecco’s modified eagles medium (DMEM) medium supplemented with 10% FBS and 1% penicillin/streptomycin (Invitrogen; Carlsband, CA, USA).
Tissue samples
Human breast tissue samples were collected from Regional Cancer Centre (RCC), Thiruvananthapuram, India after obtaining approval from the Institutional Human Ethics Committee. The samples were collected from patients who underwent primary surgery without prior chemotherapy or radiotherapy. Informed consent was taken from all the patients prior surgery/excision of tissue.
siRNA and shRNA transfections
siRNA against Smurf2, in the form of either a mixture of three siRNAs targeting different regions of Smurf2 (Santa Cruz Biotechnology; Texas, USA), or the negative control siRNA included in the kit (Santa Cruz Biotechnology; Texas, USA) was transfected into 60–70% confluent MDA-MB-231 cells with siLentFect (Bio-Rad; CA, USA) according to the manufacturer’s instructions. Three days later, cells were subjected to western blotting, immunoprecipitation, and quantitative RT-PCR analysis.
Stable knockdown of Smurf2 in MDA-MB-231 cells was achieved by transfecting Smurf2 shRNA plasmid containing a pool of three target-specific lentiviral vector plasmids each encoding 19–25 nucleotide (plus hairpin) shRNAs designed to knockdown Smurf2 gene expression (Santa cruz biotechnology; Texas, USA) using Amaxa® Cell Line Nucleofector® Kit V (Lonza; Basel, Switzerland) according to the manufacturer’s instructions. Briefly, 1 × 106 cells were transfected with 2 μg plasmid DNA in 100 μl cell Line Nucleofector® Solution V (Lonza; Basel, Switzerland) and seeded in 6-well plate containing pre-equilibrated DMEM medium. After 24 h, cells were selected for stable integration using 1 μg/ml Puromycin (Sigma; St. Louis, MO, USA) for 3 weeks and were used for further experiments. MDA-MB-231 cells stably expressing Control shRNA Plasmid-A (Santa Cruz Biotechnology; Texas, USA) were also generated in the same fashion to serve as negative control.
Immunoprecipitation
For cultured cells, in a 10-cm dish format, HEK293 cells transfected with empty vector, pCMV5B-Flag-Smurf2 WT, or pCMV5B-Flag-Smurf2 C716A were washed in 1 × PBS and resuspended in 1 ml of lysis buffer (70 mM NaCl, 50 mM Tris, pH 8 and 0.5% NP-40), supplemented with phosphatase inhibitor and protease inhibitor cocktails (Sigma; St. Louis, MO, USA). After rotating at 4 °C for 30 min, the cell lysate was collected and precleared by spinning at 14,000 rpm for 10 min. For co-immunoprecipitation of endogenous Smurf2 with CNKSR2, total protein was isolated from MDA-MB-231 as described previously. For each pull down, 3 μg of corresponding antibody was added to the normalized lysate (1 mg of total protein by Bradford’s Assay, Bio- Rad; CA, USA)) and the mixture was incubated overnight at 4 °C. Normal rabbit IgG was used as negative control. Immune complexes were then precipitated with protein A sepharose macrobeads (Sigma; St. Louis, MO, USA). Beads were washed four times with lysis buffer and immunoprecipitated samples were resolved by SDS-PAGE for immunoblotting.
Surface plasmon resonance (SPR)
Binding kinetics was determined by SPR using a ProteOn XPR36™ SPR instrument (Bio-Rad; CA, USA) with a GLC (General Layer Compact) biosensor chip (Bio-Rad; CA, USA) with its unique “one-shot” kinetics approach. The apparatus has six parallel flow channels for uniformly immobilizing strips of six ligands on the sensor surface. The fluidic system can be rotated automatically to 90° so that up to six different analytes can be injected, thereby simultaneously monitoring up to 36 individual molecular interactions in a single run on a single chip [
11]. In our study, the ligand (recombinant Smurf2 protein; Sigma; St. Louis, MO, USA) was immobilised to the biosensor surface layer by amine coupling chemistry, followed by capture of the analyte (recombinant CNKRS2 protein; Cusabio Biotech, China) as described previously (Bravman et al., [
11]). One of the six ligand channel (L4) was used to capture the Smurf2 protein and one ligand channel was left blank (L6) by injecting the 10 mM sodium acetate buffer, pH 4.0 (without Smurf2), which served as the ligand reference channel. The chip surface was activated for the two ligand channels (L4 and L6) using a mixture of 5 mM N-hydroxysulfosuccinimide (sulfo-NHS) and 20 mM 1-ethyl-3(3-dimethylaminopropyl)-carbodiimide (EDC), ethanolamine-HCl) mixed in 1:1 ratio to maintain a uniform composition throughout the chip. This was followed by an immediate injection of 300 μl of 7.5 μg/ml of the recombinant Smurf2 protein in 10 mM sodium acetate buffer, pH 4.0 for 600 s. Immobilisation was performed at 25 °C at a flow rate of 30 μl/min using PBS/Tween (Phosphate buffered saline, pH 7.4, 0.005% Tween 20), as the running buffer. The sensor chip surface was then inactivated using 120 μl of 1 M ethanolamine-HCl to block residual NHS-activated carboxyl groups on the chip surface. The final immobilization levels of the Smurf2 ligand was ~ 2000 response units (RU) (1 RU = 1 pg of protein/mm
2).
During the interaction step the running buffer used was 20 mM Tris-HCl, pH -8.0, 0.5 M NaCl. The system was primed with the running buffer before the start of the experiment. Four of the six analyte channels (A1-A4) were used to capture CNKRS2 at four different concentrations using a twofold dilution series ranging from 0.25 μM to 2 μM. All the four different analyte concentrations were passed simultaneously (a single injection) in the analyte direction at a flow rate of 100 μl/min for 60 s of association with Smurf2, followed by dissociation for 600 s. In the sixth channel (A6), ‘analyte running buffer’ without any protein was injected for double referencing. All these assays were done at 25 °C. All binding sensorgrams were collected, processed, and analyzed using the integrated ProteOn Manager software version 3.1.0.6 (Bio-Rad; CA, USA). Response data from the ProteOn instrument were normalized to a baseline value of zero just prior to the start of the analyte injection. Responses from the reference positions before and after each reaction spot were corrected by subtracting the non-specific response observed in the reference surface to remove non-specific binding and the large step changes in response due to changes in buffer refractive index (‘buffer shift’). Finally, the binding curves were grouped fitted using the Langmuir model describing 1:1 binding stoichiometry with one local parameter for surface capacity (Rmax), one association rate constant (ka or kon), and one dissociation rate constant (kd or koff). The ratio of the rate constants (kd/ka) yielded the value for the equilibrium or affinity constant (KD).
Immunofluorescence microscopy
HEK293 cells were transfected with pCMV5B-Flag-Smurf2 WT, fixed with ice-cold acetone/methanol (1:1) and blocked with 3% BSA in PBS for 1 h at room temperature. Cells were then incubated with mouse Flag (Sigma; St. Louis, MO, USA) and rabbit CNKSR2 (Abcam; Cambridge, UK) antibodies, followed by incubation with FITC-conjugated anti-mouse and PE-conjugated anti-rabbit (Santa cruz Biotechnology; Texas, USA) secondary antibodies. For immunofluorescence of endogenous Smurf2 and CNKSR2, cells were fixed with ice-cold acetone/methanol (1:1) and blocked with 3% BSA in PBS for 1 h at room temperature. Cells were then incubated with goat Smurf2 (Santa Cruz Biotechnology; Texas, USA) and rabbit CNKSR2 (Abcam; Cambridge, UK) antibodies, followed by incubation with FITC-conjugated anti-goat and PE-conjugated anti-rabbit (Santa Cruz Biotechnology; Texas, USA) secondary antibodies. Cells were also stained with Propidium Iodide (PI) and Hoecht dye to reveal nuclei. Immunofluorescence was monitored by both conventional and confocal microscopy.
Ubiquitination assay
For ubiquitination of endogenous CNKSR2 after Smurf2 overexpression, HEK293 cells were transfected with Smurf2WT, Smurf2C716A, and empty vector and disrupted in IP lysis buffer (70 mM NaCl, 50 mM Tris, pH 8, 0.5% NP-40) supplemented with 1 × phosphatase inhibitor and 1 × protease inhibitor cocktails (Sigma; St. Louis, MO, USA). Lysates were denatured by boiling for 15 min and diluted 10-fold in IP buffer (70 mM NaCl, 50 mM Tris, pH 8, 0.5% NP-40) supplemented with 1 × phosphatase inhibitor and 1 × protease inhibitor cocktails (Sigma; St. Louis, MO, USA). The lysates were immunoprecipitated with rabbit CNKSR2 (Abcam; Cambridge, UK) followed by incubation with protein A sepharose macrobeads (Sigma; St. Louis, MO, USA). Ubiquitinated CNKSR2 was revealed by mouse ubiquitin (Santa Cruz Biotechnology; Texas, USA) immunoblotting. 1% total cell lysate was also included as Input control.
For ubiquitination of endogenous CNKSR2 after Smurf2 knockdown, MDA-MB-231 cells were seeded at 2 × 105 cells/well in 6-well format and treated with Control siRNA or Smurf2-specific siRNA. Cells were treated for 4 h with 10 μM MG132 (Calbiochem; CA, USA) 44 h post-transfection, and disrupted in the lysis buffer (70 mM NaCl, 50 mM Tris, pH 8 and 0.5% NP-40). Lysates were denatured by boiling for 15 min and diluted 10-fold in IP buffer (70 mM NaCl, 50 mM Tris, pH 8, 0.5% NP-40 1Xprotease inhibitor, 1XSer/Thr and Tyr phosphatase inhibitor). The lysates were immunoprecipitated with either control rabbit IgG (Santa cruz biotechnology; Texas, USA), or rabbit CNKSR2 (Abcam; Cambridge, UK). Entire immunoprecipitates were loaded onto gel; gel was transferred onto PVDF membrane as usual. Ubiquitinated CNKSR2 was revealed by mouse ubiquitin (Santa Cruz Biotechnology; Texas, USA) immunoblotting. 1% total cell lysate was also included as Input control.
Immunohistochemistry
Breast cancer tissue arrays with progressive changes- BRC961 and BRC962, non-overlapping with each other and each containing 48 cases from normal, premalignant and cancer tissues with progressive grades and stages in duplicates were purchased from Pantomics. Expression of Smurf2 and CNKSR2 was evaluated using an indirect immunoperoxidase procedure (ABC-Elite, Vector Laboratories; CA, USA). Briefly, TMA slides were baked at 60 °C for 30 min before use. The slides were then deparaffinized in xylene and rehydrated in graded ethanol/isopropanol. Endogenous peroxidase activity was blocked by incubating the slides in 1.5% H2O2 in methanol for 20 min at room temperature. Antigen retrieval was done by dipping the slides for 12 min in boiling citrate buffer. The slides were then blocked with 3% BSA-PBS for 1 h at room temperature. After blocking, the sections were incubated with primary antibodies; rabbit Smurf2 (Santa cruz Biotechnology; Texas, USA) and rabbit CNKSR2 (Abcam; Cambridge, UK), in 1% BSA-PBS at 4 °C overnight. Bound antibody was detected using Vectastain ABC kit (ABC-Elite, Vector Laboratories; CA, USA). Briefly, post PBS wash, the sections were incubated with universal biotinylated anti-mouse IgG/rabbit IgG secondary antibody (C) in normal horse serum blocker (supplied in the kit) for 30 min at room temperature. The sections were then washed thrice in 1XPBS followed by incubation with A + B mixture (Avidin+Biotinylated Horseradish Peroxidase) for another half an hour at room temperature. The sections were again washed with 1XPBS and incubated in DAB (Sigma; St. Louis, MO, USA) solution till they developed brown color. Further, the sections were washed in distilled water and counterstained with hematoxylin for 1 min for nuclear staining. The sections, after bluing for 10 min under running tap water were dehydrated in graded ethanol changes and finally, cleared in xylene. The sections were then mounted using DPX (Qualigens; India) and visualized using a digital up-right microscope, Leica DMI 1000, at 40× magnification.
For scoring, at least five randomly selected high power fields were chosen and a total number of 1000 cells were considered. Only epithelial cells were evaluated. The cells were considered nuclear positive if the nuclei stained positive irrespective of the intensity of staining. Cytoplasmic positivity of Smurf2 and CNKSR2 were measured depending on the intensity of immunoreactivity and scored as mild (+), moderate (++), and intense (+++). The staining intensity in each core biopsy was recorded separately. Slides were read blinded by two readers (D.D. and A.N.) along with discussions made with an expert pathologist. The IHC data regarding ER, PR, and HER2 staining intensity of the corresponding tissues were obtained from Pantomics.
Quantitative real time PCR
shRNAs against Smurf2 (Santa Cruz Biotechnology; Texas, USA) or the negative control shRNA (Santa cruz biotechnology; Texas, USA) was transfected into MDA-MB-231 cells using Amaxa® Cell Line Nucleofector® Kit V (Lonza; Basel, Switzerland) according to the manufacturer’s instructions. Cells were selected for stable integration using 1 μg/ml Puromycin (Sigma; St. Louis, MO, USA) for 3 weeks and total RNA was isolated with TRIzol (Ambion/Invitrogen; Carlsband, CA, USA), and quantitative real-time RT-PCR was performed in three triplicates with primer sets specific for Smurf2(forward,TGGATCAGGAAGTCGGAAAA and reverse, GGACATGTCTAACCCCGGA) and CNKSR2 (forward, CGACCTCCCTCGATGAGTTG and reverse, CACTGCACTGCTCCCAGTTA) and the control gene GAPDH (forward, TTGGTATCGTGGAAGGACTCA and reverse, TGTCATCATATTTGGCAGGTT). Products were amplified and detected with the Power SYBR Green PCR Master Mix (Applied Biosystems; Massachusetts, USA) on an.
ABI 7900HT Fast Real-Time PCR System (Applied Biosystems; Massachusetts, USA) according to the manufacturer’s instructions. The real time analyses were done by 2
-ΔΔCt method [
12] using SDS 2.1 software (Applied Biosystems; Massachusetts, USA) after normalization with the GAPDH control.
Clonogenic assay
Clonogenic assay or foci formation assay was performed as described previously [
13]. Briefly, cells were seeded at a density of 500 cells/well in six-well plates and incubated for 14 days with intermittent media change at every 2 days. After 14 days, plates were rinsed carefully twice with 1 × PBS and fixed with 4% paraformaldehyde. After fixation, cells were washed thrice with 1 × PBS and stained with 0.5% crystal violet for 30 min at room temperature on a rocking platform. The dishes were rinsed three times with water, air-dried and analyzed. Colonies comprising more than 50 cells were counted manually and images were obtained using a digital camera. The experiments were done in duplicate at least three times.
Soft agar colony assay
Anchorage-independent growth was determined by soft agar analysis as follows: Smurf2 knockdown MDA-MB-231 cells were trypsinized and 1 × 104 cells per 35-mm dish were seeded in 0.35% agar on top of a base layer containing 0.8% agar. Plates were incubated at 37 °C at 5% CO2 in a humidified incubator for 21 days. After 21 days, the colonies were fixed with 4% paraformaldehyde and stained with 0.05% crystal violet for 1 h. Colonies with > 50 cells were counted under a microscopic field at 10× magnifications. The experiments were done in duplicates at least three times.
Statistical analysis
Statistical analysis was performed using Intercooled Stata software (Intercooled Stata 11.2 version). The expression of Smurf2, CNKSR2, ER, PR, and HER2 were compared between tumor grades using Chi squared test. The effects of Smurf2 and CNKSR2 expression on histopathological grade of the tumor were estimated with Odds Ratio (OR) (unadjusted Odds ratio) and their 95% Confidence Interval (CI) derived from binary logistic regression model. In order to assess the independent effect of Smurf2, CNKSR2, ER, PR, and HER2 expression on tumor grade, we included all the five factors in the multiple logistic regression model and estimated the Odds Ratio (OR) (adjusted Odds ratio) and their 95% Confidence Interval (CI). All data were obtained by at least three independent experiments and represented as mean ± standard deviation. The limit of statistical significance was set at P < 0.05.
Discussion
Smurf2 E3 ubiquitin ligases plays crucial physiological roles in regulating multiple cellular processes including TGF-β/BMP signaling, cell polarity, cell migration, and mitotic regulation by specifically targeting corresponding cellular substrates for ubiquitination which can either lead to stabilization and or proteasome mediated degradation [
15‐
18]. In the present study we have demonstrated the novel regulation of the multi-functional scaffold protein CNKSR2 by the WW-HECT protein Smurf2. Consistent with the presence of the Smurf2-interacting PY motif in CNKSR2, Smurf2 was found to be immunoprecipitated with CNKSR2 and vice versa, which demonstrated a direct physical interaction between CNKSR2 and Smurf2 (Fig.
1). Interestingly, we observed that though CNKSR2 interacts with Smurf2, it does not bind to Smurf1 (Fig.
1a), which also belong to the Nedd4 family of E3 ubiquitin ligases possessing WW domains in its structure indicating specific Smurfs can recognize CNKSR2 specifically, which also provide evidence of CNKSR2 as a novel class of scaffold proteins regulated by Smurf2.
Functional activity and accessibility of Smurfs towards different substrates is mainly regulated by its subcellular localization [
5,
10]. As shown in Fig.
3 we observed that both Smurf2 and CNKSR2 co-localizes particularly in the cytoplasm, eventhough a recognizable expression of Smurf2 was also observed in the nucleus. Because Smurf2 localization is highly dynamic and as previously reported Smurf2 interacts with TGF-β receptor in distinct endosomal compartments [
19], we hypothesize that Smurf2 may be interacting mainly with the cytoplasmic CNKSR2 pool which may further reinforce the physical interaction between Smurf2 and CNKSR2, leading to its increased stability.
In order to further confirm the physical interaction between Smurf2 and CNKSR2 and to measure the kinetics of interaction, we have described for the first time, a simple and rapid, label-free, real-time, multiplex surface plasmon resonance (SPR) assay [
20,
21] by which Smurf2-CNKSR2 interactions could be studied. We observed that, the results for Smurf2-CNKSR2 interaction fit well with a Langmuir model and give K
D values that indicate a high affinity, and stable interaction between Smurf2 and CNKSR2 (Fig.
6 and Table
1). This is the first time, to our knowledge, that real-time Smurf2-CNKSR2 interactions have been demonstrated using SPR. We believe this approach could be extended to study the effect of a variety of other factors on the interaction.
Protein ubiquitylation can control signaling pathways by regulating the cellular levels of a protein, by controlling its subcellular localization, by preventing accumulation of defective or damaged proteins, and in some instances, by preventing protein degradation [
22]. Together with, our studies explored the consequence of the physical interaction between Smurf2 and CNKSR2, which indicated the presence of a ubiquitinated CNKSR2 in the presence of wild type Smurf2 compared with ligase deficient Smurf2 (Fig.
2c). However, Smurf2 mediated ubiquitination of CNKSR2 leads to stabilization of CNKSR2 instead of degradation as indicated by increased expression of CNKSR2 in Smurf2WT transfected cells. Unlike the interaction of Smurf2 with Mad2 [
23] and EGFR [
24], we observed that CNKSR2 can interact with both wild type and ligase inactive mutant of Smurf2. However, the expression of endogenous CNKSR2 was found to be significantly upregulated in Smurf2WT transfected cells (Fig.
2a, b) which may be due to the editing of CNKSR2 ubiquitination by Smurf2WT, leading to increased stabilization of CNKSR2 expression. Furthermore, we demonstrated that the loss of Smurf2 caused rapid CNKSR2 degradation mediated by enhanced polyubiquitination and proteasomal degradation (Fig.
7e).
The mechanism by which Smurf2 controls CNKSR2 stability remains to be elucidated. We hypothesized that regulation of CNKSR2 by Smurf2 occurs through diverse mechanisms. Smurf2-CNKSR2 interaction exhibit a significantly high binding affinity as observed during SPR analysis and further leads to ubiquitination, which might lead to increased stability of CNKSR2 in Smurf2 overexpressed cells. Eventhough the ubiquitination of CNKSR2 by Smurf2 may increase the stability of CNKSR2; Smurf2 may target an intermediary E3 ligase for degradation to stabilize CNKSR2. This might be responsible for the comparatively increased expression of CNKSR2 in ligase deficient Smurf2C716A transfected cells also compared to the control plasmid transfected cells (Fig.
2b). Consistent with the ligase-independent function of Smurf2 is a previous report that overexpression of wild-type or ligase dead Smurf2 induces senescence [
25]. In addition, Smurf2 may sequester CNKSR2 away from cellular locations where it could encounter its E3 ligase. Alternatively, Smurf2 may instead mask regulatory epitopes for ubiquitination by other E3 ubiquitin ligases. On the other hand, Smurf2 may serve as an adaptor for an unidentified regulator that counteracts with another E3 ligase promoting CNKSR2 degradation.
Because Smurf2 and CNKSR2 were found to be involved in regulating the proliferation and invasion of breast cancer cells [
9], we next examined the functional association of Smurf2 and CNKSR2 in the progression of breast cancer by analyzing the cellular level expression using MCF10 breast cancer progression model cell lines. Interestingly, we observed that both Smurf2 and CNKSR2 showed an integrated expression in breast cells with increasing tumorigenic potential (Fig.
8). Eventhough it is reported that Smurf2 is highly expressed in breast cancer [
14], clinical information regarding the frequency and level of Smurf2 and CNKSR2 expression in breast cancer is incomplete, especially as regards differentiating the importance of Smurf2 and CNKSR2 expression in breast cancer subtypes with increasing tumorigenic potential including normal, usual hyperplasia, fibrocystic changes, fibroadenoma, carcinoma-in-situ (DCIS), and invasive ductal carcinoma (IDC) and its association with the ER, PR, and HER2 status of the breast tissue samples.
Using tissue microarray, which permits rapid high-throughput immunohistochemical analysis of several breast tissue samples simultaneously [
26], Smurf2 and CNKSR2 expression was analyzed in breast cancers as well as in pre-invasive, and pre-neoplastic breast lesions and observed that our results extend the limited findings in previous reports [
14]. Concomitantly, we observed that expression of Smurf2 and CNKSR2 was associated with stage of breast tumor progression, showing increased expression from normal, hyperplastic breast samples, and fibroadenoma to in situ carcinoma to invasive breast carcinoma (Fig.
10). Eventhough we found a statistically significant independent association between Smurf2 and CNKSR2 expression with progressive breast tumor cases by means of descriptive analysis and logistic regression analysis, we further studied the association between expression status of Smurf2 relative to CNKSR2 in non-malignant (Normal/Benign) and malignant tumors (in situ and invasive breast carcinoma) (Additional file
1: Table S1, Tables
2 and
3). Despite the independent association between Smurf2 and CNKSR2 expression with progressive breast tumor stages, the expression pattern of Smurf2 and CNKSR2 showed a significant positive association (
P < 0.001) between each other among non-malignant and malignant tumors (Table
4). These findings agree with the putative role of Smurf2 and CNKSR2 in proliferation and invasiveness of breast cancer.
Breast cancer is a heterogeneous disease with a wide spectrum of clinical, pathologic, and molecular features [
27,
28]. Recently, Liu et al.
, [
29] reported that expression of Smurf2 is associated with the ER, PR, and HER2 status of breast tumors, which instigated us to analyze the crosstalk between Smurf2 and CNKSR2 expression with the hormonal (ER and PR) and HER2 status of breast tissue samples. Interestingly, we observed a statistically significant association between Smurf2 and CNKSR2 expression with the ER, PR, and HER2 status of non-malignant and malignant tumors (Tables
5 and
6). Overall, we observed a comparatively high expression of Smurf2 and CNKSR2 when the expression of ER and PR was low, and HER2 was high. Importantly, Smurf2 and CNKSR2 expression has been associated with features of tumor aggressivity, ER and PR negativity and HER2 overexpression.
In order to assess the physiological significance of Smurf2-CNKSR2 interaction, we stably knocked down Smurf2 using Smurf2 shRNA in MDA-MB-231 cells, and observed that Smurf2 knockdown induced CNKSR2 degradation (Fig.
11), which leads to reduced pro-proliferative and tumorigenic properties of breast cancer cells.