siRNA/short hairpin RNA

siRNA for PTK6 (catalog no. D-003166-06) and MARCH2 (SMART Pool and individual sequences, catalog no. M-006986-02, D-006986-02, D-006986-18) were purchased from Dharmacon and transfection was performed using Oligofectamine (Thermo Fisher Scientific), following the manufacturer's protocol and as described previously (21).

Human PTK6 (NM_005975) short hairpin RNAs (shRNA) were purchased from Sigma-Aldrich and are identified in the figures by the last two digits of the TRC numbers: TRCN0000196912 [12] and TRCN0000021552 [52]. Human MARCH2 shRNA viral particles were purchased from Origene (TL315142V) with the following targeting sequences: (i) CTGTCTGGAGAAGTGGC TTTCCTCATCTA; (ii) GCTACTGCGAGCTGTGCCACACGGAGTTT.

Cell Culture, Transfection, Virus Generation, and Infection

MDA-MB-231 cells were obtained from ATCC (RRID:CVCL_0062) and cultured in RPMI 1640 media supplemented with 10% FBS, HEPES buffer, and Pen/Strep antibiotics. MDA-MB-231/LM2-4 cell lines were obtained from Robert Kerbel (Sunnybrook Institute, Ontario) and cultured in RPMI 1640 media supplemented with 5% FBS and HEPES buffer. MMTV-myc cells were obtained from Eduardo Farias (Mount Sinai) and cultured in DMEM F12 media supplemented with 5% FBS, insulin, and 1% Pen/Strep. All cell lines were routinely tested for Mycoplasma using MycoStrip (Invivogen) and were confirmed negative as of December 2023 using MycoStrip (Invivogen). Cells lines were authenticated using ATCC's short tandem repeat Profiling Service and confirmed to be triple negative by Western blot analysis. All cells were used within 10 passages after thawing.

Plasmids, Transfection, and Virus Production

Plasmids or viral supernatants were obtained from the following: pPGS-SNAIL-Flag (RRID:Addgene_25695); pLEX-MSC-SNAIL Firefly-luciferase (F-Luc) and pMSCV-hygro-Renilla-Luc (Yibin Kang, Princeton University, Princeton, NJ); pLv117-Control-3HA (GeneCopoeia EX-NEG-Lv117); pLv117-MARCH2-3HA (GeneCopoeia EX-K4473-Lv117); pcDNA3-HA-ubiquitin (RRID:Addgene_18712); pCMV-ubiquitin-3HA (Zhen-Qiang Pan, Mount Sinai); pLenti-puro-3HA-ubiquitin (RRID:Addgene_742198); pEB-GST-MARCH2-WT, pEB-GST-MARCH2-A97 and pEB-GST-MARCH2-tm (Ramnik Xavier, Massachusetts General Hospital); pCMV-MARCH2 (Origene SC114251); pCMV-BTRCP (Origene SC110015); pFUO-mCherry-luc (Adolfo Ferrando, Columbia University, New York, NY); pLenti-C-mGFP-MARCHF2 (Origene RC207517L4V).

To generate mutant MARCH2 lentiviral vectors, QuikChange II XL Site-Directed mutagenesis kit (Agilent Technologies) was used according to the manufacturer's protocol. Primers for mutagenesis were: (Forward) GGT GTT AGA TGA GGA AAG CGC CTT CTC CAG ACA GCT CTT and (Reverse) AAG AGC TGT CTG GAG AAG GCG CTT TCC TCA TCT AAC ACC. All mutagenesis was sequence confirmed (Genewiz).

Transfection of plasmids was performed using Lipofectamine 2000 (Thermo Fisher Scientific), following the manufacturer's protocol. To generate lentivirus, HEK293T cells (ATCC, catalog no. CRL-3216) were cotransfected with lentiviral vector and vectors encoding VSV-G envelope and delta8.9, as described previously (21). Retrovirus was generated using GPG-293T cells, as described previously (21, 23). Culture supernatant containing virus was collected daily, filtered and applied to target cells, along with 2 µg/mL hexadimethrine bromide (polybrene) and cells were subsequently placed in antibiotic selection medium.

Ubiquitin Assay, Co-immunoprecipitation, and Western Blot Analysis

MDA-MB-231 or HEK293T cells overexpressing FLAG-SNAIL were transfected with MARCH2 or β-TRCP as well as 3HA-tagged ubiquitin plasmids. After 48 hours, cells were treated with 20 µmol/L MG132 (Sigma-Aldrich) for 4 hours and then lysed in denaturing lysis buffer (1.5% SDS, N-ethylmaleimide and phenylmethylsulfonylfluoride in TBS, pH7.4). Lysates were incubated at 95°C for 15 minutes, followed by immunoprecipitation procedure. For co-immunoprecipitation, FLAG-SNAIL–expressing MDA-MB-231 cells were lysed in NP40 lysis buffer (Boston Bioscience). Cell lysates (0.5 µg total protein in 1 ml of lysis buffer) were precleared with mouse IgG-conjugated agarose (Sigma-Aldrich), followed by incubation with anti-flag EZView Red M2 affinity gels (catalog no.: F2426, Sigma-Aldrich). After 1-hour incubation at room temperature, the beads were washed three times and incubated at 95°C for 10 minutes in 2x sample buffer. Eluates were resolved on NuPAGE 4%–12% Bis-Tris protein gels (Thermo Fisher Scientific) and transferred onto polyvinylidene difluoride membrane. Western blot analysis was performed with primary antibody according to manufacturer's protocol. Quantification of band intensity was performed using ImageJ software.

The in vitro ubiquitination reactions were performed using 2 µmol/L recombinant human SNAIL protein (Origene TP304581), 3 µmol/L recombinant human March2 E3 ligase (Origene TP307517), 100 µmol/L Ubiquitin (R&D Systems U-100H-10M), 100 nmol/L recombinant E1(UBE1, Biotechne E-305), 500 nmol/L recombinant E2 (UBE2D3, Biotechne E2-627), 1X ATP Energy Regeneration buffer (Enzo BML-EW9810-0100) in a 25 µL reaction buffer containing 50 mmol/L Hepes, pH 8.0, 50 mmol/L NaCl, and 1 mmol/L TCEP. After 1 hour at 37°C, reactions were stopped with 4X Laemmli SDS-Sample Buffer and boiled at 95°C for 5 minutes.

Proximity Ligase assay with Duolink

Proximity ligase assay (PLA) was performed using Duolink In Situ Kit (Sigma-Aldrich) according to the manufacturer's protocol. MDA-MB-231 cells cultured on coverslips in 12-well plates were fixed in 4% paraformaldehyde/PBS for 20 minutes and blocked/permeabilized with 5% BSA/0.1% Triton X in PBS for 1 hour at room temperature. Cells were incubated with primary antibody overnight at 4°C, washed and incubated with PLA probe anti-rabbit PLUS and anti-mouse MINUS Donkey IgG (RRID: AB_2713942) at 37°C for 1 hour. After wash, ligase reaction was performed followed by amplification with Red detection reagent. AlexaFluor 488 Phalloidin (Thermo Fisher Scientific) was used for counter staining actin and mounted with VECTASHIELD with DAPI (Vector Laboratories). Images were captured by a SP5 Leica confocal microscope with a fixed setting for PLA signal (red). The number of PLA signals per cells (n = 90 to 300 cells) was calculated using ImageJ software.

IHC

Tumor tissue microarrays were obtained from Icahn School of Medicine at Mount Sinai Department of Pathology. Mouse tumor tissues were fixed in 4% paraformaldehyde and embedded in paraffin (FFPE) or directly frozen in optimal cutting temperature compound. FFPE slides were deparaffinized in xylene substitute (Sigma-Aldrich) and hydrated in ethanol/H₂O solution (100%, 90%, 75%, 50% ethanol/H₂O and 100% H₂O). Slides were incubated in citrate-based antigen retrieval buffer (Sigma-Aldrich) in a steamer and blocked with BLOXALL (Vector Laboratories) for 15 minutes at room temperature. Slides were incubated with primary antibody, washed, then incubated with ImmPRESS horseradish peroxidase–conjugated secondary antibody (Vector Laboratories), followed by ImmPACT 3,3′-Diaminobenzidine (DAB) peroxidase substrate (Vector Laboratories). Peptide competition assays were performed using PrEST antigens (PTK6: catalog no. APREST78020, Sigma-Aldrich). The slide images were captured by Panoramic 250 FLASH II digital scanner (Perkin Elmer) or Leica DM550 fluorescence microscope and analyzed by CaseViewer (3DHISTECH Ltd.) or LAS AF software. The staining intensity was scored by pathologists as low (negative or +) or high (++ or +++).

siRNA Screening

MDA-MB-231 cells expressing both pLEX-SNAIL-F-luc and pMSCV-R-luc (2.5 × 10⁴ cells per well) were seeded into 96-well plates. siRNA transfection was performed using Oligofectamine and siRNA library (ubiquitin conjugate sets 1–3, Dharmacon, catalog nos. G-005615-01, G-005625-01, G-005635-01) or siRNA controls (OTC non-targeting control and siGENOME non-targeting control #5, Dharmacon). Media was replaced with DMSO or P21d-containing media. After 24 hours, the dual luciferase assay (DLA) was performed with Dual-Glo Luciferase Reagent and GloMAX Discover System (Promega). The F-luc/R-luc (Renilla-luciferase) values of P21d-treated wells were normalized to DMSO-treated wells. SNAIL recovery rate by siRNA gene X was calculated by follow (control in each plate). Recovery rate (fold change) = [Snail reduction % by siRNA X)]/[Snail reduction % by siRNA Control #5). For secondary screening, Western blot analysis was performed using antibodies against FLAG (RRID:AB_2572291) and GAPDH. Tertiary screening was performed using DLA of parental MDA-MB-231 cell transfected with top candidate siRNA.

Transwell Migration Assay and Boyden Chamber Invasion Assay

Transwell migration assay was performed as described previously (21). Migrated MDA-MB-231 cells were stained with Protocol HEMA3 (Thermo Fisher Scientific), and cell number was counted by ImageJ software. Invasion assay was performed with 96-well CytoSelect 96-well Cell Invasion assay Basement Membrane kit (Cell BioLabs)

Anoikis Assay

MDA-MB-231 cells were incubated in suspension condition for 24 or 48 hours at 5% CO₂ at 37°C. The cells were washed in PBS and stained with FITC-AnnexinV/PI (propidium iodide) staining kit (BD Biosciences). The cells were then analyzed by flow cytometer (BD FACSCanto or LSRFortessa) and De Novo FCS Express software.

Three-dimensional Matrigel Culture Assay

For three-dimensional (3D) culture assay, 96-well plates were coated with 55 µL of growth factor reduced Matrigel Matrix (Corning). A total of 3,000 cells were seeded on top of the solidified Matrigel. After 5 days, the cells were visualized, and the cell viability was measured by CellTiter-Glo 3D Viability Assay (Promega).

Quantitative RT-PCR

Total RNA was isolated using RNEasy Mini Kit (Qiagen). Synthesis of cDNA was carried out using Superscript IV (Life Technologies) using the manufacturer's instructions. Total RNA (1 µg) was reverse transcribed with SuperScript IV (Invitrogen) using random hexamers (Promega) according to the manufacturer's protocol. The resulting cDNA was diluted 1/1,000 and 1 µL cDNA was used as a PCR template. Primers used to amplify March2, and the loading control β-actin are listed below. qPCR reactions were assembled using Power SYBR Green PCR Master Mix (Life Technologies) according to the manufacturer's protocols and quantified using Quant Studio 7 Pro (Applied Biosystems) qPCR machine. Expression fold changes were calculated using the Ct method. Primers used were: MARCH2 (forward) CAGAGGAGACACCAGTATGAATG, MARCH2 (reverse) GTTGAAGTGGAAGTGTTGAAGTG, β-actin (forward) CACCATTGGCAATGAGCGGTTC, β-actin (reverse) AGGTCTTTGCGGATGTCCACGT.

Patient Samples

Samples from patients treated at the Mount Sinai Hospital were collected for patient-derived xenograft (PDX) generation and expansion. Written informed consent was obtained for all subjects on Icahn School of Medicine at Mount Sinai (ISMMS) Institutional Review Board (IRB)-approved protocol (HSM# 14-00330) prior to the procedure at which the specimen was obtained. The studies were conducted in accordance with the Belmont Report and U.S. Common Rule and approved by the ISMMS IRB.

Metastatic Lung Colonization and Spontaneous Metastases Assays

All animal procedures were conducted in compliance with the guidelines of the Institutional Animal Care and Use Committee of Icahn School of Medicine at Mount Sinai under approved protocol LA12-00113. In our experimental metastasis models, 8–10 weeks old female NSG mice (Jackson Laboratory, RRID: IMSR_JAX:005557) were randomized and injected intravenously (tail vein) with 1 × 10⁶ cells/100 µL of either MDA-MB-231/Luciferase cells overexpressing control vector (N = 8), wild-type (WT) MARCH2 (N = 9) or RING domain mutant (W97A) MARCH2 (N = 12). Lung metastases were imaged using in vivo imaging system (IVIS) 7 and 14 days after injection (Translational and Molecular Imaging Institute at Mount Sinai). During IVIS imaging, mice were anesthetized with inflowing isoflurane (2%–2.5%) in 2 to 2.5 L/minute oxygen. After anesthesia, 10 µL/g mouse body weight of Luciferin solution (Revvity Inc.) was intraperitoneally injected into mouse. IVIS images taken 15 minutes after Luciferin injection were used for image analysis. The IVIS images were analyzed using Living image software (Version 4.7.3, Revvity Inc. RRID: SCR_014247). A 3.3 × 3.3 cm² area in the lungs was set as region of interest (ROI) for each mouse, and then total photons/second in the ROI was obtained. Statistical difference between the groups was assessed by Student t test, and the P value less than 0.05 was regarded as statistically significant.

For the spontaneous metastasis models, 8–10 weeks old female NSG mice were randomized to be injected with 5 × 10⁵ MDA-MB-231/LM2-4 Luciferase cells overexpressing control vector (N = 10) or WT MARCH2 (N = 10) into the left fourth mammary fat pad. Primary tumor growth was monitored by caliper measurements every day beginning the seventh day after implantation and tumor volume was calculated using the formula: ½ (width² × length). Lung metastases were imaged 12, 15, and 18 days after injection using IVIS and analyzed as above. During IVIS imaging, the primary tumor site of the mice was shielded with opaque black paper.

Data Analysis and Statistical Analysis

All values were expressed as mean ± SEM unless indicated otherwise, and the number of samples (n) was shown. Experiments were repeated minimum of three, unless indicated otherwise. The statistical significance of differences between groups was determined by Student t test with P < 0.05 considered significant (*, P < 0.05; **, P < 0.01; ***, P < 0.001). F-test was also performed to confirm similar variances of groups. Two-way ANOVA with the Bonferroni post hoc test was performed on the tumor growth experiments. All statistical analyses were performed using GraphPad Prism v6.0.

Survival Curves

PTK6 Inhibition Promotes Ubiquitination and Degradation of SNAIL

PTK6 is activated and autophosphorylated in patient with breast cancers and breast cancer cell lines (21, 25). Downregulation of PTK6 or inhibition of its kinase activity promotes degradation of SNAIL in a proteasome-dependent manner and partially reverses EMT of TNBC cells (21). Treatment with PTK6 kinase inhibitor (P21d or 4f) or PTK6 shRNA downregulates SNAIL expression in human and mouse TNBC cells (Fig. 2A; Supplementary Fig. S2; ref. 21). Treatment with P21d promotes ubiquitination of SNAIL in triple-negative MDA-MB-231 cells, as assessed biochemically and by immunofluorescence using Duolink PLAs (Fig. 2B–D). With PLA, a signal is detected only when there is close interaction or colocalization of molecules. PLA can also be adapted to detect posttranslational modifications on individual molecules such as ubiquitination and SUMOylation (26, 27).

Open in a separate window

FIGURE 2

PTK6 inhibition increases ubiquitination and degradation of SNAIL in TNBC. A, PTK6 inhibitor (P21d) treatment or PTK6 shRNA downregulates SNAIL in TNBC cells. Human MDA-MB-231 or mouse MMTV-myc cells were treated with P21d for 24 hours and SNAIL expression was assessed by Western blot analysis. Lysates of MDA-MB-231 cells expressing vector control or either of two independent PTK6 shRNA vectors were assessed for SNAIL expression. B, MDA-MB-231 cells overexpressing HA-ubiquitin and Flag-SNAIL were treated with P21d for 8 hours with MG132 cotreatment for the last 4 hours. Lysates were immunoprecipitated with anti-FLAG antibody and probed with anti-HA or FLAG antibody. The result is representative of three independent experiments. C, Representative images of Duolink PLA using MDA-MB-231 cells coexpressing HA-ubiquitin and Flag-SNAIL. Cells were treated with P21d (5 µmol/L) for 24 hours with MG132 cotreatment for the last 4 hours and incubated with antibodies against HA (Ubiquitin) and FLAG (SNAIL) for PLA (red) and Phalloidin for F-actin and counterstaining. PLA signals (red)/cell were quantitated using ImageJ (left). Scale bar, 15 µm. D, Representative images of PLA using MDA-MB-231 cells overexpressing HA-ubiquitin only. Cells were treated with P21d for 24 hours and with MG132 cotreatment for the last 4 hours and incubated with antibodies against HA (Ubiquitin) and SNAIL for PLA (red) and Phalloidin for F-actin counterstaining. PLA signals/cell were quantitated using ImageJ (left). Scale bar, 15 µm. E, Representative images of PLA using MDA-MB-231 cells overexpressing HA-ubiquitin only that were transfected with control or PTK6 siRNA for 48 hours. Cells were treated with MG132 for the last 4 hours and incubated with antibodies against HA (Ubiquitin) and SNAIL for PLA (red) and Phalloidin for F-actin counterstaining. PLA signals/cell were quantitated using Image (left). Scale bar, 15 µm. All the PLA signals were counted by ImageJ with a fixed setting. Data represent mean ± SEM. *, P = 0.05; ***, P = 0.0001 by Student t test.

P21d treatment increases ubiquitination of FLAG-tagged SNAIL immunoprecipitated from cells MDA-MB-231 cells coexpressing HA-Ubiquitin (Fig. 2B). In addition to these biochemical assessments, PLA was used to quantitate the levels of SNAIL ubiquitination by immunofluorescence. We validated PLA as a methodology to detect SNAIL ubiquitination in cells overexpressing an established SNAIL E3 ligase, β-TRCP, using antibodies for SNAIL and Ubiquitin (16). β-TRCP overexpression increases ubiquitination of immunoprecipitated FLAG-SNAIL (Supplementary Fig. S3A). In β-TRCP–overexpressing cells, increased ubiquitin/SNAIL PLA signals were detected using two independent mixes of anti-HA (for ubiquitin) and anti-SNAIL antibodies (Supplementary Fig. S3B and S3C). Furthermore, these PLA signals were suppressed by GSK3β inhibitor (TWS119) treatment, as GSK3β phosphorylation is a requisite step for β-TRCP–dependent ubiquitination (Supplementary Fig. S3D).

PTK6 kinase activity inhibition or PTK6 downregulation increased SNAIL ubiquitination, as assessed by PLA. PTK6 kinase inhibitor P21d treatment increased SNAIL:Ubiquitin signals in cells overexpressing FLAG-SNAIL and HA-Ubiquitin by 2.3-fold (Fig. 2C). A 6.0-fold increase in endogenous SNAIL:Ubiquitin PLA signals was detected in MDA-MB-231 cells expressing only HA-Ubiquitin treated with P21d compared with vehicle control (Fig. 2D). Finally, increased endogenous SNAIL:Ubiquitin PLA signals were detected in MDA-MB-231 cells expressing HA-Ubiquitin transfected with PTK6 siRNA compared with control siRNA (Fig. 2E). These data collectively support an increase in SNAIL ubiquitination following PTK6 inhibition or downregulation.

Novel PTK6-dependent Regulators of SNAIL Ubiquitination and Degradation

SNAIL ubiquitination and proteasome-dependent degradation downstream of PTK6 inhibition is independent of known mechanisms. We previously reported that established SNAIL E3 ligases such as GSK3β/β-TRCP, FBXL5, FBXL14, and FBXO11 have no role in PTK6-dependent regulation of SNAIL; downregulation of these E3 ligases did not impact SNAIL levels in P21d-treated TNBC cells in our previous report (21). To identify the novel E3 ligase(s) responsible for PTK6-dependent SNAIL regulation, an siRNA library of 588 E3 ligases and related proteins was screened to identify those siRNAs that restored Snail levels in P21d-treated TNBC cells. This strategy would identify molecules that are critical for P21d-induced, but not basal, SNAIL ubiquitination and degradation (Fig. 3A).

Open in a separate window

FIGURE 3

MARCH2 is a novel PTK6-regulated SNAIL E3 ligase. A, Schema of siRNA screening strategy used to identify PTK6-regulated SNAIL E3 ligases. B, MARCH2 siRNA (SMART Pool) identified as top candidate siRNA that restores SNAIL expression in P21d-treated MDA-MB-231/SNAIL- Luciferase cells by Western analysis. C, MARCH2 siRNA (SMART POOL or individual) prevents P21d-induced SNAIL downregulation in MDA-MB-231 cells.

To enable primary screening using luciferase assays, MDA-MB-231 cells were engineered to express SNAIL protein fused to F-luc in addition to R-luc, which serves as control for changes in viability (gift of Yibin Kang, Princeton). P21d treatment of these cells was confirmed to decrease levels of exogenously expressed SNAIL-F-luc as assessed by Western blot analysis and Dual-glo luciferase assay (Supplementary Fig. S4A). SNAIL-F-luc/R-luc–expressing MDA-MB-231 cells were transfected with the libraries of SMART Pool siRNAs. Beginning 24 hours after transfection, cells were treated with P21d or DMSO vehicle control for 24 hours and the dual Luciferase reporter assay (Promega) was used to assess levels of SNAIL-F-Luc and R-Luc. From the primary screen, 26-candidate siRNAs were identified that prevented downregulation of SNAIL-F-Luc in P21d-treated cells but did not affect basal SNAIL-F-Luc levels in DMSO-treated cells. The effect of these 26-candidate siRNAs on SNAIL-F-luc protein expression in P21d-treated cells was confirmed by Western blot analysis (secondary screening; Fig. 3B). Quantification of the immunoblot intensity shows that 13-candidate siRNAs led to recovery of SNAIL-F-Luc levels in P21d-treated cells (Fig. 3B; Supplementary Fig. S4B).

The 13 prioritized SMART Pool siRNAs were assessed for effects on endogenous SNAIL levels in P21d- or DMSO-treated MDA-MB-231 cells (tertiary screen) and MARCH2 emerged as the top PTK6-regulated SNAIL E3 ligase candidate (Fig. 3C). Transfection with either MARCH2 SMART Pool siRNA or either of two individual MARCH2 siRNA (02 and 18) restores SNAIL expression in P21d-treated cells. These results support MARCH2 as a novel PTK6-regulated E3 ligase candidate that promotes proteasome-dependent SNAIL degradation.

MARCH2 is a Novel SNAIL E3 Ligase Regulated by PTK6 Kinase Activity

MARCH2 is a member of the RING family of E3 ligases with an increasing number of target substrates. The MARCH family of proteins, first discovered as proteins encoded by Kaposi sarcoma–associated herpes viruses, now consists of 11 members that contain a variable number of putative transmembrane domains and an N-terminal cytoplasmic RING-CH domain with E3 ubiquitin ligase activity (28–33). MARCH2 substrates identified thus far include β2 adrenergic receptor, cystic fibrosis transmembrane regulator (CFTR), DLG1, syntaxin-6, and Disheveled (34–38).

Overexpression of WT MARCH2, but not RING domain–mutated MARCH2 (W97A), downregulates endogenous SNAIL expression in MDA-MB-231 cells (Fig. 4A). MARCH2 expression did not affect levels of other transcriptional drivers or markers of EMT, such as Zeb1, Slug, or Vimentin (Supplementary Fig. S5). HA-tagged MARCH2 coprecipitates with FLAG-SNAIL, supporting a physical interaction between these two molecules (Fig. 4B). Overexpression of WT MARCH2, but not RING domain–mutant (W97A) or C-terminal domain–mutant MARCH2 (35), increased ubiquitination of immunoprecipitated FLAG-SNAIL (Fig. 4C). In in vitro ubiquitination assays, recombinant SNAIL is polyubiquitinated in the presence of E1, E2, and MARCH2, further confirming MARCH2’s E3 ligase activity toward SNAIL (Fig. 4D).

Open in a separate window

FIGURE 4

MARCH2 promotes ubiquitination and downregulation of SNAIL. A, HA-tagged WT, but not RING domain–mutant (W97A), MARCH2 downregulates SNAIL levels in MDA-MB-231 cells. B, MARCH2 coprecipitates with SNAIL in MDA-MB-231 cells coexpressing FLAG-SNAIL and HA-MARCH2. C, WT, but not RING domain– or transmembrane domain–mutant, MARCH2 ubiquitinates SNAIL. FLAG-Snail was immunoprecipitated from MDA-MB-231 cells coexpressing FLAG-Snail, HA-Ubiquitin and GST-tagged MARCH2 constructs. Immunoprecipitates were blotted with antibodies to HA to assess ubiquitination. D, MARCH2 ubiquitinates SNAIL in in vitro ubiquitination assay. Reactions were performed with recombinant SNAIL (Myc-tagged) in the presence of ubiquitin, recombinant E1 (UBE1), recombinant E2 (UBE2D3) and/or recombinant MARCH2.

PTK6 inhibition increases the association of MARCH2 with SNAIL. In cells coexpressing FLAG-SNAIL and HA-MARCH2, P21d treatment increased the amount of MARCH2 that coprecipitated with FLAG-SNAIL (Fig. 5A). With P21d treatment, an increase in PLA signals using antibodies for MARCH2 and SNAIL was also observed supporting an increase in association between endogenous SNAIL and endogenous MARCH2 in TNBC cells (Fig. 5B).

Open in a separate window

FIGURE 5

PTK6 inhibition enhances MARCH2:SNAIL association. A, MDA-MB-231 cells expressing FLAG-Snail and HA-MARCH2 (WT) were treated with vehicle control or P21d for 4 or 18 hours in the presence of MG132. MARCH2–SNAIL interaction was assessed by immunoblotting FLAG immunoprecipitates with anti-HA antibody. B, Endogenous MARCH2–SNAIL interaction following P21d treatment of MDA-MB-231 was assessed by Duolink PLA. Cells were treated with DMSO or P21d for 8 hours and MG132 cotreatment for the last 4 hours. Cells were stained with antibodies to Snail and MARCH2. PLA signals/cell were quantitated using ImageJ (left). Scale bar, 15 µm.

Collectively, these data support MARCH2 as a novel SNAIL E3 ligase. The RING domain of MARCH2 is critical for SNAIL ubiquitination and its association with SNAIL is increased by PTK6 inhibition, leading to SNAIL downregulation.

The authors thank Dr. Yibin Kang (Princeton University) for providing pLEX-F-luc-SNAIL and pMSCV-hygro-Renilla-luc plasmids for siRNA screening. The authors also thank Dr. Ramnik Xavier for MARCH2 mutant plasmids and Dr. Zhen-Qiang Pan for HA-ubiquitin plasmids and advice about ubiquitination assays. The authors acknowledge cell line providers, Dr. Robert S. Kerbel for MDA-MB-231/LM2-4 cells and Dr. Eduardo Farias for MMTV-myc cells. The authors acknowledge the Tisch Cancer Institute Biorepository and Histopathology Core, supported by the Tisch Cancer Institute Cancer Center Support Grant (P30 CA196521), and the Translational and Molecular Imaging Core at Icahn School of Medicine at Mount Sinai.

This work was supported in part by Susan G. Komen Postdoctoral Fellowship (PDF17488499, to K. Ito), Susan G. Komen for the Cure Career Catalyst Award (CCRCR17499360, to H.Y. Irie), American Cancer Society (RSG-18-044-01-LIB, to H.Y. Irie), the Breast Cancer Research Foundation (E. Port and H.Y. Irie) and the New York State Peter T. Rowley Breast Cancer Research award (C36612GG, to H.Y. Irie).