Src kinase inhibits C/EBPδ protein expression in breast epithelial cells. (A) Western analysis of C/EBPδ protein levels in comparison to an antibody to the autophosphorylation site of Src family proteins (Y419 in human Src [pSrc^Y419]) in human mammary epithelial cell lines (Normal, nontumorigenic; Cancer, tumorigenic). (B) Luciferase reporter gene activity in MCF-10A cells transfected with the indicated promoter-reporter constructs (**, P < 0.01; ****, P < 0.0001). (C) Western analysis of the indicated proteins in MDA-MB-231 cells treated with SKI-606 for various times. (D) MDA-MB-468 cells were treated for 24 h with SKI-606 and analyzed as described for panel C. Data are from the same blot with intermediate lanes deleted. (E) Quantitative PCR analysis of C/EBPδ mRNA levels in MDA-MB-231 and MDA-MB-468 cells treated for 24 h with SKI-606. Normalized data are relative to untreated controls. (F) Western analysis of the indicated proteins in MCF-10A cells transfected with a Src-531 expression construct. (G) Quantitative PCR analysis of C/EBPδ expression in MCF-10A cells from the experiment shown in panel F. (H) Western analysis of MCF-10A cells transfected with expression constructs for Src-531, Src^Y529F (chicken-mouse chimera), Src^Y527F (chicken-human chimera), K295M inactivated Src (kinase dead), or vector control. Endogenous Src is not seen because the exposure was optimized for overexpressed protein.

Src kinase mediates proteasome-dependent C/EBPδ degradation. (A) MCF-10A cells were transfected with expression constructs for HA-tagged C/EBPδ and Src-531 as indicated, and 24 h later cycloheximide (100 μM) was added for the indicated times. Data are from the same blot with the intermediate lane deleted. (B) Western analysis of MCF-10A cells with or without Src-531 and lactacystin (1 μM, 6 h) treatment or from MDA-MB-468 cells treated with MG132 (50 μM, 3 h). (C) MCF-10A cells were transfected with expression vectors for HA-tagged ubiquitin (Ub) and Src-531 and 24 h later treated with MG132. Denatured lysates were immunoprecipitated with anti-C/EBPδ antibody and immunoblotted for HA tag and for C/EBPδ as an IP control. Inputs were analyzed as indicated. Numbers on the right indicate the position of MW markers.

SIAH2 is necessary and sufficient for C/EBPδ downregulation. (A) Western analysis of MCF-10A cells transfected with expression constructs for the indicated proteins. (B) MCF-10A cells were transfected with SIAH2, treated with MG132, and analyzed as described in panel A. (C) Western analysis of the indicated cell lines for SIAH1 and SIAH2 expression. (D) Western analysis of MDA-MB-231T (top) and MDA-MB-468 (bottom) cells transfected with the indicated SIAH proteins, which were detected with anti-Flag antibodies. (E) Western analysis of MDA-MB-231T cells after nucleofection of siRNA against SIAH2. (F) Western analysis of MDA-MB-231T cells transfected with a DN-SIAH2 expression construct or nucleofected with siRNA against SIAH2. Anti-Flag was used to detect specifically DN-SIAH2, which has a higher MW than endogenous SIAH2. (G) Western analysis of MCF-10A cells transfected with Src-531. (H) Western analysis of MCF-10A cells cotransfected with Src-531 and/or DN-SIAH2. (I) Western analysis of the indicated cell lines treated with SKI-606 for 24 h. The same lysates were immunoprecipitated with anti-SIAH2 and immunoblotted with anti-phospho-tyrosine. (J) In vitro phosphorylation assay conducted by incubating affinity purified Flag-SIAH2 with (+) or without (−) purified Src-530 in the presence of [³²P]ATP. Samples were resolved by SDS-PAGE, transferred to nitrocellulose, and subjected to autoradiography (left panel). Labeled SIAH2 was digested from the membrane and subjected to phosphoamino acid analysis (PAA; right panel). Expression of the input SIAH2 by Western analysis is also shown. (K) Results from HPLC analysis of SIAH2 with or without individual tyrosine-to-phenylalanine mutations (see panel L) after in vitro phosphorylation assays as in panel J, compared to unphosphorylated WT SIAH2, Edman degradation, and phosphoamino acid analysis. T, trypsin; T+C; trypsin followed by chymotrypsin. (L) Western analysis of MCF-10A cells transfected with expression constructs for the indicated SIAH2 proteins. (M) Western analysis of MDA-MB-231 cells transfected as described for panel L.

SIAH2 interacts with C/EBPδ, leading to polyubiquitination. (A) Western analysis of SIAH2 immunoprecipitates and input samples from MDA-MB-468 cells treated with vehicle or MG132. (B) MDA-MB-468 and MDA-MB-231T cells were transfected with HA-ubiquitin along with DN-SIAH2 as indicated and treated with MG132. Anti-HA immunoprecipitates and input were analyzed for C/EBPδ and input as indicated. The asterisk indicates IgG at 50 kDa. Numbers on the right indicate the position of molecular mass markers. (C) Src-531 kinase-mediated (top) and SIAH2-mediated (bottom) degradation of C/EBPδ requires amino acid K120. Whole-cell lysates of MCF-10A cells cotransfected with expression constructs were analyzed as indicated. (D) K120 of C/EBPδ is not required for interaction with SIAH2. MCF-10A cells were transfected with polyomavirus epitope (Pyo)-tagged wild-type C/EBPδ or HA-tagged K120A-C/EBPδ and Flag-tagged SIAH2 as indicated. Whole-cell lysates were immunoprecipitated and analyzed as indicated. SIAH2 protein was detected with Flag-specific antibodies. Cells had been treated with MG132 to stabilize C/EBPδ. (E) K120 of C/EBPδ is necessary for SIAH2-mediated polyubiquitination. MCF-10A cells were transfected with HA-tagged WT-C/EBPδ or K120A-C/EBPδ along with Myc-tagged ubiquitin and SIAH2, as indicated, and treated with vehicle or MG132. Immunoprecipitates from denatured lysates with anti-HA antibodies or IgG and input were analyzed as indicated. Numbers on the right indicate the positions of molecular mass markers. IB, immunoblotting.

Identification of the SIAH2 interaction domain in C/EBPδ. (A) Six different C/EBPδ-C/EBPβ chimeric proteins, as shown in the schematic and the full-length proteins, were cotransfected with Flag-SIAH2 in MCF-10A cells. Immunoprecipitates with anti-Flag antibody and input were analyzed with a pan-C/EBP and anti-Flag antibody for SIAH2. Cells had been treated with MG132 to stabilize C/EBPδ. TA, transactivation domain; BR, basic region; LZ, leucine zipper. (B) Schematic of the sequence of human C/EBPδ and C/EBPβ aligned with the consensus for SIAH2 interaction domains and two point mutations in C/EBPδ (underlined amino acids match the consensus motif). Immunoprecipitates with anti-Flag antibody and input were analyzed with antibodies specific for C/EBPβ and C/EBPδ and with anti-Flag to detect SIAH2. Cells had been treated with MG132 to stabilize C/EBPδ. (C) Western analysis of MCF-10A cells after transfection with the indicated C/EBPδ expression constructs and Flag-SIAH2.

C/EBPδ mediates tumor cell responses to SKI-606. (A) Western analysis of MDA-MB-468 cells nucleofected with siRNA against C/EBPδ and treated with SKI-606 for 48 h. (B) MDA-MB-231T cells with stable shRNA-mediated depletion of C/EBPδ or controls were treated with SKI-606 for 24 h and analyzed as indicated. (C) Characterization of cell growth of MDA-MB-468 cells in response to various SKI-606 concentrations as indicated. ****, P < 0.0001, by mixed-effects analysis of variance. (D) Analysis of cell growth as described for panel C of stable clones of MDA-MB-231T cells with control shRNA (n = 10) or C/EBPδ-shRNA (n = 9). Data from two independent sets of clones are combined. (E) Cell motility assay. Control and C/EBPδ-silenced MDA-MB-231T cells were grown in culture inserts until confluent at which point cell inserts were removed, and SKI-606 was added for 8 h. Representative images are shown along with quantification of wound closure from four independent experiments. **, P = 0.0021. (F) Transwell Matrigel invasion assay. Quantification of control- and C/EBPδ-silenced MDA-MB-231T cells that had migrated through a Matrigel-coated transwell filter in the presence or absence of SKI-606. ***, P = 0.0007. (G) Soft agar colony assay. Quantification of colonies by control- and C/EBPδ-silenced MDA-MB-231T cells with or without SKI-606 for 10 to 14 days. ****, P < 0.0001; ***, P = 0.0001. (H) MDA-MB-231T cells were nucleofected with siRNA against SIAH2. Cells were lysed for Western analysis while another aliquot was transferred to dishes with culture inserts for cell motility assays as described for panel E. ***, P < 0.0001; *, P = 0.015. (I and J) Cell motility assay as described for panel E of MDA-MB-231T (I) and WT-C/EBPδ-expressing MCF-10A (J) cells transfected with the indicated cyclin D1 expression constructs. Wound closure was assessed 10 h after removal of the inserts. ***, P < 0.001; *, P = 0.02. (K) Western analysis of MCF-10A cells as described for panel J, harvested 24 h after transfection.

C/EBPδ inhibits transformation of MCF-10A cells by Src-531. (A) Western analysis of MCF-10A cells transfected with the indicated expression constructs. Endogenous Src is not seen because the exposure was optimized for detection of overexpressed protein. Ctrl, MCF-10A cells transfected with WT-C/EBPδ and Src-531 as a positive control for pRb (on the same blot). (B) Acini of MCF-10A cells cotransfected with WT- or K120A-C/EBPδ along with Src-531 as indicated and grown on Matrigel for 2 weeks before imaging by light microscopy. (C) Confocal microscopy of acini of MCF-10A cells with stable expression of ectopic HA-tagged WT- or K120A-C/EBPδ or transient expression of A137Y-C/EBPδ and with Src-531. Acini were stained for laminin 5 (green) and with DAPI (blue). (D) Ten-day-old acini of MCF-10A cells as in panel C were stained for active caspase-3 (green) and Ki67 (red) as well as with DAPI (blue). (E) Soft-agar colony assay with MCF-10A cells transfected with the indicated expression constructs and cultured for 14 days. Light microscopy images show representative colonies of Src-531-transfected cells containing ectopic C/EBPδ as indicated.

Schematic summarizing the molecular pathway described in this study. Activated Src kinase, which can be inhibited by SKI-606, induces expression and/or activity (P, phosphorylation) of the SIAH2 E3 ligase. SIAH2 interacts with C/EBPδ and promotes its polyubiquitination (Ub) and proteasomal degradation (19S/20S depict the proteasome). In the presence of C/EBPδ, cyclin D1 becomes downregulated, and cells exhibit less “transformed” features, such as reduced motility and invasion in two-dimensional cultures, less invasive acini and anoikis in three-dimensional cultures, and reduced colony formation in soft agar.