2.2. SOSTDC1 Promotes TNBC Malignancy and Olaparib Resistance

SOSTDC1 function was assessed by silencing or overexpressing in TNBC cells. Stable‐SOSTDC1‐knockdown cell lines were established for SUM149 and HCC1937 (Figure S2A, Supporting Information). Stable‐SOSTDC1‐overexpression cell lines were established for SUM159 and MDA‐MB‐231 (Figure S2B, Supporting Information). SOSTDC1 expression was found to be positively related to TNBC cell proliferation (Figure S2C–F, Supporting Information), mammosphere formation (Figure 1G,H; Figure S2G,H, Supporting Information), BTIC population enrichment (Figure 1I,J, Figure S2I–L, Supporting Information), and Olaparib resistance (Figure 1K–L; Figure S2M, Supporting Information). Importantly, SOSTDC1‐knockdown rendered the BTIC population sensitive to Olaparib as judged by a decrease in absolute BTIC numbers (Figure 1M,N; Figure S2N, Supporting Information). Moreover, in vivo xenograft experiments showed that SOSTDC1 knockdown significantly inhibited tumor growth (Figure 1O; Figure S2O), meanwhile SOSTDC1 overexpression promoted tumor outgrowth (Figure 1P; Figure S2P, Supporting Information). BTIC population was decreased in SOSTDC1‐knockdown xenografts but increased in SOSTDC1‐overexpressing xenografts (Figure 1Q–R; Figure S2Q–R, Supporting Information).

To investigate whether BRCA1 status affected the function of SOSTDC1, we established SOSTDC1‐overexpressing SUM149 cells (BRCA1‐mutant) and SOSTDC1‐knockdown SUM159 cells (BRCA1‐wildtype, Figure S3A,B, Supporting Information). Consistent with our results above, SOSTDC1 overexpression made SUM149 cells more insensitive to Olaparib and SOSTDC1 knockdown sensitized SUM159 cells to Olaparib (Figure S3C, Supporting Information), regardless of BRCA1 status. These results indicated that cell sensitivity to Olaparib regulated by SOSTDC1 was not entirely dependent on BRCA1 status.

The therapeutic effect of Olaparib was also evaluated in SOSTDC1‐knockdown xenografts. We found that SOSTDC1‐knockdown and treatment with Olaparib significantly decreased tumor volume (Figure 1S), the BTIC population (Figure 1T; Figure S2S, Supporting Information), and absolute BTIC numbers (Figure 1U). These results demonstrated that SOSTDC1 mediated BTIC maintenance and Olaparib resistance in TNBC. Targeting SOSTDC1 might increase the therapeutic effect of PARPi.

2.3. Importin‐α‐Mediated SOSTDC1 Nuclear Translocation Promotes BTIC Maintenance and Olaparib Resistance

Since SOSTDC1 has been reported to function as a secreted protein, we treated the parental BC cells with culture supernatants of SOSTDC1‐overexpressing cells (Figure S4A, Supporting Information). However, the culture supernatants had no significant effect on cell proliferation (Figure S4B,C, Supporting Information), BTIC enrichment (Figure S4D, Supporting Information), or Olaparib resistance (Figure S4E, Supporting Information), indicating that the observed SOSTDC1 biological functions (those described above) are dependent on intracellular expression.

Intriguingly, SOSTDC1 could be detected in both cytoplasm and cell nucleus, and nuclear SOSTDC1 expression was significantly upregulated after Olaparib treatment in xenografts (Figure  2A), which was further confirmed in TNBC cell lines (Figure 2B,C) as well as in U2OS cells (Figure S5A, Supporting Information). Secreted proteins have been shown to transport to the nucleus to regulate some important processes such as cell proliferation and differentiation, DNA replication, DNA repair, etc.^{[ 46 ]} However, there were few reports about the nuclear translocation of secreted proteins involved in the WNT and BMP signal pathway. We here confirmed the nuclear translocation of SOSTDC1 in parental TNBC cells (Figure 2D). Moreover, we observed a higher level of nuclear SOSTDC1 protein in BTICs (Figure 2E,F). These results suggested a biological association of nuclear SOSTDC1 with BTIC maintenance and Olaparib resistance.

Open in a separate window

Figure 2

Importin‐α‐mediated SOSTDC1 nuclear translocation promotes BTIC maintenance and Olaparib resistance of TNBC cells. A) Immunohistochemistry analysis of nuclear SOSTDC1 expression in SUM149 xenografts treated with DMSO or Olaparib. Representative images were shown and H‐Scores were calculated. Data were presented as mean ± SEM. Significance was calculated using unpaired Student's two‐sided t‐tests. ^** p < 0.01. Scale bar, 100 µm. B) SOSTDC1‐overexpressing SUM159 cells were treated with DMSO or 5 µmol L⁻¹ Olaparib for 12 h before immunofluorescence analysis using anti‐FLAG antibody. Representative images were shown and percentages of SOSTDC1 nuclear translocation were calculated. Data were presented as mean ± SEM, following unpaired Student's two‐sided t‐tests. ^*** p < 0.001. Scale bar, 5 µm. C) Time‐dependent analysis of the expression of cytoplasmic and nuclear SOSTDC1 by western blotting after 1 µmol L⁻¹ Olaparib treatment in SOSTDC1‐overexpressing SUM159 cells. D) Analysis of the endogenous expression of cytoplasmic and nuclear SOSTDC1 by western blotting in SUM149 cells. E) Analysis of the expression of cytoplasmic and nuclear SOSTDC1 by western blotting in sorted BTICs (B) and the rest non‐BTICs (R) of SOSTDC1‐overexpressing SUM159 cells. F) Immunofluorescence analysis in sorted BTICs (B) and the rest non‐BTICs (R) of SUM149 cells using anti‐SOSTDC1 antibody. Representative images were shown and the fluorescence intensities of SOSTDC1 and nuclear SOSTDC1 were measured. Data were presented as mean ± SEM. Significance was calculated using unpaired Student's two‐sided t‐tests. ^* p < 0.05; ^*** p < 0.001. Scale bar, 20 µm. G) Schematic of the location and sequence of the predicted SOSTDC1 nuclear localization sequences. H,I) Wild‐type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing SUM159 cells were treated with DMSO or 5 µmol L⁻¹ Olaparib for 12 h. Immunofluorescence staining using anti‐FLAG antibody H) analyzed the percentage of SOSTDC1 nuclear translocation (Data were presented as mean ± SEM, following unpaired Student's two‐sided t‐tests. ^** p < 0.01. Scale bar, 5 µm) and western blotting I) analyzed the expression of cytoplasmic and nuclear SOSTDC1. J) Mammosphere formation assays of wild‐type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing SUM159 cells. Representative images were shown and data were presented as mean ± SEM. Statistical significance was determined using the one‐way ANOVA. ^** p < 0.01; ns, not significant. Scale bar, 100 µm. K) The percentage of BTIC population in wild type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing SUM159 cells. Data were presented as mean ± SEM, following the one‐way ANOVA. ^* p < 0.05; ns, not significant. L) Survival analysis of wild type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing SUM159 cells in response to Olaparib; n  = 3 independent biological samples. M) Co‐IP assay detected the interaction between SOSTDC1 and protein members of the importin‐α. 293T cells were co‐transfected with SOSTDC1 and the indicated FLAG‐tagged KPNA vectors for 36 h, then exposed to 1µµ Olaparib for 4 h. The cell lysates were subjected to co‐IP experiments with anti‐FLAG magnetic beads; IB, immunoblot. N–P) mRNA expression of KPNA1 N), KPNA2 O) or KPNA3 P) in SOSTDC1‐overexpressing SUM159 cells with KPNA1 and KPNA2 and KPNA3 knockdown separately or simultaneously. Data were presented as mean ± SEM, following the one‐way ANOVA. ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001. Q) Analysis of the expression of cytoplasmic and nuclear SOSTDC1 after 1 µmol L⁻¹ Olaparib treatment for 12 h by western blotting in SOSTDC1‐overexpressing SUM159 cells with KPNA1, KPNA2, KPNA3 knockdown separately or simultaneously. R) Analysis of the expression of cytoplasmic and nuclear SOSTDC1 after DMSO or 1 µmol L⁻¹ Olaparib treatment for 12 h by western blotting in SOSTDC1‐overexpressing SUM159 cells with KPNA1, KPNA2, and KPNA3 knockdown.

Bioinformatic analysis identified a conserved nuclear localization signal (NLS) sequence within SOSTDC1 (Figure 2G). Deletion of the NLS not only significantly blocked the nuclear translocation of SOSTDC1 (Figure 2H,I), also reversed full‐length SOSTDC1‐induced mammosphere formation (Figure 2J), BTIC enrichment (Figure 2K; Figure S5B, Supporting Information) and Olaparib resistance (Figure 2L). It is widely accepted that importin‐α plays an indispensable role in protein nuclear import by recognition of the NLS.^{[ 47 ]} We found that SOSTDC1 may interact with protein members of the importin‐α family karyopherin (KPNA), including KPNA1, KPNA2, and KPNA3 (Figure S5C, Supporting Information). We verified the interactions upon Olaparib treatment (Figure 2M). Furthermore, we observed nuclear import of SOSTDC1 was modestly inhibited after knocking down KPNA proteins separately (Figure S5D,E, Supporting Information), but was clearly inhibited after knocking down these three importins simultaneously with or without Olaparib treatment (Figure 2N–R; Figure S5F, Supporting Information). Collectively, these results demonstrated importin‐α‐mediated SOSTDC1 nuclear translocation promotes BTIC maintenance and Olaparib resistance of TNBC cells.

2.4. Nuclear SOSTDC1 Promotes HR Repair

To further explore the biological significance of SOSTDC1 nuclear translocation and its association with PARPi resistance, we analyzed the transcriptional profiles of SUM159‐CTRL and SUM159‐SOSTDC1 cells. A series of DNA repair‐related genes were upregulated in SOSTDC1 overexpressing cells (Figure  3A). Gene set enrichment analysis (GSEA) and KEGG pathway analysis indicated that the HR repair pathway was significantly upregulated with SOSTDC1 overexpression (Figure 3B; Figure S6A, Supporting Information). However, we did not observe a consistent dysregulation of several classic HR regulator genes, including BRCA1, BRCA2, BARD1, and RAD51 (Figure S6B,C, Supporting Information). We examined γH2AX foci formation, which is a sensitive indicator of DSBs arising from uncompleted repair of base damage, in SOSTDC1‐overexpressing cells exposed to ionizing radiation (IR). In contrast to control cells, at 12 h after IR, SOSTDC1‐overexpressing cells exhibited shortened and diminished γH2AX foci formation (Figure 3C). Moreover, we observed that upon Olaparib treatment, SOSTDC1 overexpressing cells showed decreased γH2AX foci (Figure 3D), while SOSTDC1 knockdown resulted in an elevated level of γH2AX foci (Figure S6D,E, Supporting Information), suggesting that SOSTDC1 contributes to Olaparib induced DNA damage repair. In particular, we observed that SOSTDC1‐overexpressing cells showed better viability (Figure S6F,G, Supporting Information) and decreased γH2AX foci with Cisplatin treatment, indicating that SOSTDC1 may also function in the HR process in response to other DNA damaging drugs (Figure S6H, Supporting Information). Next, we utilized the integrated reporter assay for HR and non‐homologous end‐joining (NHEJ)^{[ 48 ]} to examine how SOSTDC1 promotes DNA repair. We observed significantly activated HR repair in SOSTDC1 overexpressing cells (Figure 3E; Figure S6I, Supporting Information), while only a minor increase in NHEJ efficiency (Figure 3F; Figure S6J, Supporting Information), declaring an important role of SOSTDC1 in the HR pathway. Importantly, deleting of the NLS recovered SOSTDC1 suppressed γH2AX foci formation upon Olaparib treatment (Figure 3G; Figure S6K, Supporting Information) and decreased the HR efficiency (Figure 3H; Figure S6L, Supporting Information). In addition, SOSTDC1 knockdown did not affect cell‐cycle distribution (Figure S7A, Supporting Information). Compared with PARP inhibitor or Cisplatin, SOSTDC1 knockdown did not impact cell sensibility to Docetaxel (DTX), which stops the growth of cancer cells by interfering with microtubules and blocking cell division (Figure S7B, Supporting Information). Collectively, these results suggested that nuclear SOSTDC1 promoted DNA repair, specifically HR repair, upon genotoxic stress.

Open in a separate window

Figure 3

Nuclear SOSTDC1 promotes HR repair. A) Transcriptional profile analysis of the mRNA expression of genes associated with DNA repair in control and SOSTDC1‐overexpressing SUM159 cells. B) GSEA showed the enrichment for HR repair‐related genes with increased expression in SOSTDC1‐overexpressing SUM159 cells. C) Immunofluorescence analysis of γ‐H2AX foci at indicated times after irradiation (5 Gy) in SOSTDC1‐overexpressing SUM159 cells. Representative images were shown and percentages of γ‐H2AX foci‐positive cells (>10 foci) were quantified. ^* p < 0.05; ^** p < 0.01 (unpaired Student's two‐sided t‐tests). Scale bar, 10 µm. D) Immunofluorescence analysis of γ‐H2AX foci in SOSTDC1‐overexpressing SUM159 cells treated with DMSO or Olaparib (0.1 µmol L⁻¹, 1 µmol L⁻¹) for 12 h. Representative images were shown and percentages of γ‐H2AX foci‐positive cells (>10 foci) were quantified. ^** p < 0.01; ^*** p < 0.001 (unpaired Student's two‐sided t‐tests). Scale bar, 10 µm. E) Analysis of HR efficiency in SOSTDC1‐overexpressing cells using the DR‐GFP reporter assay. ^** p < 0.01 (unpaired Student's two‐sided t‐tests). F) Analysis of NHEJ efficiency in SOSTDC1‐overexpressing cells using the pBigT‐neo‐GFP reporter assay. Data were presented as mean ± SEM. Significance was calculated using unpaired Student's two‐sided t‐tests. ns, not significant. G) Immunofluorescence analysis of γ‐H2AX foci in wild type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing SUM159 cells treated with DMSO or Olaparib (0.1 µmol L⁻¹, 1 µmol L⁻¹) for 12 h. Percentages of γ‐H2AX foci‐positive cells (>10 foci) were quantified. ^* p < 0.05; ^** p < 0.01; ns, not significant (the one‐way ANOVA). Scale bar, 10 µm. H) Analysis of HR efficiency in wild type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing cells using the DR‐GFP reporter assay. ^* p < 0.05; ns, not significant (the one‐way ANOVA).

2.5. Nuclear SOSTDC1 Interacts with CHD1 to Facilitate HR Repair, BTIC Maintenance, and Olaparib Resistance

To explore how nuclear SOSTDC1 participates in HR, we utilized the co‐IP/MS assay to screen for SOSTDC1‐interacting proteins, and identified a potential SOSTDC1‐interacted nucleoprotein, CHD1 (Figure S8A,B, Supporting Information). CHD1 chromatin helicase DNA‐binding factor (CHD1) regulates open chromatin around the DSB to facilitate the recruitment of HR proteins. CHD1 protects genome integrity at promoters to sustain hypertranscription and pluripotency of embryonic stem cells.^{[ 49 , 50 ]} Loss of CHD1 elicits cellular sensitivity to PARP inhibition.^{[ 51 , 52 , 53 ]} The direct interaction between SOSTDC1 and CHD1 was confirmed by co‐IP assay (Figure  4A) and especially by in vitro co‐IP assay (Figure 4B). CHD1 consists of three main signature domains: Double chromodomain located in the N‐terminal region, ATPase motor centered in the middle of the protein, and DNA‐binding domain located in the C‐terminal region.^{[ 54 ]} Three CHD1 truncated mutants (a, b, c) were constructed to verify the SOSTDC1‐binding domain (Figure S8C, Supporting Information). Only truncations containing the N‐terminus of CHD1 (Fragment a and b) could interact with SOSTDC1 (Figure 4C). IF staining showed a nuclear co‐localization of SOSTDC1 with CHD1 upon Olaparib treatment (Figure 4D). Nuclear rather than cytoplasmic SOSTDC1 interacted with endogenous CHD1 (Figure 4E). Upon Olaparib treatment, CHD1 protein level was upregulated (Figure 4F) and the SOSTDC1‐CHD1 interaction dramatically enhanced (Figure 4G). Taken together, these results manifested the nuclear SOSTDC1‐CHD1 interaction.

Open in a separate window

Figure 4

Nuclear SOSTDC1 interacts with CHD1 to promote HR repair, BTIC maintenance, and Olaparib resistance. A) Co‐IP assay detected the interaction between SOSTDC1 and CHD1. Vectors encoding HA‐tagged CHD1 or FLAG‐tagged SOSTDC1 were transfected into 293T cells, which were subjected to co‐IP experiments with magnetic beads coated with antibodies to FLAG or HA, followed by western blotting with the indicated antibodies. B) In vitro co‐IP assay detected the interaction between SOSTDC1 and CHD1. Purified HA‐tagged SOSTDC1 or FLAG‐tagged CHD1 recombinant proteins were subjected to co‐IP experiments with anti‐FLAG magnetic beads, followed by western blotting with the indicated antibodies. C) Co‐IP assay detected the interaction between SOSTDC1 and full length CHD1 (fl) or its three truncated mutants (a, b, c). D) Immunofluorescence analysis of the distribution of endogenous SOSTDC1 and CHD1 in SUM149 cells upon 1 µmol L⁻¹ Olaparib treatment for 12 h. Scale bar, 5 µm. E) Co‐IP assay detected the interaction between nuclear SOSTDC1 and CHD1. Cytoplasmic and nuclear lysates from FLAG‐tagged SOSTDC1‐overexpressing SUM159 cells were prepared for co‐IP experiments with anti‐FLAG magnetic beads, followed by western blotting with the indicated antibodies. F) Time‐dependent analysis of the expression of CHD1 by western blotting after 1 µmol L⁻¹ Olaparib treatment in SUM159 cells. G) Co‐IP assay detected the interaction between SOSTDC1 and CHD1 upon 1 µmol L⁻¹ Olaparib treatment for 12 h. H) CHD1 protein expression in sorted BTICs (B) and the rest non‐BTICs (R) of SUM149 and SUM159. I) Immunohistochemistry analysis of CHD1 expression in para‐tumor tissues and tumor tissues from BC patients. Representative images of different molecular subtypes were shown and H‐Scores were calculated. Data were presented as mean ± SEM. Significance was calculated using unpaired Student's two‐sided t‐tests. ^* p < 0.05; ^*** p < 0.001; ns, not significant. Scale bar, 100 µm. J) Analysis of the correlation between SOSTDC1 protein levels (related to Figure 1C) and CHD1 protein levels (related to Figure 4I) in BC patient tumor tissues. Statistical significance was determined by Pearson's correlation test. K) Immunohistochemistry analysis of CHD1 expression in SOSTDC1‐knockdown SUM149 xenografts. Representative images were shown and H‐Scores were calculated. Data were presented as mean ± SEM, following unpaired Student's two‐sided t‐tests. ^** p < 0.01. Scale bar, 100 µm. L) SOSTDC1‐overexpressing SUM159 cells with CHD1‐knockdown were treated with DMSO or Olaparib (0.1 µmol L⁻¹, 1 µmol L⁻¹) for 12 h, following immunofluorescence analysis of γ‐H2AX foci. Percentages of γ‐H2AX foci‐positive cells (>10 foci) were quantified. Data were presented as mean ± SEM, following the one‐way ANOVA. ^* p < 0.05; ns, not significant. M) Analysis of HR efficiency in SOSTDC1‐overexpressing cells with CHD1‐knockdown using the DR‐GFP reporter assay. Data were presented as mean ± SEM. Significance was calculated using the one‐way ANOVA. ^* p < 0.05, ^** p < 0.01. N) Mammosphere formation assays of SOSTDC1‐overexpressing SUM159 cells with CHD1‐knockdown. Data were presented as mean ± SEM, following the one‐way ANOVA. ^*** p < 0.001. Scale bar, 100 µm. O) The percentage of BTIC population in SOSTDC1‐overexpressing SUM159 cells with CHD1‐knockdown. Data were presented as mean ± SEM, following the one‐way ANOVA. ^*** p < 0.001. P) Cell viability determined by MTT assay in SOSTDC1‐overexpressing SUM159 cells with CHD1‐knockdown. n  = 3 independent biological samples.

Although CHD1 mRNA level in TNBC was lower than in other subtypes (Figure S8D, Supporting Information), the CHD1 protein level was significantly upregulated in TNBC cell lines (Figure S8E, Supporting Information) especially in BTICs (Figure 4H), and in TNBC patient samples (Figure 4I). There was a strong correlation between SOSTDC1 and CHD1 in BC patient samples (Figure 4J). Besides, CHD1 protein was markedly downregulated in SOSTDC1‐knockdown xenografts (Figure 4K). Survival analysis showed that high CHD1 expression associated with poor OS (p = 0.089) in BC patients (Figure S8F, Supporting Information). These results revealed that CHD1 protein level is positively correlated with SOSTDC1 expression and TNBC malignancy.

To investigate whether SOSTDC1‐CHD1 interaction is important for SOSTDC1 function, we knocked down CHD1 in SOSTDC1‐overexpressing SUM159 cells (Figure S8G, Supporting Information). CHD1 knockdown reversed the effect of SOSTDC1 on cell proliferation (Figure S8H, Supporting Information), DNA repair (Figure 4L; Figure S8I, Supporting Information), and in particular HR repair (Figure 4M; Figure S8J, Supporting Information), mammosphere formation (Figure 4N), BTIC enrichment (Figure 4O; Figure S8K, Supporting Information), and Olaparib resistance (Figure 4P). These results suggested that CHD1 is the main downstream pathway of SOSTDC1 function.

2.6. Nuclear SOSTDC1 Stabilizes CHD1 by Blocking Ubiquitin‐Dependent Degradation

We noticed that SOSTDC1 overexpression promoted while SOSTDC1 knockdown reduced CHD1 protein abundance, with little influence on CHD1 mRNA level (Figure  5A–D), suggesting post‐translation effects of SOSTDC1 on CHD1. Consistent with our speculation, we found the degradation of endogenous CHD1 was accelerated by SOSTDC1 knockdown (Figure 5E) and decelerated by SOSTDC1 overexpression (Figure 5F). Further, we found that the inhibition of CHD1 induced by SOSTDC1 knockdown could be reversed by the proteasome inhibitor MG132 (Figure 5G), indicating the ubiquitin‐proteasome proteolysis regulation of CHD1, which is compatible with previous report.^{[ 55 ]} Moreover, the ubiquitylated CHD1 was significantly increased by SOSTDC1 knockdown (Figure 5H) and decreased by SOSTDC1 overexpression, which was counted on SOSTDC1 nuclear localization (Figure 5I). These results implied that nuclear SOSTDC1 stabilized CHD1 by blocking ubiquitin‐dependent degradation.

Open in a separate window

Figure 5

Nuclear SOSTDC1 inhibits β‐TrCP‐mediated ubiquitination and degradation of CHD1. A,B) CHD1 mRNA expression in SOSTDC1‐knockdown SUM149 cells (A) or SOSTDC1‐overexpressing SUM159 cells (B). C,D) CHD1 protein expression in SOSTDC1‐knockdown SUM149 cells C) or SOSTDC1‐overexpressing SUM159 cells D). E,F) Analysis of CHD1 protein level in SOSTDC1‐knockdown SUM149 cells E) or SOSTDC1‐overexpressing SUM159 cells F) treated with 100 µg mL⁻¹ CHX for the indicated time. G) Analysis of CHD1 protein level in SOSTDC1‐knockdown SUM149 cells treated with 10 µmol L⁻¹ MG132 for 4 h. H) 293T cells with HA‐tagged CHD1 overexpressing were transfected with SOSTDC1‐knockdown and ubiquitin‐encoding vectors. Cells were treated with 10 µmol L⁻¹ MG132 for 8 h and collected for HA‐immunoprecipitation and analysis of CHD1 ubiquitination by western blotting. I) 293T cells with HA‐tagged CHD1 overexpressing were transfected with ubiquitin‐encoding, wild‐type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing vectors. Cells were treated with 10 µmol L⁻¹ MG132 for 8 h and collected for HA‐immunoprecipitation and analysis of CHD1 ubiquitination by western blotting. J) Analysis of CHD1 protein level in Flag‐tagged β‐TrCP‐overexpressing cells with wild type (WT) or NLS‐deleted mutant (ΔNLS) SOSTDC1‐overexpressing. K) 293T cells with HA‐tagged CHD1 overexpressing were transfected with ubiquitin‐encoding, SOSTDC1‐encoding or FLAG‐tagged β‐TrCP‐overexpressing vectors. Cells were treated with 10 µmol L⁻¹ MG132 for 8 h and collected for HA‐immunoprecipitation and analysis of CHD1 ubiquitination by western blotting. L) In vitro ubiquitination assay of β‐TrCP mediated CHD1 ubiquitination in the presence or absence of SOSTDC1. M) Co‐IP assay of the interaction between β‐TrCP and CHD1 in the presence or absence of SOSTDC1. N) In vitro co‐IP assay of the interaction between CHD1 and β‐TrCP in the presence or absence of SOSTDC1. O) Co‐IP assay of the interaction between SOSTDC1 and CHD1 mutants. Vectors encoding HA‐tagged SOSTDC1, FLAG‐tagged wild type, or the two β‐TrCP‐binding‐motif mutants of CHD1 were transfected into 293T cells, which were subjected to co‐IP experiments with anti‐HA magnetic beads, followed by western blotting with the indicated antibody.

CHD1 has been reported to induce trimethyl lysine‐4 histone H3 (H3K4me3) DNA modifications, which activate transcription of the pro‐tumorigenic TNF–NF‐κB gene network.^{[ 55 ]} We found that SOSTDC1 positively regulated CHD1 downstream gene expression (Figure S9A,B, Supporting Information). Taken together, these results demonstrated that SOSTDC1 maintained CHD1 stabilization and promoted CHD1‐regulating genes expression.

Tumorigenicity in Nude Mice

Breast cancer cells were injected with Matrigel into the fourth mammary fat pads of four‐week‐old female nude mice purchased from Vitalriver. The mice were housed in standard animal cages under a Specific‐pathogen‐free facility at 23–25 °C on a 12 h light/ dark cycle in the Department of Laboratory Animal Science of Fudan University. The in vivo treatment of Olaparib started when the average diameter of xenograft tumors reached to 3 mm. Mice in each group were further divided into another two groups and treated with DMSO or Olaparib (50 mg kg⁻¹, once a day) for 29 days intraperitoneally. The tumor sizes were measured with a caliper and calculated as length × width × width/2 and monitored weekly. Mice were sacrificed when the diameter of tumors reached 10–15 mm.

Plasmid Construction and Lentivirus Transfection

SOSTDC1 full‐length cDNA and its NLS‐deleted mutant with FLAG tag were cloned into the lentiviral vector pSIN (Addgene). FLAG‐tagged β‐TrCP, SOSTDC1, and Ubi were cloned into the lentiviral vector pSIN. CHD1 full‐length cDNA and its truncated mutants with HA tag were cloned into the lentiviral vector pLVX (Addgene). shRNA sequence of SOSTDC1, KPNA1, KPNA2, KPNA3, and CHD1 were cloned into lentiviral vector pLKO.1 (Addgene). HA‐tagged SOSTDC1, FLAG‐tagged CHD1, and β‐TrCP were cloned into bacterial expression vector pGEX‐6p‐1 (Addgene). DR‐GFP (HR report vector) and pCBAScel (I‐Scel expression vector) were kind gifts from Dr. Weixing Zhao. FLAG‐KPNA1, FLAG‐KPNA2, FLAG‐KPNA3, FLAG‐KPNA4, FLAG‐KPNA5, and FLAG‐KPNA6 were kind gifts from Dr. Daming Gao. pBigT‐neo (NHEJ report vector) was a gift from Dr. Anyong Xie. A highly efficient lentiviral system was used to generate plasmid DNA. The breast cancer cell lines were then infected with the lentiviruses to establish stable cell lines. The bacterial expression plasmid was transduced into BL21 E. coli bacteria to produce protein. The primers used for cloning are listed in Table S2 (Supporting Information).

Total RNA Isolation and qRT‐PCR

Total RNA was extracted from cells using TRIzol reagent (Takara Bio) and then reverse‐transcribed to complementary DNA (cDNA) using the HiScript II 1st Strand cDNA Synthesis kit (Vazyme Biotech) according to the manufacture's recommendation. qRT‐PCR was performed to detect the expression levels of target genes using AceQ qPCR SYBR Green Master Mix (Vazyme Biotech) in 7300 Plus Real‐Time PCR System (Applied Biosystems). The relative mRNA level of the target gene was analyzed by the Equation (2)^{–Δ Ct} (ΔCt = Ct of target gene minus Ct of TBP). The primers used are shown in Table S3 (Supporting Information).

Western Blotting and Separation of Nuclear and Cytoplasmic Proteins

Protein samples were extracted from cells or tissues using RIPA buffer (Beyotime Biotechnology). After determining the concentration by a BCA kit (Thermo Fisher), the protein lysates were denatured in 5× loading buffer in 100 °C for 10 min, separated by SDS‐PAGE, and transferred onto PVDF membranes (Millipore). The membranes were blocked with 5% de‐fat milk, incubated with primary antibody overnight at 4 °C, and probed with HRP‐conjugated secondary antibody. Chemiluminescence detection was carried out with an ImageQuant LAS 4000 mini‐imaging system (GE) with western HRP Substrate (Millipore). Nuclear and Cytoplasmic Protein Extraction Kit (P0028, Beyotime) was used to separate cytoplasmic and nuclear protein according to the manufacture's instructions. Phenylmethylsulfonyl fluoride (PMSF) (ST506, Beyotime) was added to inhibit protein degradation. ImageJ was used to quantify the expression levels of related proteins, which were presented as the ratio of test protein‐integrated density to control protein‐integrated density.

The following antibodies and dilutions were used:

SOSTDC1 (1:200, AF7034, R&D systems), CHD1 (1:1000, A7883, Abclonal), β‐TrCP (1:1000, A1656, Abclonal), γH2AX (1:1000, 60566, CST), Ub (1:500, Santa Cruz Biotechnology), Lamin A/C (1:1000, 4777, CST), Tubulin (1:2000, HC101, TransGen), FLAG (1:2000, F7425, Sigma‐Aldrich), HA (1:1000, C29F4, CST), goat anti‐mouse immunoglobulin G (IgG)–horseradish peroxidase (HRP) (1:5000, HS201‐01, TransGen), goat anti‐rabbit IgG‐HRP (1:5000, HS101‐01, TransGen), donkey anti‐Goat IgG‐HRP (1:5000, SA00001‐3, Proteintech).

Flow Cytometry Analysis and Sorting

The ALDEFLUOR assay (Stem Cell Technologies) was performed following the manufacturer's protocols. A CD24/CD44 assay was performed with anti‐CD24‐APC (Biolegend,1:40) and anti‐CD44‐APC‐H7 (BD Bioscience,1:100). PE‐conjugated anti‐mouse H2kd antibody (Biolegend, 1:100) was used to discriminate human breast tumor cells from mouse cells in single‐cell suspensions digested by collagenase from xenografted tumors. Four populations from the total tumor cells (ALDH⁺CD24⁻CD44⁺, ALDH⁺non‐CD24⁻CD44⁺, ALDH⁻CD24⁻CD44⁺ and ALDH⁻non‐CD24⁻CD44⁺) were sorted according to the strategy reported previously.^{[ 30 ]} The cells were collected and divided into two populations (BTICs (ALDH⁺CD24⁻CD44⁺)) and the rest non‐BTICs (ALDH⁺non‐CD24⁻CD44⁺, ALDH⁻CD24⁻CD44⁺, ALDH⁻non‐CD24⁻CD44⁺)) for further analysis. Flow cytometry analysis and cell sorting were conducted by MOFLO ASTRIOS (Beckman Coulter) instrument and analyzed by Summit 6.3 software.

Immunohistochemistry

The breast cancer patient tumor tissues and para‐tumor tissues were obtained from Shanghai Cancer Hospital affiliated with Fudan University. Informed consent was obtained from the involved patients, and the study was approved by the institution's ethics committee. The tumor tissues of the mouse were fixed in formalin and processed for paraffin embedding. Section samples were dewaxed in xylene and rehydrated in graded alcohol. The immunohistochemistry and signal evaluation (H‐score) were performed according to the procedures reported previously.^{[ 64 ]}

The following antibodies and dilutions were used: SOSTDC1 (1:100, AF7034, R&D systems), Ki67 (1:100, ZM‐0166, ZSGB‐BIO) and peroxidase‐conjugated secondary antibody (KIT‐5010, KIT‐5107, Maxvision).

MTT Assay

To evaluate the cell proliferation, exponentially growing cells were seeded in 96‐well culture plates (200 to 800 cells per well) and cultured for 3, 5, or 7 days. To determine the cell viability, exponentially growing cells were seeded in 96‐well culture plates (2000 to 10000 cells per well) and incubated overnight to facilitate cell attachment. The following day, the cell cultures were treated with Olaparib (HY‐10162, MCE) in concentrations ranging from 0.0001 to 200 µm. After 72 h, MTT (Biosharp Life Science) was added to each well (final concentration: 0.5 mg mL⁻¹) and incubated at 37 °C for 4 h. Then the supernatant was removed, and 100 µL DMSO was added. The optical density value was measured at 490 nm after shaking plates for 10 min. Cell viability was set as 100% for cells not treated with Olaparib.

Colony Formation Assay

Breast cancer cells were seeded in 6‐well culture plates (200 to 800 cells per well) and cultured at 37 °C for 14 days to allow colonies to form. Cells were fixed with 4% formaldehyde for 30 min and stained with crystal violet. After washing, the cell colonies were quantified.

Mammosphere Formation Assay

Breast cancer cells were seeded in ultra‐low attachment 96‐well plates (Corning) (100 to 200 cells per well) and cultured with a MammoCult Human Medium kit (StemCell Technologies) supplemented with 4 µg mL⁻¹ heparin (StemCell Technologies), 1% pen‐strep, 1 µg mL⁻¹ hydrocortisone for 10–14 days. Fresh complete medium was added every four days. The images of mammospheres were collected by microscope for further statistical analysis.

Immunofluorescence Staining and Confocal Imaging

Breast cancer cells were seeded in 4‐well chamber (Thermo Scientific) (30 000 cells per well) and cultured for 2 days. Cells were fixed with 4% paraformaldehyde (Sangon Biotech) for 15 min at room temperature (RT), membrane perforated with 0.15% Triton X‐100 (Sangon Biotech) for 10 min at RT, blocked with 5% bovine serum albumin (Sangon Biotech) solution for 1 h at RT, incubated with primary antibody overnight at 4 °C and secondary antibody for 40 min at RT. The cell nucleus was stained with DAPI (Invitrogen). Images were captured by confocal microscope (LEICA SP5) with a 63× objective lens.

The following antibodies and dilutions were used: SOSTDC1 (1:20, AF7034, R&D systems), CHD1 (1:100, A7883, Abclonal), FLAG (1:100, F7425, Sigma–Aldrich), γH2AX (1:100, 60566, CST), goat anti‐rabbit IgG secondary antibody Alexa Fluor 546 (1:100, A‐11035, Invitrogen), donkey anti‐goat IgG secondary antibody Alexa Fluor 488 (1:100, A‐11055, Invitrogen).

HR and NHEJ Reporter Assay

293T cells integrated with DR‐GFP or pBigT‐neo‐GFP cassettes were used to determine HR or NHEJ efficiency, respectively. Cells transiently transfected with report vectors or control‐mcherry vector, control or SOSTDC1‐overexpressing plasmids were then transfected with I‐Scel expression vector pCBAScel. 36 h after transfection, the percentage of GFP⁺ or mCherry⁺ cells was analyzed by flow cytometry. HR or NHEJ efficiency was presented as the percentage of GFP⁺/mCherry⁺. HR or NHEJ efficiency presented in the figures are the mean ± SEM of three independent experiments.

Chemical Inhibitors

PARP inhibitor Olaparib (HY‐10162, MCE), treated breast cancer cells at concentrations of 0.1 nmol L⁻¹‐200 µmol L⁻¹ in vitro and treated xenograft tumors (50 mg kg⁻¹, once a day) for 29 days intraperitoneally in vivo. Eukaryote protein synthesis inhibitor CHX (HY‐12320, MCE), treated breast cancer cells at a concentration of 100 µg mL⁻¹. Proteasome inhibitor MG132 (HY‐13259, MCE), treated breast cancer cells at a concentration of 10 µmol L⁻¹.

Immunoprecipitation Assay and Ubiquitylation

Breast cancer cells were collected and lysed by the mild RIPA lysis containing protease inhibitor cocktail (Roche) for 1 h at 4 °C. Except for a small fraction of input group, the remaining supernatant was diluted with the week RIPA lysis and incubated with magnetic FLAG‐Beads (Sigma–Aldrich) or anti‐HA antibody‐conjugated protein A/G agarose beads (Smart‐Lifesciences) overnight at 4 °C. After removing the supernatant, beads were washed 4 times with NETN buffer (20 mmol L⁻¹ Tris, 100 mmol L⁻¹ NaCl, 0.5% NP40, 1 mmol L⁻¹ EDTA, pH8.0). Proteins were then eluted by competitive elution of 3× FLAG fusion proteins (0.4 mg ml⁻¹; F4799, Sigma‐Aldrich) and boiled or directly denatured in 2× loading buffer. Boiled samples can be followed by silver staining and western blotting. To examine the ubiquitylation level of CHD1, cells were treated with MG132 (10 µmol L⁻¹, 8 h, MCE) before harvested. Protein lysates were performed with immunoprecipitation and western blotting.

Protein Purification and GST Pull‐Down Assay

BL21 E. coli bacteria were transfected with GST‐tagged vectors and grown in Luria–Bertani liquid medium to an OD600 of ≈0.8, and then adding with isopropyl β‐d‐1‐thiogalactopyranoside (IPTG, 1 mmol L⁻¹) to induce the protein expression overnight at 16 °C. Proteins were purified using the GST agarose beads (TransGen Biotech) from the cell lysates and then eluted with 10 mmol L⁻¹ GSH (BBI). Samples were subject to co‐IP assay or western blotting.

In Vitro Ubiquitination Assay

GST‐fusion proteins (β‐TrCP, FLAG‐tagged CHD1, HA‐tagged SOSTDC1) were purified from BL21 strain and added into the ubiquitination buffer system (10x reaction buffer, 10x E1 enzyme, 5x Ubiquitin, 20x UBE2D3/UbcH5c, 10x Mg2⁺‐ATP) provided by E2Select Ubiquitin Conjugation Kit (ubbiotech). The reaction was conducted at 37 °C for 3 h and then mixed with loading buffer subjected to following western blotting.

Mass Spectrometry

The co‐IP protein samples of SUM159‐CTRL or SOSTDC1 and MDA‐MB‐231‐CTRL or SOSTDC1 cells were concentrated in stacking gel. Gel‐containing samples were decolorized and washed to make it transparent and then freeze‐dried. Samples were reduced of disulfide bonds before enzymatic hydrolysis. Then, the peptide segment was extracted and dried in a vacuum. Samples were desalted, and the supernatant was added to the sample bottle for mass spectrometry (Q Exactive) detection. The search database was Maxquant.

RNA Sequencing

After RNA extraction, RNA concentration and quality were measured by Agilent 2100. RNA‐seq libraries were established and sequenced on HiSeq3000 platform (Illumina). For analyzing the RNA‐sequencing results, the differential gene expression was determined by DEseq using 1.5‐fold change, with p‐value < 0.05 as threshold.

Quantification and Statistical Analysis

All experiments for quantitative analysis and representative images were reproduced with similar results for at least three times. Bar graphs were generated with GraphPad Prism 6.0, and all values are reported as the means ± SEM. Comparisons between two groups were performed using an unpaired Student's t‐test or Mann–Whitney U‐test. Differences among three or more groups were analyzed by one‐way/two‐way ANOVA. Two‐sided log‐rank (Mentel–Cox) test was used to evaluate the survival analysis. Pearson Chi‐square test was used to evaluate IHC score levels between different clinicopathological variable groups. Bivariate correlation analysis was performed using the Pearson correlation method. p < 0.05 was considered statistically significant.

Public Data Analysis

GSE165914 for Olaparib‐resistant gene cluster^{[ 44 ]} was obtained from Gene Expression Omnibus of the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/geo). PRJNA739366 and PRJNA376644 data were obtained from the Sequence ReadArchive of NCBI. Results were obtained with the online tool bc‐GenExMiner v4.5, for which the available URL is http://bcgenex.ico.unicancer.fr/BC‐GEM/GEM‐requete.php. Kaplan–Meier plotter was used to analyze the OS and RFS of different molecular subtypes of breast cancer patients based on SOSTDC1 expression. A total of 1220 BC patients from The Cancer Genome Atlas (https://tcga‐data.nci.nih.gov/tcga) were used for gene expression analysis for SOSTDC1 and CHD1. The NLS sequence was analyzed by NLS Mapper (http://nls‐mapper.iab.keio.ac.jp/cgi‐bin/NLS_Mapper_form.cgi).

The authors thank Dr. Weixing Zhao, Dr. Daming Gao, and Dr. Anyong Xie for providing the plasmids. The authors acknowledge the staff members of the Institutional Center for Shared Technologies and Facilities of SIMM for technical assistance. The authors thank the members of the Liu laboratory for their technical assistance. This work was supported by National Key Research and Development Program of China (2023YFC2506400, 2020YFA0112300); National Natural Science Foundation of China (82230103, 81930075, 82073267, 82203399, 82372689); Program for Outstanding Leading Talents in Shanghai; Program for Outstanding Medical Academic Leader in Shanghai (2019LJ04); Program of Shanghai Academic/Technology Research Leader (20XD1400700); The innovative research team of high‐level local university in Shanghai.