2.1 Clinical data and samples

The recently published expression profiles of quantitative proteomic dataset (normal samples: n = 72; TNBC samples: n = 90)²² and RNA-sequencing (RNA-Seq) dataset (normal samples: n = 88; TNBC samples: n = 360)²³ were obtained from Fudan University Shanghai Cancer Center (FUSCC) TNBC cohort. The tissues used in this study were collected from TNBC patients who underwent surgery in FUSCC. The FUSCC Ethics Committee approved all experimental procedures.

2.2 Cell culture

We obtained HMEC and MCF10A cell lines, which are normal human mammary epithelial cells, as well as LM2-4175, HCC1806, SUM159, BT549, HCC1937, MDA-MB-231 and Hs578T cell lines, which are human TNBC cells used in this study from Shanghai Key Laboratory of Breast Cancer. The above cells were cultured in medium as described previously.³¹

2.3 DNA plasmids and vector transfection

The cDNA was subcloned into the pCDH-CMV-3× Flag-Puro vectors. The specific short hairpin RNA (shRNA) sequences targeting C9orf142 and MTBP were acquired from Sigma-Aldrich Advanced Genomics (www.sigmaaldrich.cn) (Table S1). The primer sequences were synthesized by HuaGene Biotech. The primers were subsequently annealed and then inserted into the pLKO.1 vectors. The complete human C9orf142 cDNA was amplified using the primers specfied in Table S2. Vector transfection, lentivirus packaging, infection and generation of stable cell lines were carried out according to the methods outlined in previous studies.^{32, 33}

2.4 LC-MS/MS analysis

Liguid chromatograph tandem mass spectrometry (LC-MS/MS) analysis was conducted in the Institutes of Biomedical Sciences, Fudan University as described previously.³² For analyzing label-free quantitative proteomic results, the fold change of 1.5-fold was established as the threshold for differentially expressed proteins, requiring at least two unique peptides.

2.5 Immunoblotting assays

Proteins were lysed and quantified as previously described.³² A total of 20−50-μg proteins, isolated from cells or tissues, were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes. The detailed information of the primary antibodies is provided in Table S3. The protein signals were detected using an enhanced chemiluminescence (ECL) kit (Tanon, No. 180−5001E) after incubation with corresponding secondary antibodies.

2.6 RT-qPCR assays

Total RNA extraction, cDNA synthesis, removal of genomic DNA, and real time-quantitative PCR (RT-qPCR) assays were carried out according to the methods outlined in previous studies.³² Detailed information of the primers utilized for RT-qPCR in this investigation is provided in Table S4.

2.7 Cell proliferation, colony formation and flow cytometry assays

Briefly, for cell proliferation assays, cells were seeded into 96-well plates in triplicate. Subsequently, they were exposed to either DMSO or abemaciclib (Sellect, No. S5716) at the specified concentrations. Cell viability was assessed using cell counting kit-8 (CCK-8) kit. For colony formation assays, in short, cells were seeded into 6-well plates in triplicate. Subsequently, they were treated with either DMSO or abemaciclib at the indicated concentrations. After fixed with paraformaldehyde, the survival colonies clones were stained with crystal violet. Flow cytometry assay was used to determine cell cycle, following the previously described protocol.³²

2.8 Cell migration and invasion assays

Cells were re-suspended in 200 μL of complete medium and then seeded into Transwell inserts without Matrigel or coated with Matrigel for cell migration and invasion assays, respectively. Twenty four-well plates were filled with 800 μL of complete medium. After 20–24 h of incubation, the migrated and invaded cells attached to the subventricular layer of Transwell inserts were fixed using methanol and stained with crystal violet.

2.9 Xenograft tumour experiments

All animal experiments were approved by the Ethics Committee of Fudan University Shanghai Cancer Center (FUSCC) on July 9^th, 2021, and the registration number of the ethics is No. FUSCC-IACUC-2021383. The animals using in this study were all 6-week-old female BALB/c nude mice. For xenograft tumour experiments, cells were inoculated into the mammary fat pads of mice. Tumour size was measured every 2 days, and tumour volume was calculated using the method described previously.³² To determine the effects of CDK4/6 inhibitor abemaciclib on xenograft tumour growth, phosphate buffered solution (PBS) or abemaciclib (25 mg/kg) was administered by oral gavage daily when the average tumour volumes reached 100 mm³. For spontaneous metastasis experiments, cells were inoculated into the mammary adipose tissues of mice. For experimental lung metastasis experiments, cells were injected into the caudal veins of mice. Tumour weight and physical condition of mice were monitored every week. Nude mice were sacrificed once they lost weight, developed dysphagia or cachexia. For experimental axillary lymph node metastasis, cells were inoculated into the mammary adipose tissues of mice. Nude mice were sacrificed once tumour volumes reached 2000 mm³ or the maximum diameter of the tumour reached 15 mm.

2.10 Chromatin immunoprecipitation assays

Chromatin immunoprecipitation (ChIP) assays were performed using Simple ChIP kit (Cell Signaling Technology, No. 9003). Histone H3 antibody was served as a positive control, whereas normal IgG antibody was employed as a negative control. The primers used for ChIP-PCR in this investigation are provided in Table S4.

2.11 Dual-luciferase reporter assays

MTBP promoters (four sections ranging from −2000 bp before the transcription start site to 100 bp after it) were cloned into the pGL3 vector to generate pGL3-MTBP. The primers used to clone pGL3-MTBP plasmids are provided in Table S5. Subsequently, HEK293T cells were transfected with 50 ng of pRL-TK Renilla luciferase expression vector and 1 μg of either pGL3 basic vector or pGL3-MTBP plasmids, following the previously described protocol.³⁴ Luciferase assays were conducted with the Dual Luciferase Reporter kit (Yeasen, number: 11402ES60) after transfection for 48 h.

2.12 Immunofluorescent staining assays

Cells were re-suspended in 300 μL of complete medium and then seeded into 24-well plates with cover glass. The next day, cells were fixed with 4% paraformaldehyde. Permeabilization, antibody incubation and imaging were performed as described previously.^{32, 33}

2.13 Statistical analysis

Statistical analyses of two group were conducted with unpaired Student's t tests or one-way analysis of variance (ANOVA). The log-rank test was utilized to calculate the survival curve of Kaplan-Meier plot. The statistical analyses were performed using R software, GraphPad and SPSS. The p value less than 0.05 was deemed statistically significant.

3.1 C9orf142 is aberrantly up-regulated in TNBC tissues and its elevated expression levels are indicatives of poor prognosis for TNBC patients

In order to identify possible therapeutic targets that promote the progression of TNBC, we initially conducted an integration analysis of both TNBC proteomic²² and transcriptomic dataset²³ from our center project cohort (Figure 1A). A total of 2568 proteins in the proteomic dataset were abnormally increased in TNBC samples when compared to adjacent normal samples. According to the status of lymph node metastasis, we found that 81 proteins were upregulated in TNBC patients with lymph node metastasis (n = 36) relative to those without lymph node metastasis (n = 54). Further examination revealed that 34 proteins were elevated in both TNBC patients and those with lymph node metastasis (Figure 1A, upper panel). Similarly, a total of 7512 genes in RNA-Seq dataset were overexpressed in TNBC tissues as compared with normal tissues. Moreover, 974 genes were elevated in patients with lymph node metastasis (n = 143) relative to those without lymph node metastasis (n = 217). Further analysis demonstrated that a total of 476 oncogenes were positively correlated with lymph node metastasis of TNBC patients (Figure 1A, lower panel). According to the p value, the top 30 metastasis-related proteins and genes are presented in Figures S1A–D.

Interestingly, we noticed that C9orf142 was upregulated in total and paired TNBC specimens in both the quantitative proteomic dataset (Figure 1B,C) and the RNA-Seq dataset (Figure 1D,E). Moreover, C9orf142 expression was higher in TNBC patients with lymph node metastasis than those without lymph node metastasis (Figure 1F,G) and was positively correlated with tumour size (Figure 1H) and clinical stage (Figure 1I) of TNBC patients. Consistently, analysis of the TCGA breast cancer database also demonstrated that C9orf142 was upregulated in TNBC samples, and its expression levels in TNBC were slightly higher than that in luminal and HER2-positive subtypes of breast cancer (Figure S1E). To validate the above results, we detected the protein expression levels of C9orf142 in paired TNBC tissues. Results showed that in comparison with the normal counterparts, C9orf142 exhibited a notable increase in 83.3% (15/18) TNBC tissues (Figure 1J and Figure S1F). Analysis of survival using Kaplan–Meier Plotter database revealed that the difference of C9orf142 expression was not related to the overall survival and relapse-free survival prognosis of all types of breast cancer patients (Figures S1G,H). However, increased C9orf142 expression correlated with worse outcomes in TNBC patients, including overall survival (OS) and relapse-free survival (RFS), suggesting that C9orf142 may serve as a prognostic factor for TNBC patients (Figure 1K,L). Taken together, these findings suggest that C9orf142 could function as a novel oncogene to drive the advancement of TNBC, and its elevated levels are indicatives of unfavorable prognosis in individuals with TNBC.

3.2 C9orf142 promotes TNBC growth both in vitro and in vivo

In order to verify the possible functions of C9orf142 in the progression of TNBC, we initially examined the expression levels of C9orf142 in two normal mammary epithelial cell lines as well as seven representative TNBC cell lines. The findings indicated that C9orf142 exhibited significant up-regulation in TNBC cell lines when compared to normal cell lines (Figure 2A). Based on the relative expression levels of C9orf142 in these cell lines and their tumourigenic and metastatic properties, we stably overexpressed Flag-C9orf142 in cells with low expression of C9orf142 (Figure 2B), whereas knocked down endogenous C9orf142 in cells with high expression of C9orf142 by lentiviral infection, respectively (Figure 2C). Results of in vitro functional experiments demonstrated that overexpression of C9orf142 in SUM159 and MDA-MB-231 cells enhanced cell growth (Figure 2D) and generated a greater number of viable colonies (Figure 2E,F). Conversely, the reduction of cell viability (Figure 2G) and colony formation capacity (Figure 2H,I) was observed when C9orf142 was knocked down using shRNA. Additionally, flow cytometry assays revealed that depletion of C9orf142 led to a rise in cells within the G1 phase, coupled with a decline in cells within the S phase (Figure 2J).

To further determine whether C9orf142 enhances the tumourigenic ability of TNBC cells in animal models, LM2-4175 cells expressing shNC empty vector and shC9orf142 (#1 and #2) were injected into the mammary fat pads of female BALB/c nude mice (n = 10) (Figure 2K). Consistent with in vitro results, knockdown of C9orf142 significantly reduced tumour volume (Figure 2L, M) and weight (Figure 2N). The expression levels of C9orf142 in xenograft tumours were basically consistent with its in vitro validation results in cell lines (Figure S2A). Collectively, these results imply that C9orf142 enhances TNBC growth both in vitro and in vivo.

3.3 C9orf142 promotes the metastatic potential of TNBC cells

Due to the highly invasive and metastatic potential of TNBC, we then investigated the impact of C9orf142 on the metastatic capacities of these cells. The results showed that overexpression of C9orf142 increased the migratory and invasive potential of TNBC cells (Figure 3A–D). Conversely, the opposite effects were observed in cells with C9orf142 knockdown (Figure 3E–H).

To validate these results in a living organism, LM2-4175 that were genetically modified to express shNC and shC9orf142 (#1 and #2) were injected into the mammary fat pads of female BALB/c nude mice (n = 10) for 8 weeks to generate the spontaneous metastatic models (Figure 3I). As expected, knockdown of C9orf142 dramatically reduced the quantity of metastatic lung nodes in the mice, the incidence of lung metastasis and the weight of the lungs (Figure 3J–M, and Figure S2B). Additionally, experimental murine model of axillary lymph node metastasis reconfirmed that C9orf142 knockdown dramatically reduced the occurrence of axillary lymph node metastasis, which was confirmed by the HE staining (Figure S2C,E). Collectively, these results suggest that C9orf142 boosts the invasive and metastatic potential of TNBC cells.

3.4 C9orf142 transcriptionally activates MTBP expression

To investigate the underlying molecular mechanism of C9orf142 that promotes TNBC progression, we conducted label-free quantitative proteomic assays. According to the cut-off value of 1.5-fold change, a total of 683 proteins were found to be upregulated and 344 proteins were down-regulated following C9orf142 knockdown (Figure 4A), and the top 30 up-regulated and down-regulated proteins following C9orf142 knockdown are presented in Figure 4B. In order to acquire a comprehensive understanding of the biological roles of these proteins that were expressed differentialy, we conducted gene ontology-biological process (GO-BP) and GO-molecular function (GO-MF) analysis. It was found that these differentially expressed proteins were primarily participated in the following biological processes, such as cell division, cell cycle, cell migration, DNA repair and replication (Figures 4C). Molecular functions of these proteins were primarily related to protein/RNA/chromatin/enzyme binding, and transcriptional coactivator activity (Figure 4D). Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed that these proteins were mainly concentrated in cell cycle, DNA replication and ubiquitin proteolysis pathways (Figure 4E).

Given that the functional role of the top one down-regulated protein LINGO2 (leucine rich repeat and Ig domain containing 2) following C9orf142 knockdown in human cancer is ambiguous, we selected MTBP, the second down-regulated protein following C9orf142 knockdown, for further verification. Immunoblotting and RT-qPCR analyses confirmed that knockdown of C9orf142 in Hs578T and LM2-4175 cells resulted in a decrease in the protein and mRNA expression levels of MTBP (Figure 4F,G), whereas the opposite results were obtained in cells with overexpression of C9orf142 (Figure 4H,I). Furthermore, immunofluorescent staining assays also demonstrated a significant decrease in the expression levels of MTBP following C9orf142 knockdown (Figure S3A). These results suggest that C9orf142 may transcriptionally activate MTBP expression.

Previous studies have shown that MTBP enhances the stability of MDM2 and facilitates MDM2-induced degradation of p53.²⁵ In addition, p53 directly up-regulates the expression levels of cyclin-dependent kinase inhibitor p21, thereby arresting cells in the G1 phase to slow malignant progression.^{35, 36} Indeed, immunoblotting assays demonstrated that C9orf142 had a positive regulatory effect on the expression levels of MTBP and MDM2, while it had a negative regulatory effect on the expression levels of p53 and p21 (Figure 4F,H). These results collectively indicate that C9orf142 activates MTBP/MDM2/p53/p21 signaling axis in TNBC cells.

3.5 C9orf142 is enlisted to the MTBP promoter and boosts its promoter activities

To examine whether C9orf142 is recruited to the MTBP promoter, we performed immunoprecipitation using formaldehyde crosslinked chromatin (ChIP) from TNBC cells stably expressing empty vectors pCDH or Flag-C9orf142. Subsequent RT-qPCR tests were conducted with primers specifically designed for approximately every 500 base pair (bp) segment of the MTBP promoter (−2000 to +100) (Figure 5A). The findings indicated that C9orf142 was enlisted in two sections (−504 to −398 and −241 to −73) of the MTBP promoter (Figure 5B,C, Figure S3B,C). In order to investigate the impact of C9orf142 on MTBP promoter activities, about every 500 bp region of MTBP promoter (−2000 to +100) was amplified. These segments were then subcloned into the pGL3 vector, respectively, and transfected into HEK293T cells alone or in combination with Flag-C9orf142 (Figure 5D). As shown in Figure 5E,F, C9orf142 enhanced MTBP promoter activities at those two regions (#3 and #4). Conversely, knockdown of C9orf142 reduced MTBP promoter activities (Figure 5G). Collectively, these results demonstrated that C9orf142 is enlisted to the MTBP promoter and boosts its promoter activities.

3.6 C9orf142 augments TNBC progression partially through regulating MTBP

To verify whether C9orf142 promotes TNBC progression through regulating MTBP, we knocked down MTBP in TNBC cells with overexpression of Flag-C9orf142. Immunoblotting assays reconfirmed that overexpression of C9orf142 resulted in an increase in the expression levels of MTBP and MDM2, whereas knockdown of MTBP partially abrogated these effects caused by overexpression of C9orf142 (Figure 6A and Figure S4A). Functional assays showed that knockdown of MTBP impaired the growth of TNBC cells (Figure 6B) and formation of colonies (Figure 6C and Figure S4B) triggered by ectopic expression of C9orf142. Similarly, the results revealed that the enhanced migration and invasion capacity of TNBC cells, which resulted from the overexpression of C9orf142, was partially attenuated following knockdown of endogenous MTBP (Figure 6D,E, Figure S4C,D). In the xenograft tumour mouse models, the C9orf142-mediated enhancement of tumour growth was partially impaired after MTBP knockdown (Figure 6F–H). Furthermore, in the experimental lung metastasis models, the C9orf142-mediated increase in the quantity of metastatic lung nodes and the occurrence of lung metastasis was compromised following knockdown of MTBP (Figure 6I–K, Figure S4E,F). In summary, these results indicate that C9orf142 contributes to TNBC progression by, at least partially, facilitating the transcriptional transactivation of MTBP.

3.7 C9orf142 promotes resistance of TNBC cells to CDK4/6 inhibitor abemaciclib

The mechanism of CDK4/6 inhibitors against cancer is to prevent the transition of cell cycle from G1 phase to S phase.³⁷ Our above results demonstrated that C9orf142 had a positive effect on the expression of MTBP and MDM2, while it had a negative effect on the expression of p53 and p21 (Figures 4F and 4H), thus promoting transition of cell cycle through G1-to-S phase. Flow cytometry assays also demonstrated that depletion of C9orf142 led to an increase in cells in the G1 phase, along with a decrease of cells in the S phase (Figure 2J). Therefore, we proceeded to examine the influence of C9orf142 on the sensitivity of TNBC cells to CDK4/6 inhibitor abemaciclib. We found that C9orf142-depleted cells were more sensitive to abemaciclib compared to control cells (Figure 7A). Similarly, the colony formation assays yielded comparable outcomes (Figure 7B,C). Moreover, in xenograft tumour models, the tumour growth inhibitory effects (including tumour volume and weight in mice) caused by C9orf142 knockdown were significantly greater than that observed in control mice following abemaciclib administration (Figures 7D–7F). Collectively, these results indicate that C9orf142 promotes resistance of TNBC cells to the CDK4/6 inhibitor abemaciclib.