Cancer stem cells (CSC) are supported by the tumor microenvironment, and non-CSCs can regain CSC phenotypes in certain niches, leading to limited clinical benefits of CSC-targeted therapy. A better understanding of the mechanisms governing the orchestration of the CSC niche could help improve the therapeutic targeting of CSCs. Here, we report that Rab13, a small GTPase, is highly expressed in breast CSCs (BCSC). Rab13 depletion suppressed breast cancer cell stemness, tumorigenesis, and chemoresistance by reducing tumor-stroma cross-talk. Accordingly, Rab13 controlled the membrane translocation of C-X-C chemokine receptor type 1/2 (CXCR1/2), allowing tumor cells to interact with tumor-associated macrophages and cancer-associated fibroblasts to establish a supportive BCSC niche. Targeting the Rab13-mediated BCSC niche with bardoxolone-methyl (C-28 methyl ester of 2-cyano-3, 12-dioxoolen-1, 9-dien-28-oic acid; CDDO-Me) prevented BCSC stemness in vitro and in vivo. These findings highlight the novel regulatory mechanism of Rab13 in BCSC, with important implications for the development of therapeutic strategies for disrupting the BCSC niche.

Significance:

Targeting Rab13 perturbs formation of the breast cancer stem cell niche by inhibiting cross-talk between cancer cells and the tumor microenvironment, providing a therapeutic opportunity for niche-targeted breast cancer treatment.

Targeting breast cancer stem cells (BCSC) is critical for breast cancer treatment. However, evidence indicates that cancer stem cells (CSC) and non-CSCs are dynamic and plastic (1). Non-CSCs can regain stemness phenotypes and thereby reenter the CSC pool under the stimulation of the microenvironment (1), suggesting that therapeutically eradicated CSC populations could be replenished by non-CSCs in the tumor after treatment (2). Therefore, cell plasticity is a considerable challenge in the development of CSC-targeted therapies. Tumor cells, especially CSCs, proactively remodel their microenvironment to maintain CSC stemness, which contributes to tumor survival and treatment resistance (1, 3, 4). Therefore, compared with targeting BCSCs themselves, targeting the BCSC niche (2), or severing the connection between BCSCs and their supportive niche may be a potent alternative strategy. Identifying the underlying mechanisms of niche-mediated BCSC stemness and niche orchestration is essential for the development of therapeutic strategies for targeting the BCSC niche.

Stromal cells secrete a variety of protumorigenic factors in BCSC niche (4), and the ability of tumor cells engaging with the surrounding niche successfully is dictated by the protein composition of the cell surface, which governs the binding of extracellular factors and cross-talk with stroma cells (4–6). Vesicular trafficking-mediated cell surface protein composition serves as a promising regulatory mechanism in several cancer hallmarks (7, 8). Surface protein transportation is known to regulate many cancer behaviors, e.g., E-cadherin during tumorigenesis (9), EGFR during cell migration and hepatocellular carcinoma metastasis (10), and integrins during cell migration (10). However, the regulatory roles and machinery of vesicle trafficking-mediated cell surface protein in BCSC niche orchestration remain poorly defined.

In this study, we investigated the regulatory factors controlling transportation of BCSC surface proteins. Rab13 is a member of the Rab GTPase family (11). We found that Rab13 was highly expressed in BCSC of triple-negative breast cancer (TNBC). Rab GTPases are master regulators of vesicle-mediated protein transportation (12, 13). Transcriptomic analysis showed that Rab13 knockdown (KD) impaired cytokine–cytokine receptor interactions, suggesting a regulatory function of Rab13 in ligand-receptor interactions and tumor-stroma cross-talk. Furthermore, Rab13 mediated cell surface trafficking of two receptors of IL8, i.e., C-X-C chemokine receptor type 1 (CXCR1) and CXCR2. The IL8/CXCR1/2 axis contributes to maintenance of BCSC stemness and chemoresistance (14–16). By controlling membrane transportation of the above receptors, Rab13 sustained the cross-talk between tumor cells and stromal cells in the BCSC niche. Phenotypically, our results showed that Rab13 is involved in BCSC stemness, tumorigenesis, chemoresistance, and TNBC relapse after treatment. These findings suggest that Rab13 is a constructor and maintainer of the BCSC niche, thus providing new evidence for Rab13-mediated BCSC niche as a clinical therapeutic target. Furthermore, we have identified that bardoxolone methyl (C-28 methyl ester of 2-cyano-3, 12-dioxoolen-1, 9-dien-28-oic acid; CDDO-Me) decreases Rab13 expression. CDDO-Me, a synthetic triterpenoid, has been under clinical investigation for the treatment of chronic kidney disease associated with type 2 diabetes and pulmonary hypertension (17). Here, we found that CDDO-Me targets Rab13, remodels BCSC niche, disrupts the connection between tumor cells and stromal cells, and then prevents BCSC stemness in vitro and in vivo.

Patients and tissue samples

Breast cancer tissue chips used for immunofluorescence (IF) and IHC staining were obtained from AlenaBio. Tumor samples used for isolation of cancer-associated fibroblasts (CAF) were obtained from three patients with breast carcinoma (TNBC) at the Yunnan Tumor Hospital (Kunming, Yunnan, China) between 2018 and 2019. Sample information is provided in Supplementary Table S1. All samples were collected from patients with informed consent, and all related procedures were performed with the approval of the Internal Review Board of Kunming Institute of Zoology, Chinese Academy of Sciences (Kunming, Yunnan, China; SMKX-20171110-01).

Bioinformatics analysis of single-cell RNA sequencing data

The single-cell RNA sequencing (scRNA-seq) data for the TNBC cohort were obtained from previous studies (18, 19). In total, five cohorts were used. The original expression matrix files were downloaded from the NCBI Gene Expression Omnibus (GEO) database. More details are provided in Supplementary Methods and Materials.

Transcriptomic RNA-seq and data analysis.

Total RNA was extracted from the HCC1806 cells for next-generation sequencing using TRIzol reagent (Life Technologies). After certain protocols for mRNA sequencing sample preparation, the RNA samples were sequenced on an Illumina X-Ten platform. In total, three biological repetitions were sequenced for the Rab13-KD or control groups, respectively. More details about data analysis are provided in Supplementary Methods and Materials.

Overall survival analysis.

An online database (http://kmplot.com/analysis/) was used to determine the relevance of individual Rab13 protein expression to the overall survival (OS) in patients with basal breast cancer.

Animal study

Female heterozygous mice expressing the polyomavirus middle T-antigen (PyMT) driven by the mouse mammary tumor virus (MMTV) promoter in C57BL/6J background were obtained from Shanghai Model Organisms Center, Inc. and genotyped as previously described (20). Female athymic BALB/c (nu/nu) nude mice (7–8 weeks old) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. All animal studies were approved by the Internal Review Board of Kunming Institute of Zoology, Chinese Academy of Sciences (SMKX-20160506-02).

Tumor growth and limiting dilution transplantation assays

To measure tumor-initiating cell frequency, serial dilutions of TNBC cells were suspended in a 1:1 mixture of PBS and Matrigel (total volume 100 μL) were injected into the mammary fat pads of nude mice. The final count for tumor formation was performed at 8 weeks after tumor cell transplantation. The number of tumor-initiating cells was calculated using the extreme limiting dilution analysis (ELDA) web interface (http://bioinf.wehi.edu.au/software/elda/). Xenografts were dissociated by collagenase type I (1.5 mg/mL) and collagenase type III (1.5 mg/mL) at 37°C with agitation for 30 minutes in DMEM with 10% FBS. Breast cancer cells were subjected to flow cytometry, tumorsphere formation assay, and limiting dilution transplantation assay for a second time. For tumor growth in vivo, luciferase-expressing MDA-MB-231 cells were injected into the mammary fat pads of mice. Bioluminescence signals were imaged, as described in our pervious study (21).

Chemoresistance and relapse in xenografts

5 × 106 MDA-MB-231 cells with or without Rab13-KD were injected into the #4 mammary fat pad of female nude mice, which all developed tumors (∼50 mm3 in size) within 10 days. The mock cell-injected mice (sh-Control) were randomly divided into two groups (untreated or treated). The treated sh-Control group and sh-Rab13 group were intraperitoneally injected with 5 mg/kg doxorubicin every 5 days (3 cycles). The untreated group was intraperitoneally injected with the same volume of DMSO alone. Treatment was started 3 weeks after tumor cell transplantation. The tumor volume was measured every week and calculated by: 0.52 × (width)2 × length.

CDDO-Me treatment

For in vivo drug treatment, 8- to 10-week–old MMTV-PyMT mice with tumor size of 200 to 300 mm3 were divided into two groups (n = 8 tumors from 2 mice/group). CDDO-Me, dissolved in DMSO, was intraperitoneally injected with 5 mg/kg every 3 days (five cycles) into mice of treatment group, and the control mice with DMSO alone. For treatment with CDDO-Me in xenograft model, implantation of MDA-MB-231 was performed as above description. The mice were randomly divided into a control group and a treatment group (n = 10/group) at 2 weeks after tumor cell transplantation, which were intraperitoneally injected with the same volume of DMSO alone or CDDO-Me (2 mg/kg every 3 days, 10 cycles), respectively. At the endpoint of the experiment, the tumors were dissected and subjected to IHC, aldehyde dehydrogenase (ALDH) activity detection, and tumorsphere assay.

Cell culture

Primary CAFs were isolated from fresh TNBC breast cancer samples as previously described (22). When the fibroblast-like cells showed at least 50% confluence, they were detached with trypsin and purified using anti-Fibroblast MicroBeads as described in a previous study (15) according to the manufacturer's instructions. The detailed isolation and purification of primary CAFs are provided in supplemental methods and materials. RAW264.7 (RRID: CVCL_0493), THP-1 (RRID: CVCL_0006), HCC1806 (RRID: CVCL_1258), MDA-MB-231 (RRID: CVCL_0062), and HEK293T cells (RRID: CVCL_0063) were obtained from ATCC. MDA-MB-231-ADR cells were obtained from Professor Xiyun Deng in Central South University (Changsha, Hunan, China). We have validated that MDA-MB-231-ADR cells are resistant to doxorubicin. All cells were authenticated by autosomal short tandem repeat (STR) profiling (Kunming Cell Bank, Kunming Institute of Zoology, Chinese Academy of Sciences) and were free of Mycoplasma contamination as determined by PCR every 6 months. The culture of above cell lines is provided in Supplementary Methods and Materials.

Coculture experiments

Coculture experiments were performed by seeding TNBC breast cancer cells (1 × 105) in the lower chamber and CAF (1 × 105), RAW264.7 (5 × 104), or PMA-primed THP-1 cells (1 × 105) in the upper chamber of a six-well transwell apparatus with a 0.4-μm pore size. Cells cocultured with CAFs for 3 days or with macrophages for 24 to 36 hours were subjected to further analysis.

Conditioned media preparation and cell treatment

For conditioned media (CM) preparation, TNBC cells or primary CAFs that reached 70% to 80% confluence were refed with serum-free DMEM/F12 or DMEM containing antibiotic–antimycotic solution. The cell-cultured CM were collected after 48 hours of incubation. Preparation of tumor-cocultured macrophages (TCM)-CM (THP-CM and RAW-CM) and cell treatment with above CM, IL8, GM-CSF, IL10, or reparixin are provided in Supplementary Methods and Materials. After the cell treatment, cells were subjected to ALDH activity detection or tumorsphere formation assay.

Cell migration assay

To induce differentiation of THP-1 monocytes into macrophages, cells were incubated with phorbol 12-myristate 13-acetate (PMA; 50 nmol/L) at 37°C for 3 hours. Then, THP-1 cells, RAW264.7 cells, or CAFs were placed on the top layer of a cell culture insert with a permeable membrane (5-μm pore size for macrophages or 8-μm pore size for CAFs) and HCC1806 or MDA-MB-231 cells were placed below the cell permeable membrane. More details are provided in Supplementary Methods and Materials.

Plasmid construction and lentivirus transduction

The construction and transduction of short hairpin RNA (shRNA) were conducted as previously described (21). The shRNA sequences are listed in Supplementary Table S2. More details and construction of the wild-type expression plasmid for Rab13 are available in Supplementary Methods and Materials.

Cell-surface protein extraction and detection.

Cell-surface protein extraction and detection were performed as described in the previous study (23). More details are available in Supplementary Methods and Materials.

Western blot analysis

Cell lysates were prepared by lysing cells with RIPA buffer plus complete protease inhibitor. Briefly, protein (30 mg) from each sample was resolved by gel electrophoresis and transferred to a polyvinylidene fluoride (PVDF) membrane. Signals were detected using the super-signal–enhanced chemiluminescence system. Information of all antibodies used for Western blot analysis is listed in Supplementary Methods and Materials.

Immunoprecipitation

Immunoprecipitation was conducted as previously described (21). More details are available in Supplementary Methods and Materials.

Total RNA extraction, reverse transcription, and qRT-PCR

Total RNA was extracted using RNAiso Plus Reagent according to the manufacturer's protocols. Reverse transcription was performed using a PrimeScript RT Reagent Kit. cDNA was subjected to qRT-PCR using a SYBR Premix Ex Taq Kit according to the manufacturer's instructions. Primer sequences are listed in Supplementary Table S3.

Flow cytometry

The detection of CD44+CD24–/low and ALDH+ BCSC population was performed as previously described (21). More details are available in Supplementary Methods and Materials.

Tumorsphere formation assay and colony formation assay

Tumorsphere formation assay and colony formation assay were performed as previously described (21). More details are available in Supplementary Methods and Materials.

Apoptosis analysis

Cells treated with the indicated chemotherapeutic agents (paclitaxel 5 nmol/L, doxorubicin 5 nmol/L, or cisplatin 100 nmol/L) for 12 hours. Apoptosis was determined using an Annexin V Apoptosis Detection Kit according to the manufacturer's instructions. Cells were analyzed using a FACSCanto analyzer (BD Biosciences). Flow cytometry data processing was performed with FlowJo.

Cell viability assay

3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay was conducted to detect the cell viability in cell cytotoxicity assays according to the manufacturer's instructions (CellTiter 96 AQueous One Solution Reagent, Promega). More details are available in Supplementary Methods and Materials.

IHC and IF analysis

Xenograft tumor samples were handled and sectioned as described previously (24). Then, IHC and IF analysis were performed according to previous study (24). The detailed information is provided in Supplementary Methods and Materials.

Compound screening based on luciferase reporter assay

Based on the promoter activity of Rab13, the compounds in DiscoveryProbe FDA-approved Drug Library (APExBIO, catalog no. L1021) were screened. More details are provided in Supplementary Methods and Materials.

Antibodies and reagents

The source of antibodies, key reagents, cell lines, and software were listed in Supplementary Tables S4 and S5.

Statistical analysis

All quantitative data are shown as mean ± SEM. Statistical differences were determined using two-sided t tests or one-/two-way ANOVA followed by Dunnett posthoc multiple comparison test. P values less than 0.05 were considered significant. All statistical analyses were performed using GraphPad Prism v8.0.

Data availability

The raw sequence data reported in this paper were deposited at the NCBI Sequence Read Archive (SRA) portal with bioproject ID (PRJNA699049).

Identification of Rab13 as an upregulated gene associated with BCSCs and BCSC niche

Both CD44+CD24–/low and ALDH+ cell fractions are defined as BCSC populations (25, 26), and the overlapping population (CD44+CD24–/lowALDH+ cells) displays higher tumor initiating capacity than the separate populations themselves (27). TNBC is the most malignant breast cancer subtype and harbors enriched BCSCs (2, 28, 29). To determine the regulatory factors in BCSC, we identified highly expressed genes in CD44+CD24–/low cells, ALDH+ cells and TNBC cells based on published RNA-seq data of single-cell and patient-derived xenografts of metastatic TNBC (18, 19). Sixty-four highly expressed genes were identified in the above three groups (Fig. 1A). Among them, four genes, that is, AP2M1, ATP1B3, BSG, and Rab13, are known regulatory factors of protein localization to the plasma membrane (11, 30–32), which potentially play a role in tumor-stroma cross-talk. Although these four genes were highly expressed in the tumorsphere that is enriched with BCSCs, Rab13 expression showed the largest increase range (Fig. 1B; Supplementary Fig. S1A and S1B). Furthermore, we found that stemness-promoting factors, including c-Myc, Nanog, Sox9, CD44, ALDH1, and Stat3, were highly expressed in the Rab13high population compared with the Rab13low/– population (Fig. 1C). These results imply that Rab13 plays a regulatory role in BCSCs.

Figure 1.

Upregulation of Rab13 is correlated with BCSC niche and poor prognosis of TNBC. A, Venn diagram showing intersection of highly expressed genes among stemness-related groups and TNBC highly expressed group. B, mRNA expression levels of BSG, ATP1B3, AP2M1, and Rab13 in adherent cells (Ac) versus CSC-enriched tumorsphere (Ts; BCSC enrichment) of MDA-MB-231 cells. C, mRNA expression levels of stemness promoting factors in CSCs in Rab13low/– versus Rab13high populations of HCC1806 cells (n = 3). D and E, Correlation of Rab13 and CD206 (D) or αSMA (E) based on IHC staining density in patient samples from local hospital (n = 17 patients). F, mRNA expression levels of Rab13 in TCGA breast cancer cohort. Normal (n = 113 patients), LumA (luminal A, n = 567 patients), LumB (luminal B, n = 207 patients), Her2 (n = 81 patients), and TNBC (n = 234 patients). G, Gene amplification of Rab13 in different types of TCGA breast cancer. LumA (n = 561 patients), LumB (n = 207 patients), Her2 (n = 80 patients), TNBC (n = 227 patients). H, Kaplan–Meier survival analysis showing correlation between OS and Rab13 protein levels in patients with basal breast cancer. I, Representative IHC images of Rab13 and ALDH1 in breast cancer tissues (n = 114 patients). Scale bar, 200 μm. J, ALDH1 expression levels in breast cancer tissues with different Rab13 expression (n = 114 patients). K and L, Expression correlation between Rab13 and ALDH1 based on IHC intensities in breast cancer tissues (n = 114 patients; K), or in subtypes of TNBC (n = 58 patients), luminal (n = 28 patients), and HER2 (n = 22 patients) breast cancer tissues (L). Correlations were calculated using Pearson correlation coefficients. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 1.

Upregulation of Rab13 is correlated with BCSC niche and poor prognosis of TNBC. A, Venn diagram showing intersection of highly expressed genes among stemness-related groups and TNBC highly expressed group. B, mRNA expression levels of BSG, ATP1B3, AP2M1, and Rab13 in adherent cells (Ac) versus CSC-enriched tumorsphere (Ts; BCSC enrichment) of MDA-MB-231 cells. C, mRNA expression levels of stemness promoting factors in CSCs in Rab13low/– versus Rab13high populations of HCC1806 cells (n = 3). D and E, Correlation of Rab13 and CD206 (D) or αSMA (E) based on IHC staining density in patient samples from local hospital (n = 17 patients). F, mRNA expression levels of Rab13 in TCGA breast cancer cohort. Normal (n = 113 patients), LumA (luminal A, n = 567 patients), LumB (luminal B, n = 207 patients), Her2 (n = 81 patients), and TNBC (n = 234 patients). G, Gene amplification of Rab13 in different types of TCGA breast cancer. LumA (n = 561 patients), LumB (n = 207 patients), Her2 (n = 80 patients), TNBC (n = 227 patients). H, Kaplan–Meier survival analysis showing correlation between OS and Rab13 protein levels in patients with basal breast cancer. I, Representative IHC images of Rab13 and ALDH1 in breast cancer tissues (n = 114 patients). Scale bar, 200 μm. J, ALDH1 expression levels in breast cancer tissues with different Rab13 expression (n = 114 patients). K and L, Expression correlation between Rab13 and ALDH1 based on IHC intensities in breast cancer tissues (n = 114 patients; K), or in subtypes of TNBC (n = 58 patients), luminal (n = 28 patients), and HER2 (n = 22 patients) breast cancer tissues (L). Correlations were calculated using Pearson correlation coefficients. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Close modal

To explore the potential association between Rab13 and the BCSC niche, we investigated the cell types around BCSCs in breast cancer patient samples. Tumor-associated macrophages (TAM) and CAFs are important stromal components of the BCSC niche (15, 33). CD206 is a marker of M2-like TAMs with tumor-promoting properties, and alpha-smooth muscle actin (αSMA) is a marker of CAFs (22, 34). We found that Rab13 was highly expressed in the ALDH1+ BCSC population surrounded by CD206+ TAMs and αSMA+ CAFs (Supplementary Fig. S1C). Moreover, both CD206 and αSMA showed high correlation with Rab13 expression in the patients with breast cancer (Fig. 1D and E). These data further highlight the close association of Rab13 with the BCSC niche.

Association of Rab13 expression with breast cancer patient outcome

To further clarify the potential prognostic value of Rab13 in breast cancer, we investigated its expression in clinical samples. Analysis of The Cancer Genome Atlas (TCGA) data and qRT-PCR detection in patient samples showed that expression level of Rab13 was significantly higher in breast tumor samples, including luminal (estrogen receptor and/or progesterone receptor positive), HER2, and TNBC subtypes, than that in normal breast tissue (Fig. 1F; Supplementary Fig. S1D and S1E). Rab13 expression also increased in advanced breast cancer stages (Supplementary Fig. S1F), suggesting a positive correlation between Rab13 and breast cancer malignancy. Copy-number variation (CNV) change drives tumorigenesis, chemoresistance, and cancer progression (35, 36). We calculated the Rab13 amplification percentages using the CNV dataset from TCGA. Rab13 amplification was detected in approximately 35% of TNBC samples and in approximately 10% to 20% of other breast cancer subtypes (Fig. 1G). Meanwhile, TNBC cells showed the highest Rab13 expression levels (Supplementary Fig. S1G). Consistently, Kaplan–Meier plots showed that patients with TNBC with high Rab13 expression exhibited significantly poorer survival than patients with low Rab13 expression (Fig. 1H). Patients with high expression of ALDH1 generally exhibit poor prognosis (26). Thus, we next detected the protein levels of Rab13 and ALDH1 using a tissue microarray (n = 114) and assessed their correlations. Samples were divided into three groups based on Rab13 expression (Fig. 1I). A higher ALDH1 expression level was observed in the high Rab13 expression group (Fig. 1J). Furthermore, a positive correlation between Rab13 and ALDH1 (R = 0.589) was observed in patients with breast cancer (Fig. 1K). Comparing different breast cancer subtypes, a higher correlation was observed between Rab13 and ALDH1 in TNBC (R = 0.675) than in the luminal (R = 0.496) and HER2 (R = 0.432) subtypes (Fig. 1L). These results suggest that Rab13 is clinically associated with BCSC stemness and poor prognosis in breast cancer, especially in TNBC.

Rab13 promotes BCSC stemness

These above results suggested that Rab13 played a regulatory role in BCSCs. Accordingly, we evaluated the functions of Rab13 in BCSC stemness using in vitro and in vivo assays. We first generated Rab13-specific shRNA to silence endogenous Rab13 expression (Rab13-sh) in TNBC cells. Both Rab13-sh#1 and Rab13-sh#2, which induced significant KD effects, were adopted for further study (Supplementary Fig. S2A). Furthermore, using BCSC population detection assay, tumorsphere formation assay, and colony formation assay, we further observed that Rab13-KD resulted in a considerable inhibitory effect on BCSC stemness (Fig. 2AC; Supplementary Fig. S2B–S2G). In addition, we overexpressed Rab13 in HCC1806 and MDA-MB-231 cells using pCDH lentivirus carrying a wild-type Rab13 gene (Supplementary Fig. S2H). Exogenous upregulation of Rab13 increased the proportion of CD44+CD24–/low BCSCs and tumorsphere formation ability of HCC1806 cells (Supplementary Fig. S2I). Moreover, the downregulation of BCSC enrichment was rescued by Rab13 overexpression (Supplementary Fig. S2F and S2G). The expression levels of well-recognized stemness-promoting genes, including Nanog, Sox9, CD44, Stat3, and c-Myc, were inhibited by the silencing of Rab13 in both TNBC lines (Supplementary Fig. S2J and S2K). These results further confirmed the supporting role of Rab13 in BCSC stemness. In addition to in vitro analysis, we performed a tumor xenograft assay using luciferase-labeled MDA-MB-231 cells, and thereafter monitored tumor formation and progression in vivo. Interestingly, the tumor diminished (5/12) or disappeared (7/12) in the stable Rab13-KD group, whereas the tumor grew continuously in the control group (9/10; Fig. 2D), showing the potential inhibitory effects of Rab13-KD on tumor initiation and progression. Accordingly, transplantation of TNBC cells (MDA-MB-231 or HCC1806, respectively) with limiting dilution revealed that Rab13-KD reduced tumor formation frequency (Fig. 2E; Supplementary Fig. S2L; Supplementary Table S6). We further digested the above xenograft tumors (first transplantation) into single cells for second transplantation, with results indicating the same patterns in terms of tumor formation frequency (Fig. 2F; Supplementary Fig. S2L; Supplementary Table S7), confirming the promoting effect of Rab13 on breast tumor initiation. Additionally, we detected BCSC stemness in above xenograft tumors and found that the tumorsphere formation ability as well as ALDH+ and CD44+CD24–/low BCSC proportions decreased markedly in the KD groups (Fig. 2G and H; Supplementary Fig. S2M and S2N). Thus, Rab13 appears to promote both BCSC stemness properties and tumor initiation in vitro and in vivo.

Figure 2.

Promoting role of Rab13 in BCSC stemness. A, Percentage of ALDH+ BCSC population upon Rab13-KD in MDA-MB-231 cells (n = 3). B and C, Tumorsphere formation (B) and clonal formation assay (C) upon Rab13-KD in MDA-MB-231 cells (n = 3). Quantification (top) and representative images (bottom) of tumorsphere and colonies. D, Luciferase-labeled MDA-MB-231 cells were injected into fat pads of nude mice. Top, luciferase signal intensities in mice were recorded every 2 weeks. Bottom, quantification of average luciferase signal for each condition over animal experiment duration. sh-Control (n = 10); sh-Rab13 (n = 12). Limiting dilution xenografts of MDA-MB-231 cells with Rab13-KD. E and F, Representative dissection images (left) as well as tumor incidence (right) of first transplantation (n = 8; E) and second transplantation (n = 6; F). G and H, Xenografts from first (G) and second tumorigenesis (H) were dissociated into single cells, then subjected to FACS analysis of ALDH+ BCSC population (left) or tumorsphere formation assay (right). Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (versus sh-Control group). P/S, p/sec/cm2/sr; wk, weeks.

Figure 2.

Promoting role of Rab13 in BCSC stemness. A, Percentage of ALDH+ BCSC population upon Rab13-KD in MDA-MB-231 cells (n = 3). B and C, Tumorsphere formation (B) and clonal formation assay (C) upon Rab13-KD in MDA-MB-231 cells (n = 3). Quantification (top) and representative images (bottom) of tumorsphere and colonies. D, Luciferase-labeled MDA-MB-231 cells were injected into fat pads of nude mice. Top, luciferase signal intensities in mice were recorded every 2 weeks. Bottom, quantification of average luciferase signal for each condition over animal experiment duration. sh-Control (n = 10); sh-Rab13 (n = 12). Limiting dilution xenografts of MDA-MB-231 cells with Rab13-KD. E and F, Representative dissection images (left) as well as tumor incidence (right) of first transplantation (n = 8; E) and second transplantation (n = 6; F). G and H, Xenografts from first (G) and second tumorigenesis (H) were dissociated into single cells, then subjected to FACS analysis of ALDH+ BCSC population (left) or tumorsphere formation assay (right). Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (versus sh-Control group). P/S, p/sec/cm2/sr; wk, weeks.

Close modal

Rab13 is required for cross-talk between cancer cells and surrounding stromal cells

In view of its correlation with the BCSC niche, we analyzed whether Rab13 was involved in microenvironment-mediated BCSC stemness using a coculture assay. We isolated primary CAFs from 3 patients with breast cancer, and CAFs were validated based on highly expression levels of CAF markers (vimentin and αSMA) and lack of somatic mutations that frequently observed in tumor cells using Western blot and exon sequencing (Supplementary Fig. S3A and S3B). ALDH+ and CD44+CD24–/low BCSC enrichment under coculture was inhibited by the silencing of Rab13 in TNBC cells (Fig. 3A and B; Supplementary Fig. S3C), indicating that tumor cells with Rab13-KD did not respond to the stimulation of macrophage/CAF. We also measured the tumorsphere and colony formation abilities of TNBC cells under the above coculture conditions. Consistently, macrophages/CAFs coculture did not exhibit stemness-promoting effects on TNBC with Rab13-KD (Fig. 3CF; Supplementary Fig. S3D–S3H). To detect whether Rab13 is involved in non-CSC plasticity mediated by stromal cells, non-CSCs (CD44+CD24+ population) were isolated and cocultured with CAFs or THP-1 (Fig. 3G). As shown in Fig. 3H, coculture increased the dedifferentiation of CD44+CD24+ non-CSCs into CD44+CD24–/low BCSCs, confirming that non-CSCs are plastic and can regain the BCSC phenotype under the stimulation of microenvironment (1). Following Rab13-KD, the above dedifferentiation was suppressed (Fig. 3H). These results suggest that Rab13 ensures that tumor cells efficiently engage with macrophages/CAFs in the tumor microenvironment, then supports BCSC stemness.

Figure 3.

Loss of Rab13 impairs macrophage/CAF-mediated BCSC stemness. A and B, ALDH+ (left) and CD44+CD24–/low (right) BCSC population analysis in TNBC cells with Rab13-KD or their mock cells (sh-Control) cocultured with PMA-primed THP-1 macrophages (A) and CAFs (B), respectively (n = 3). C–F, Tumorsphere (C and E) and colony formation assays (D and F) in MDA-MB-231 cells upon Rab13-KD or their mock cells (sh-Control) cocultured with PMA-primed THP-1 macrophages, CAFs, and RAW264.7, respectively (n = 3). G, CD44+CD24+ non-CSCs were isolated and subjected to coculture assay. H, Percentage of CD44+CD24–/low population in Rab13-KD non-CSCs or their mock non-CSCs (sh-Control) cocultured with CAFs and THP-1, respectively (n = 3). Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001. NS, not significant.

Figure 3.

Loss of Rab13 impairs macrophage/CAF-mediated BCSC stemness. A and B, ALDH+ (left) and CD44+CD24–/low (right) BCSC population analysis in TNBC cells with Rab13-KD or their mock cells (sh-Control) cocultured with PMA-primed THP-1 macrophages (A) and CAFs (B), respectively (n = 3). C–F, Tumorsphere (C and E) and colony formation assays (D and F) in MDA-MB-231 cells upon Rab13-KD or their mock cells (sh-Control) cocultured with PMA-primed THP-1 macrophages, CAFs, and RAW264.7, respectively (n = 3). G, CD44+CD24+ non-CSCs were isolated and subjected to coculture assay. H, Percentage of CD44+CD24–/low population in Rab13-KD non-CSCs or their mock non-CSCs (sh-Control) cocultured with CAFs and THP-1, respectively (n = 3). Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001. NS, not significant.

Close modal

As BCSC with high Rab13 expression are surrounded by TAMs and CAFs (Supplementary Fig. S1C), we explored the role of Rab13 in recruitment and education of stromal cells. IHC staining showed that CD68+ TAMs, CD206+ TAMs, and αSMA+ CAFs were enriched in tumor tissues of first and second transplantation, but this enrichment was significantly diminished upon Rab13-KD (Fig. 4A; Supplementary Fig. S4A and S4B). Using transwell assays, TNBC cells significantly increased the migration of macrophages (THP-1 and RAW264.7 cells) and CAFs (Fig. 4B; Supplementary Fig. S4C). Rab13 silencing in TNBC cells markedly reduced the above stromal cell attraction to the control level (Fig. 4B; Supplementary Fig. S4C), indicating that Rab13 is required for tumor cell-mediated infiltration of macrophages and CAFs. In tumor tissue, cancer cells educate macrophages towards the M2 activation status (37), then build a supportive niche (3, 4). We observed that the ratios of CD206+ (M2 TAM marker) TAMs/CD68+ (pan marker of macrophage) TAMs were decreased upon Rab13-KD in tumor tissues (Fig. 4A; Supplementary Fig. S4A and S4B), indicating Rab13 is involved in macrophage education. Therefore, we investigated the ratios of CD206+ M2-like macrophages to total macrophages under coculture conditions. TNBC cells (MDA-MB-231 and HCC1806) increased the ratios of CD206+ cells in both macrophage cell lines (PMA-primed THP-1 and RAW264.7; Fig. 4C; Supplementary Fig. S4D). However, after Rab13-KD in TNBC cells, these elevations were almost completely repressed in the cocultured macrophages (Fig. 4C; Supplementary Fig. S4D), suggesting that Rab13 promotes the educating effect of tumor cells on macrophages. Therefore, Rab13 promotes BCSC stemness and regulates tumor cell plasticity by establishing the BCSC niche.

Figure 4.

Rab13 depletion reduces breast cancer cell–mediated stromal cell recruitment and education. A, IHC staining of αSMA, CD68, and CD206 as well as hematoxylin and eosin (H&E) staining in sections of HCC1806 tumor xenografts (the first transplantation). Representative images of IHC staining (left) and quantification (right) of αSMA+, CD68+, and CD206+ signals by percentages of brown color staining intensity in IHC images as well as ratio of CD206+ TAMs/CD68+ TAMs (CD206/CD68 ratio). IHC staining of CD68 and CD206 as well as hematoxylin and eosin staining were performed on adjacent section. Scale bar, 100 μm. n = 5–9. B, Transwell migration of macrophage (M; THP-1 macrophages and RAW264.7 macrophages) or CAFs cocultured with TNBC cells expressing indicated constructs. Representative photographs (top) and statistical results of migrating cells number (bottom). C, Percentages of CD206+ cells in THP-1 macrophages cocultured with TNBC cells expressing indicated constructs (n = 3). Top, the representative images; bottom, the quantification data. D and E, mRNA (D) and protein (E) expression levels of GM-CSF and IL10 in TNBC cells with Rab13-KD or their mock cells (sh-Control; n = 3). F and G, Transwell migration of THP-1 macrophages, RAW264.7 macrophages, or CAFs (F) as well as the percentages of CD206+ cells in macrophages (G) treated with CM of TNBC (expressing indicated constructs) with or without GM-CSF or IL10, respectively. NC, macrophages or CAFs only; sh-C, sh-Control; sh-1, Rab13-sh#1; sh-2, Rab13-sh#2. P, versus sh-C-CM group. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (A–E, versus sh-Control group).

Figure 4.

Rab13 depletion reduces breast cancer cell–mediated stromal cell recruitment and education. A, IHC staining of αSMA, CD68, and CD206 as well as hematoxylin and eosin (H&E) staining in sections of HCC1806 tumor xenografts (the first transplantation). Representative images of IHC staining (left) and quantification (right) of αSMA+, CD68+, and CD206+ signals by percentages of brown color staining intensity in IHC images as well as ratio of CD206+ TAMs/CD68+ TAMs (CD206/CD68 ratio). IHC staining of CD68 and CD206 as well as hematoxylin and eosin staining were performed on adjacent section. Scale bar, 100 μm. n = 5–9. B, Transwell migration of macrophage (M; THP-1 macrophages and RAW264.7 macrophages) or CAFs cocultured with TNBC cells expressing indicated constructs. Representative photographs (top) and statistical results of migrating cells number (bottom). C, Percentages of CD206+ cells in THP-1 macrophages cocultured with TNBC cells expressing indicated constructs (n = 3). Top, the representative images; bottom, the quantification data. D and E, mRNA (D) and protein (E) expression levels of GM-CSF and IL10 in TNBC cells with Rab13-KD or their mock cells (sh-Control; n = 3). F and G, Transwell migration of THP-1 macrophages, RAW264.7 macrophages, or CAFs (F) as well as the percentages of CD206+ cells in macrophages (G) treated with CM of TNBC (expressing indicated constructs) with or without GM-CSF or IL10, respectively. NC, macrophages or CAFs only; sh-C, sh-Control; sh-1, Rab13-sh#1; sh-2, Rab13-sh#2. P, versus sh-C-CM group. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (A–E, versus sh-Control group).

Close modal

We next detected the expression levels of cytokines that contributes to the recruitment and polarization of macrophages (3, 38). Among the cytokines, GM-CSF and IL10 showed the largest fold changes upon Rab13-KD in both TNBC cell lines (Fig. 4D and E). The mRNA expression of other factors, i.e., CC chemokine ligand 2 (CCL2), IL6, IL8, oncostatin M (OSM), TGFβ, macrophage inhibitory cytokine-1 (MIC1), CCL17, IL4, macrophage colony-stimulating factor-1 (CSF1), CXC chemokine ligand 4 (CXCL4), IL13, and bone morphogenetic protein 7 (BMP7), did not show conclusive changes (Supplementary Fig. S4E). We further observed whether exogenous GM-CSF and IL10 proteins could recover the deduction of stromal recruitment and education induced by Rab13-KD. Treatment with CM of MDA-MB-231 (TNBC-CM) significantly increased the migration of macrophages and CAFs, and incubation with CM of TNBC with Rab13-KD (Rab13-KD-CM, sh-1/2-CM) markedly reduced stromal cell's attraction to the control levels (negative control; NC group; Fig. 4F). However, the addition of GM-CSF to Rab13-KD-CM markedly increased stromal cell's attraction to the levels of TNBC-CM group (sh-Control-CM; sh-C-CM; Fig. 4F). These functional data support that Rab13 mediates migration of stromal cells during tumor cell-macrophage interactions through GM-CSF. Moreover, compared with sh-C-CM treatment, the ratios of the CD206+ cells in macrophage cell lines decreased under Rab13-KD-CM treatment (Fig. 4G). These reductions were almost completely rescued by the addition of recombinant IL10 protein to Rab13-KD-CM (Fig. 4G), thus suggesting the role of Rab13 in TAM M2 activation mediated by IL10. Taken together, these results suggest that Rab13 in tumor cells promotes the recruitment and education of macrophages through GM-CSF and IL10, respectively, thereby demonstrating the supporting role of Rab13 in tumor-stroma cross-talk.

Rab13 controls plasma membrane trafficking of CXCR1/2

To elucidate the molecular mechanisms of Rab13 in tumor-stroma cross-talk and BCSC stemness promotion, we performed transcriptomic analysis following Rab13 silencing. Results showed that cytokine–cytokine receptor interactions, which are responsible for interactions between cancer cells and their microenvironment, were clearly downregulated upon Rab13-KD in TNBC cells (Fig. 5A). Based on the important roles of Rab GTPases in vesicular trafficking (5), we speculated that Rab13 potentially controlled the cell surface translocation of certain receptors, which made tumor cells engage with the surrounding niche. Then, we clarified potentially extracellular factors involved in Rab13-mediated cytokine–cytokine receptor interactions by qRT-PCR detection (Supplementary Fig. S4F–S4J). Under coculture condition, IL8 enrichment in macrophage was inhibited by Rab13-KD in tumor cells (Supplementary Fig. S4F). Gene set enrichment analysis (GSEA) further revealed that the chemokine signaling pathway was reduced upon Rab13-KD in HCC1806 cells (Fig. 5B). The IL8/CXCR1/2 axis belongs to the chemokine signaling pathway. We investigated whether Rab13 controlled the membrane translocation of CXCR1/2 via mediating vesicular trafficking. Results showed that loss of Rab13 did not change the amount of protein in the total cell extract, whereas the plasma membrane fractions of CXCR1 and CXCR2 decreased markedly (Fig. 5C). IF staining showed that both CXCR1 and CXCR2 were ubiquitously transported from perinuclear regions to the cell periphery under stimulation by ligand IL8. Silencing of Rab13 led to the retention of CXCR1 and CXCR2 in small perinuclear regions (Fig. 5D and E), suggesting that neither CXCR1 nor CXCR2 could be translocated to other cell compartments, including the cell membrane. Furthermore, coimmunoprecipitation (co-IP) demonstrated clear interactions between the Rab13 and CXCR1/2 proteins in both HCC1806 and MDA-MB-231 cells (Fig. 5F). The trafficking of adaptor protein complex-1 (AP-1) /clathrin-coated vesicles is controlled by multiple Rab GTPases (39). Accordingly, our results showed that membrane localization of CXCR1/2 was completely diminished by clathrin inhibitor treatment (Supplementary Fig. S5A). As an adapter protein in clathrin-coated vesicles, AP1B1 destines cargo translocation to the plasma membrane by controlling cargo sorting in the endosome (40). Rab13 is colocalized with AP1B1 in activated endosomes (perinuclear regions; ref. 41). Therefore, we examined whether AP1B1 is a partner of Rab13 during the cell surface translocation of CXCR1/2. Co-IP experiments showed obvious interactions between AP1B1 and CXCR1/2 in both HCC1806 and MDA-MB-231 cells (Fig. 5G). After silencing of AP1B1 by shRNAs (Supplementary Fig. S5B; Fig. 5H), the interactions between Rab13 and CXCR1/2 were weakened (Fig. 5H), and the membrane expression of CXCR1/2 decreased (Supplementary Fig. S5C). Furthermore, we found that tumorsphere formation ability decreased significantly in AP1B1-KD cells (Supplementary Fig. S5D), indicating the promoting effect of AP1B1 on BCSC stemness. Therefore, Rab13 regulates the trafficking of AP1B1/clathrin-coated vesicles and controls cell surface translocation of CXCR1/2, thereby leading to promotion of BCSC stemness.

Figure 5.

Rab13 controls membrane translocation of CXCR1/2. A and B, GSEA plots of cytokine–cytokine receptor interactions (A) and chemokine signaling pathways (B) in HCC1806 cells with Rab13-KD and their mock cells (sh-Control). C, Protein expression levels of CXCR1/2 in plasma membrane fraction (PM) and total cell lysate (total protein) of TNBC cells with Rab13-KD and their mock cells (sh-Control). D and E, Representative images of Rab13 and CXCR1 (D) or CXCR2 (E) immunofluorescent staining in TNBC cells with Rab13-KD and their mock cells (sh-Control) treated with or without IL8 (20 ng/mL). Scale bar, 10 μm. F, Co-IP of Rab13 and CXCR1/2 in HCC1806 cells. G, Co-IP of indicated proteins in HCC1806 and MDA-MB-231 cells. H, Co-IP of indicated proteins in MDA-MB-231 cells with AP1B1-KD and their mock cells (sh-Control). ES, enrichment score; NES, normalized enrichment scale; WT, wild-type; min, minutes.

Figure 5.

Rab13 controls membrane translocation of CXCR1/2. A and B, GSEA plots of cytokine–cytokine receptor interactions (A) and chemokine signaling pathways (B) in HCC1806 cells with Rab13-KD and their mock cells (sh-Control). C, Protein expression levels of CXCR1/2 in plasma membrane fraction (PM) and total cell lysate (total protein) of TNBC cells with Rab13-KD and their mock cells (sh-Control). D and E, Representative images of Rab13 and CXCR1 (D) or CXCR2 (E) immunofluorescent staining in TNBC cells with Rab13-KD and their mock cells (sh-Control) treated with or without IL8 (20 ng/mL). Scale bar, 10 μm. F, Co-IP of Rab13 and CXCR1/2 in HCC1806 cells. G, Co-IP of indicated proteins in HCC1806 and MDA-MB-231 cells. H, Co-IP of indicated proteins in MDA-MB-231 cells with AP1B1-KD and their mock cells (sh-Control). ES, enrichment score; NES, normalized enrichment scale; WT, wild-type; min, minutes.

Close modal

Rab13 promotes microenvironment-mediated BCSC stemness via IL8-CXCR1/2 axis

Given the important role of the IL8/CXCR1/2 axis in BCSC stemness (14, 15), we confirmed whether the translocation of CXCR1/2 regulated by Rab13 is responsible for maintaining BCSC stemness. Using BCSC population enrichment (CD44+CD24–/low and ALDH+) assay, tumorsphere formation assay, and colony formation assay, we found that IL8 ligand stimulation increased BCSC stemness, while tumor cells with Rab13-KD showed a limited response to IL8 stimulation (Fig. 6AD; Supplementary Fig. S6A and S6B). We also stimulated a non-CSC population using IL8 and found that treatment promoted the dedifferentiation of CD44+CD24+ non-CSCs into CD44+CD24–/low BCSCs. Following Rab13-KD, the promoting effects of IL8 on non-CSC dedifferentiation were inhibited (Fig. 6E). Since STAT3 and AKT are well-recognized downstream signaling of IL8 (14, 16, 42), we measured the phosphorylation levels of STAT3 at Y705 and AKT at Serine 473 by specific antibodies. The loss of Rab13 prevented activations of STAT3 and AKT signaling pathways in both HCC1806 and MDA-MB-231 cells (Fig. 6F; Supplementary Fig. S6C). Based on Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of RNA-seq data, the JAK/STAT signaling pathway was also downregulated in the HCC1806 cells following Rab13-KD (Fig. 6G). These findings suggest that Rab13 is required for IL8/CXCR1/2 axis activating its downstream STAT3 and AKT, confirming that Rab13 controls the response of tumor cells to IL8 stimulation. We further confirmed the above observations using a coculture assay. Specifically, coculture with macrophages and CAFs activated STAT3 and AKT phosphorylation in the sh-Control group, but these activations were inhibited in Rab13-KD TNBC cells (Fig. 6H). To further demonstrate the mediating role of CXCR1/2 controlled by Rab13, we carried out rescue experiments by overexpressing Rab13 and blocking CXCR1/2 using its inhibitor, reparixin. Rab13 overexpression and CM treatment promoted BCSC stemness (BCSC marker enrichment and tumorsphere formation ability), while reparixin administration markedly decreased CAF/TCM-CM-induced BCSC enrichment to the control level (vector group; Fig. 6I; Supplementary Fig. S6D and S6E). Moreover, in wild-type TNBC cells, addition of reparixin inhibited the promoting effects of IL8 and CAF/TCM-CM on BCSC stemness (Fig. 6J; Supplementary Fig. S6F–S6H), whereas, in the Rab13-KD TNBC cells, administration of reparixin did not have an inhibitory effect under IL8 or CAF/TCM-CM treatment (Fig. 6J; Supplementary Fig. S6F–S6H). These results suggest that Rab13 guarantees the response of tumor cells to autocrine/paracrine IL8 in the BCSC niche, leading to the promotion of BCSC stemness.

Figure 6.

Rab13 maintains BCSC stemness by facilitating IL8/CXCR1/2 axis. A, Percentages of CD44+CD24–/low and ALDH+ populations in Rab13-KD TNBC cells or their mock cells treated with or without IL8 (20 ng/mL) (n = 3). B and C, Tumorsphere (B) and colony formation (C) of Rab13-KD TNBC cells or their mock cells treated with or without IL8 (20 ng/mL). Left, representative images; right, quantification data. (n = 3). D, Quantification of soft agar colony formation in Rab13-KD TNBC cells or their mock cells treated with or without IL8 (20 ng/mL; n = 3). E, Percentages of CD44+CD24–/low populations in Rab13-KD non-CSCs or their mock non-CSCs (sh-Control) treated with or without IL8 (20 ng/mL; n = 3). F, Western blotting showing expression levels of indicated proteins or phosphorylation sites in Rab13-KD cells or their mock cells treated with or without IL8 (20 ng/mL). G, GSEA plot of JAK/STAT signaling pathway in HCC1806 cells with Rab13-KD and their mock cells (sh-Control). H, After coculture with macrophages or CAFs, expression levels of indicated proteins or phosphorylation sites were detected by Western blotting in HCC1806 or MDA-MB-231 cells with or without Rab13-KD. I, Percentages of ALDH+ BCSC populations in MDA-MB-231 cells with or without Rab13-overexpression (Rab13-OE) treated/untreated with CM of CAFs (CAF-CM) or CM of tumor cell–cocultured macrophage (TCM; RAW264.7-CM, RAW-CM) or CM plus reparixin (n = 3). J, Percentages of CD44+CD24–/low BCSC populations in Rab13-KD HCC1806 cells or their mock cells treated with or without IL8/CAF-CM/TCM-CM (THP-1-CM: THP-CM) or IL8/CM plus reparixin. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001. ES, enrichment score; NES, normalized enrichment scale; NS, not significant; WT, wild-type.

Figure 6.

Rab13 maintains BCSC stemness by facilitating IL8/CXCR1/2 axis. A, Percentages of CD44+CD24–/low and ALDH+ populations in Rab13-KD TNBC cells or their mock cells treated with or without IL8 (20 ng/mL) (n = 3). B and C, Tumorsphere (B) and colony formation (C) of Rab13-KD TNBC cells or their mock cells treated with or without IL8 (20 ng/mL). Left, representative images; right, quantification data. (n = 3). D, Quantification of soft agar colony formation in Rab13-KD TNBC cells or their mock cells treated with or without IL8 (20 ng/mL; n = 3). E, Percentages of CD44+CD24–/low populations in Rab13-KD non-CSCs or their mock non-CSCs (sh-Control) treated with or without IL8 (20 ng/mL; n = 3). F, Western blotting showing expression levels of indicated proteins or phosphorylation sites in Rab13-KD cells or their mock cells treated with or without IL8 (20 ng/mL). G, GSEA plot of JAK/STAT signaling pathway in HCC1806 cells with Rab13-KD and their mock cells (sh-Control). H, After coculture with macrophages or CAFs, expression levels of indicated proteins or phosphorylation sites were detected by Western blotting in HCC1806 or MDA-MB-231 cells with or without Rab13-KD. I, Percentages of ALDH+ BCSC populations in MDA-MB-231 cells with or without Rab13-overexpression (Rab13-OE) treated/untreated with CM of CAFs (CAF-CM) or CM of tumor cell–cocultured macrophage (TCM; RAW264.7-CM, RAW-CM) or CM plus reparixin (n = 3). J, Percentages of CD44+CD24–/low BCSC populations in Rab13-KD HCC1806 cells or their mock cells treated with or without IL8/CAF-CM/TCM-CM (THP-1-CM: THP-CM) or IL8/CM plus reparixin. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001. ES, enrichment score; NES, normalized enrichment scale; NS, not significant; WT, wild-type.

Close modal

Loss of Rab13 increases chemotherapy sensitivity of TNBC through BCSCs

As failure to eliminate CSCs is a major reason for resistance to therapy (2, 43, 44), we investigated whether Rab13 silencing improves therapeutic efficacy in TNBC. Kaplan–Meier plots showed that Rab13 upregulation was correlated with poor prognosis in patients with basal breast cancer exposed to chemotherapy (Fig. 7A). We next explored whether Rab13 silencing increased chemotherapy sensitivity. Results showed that Rab13-KD led to a much lower cell viability dosage (IC50) for chemotherapy (Fig. 7B; Supplementary Fig. S7A). Furthermore, TNBC cells with Rab13-KD were more susceptible to chemotherapy-induced cell apoptosis (Fig. 7C; Supplementary Fig. S7B). These results indicated that Rab13 silencing increased the sensitivity of TNBC cells to chemotherapies, including doxorubicin, paclitaxel, and cisplatin, compared with that of the sh-Control cells. We further examined whether Rab13 promoted chemoresistance by sustaining BCSC stemness. Consistent with previous reports that CSCs were enriched after chemotherapies (21, 45–48), we discovered that treatments with doxorubicin, paclitaxel, and cisplatin promoted ALDH+ and CD44+CD24–/low BCSC enrichment as well as tumorsphere and colony formation abilities of TNBC cells (Fig. 7DF; Supplementary Fig. S7C). However, the above BCSC-promoting effects induced by chemotherapies were prevented upon Rab13-KD in the TNBC cells (Fig. 7DF; Supplementary Fig. S7C). To confirm whether Rab13 contributes to chemoresistance in vivo, we measured tumor volume in a xenograft mouse model exposed to single doxorubicin or Rab13 shRNA plus doxorubicin. The scheme of doxorubicin treatment is described in Fig. 7G. Although tumor volume significantly decreased under doxorubicin treatment alone, the tumors rapidly regrew after the cessation of treatment (Fig. 7H), and tumor volume was finally indistinguishable between the doxorubicin-treated and untreated sh-Control groups at the experimental end point (Fig. 7H and I). These results suggest that tumors initiated by intact MDA-MB-231 cells are sensitive to doxorubicin, but relapse occurs from the original site after treatment ends. Notably, loss of Rab13 not only blocked tumor growth but also inhibited or delayed tumor relapse (Fig. 7H). The dissected tumors confirmed that tumor volume was smallest in the Rab13-KD group (sh-Rab13 + doxorubicin; Fig. 7I). The survival rates of the above three groups were also calculated. The Rab13-KD mice exhibited a much better survival rate compared with the other groups (Fig. 7J). To determine whether BCSCs were diminished within above tumors, we detected the BCSC population proportion and tumorsphere formation ability in vivo. As expected, Rab13-KD decreased doxorubicin-induced BCSC enrichment in tumors formed by both HCC1806 and MDA-MB-231 cells (Fig. 7K and L). These observations indicate that combining Rab13-KD with chemotherapy may exhibit better therapeutic efficacy than a single chemodrug treatment, suggesting that Rab13-KD targets BCSCs and increases the sensitivity of TNBC cells to chemotherapy.

Figure 7.

Loss of Rab13 increases chemotherapy sensitivity and inhibits tumor relapse of breast cancer through BCSC. A, Kaplan–Meier survival analysis showing correlation between OS and Rab13 expression levels in patients with basal breast cancer exposed to chemotherapy. B, Cell viability after treatment with different concentrations of doxorubicin in HCC1806 cells (n = 6). C, FACS analysis of apoptotic cells following treatment with doxorubicin (DOX), paclitaxel (PTX), and cisplatin (CIS) in HCC1806 cells with Rab13-KD and their mock cells (sh-Control). D–F, FACS analysis of CD44+CD24–/low (D) and ALDH+ (E) BCSC populations as well as tumorsphere formation assay (F) following treatment with doxorubicin, paclitaxel, or cisplatin in Rab13-KD breast cancer cells and their mock cells. G, The scheme of doxorubicin treatment. H, Volume of MDA-MB-231 tumor xenografts with or without Rab13-KD (sh-Control or sh-Rab13) untreated or treated with doxorubicin (n = 12). I, Photographs of MDA-MB-231 tumor xenografts at experimental endpoint. J, Kaplan–Meier survival analysis. P values were calculated by log-rank test (n = 12; P < 0.001). K and L, FACS analysis of CD44+CD24–/low and ALDH+ BCSC populations as well as tumorsphere assay using xenografted tumors generated by HCC1806 cells (K) and MDA-MB-231 cells (L) with or without Rab13-KD under doxorubicin treatment. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (versus control group). conc., concentration; NS, not significant.

Figure 7.

Loss of Rab13 increases chemotherapy sensitivity and inhibits tumor relapse of breast cancer through BCSC. A, Kaplan–Meier survival analysis showing correlation between OS and Rab13 expression levels in patients with basal breast cancer exposed to chemotherapy. B, Cell viability after treatment with different concentrations of doxorubicin in HCC1806 cells (n = 6). C, FACS analysis of apoptotic cells following treatment with doxorubicin (DOX), paclitaxel (PTX), and cisplatin (CIS) in HCC1806 cells with Rab13-KD and their mock cells (sh-Control). D–F, FACS analysis of CD44+CD24–/low (D) and ALDH+ (E) BCSC populations as well as tumorsphere formation assay (F) following treatment with doxorubicin, paclitaxel, or cisplatin in Rab13-KD breast cancer cells and their mock cells. G, The scheme of doxorubicin treatment. H, Volume of MDA-MB-231 tumor xenografts with or without Rab13-KD (sh-Control or sh-Rab13) untreated or treated with doxorubicin (n = 12). I, Photographs of MDA-MB-231 tumor xenografts at experimental endpoint. J, Kaplan–Meier survival analysis. P values were calculated by log-rank test (n = 12; P < 0.001). K and L, FACS analysis of CD44+CD24–/low and ALDH+ BCSC populations as well as tumorsphere assay using xenografted tumors generated by HCC1806 cells (K) and MDA-MB-231 cells (L) with or without Rab13-KD under doxorubicin treatment. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (versus control group). conc., concentration; NS, not significant.

Close modal

Next, doxorubicin-resistant MDA-MB-231 cells (MDA-MB-231-ADR) were used for gene expression profiling. In comparison with parental cells (MDA-MB-231-P), the expression levels of Rab13 and CXCR1/2 were much higher in the MDA-MB-231-ADR cells (Supplementary Fig. S7D). The tumorphere formation ability of MDA-MB-231-P and MDA-MB-231-ADR was inhibited upon Rab13-KD or AP1B1-KD, respectively (Supplementary Fig. S7E and S7F). These results further confirm the promoting role of Rab13 in BCSC-mediated chemoresistance.

Identification of CDDO-Me inhibiting Rab13 expression and BCSC stemness

To search for inhibitors targeting Rab13, we screened a library of FDA-approved drugs using a luciferase reporter assay for evaluating Rab13 promoter activity. Our lead compound was CDDO-Me with the formula C32H43NO4 (Fig. 8A). The administration of CDDO-Me decreased the luciferase activity in a dose-dependent manner in HEK293 cells transfected with Rab13-promoter luciferase reporter (Fig. 8B). Furthermore, CDDO-Me treatment decreased Rab13 expression level in breast cancer cells and breast tumor tissues of MMTV-PyMT mice (Fig. 8C and D; Supplementary Fig. S8A), confirming the inhibitory effect of CDDO-Me on Rab13 expression. Moreover, the ALDH+ BCSC population and tumorsphere formation ability clearly decreased in a dose-dependent manner when breast cancer cells were treated with CDDO-Me (Fig. 8E and F). These results suggested the inhibitory effect of CDDO-Me on BCSC stemness, which was consistent with the observation in Rab13-KD cells (Fig. 2). To determine the effect of CDDO-Me on niche-mediated BCSC stemness, we conducted BCSC population detection and tumorsphere formation assay under the treatment of TCM/CAF-CM/IL8 plus CDDO-Me. The results showed that CDDO-Me administration depleted the promotion effects of TCM/CAF-CM and IL8 on BCSC stemness, respectively (Fig. 8GI; Supplementary Fig. S8B–S8E), providing further confirmation that CDDO-Me treatment limited the response of BCSCs to stromal stimulation. In addition, CDDO-Me treatment led to a much lower cell viability dosage (IC50) for chemotherapy (Fig. 8J; Supplementary Fig. S8F), indicating that CDDO-Me treatment increases chemotherapy sensitivity in breast cancer. To further evaluate the effect of CDDO-Me on BCSCs in vivo, we administered CDDO-Me to MDA-MB-231 breast tumor xenografts and MMTV-PyMT mice with spontaneous breast tumors. As shown in Fig. 8K and Supplementary Fig. S8G, CDDO-Me treatment decreased the size of above tumors. Furthermore, the inhibitory effects of CDDO-Me treatment on tumorsphere formation ability and ALDH activity were observed in above xenograft tumor cells and primary MMTV-PyMT breast tumor cells (Fig. 8L and M). Finally, IHC staining showed that CDDO-Me administration not only decreased Rab13 expression but also significantly prevented the enrichment of CD68+ TAMs, CD206+ TAMs, and αSMA+ CAFs as well as declined the ratio of CD206+ TAMs/CD68+ TAMs in MMTV-PyMT breast tumors (Fig. 8D and 8N; Supplementary Fig. S8H), suggesting the inhibitory effects of CDDO-Me on TAM and CAF recruitment and niche orchestration. Taken together, these results suggest that CDDO-Me targets Rab13, prevents tumor-stromal cross-talk, and then inhibits TAM/CAF-induced BCSC stemness, raising the possibility that CDDO-Me serves as an agent targeting Rab13 and tumor microenvironment.

Figure 8.

CDDO-Me decreases Rab13 expression and BCSC stemness in vitro and in vivo. A, Chemical structure of CDDO-Me. B, The luciferase activity of Rab13-promoter reporter treated by CDDO-Me (n = 6). C, Protein expression levels of Rab13 treated with CDDO-Me with indicated concentrations for 48 hours in MDA-MB-231 cells. D, IHC staining showing Rab13 expression treated with CDDO-Me in breast tumor tissue of MMTV-PyMT mice. Left, quantification of Rab13+ signals by percentages of brown color staining intensity in IHC images. Right, representative images of IHC staining. Scale bar, 100 μm. (Control, n = 6; CDDO-Me, n = 4). E and F, The effects of CDDO-Me treatment on ALDH activity (n = 3; E) and tumorsphere formation ability (n = 8; F) of MDA-MB-231 cells. G–I, The effects of CDDO-Me treatment on ALDH activity (n = 3; G) and tumorsphere formation ability (H and I) of MDA-MB-231 cells treated with CM of tumor cell–cocultured macrophage (TCM, THP-CM, THP-1-CM; RAW-CM, RAW264.7-CM) or IL8 (20 ng/mL).H,n = 6; I,n = 9. J, Cell viability after treatment with various concentrations of doxorubicin (DOX) treated/untreated with RAW-CM/IL8 or RAW-CM/IL8 plus CDDO-Me (160 nmol/L; n = 6) in MDA-MB-231 cells. Xenografted tumors generated by MDA-MB-231 cells and spontaneous tumors in MMTV-PyMT mice treated with CDDO-Me. K, The tumor volume (left) during tumor growth as well as tumor weight (right) and representative dissection images of tumors xenografts (bottom) at the end point of treatment in MDA-MB-231 xenograft model (n =10). L and M, ALDH activity (n = 3; L) and tumor sphere formation ability (M) in primary cells of MDA-MB-231 tumor xenografts (left; n = 7) or MMTV-PyMT breast tumor (right; n = 9). N, Quantification of αSMA+, CD68+, and CD206+ signals by percentages of brown color staining intensity in IHC images as well as ratios of CD206+ TAMs/CD68+ TAMs (CD206/CD68 ratio) in MMTV-PyMT tumors tissue (Control, n = 6; CDDO-Me, n = 4). O, Working model. Rab13 controls plasma membrane translocation of CXCR1/2 by AP1B-mediated vesicle trafficking, activates STAT3 and AKT pathways, and upregulates GM-CSF, IL10, and IL8 in tumor cells and niche cells, respectively, which forms a positive feedback loop for BCSC niche orchestration, leading to promotion of BCSC stemness and chemoresistance. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001. conc., concentration; NS, not significant.

Figure 8.

CDDO-Me decreases Rab13 expression and BCSC stemness in vitro and in vivo. A, Chemical structure of CDDO-Me. B, The luciferase activity of Rab13-promoter reporter treated by CDDO-Me (n = 6). C, Protein expression levels of Rab13 treated with CDDO-Me with indicated concentrations for 48 hours in MDA-MB-231 cells. D, IHC staining showing Rab13 expression treated with CDDO-Me in breast tumor tissue of MMTV-PyMT mice. Left, quantification of Rab13+ signals by percentages of brown color staining intensity in IHC images. Right, representative images of IHC staining. Scale bar, 100 μm. (Control, n = 6; CDDO-Me, n = 4). E and F, The effects of CDDO-Me treatment on ALDH activity (n = 3; E) and tumorsphere formation ability (n = 8; F) of MDA-MB-231 cells. G–I, The effects of CDDO-Me treatment on ALDH activity (n = 3; G) and tumorsphere formation ability (H and I) of MDA-MB-231 cells treated with CM of tumor cell–cocultured macrophage (TCM, THP-CM, THP-1-CM; RAW-CM, RAW264.7-CM) or IL8 (20 ng/mL).H,n = 6; I,n = 9. J, Cell viability after treatment with various concentrations of doxorubicin (DOX) treated/untreated with RAW-CM/IL8 or RAW-CM/IL8 plus CDDO-Me (160 nmol/L; n = 6) in MDA-MB-231 cells. Xenografted tumors generated by MDA-MB-231 cells and spontaneous tumors in MMTV-PyMT mice treated with CDDO-Me. K, The tumor volume (left) during tumor growth as well as tumor weight (right) and representative dissection images of tumors xenografts (bottom) at the end point of treatment in MDA-MB-231 xenograft model (n =10). L and M, ALDH activity (n = 3; L) and tumor sphere formation ability (M) in primary cells of MDA-MB-231 tumor xenografts (left; n = 7) or MMTV-PyMT breast tumor (right; n = 9). N, Quantification of αSMA+, CD68+, and CD206+ signals by percentages of brown color staining intensity in IHC images as well as ratios of CD206+ TAMs/CD68+ TAMs (CD206/CD68 ratio) in MMTV-PyMT tumors tissue (Control, n = 6; CDDO-Me, n = 4). O, Working model. Rab13 controls plasma membrane translocation of CXCR1/2 by AP1B-mediated vesicle trafficking, activates STAT3 and AKT pathways, and upregulates GM-CSF, IL10, and IL8 in tumor cells and niche cells, respectively, which forms a positive feedback loop for BCSC niche orchestration, leading to promotion of BCSC stemness and chemoresistance. Data are mean ± SEM. *, P < 0.05; **, P < 0.01; ***, P < 0.001. conc., concentration; NS, not significant.

Close modal

Tumor microenvironment contributes to therapy resistance and tumor recurrence by maintaining CSC stemnesss. Here, we showed that Rab13 ensured cell surface delivery of CXCR1/2 by mediating AP1B1/clathrin vesicle trafficking, and established cross-talk between tumor cells and surrounding TAMs/CAFs, and orchestrated a supportive niche to maintain BCSC stemness (Fig. 8O). Furthermore, silencing of Rab13 increased chemotherapy sensitivity in vitro and impaired BCSC stemness, tumorigenesis, and tumor relapse after treatment in vivo. These results indicate that Rab13 is one of the architects and maintainers of tumor-stroma cross-talk, suggesting a potential target of the BCSC niche in TNBC treatment.

The IL8/CXCR1/2 axis maintains CSC stemness by activating downstream pathways, including STAT3 and AKT (14, 16, 42). Consistently, Rab13-KD inhibited activation of STAT3 and AKT in the TNBC cells treated with IL8 or cocultured with macrophages/CAFs, suggesting that Rab13 play a promoting role in the niche-activated STAT3 and AKT pathways. In addition, the protein level of Rab13 was upregulated in the TNBC cells upon IL8 treatment or macrophages/CAF coculture (Fig. 6F and H). By analyzing chromatin immunoprecipitation sequencing (ChIP-seq) data (49), we found that STAT3 was a potential transcription factor of Rab13 in TNBC (Supplementary Fig. S8I), suggesting the possibility that STAT3 may promote Rab13 expression. Rab13 silencing inhibited the expression of GM-CSF and IL10 in the TNBC cells. In addition, IL8 expression in cocultured macrophages was impaired by Rab13-KD in the TNBC cells under coculture conditions. The AKT pathway promotes the expression of inflammatory factors, including GM-CSF and IL10 in cancer, which facilitates the infiltration and polarization of macrophages in the BCSC niche (38, 50–54). Thus, Rab13 most likely increased the expression of GM-CSF and IL10 by activating the AKT pathway in the tumor cells, thereby enhancing the recruitment and education of macrophages and indirectly leading to the upregulation of IL8 in macrophages under tumor cell/macrophages coculture. In turn, upregulation of IL8 in the BCSC niche activated STAT3, increased Rab13 expression, and maintained BCSC stemness. Therefore, Rab13 mediated the cell surface expression of CXCR1/2, activation of STAT3 and AKT, and upregulation of GM-CSF and IL10 in tumor cells and IL8 in macrophages, which formed a positive feedback loop to construct the BCSC niche and maintain BCSC stemness (Fig. 8O).

By screening a clinical drug library containing FDA-approved drugs, we found that CDDO-Me treatment markedly downregulated Rab13 expression and significantly reduced BCSC stemness in vitro and in vivo. Accordingly, previous studies have reported that CDDO-Me delays breast cancer development in BRCA1-mutated mice (55) and inhibits BCSCs in TNBC (56). In addition, CDDO-Me treatment prevents STAT3 phosphorylation in multi-drug resistant ovarian cancer cells (57). Our ChIP-seq analysis showed that as a transcription factor, STAT3 bound to Rab13 promoter (Supplementary Fig. S8G), providing the possibility that CDDO-Me treatment inhibits Rab13 expression by preventing STAT3 activation. Previous studies have shown that CDDO-Me treatment converts states of breast TAMs from tumor-promotion to tumor-inhibition in vitro (58), and redirects TAM activation and T-cell tumor infiltration in vivo (59). Here, we found that tumor cells with CDDO-Me treatment did not respond to stimulation of TCM/CAF-CM or IL8. The murine Il8 gene is named as Cxcl15 in mice (Gene ID: 20309 in NCBI) and the sequence similarity between human IL8 and mouse Il8 is relatively high (84%). Previous study also observed that mouse Il8 (Cxcl15) expressed in mouse breast tumor tissue (60), demonstrating the conservation of working model in mice as well. Accordingly, we observed that CDDO-Me administration inhibited tumor growth, BCSC stemness, and enrichment of TAMs and CAF in MMTV-PyMT breast tumors. These results indicate that CDDO-Me targets Rab13 and disrupts the connection between tumor cells and stromal cells, remodels BCSC niche in vivo, which further supports the benefit of targeting Rab13 in breast cancer treatment.

For the specificity of CDDO-Me, it is true that a few reports have demonstrated its inhibitory effects on Wnt, AKT, NF-κB, Hsp90, mTOR, and JAK/STAT signaling (56, 59, 61–63). Our data indicated that Rab13 regulated JAK/STAT signaling (Fig. 6), suggesting that Rab13 may mediate some of the above inhibitory effects of CDDO-Me, which needs further investigation. Despite all this, it is possible that Rab13 may be partially responsible for CDDO-Me–mediated tumor inhibition, we cannot exclude other potential targets of CDDO-Me.

Collectively, our findings reveal a previously unsuspected role of Rab13 in maintaining BCSC stemness and controlling cell plasticity by building and safeguarding the association between tumor cells and stromal cells in BCSC niche.

B. Jiao reports grants from National Science Foundation of China Grant during the conduct of the study; in addition, B. Jiao has a patent for An inhibitor of Rab13 in cancer pending. No disclosures were reported by the other authors.

H. Wang: Conceptualization, data curation, formal analysis, funding acquisition, validation, investigation, writing–original draft, writing–review and editing. H. Xu: Software, bioinformatic analysis. W. Chen: Investigation, methodology. M. Cheng: Investigation, methodology. L. Zou: Project administration. Q. Yang: Project administration. C.B. Chan: Formal analysis. H. Zhu: Formal analysis. C. Chen: Formal analysis. J. Nie: Formal analysis. B. Jiao: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, writing–original draft, writing–review and editing.

This work was supported by the National Key Research and Development Program of China (grant no. 2020YFA0112300 to C. Chen), Strategic Priority Research Program of the Chinese Academy of Sciences (grant no. XDPB17 to B. Jiao), National Natural Science Foundation of China (grant nos. 31701306 to H. Wang, U1802285 to B. Jiao; and 31970612 to B. Jiao), and Chinese Academy of Sciences “Light of West China” Program (xbzg-zdsys-201913 to B. Jiao).The authors also thank Prof Xiyun Deng from Central South University for providing doxorubicin-resistant cells (MDA-MB-231-ADR) as well as Guolan Ma and Shuangjuan Yang from the Kunming Institute of Zoology for flow cytometry analysis. They thank Dr. Christine Watts for English editing.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Schwitalla
S
,
Fingerle
AA
,
Cammareri
P
,
Nebelsiek
T
,
Goktuna
SI
,
Ziegler
PK
, et al
.
Intestinal tumorigenesis initiated by dedifferentiation and acquisition of stem-cell-like properties
.
Cell
2013
;
152
:
25
38
.
2.
Batlle
E
,
Clevers
H
.
Cancer stem cells revisited
.
Nat Med
2017
;
23
:
1124
34
.
3.
Hill
BS
,
Sarnella
A
,
D'Avino
G
,
Zannetti
A
.
Recruitment of stromal cells into tumour microenvironment promote the metastatic spread of breast cancer
.
Semin Cancer Biol
2020
;
60
:
202
13
.
4.
Korkaya
H
,
Liu
S
,
Wicha
MS
.
Breast cancer stem cells, cytokine networks, and the tumor microenvironment
.
J Clin Invest
2011
;
121
:
3804
9
.
5.
Pfeffer
S
,
Aivazian
D
.
Targeting RAB GTPases to distinct membrane compartments
.
Nat Rev Mol Cell Bio
2004
;
5
:
886
96
.
6.
Brooks
MD
,
Wicha
MS
.
Tumor twitter: cellular communication in the breast cancer stem cell niche
.
Cancer Discov
2015
;
5
:
469
71
.
7.
Mughees
M
,
Chugh
H
,
Wajid
S
.
Vesicular trafficking-related proteins as the potential therapeutic target for breast cancer
.
Protoplasma
2020
;
257
:
345
52
.
8.
Mosesson
Y
,
Mills
GB
,
Yarden
Y
.
Derailed endocytosis: an emerging feature of cancer
.
Nat Rev Cancer
2008
;
8
:
835
50
.
9.
Le
TL
,
Yap
AS
,
Stow
JL
.
Recycling of E-cadherin: a potential mechanism for regulating cadherin dynamics
.
J Cell Biol
1999
;
146
:
219
32
.
10.
Caswell
PT
,
Chan
M
,
Lindsay
AJ
,
McCaffrey
MW
,
Boettiger
D
,
Norman
JC
.
Rab-coupling protein coordinates recycling of alpha5beta1 integrin and EGFR1 to promote cell migration in 3D microenvironments
.
J Cell Biol
2008
;
183
:
143
55
.
11.
Ioannou
MS
,
McPherson
PS
.
Regulation of cancer cell behavior by the small GTPase Rab13
.
J Biol Chem
2016
;
291
:
9929
37
.
12.
Hutagalung
AH
,
Novick
PJ
.
Role of Rab GTPases in membrane traffic and cell physiology
.
Physiol Rev
2011
;
91
:
119
49
.
13.
Wandinger-Ness
A
,
Zerial
M
.
Rab proteins and the compartmentalization of the endosomal system
.
Cold Spring Harb Perspect Biol
2014
;
6
:
a022616
.
14.
Ha
H
,
Debnath
B
,
Neamati
N
.
Role of the CXCL8-CXCR1/2 axis in cancer and inflammatory diseases
.
Theranostics
2017
;
7
:
1543
88
.
15.
Su
S
,
Chen
J
,
Yao
H
,
Liu
J
,
Yu
S
,
Lao
L
, et al
.
CD10(+)GPR77(+) cancer-associated fibroblasts promote cancer formation and chemoresistance by sustaining cancer stemness
.
Cell
2018
;
172
:
841
56
.
16.
Ginestier
C
,
Liu
S
,
Diebel
ME
,
Korkaya
H
,
Luo
M
,
Brown
M
, et al
.
CXCR1 blockade selectively targets human breast cancer stem cells in vitro and in xenografts
.
J Clin Invest
2010
;
120
:
485
97
.
17.
de Zeeuw
D
,
Akizawa
T
,
Agarwal
R
,
Audhya
P
,
Bakris
GL
,
Chin
M
, et al
.
Rationale and trial design of bardoxolone methyl evaluation in patients with chronic kidney disease and type 2 diabetes: the occurrence of renal events (BEACON)
.
Am J Nephrol
2013
;
37
:
212
22
.
18.
Merino
D
,
Weber
TS
,
Serrano
A
,
Vaillant
F
,
Liu
K
,
Pal
B
, et al
.
Barcoding reveals complex clonal behavior in patient-derived xenografts of metastatic triple negative breast cancer
.
Nat Commun
2019
;
10
:
766
.
19.
Davis
RT
,
Blake
K
,
Ma
D
,
Gabra
MBI
,
Hernandez
GA
,
Phung
AT
, et al
.
Transcriptional diversity and bioenergetic shift in human breast cancer metastasis revealed by single-cell RNA sequencing
.
Nat Cell Biol
2020
;
22
:
310
20
.
20.
Lin
EY
,
Jones
JG
,
Li
P
,
Zhu
L
,
Whitney
KD
,
Muller
WJ
, et al
.
Progression to malignancy in the polyoma middle T oncoprotein mouse breast cancer model provides a reliable model for human diseases
.
Am J Pathol
2003
;
163
:
2113
26
.
21.
Zhao
L
,
Qiu
T
,
Jiang
D
,
Xu
H
,
Zou
L
,
Yang
Q
, et al
.
SGCE promotes breast cancer stem cells by stabilizing EGFR
.
Adv Sci
2020
;
7
:
1903700
.
22.
Costa
A
,
Kieffer
Y
,
Scholer-Dahirel
A
,
Pelon
F
,
Bourachot
B
,
Cardon
M
, et al
.
Fibroblast heterogeneity and immunosuppressive environment in human breast cancer
.
Cancer Cell
2018
;
33
:
463
79
.
23.
Huang
Y
,
Shi
H
,
Zhou
H
,
Song
X
,
Yuan
S
,
Luo
Y
.
The angiogenic function of nucleolin is mediated by vascular endothelial growth factor and nonmuscle myosin
.
Blood
2006
;
107
:
3564
71
.
24.
Zhao
L
,
Ke
H
,
Xu
H
,
Wang
GD
,
Zhang
H
,
Zou
L
, et al
.
TDP-43 facilitates milk lipid secretion by post-transcriptional regulation of Btn1a1 and Xdh
.
Nat Commun
2020
;
11
:
341
.
25.
Al-Hajj
M
,
Wicha
MS
,
Benito-Hernandez
A
,
Morrison
SJ
,
Clarke
MF
.
Prospective identification of tumorigenic breast cancer cells
.
Proc Natl Acad Sci U S A
2003
;
100
:
3983
8
.
26.
Ginestier
C
,
Hur
MH
,
Charafe-Jauffret
E
,
Monville
F
,
Dutcher
J
,
Brown
M
, et al
.
ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome
.
Cell Stem Cell
2007
;
1
:
555
67
.
27.
Liu
S
,
Cong
Y
,
Wang
D
,
Sun
Y
,
Deng
L
,
Liu
Y
, et al
.
Breast cancer stem cells transition between epithelial and mesenchymal states reflective of their normal counterparts
.
Stem Cell Reports
2014
;
2
:
78
91
.
28.
Bianchini
G
,
Balko
JM
,
Mayer
IA
,
Sanders
ME
,
Gianni
L
.
Triple-negative breast cancer: challenges and opportunities of a heterogeneous disease
.
Nat Rev Clin Oncol
2016
;
13
:
674
90
.
29.
Idowu
MO
,
Kmieciak
M
,
Dumur
C
,
Burton
RS
,
Grimes
MM
,
Powers
CN
, et al
.
CD44(+)/CD24(-/low) cancer stem/progenitor cells are more abundant in triple-negative invasive breast carcinoma phenotype and are associated with poor outcome
.
Hum Pathol
2012
;
43
:
364
73
.
30.
Zhang
SH
,
Liu
DX
,
Wang
L
,
Li
YH
,
Wang
YH
,
Zhang
H
, et al
.
A CASPR1-ATP1B3 protein interaction modulates plasma membrane localization of Na(+)/K(+)-ATPase in brain microvascular endothelial cells
.
J Biol Chem
2019
;
294
:
6375
86
.
31.
Helbig
I
,
Lopez-Hernandez
T
,
Shor
O
,
Galer
P
,
Ganesan
S
,
Pendziwiat
M
, et al
.
A recurrent missense variant in AP2M1 impairs clathrin-mediated endocytosis and causes developmental and epileptic encephalopathy
.
Am J Hum Genet
2019
;
104
:
1060
72
.
32.
Rusu
V
,
Hoch
E
,
Mercader
JM
,
Tenen
DE
,
Gymrek
M
,
Hartigan
CR
, et al
.
Type 2 diabetes variants disrupt function of SLC16A11 through two distinct mechanisms
.
Cell
2017
;
170
:
199
212
.
33.
Lu
H
,
Clauser
KR
,
Tam
WL
,
Frose
J
,
Ye
X
,
Eaton
EN
, et al
.
A breast cancer stem cell niche supported by juxtacrine signalling from monocytes and macrophages
.
Nat Cell Biol
2014
;
16
:
1105
17
.
34.
Im
JH
,
Buzzelli
JN
,
Jones
K
,
Franchini
F
,
Gordon-Weeks
A
,
Markelc
B
, et al
.
FGF2 alters macrophage polarization, tumour immunity and growth and can be targeted during radiotherapy
.
Nat Commun
2020
;
11
:
4064
.
35.
Hull
RM
,
Cruz
C
,
Jack
CV
,
Houseley
J
.
Environmental change drives accelerated adaptation through stimulated copy number variation
.
PLoS Biol
2017
;
15
:
e2001333
.
36.
Etemadmoghadam
D
,
deFazio
A
,
Beroukhim
R
,
Mermel
C
,
George
J
,
Getz
G
, et al
.
Integrated genome-wide DNA copy number and expression analysis identifies distinct mechanisms of primary chemoresistance in ovarian carcinomas
.
Clin Cancer Res
2009
;
15
:
1417
27
.
37.
Sousa
S
,
Brion
R
,
Lintunen
M
,
Kronqvist
P
,
Sandholm
J
,
Monkkonen
J
, et al
.
Human breast cancer cells educate macrophages toward the M2 activation status
.
Breast Cancer Res
2015
;
17
:
101
.
38.
Muller
L
,
Tunger
A
,
Plesca
I
,
Wehner
R
,
Temme
A
,
Westphal
D
, et al
.
Bidirectional crosstalk between cancer stem cells and immune cell subsets
.
Front Immunol
2020
;
11
:
140
.
39.
Maib
H
,
Smythe
E
,
Ayscough
K
.
Forty years on: clathrin-coated pits continue to fascinate
.
Mol Biol Cell
2017
;
28
:
843
7
.
40.
Fields
IC
,
Shteyn
E
,
Pypaert
M
,
Proux-Gillardeaux
V
,
Kang
RS
,
Galli
T
, et al
.
v-SNARE cellubrevin is required for basolateral sorting of AP-1B-dependent cargo in polarized epithelial cells
.
J Cell Biol
2007
;
177
:
477
88
.
41.
Ioannou
MS
,
Bell
ES
,
Girard
M
,
Chaineau
M
,
Hamlin
JN
,
Daubaras
M
, et al
.
DENND2B activates Rab13 at the leading edge of migrating cells and promotes metastatic behavior
.
J Cell Biol
2015
;
208
:
629
48
.
42.
Zhou
J
,
Yi
L
,
Ouyang
Q
,
Xu
L
,
Cui
H
,
Xu
M
.
Neurotensin signaling regulates stem-like traits of glioblastoma stem cells through activation of IL-8/CXCR1/STAT3 pathway
.
Cell Signal
2014
;
26
:
2896
902
.
43.
Adorno-Cruz
V
,
Kibria
G
,
Liu
X
,
Doherty
M
,
Junk
DJ
,
Guan
D
, et al
.
Cancer stem cells: targeting the roots of cancer, seeds of metastasis, and sources of therapy resistance
.
Cancer Res
2015
;
75
:
924
9
.
44.
Palomeras
S
,
Ruiz-Martinez
S
,
Puig
T
.
Targeting breast cancer stem cells to overcome treatment resistance
.
Molecules
2018
;
23
:
2193
.
45.
Chen
J
,
Li
Y
,
Yu
TS
,
McKay
RM
,
Burns
DK
,
Kernie
SG
, et al
.
A restricted cell population propagates glioblastoma growth after chemotherapy
.
Nature
2012
;
488
:
522
6
.
46.
Liu
YP
,
Yang
CJ
,
Huang
MS
,
Yeh
CT
,
Wu
AT
,
Lee
YC
, et al
.
Cisplatin selects for multidrug-resistant CD133+ cells in lung adenocarcinoma by activating Notch signaling
.
Cancer Res
2013
;
73
:
406
16
.
47.
Calcagno
AM
,
Salcido
CD
,
Gillet
JP
,
Wu
CP
,
Fostel
JM
,
Mumau
MD
, et al
.
Prolonged drug selection of breast cancer cells and enrichment of cancer stem cell characteristics
.
J Natl Cancer Inst
2010
;
102
:
1637
52
.
48.
Creighton
CJ
,
Li
X
,
Landis
M
,
Dixon
JM
,
Neumeister
VM
,
Sjolund
A
, et al
.
Residual breast cancers after conventional therapy display mesenchymal as well as tumor-initiating features
.
Proc Natl Acad Sci U S A
2009
;
106
:
13820
5
.
49.
McDaniel
JM
,
Varley
KE
,
Gertz
J
,
Savic
DS
,
Roberts
BS
,
Bailey
SK
, et al
.
Genomic regulation of invasion by STAT3 in triple negative breast cancer
.
Oncotarget
2017
;
8
:
8226
38
.
50.
Cho
H
,
Kim
JH
,
Jun
CD
,
Jung
DW
,
Williams
DR
.
CAF-derived IL6 and GM-CSF cooperate to induce M2-like TAMs-response
.
Clin Cancer Res
2019
;
25
:
894
5
.
51.
Cho
H
,
Seo
Y
,
Loke
KM
,
Kim
SW
,
Oh
SM
,
Kim
JH
, et al
.
Cancer-stimulated CAFs enhance monocyte differentiation and protumoral TAM activation via IL6 and GM-CSF secretion
.
Clin Cancer Res
2018
;
24
:
5407
21
.
52.
Li
Y
,
Li
P
,
Lin
SH
,
Zheng
YQ
,
Zheng
XX
.
Paeonol inhibited TNFalpha-induced GM-CSF expression in fibroblast-like synoviocytes
.
Int J Clin Pharmacol Ther
2014
;
52
:
986
96
.
53.
Song
M
,
Bode
AM
,
Dong
Z
,
Lee
MH
.
AKT as a therapeutic target for cancer
.
Cancer Res
2019
;
79
:
1019
31
.
54.
Lenart
M
,
Rutkowska-Zapala
M
,
Baj-Krzyworzeka
M
,
Szatanek
R
,
Weglarczyk
K
,
Smallie
T
, et al
.
Hyaluronan carried by tumor-derived microvesicles induces IL-10 production in classical (CD14(++)CD16(-)) monocytes via PI3K/Akt/mTOR-dependent signalling pathway
.
Immunobiology
2017
;
222
:
1
10
.
55.
Kim
EH
,
Deng
C
,
Sporn
MB
,
Royce
DB
,
Risingsong
R
,
Williams
CR
, et al
.
CDDO-methyl ester delays breast cancer development in BRCA1-mutated mice
.
Cancer Prev Res
2012
;
5
:
89
97
.
56.
So
JY
,
Lin
JJ
,
Wahler
J
,
Liby
KT
,
Sporn
MB
,
Suh
N
.
A synthetic triterpenoid CDDO-Im inhibits tumorsphere formation by regulating stem cell signaling pathways in triple-negative breast cancer
.
PLoS One
2014
;
9
:
e107616
.
57.
Ahmad
R
,
Raina
D
,
Meyer
C
,
Kufe
D
.
Triterpenoid CDDO-methyl ester inhibits the Janus-activated kinase-1 (JAK1)→signal transducer and activator of transcription-3 (STAT3) pathway by direct inhibition of JAK1 and STAT3
.
Cancer Res
2008
;
68
:
2920
6
.
58.
Ball
MS
,
Shipman
EP
,
Kim
H
,
Liby
KT
,
Pioli
PA
.
CDDO-Me redirects activation of breast tumor associated macrophages
.
PLoS One
2016
;
11
:
e0149600
.
59.
Ball
MS
,
Bhandari
R
,
Torres
GM
,
Martyanov
V
,
ElTanbouly
MA
,
Archambault
K
, et al
.
CDDO-Me alters the tumor microenvironment in estrogen receptor negative breast cancer
.
Sci Rep
2020
;
10
:
6560
.
60.
Lin
L
,
Luo
X
,
Wang
L
,
Xu
F
,
He
Y
,
Wang
Q
, et al
.
BML-111 inhibits EMT, migration and metastasis of TAMs-stimulated triple-negative breast cancer cells via ILK pathway
.
Int Immunopharmacol
2020
;
85
:
106625
.
61.
Deeb
D
,
Gao
X
,
Dulchavsky
SA
,
Gautam
SC
.
CDDO-me induces apoptosis and inhibits Akt, mTOR and NF-kappaB signaling proteins in prostate cancer cells
.
Anticancer Res
2007
;
27
:
3035
44
.
62.
Qin
DJ
,
Tang
CX
,
Yang
L
,
Lei
H
,
Wei
W
,
Wang
YY
, et al
.
Hsp90 is a novel target molecule of CDDO-Me in inhibiting proliferation of ovarian cancer cells
.
PLoS One
2015
;
10
:
e0132337
.
63.
Liby
KT
,
Sporn
MB
.
Synthetic oleanane triterpenoids: multifunctional drugs with a broad range of applications for prevention and treatment of chronic disease
.
Pharmacol Rev
2012
;
64
:
972
1003
.

Supplementary data