Thymosin beta 10 is a key regulator of tumorigenesis and metastasis and a novel serum marker in breast cancer

HumanTMSB10 cDNA was purchased form (Vigene Biosciences, Shandong, China) and cloned into the pSin-EF2 plasmid (addgene #16578, Cambridge, MA, USA). Knockdown of endogenous TMSB10 was performed by cloning two short hairpin RNA (shRNA) oligonucleotides into the pSUPER-puro-retro vector (OligoEngine, Seattle, WA, USA). The sequences of the two separate shRNA fragments are: RNAi#1: CCGGCCCAGTCGTGATGTGGAGGAACTCGAGTTCCTCCACATCACGACTGGGTTTTTG and RNAi#2: CCGGGCCGACCAAAGAGACCATTGACTCGAGTCAATGGTCTCTTTGGTCGGCTTTTTG (synthesized by Invitrogen). Retroviral production and infection were performed according to Weinberg et al. [18]. The reporter plasmid for the quantitative detection of forkhead box O (FOXO) transcriptional activity was generated using the pGL3-Enhancer plasmid (Promega, Madison, WI, USA) [19]. Cells were treated with Perifosine (KRX-0401) (Cell Signaling, Beverly, MA, USA) at the indicated final concentrations (30 μM).

The total 253 paraffin-embedded, archived invasive breast cancer samples used in this study were histopathologically and clinically diagnosed at the Sun Yat-sen University Cancer Center (Guangzhou, China) between 1995 and 2009. A total of 100 invasive samples of fresh breast cancer tissue (with 9 paired samples from adjacent normal tissue) and serum specimens were obtained from age-matched women during surgery at the Sun Yat-sen University Cancer Center between January 2015 and December 2015. Clinicopathological classification and staging were determined according to the criteria of the American Joint Committee on Cancer (AJCC) [20]. Patient consent and approval from the Sun Yat-Sen University Cancer Center Institutional Review Board (IRB) were obtained prior to the use of these clinical materials for research purposes (approval number GZR2013-024). The clinicopathological features of the patients are summarized in Additional files 1 and 2: Table S1 and Table S2.

All statistical analyses were performed using SPSS version 19.0 (SPSS Inc., Chicago, IL, USA) [18]. Associations between TMSB10 expression and clinicopathological characteristics of the patients were analyzed using the chi-squared test. Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. Survival data were evaluated using univariate and multivariate Cox regression analyses [21]. A two-tailed P value below 0.05 was considered statistically significant in all experiments.

We first analyzed expression levels of primary thymosin-associated proteins in the RNA sequencing data from E-GEOD-58135 and The Cancer Genome Atlas (TCGA) datasets, including TMSB4, TMSB10, TMSB15, prothymosin, alpha (PTMA) and parathymosin (PTMS), which have been reported to be implicated in the development and progression of different types of cancer, including breast cancer [11‐13, 17, 22, 23] and found that expression levels of TMSB10, PTMS and PTMA were upregulated to varying degrees in breast cancer tissues compared with normal tissues, particularly TMSB10 at the highest level (2.03-fold and 1.45-fold change, respectively) (Fig. 1a and Additional file 3: Figure S1a).

Subsequent analysis of TMSB10 expression in breast cancer datasets revealed that TMSB10 was upregulated in breast cancer tissues and breast cancer cells compared with normal breast tissue samples and epithelium cells (Fig. 1b, c and Additional file 3: Figure S1b). To further confirm the results from public datasets, we examined the expression of TMSB10 in two normal breast epithelial cell lines and eight breast cancer cell lines by real-time PCR and western blotting, and found that mRNA and protein levels of TMSB10 were upregulated in breast cancer cells compared with NMEC1 and NMEC2 (Fig. 1d and e). Moreover, TMSB10 expression was markedly increased in nine paired breast cancer tissue samples compared with the matched adjacent normal tissues (Fig. 1f and g).

We further analyzed the mRNA expression of TMSB10 in the breast cancer tissue samples of TCGA different molecular subtypes, which contribute to the significant heterogeneity of breast cancer and found that compared with normal tissues, TMSB10 was differentially upregulated in all subtypes except the luminal A subtype, and was strikingly higher in the basal-like and human epidermal growth factor receptor 2 (Her2)-enriched subtypes (Additional file 3: Figure S1c and d). Taken together, these results indicate that TMSB10 may be involved in human breast cancer progression.

We further investigated the specific mechanism underlying the overexpression of TMSB10 in breast cancer. Through analyzing the methylation array dataset of breast cancer from TCGA, we found that in the methylation levels of TMSB10 there was no obvious discrepancy between tumor and the matched adjacent normal tissue (Additional file 3: Figure S1e). On analysis of breast cancer datasets from TCGA and Tumorscape, recurrent gains (amplification) were found in approximately 15% of patients with breast cancer (Additional file 3: Figure S1f and g). The average expression level of TMSB10 in patients with breast cancer with gains was dramatically higher than those without gains (Fig. 1h) (Additional files 4, 5 and 6). We further examined the TMSB10 in our nine paired breast cancer tissue samples using RT-PCR and identified amplification in two of nine breast cancer tissue samples (Fig. 1i). The mRNA expression level of TMSB10 in patients with breast cancer with gains was significantly elevated compared those without gains (Fig. 1j). These results indicate that recurrent gains are involved in the TMSB10 overexpression in breast cancer.

We further examined TMSB10 expression by immunohistochemical analysis of 253 human breast cancer tissue samples (Additional file 1: Table S1) and found TMSB10 expression was primarily detected within the cytoplasm and the expression levels of TMSB10 positively correlated with clinical stages (Fig. 2a and b). High expression of TMSB10 was observed in 154/253 breast cancer tissue samples (60.9%) (Additional file 7: Figure S2a). In addition, there were strong positive associations between TMSB10 expression and clinical stage, tumor (T) classification, node (N) classification, metastasis (M) classification, pathological grade, estrogen receptor (ER) status and progesterone receptor (PR) status (Additional file 8: Table S3). Kaplan-Meier survival analysis revealed that patients with high TMSB10 expression had shorter overall survival (P < 0.001; hazard ratio = 3.50, 95% CI = 2.38 to 5.17; Fig. 2c) and distant metastasis-free survival (P < 0.001; hazard ratio = 3.28, 95% CI = 2.28 to 4.73; Fig. 2d), which were consistent with TCGA and Kaplan-Meier Plotter (BC) profiles (Additional file 7: Figure S2b-f). Univariate Cox regression analysis indicated that patients with high TMSB10 expression had shorter overall survival (P < 0.001; hazard ratio = 4.62, 95% CI = 2.75 to 7.79) and metastasis-free survival (P < 0.001; hazard ratio = 3.89, 95% CI = 2.48 to 6.11) compared to patients with low TMSB10 expression (Additional file 9 and 10: Table S4 and S5).

Multivariate Cox regression analysis suggested that TMSB10 may be an independent factor for predicting poor survival (P = 0.010; hazard ratio = 2.10, 95% CI = 1.20 to 3.67; Fig. 2e) and distant metastasis status (P = 0.003; hazard ratio = 2.07, 95% CI = 1.27 to 3.35; Fig. 2f). Furthermore, the analysis of breast cancer datasets from Kaplan-Meier Plotter revealed that patients with high TMSB10 expression exhibited longer disease-free survival compared to patients with low TMSB10 expression after chemotherapy, but not seen in patients treated with endocrine therapy or without systemic treatment (Additional file 7: Figure S2g-i), indicating that TMSB10 may be used as a potential indicator of chemotherapeutic response to stratify patients for the treatment of breast cancer. Collectively, these results demonstrate that the overexpression of TMSB10 correlates with poor prognosis and distant metastatic status in patients with breast cancer.

To determine the biological roles of TMSB10 in breast cancer, we constructed TMSB10 stably expressing BT-549 and SK-BR-3 breast cancer cell lines by ectopically overexpressing TMSB10 and endogenously knocking down TMSB10 by retrovirus infection. Real-time PCR and western blot were performed to measure the mRNA and protein levels of TMSB10 expression (Additional file 11: Figure S3a and b). MTT assays revealed that ectopic expression of TMSB10 significantly increased, while silencing of TMSB10 reduced, the cell numbers in breast cancer cells (Fig. 3a). Colony formation assay indicated that the upregulation of TMSB10 enhanced the colony-forming abilities of breast cells. Conversely, downregulation of TMSB10 decreased the colony-forming ability (Fig. 3b).

To investigate the effects of TMSB10 on the tumorigenic activity of breast cancer cells, we performed an anchorage-independent growth assay and found overexpression of TMSB10 enhanced, while silencing of TMSB10 reduced, the anchorage-independent growth ability of breast cancer cells (Fig. 3c). We next evaluated the effect of TMSB10 on tumorigenesis in vivo. As shown in Fig. 3d-f, tumor volumes and weight were increased significantly in the TMSB10-overexpression group compared with the control group. In contrast, the tumors formed by the TMSB10-silenced cells were smaller, with lower tumor volume and weight compared to the control group. On IHC the TMSB10 overexpressing tumor tissues had higher Ki67 proliferation indexes, whereas the TMSB10-silenced tumor tissues had reduced numbers of Ki67-positive cells (Additional file 11: Figure S3c), which is consistent with the results of analysis of clinical correlation between TMSB10 and Ki67 (Additional file 8: Table S3). Taken together, these results suggest that TMSB10 promotes the tumorigenicity of breast cancer cells by modulating proliferation in breast cancer.

We next explored the specific mechanism underlying the pro-proliferative role of TMSB10 in breast cancer cells. The bromodeoxyuridine (BrdU) incorporation assay demonstrated that the percentages of cells with incorporated BrdU were increased in TMSB10-overexpressing cells, but reduced in TMSB10-silenced cells (Additional file 11: Figure S3d). Moreover, real-time PCR and western blotting analysis revealed that cyclinD1, cyclinE1 and p-Rb were upregulated in TMSB10-overexpressing cells but decreased in TMSB10-silenced cells. Conversely, the cell cycle inhibitors p21Cip1 and p27Kip1 were decreased in TMSB10-overexpressing cells but increased in TMSB10-silenced cells (Additional file 11: Figure S3e and S3f). Collectively, these results indicate that TMSB10 promotes proliferation and tumorigenesis of breast cancer cells by accelerating the cell cycle.

On gene set enrichment analysis (GSEA) based on data on mRNA expression of TMSB10 from the TCGA, high expression of TMSB10 significantly correlated with metastasis-associated gene signatures (Additional file 12: Figure S4a). To investigate the effects of TMSB10 on the metastatic ability of breast cancer cells, we performed cell wound healing and migration assay and found overexpression of TMSB10 enhanced, while silencing of TMSB10 reduced, the migration ability of breast cancer cells (Fig. 4a and b). Cell invasion and 3D spheroid invasion assay demonstrated that upregulating TMSB10 increased the invasive ability of breast cancer cells with typical highly aggressive and invasive cell morphology, which presented a greater number of outward projections (Fig. 4c and Additional file 12: Figure S4b). However, silencing TMSB10 displayed the opposite effect on the invasive ability of breast cancer cells. We further examined the effects of TMSB10 on breast cancer metastasis in lung colonization models. TMSB10-overexpressing and vector-transduced BT-549 cells were injected into the BALB/c nude mice via the lateral tail veins respectively and metastatic activity was examined by bioluminescence imaging (BLI) of luciferase-transduced cells and H&E staining of lung metastasis sections. We observed that upregulating TMSB10 significantly enhanced the lung metastasis burden of BT-549 and shortened the survival of mice (Fig. 4d-g). In contrast, silencing TMSB10 attenuated the development of lung metastasis and increased the survival of mice (Fig. 4d-g). Collectively, these results indicate that TMSB10 promotes the metastasis capability of breast cancer cells.

GSEA was further performed to identify the pathways involved in TMSB10-mediated breast cancer progression. TMSB10 expression level positively correlated with the AKT-activated gene signatures but negatively correlated with the FOXO(s)-activated gene signatures (Additional file 13: Figure S5a), suggesting that the AKT/FOXO pathway may mediate the pro-tumor effects of TMSB10. On analysis of the TCGA breast cancer datasets, TMSB10 was positively associated with the T308 and S473 phosphorylation levels of AKT, but not with AKT expression level (Additional file 13: Figure S5b). Western blot revealed that upregulating TMSB10 increased, while silencing TMSB10 decreased the phosphorylation levels of AKT, GSK3β, FOXO4 and FOXO3a (Fig. 5a). The analysis of AKT activity revealed that AKT activity was indeed increased in the TMSB10-overexpressing cells and decreased in the TMSB10-silenced cells (Fig. 5b). However, TMSB10 had the opposite effect on the FOXO transcriptional activity (Fig. 5c). These results suggest that the activity of AKT signaling is modulated by TMSB10.

To further determine whether AKT signaling is involved in the pro-tumor role of TMSB10 in breast cancer, we applied the AKT kinase inhibitor perifosine to TMSB10-overexpressing breast cancer cells. As shown in Fig. 5d and e, the colony formation and anchorage-independent growth assays revealed that inhibition of AKT activity decreased the colony and anchorage-independent growth ability of breast cancer cells. Meanwhile, inhibition of AKT kinase activity abrogated the stimulatory effect of TMSB10 on the migration and invasion abilities of breast cancer cells (Fig. 5f and g). Taken together, these findings suggest that TMSB10 promotes tumor progression and metastasis in breast cancer by activating the AKT/FOXO pathway.

Mounting evidence indicates that several members of thymosin, including TMSB10, are secreted proteins both in the physiological and pathological condition [24‐28]. Therefore, we examined the protein levels in the supernatants of breast cancer cells using ELISA, and found TMSB10 was highly expressed in breast cancer cells compared with NMECs (Fig. 6a). The previous studies indicated that TMSB4 can be identified in the human serum [29, 30]. So, we further measured the expression levels of TMSB10 in the serum of nine patients with breast cancer and detected TMSB10 at the approximate concentration range of 100–400 ng/ml (Fig. 6b). Moreover, mRNA and protein expression levels of TMSB10 in the separate cell and tissue were consistent with the concentration of TMSB10 in their respective supernatants (Fig. 6c-f).

We further examined the TMSB10 expression in the serum of a large set of patients with breast cancer and found that expression of TMSB10 in the serum was dramatically increased compared with healthy serum (Fig. 6g). Receiver operating characteristic (ROC) analysis of patients with breast cancer gave an area under the curve (AUC) of 0.80 (95% CI = 0.73 to 0.87), Fig. 6h). It is of great interest to find that TMSB10 expression in the serum of patients with breast cancer positively correlated with the clinical stage of breast cancer (Fig. 6i) and on ROC analysis of patients with breast cancer of different stages, the AUC was 0.71 (95% CI = 0.60 to 0.82), Fig. 6j). Therefore, our results indicate that TMSB10 may be identified as a valuable serum biomarker for the diagnosis and clinical staging of breast cancer. Together, these data indicate that TMSB10 plays an important role in tumorigenesis and metastasis in patients with breast cancer (Fig. 6k).