Transgelin 2 Promotes Paclitaxel Resistance, Migration, and Invasion of Breast Cancer by Directly Interacting with PTEN and Activating PI3K/Akt/GSK-3β Pathway Free

Breast cancer is the most common malignancy among women worldwide (1). Although treatment with a combination of chemotherapy and surgery has made remarkable progress in patients with breast cancer during the past few years, the recurrence and cancer death remain very high due to high frequency of metastasis and recurrence. Even though additional drugs were used, breast cancer cells generally gained MDR, finally causing tumor migration and invasion over time (2, 3). Therefore, exploring the molecular mechanisms of MDR and tumor metastasis is significant for patients with breast cancer to develop new effective therapeutic strategies.

Paclitaxel (PTX) is widely applied in first-line chemotherapies for treating breast cancer (4, 5). However, the efficacy of PTX is often hampered to acquired resistance in patients. Various mechanisms have been proposed to explain PTX resistance, including alterations in the level of expressed tubulin, overexpression of ATP-binding cassette (ABC) transporter proteins, and apoptotic modulation (6–8).

Several studies have found that the epithelial–mesenchymal transition (EMT) is related to acquisition of the MDR phenotype (9–11). The EMT is known to be involved in cancer metastasis and is a process in which epithelial cancer cells lose their typical epithelial characteristics and acquire mesenchymal traits. Cancer cells undergoing the EMT become migratory and invasive, enabling metastasis and chemotherapy resistance. Therefore, identifying the molecular mechanism underlying tumor progression may ultimately lead to innovative therapeutic strategies against breast cancer.

Transgelin 2, which is encoded by TAGLN2, belongs to the family of actin-binding proteins and is a 22,391-Da protein containing 199 amino acids. Deregulation of Transgelin 2 is considered to be associated with tumorigenesis and tumor development in various human malignancies (12–14). We previously used proteomic technologies to reveal the expression of Transgelin 2 to be 15.48-fold higher in PTX-resistant human breast cancer cells (MCF-7/PTX) established by our laboratory than in breast cancer drug-sensitive cells (MCF-7/S; ref. 15). Furthermore, our data demonstrated that Transgelin 2 can modulate the expression levels of MDR1, MRP1, and BCRP, suggesting that Transgelin 2 is critical for the occurrence of MDR in breast cancer cells (16). However, whether Transgelin 2 mediates the migration and invasion of breast cancer cells remains unclear.

The PI3K/Akt/GSK-3β pathway is known to participate in the regulation of various biological processes, such as the inhibition of apoptosis, induction of cell proliferation, and promotion of cancer metastasis (17). Our previous study demonstrated that Transgelin 2 activated the PI3K/Akt pathway in MCF-7/PTX cells, thereby inhibiting mitochondrial apoptosis and inducing breast cancer resistance (18). Nevertheless, the role of Transgelin 2 in the PI3K/Akt/GSK-3β pathway is still unknown and needs to be explored further.

In this study, we found that Transgelin 2 was upregulated in breast cancer tissues, and that a strong expression of Transgelin 2 indicated a poor prognosis in patients with breast cancer. Moreover, Transgelin 2 has a strong effect on the PTX resistance and the invasion and metastasis of breast cancer cells both in vitro and in vivo. The molecular mechanism assays demonstrated that Transgelin 2 could suppress the expression of PTEN and activate the PI3K/Akt/GSK-3β pathway.

Paclitaxel and MK-2206 were purchased from Sike Pharmaceutical and Selleck Chemicals, respectively. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) was from Sigma. Annexin-V FLUOS staining kit was obtained from Invitrogen. The human monoclonal anti-N-cadherin, anti-P-gp, and anti-BCRP were purchased from Abcam. The human monoclonal anti-vimentin, anti-MRP1, and anti-Transgelin 2 antibodies were purchased from GeneTex. The human anti-Bcl-2 and anti-Bcl-2–associated X (Bax) were from Epitomics. The human monoclonal anti-E-cadherin, anti-PTEN, anti-Akt, anti-phospho-Akt (p-Akt), anti-caspase-3, anti-caspase-9, anti-PARP, and anti-Snail were purchased from Cell Signaling Technology. The human polyclonal anti-β-actin antibody was obtained from Biosynthesis Biotechnology and horseradish peroxidase–conjugated secondary antibody was obtained from Cwbiotech.

Fifty-five tissue samples of formalin-fixed, paraffin-embedded breast cancer and adjacent nontumor tissue specimens from female patients who had been diagnosed with breast cancer and received primary surgical treatment between 2008 and 2013 were obtained from the First Affiliated Hospital of Xi'an Jiaotong University (Xi'an, China), whose ethics committees approved the protocol. The mRNA expression microarray data and corresponding overall survival information of 1,904 breast cancer samples were downloaded from the METABRIC (Molecular Taxonomy of Breast Cancer International Consortium) databases to analyze the association between Transgelin 2 expression and clinicopathologic features as well as the prognostic value of Transgelin 2 in breast cancer.

MCF-7/S, MDA-MB-231, and T-47D human breast cancer cell lines were purchased from the Cell Bank of Shanghai, Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, and MCF-7/PTX cell line was established as previously described (19). All cell lines were grown in DMEM supplemented with 10% FBS, 1% penicillin, and streptomycin at 37°C in a humidified atmosphere of 5% CO₂. The other culture conditions of MCF-7/PTX cells were the same as the MCF-7/S cells except for maintaining in 30 nmol/L paclitaxel (19). Cells in exponential phase growth were observed under the inverted light microscope (Olympus).

MCF-7/PTX cells were seeded in 6-well plate (5 × 10⁵ cells/well). For siRNA transfection, cells were incubated in RPMI1640 without antibiotics for 24 hours, after which 0.1 nmol/L double-stranded siRNA against TAGLN2 or nonspecific control siRNA (GenePharma) was transfected with Lipofectamine 2000 (Invitrogen). For plasmid transient transfection, 2 μg eukaryotic expression vector alone (pcDNA3.1) or pcDNA3.1-TAGLN2 (GenePharma) was transfected to MCF-7/S cells with the same method. All the processes above were conducted according to the manufacturer's instruction.

MDA-MB-231 cells (2 × 10⁶ cells) were prepared and infected at a multiplicity of infection of 10 with negative control (Lenti-EV) or Transgelin 2-overexpressing lentiviruses (Lenti-TAGLN2) for 10 hours at 37°C in the presence of 5 μg/mL of polybrene. Selecting process was performed in 3 mg/mL puromycin 96 hours after transfection. The lentivirus was purchased from GeneChem and labeled with GFP.

The expression level of Transgelin 2 and transfection efficiency were detected by qPCR and Western blot assays, supplemented by fluorescent inverted microscope (ZEISS) to observe GFP expression after lentiviral infection.

Cells (5 × 10⁵/mL) were seeded in 96-well plates with different treatments for indicated duration. A solution of 20 μL of MTT (5 mg/mL) was added into each well and incubated for 4 hours. Then 150 μL of dimethyl sulfoxide was added into each well to solubilize the formazan for 15 minutes. The absorbance was read at 490 nm on a microplate reader (BioTek). Each experiment was repeated three times. The following equation was used to calculate the reversal index (RI): RI = IC₅₀ of paclitaxel/IC₅₀ of paclitaxel plus MK-2206.

Cells with different treatments after 48 hours were harvested and washed and resuspended with cold PBS. Then staining process was conducted using the Annexin V-FITC/PI Apoptosis Detection Kit (KeyGen Biotech) or Annexin V-APC/7-AAD Apoptosis Detection Kit (KeyGen Biotech) according to manufacturer's instructions. The stained cells were analyzed using FACSCanto II flow cytometry (Becton Dickinson Company).

Total mRNA of cells was extracted using RNAfast2000 kit (Fastagen). All PCR reactions were performed using the Prime Script RT Master Mix Perfect Real Time kit (DRR036A, TaKaRa) and SYBR Premix Ex Taq II (TaKaRa) according to manufacturer's instructions. Each sample was run independently in triplicate. The primer sequences and product length are listed in Supplementary Table S1. The experiments were run on the Bio-Rad CFX96 Real-time system (Bio-Rad). β-actin was used as an internal control.

Cells with different treatments were lysed in RIPA buffer containing protease inhibitor on ice. Then equal amount of protein lysates were electrophoretically separated by 10% SDS-PAGE and transferred to polyvinylidene fluoride membranes (Millipore). After blocking with 5% nonfat dried milk for 2 hours, the membranes were incubated with primary antibodies overnight at 4°C. Another incubation with a horseradish peroxidase–conjugated secondary antibody was performed in the following day for 2 hours at 37°C, after when the protein bands were detected using the Super Signal West Pico kit (Thermo Fisher Scientific). All Western blot experiments were repeated at least 3 times.

IHC staining was applied to 3-μm thick sections of paraffin-embedded tissue specimens using the PV-9001 Detection Kit (ZSGB-BIO) according to the manufacturer's instructions. Briefly, formalin-fixed paraffin-embedded tissues were baked at 60°C for 2 hours and then deparaffinized in xylene for 10 minutes and rehydrated in a graded series of ethanol solutions. The tissues were immersed in 0.01 M citric acid buffer at 121°C for 5 minutes and then cooled and washed with 0.1 M PBS 3 times. By being treated with 3% hydrogen peroxide for 10 minutes, endogenous peroxidase was blocked. Then the tissues were incubated with primary antibodies overnight at 4°C, followed by the secondary antibody for 30 minutes at room temperature. Sections were stained with 3,3′-diaminobenzidine and counterstained using hematoxylin for 5 seconds, dehydrated in a graded series of ethanol solutions, immersed in xylene, and examined with the aid of a microscope (Axio Image M2, Zeiss).

Transgelin 2 expression was evaluated by a pathologist in a blinded manner on the basis of the staining intensity and the percentage of positive cells. Staining intensity was scored on the following 4-point scale: negative, 0; weak, 1; intermediate, 2; and strong, 3. The percentage of positive cells was classified as follows: 0%–5%, 0; 6%–25%, 1; 26%–50%, 2; 51%–75%, 3; and >75%, 4. The final result was calculated by multiplying the percentage of positive cells by the staining intensity to obtain a total score ranging from 0 to 12. Total scores of 0, 1–4, 5–8, and 9–12 were regarded as indicating negative expression and low, moderate, and high expression levels, respectively.

Mammosphere culture was done in a serum-free DMEM/F12 (Invitrogen) supplemented with 2% B27 (Invitrogen), 20-μg/L human epidermal growth factor (Invitrogen), 10-μg/L human basic fibroblast growth factor (Invitrogen), and 5-mg/L insulin. Single cells prepared from mechanical and enzymatic dissociation were plated in 6-well ultralow attachment plates at a density of 10⁴ cells/mL in culture. Single cell status was confirmed under microscope. After 14 days, the number of mammospheres (420 μm) was counted under an upright microscope. Experiments were repeated in triplicate, independently.

Cells were grown in 6-well plate until confluent. An artificial scratch wound was created, followed by being serum-starved overnight. Then cells were maintained in serum-free culture at 37°C in a humidified atmosphere of 5% CO₂. Migration photographs were captured at 0, 12, 24, and 48 hours after scratching. Experiments were repeated in triplicate independently. The percent wound closure were calculated using the following equation: percent wound closure (%) = [1 − (L_t/L₀)] × 100%.

The invasiveness of cells was evaluated by Boyden chamber method. Coated with Matrigel Matrix (BD Biosciences), the polycarbonate filters (8-μm pore size, Corning) were incubated at 37°C for 5 hours. Next, 5 × 10⁵ cells suspended in 200-μL serum-free RPMI1640 were added into the top chamber, whereas 800 μL of complete media was added to the bottom chamber. After 48 hours, cells migrated through Matrigel and adhered onto the bottom chamber were fixed in 4% paraformaldehyde for 30 minutes, stained with 0.1% crystal violet, and counted under upright microscope (five fields per chamber). Each invasion assay was repeated in three independent experiments.

BALB/C nude mice aged 4 weeks were acquired from Shanghai SLAC Laboratory Animal Center of the Chinese Academy of Sciences. All animal experiments were approved by the ethical committee for animal care of Xi'an Jiaotong University. For in vivo tumor growth, 14 female BALB/C nude mice were used to establish a subcutaneous tumor model. MDA-MB-231 cells (1 × 10⁶) infected with Lenti-TAGLN2 (or Lenti-EV as a control) were injected subcutaneously into the flank of mice (n = 10 per group). Tumor growth was monitored by estimating the tumor volume as length × width² × 0.5. For the metastasis model, MDA-MB-231 cells (1 × 10⁶) were injected into the tail vein (n = 5 per group). The mice were killed 6 weeks later, whose lung tissues were removed and fixed in formalin. The presence of lung metastases was then assessed by hematoxylin and eosin staining.

Protein lysates from cells were extracted and incubated with 20-μL protein A/G-agarose beads at 4°C; 0.5 hours later, beads were removed and the residue was incubated with 8-μL antibody against PTEN or Transgelin 2 or normal IgG and 20-μL proteinA/G beads with gentle rocking at 4°C overnight. The next day, isolated beads were washed 3 times with PBS buffer. The supernatant (30 μL) was examined by Western blot assay using relative antibodies. Each experiment was repeated at least 3 times.

The equilibrium-binding constant (KD) of Transgelin 2 and PTEN was determined by OpenSPR. Briefly, the PTEN (40 μg/mL) was covalently immobilized on COOH-sensor chips by the N-3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride and N-hydroxysuccinimide chemistry. Then, Transgelin 2 was continuously diluted into several different concentrations using the running buffer and injected into the chip from low to high concentrations. BSA was used as a negative control. In each cycle, a 250-μL sample was flowed through the chip for 5 minutes at a constant flow rate of 20 μL/minute. After detection, 0.02% SDS was added to dissociate the peptides from target protein. The kinetic parameters of the binding reactions were calculated and analyzed by TraceDrawer software (Ridgeview Instruments AB).

Statistical analysis was performed using SPSS (version 19.0) software. Data obtained in triplicate experiments performed in a parallel manner were expressed as mean ± standard deviation values. One-way ANOVA was used for comparison tests. The relationships between Transgelin 2 expression levels and clinicopathologic parameters were tested using the χ² test. The prognostic value of Transgelin 2 expression in the prognosis of breast cancers was analyzed by Kaplan–Meier curves and compared by the log-rank test using GraphPad prism software. The cut-off value for the Transgelin 2 expression level was set using X-tile (version 3.6.1) software. Probability values of P < 0.05 were considered statistically significant.

To explore the clinical significance of Transgelin 2, we applied IHC to the 54 collected pairs of breast cancer tissues and adjacent nontumor tissues. The results showed that Transgelin 2 was mainly expressed in breast cancer tissues in the cell cytoplasm, and the expression level was significantly higher than that in the adjacent nontumor tissues (Fig. 1A and B). We then obtained 1,904 breast cancer cases with detailed clinical information and RNA-seq results from the METABRIC databases. We analyzed the associations between Transgelin 2 mRNA expression and clinicopathologic features in the patients with breast cancer (Supplementary Table S2). As Supplementary Figure S1 shows, the Transgelin 2 mRNA expression in breast cancer was not associated with age or menopausal status; however, a high Transgelin 2 mRNA expression was significantly associated with larger tumors, poor histologic grade, high tumor stage, negative estrogen receptor (ER) expression, progesterone receptor (PR), positive Her-2 expression, and invasive ductal tumors. A subsequent multivariate logistic regression analysis found that the cancer type, ER status, and histologic grade were independently associated with Transgelin 2 mRNA expression in breast cancers (Supplementary Table S3). Furthermore, Kaplan–Meier survival analysis was used to reveal the prognostic value of Transgelin 2. Breast cancer patients with a high Transgelin 2 mRNA expression had poorer prognosis than those with low Transgelin 2 expression in breast cancers (184.8 vs. 159.0 months, P = 0.0111; Fig. 1C).

These results indicate that Transgelin 2 can be considered a diagnostic marker for breast cancer.

To investigate the biological functions of overexpressed Transgelin 2 in breast cancer, we detected the expression levels of Transgelin 2 in three breast cancer cell lines. We found that the expression level was low in MCF-7 and MDA-MB-231 cells (Supplementary Fig. S2A). We then overexpressed Transgelin 2 in these two cell lines using plasmid transient transfection and lentivirus stable transfection, respectively. As shown in Supplementary Figure S2B, mRNA and protein expression levels of Transgelin 2 were both substantially increased compared with the control group in both two cell lines.

To identify the effect of Transgelin 2 in paclitaxel resistance, we measured the sensitivity to paclitaxel in breast cancer cells. Surprisingly, Transgelin 2 overexpression decreased the chemosensitivity of breast cancer cells to paclitaxel (Fig. 1D). Consistently, flow cytometry analysis indicated that after treatment with 20 nmol/L paclitaxel, the percentage of apoptotic cells in Transgelin 2-overexpressing cells (23.82 %) was less than that in control cells (41.35 %; Fig. 1E). In addition, qPCR and Western blotting revealed that the overexpression of Transgelin 2 in MCF-7 and MDA-MB-231 cells significantly increased the mRNA and protein levels of P-gp, MRP1, and BCRP (Fig. 1F and G). Together these findings strongly indicate that strong expression of Transgelin 2 contributed to PTX resistance, which leads us to propose that targeting Transgelin 2 could be a useful strategy for improving the sensitivity of breast cancer cells to PTX.

Mammosphere formation, qPCR, Western blotting, scratch wound healing, and Transwell invasion assays were performed to clarify whether overexpression of Transgelin 2 could alter the migration and invasion abilities of MCF-7/S cells. Transgelin 2 overexpression led to a distinct increase in the mammosphere-forming ability of breast cancer cells (Fig. 2A). In accordance with these observations, the expression of epithelial marker E-cadherin was dramatically decreased, whereas the levels of mesenchymal markers N-cadherin and vimentin were increased (Fig. 2B and C). In addition, cells that overexpressed Transgelin 2 exhibited significantly enhanced migration and invasion characteristics (Fig. 2D and E). Our results imply that upregulation of Transgelin 2 induces the EMT process, leading to enhancement of the migratory and invasive ability of breast cancer cells.

We next explored the effect of Transgelin 2 in vivo. To investigate the effect of Transgelin 2 on cancer progression, MDA-MB-231 cells infected with Lenti-TAGLN2 (or Lenti-EV as a control) were subcutaneously injected into nude mice. The tumor volumes were significantly higher in the Lenti-TAGLN2 group than in the control group (Fig. 3A). Moreover, Kaplan–Meier survival analysis indicated that survival was worse in the Transgelin 2-overexpressing group than in the control group (Fig. 3B). A metastasis model was established to further validate the role of Transgelin 2 in tumor metastasis, with the results showing that overexpression of Transgelin 2 increased lung metastases of MDA-MB-231 cells (Fig. 3C and D).

We previously revealed that knockdown of Transgelin 2 reversed the resistance to PTX in MCF-7/PTX cells by inactivating the PI3K/Akt pathway and then promoting activation of the mitochondrial apoptotic pathway (18). Akt activation plays a pivotal role in inducing the EMT by inhibiting GSK-3β, leading to the stabilization and nuclear localization of Snail, thereby triggering tumor migration (20). To confirm whether the PI3K/Akt/GSK-3β pathway can regulate the EMT and the migration of breast cancer cells, we used the Western blot assay to measure the expression of several proteins involved in the PI3K/Akt/GSK-3β pathway. Remarkably, downregulated PTEN and upregulated phospho-Akt, phospho-GSK-3β (inactive form), and Snail were observed in MCF-7/PTX cells, along with the elevation of Transgelin 2, demonstrating that the PI3K/Akt/GSK-3β pathway was activated in MCF-7/PTX cells. However, no significant change was found in the expression levels of Akt and GSK-3β (Fig. 4A).

In view of the above findings, we hypothesized that the PI3K/Akt/GSK-3β pathway is involved in the Transgelin 2-induced migration and invasion of breast cancer cells. To verify this hypothesis, Transgelin 2 was knocked down by siRNA in MCF-7/PTX cells. As expected, compared with the control, TAGLN2 knockdown resulted in distinct decreases in the expression levels of Transgelin 2, phospho-Akt, phospho-GSK-3β, and Snail, whereas that of tumor suppressor PTEN protein was significantly increased (Fig. 4B). In contrast, overexpression of Transgelin 2 in MCF-7/S cells decreased the PTEN protein level and upregulated the expression levels of phospho-Akt and phospho-GSK-3β (Fig. 4C). Meanwhile, activation of the Akt signaling pathway subsequently elevated the expression of prosurvival factor Bcl-2 but suppressed proapoptotic factor Bax and the cleavage of caspase-3, caspase-9, and PARP. In contrast, the downstream molecule Snail was strongly upregulated (Fig. 4D).

In brief, these observations provide strong evidence that the PI3K/Akt/GSK-3β pathway was activated in MCF-7/PTX cells, which exhibit a high expression level of Transgelin 2, and that the level of tumor suppressor PTEN protein was regulated by Transgelin 2.

PTEN is another very important tumor suppressor factor following p53, which functions in many cellular processes (21). Our results indicate that Transgelin 2 can negatively regulate PTEN expression, suggesting relevance between PTEN and Transgelin 2. A coimmunoprecipitation assay was carried out to confirm the interaction between PTEN and Transgelin 2 protein in cells. Anti-PTEN and anti-Transgelin 2 antibodies pulled down Transgelin 2 protein and PTEN protein, respectively (Fig. 4E). However, the regulatory relationship between Transgelin 2 and PTEN remains unclear. We therefore simultaneously upregulated or knocked out the expression of Transgelin 2 and PTEN in MCF-7/S or MCF-7/PTX cells, respectively, and then detected the expression of the PI3K/Akt signaling pathway with the aim of defining the relative positions of Transgelin 2 and PTEN. We found that when Transgelin 2 was overexpressed in MCF-7/S cells, the expression of PTEN was inhibited and the PI3K/Akt signaling pathway was activated; however, when Transgelin 2 and PTEN were both upregulated, the PI3K/Akt pathway was comparatively inhibited (Fig. 4F). Similarly, knocking out Transgelin 2 led to the restoration of PTEN and the inhibition of the PI3K/Akt pathway, whereas the simultaneous knockout of Transgelin 2 and PTEN reactivated the PI3K/Akt pathway (Fig. 4G). These results indicate that Transgelin 2 is upstream of PTEN. Furthermore, to understand the regulation of Transgelin 2 on PTEN, we conducted ChIP-seq to detect the transcriptional regulatory sites of Transgelin 2 on PTEN. The result showed that Transgelin 2 might be a transcriptional regulatory factor of PTEN; furthermore, it could regulate PTEN expression through direct binding of PTEN (Supplementary Table S4 and Supplementary Table S5). Then the affinity of Transgelin 2 and PTEN was studied at the protein structure level. Their affinity was detected by localized surface plasmon resonances (LSPR) and the result showed a strong interaction between Transgelin 2 and PTEN with a KD of 2,610 nmol/L (Fig. 4H). These results showed that Transgelin 2 could directly interact with PTEN at the level of transcription and protein.

MK-2206 is a potent AKT inhibitor that has been proved to have good tolerance at the bioactive dose of inhibiting AKT signal (22). To confirm the role of the PI3K/Akt/GSK-3β pathway, MK-2206 was used to investigate the correlation between activation of Akt and increased PTX resistance and cell motility induced by Transgelin 2 in breast cancer cells. First of all, the cytotoxicity and effect of resistance reversal of MK-2206 toward cells were evaluated using the MTT assay. As shown in Figure 5A, MK-2206 inhibited the growth of both MCF-7/S and MCF-7/PTX cells in a dose-dependent manner. The IC₅₀ values were 3.11 ± 3.43 and 3.81 ± 5.18 μmol/L (Fig. 5A) respectively, indicating that MCF-7/PTX cells did not produce resistance to MK-2206. Consequently, MK-2206 was used at three nontoxic concentrations (20, 40, and 80 nmol/L, which produced inhibitions <10%) in the subsequent experiments. The growth curves showed that MK-2206 at these three concentrations augmented the sensitivity of MCF-7/PTX cells to PTX by 2.02-, 3.14-, and 4.67-fold, respectively (Fig. 5B; Supplementary Table S6). We also found that the apoptosis rate (69.67%) was effectively increased in MCF-7/PTX cells following treatment with PTX (1,000 nmol/L) combined with MK-2206 (80 nmol/L; Fig. 5C). Moreover, the mRNA and protein levels of P-gp, MRP1, and BCRP were consistently reduced (Fig. 5D and E). Our data supported that MK-2206 has a strong ability to reverse PTX resistance in MCF-7/PTX cells.

qPCR, Western blotting, scratch wound healing, and Transwell invasion assays were performed to elucidate if MK-2206 can inhibit the migration and invasion of breast cancer. The number of mammospheres of these cells decreased by about 2-fold compared with the control group (Fig. 6A). Meanwhile, E-cadherin expression increased sharply, whereas N-cadherin and vimentin were both clearly reduced (Fig. 6B and C). Similarly, the migration and invasion abilities of MCF-7/PTX cells treated with MK-2206 alone for 48 hours were both reduced (Fig. 6D and E), indicating that MK-2206 can attenuate the Transgelin 2-mediated EMT and the motility and invasiveness of MCF-7/PTX cells.

Western blotting was used to detect the effects of MK-2206 on the PI3K/Akt/GSK-3β pathway and Transgelin 2. After treating MCF-7/PTX cells with MK-2206 for 48 hours, the levels of Transgelin 2, phospho-Akt, and phospho-GSK-3β were downregulated with a concomitant increased expression of PTEN, whereas the levels of Akt and GSK-3β were unaffected (Fig. 6F). Meanwhile, the expression levels of its downstream factor Bcl-2 and Snail were reduced. A particularly interesting finding was that the expression levels of Bax, cleaved-caspase-3, cleaved-caspase-9, and cleaved-PARP were increased notably after MK-2206 treatment (Fig. 6G).

Transgelin 2, located on chromosome 1q21-q25, is an important cytoskeletal actin–binding protein that influences the dynamics of the actin cytoskeleton. Some recent studies have described Transgelin 2 as a potential oncogenic factor in many human malignancies (14). Studies have also shown that overexpression of Transgelin 2 was associated with tumor migration and invasion (23). Moreover, Transgelin 2 was reported to be correlated with lymph node metastasis, distant metastasis, and the TNM classification in colorectal cancer, and it may be useful as a new biomarker for predicting the progression and prognosis of colorectal cancer (24). On the other hand, there is accumulating evidence that a high level of Transgelin 2 in cancer cells could contribute to drug resistance (25).

This study is the first to find a high expression level of Transgelin 2 in MCF-7/PTX cells and its association with cell PTX resistance, migration, and invasion. In this study, we were planning to collect clinical samples from patients who had undergone PTX chemotherapy treatment prior to breast surgery to explore the effect of chemotherapy on Transgelin 2 expression. However, the difficulty of collecting such specimens meant that we collected only three specimens in which the Transgelin 2 expression level was significantly higher than that in the adjacent nontumor tissue. Besides, considering hormone levels, we found that the expression of Transgelin 2 was inversely correlated with ER and PR status. We also found that ER status was independently associated with Transgelin 2 mRNA expression in breast cancers. What is more, we had detected the expression of ER in drug-resistant cell lines and found that ER expression was low in drug resistant cells while the expression of human epidermal growth factor receptor 2 (Her-2) was upregulated (19). These suggest that Transgelin 2 might be a possible candidate as therapeutic target for ER-negative breast cancer. However, whether hormone status has effect on the expression and function of Transgelin 2 and the therapeutic effect of Transgelin 2 in ER-negative breast cancer remain unclear. Therefore, ER status should be taken into account in the follow-up study of Transgelin 2.

It is widely believed that the PI3K/Akt/GSK-3β pathway plays an important role in physiologic processes in cancer cells, such as the cell cycle, migration, differentiation, and apoptosis (26). Multiple investigations have discovered hyperactivity of Akt/GSK-3β/Snail signaling to be necessary for cell proliferation, the EMT process, and also migration and invasion (27–32). PTEN, a negative regulator of the PI3K/Akt pathway, can dephosphorylate phosphatidylinositol (3,4,5)-triphosphate to phosphatidylinositol (4,5)-bisphosphate, and it therefore decreases the activation of the PI3K/Akt pathway (33). In this study, we observed that the PI3K/Akt/GSK-3β pathway was activated when the negative regulator PTEN was downregulated in MCF-7/PTX cells. Overexpression of Transgelin 2 in MCF-7/S cells via plasmid transfection immensely greatly increased ABC transporters and decreased PTEN expression, resulting in the activation of the PI3K/Akt/GSK-3β pathway and upregulation of Snail expression, thereby facilitating cell proliferation, migration, and invasion. However, the opposite results were found when TAGLN2 was knocked down in MCF-7/PTX cells, indicating that the mechanism underlying Transgelin 2-mediated PTX resistance, migration, and invasion in breast cancer involves the PI3K/Akt/GSK-3β pathway. It is particularly interesting that we also found that Transgelin 2 directly interacted with PTEN and was located upstream of PTEN, thus activating the PI3K/Akt pathway. We also detected the effect of Transgelin 2 on PTEN stability, but the results showed that the overexpression of Transgelin 2 did not have significant effect on the half-life or the proteasome-mediated degradation of PTEN (Supplementary Fig. S3). We therefore speculate that Transgelin 2 might inhibit the expression of PTEN at the translation level via the non–proteasome–ubiquitin pathway; this mechanism requires further investigation. To confirm the role of Transgelin 2 in the PI3K/Akt/GSK-3β pathway, we treated MCF-7/PTX cells with the Akt inhibitor MK-2206 and found that inactivation of the PI3K/Akt/GSK-3β pathway exerted a feedback effect on Transgelin 2 expression, which was essential to reverse the PTX resistance and suppress migration and invasion (Supplementary Fig. S4).

In summary, this study found that Transgelin 2 plays an important role in the PTX resistance, migration, and invasion of breast cancer cells. A high level of Transgelin 2 could activate the PI3K/Akt/GSK-3β pathway by interacting with PTEN and thus participate in the PTX resistance, migration, and invasion of breast cancer. These results provide insights into the mechanism of drug resistance mediated by Transgelin 2 and indicate that Transgelin 2 is a potential therapeutic target in breast cancer.

No potential conflicts of interest were disclosed.

Conception and design: L. Liu, J. Xing, Y. Dong

Development of methodology: L. Liu, T. Meng, X. Zheng, Y. Liu, Y. Yan

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T. Meng, R. Hao

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L. Liu, X. Zheng, R. Hao, S. Chen

Writing, review, and/or revision of the manuscript: L. Liu, T. Meng, S. Chen

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): T. Meng

Study supervision: H. You, J. Xing, Y. Dong

We thank Hongying Wang (Xi'an Jiaotong University, Shaanxi, China) for her support in the establishment of xenograft tumor model. The ChIP-Seq high throughput sequencing and subsequent bioinformatics analysis were all done by Cloud-Seq Biotech (Shanghai, China).

The work is supported by National Natural Science Foundation of China (no. 81473177, 81672954).

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.