miR-302a/b/c/d is downregulated in MCF-7/ADR cells overexpressing P-gp
To better understand the biological mechanisms of chemoresistance in breast cancer cells and search for the reversion opportunities, we selected adriamycin sensitive and derived resistant breast cancer cell line pair(MCF-7 and MCF-7/ADR). To identify the differential sensitivity of the parental MCF-7 and MCF-7/ ADR breast cancer cell line to chemodrugs, we first determined the cytotoxicity of chemotherapeutic drugs, including adriamycin (ADR), paclitaxel(PAC) and etoposide(VP-16) by MTS assay, all of which are currently used for the treatment of breast cancer. As shown in Fig. 1a, MCF-7/ADR cells showed resistance to ADR, PAC and VP-16 with higher IC50 values (ADR:47.2 ± 4.33 μM, PAC:112.5 ± 10.16nM, VP-16: 1.072 ± 0.099 mM) than MCF-7 cells(ADR:1.02 ± 0.09 μM, PAC:3.57 ± 0.35nM, VP-16:75.7 ± 4.65 μM). The results showed that MCF-7/ADR cells had cross-resistance ADR, PAC and VP-16. We then characterized the differential expression of MDR-related ABC transporters, including MRP, P-gp, LRP, and BCRP, between the parental MCF-7 and its derivative ADR-resistant MCF-7/ADR breast cancer cells, using Western blot. As shown in Fig. 1b, the results showed that the MCF-7/ADR cell line displayed obviously increased levels of P-gp expression compared with the parental MCF-7 cell line, while the other three detected MDR proteins only showed slight upregulation in the MCF-7/ADR cell line. These data suggest that the overexpression of P-gp is one of the reasons that MCF-7/ ADR breast cancer cells are resistant to chemodrugs. Figure 1c shows that miR-302 members (302a, 302b, 302c, and 302d) share a high sequence homology, differing only in the 3′ hexanucleotides. We further examined the miR-302 in MCF-7 and MCF-7/ADR cells. qRT-PCR results showed that miR-302a, miR-302b, miR-302c and miR-302d were significantly downregulated in MCF-7/ADR cells compared with MCF-7 cells (Fig. 1d, P < 0.05).
miR-302 overexpression sensitizes MCF-7 and MCF-7/ADR cells to ADR
To investigate the association of miR-302 expression with breast cancer chemoresistance, we determined the sensitivity of MCF-7 and MCF-7/ADR cells to ADR, PAC and VP-16 after treatment with miR-302a, miR-302b, miR-302c, miR-302d, and miR-302S mimics. The expression level of miR-302 was confirmed by qRT-PCR, showing that MCF-7 and MCF-7/ADR cells transfected with miR-302 mimics had higher miR-302 expression compared with those transfected with negative controls (Fig. 2a). miR-302 mimics alone had no significant effects on proliferation in MCF-7 and MCF-7/ADR cells (Fig. 2b). Following transfection of miR-302 mimics into MCF-7 and MCF-7/ADR, we treated the cells with a series of concentrations of ADR, PAC and VP-16. The inhibition of cell proliferation by the chemotherapeutic drugs was significantly increased by overexpression of miR-302a, miR-302b, miR-302d, and miR-302d in MCF-7 and MCF-7/ADR (Fig. 2b), suggesting that miR-302 sensitized MCF-7 and MCF-7/ADR cells to ADR, PAC and VP-16 cytotoxicity. Coexpression of miR-302a, miR-302b, miR-302c and miR-302d mimics combination (ie, four miRNA combination) had a stronger effect on reducing the cell growth compared with individual mimic (Fig. 2b).
miR-302 increased ADR accumulation in MCF-7 and MCF-7/ADR cells
We also investigated the effect of miR-302a, miR-302b, miR-302c and miR-302d on ADR accumulation in MCF-7 and MCF-7/ADR cells. The intracellular ADR accumulation was evaluated after 2 h incubation with ADR. As shown in Fig. 3a and b, overexpression of miR-302a, miR-302b, miR-302c, and miR-302d significantly increased intracellular ADR accumulation in MCF-7 and MCF-7/ADR cells compared with controls. miR-302S treatment resulted in a significantly increase in intracellular ADR accumulation in the MCF-7 and MCF-7/ ADR cells than each individual member alone.
MiR-302 inhibits P-gp expression more strongly than each individual member in MCF-7 and MCF-7/ADR cell
Overexpression of the ATP-binding cassette transporter (ABC transporter) proteins P-gp in the cell membrane is known to promote active transport of ADR out of the cells. These processes decrease intracellular drug concentrations, and lead to multidrug resistance (MDR) of breast cancer cells. Therefore, expression levels of P-gp proteins are considered a useful clinical indicator of tumor cells’ drug sensitivity and patient prognosis [20]. Since we observed that miR-302 increased intracellular ADR accumulation, we sought to determine whether P-gp proteins were involved in the miR-302 overexpression-induced effects on MCF-7 and MCF-7/ADR cells. To investigate whether P-gp expression was regulated by miR-302, we transfected miRNAs mimics into MCF-7 and MCF-7/ADR cells and then evaluated P-gp expression levels. We found that the mRNA and protein expression of P-gp was significantly down-regulated in the miR-302a, miR-302b, miR-302c and miR-302d transfected cells compared with cells transfected with control miRNA. Further, combinations of four mimics had significant effects on the mRNA (Fig. 4a) and protein (Fig. 4b) level of P-gp compared with individual miRNA .
To investigate whether or not the 3′UTR of P-gp carries binding sites for the four miRNAs, we constructed a P-gp 3′UTR luciferase reporter vector, and co-transfected the four mimics individual or combination with P-gp mRNA 3′UTR reporter into HEK-293T cells. miR-302 had no significant effects on luciferase activity compared with control miRNA (Fig. 4c). We next tested whether or not the miRNA mimics regulate degradation of P-gp mRNA. We analyzed the stability of P-gp mRNA in MCF-7/ADR cells. Cells transfected with miR-302b or miR-302c mimic, were treated with vehicle or 5 μg/mL actinomycin D (to inhibit denovo RNA synthesis), and qRT-PCR were performed with primers amplifying P-gp mRNA. As shown in Fig. 4d, the half-life of P-gp mRNA in cells was not altered when miR-302 was overexpressed in MCF-7/ADR cells. These results suggest that miR-302b, miR-302c and miR-302S does not enhance the degradation of P-gp mRNA, implying that a mechanism other than direct mRNA degradation may be involved in miR-302 mediated P-gp inhibition.
MiR-302 members directly target the 3′-UTR of MEKK1 and significantly suppress luciferase activity cooperatively
Because our results showed that suppression of P-gp expression by miR-302a, miR-302b, miR-302c and miR-302d is not through typical miRNA-mediated mRNA degradation but at the level of protein and transcription, we hypothesized that a miR-302 target gene may function as a regulator of P-gp. To elucidate the mechanisms by which the miR-302a, miR-302b, miR-302c and miR-302d regulates P-gp expression, we used mRNA target-predicting algorithms (TargetScan, miRanda, and TargetRank) based on the presence of binding sites in the 3′UTR. Of these genes that overlapped among these algorithms, we found that MAP/ERK kinase kinase 1 (MEKK1), was one of the targets of miR-302a, miR-302b, miR-302c and miR-302d. Previous studies also have shown that MEKK1 phosphorylates the MAPK c-Jun amino-terminal kinase (JNK) [21], and activation of the JNK pathway can upregulate P-gp expression [22]. Therefore, in the next study, we choose MEKK1 as our target gene to study the mechanism of downregulated P-gp by miR-302. The 3′-UTRs of MEKK1 mRNA contained one putative miR-302 binding sites (Fig. 5a). The minimum free energy between miR-302 and the putative binding sites in the 3′-UTRs of MEKK1 mRNA suggested that miR-302 may target MEKK1 via binding these putative sites (Fig. 5b). To investigate whether MEKK1 is a direct target of miR-302, we generated a firefly luciferase reporter vector containing the MEKK1 3′UTR (Fig. 5a). The vector was transfected along with miRNA mimics into HEK-293T cells. Compared with the negative control, overexpression of each individual miRNAs significantly reduced luciferase activity by approximately 19 to 25 % (Fig. 5c). At the same dose, miR-302S inhibited luciferase activity by approximately 32 %, which was significantly more than each individual member alone (P < 0.05, Fig. 5c). Further, we generated mutant MEKK1 luciferase constructs, carrying base pair changes in miRNA putative binding site located at the MEKK1 3′UTR (Fig. 5a). All four miRNA mimics were transfected along with mutated MEKK1 vectors into HEK-293T cells to test the binding ability of different miRNA combinations. Mutation of the four miRNA-binding sequences reduced the ability of inhibiting luciferase activity by miR-302 (Fig. 5d), suggesting that they directly bind to the 3′UTR of MEKK1.
We further studied the effect of miR-302 and each individual member on MEKK1 expression by overexpressing these miRNAs in MCF-7 and MCF-7/ADR cells. MEKK1 mRNA levels were confirmed by quantitative PCR analysis 48 h post-transfection. A significant decrease in MEKK1 mRNA levels was found in cells transfected with each individual miRNAs, compared with controls (miR-NC). In contrast, when the four miRNAs were transfected together, we observed a more significant decrease in MEKK1mRNA levels (P < 0.05, Fig. 6a), which is consistent with the findings of the luciferase experiments described above. Similarly, a 50 % reduction in MEKK1 protein levels was observed when the four miRNAs were transfected together, suggesting again that they act cooperatively (Fig. 6b).
RNA degradation experiments were conducted to assess if miR-302 affect MEKK1 mRNA stability. The data showed that miR-302b and miR-302c accelerated the decay of MEKK1 mRNA, compared with the control. The higher rate of mRNA decay caused by miR-302S than individual was also found (Fig. 6c). The results suggest that mRNA degradation could be an important mechanism underlying miR-302-mediated posttranscriptional regulation of MEKK1.
MiR-302 decreases MEKK1-mediated ERK MAP kinase signaling pathway in MCF-7 and MCF-7/ADR cells
We found that miR-302 negatively regulates MEKK1 by targeting the MEKK1 mRNA-3′UTR in MCF-7 and MCF-7/ADR cells. In addition, MEKK1 can phosphorylate and activate ERK and JNK MAPK pathway. We then sought to determine whether the MEKK1-mediated downstream signal pathway was also impacted by miR-302. Therefore, we examined changes in the expression of proteins association with the MAPK pathway, including p38, JNK and ERK in breast cancer cells following miR-302 overexpression. The results showed that after transfection with miR-302 mimics in MCF-7 and MCF-7/ADR cells, the expression of phosphorylation levels of ERK was decreased (Fig. 7a and b). Moreover, the four miRNAs combination effects were notably higher than those treated with individual miRNA. The total and phosphorylated protein levels of JNK and p38 were unaffected by miR-302 mimics transfection in any of the cell types (Fig. 7a and b). These results indicated that miR-302 may be an important regulator of ERK signaling pathway.
MEKK1-mediated ERK pathway activation accounts for the decreased P-gp expression induced by overexpression of miR-302
In the previous experiments, we found that induced expression of miR-302 led to an increased drug sensitivity. These changes appeared to be due to inhibition of P-gp expression. ERK signaling pathway is known to cause activation and expression of P-gp [23]. We also found that miR-302 suppressed ERK pathway in breast cancer cells. To further investigate whether inhibition of MEKK1-mediated ERK pathways was involved in the miR-302 induced P-gp suppression, MCF-7 and MCF-7/ADR cells were co-transfected miR-302 mimics with MEKK1 cDNA(lacking the 3′UTR sequence) inducing expression of MEKK1. The results showed that co-transfection of pcDNA3.1(+)-MEKK1 and miR-302 mimics increased MEKK1 expression (Fig. 8a and b). miR-302 suppressed P-gp suppression and ERK pathway, which was reversed by MEKK1 overexpression (Fig. 8a and b). These results demonstrated that P-gp suppression by miR-302 is mediated by down-regulation of MEKK1-mediated ERK pathway. Collectively, these results indicate that the synergistic overexpression of these four miRNAs sensitizes breast cancer cells to ADR more effectively than individual miRNA alone. This might be due in part to the indirectly suppressing P-gp by targeting MEKK1-mediated ERK pathway.