Downregulated miR-495 [Corrected] Inhibits the G1-S Phase Transition by Targeting Bmi-1 in Breast Cancer.

Bmi-1 (B cell-specific Moloney murine leukemia virus integration site 1) is upregulated in breast cancer and was involved in many malignant progressions of breast cells, including cell proliferation, stem cell pluripotency, and cancer initiation. However, the epigenetic regulatory mechanism of Bmi-1 in breast cancer remains unclear. After analysis of the ArrayExpress dataset GSE45666, we comparatively detected the expression levels of miR-495 in 9 examined breast cancer cell lines, normal breast epithelial cells and 8 pairs of fresh clinical tumor samples. Furthermore, to evaluate the effect of miR-495 on the progression of breast cancer, MCF-7 and MDA-MB-231 were transduced to stably overexpress miR-495. The 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide assay, colony formation assays, 5-Bromo-2-deoxyUridine labeling and immunofluorescence, anchorage-independent growth ability assay, flow cytometry analysis, and luciferase assays were used to test the effect of miR-495 in MCF-7 and MDA-MB-231 cells in vitro. Xenografted tumor model was also used to evaluate the effect of miR-495 in breast cancer. Herein, we found that miR-495, a predicted regulator of Bmi-1, was frequently downregulated in malignant cells and tissues of breast. Upregulation of miR-495 significantly suppressed breast cancer cell proliferation and tumorigenicity via G1-S arrest. Further analysis revealed that miR-495 targeted Bmi-1 through its 3' untranslated region. Moreover, Bmi-1 could neutralize the suppressive effect of miR-495 on cell proliferation and tumorigenicity of breast cancer in vivo. These data suggested that miR-495 could inhibit the G1-S phase transition that leads to proliferation and tumorigenicity inhibition by targeting and suppressing Bmi-1 in breast cancer.

INTRODUCTION

Bmi-1, a polycomb group protein, was originally identified as a transgene that could co-operate with c-Myc (v-myc avian myelocytomatosis viral oncogene homolog) to induce hematopoietic malignancies,[1] and acts as master switch for cell proliferation,[2] early embryogenesis,[3] stem cell pluripotency,[4] and cancer initiation.[5] Bmi-1 is necessary for the progression of many types of cancer, and increased Bmi-1 expression may contribute to tumorigenesis, such as in hepatocellular carcinoma,[6] oral squamous cell carcinoma,[7] small-cell lung cancer,[8] and gastric cancer.[9] Nacerddine reported that Bmi-1 showed its oncogenic potential, E3 ligase activity, and DNA damage repair activity in mouse prostate cancer.[10] Downregulation of the Bmi-1 gene by RNAi inhibited the proliferation and invasiveness of breast cancer cells and laryngeal carcinoma cells.[4,11]

The clinicopathological characteristics of Bmi-1 indicated its significance in clinical diagnosis and potential therapy.[4,12] Hayry et al[13] reported that Bmi-1 is an independent marker for poor prognosis in oligodendroglial tumors. Bmi-1 expression displayed a significant inverse association with patient overall survival (P = 0.006) and mean survival (P < 0.001).[4] A high level of Bmi-1 indicates unfavorable overall survival and serves as a high-risk marker for breast cancer.[14] Song et al[15] found that Bmi-1 transcriptionally downregulated the expression of the tumor suppressor PTEN in tumor cells through direct association with the PTEN locus. Sp1 and c-Myc were observed to regulate the transcription of Bmi-1 in nasopharyngeal carcinoma in tumor sample;[16] VEGF/neuropilin-2 regulation of Bmi-1 defines a novel mechanism of aggressive prostate cancer.[17] In a recent study, Jiang et al[18] revealed the Bmi-1 was involved in the nuclear factor kappa B (NFκB) pathway. However, an epigenetic regulatory mechanism for the overexpression of Bmi-1 in breast cancer has not been fully clarified.

MicroRNAs (miRNAs) are proven to inhibit gene translation or facilitate mRNA degradation, resulting in repression of target genes expression.[19] According to miRBase, >1000 different mature miRNAs have been identified in human.[20] As important epigenetic regulators, miRNAs have vital roles in cancer progression.[21] Many miRNAs function as oncogenes, such as miR-34a,[22] miR-320,[23] and miR-21,[24] whereas the others function as tumor suppressor genes such as miR-154,[25] miR-126,[26] and miR-203.[27] MiRNAs are involved in many important signal pathways, such as the TGFβ pathway,[28] AKT pathway[29] and WNT pathway.[30] MiR-22 overexpression induces phosphatase and tensin homolog (PTEN) downregulation and phosphoinositide 3-kinase (PI3K)/AKT pathway activation.[31] MiR-7 inhibits tumor metastasis and reversed the epithelial-mesenchymal transition through AKT and ERK1/2 pathway inactivation.[32]

Here, we reported that miR-495 was frequently downregulated in malignant cells and tissues of breast. Upregulation of miR-495 significantly suppressed breast cancer cell proliferation, possibly through G1-S arrest. We demonstrated that miR-495 directly targeted the 3′-untranslated region (3′ UTR) of the Bmi-1 mRNA and regulated the expression of PTEN, p21Cip1and p27Kip1, cyclin D1, and phosphorylated AKT. In vivo xenograft formation assays supported the phenotype observed with miR-495-transfected cells and Bmi-1 replenished cells. Our results suggest that frequent downregulation of miR-495 in breast cancer may influence the G1-S phase transition by targeting Bmi-1.

METHODS

Cell Culture

Normal breast epithelial cells (NBECs), breast cancer cell lines, and stably transfected cells were maintained according to our previous report.[33]

Real-time Polymerase Chain Reaction

The mirVana miRNA Isolation Kit (Ambion, Austin, TX), the Taqman miRNA reverse transcription kit (Applied Biosystems, Foster City, CA), the miRNA-specific TaqMan MiRNA Assay Kit (Applied Biosystems, Foster City, CA) and the Applied Biosystems 7500 Sequence Detection system were used to perform real-time quantitative polymerase chain reaction (PCR) as previously described.[34] The primers used were as follows: Bmi-1, forward, 5′-TCGTTGTTCGATGCATTTCT-3′, and reverse, 5′-CTTTCATTGTCTTTTCCGCC-3′; p21Cip1, forward, 5′-CGATGCCAACCTCCTCAACGA-3′, and reverse, 5′-TCGCAGACCTCCAGCATCCA-3′; p27Kip1, forward, 5′-TGCAACCGACGATTCTTCTACTCAA-3′, and reverse, 5′-CAAGCAGTGATGTATCTGATAAACAAGGA-3′; cyclin D1, forward, 5′-AAC TACCTGGACCGCTTCCT-3′, and reverse, 5′-CCACTT GAGCTTGTTCACCA-3 GAPDH, forward, 5′-GACTCATGACCACAGTCCATGC-3′, and reverse, 3′-AGAGGCAGGGATGATG TTCTG-5′.

Plasmid and Transfection

The only 1 miR-495 binding site of Bmi-1 3′ UTR is from 326 base pairs (bp) to 333 bp. The 230 bp-length sequence of human Bmi-1–3′ UTR (from 204 to 453) was cloned into the pGL3-basic luciferase reporter plasmid (Promega, Madison, WI) and pGFP-C3 (Clontech, Mountain View, CA).[34] The p3xIRS-MLP-luc plasmid, pMSCV/Bmi-1(with 3′ UTR or without 3′ UTR), and pMSCV /miR-495 were constructed as previously described.[34-36] The primers selected were as follows: miR495-up: 5′-GCCAGATCTGCTTTATCCGTCATGACTGT-3′; miR495-dn: 5′-GCCGAATTCTGGCTGCTATGATCTGAACT-3′; Bmi-1–3′ UTR-wt-up: 5′-GCCCCGCGGGAATTGGTTTCTTGGAAAGCAG-3′; Bmi-1–3′ UTR-wt-dn: 5′-GCCCTGCAGATGCATATTCTTTCCGTTGG-3′; Bmi-1–3′ UTR-mut-up: 5′-TAAGCATTGGGCCATAGTTTCAAAATCTCAACTAACGCCTACATTACATTCTCC-3′; Bmi-1–3′ UTR-mut-dn: 5′-GGAGAATGTAATGTAGGCGTTAGTTGAGATTTTGAAACTATGGCCCAATGCTTA-3′.

Western Blot

Western blot analysis was performed as previously described,[33] using anti-α-tubulin, anti-Bmi-1(Millipore, Billerica, MA), anti-p21, anti-p27, anti-cyclin D1, anti-Rb, anti-phosphorylated Rb (Abcam, Cambridge, MA), anti-PTEN, anti-AKT, anti-phosphorylated AKTThr308 and anti-phosphorylated AKTSer473 (Sigma, St. Louis, MO) antibodies.[34]

3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium Bromide Assay

Ninety-six-well plates were used to culture 1 × 104 cells each hole and all experiments were performed in triplicate as to our previous report.[33,34]

Colony Formation Assays

Five hundred cells per plate were incubated on 6-well plates for 10 days. The colonies were fixed, stained, and counted as to our previous report.[33,34]

Anchorage-independent Growth Ability Assay

Five hundred cells/well were seeded, counted, and measured according to our previous report.[33,34]

5-Bromo-2-deoxyuridine Labeling and Immunofluorescence

The images of cells stained with anti-5-bromo-2-deoxyuridine (BrdUrd) antibody (Upstate, Temecula, CA) after incubation with BrdUrd for 1 h were acquired by a laser scanning microscope (Axioskop 2 plus, Carl Zeiss Co. Ltd.).[34]

Flow Cytometry Analysis

According to previous report,[34] 20,000 harvested cells were washed, fixed, pelleted, re-suspended, incubated with bovine pancreatic RNAase (Sigma, Saint Louis, MO), and stained with propidium iodide (Sigma-Aldrich) before analyzed on a flow cytometer (FACSCalibur; BD Biosciences).[34]

Luciferase Assays

According to the manufacturer's recommendation of the Lipofectamine 2000 reagent (Invitrogen Co, Carlsbad, CA), 100 ng of p3x IRSMLP -luciferase plasmid, or pGL3-Bmi-1–3′ UTR (wt/mut), or the control-luciferase plasmid, plus 1ng of pRL-TK renilla plasmid (Promega, Madison, WI) was transfected into the indicated cells.[34] The Dual Luciferase Reporter Assay Kit (Promega, Madison, WI) was used to detect luciferase and renilla signals 48 h after transfection.[34]

Xenografted Tumor Model, Immunohistochemistry, and Hematoxylin and Eosin Staining

BALB/c-nu mice (4–5 weeks of age, 18–20 g) were purchased from Hunan SJA Laboratory Animal Co, Ltd, Hunan, PR China. According to our previous report, the BALB/c nude mice were randomly divided into 2 groups (5 per group): the first group was inoculated in situ with MDA-MB-231-NC cells (5 × 106) in the left breast and with MDA-MB-231-miR-495 cells(5 × 106) in the right breast per mouse; the second group was inoculated in situ with MDA-MB-231–495-Bmi-1-vector cells (5 × 106) in the left breast and with MDA-MB-231–495-Bmi-1-ORF cells (5 × 106) in the right breast per mouse.[37] Tumors were measured and calculated every 3 days until the 37th day. The acquired tumors were subjected to immunohistochemistry using an anti-Bmi-1 antibody (1:500; Upstate, Temecula, CA) or hematoxylin-eosin (H&E) stain with Mayer hematoxylin solution. The Institutional Animal Care and Use Committee of Sun Yat-sen University approved all the experimental procedures.

Statistical Analysis

Student t test in SPSS13.0 (IBM, Armonk, NY) was used to evaluate the significant difference of 2 groups of data in all the pertinent experiments. A P value <0.05 (using a 2-tailed paired t test) was considered statistically significant.[33,34]

RESULTS

MiR-495 Expression is Downregulated in Breast Cancer

Analysis of the ArrayExpress dataset GSE45666 indicated that miR-495 expression was frequently decreased in 101 breast cancer tissues compared with 15 adjacent noncancerous breast tissues. Among the 101 cancer samples, the expression of miR-495 in 99 samples was lower than the mean of the 15 adjacent noncancerous breast tissues (Figure 1A). The details of these 116 samples are listed in the supplementary Table 1, http://links.lww.com/MD/A245 (see table, supplemental content, http://links.lww.com/MD/A245, which illustrates the detail information of the samples). Further analysis showed that miR-495 expression is significantly decreased in 10 tissues and slightly decreased in 3 tumor tissues (a total of 15 pairs) compared with their adjacent noncancerous tissues in GSE45666 (P < 0.001) (Figure 1B). As compared with NBEC, real-time PCR analyses revealed that miR-495 was significantly downregulated in all 9 examined cell lines (changing fold was from 3% to 33%, Figure 1C). To further confirm these differences, comparative analysis of expression of miR-495 was conducted by real-time PCR. As shown in Figure 1D, miR-495 was downregulated in malignant tumor samples compared with the paired adjacent tissues (Figure 1D). Taken together, these data revealed that miR-495 expression is frequently downregulated in breast cancer.

FIGURE 1

The expression of miR-495 in breast cancer. (A) The expression of miR-495 was frequently downregulated in 101 breast tumor tissues (Tumor) compared with 15 adjacent noncancerous breast tissue samples (Adjacent) in the GSE45666 dataset. (B) MiR-495 expression was markedly reduced in 15 paired breast tumor tissues (Tumor) and their adjacent noncancerous tissues (Adjacent) in the GSE45666 database. (C) Real-time PCR analysis of miR-495 expression in indicated cells. (D) The expression of miR-495 in eight collected fresh tumor tissues (Tumor) and their adjacent noncancerous tissues (Adjacent). U6 was used as the normalized gene. Three independent experiments were conducted in each assay.

Overexpression of miR-495 Suppresses Proliferation

To examine the effect of miR-495 on the progression of breast cancer, breast cancer cells MCF-7 and MDA-MB-231 stably overexpressing miR-495 were established (Figure 2A). 3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide (MTT) assays showed that miR-495-transfected cells had a significant decrease in the growth rate compared with that of negative-control-transduced cells (NC) (Figure 2B). The results showed that 421 colonies were present in miR-495-transfected MCF-7 cells, compared with the 112 colonies in NC-transfected MCF-7. The same trend was found in the MDA-MB-231 cells (Figure 2C). As Ki-67 is considered to be a proliferation marker, the staining of Ki-67-positive cells was employed to detect the rapidly proliferating cells.[38] As the Figure 2D shown, the proportion of Ki-67-positive cells was dramatically decreased in miR-495-transfected MCF-7 and MDA-MB-231 cells, further suggesting that upregulation of miR-495 suppressed the proliferation of breast cancer cells. Moreover, colony number and size were both decreased in miR-495-transfected MCF-7 and MDA-MB-231 cells, further supporting the suppressive effect of miR-495 on proliferation in breast cancer cells (Figure 2E).

FIGURE 2

The effect of miR-495 on the proliferation and tumorigenicity of breast cancer cells MCF-7 and MDA-MB-231. (A) Increasing levels of miR-495 expression as assessed by real-time PCR. (B) MTT assays showing that miR-495-transfected cells grow slower than the NC-transfected cells. (C) Representative micrographs and quantification of formative colonies (MCF-7-NC, MCF-7-miR-495, MDA-MB-231-NC, MDA-MB-231-miR-495), which were crystal violet-stained. (D) The proportion of Ki-67 positive cells in cells transfected with miR-495 decreased. (E) Representative micrographs (left) and numbers of colonies in the anchorage-independent growth assay. Three independent experiments were conducted in each assay. ∗P < 0.05. MTT = 3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide.

MiR-495 Inhibits the G1-S Phase Transition

We further investigated the mechanism by which miR-495 suppressed cell proliferation. Flow cytometry analyses revealed that miR-495 could reduce the proportion of cells entering S phase and increase the proportion of cells staying at G1/G0 phase. The results showed that 66.93% G1/G0 phase cells and only 17.03% S phase cells were present in miR-495-transfected MCF-7 cells, compared with the 53.17% G1/G0 phase cells and 34.08% S phase cells in NC-transfected MCF-7. Similar results were found in the MDA-MB-231 cells (Figure 3A). Furthermore, BrdUrd incorporation assay showed a markedly decreased percentage of newly synthesized DNA in miR-495-transfected MCF-7 cells (9.15%) and miR-495-transfected MDA-MB-231 cells (11.01%) compared with NC-transfected cells (26.04% in MDA-MB-231; 31.12% in MCF-7) (Figure 3B). Collectively, our data suggested that the suppressive effect of miR-495 on breast cancer cell growth probably attributes to the G1-S transitional arrest.

FIGURE 3

The inhibitory role of miR-495 in the G1-S phase transition. (A) Obvious changes in the percentage of cells in the G1/G0 peak or S peak in miR-495-transfected breast cancer cells was detected by the flow cytometry analysis. (B) Representative micrographs (left) and quantification (right) of BrdUrd incorporating cells indicated miR-495-mediated proliferation inhibition. (C) The protein expression and mRNA levels of indicated cell-cycle-associated genes were tested by real-time PCR and western blot analysis. GAPDH and α-tubulin were used as a loading controls, separately. ∗P < 0.05.

MiR-495 Modulates Cell-cycle Regulators

Cell cycle is verified to be modulated by many cell-cycle promotors and inhibitors, such as Rb and cyclin D1 in the regulation of the G1/S transition.[39,40] Compared with the NC-transfected cells, the protein and mRNA expressions of p21Cip1 and p27Kip1, 2 cell-cycle inhibitors, were increased in miR-495-transfected cells. However, the expression of cyclin D1, a cell-cycle promoter, was decreased in miR-495-transfected cells. The phosphorylation level of Rb, positively correlating with cell-cycle, was dramatically decreased in miR-495-transfected cells (Figure 3C). These results further suggested that miR-495 inhibits the G1-S transition in breast cancer cells.

MiR-495 Directly Targets the Oncogene Bmi-1

We screened 3 publicly algorithms (Pictar, miRANDA, and TargetScan) searching for a potential target gene of miR-495. As shown in Figure 4A, diagrammatic sketch of predicted consequential pairing of target region and miRNA indicated that Bmi-1 is a potential target gene of miR-495. In consideration of the effects of miRNAs on repression of target genes expression, we detected the predicted Bmi-1 gene expression by western blot. Western blot revealed that ectopic expression of miR-495 in MCF-7 and MDA-MB-231cells reduced the expression of the Bmi-1 protein (Figure 4B). AKT, a gene typically affected by Bmi-1, showed changes in its phosphorylation levels consistent with reduced Bmi-1 expression (Figure 4B). PTEN, a tumor suppressor included in the AKT pathway, was upregulated in miR-495-transfected cells (Figure 4B). Furthermore, plasmids of Bmi-1 3′ UTR fragment containing predicted sequences (vectors are pEGFP-C3 and pGL3) were transfected in MCF-7 and MDA-MB-231 cells. As shown in Figure 4C, the expression of the GFP protein, but not the expression of GFP-γ-tubulin, was dramatically decreased in miR-495-transfected cells compared with in the NC-transfected cells. These data indicated an immediate relevance between miR-495 and Bmi-1–3′ UTR. Luciferase activity was also reduced in miR-495-transfected cells in a dose-dependent manner. Nevertheless, the mutation in the specific region of Bmi-1–3′ UTR (GUU to CAA) abrogated the inhibitory effect of miR-495 (Figure 4A and D). Taken together, we concluded that Bmi-1 is a genuine target of miR-495.

FIGURE 4

MiR-495 directly targeting the Bmi-1-3′ UTR. (A) Diagrammatic sketch indicated the miR-495 target sequences (wt: wild-type; mut: mutant) in the 3′ UTR of Bmi-1 (B) Western blot analysis of Bmi-1, PTEN, p-AKT(Thr308), p-AKT(Ser473), and total AKT expression in indicated cells. (C) Western blotting analysis of GFP expression in the indicated cells. (D) The luciferase activity of Bmi-1-3′ UTR was tested in the indicated cells with increasing amounts of miR-495 plasmid (20 nmol/L, 50 nmol/L). Error bars correspond to the mean ± standard deviation. ∗P < 0.05. UTR = untranslated region.

Bmi-1 is Intimately Involved in miR-495-induced G1-S Arrest of Breast Cancer Cells

To determine the role of Bmi-1 in miR-495-mediated G1-S arrest, the plasmids of Bmi-1 open reading frame (ORF, without the 3′ UTR), and Bmi-1–3′ UTR (ORF with the 3′ UTR) were created and used to process miR-495-transfected cells. As expected, overexpressing Bmi-1-ORF, but not Bmi-1–3′ UTR, could indeed dramatically reduce the expression of p21Cip1/ p27Kip1 and increase the expression of cyclin D1, suggesting an obvious attenuated effect of Bmi-1-ORF on the modulation of cell-cycle regulators by miR-495 overexpression (Figure 5A). Consistently, ectopically expressing Bmi-1-ORF, but not Bmi-1-3′ UTR, could rescue the luciferase activity of Bmi-1 reporter modulated by miR-495 (Figure 5B). Furthermore, we investigated the effect of Bmi-1-ORF or Bmi-1-3′ UTR on the growth rate in the miR-495-transfected cells by the MTT assay. As shown in the Figure 5C, only Bmi-1-ORF dramatically accelerated the proliferation of miR-495-transfected cells. Meanwhile, flow cytometry showed that only co-transfection of miR-495 and Bmi-1-ORF dramatically increased the S phase percentage of MCF-7 and MDA-MB-231 cell lines (Figure 5D). These results indicated that Bmi-1 is intimately involved in miR-495-induced G1-S arrest of breast cancer cells.

FIGURE 5

The role of Bmi-1 replenishment in miR-495-induced G1-S arrest. (A) Real-time PCR analysis of the expression of indicated genes in MCF-7 (left) or MDA-MB-231 (right) cells. GAPDH was used as a loading control. (B) The luciferase activities of Bmi-1 reporter were tested in indicated cells. (C) MTT assays revealed the growth rate of breast cancer cells transfected with NC, miR-495, or mixed miR-495 and Bmi-1-ORF (without 3′ UTR), or mixed miR-495 and Bmi-1-3′ UTR (ORF + 3′ UTR). (D) Flow cytometry analysis of indicated breast cancer cells. MTT = 3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2H-tetrazolium bromide, UTR = untranslated region.

The Relationship Between miR-495 and Bmi-1 Pathway Is Verified in Clinical Breast Samples

Human primary tumor samples were used to further confirm the above conclusion according to researches in subculture breast cancer cells. The expression levels of miR-495, Bmi-1, p21Cip1, p27Kip1, and cyclin D1 were examined in 8 paired noncancerous human breast tissues (N1–N8) and breast tumor tissues (T1–T8). Colors in Figure 6 depict the expression of indicated genes tested by real-time PCR (red represents higher expression; green represents lower expression). The correlation coefficient and significance between miR-495 and indicated genes were listed in Figure 6. MiR-495 expression was negatively correlated with Bmi-1 (r = −0.802, P < 0.001) and cyclin D1 (r = −0.729, P = 0.001), whereas it was positively correlated with p21Cip1 (r = 0.713, P = 0.002) and p27Kip1 (r = 0.772, P < 0.001). Collectively, our results strengthened the hypothesis that miR-495 mediates G1-S arrest of breast cancer through suppression of pro-oncogene Bmi-1.

FIGURE 6

The relationship between miR-495 and Bmi-1, cyclin D1, p21Cip1, and p27Kip1 in 8 paired normal human breast tissues (N1–N8) and human breast cancer tissues (T1–T8). (A) Red represents higher expression and green represents lower expression tested by real-time PCR. (B) Correlation analysis between miR-495 and Bmi-1 or other indicated genes by the SPSS13.0 software. PCR = polymerase chain reaction.

In Vivo Assay Reveals the Suppressive Role of miR-495 on Tumorigenicity

BALB/C nude mice were used to test and verify the above conclusion using MDA-MB-231. As shown in Figure 7A, miR-495-tranfected cells showed an anti-proliferative tendency in nude mice. Tumor volume and tumor weight generated from miR-495-transfected cells were decreased compared with that from vector-transfected cells (Figure 7B and Figure 7C). The expression of miR-495 was determined in all transplanted tumors after the mice were killed. Real-time PCR revealed miR-495 was intensively expressed in that tumors generated from miR-495-transfected MDA-MB-231 cells, but weakly expressed in the vector-transfected group (Figure 7D). Supplementation of Bmi-1 attenuated the repressive effect of miR-495 in the nude mice transplant assay. The histological statuses of these tumors were validated by H&E staining and immunohistochemical staining for Bmi-1 in the indicated excised tumors was performed (Figure 7E). As expected, Bmi-1 was decreased in the miR-495-transfected group compared with the vector-transfected group and was increased in the Bmi-1 complementary transfected cells. Taken together, our results demonstrated that Bmi-1 plays an important role in the reduced tumorigenicity induced by miR-495 in nude mice.

FIGURE 7

The inhibitory effect of miR-495 and the neutralizing effect of Bmi-1 on proliferation indicated by in vivo assays. (A) Images of excised tumors 37 days after injection. (B) The growth curve of tumor measured every 3 days from the beginning of the 10th day. Data are means ± standard error mean of all tumor volumes per group. (C) The weight of excised tumors measured on day 37 by electronic weighing. (D) The expression of miR-495 of indicated excised tumors as assessed by real-time PCR. (E) Hematoxylin and eosin staining (H&E staining) and immunohistochemical staining for Bmi-1 of indicated excised tumors. Three independent experiments were conducted in each assay. Error bars correspond to the mean ± standard deviation. ∗P < 0.05. PCR = polymerase chain reaction.

DISCUSSION

A report by National Cancer Institute pointed out that breast cancer will be the secondary diagnosed nonskin malignancy in US women in 2015.[41-43] Large tumor size, growing by cell proliferation, is commonly considered an established prognostic marker.[44] More and more evidence shows that gene disorders in cells deeply influence proliferation, tumorigenesis and treatment strategies.[45-47] Both loss and gain of critical genes in key pathways, especially those related to proliferation, are involved in cancer development and the selection of therapeutic strategies.[48] The AKT pathway is a critical regulatory mechanism mediated by many key gene products, such as Bmi-1,[36] NFκB,[39] Snail (snail family zinc finger 1),[40] and BCL2 (B-cell CLL/lymphoma 2).[41] In breast cancer, high Bmi-1 expression is correlated with advanced clinicopathological classifications (T, N, and M) and clinical stages.[14] Furthermore, a high level of Bmi-1 indicates an unfavorable overall survival and serves as a high risk marker for breast cancer.[14] Similarly, Hex-protein docking analysis revealed that the Bmi-1 gene was co-regulated (down) with the progesterone receptor in invasive ductal breast carcinoma, which supported Bmi-1 as a biomarker of breast cancer.[42]

Many studies have shown that Bmi-1 plays vital roles in multiple tissues. In Bmi-1−/− mice, DNA damage, impaired mitochondrial structure, decreased mitochondrial numbers, and increased oxidative stress occurred.[49] When the expression of Bmi-1 was blocked by siRNA transduction, the proliferation of oral epithelial cells was inhibited and immortalized cells was decreased.[50] Endogenously inhibition of Bmi-1 reduced the invasiveness of 2 HCC cell lines in vitro by upregulating PTEN expression and inhibiting the phosphatidylinositol 3-kinase (PI3K)/Akt signaling pathway.[51] However, how Bmi-1 is regulated in breast cancer remains to be investigated. It is reasonable to expect that the expression of Bmi-1 mediated by epigenetic changes, such as miRNAs, may contribute directly to the disease progression of various cancers.

Several miRNAs are involved in the regulation of the oncogene Bmi-1, such as miR-200c in radiation-induced thymic lymphoma,[52] miR-15a and miR-16 in ovarian cancer,[53] miR200b in prostate cancer,[53] and miR-203 in esophageal cancer;[27] For the first time, we revealed that miR-495 was dramatically decreased in breast cancer cell lines compared with NBECs. Furthermore, we demonstrated that ectopic expression of miR-495 drastically retarded the proliferation and tumorigenicity in in vitro and in vivo assays, suggesting that downregulation of miR-495 may associate with features of breast cancer and that it functions as an antimir. Consistent with present findings in breast cancer, the expression level of miR-495 is downregulated in gastric cancer,[54] prostate cancer,[55] and non-small cell lung cancer.[56] However, it has been reported that miR-495 directly suppressed E-cadherin expression to promote cell invasion, indicating that upregulated miR-495 plays important roles in epithelial-to-mesenchymal transition (EMT), which was further confirmed by Cao et al that miR-495 could induce breast cancer cell migration.[57,58] Interestingly, Dr. Robert A. Weinberg has documented that activation of the EMT program in carcinoma cells can give rise to cells with stem-like properties.[59] Indeed, miR-495 is found to be upregulated in breast cancer stem cells.[57] Moreover, Cao et al[58] found that miR-495 also upregulated in breast cancer samples. Therefore, miR-495 might be expressed in the different levels in different subpopulation of breast cancer cells, which miR-495 is upregulated in the cancer stem cells that also showed highly invasive ability, but is downregulated in differentiated breast cancer cells that showed high proliferative rate. Our data may further confirm and explain the consistent detection that miR-495 inhibits proliferation of glioblastoma multiforme cells by downregulating CDK6.[60] Overall, these studies prompted us to investigate further the relationship between miR-495 and clinical characteristics, such as tumor, node, and metastasis in extensive clinical samples.

Why is miR-495 frequently downregulated in breast cancer? One study observed that miR-495 is upregulated by E12/E47, which prompted us to look for a regulator of miR-495.[57] By analyzing the position of miR-495 in the genome, the promoter regulatory region of miR-495 might be: chromosome 14:101,498,500–101, 500,091. The ECR browser, a tool for visualizing and accessing data from comparisons, was used to predict conserved transcription factor-binding sites in the identified region. Growth factor independent 1 transcription repressor (GFI1), known to form a large protein complex with co-repressors that recruit histone deacetylases, has 3 binding sites in the miR-495 promoter (see supplementary Figure S1, http://links.lww.com/MD/A245). GFI1 is overexpressed and acts as a proliferation repressor in chronic myelogenous leukemia,[53] neuroendocrine lung cancer cell lines,[61] and prostate cancer cells.[62] By searching the GSE profiles, it was indicated that GFI1 is upregulated in breast cancer compared with normal breast tissue, according to profile GDS2617. Therefore, we speculated that GFI1 is a direct regulator of miR-495 and plan to verify this hypothesis in the future experiments.

Firstly, we elucidated that miR-495-mediated G1-S arrest was associated with downregulation of oncogene Bmi-1. The current findings suggest a significant biological role of miR-495 in the regulation of proliferation and tumorigenicity of breast cancer cells. Our study suggests a potential usefulness of miR-495-targeting strategies to deliver an anti-proliferative therapeutic effect.