Background
Micro RNAs (miRNAs) are small non-coding, cellular RNAs (17-27 bp) that post-transcriptionally regulate gene expression by inducing the degradation or translational repression of target mRNAs. The discovery of miRNAs and their mode of action has revealed an entirely new level of gene regulation. miRNAs must be assembled into a complex termed the RNA induced silencing complex (RISC) in order to regulate expression of their mRNA targets. Once assembled they act by binding to the 3'untranslated region (3'-UTR) and inducing degradation or transcriptional repression [
1]. An individual miRNA is capable of regulating hundreds of distinct mRNAs, and more than 1,000 human miRNAs have been identified that could potentially modulate close to one-third of the coding genes in human genome [
2].
Aberrant expression of miRNAs has been correlated with various human diseases including cancers. miRNAs have been identified which have oncogenic or tumor suppressor properties because the target genes they regulate are oncogenes or tumor suppressor genes. The abnormal expression profiles of miRNAs have been examined in many different cancers including breast cancer [
3,
4] and their roles in the proliferation, apoptosis, invasion/metastasis and angiogenesis of normal and cancer cells are being investigated aggressively [
5‐
9].
The function of the miRNA miR-135a is currently under investigation in our laboratory. Processes known to be under the control of miR-135a include megakaryocytopoiesis[
10], bone and muscle development, hypertension, colorectal cancer through its target gene
Adenomatous Polyposis Coli (APC)[
11‐
13], epithelial ovarian cancer and endometriosis through its target gene
HOXA10[
14], portal vein tumor thrombus through its target gene
metastasis suppressor 1 and Hodgkin disease and gastric cancer through its target JAK2[
15,
16]. At present however, its role in breast cancer is unknown.
Normal development and tumorigenesis both depend on shifts in the delicate balance between cell growth and differentiation. Altered expression of genes that are involved in the transcriptional control of developmental pathways often contribute significantly to oncogenesis because cancer can arise from the misappropriation of signalling pathways normally used to control cell fate [
17]. The
Homeobox genes encode transcription factors that are critical for the proper placement of segment structures during embryonic development. Analysis of numerous tumors have revealed that the expression of specific
HOX genes is often increased or decreased, indicating that they can influence tumor suppression or tumor development, invasion, and metastasis [
18]. Several
HOX genes have been shown to be partially regulated by miRNAs [
6,
19‐
21], indicating the potential for oncogenic pathways that begin with miRNA dysregulation, leading to altered
HOX gene expression and ultimately tumorigenesis or tumor suppression. The
HOXA10 gene is a regulator of embryonic morphogenesis and differentiation and is aberrantly expressed in several types of cancers [
22‐
29]. Recently it was reported to induce p53 expression in breast cancer cells and to reduce their invasiveness [
30]. Of specific interest,
HOXA10 has been shown to be under the control of miR-135a in endometrial and epithelial cancer tissues [
14] which led us to ask if miR-135a might also be associated with breast cancer.
In this study, we found that miR-135a levels are elevated in breast cancer with metastasis. By manipulating the expression level of miR-135a in vitro, we showed that miR-135a could promote the migration and invasiveness of breast cancer cells. We used bioinformatic tools and the literature to identify HOXA10 as a miR-135a target gene candidate and verified it is directly regulated by miR-135a in breast cancer cells. We found that over expression of HOXA10 could partially reduce the invasive property mediated by elevated miR-135a levels in the breast cancer cell line BT549.
Methods
Cell culture and tissue samples
All cell lines were obtained from the American Type Culture Collection. HEK293, human breast cancer cell lines BT549, SKBr3 and MDA-MB-231 were cultured in Dulbcco"s Modifed Eagle Medium (Gibco, Grand Island, NY, USA). All cell lines were incubated at 37°C in 5% CO2. Patients samples are collected from Zhongshan Hospital, Fudan University. This research was taken under consent of all patients for the use of their samples. This program is got approval by the Institute of Biomedical Sciences ethics committee of Fudan University. Ten Benign patient diagnosed as adenosis or fibroadenoma, fifteen invasive breast cancer as malignant samples whose ER(-), PR(-) and CerbB-2 (-) markers are all negative and all have lymph node metastasis. Fresh tissue samples were harvested from patients, and preserved at -80°C.
Detection of mRNAs and miRNAs
Total RNA from cells and human tissue samples were extracted using Trizol (Invitrogen) according to the manufacturer"s instructions and 2 μg of each total RNA sample was aliquoted to synthesize cDNA using the Reverse Transcription System (Promega). For mRNA detection and expression, HOXA10 and β-actin were analyzed by RT-PCR or qRT-PCR. All qRT-PCR products were amplified using a SYBR green PCR Master Mix kit (Qiagen) according to the manufacturer"s instructions on the ABI Prism 7,500 Detection System (Applied Biosystems). For quantification of HOXA10 mRNA in transfected cells, β-actin was used as the internal control. Levels of mRNA were quantified based on the ratio of HOXA10 mRNA/β-actin mRNA using the 2 - ΔΔCt method where ΔΔCt = ΔCtexp - ΔCtnc = (Ctexp-target - Ctexp-actin) - (Ctnc-target - Ctnc-actin), in which "exp" represents the experimental group, "nc" the negative control group, and "target" the gene of interest.
For miRNA detection, we employed Poly (A) RT-PCR method using specific forward primers and a universal reverse primer complementary to an anchor primer as previously described [
31]. The anchor RT primer was used as the template for the negative control and the U6 small nuclear RNA was used as the control to determine relative miRNA expression. Levels of miR-135a in human tissue samples were quantified and normalized to 18S rRNA using the 2 - ΔΔCt method formula as described above. The PCR profile was one cycle of 95°C for 5 min, then 40 cycles of 95°C for 5 s and 60°C for 50 s. The primers used for detection are listed in Table
1.
Table 1
Primers for detection of HOXA10 mRNA and miR-135a by RT-PCR
miR-135 RF
| cgcgtctatggctttttattccta |
anchor RT primer
| cgactcgatccagtctcagggtccgaggtattcgatcgagtcgcactttttttttttt |
Universal rev primer
| ccagtctcagggtccgaggtattc |
U6F
| ctcgcttcggcagcaca |
U6T
| aacgcttcacgaatttgcgt |
HOXA10 RT F
| ctggtcccctccctctgtc |
HOXA10 RT T
| acaacaaataaaccagcaccaag |
β-actin F
| ccttcctgggcatggagtcct |
β-actin T
| aatctcatcttgttttctgcg |
Plasmid construction and transfection
A DNA fragment encoding the miR-135a pre-miRNA was amplified by PCR from HEK293 genomic DNA and cloned into a modified pSilencer 4.1-CMVneo vector (Ambion) as previously described [
31]. Positive clones were identified by PCR screening and DNA sequencing. Using genomic similarly, 3"-UTRs from the predicted mir-135a target genes
HOXB7, APC and
HOXA10 were PCR amplified from HEK293 DNA and cloned into the XbaI site immediately downstream of the stop codon in the pGL3-promoter vector (Promega). To produce mutant
HOXA10 3'-UTR pGL3-reporter plasmids, the predicted miR-135a binding sites were replaced with 18 bp-long fragments (Table
2) by overlapping PCR. Fragments of
HOXA10 containing 3' UTR regions were cloned from a
HOXA10 cDNA into pEGFP-C1 for pEGFP-
HOXA10 plasmid construction. The
HOXA10 expression vector pcDNA3-hisC-
HOXA10 and vector pcDNA-
HOXA10-de-3'UTR were kindly provided by Dr Herring (Indiana University, USA). The miR-135a anti-sense oligonucleotide inhibitor (miR-135a inhibitor) and a mismatched sequence (N.C) were purchased from GenePharma (Shanghai, China).
HOXA10 and Control siRNAs were from Santa Cruz (sc-38685, sc-37007). The primers used were listed in Table
2. All plasmid DNAs used for transfection were prepared using a Qiagen DNA Miniprep kit following the manufacturer"s instructions. Cells were transiently transfected with vectors or anti-miR inhibitor using Lipofectamine 2000 from Invitrogen (CA, USA) according to the manufacturer"s instructions.
Table 2
Sequences of PCR primers for plasmid construction
HOXA10 UTR F
| ccTCTAGActggtcccctccctctgtc |
HOXA10 UTR T
| ccTCTAGAgatagggagaattgtggtgtgc |
HOXB7 UTR F
| ggTCTAGAgggcagaggaagagacatgag |
HOXB7 UTR T
| ggTCTAGAgggttagtccagacccacag |
APC UTR F
| ggTCTAGAttaaaagagaggaagaatgaaactaag |
APC UTR T
| ggTCTAGAgcatgtatctccattgtttatgg |
muHOXA10 F
| TCTTGGATCCTTCAAGTCtcatgctaaaattctatagagatag |
muHOXA10 T
| GACTTGAAGGATCCAAGAacaacaaataaaccagcaccaag |
In vitro luciferase assay
HEK293 cells (1 × 105) were seeded into a 96-well plate and co-transfected with 5 ng of internal control vector pRL-renilla (Promega), 50 ng of a pGL3-promoter reporter with either the HOXA10, mutant HOXA10, HOXB7 or APC 3'-UTR and 150 ng of the pSilencer-135a (pS-135a) or pSilencer-4.1-CMV-negative (pS-negative) vector. Forty eight hours after transfection, the firefly and Renilla luciferase activities were assayed using the Dual-Glo Luciferase assay system from Promega according to the manufacturer"s protocol. For each sample firefly luciferase activity was normalized to the Renilla luciferase activity value.
Western blotting
Whole-cell extracts were prepared using RIPA lysis buffer and SDS-PAGE followed by Western blotting were performed as previously described [
31] and signals were visualized with Super Signal West Femto chemiluminescent substrate. For each condition, samples were analyzed three times independently. The primary antibodies used were from Santa Cruz Biotechnologies for
HOXA10 (sc-28602), Sigma for β-actin (A7441), and Cell Signaling for GFP (#2956). The peroxidase-conjugated anti-mouse or anti-rabbit IgG secondary antibodies were from Santa Cruz Biotechnology.
Cell proliferation assay
MDA-MB-231 or BT549 (1 × 104 cells per well) were plated into 96-well plates in triplicate prior to pSilencer 4.1-CMV-135a or pS-negative transfection and then cultured for approximately 16 h. Cell proliferation was assessed 72 h after transfection using CCK8 (Dojindo, Tokyo, Japan) according to the manufacturer"s instructions.
Migration and invasion assay
Cell migration and invasive ability was examined using a 24-well transwell plate with 8 mm pore polycarbonate membrane inserts, according to the manufacturer"s protocol (Corning, NY, USA). The matrigel (14.8 μg/ml) employed for the invasion assays was applied to the upper surface of the membranes. Forty eight hours after transfection, 5 × 104 cells per well were seeded into the top chamber in serum-free media and this was replaced with complete growth media for 12 h. Cells that migrated or invaded through surface of the membrane were fixed with methanol and stained with hematoxylin. Migrating or invasive cells from three random microscope fields per filter were selected for cell counting.
We computationally screened proper targets for miR-135a by miRNAMap
http://mirnamap.mbc.nctu.edu.tw/. Statistical analysis for miR-135a expression in tissues was performed using a Jonckheere-Terpstra exact test for trend to compare the distribution of expression levels (high, medium, low, none) across breast tumors and benign tissues. A Bonferroni adjustment was applied to the
p values for the pair-wise comparisons. Results were delineated as means ± S.D., differences were tested for significance using 2-sided Student"s
t-test.
Discussion
miRNAs are small noncoding regulatory RNAs that have been studied in various types of cancers. Many miRNAs that regulate epithelial to mesenchymal transition (EMT) [
32] and pro-metastatic [
5,
6] or anti-metastatic functions [
9] have been identified. Previously miR-135 has been reported to regulate genes in many other types of cancer [
11,
14,
16,
33,
34], but its roles in breast cancer was unknown. Our current study provides the first evidence to demonstrate that miR-135a plays a role in promoting migration and invasion of breast cancer cells.
The ability of miR-135a to promote cell migration and invasion was assessed by both over-expression and down-regulation experiments. Remarkably, we observed apparent high levels of miR-135a expression in „triple negative" malignant invasive breast tumors (Figure
1A), and the highly invasive phenotype of the BT549 cell line (Figure
1B), suggesting that miR-135a might play an important role in maintaining metastatic functions. This hypothesis is supported by our experiments showing that inhibition of miR-135a activity impaired the invasion and migration of BT549 cells
in vitro (Figure
2A). To further verify, this relationship up-regulation experiments were performed in cell lines with different invasive phenotypes. In addition to the enhanced migration and invasive ability of SKBr3 cells induced by increased miR-135a expression (Figure
2B), we also observed enhanced invasive ability of miR-135a-transfected MDA-MB-231 cells (Figure
2C). Although endogenous miR-135a was not detectable in MDA-MB-231 cells they are highly invasive, and we speculate that migration and invasion may not be miR-135a-dependent processes in those cells due to genetic differences between different breast cancer cell lines. Furthermore, miR-135a could not affect the low invasive property of MCF-7 cells even though it was forced overexpressed and led to decreased endogenous HOXA10 protein expression in those cells (Figure
4D). We speculate this is also due to genetic cell specificity. For example, E-cadherin is highly expressed in MCF-7, so we believe that the effect of miR-135a over-expression is not strong enough to induce a change of the migration status of the MCF-7 cells. This suggests that the native non-metastatic character of MCF-7 may have other miR-135a-independent mechanisms that are responsible to maintain the non-aggressiveness of this particular cell type. It is also possible that HOXA10 3'UTR is mutated and cannot be targeted by miR-135a in MCF-7 cells. Future studies are required to define these two possibilities. Indeed it has been reported that HOXA10 is not regulated by miR-135a in MCF-7 cells [
14] which is in contrast to our results for other cancer cell lines tested. Unlike the migration and invasion phenotypes, up- or down-regulation of miR-135a did not affect cell proliferation (Figure
2D).
Several HOX genes are regulated by miRNAs [
6,
19‐
21]. This work is the first to implicate miR-135a down-regulation of
HOXA10 expression in breast cancer cell invasiveness. The mechanism by miR-135a targets
HOXA10 for repression was verified by in vitro 3'-UTR luciferase assays.
HOXA10 over-expression in miR-135a expressing cells was dependent on the absence of the
HOXA10 3'-UTR (Figure
3C-D) supporting our conclusion that miR-135a inhibits
HOXA10 via targeting its 3'-UTR. Our results showed that endogenous
HOXA10 in MDA-MB-231 was not detectable (Figure
3B) and this could be attributable to methylation of the
HOXA10 promoter as previously reported [
35]. Combined with the fact that individual miRNAs have the potential to modulate the expression of many mRNAs, our result showing miR-135a expression in MDA-MB-231 cells that increased migration and invasion (Figure
2C) suggest there may be miR-135a targets other than
HOXA10 that can promote migration/invasion events. Indeed, miR135a was reported to be up-regulated in portal vein tumor thrombus and these cells showed increased migration and invasion in vitro. However, the metastasis suppressor 1 gene and not
HOXA10 was found to be the direct, functional target of miR-135a in this tissue [
16]. We propose that this gene may function in some breast cancers to suppress migration and invasion rather than
HOXA10, and we intend to test this in the near future. Our results also suggested that miR-135a post-transcriptional down-regulation of the
HOXA10 target gene was not restricted to translation repression (Figure
4A-B,
4D-E) but also occurred by inducing mRNA degradation (Figure
4C), which agrees with previous reports on the action of miRNAs and highly homologous targets [
36].
To further investigate the effect of
HOXA10 on breast cancer cell invasiveness, we overexpressed
HOXA10 in already highly invasive breast cancer cell lines. Overexpression of
HOXA10 without 3' UTR targeting by miR-135a led to decreased invasion of MDA-MB-231 cells (Figure
5A), and significantly inhibited the miR-135a-regulated invasion of BT549 cells (Figure
5D). Knockdown of
HOXA10 increased invasiveness of BT549 cells and could partially rescue the decreased invasive property caused by 135a inhibitor (Figure
5B). Other members in the same gene family have also been shown to play roles in breast cancer. It has previously been reported that HOXA9, a paralog of
HOXA10, is a tumor suppressor in breast cancer [
37], and expression of HOXD10 in MDA-MB-231 significantly impaired migration [
38]. Interestingly, in BT549 cells the expression of a full-length
HOXA10 cDNA was repressed. It is highly likely that the high endogenous miR-135a level inhibits
HOXA10 expression through targeting its 3'-UTR, since deletion of 3'-UTR led to a higher expression of
HOXA10 protein than the full-length cDNA
HOXA10 (Figure
5C), and the effect of
HOXA10 overexpression on cell invasion varied depending on the absence or presence of the
HOXA10 3'-UTR (Figure
5D). These results illustrate that miR-135a promoted cell migration and invasion, at least partially, through repression of
HOXA10 via its 3'-UTR and also verified the in vitro luciferase assay results (Figure
3C-D). However, there is no evidence that miR-135a regulation of
HOXA10 is exclusive.
HOXA10 may also be targeted by other miRNAs and it appears that promoter methylation is also an important regulatory mechanism for
HOXA10 in some tissues [
35,
39]. Our result is in an agreement with a former study that reported
HOXA10 expression inhibited matrigel invasion by breast cancer cells [
30]. The study also reported that
HOXA10 expression induced p53 production. Therefore, understanding the molecular mechanism of the regulation of
HOXA10 by miR-135a may provide a method to explore upstream regulation of
HOXA10 and connect it with p53 tumor suppressor signalling pathways in breast cancer. Also, the "triple negative" invasive breast cancer type has a very poor prognosis and as yet no antibody target has been reported for the treatment of this type of breast cancer. So we believe future investigation on
in vivo studies as well as on clinical specimens will confirm the importance of miR-135 and identify additional markers for diagnosis and treatment.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YC carried out the experimental studies, drafted and completed the manuscript. JZ set up the method called Poly (A) RT PCR for miRNA detection and participated in the design of the study. HW was in charge of the clinical samples selection and performed the proofreading. ZJ, CX, YD and LX disposed the tissue samples. FZ and JZ completed sample conservation. RL refined the manuscript. HZ and DM conceived of the study and performed the statistical analysis. All authors read and approved the final manuscript.