MicroRNA-125a represses cell growth by targeting HuR in breast cancer

Abstract

MicroRNAs (miRNAs) are a class of naturally occurring, small, non-coding RNAs that control gene expression during development, normal cell function, and disease. Although there is emerging evidence that some miRNAs can function as oncogenes or tumor suppressors, there is limited understanding of the role of miRNAs in cancer. In this study, we observed that the expression of miR-125a was inversely correlated with HuR expression in several different breast carcinoma cell lines. HuR is a stress-induced RNA binding protein whose expression is elevated or localization perturbed in several different cancers. Increased cytoplasmic localization of HuR is a prognostic marker in breast cancer. Real time PCR and gene reporter assays indicated that HuR was translationally repressed by miR-125a. Re-establishing miR-125a expression in breast cancer cells decreased HuR protein level and inhibited cell growth. Using MCF-7 breast cancer cells, we further clarified that miR-125a inhibited cell growth via a dramatic suppression of cell proliferation and promotion of apoptosis. In addition, cell migration was also inhibited by miR-125a overexpression. Importantly, the repression of cell proliferation and migration engendered by miR-125a was partly rescued by HuR re-expression. Our results suggest that miR-125a may function as a tumor suppressor for breast cancer, with HuR as a direct and functional target.

INTRODUCTION

HuR, a ubiquitously expressed member of the ELAV (embryonic lethal abnormal vision) family, is an RNA binding protein (RBP) that stabilizes mRNAs of genes that regulate cell proliferation, angiogenesis, apoptosis, rapid inflammatory response and the stress response ^1–4. To date, elevated expression of HuR has been suggested to contribute to carcinogenesis in a wide variety of human carcinomas including breast, colon, lung ovary, uterine, cervical, brain, prostate, gastric, and Merkel cell carcinomas ^5–15. Although primarily a nuclear protein, HuR stabilizes mRNAs in the cytoplasm and elevated cytoplasmic localization of HuR is associated with high histologic grade and poor survival of patients with breast, ovarian, and gastric adenocarcinomas ^7,8,12,15. As such, HuR has been proposed as a critical trigger, a strong prognostic marker and a key therapeutic target for tumorigenesis.

MicroRNAs (miRNAs) are a class of non-coding RNAs whose final product is a ~22 nt functional RNA that controls gene expression by targeting mRNAs for either translational repression or destabilization. As a new layer of gene regulation, miRNAs have diverse functions, including the regulation of cellular differentiation, proliferation, death and metabolism ^16–19. Thus, deregulation of miRNAs alters normal cell growth and development, leading to a variety of disorders including human cancer. To date, aberrant expression or mutation of miRNAs has been observed in many types of human cancers ^20–28. A significant percentage of miRNA-encoding genes are located at cancer associated genomic regions or in fragile sites ^29–32. A recent report uncovered a correlation between miRNA expression patterns and patient survival in hepatocellular carcinoma, wherein lower miRNA expression was associated with poor survival while higher miRNA expression was associated with better survival ³³. MiR-21 was found to be associated with advanced clinical stage, lymph node metastasis and patient poor prognosis in human breast cancer ²². Collectively, the above reports suggest a direct correlation between aberrant miRNA expression and human malignancy. Although the role of miRNAs as oncogenes or tumor suppressors has been the subject of extensive research in recent years, the roles of miRNA in tumorigenesis and their effective targets are largely undetermined.

Here, we investigated the role of miR-125 in breast cancer based on the observation that miR-125 was decreased in human breast tumors as compared to normal breast tissue (23). We found that miR-125 expression inversely correlated with HuR protein level in the tested breast carcinoma cell lines. miR-125a but not miR-125b, targeted HuR for translational repression via a target site in its 3' untranslated region (3'UTR). Re-establishing miR-125a expression decreased HuR level, suppressed cell growth through inhibiting proliferation and promoting apoptosis, and inhibited cell migration. Suppression of cell proliferation and migration by miR-125a was partially abrogated by HuR re-expression, showing that HuR is at least one downstream target of this miRNA. Our results indicate that miR-125a may be playing a tumor suppressor role in breast cancer, with HuR as a direct and functional target.

RESULTS

The level of miR-125 inversely correlates with HuR protein level in tumorigenic and nontumorigenic breast epithelial cells

Both miR-125 and HuR may contribute to the carcinogenesis of breast cancers. Aberrant miR-125 expression has been reported in breast cancer tumor tissue and cell lines ^23,34. In SKBR3 breast cancer cells, overexpression of miR-125 impairs cell growth, migration and invasiveness by suppressing ErbB2 and ErbB3 ³⁵. To determine whether or not HuR is a potential target of miR-125, we first performed a computational screening of the HuR 3'UTR for miRNA target sites using the Sanger miRBase database (Version 7.0 release). We found that bases 686–692 of the HuR 3'UTR (Genbank no. BC003376, Figure 1A) were perfectly complementary to the first 7 nucleotides from the 5' end of miR-125, identifying HuR as a putative target of miR-125. Nucleotides 2–7, numbered from the miRNA 5' end are the “seed sequence” critical for target recognition, and increasing the mismatch in the seed sequences significantly decreases the gene regulation function of miRNAs ^30,36.

Figure 1.

HuR protein level inversely correlates with miR-125 level in breast cancer cells. A, The 1208-nt HuR mRNA 3'UTR is shown (top) along with a predicted Hsa-miR-125 target site at nt 671-692, and the predicted duplex formed with either Hsa-miR-125a (middle) or Hsa-miR125b (bottom) variants. Lines show identity and dots show conservation between the 3'UTR target site and the miR sequence. Mutant indicates the miR-125 seed sequence that was deleted from the HuR 3’UTR for experiments in Figure 6B. B, HuR protein was detected by immunoblotting in the breast epithelial cell lines noted by the numbers and indicated at the bottom of the figure. The bottom graph shows the HuR level relative to MCF10A cells after normalizing to β-actin. C, The levels of miR-125a and miR-125b were analyzed by Northern blotting in the breast epithelial cell lines noted by the numbers and indicated at the bottom of the panel. The bottom graph shows the miR-125a or miR-125b level relative to MCF10A cells after normalization to U6 snRNA. 1. MCF10A is an immortalized breast epithelial cell line. 2–6. MCF-7, T47D, SKBR3, MDA-MB-231 and HMT3522-T4-2 are breast carcinoma cell lines. Representative blots are shown, and similar results were obtained from two additional independent experiments.

Interestingly, we observed an inverse correlation between the expression level of miR-125 and HuR protein level in MCF10A immortalized breast epithelial cells and in several breast carcinoma cell lines including MCF-7, T47D, SKBR3 and HMT3522-T4-2 (Figure 1, B and C). MCF10A cells (lane 1) had less HuR and higher miR-125 (both the miR-125a and 125b homologs) compared to the breast cancer cell lines that had increased HuR and decreased miR-125 (lanes 2–4, and 6). As an exception, high levels of both miR-125a/miR-125b and HuR were observed in MDA-MB-231 cells (lane 5). These results demonstrate a concordant down-regulation of miR-125 and up-regulation of HuR protein in breast cancer cell lines with the exception of MDA-MB-231 cells.

MiR-125 expression decreases HuR and cell number

To determine if miR-125 could target HuR in breast cancer cells, we transfected miR-125a or miR-125b precursors into MCF-7, T47D, MDA-MB-231 and MCF10A cells. As expected, transfection of miR-125 precursors resulted in increased expression of mature miR-125 in all cell types (Figure 2A). MiR-125 transfection reduced HuR protein level in MCF-7 and T47D cells, with miR-125a having a greater effect than miR-125b (Figure 2B). Interestingly, sequences outside of the seed region allowed miR-125a to reduce HuR level more effectively than miR-125b in these cells. In contrast, miR-125a only slightly decreased HuR expression in MCF10A cells and neither miR-125a nor miR-125b affected HuR protein level in MDA-MB-231 cells, both cell lines with higher baseline levels of miR-125a and miR-125b. To determine if interfering with miR-125a function would increase HuR protein level in MCF10A and MDA-MB-231 cells, we transfected an antagomir specific for miR-125a. HuR protein level increased in MCF10A cells in response to miR-125a specific antagomir but did not change in MDA-MB-231 cells (Figure S1). These results suggest that HuR may not be a target of miR-125a in MDA-MB-231 cells. Alternatively, as HuR is already increased in these cells, there may be other factors limiting a further increase in its expression.

Figure 2.

Overexpression of miR-125a represses HuR protein level and cell number. The indicated cells were mock transfected or transfected with miR-125a or miR-125b precursors. A, 48 h after transfection, miR-125a or miR-125b was detected by Northern blotting. Blots were reprobed for U6 snRNA as an internal control. B, 72 h after transfection, HuR protein level was detected by immunoblotting. HuR level relative to mock transfected cells after normalization to β-actin is graphed under the blots. C, 72 h after transfection, cell number in each condition was assessed by cell counting and expressed as percent of mock-transfected cells. Data in A is representative of 3 independent experiments. The data shown in B and C are means of three independent experiments.

To determine if expression of miR-125a or miR-125b affected cellular processes regulated by HuR, we asked if cell proliferation was affected. As shown in Figure 2C, cell number was decreased after miR-125 transfection to a degree that paralleled the decrease in HuR protein. These results indicated that miR-125 not only decreased HuR level, but also had the additional functional effect of reducing cell number. The greatest effect was seen in MCF-7 cells with a 62% reduction in HuR level and a 72% decrease in cell number upon transfection of miR-125a, suggesting that miR-125a might be a key factor contributing to HuR deregulation in MCF-7 cells. Therefore, MCF-7 cells were used to determine how miR-125a regulates HuR and its downstream functions. We previously showed that the G₁/S regulator cyclin E1 was a downstream target of HuR in MCF-7 cells ³⁷. As shown in supplementary Figure S2, along with HuR repression by miR-125a transfection, cyclin E1 was also repressed in MCF-7 cells, while a control miRNA had no effect. These results indicated that HuR was functionally repressed by miR-125a.

MiR-125a inhibits cell proliferation and promotes apoptosis by its functional target HuR

To determine the mechanisms underlying miR-125a–induced reduction of cell number in MCF-7 cells, we assessed cell proliferation using immunofluorescence staining for Ki-67, a proliferation marker expressed in all phases of the cell cycle. As shown in Figure 3A, transfection with miR-125a precursor (125a+vec) reduced the number of proliferating cells by 29% as compared to mock transfected cells (mock+vec; from 77% to 48% Ki-67 positive cells). To ask if post-transcriptional down-regulation of HuR is one of the mechanisms underlying the cell growth-repressing function of miR-125a, we re-established HuR expression in miR-125a–transfected MCF-7 cells by co-transfecting a plasmid expressing myc-tagged HuR (Figure 3A, 125a+HuR; see Figure 4A for immunoblot analysis of myc-HuR). Expressing exogenous HuR in miR-125a expressing cells partially rescued cell proliferation (from 48% to 60% Ki-67 positive cells). This 12% increase in proliferation was similar to the 9% increase seen when HuR was expressed by itself (mock+vec compared to mock+HuR, 77% to 86% Ki-67 positive cells). These results showed that miR-125a overexpression suppressed cell proliferation in part due to inhibition of HuR. Figure 3B shows that the decrease in MCF-7 cell number seen upon miR-125a expression (miR-125a+vec, from 100% to 36%, a 64% decrease as compared to mock+vec) was also only partially rescued by HuR overexpression (125a+HuR, up to 56% of mock+vec). Again HuR overexpression alone increased cell number to a similar extent, by about 18% (mock+HuR as compared to mock+vec). As MCF-7 cell proliferation was reduced by 29% and cell numbers were reduced by 64–72% after miR-125a expression alone (Figure 2C and Figure 3), we speculated that miR-125a might also promote apoptotic cell death, as HuR is anti-apoptotic ⁴.

Figure 3.

miR-125a suppression of cell proliferation is partially rescued by HuR. MCF-7 cells were mock transfected or transfected with miR-125a precursor for 24 h followed by transfection of pcDNA3.1 (mock+vec, 125a+vec) or pcDNA3.1mycHuR (mock+HuR, 125a+HuR) for 48 h. A, Cell proliferation was assessed by immunofluorescence analysis of Ki-67, a marker of dividing cells. Representative fluorescence micrographs are shown. Top panels, Ki-67 was visualized with Alexa Fluor 488 conjugated secondary antibody (green). Middle panels, nuclei were stained with DAPI (blue). Bottom panels, merge of Ki-67 and DAPI channels. Magnification, 200×. Percentage of Ki-67⁺ cells was calculated by counting 500 cells for each condition. B, Cell number in each condition was assessed by cell counting and expressed as percent of mock+vec transfected cells. The data shown are representative (micrographs) or means of three independent experiments. ** p< 0.01 vs. mock+vec, # p< 0.05 vs. 125a+vec.

Figure 4.

miR-125a promotion of apoptosis is independent of HuR. MCF-7 cells were mock transfected or transfected with miR-125a precursor for 24 h followed by transfection of pcDNA3.1 (mock+vec, 125a+vec) or pcDNA3.1mycHuR (mock+HuR, 125a+HuR) for 48 h. A, HuR and PARP-1 protein levels were detected by immunoblotting. The blot was stripped and reprobed for β-actin as a loading control. B, Cell apoptosis was detected and viewed by fluorescence microscopy using Annexin V Apoptosis Detection Kit. Early apoptotic cells label with Annexin V (green) while late apoptotic cells label with propidium iodide (PI, red staining). All experiments were repeated three times.

We assessed apoptosis by detecting the cleavage of the full-length 116 kDa PARP-1 to the 85-kDa form, a key event in the process of apoptosis. As shown in Figure 4A, miR-125a expression resulted in the cleavage of PARP-1 (125a+vec). As MCF-7 cells lack caspase-3 and do not undergo DNA fragmentation ³⁸, we did not observe the characteristic morphological features of apoptosis, such as chromatin condensation. Therefore, as an alternative measure of apoptosis, we examined cell surface expression of phosphatidylserine by Annexin V staining. In Figure 4B, fluorescently labeled Annexin V (green) bound to phosphatidylserine in early apoptotic cells while DNA in late apoptotic cells labeled with propidium iodide (PI, red). An increase was seen in both early and late apoptotic cells transfected with miR-125a (125a+vec) as compared to control cells (mock+vec).

We re-established HuR expression in miR-125a–transfected MCF-7 cells by co-transfecting a plasmid expressing myc-tagged HuR (Figure 4, 125a+HuR). In the HuR blot (top panel of Figure 4A, the lower band in the lane labeled 125a+HuR is endogenous HuR, while the upper band is myc-tagged HuR. As can be seen in this figure, expressing exogenous HuR resulted in an increase in the endogenous protein. Increasing HuR level in miR-125a expressing or mock-transfected cells did not decrease apoptosis (Figure 4A and B), as neither PARP-1 cleavage nor Annexin V levels were decreased. Taken together, these results indicated that miR-125a repressed MCF-7 cell numbers through inhibiting proliferation and promoting apoptosis. Inhibition of cell proliferation was in part due to downregulation of HuR, while promotion of apoptosis was not. As an additional control, we transfected MCF-7 cells with miR-125b precursor to ensure that effects on apoptosis and proliferation were due specifically to miR-125a, and not general overexpression of a miRNA. Figure S3 shows that miR-125b had less of an effect on proliferation than miR-125a, and did not induce apoptosis, consistent with its decreased ability to reduce cell number as compared to miR-125a.

MiR-125a inhibits cell migration

Cell migration is a key process in tumor development, malignancy and metastasis ^39–41. Since HuR promotes cell migration ⁴², the effect of miR-125a expression on the migration of MCF-7 cells was assessed using a transwell assay. As shown in Figure 5A and B, the migration capacity of MCF-7 cells was reduced by 62% after miR-125a precursor transfection. Co-transfection of HuR partially abolished miR-125a–induced suppression of cell migration (Figure 5C, 125a+HuR). These results suggest that decreased miR-125a expression in MCF-7 cells facilitates their migration, in part due to inhibition of HuR.

Figure 5.

miR-125a inhibition of cell migration is partially rescued by HuR. MCF-7 cells were mock transfected or transfected with miR-125a precursor for 48 h or for 24 h followed by transfection with pcDNA3.1 (mock+vec, 125a+vec) or pcDNA3.1mycHuR (mock+HuR, 125a+HuR). 48 h following transfection, 2 × 10⁵ cells in serum-free media were transferred to the top surface of 8-mm transwell chamber inserts. Medium supplemented with 20% serum was used as a chemoattractant in the lower chamber. After 24 h incubation, cells remaining on the top were removed and the lower surface of the membrane stained with Crystal Violet, photographed and counted. A, Representative microscopic fields with stained cells are shown. B and C, Five random fields were counted for each condition and expressed as a percentage of the control cell number (bottom). Experiments were repeated 4 (B) or 3 (C) times. ** p< 0.01 vs. mock or mock+vec, # p< 0.05 vs. 125a+vec transfection.

HuR is translationally repressed by miR-125a

We next asked how miR-125a reduced expression of HuR. In general, miRNAs control gene expression by targeting mRNAs for either translational repression or mRNA decay. Using real time RT-PCR, we found that miR-125a transfection did not significantly change HuR mRNA level in MCF-7 cells (Figure 6A). However, miR-125a reduced the activity of a luciferase reporter fused to the wild-type HuR 3'UTR by 50% (Figure 6B, HuR3'UTR-WT). Moreover, deletion of the 7-nt sequence in the HuR 3'UTR complementary to the miR-125a seed sequence restored the luciferase activity of miR-125a transfected cells from 50% to 87% (Figure 6B, HuR3'UTR-mut), showing that the action of miR-125a on HuR depended on the presence of a single miR-125a cognate binding site within the 3'UTR. Thus, miR-125a likely decreased HuR level through translational repression.

Figure 6.

miR-125a targets the HuR 3’UTR for translational repression. MCF-7 cells were mock transfected or transfected with miR-125a precursor. A, 48 h after transfection, HuR mRNA level was detected by real time RT-PCR. Data were normalized to GAPDH and expressed as relative mRNA levels. B, Cells were co-transfected with miR-125a and pMIR-REPORT β-gal control vector or pMIR-REPORT luciferase vector containing either the HuR 3'UTR (HuR3'UTR-wt) or the HuR 3'UTR with the miR-125a seed sequence deleted (HuR3'UTR-mut). Cell extracts were prepared 48 h after transfection, and luciferase activity measured using the Dual-Light System. The relative luciferase activity after normalization to pMIR-REPORT β-gal control plasmid is shown. The experiments were repeated four times, ** p< 0.01 vs. mock, ## p< 0.01 vs. miR-125a plus HuR3'UTR-wt.

DISCUSSION

In this study, we observed that the expression of miR-125 inversely correlated with HuR level in several different breast carcinoma cell lines. Re-establishing miR-125 expression in these breast cancer cells decreased HuR protein level and cell growth, with miR-125a having a greater effect than miR-125b. Using MCF-7 cells, we further demonstrated that miR-125a inhibited cell growth due to a dramatic suppression of cell proliferation and promotion of apoptosis. Moreover, overexpressing miR-125a in MCF-7 cells inhibited cell migration. Real time PCR and reporter gene assays demonstrated that translation mediated by the 3'UTR of the HuR mRNA was repressed by miR-125a dependent on the presence of a single miR-125a cognate binding site. Importantly, HuR re-expression partially abrogated miR-125a–engendered repression of cell proliferation and migration, demonstrating that HuR is a functional target of miR-125a in MCF-7 breast cancer cells. These results indicate that miR-125a is a potential tumor suppressor in breast cancer, and down-regulation of HuR may represent one important mechanism for this action.

MiRNAs have been implicated in the regulation of cell differentiation, proliferation, death, metabolism and carcinogenesis ^16–19. Aberrant expression of miRNAs has been observed in many types of human cancers ^20–28. Among them, miR-21 and miR-155 (up-regulated), and miR-125, miR-145, and let-7 (down-regulated) are regarded as the most significantly deregulated miRNAs in breast cancer ^23,34,43. The diverse biological functions of these deregulated miRNAs in breast cancer are slowly being unraveled. For example, miR-21 promotes cell transformation and tumor growth, and is associated with advanced clinical stage, lymph node metastasis and poor patient prognosis in human breast cancer ^22,44–46. Let-7 regulates the stem cell-like properties and tumorigenicity of breast tumor-initiating cells by silencing several targets ⁴⁷. Coordinate suppression of ErbB2 and ErbB3 is seen in SKBR3 cells overexpressing miR-125a or miR-125b ³⁵. The current work adds to these studies by showing that re-establishing miR-125a expression in breast cancer cells repressed HuR and cell growth to various degrees, with the most profound repression in MCF-7 cells, the cell line with the lowest miR-125 expression. Decreased MCF-7 cell number was due to suppression of proliferation and promotion of apoptosis by miR-125a. In addition to circumvention of controls on proliferation and inhibition of apoptosis, increased cell migration and invasion are key steps in tumor development, malignancy and metastasis. Our results clearly show that miR-125a overexpression diminishes the migration capacity of MCF-7 cells, as was shown previously for SKBR3 breast cancer cells ³⁵, and that HuR is at least partially mediating these effects.

Overexpression of miR-125 in MDA-MB-231 cells, the breast cancer cell line with the highest miR-125 expression, did not repress HuR or cell growth (Figure 2). Similarly, expression of an antagomir for miR-125a did not increase HuR level (Figure S1). This could be due to the already high expression of both miR-125 and HuR in these cells, or due to a polymorphism in the miR-125 target site in the HuR 3’UTR, a possibility we are currently exploring. In support of this possibility, several studies have identified genetic variants in miRNA target sites that appear to be associated with different cancers ^48,49, including breast cancer ^50,51. Non-tumorigenic MCF10A cells showed only slight repression of HuR and cell growth by miR-125a, and no repression by miR-125b. This is likely due to the already abundant endogenous miR-125a/b present in these cells. T47D breast cancer cells showed an intermediate response to miR-125 overexpression, correlating with its intermediate amount of endogenous miR-125. These results show that when only a single miRNA target site is present, increasing the amount of miRNA past the amount needed for a 1:1 stoichiometry will not increase the effect on that target. It is also consistent with the fact that many different mechanisms regulate HuR level other than miR-125a targeting.

It is known that miRNAs control gene expression generally by targeting mRNAs for either translational repression or destabilization. However, the relationship between miRNAs and their targets is complex. For example, miR-21 acts as an anti-apoptotic factor through targeting different tumor suppressor genes including Bcl-2, programmed cell death 4, tropomyosin 1, and maspin in different cancers ^44,46,52,53. Bcl-2 expression is regulated by additional miRNAs including miR-15, miR-16, miR-21 and miR-34 ^27,46,54,55, illustrating that one miRNA can act on multiple target genes and one target gene can be regulated by multiple miRNAs. We showed that HuR is translationally repressed by miR-125a dependent on the presence of a single miR-125a target site within the 3'UTR of the HuR mRNA. The finding that HuR overexpression only partially restored cell proliferation and migration and failed to prevent apoptosis in miR-125a expressing MCF-7 cells, points to the likelihood of additional targets. ErbB2 and ErbB3 are known targets of miR-125a and miR-125b that regulate cell proliferation, migration and invasion in SKBR3 breast cancer cells ³⁵ and thus are likely also targeted in MCF-7 cells. Our results indicate that miR-125b does not target HuR as well as miR-125a, likely due to the better overall match between the HuR target site and the sequence of miR-125a, outside of the seed sequence, as compared to that of miR-125b (Figure 1A). HuR was recently shown to be targeted for translational repression by miR-519 in cervical, ovarian and colon cancer cell lines, similarly resulting in reduced cell proliferation ⁵⁶. In this case the HuR coding region was preferentially targeted rather than the 3'UTR, showing that miRNA regulation of HuR is likely complex.

This is one of the first reports exploring regulation of an RBP by miRNAs. Our results indicate that miR-125a may act as a tumor suppressor for breast cancer by targeting the RBP HuR for translational repression and thus suppressing cell growth, survival, and migration. mRNA translation and stability are widely known to be regulated by a variety of RBPs and miRNAs. The regulation of RBPs themselves by miRNAs thus is a logical mechanism of regulating translation of several downstream target mRNAs. Our data suggest that restoring miR-125a expression might be a future therapy for breast cancer patients with low miR-125a and HuR up-regulation. Future studies will test if low miR-125a expression is directly associated with tumor development and malignant transformation.

MATERIALS AND METHODS

Cell culture

MCF10A, MCF-7, T47D, SKBR3 and MDA-MB-231 cell lines from American Type Culture Collection (ATCC, Manassas, VA) were cultured under conditions recommended by the manufacturer. The HMT3522-T2 cell line was a generous gift from Dr. Mina Bissell (Lawrence Berkeley National Laboratory) and maintained as recommended ⁵⁷.

Constructs

A 712-nt 3'UTR segment of the HuR mRNA (Genbank accession no. BC003376) was cloned into the pMIR-REPORT luciferase construct (Ambion, Austin, TX). The following primer set was used to generate this 3’UTR: forward, 5’-GGACTAGTCCAACCTGAAGCATGC-3’ and reverse, 5’-CCCAAGCTTGGGGAGTGTTCATAC-3’. We also generated a mutant 3'UTR with the 7-nt miR-125a seed sequence deleted (mutant) using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). All constructs were sequenced.

Transfections

40 nM of miR-125a or miR-125b precursor, anti-miR-125a or negative control anti-miR (Ambion), or transfection reagent alone (mock) was transfected using lipofectamine2000 into cells seeded onto 6 well plates (Invitrogen, Carlsbad, CA). 48 h after transfection, total RNA was extracted for Northern blotting and real-time PCR. 72 h after transfection, protein was extracted for western blotting. For HuR rescue experiments, transfection reagent alone or miR-125a precursor was transfected and 24 hours later, 1 µg pcDNA3.1vector or pcDNA3.1mycHuR was then transfected in MCF-7 cells. 48 h after co-transfection, cells were harvested for cell growth, cell proliferation, apoptosis and migration assays.

Northern blotting

15 mg of total RNA was separated on a 15% denaturing polyacrylamide gel, transferred to a nylon membrane, and hybridized with QuikHyb Hybridization Solution (Stratagene). Membranes were probed with a ³²P-end-labeled oligonucleotide complementary to miR-125a (5’-CACAGGTTAAAGGGTCTCAGGGA-3’), miR-125b (5’-TCACAAGTTAGGGTCTCAGGGA-3’), or to U6 small nuclear RNA (snRNA, 5'-GCAGGGGCCATGCTAATCTTCTCTGTATCG-3') as an internal control. The blots were analyzed with a phosphorimager (Molecular Dynamics, Sunnyvale, CA).

Western blotting

Total cell lysates were resolved by SDS-PAGE, transferred to nitrocellulose membranes, blocked in 5% non-fat milk in TBS/0.05% Tween-20, and probed with antibodies specific for HuR (1;1000, Santa Cruz Biotechnology, Santa Cruz, CA), PARP-1 (1:1000, Cell Signaling Technology, Danvers, MA), or β-actin (1:1000, Santa Cruz). Probing for tubulin was also used as a loading control in place of actin with similar results (not shown),

Real-time PCR

Real-time PCR was performed as described previously ³⁷. Primers used for amplification were as follows: HuR (Genbank accession no. BC003376), forward, 5’-GACATCGGGAGAACGAATTT-3’ and reverse, 5’-TGCTGAACAGGCTTCGTAAC-3’ and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Genbank accession no. XM_006959), forward, 5'-ATGGAAATCCCATCACCATCTT-3' and reverse, 5'-CGCCCCACTTGATTTTGG-3'. Threshold cycles (Ct values) were normalized to GAPDH, and the data were expressed as relative mRNA levels.

Reporter gene assay

Cells of 50% confluence in 24-well plates were transfected using lipofectamine2000. MiR-125a with an equal amounts (200 ng) of pMIR-REPORT luciferase construct containing either wild type or mutant HuR 3'UTR and pMIR-REPORT β-gal Control vector (for normalization) were co-transfected per well. Cell extracts were prepared 48 h after transfection, and the luciferase activity was measured using the Dual-Light System (Applied Biosystems, Bedford, MA)

Cell proliferation assay

Proliferation was assessed by cell counting and by immunofluorescence analysis of Ki-67 proliferation marker. For cell counting, cells of 50% confluence in 24-well plates were transfected as indicated and cell numbers counted 72 h after transfection. For immunofluorescence, cells grown on coverslips were transfected with miR-125a and 72 h after transfection, the cells were fixed and permeabilized. Non-specific binding sites were blocked by pre-incubating cells for 1 hr at room temperature in PBS containing 0.1% Tween 20, 1% BSA and 5% normal goat serum. Cells were then incubated at 4°C overnight in a PBS solution containing mouse anti-Ki-67 antibody (BD Biosciences, San Jose, CA) diluted 1:1,000 in PBS containing 0.5% BSA. Cells were rinsed and incubated for 1 hr at room temperature with goat anti-mouse secondary antibody (1:1,000) labeled with Alexa Fluor 488, rinsed and mounted with a solution containing DAPI and anti-fade agent (Vector laboratories, Inc, Burlingame, CA). Ki-67 staining was analyzed by epifluorescence microscopy using an Olympus BH2-RFCA, inverted microscope (Olympus Optical Co. LTD, Japan). Images were collected using standard filter sets for DAPI and Alexa Fluor 488. Percentage of Ki-67⁺ cells was calculated by counting 500 cells for each condition.

Cell apoptosis assay

Apoptosis was assessed by detecting cleavage of PARP-1 by immunoblotting and with the Annexin V Apoptosis Detection Kit (Santa Cruz or Trevigen, Inc, Gaithersburg, MD). Cells grown on 6 well plates were transfected as indicated and 72 h after transfection, culture media were removed, and cells were washed with cold PBS. Samples were processed for western blotting of PARP-1 as described above or incubated in the dark for 15 minutes at room temperature with Annexin V Incubation Reagent. After washing twice in Binding Buffer at room temperature, cells were viewed immediately by fluorescence microscopy.

In vitro migration assay

For transwell migration assays, 48 h after transfection cells were trypsinized, pelleted, and resuspended in media without serum. After counting, 2 × 10⁵ cells were added to 8-mm chamber insert wells (Greiner Bio One N.A., Monroe, NC). Medium supplemented with 20% serum was used as a chemoattractant in the lower chamber. The cells were incubated for 24 h and cells that did not migrate through the pores were removed by a cotton swab. Cells on the lower surface of the membrane were fixed and stained with 0.4% crystal violet. Membranes were photographed using an Olympus microscope at 200X. Cells were counted in five random fields (each field capturing ~ 6% of total membrane area) and expressed as a percentage of the control cell number.

Statistical Analysis

Data were presented as mean ± SD. Student’s t test was used to compare two groups (p<0.05 was considered significant).

ABBREVIATIONS

ATCC

American Type Culture Collection

BSA

bovine serum albumin

ELAV

embryonic lethal abnormal vision

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

HuR

human antigen R

miRNA

microRNA

miR-125

microRNA 125

nt

nucleotide

RBP

RNA binding protein

PARP-1

poly ADP ribose polymerase 1

PI

propidium iodide

TBS

tris buffered saline

WT

wild type

3'UTR

3' untranslated region