SEARCH

SEARCH BY CITATION

Keywords:

  • breast cancer;
  • DNA methyltransferase;
  • Kindlins;
  • MicroRNA-200 family

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Kindlin 2, as a focal adhesion protein, controls integrin activation and regulates Wnt signaling in an integrin-binding independent manner. However, the association of Kindlin 2 with cancer-related microRNAs is unknown. Here, we report that Kindlin 2 markedly downregulates the expression of miR-200 family by inducing CpG island hypermethylation. Mechanistically, Kindlin 2 forms a complex with DNMT3A in the cell nucleus and the two proteins co-occupy the promoter of miRNA-200b. Functionally, repression of miR-200b is required for Kindlin 2-induced breast cancer cell invasion and tumor formation. Our data indicate that Kindlin 2 plays a novel role in epigenetic repression of miR-200 family, a mechanism that promotes breast cancer invasion.

Abbreviations
ChIP

chromatin immunoprecipitation

Co-IP

co-immunoprecipitation

DNMTs

DNA methyltransferase

EMT

epithelial to mesenchymal transition

FERMT2/Kindlin 2

Fermitin family homolog 2

ZEB1

Zinc finger E-box binding homeobox 1

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression by inhibiting protein translation or degrading target mRNAs.[1] Hundreds of miRNAs so far have been identified to play key roles in a variety of physiological pathways, including development, cell differentiation, proliferation and cell death.[2, 3] Moreover, aberrant expression of miRNA has been found in most cancers.[4] Some miRNAs are oncogenic and some function as tumor suppressors.[5, 6] Loss of miR-200 family (miR-200a, -200b, -200c, -141 and -429) is believed to be crucial in tumor progression. MiR-200 family significantly inhibits the process of epithelial to mesenchymal transition (EMT), in which it directly targets to transcriptional factors ZEB1/ZEB2 in epithelial cells.[7, 8] Interestingly, ZEB1 also regulates the transcription of miR-200 family by a double negative feedback loop.[9, 10] Although the importance of miR-200 family in tumor progression has been established, the underlying mechanism of miR-200 repression is not completely clear. Recent reports showed that epigenetic regulation, especially DNA methylation, mediates the silencing of miR-200 family in human tumorigenesis.[11, 12] However, the mechanism accounting for the epigenetic regulation of miR-200 remained unknown.

Fermitin family homolog 2 (FERMT2, Kindlin 2) is a member of Fermitin family, originally characterized in the regulation of integrin activation and cell–matrix adhesion. Beside its essential role in the activation of integrin, recent studies revealed that Kindlin 2 is related to tumor progression. Kindlin 2 was found highly expressed in human uterine leiomyomas and human malignant mesothelioma.[13, 14] Functionally Kindlin 2 was reported to promote tumor cell proliferation, adhesion, migration, and invasion.[14-17] Recently, Kindlin 2 was found to induce EMT by activating Wnt signaling.[15] However, association of Kindlin 2 with cancer-related miRNAs in the EMT process remains largely unknown.

Since both Kindlin 2 and miR-200 family function in the regulation of EMT, it is warranted to explore whether Kindlin 2 involves in the regulation of miR-200 family. Here, we report that Kindlin 2 regulates EMT via an epigenetic silencing of miR-200 family through interaction with DNA methyltransferase 3A at the miR-200b promoter.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Cell culture and establishment of stable clines

Human cancer cells were grown in DMEM supplemented with 10% fetal bovine serum (Gibico) and an appropriate amount of penicillin/streptomycin in a 37°C, 5% CO2, humidified incubator. Media were changed twice per week. For the generation of stable cell lines, MDA-MB-231 cells were transfected with plasmids of H1-hsa-mir-200b-CMV-GFP or negative control (GeneChem, Shanghai, China) using Lipofectamine 2000. Mix clones (6–10 independent clones) were obtained after being selecting by hygromycin (Sigma) for 2 week. Other stable cell lines (MCF7-Flag, MCF7-Flag-Kindlin 2, Hs578T-control shRNA and Hs578T-Kindlin 2 shRNA) have been generated as described previously.[15]

Western blotting and Co-immunoprecipitation (Co-IP)

Both Western blotting and Co-IP assays were performed as described previously.[18] The antibodies used were anti-Kindlin-2 (Millipore), anti-DNMT1 (Abcam), anti-DNMT3A (Abcam), anti-DNMT3B (Cell Signaling Technology) anti-HA (Sigma), anti-Flag (Sigma), anti-GST (Sigma), anti-Myc (Santa Cluz).

Purification of fusion proteins and GST pull-down assay

GST-DNMT3A or His-MBP-Kindlin 2 was expressed in E. coli BL21 (Tiangen Biotechnology, Beijing, China), and purified with Glutathione Sepharose 4B beads (Pharmacia Biotech), or MBP-Affinity Matrix (Amylose Resin, New England Biolabs). In a GST pull-down assay, GST-DNMT3A fusion protein was incubated with Glutathione Sepharose 4B beads by rocking at 4°C for 1 hr, and the beads were then washed three times with TEN buffer (20 mM Tris-HCl, pH 7.4, 0.1 mM EDTA, and 100 mM NaCl). His-MBP-Kindlin-2 were then added to the beads and incubated by rocking at 4°C overnight. The beads were washed three times with TENT buffer (0.5% NP40, 20 mM Tris-HCl, pH 7.4, 0.1 mM EDTA and 300 mM NaCl), and then dissolved into 2× SDS loading buffer after centrifugation, and boiled 5 min at 100°C. After centrifuging, the supernatant was extracted and analyzed by Western Blot.

RNA interference (RNAi)

Sequences of RNA interference (RNAi) oligonucleotides were as follows: DNMT1, 5′-AAGCAUGAGCACCGUUCUCCdTdT-3′;DNMT3A,5′-CAUCCACUGUGAAUGAUAAdTdT-3′; DNMT3B,5′-AGAUGACGGAUGCCUAGAGdTdT-3′. Kindlin 2 shRNA has been described previously.[15] All RNAi oligonucleotides were purchased from Shanghai GenePharma Company (Shanghai, China).

Real-time PCR (qPCR)

Total RNA was extracted using Trizol reagent (Invitrogen). cDNA was synthesized using the SuperScript kit (Invitrogen). For the analyses of miRNAs, qRT-PCR was performed using Bulge-Loop™ miRNA qRT-PCR Primer Set (RiboBio, Guangzhou, China). The quantifications of all miRNA were normalized to U6 small nuclear RNA. Primer sequences are available upon request.

Bisulfite sequencing

DNA was extracted and treated with sodium bisulfite as described previously.[19] The primers for sequencing region of each promoter were as follow: miR-200b/200a/429, forward primer, 5′-GGAGTATTGTTTTTTGTAAG-3′ and reverse primer, 5′-ACC TACACAAATAAACAACTC-3′; miR-200c/141, forward primer, 5′-GAGGTTTTG GGTTGAGTTTG-3′ and reverse primer, 5′-AATTACAATCCAAACAAACC-3′.[12] The PCR products were gel extracted (Qiagen) and the purified DNA was cloned into pGEM-T easy vector (Promega). At least 10 separate clones were performed sequencing analyses (Invitrogen).

Chromatin Immunoprecipitation (ChIP) and Re-ChIP assay

A chromatin immunoprecipitation (ChIP) assay was performed as described previously.[19] Briefly, cells were fixed with 1% formaldehyde at 37°C for 10 min. After the cells were lysed on ice for 15 min, lysates were subjected to shearing by sonication. After centrifugation, the soluble chromatin was used for immunoprecipitation with antibodies. Re-ChIP assay was performed using a Re-ChIP-IT kit (Active motif, Catalog No. 53016). The q-ChIP primers for miR-200b were as follows: forward primer, 5′-CACCGCCTCCCATTGTC-3′ and reverse primer, 5′-CACAGGAAGTCAGTTCAGACC-3′. Anti-acetyl-Histone H3 antibody was purchased from Millipore Corporation.

Luciferase reporter assay

Cells were seeded into 24-well plates the day before transfection. 200 ng of RL–control or RL–ZEB1 plasmid (a kind gift from Dr. Gregory J. Goodall) with 50 ng of pFL were transfected per well with Lipofectamine 2000. The reporter activity was measured at 48 hr post transfection using a Dual-luciferase Reporter Assay System (Promega).

Cell invasion assay

The upper wells of the modified Boydean Chamber (BD) were coated Matrigel. After incubating 30 min at 37°C, 0.5 × 105 cells were plated onto the upper well. After 2 days, the top Matrigel was removed and the cells passed through the membrane were stained with crystal violet and counted.

Tumor formation in NOD/SCID mice

Cells were counted and resuspended with 200 μl PBS, then injected subcutaneously into mammary fat pad of 4-week-old female NOD/SCID mice (Center of Experimental Animals, Peking University, Beijing, China), which were executed at 6 weeks after implantation. The mice were maintained according to the Guidelines of Animal Experiments by Peking University. The Ethics Committee of Peking University Health Science Center has approved this study (Permit No: LA2011-73). The procedures followed were in accordance with the ethical standard of the Helsinki Declaration of 1975, as revised in 1983.

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Kindlin 2 promotes tumor formation dependent on the repression of miR-200b

MicroRNA-200 family has been known to inhibit EMT. However, if miR-200 regulates tumor formation remains elusive. To this end, we firstly assessed the effect of miR-200b, a typical member of miR-200 family, on the breast tumor formation. MDA-MB-231 cells overexpressing miR-control or miR-200b were established (Fig. 1a-a′), and then implanted into NOD/SCID mice. After six weeks, tumors were formed in the both sides of mammary fat pads (Fig. 1a-b′). Then, tumor xenografts were dissected and weighted. Statistical analysis showed a significant difference between the two groups (Fig. 1a-c′), indicating that overexpression of miR-200b did inhibit tumor growth of MDA-MB-231 cells. However, no obvious changes were observed in the levels of Kindlin 2 expression between the two groups (Supporting Information Fig. 1a). Next, to compare the tumor growth of MDA-MB-231 cells with or without Kindlin 2 knockdown, control or Kindlin 2 siRNA was injected into the tail vein of mice that were implanted with MDA-MB-231 cells. The efficacy of Kindlin 2 siRNA was determined in vitro (Fig. 1b-a′). As shown in Figure 1b, knockdown of Kindlin 2 inhibited the tumor growth of MDA-MB-231 cells (Figs. 1b-b′ and 1b-c′).

image

Figure 1. Kindlin 2 promotes tumor formation dependent on the downregulation of miR-200b. (a) miR-200b inhibits tumor formation of MDA-MB-231. (a′) MDA-MB-231 stable cells of overexpressing miR-200b or miR-control were established and the level of miR-200b was detected by qPCR. (b′) 1×105 MDA-MB-231 stable cells were separately implanted into mammary fat pads bilaterally. After 6 weeks, the tumors were dissected and shown (3 mice per group, representative from two independent experiments). (c′) The average weight of tumors was quantified. Data are mean ± S.D. (n = 6); * indicates p < 0.05. (b) Knockdown of Kindlin 2 inhibits tumor formation of MDA-MB-231. (a′) The efficacy of Kindlin 2-siRNA was detected by Western Blotting (WB). (b′) 1 × 105 MDA-MB-231 cells were implanted into lateral fat pad of mice. After 2 weeks, control or Kindlin 2 siRNA was injected into the tail vein of mice for 4 weeks (6 mice per group). The tumors were dissected and shown. (c′) The average weight of tumors was quantified. Data are mean ± S.D. (n = 6); * indicates p < 0.05. (c) Overexpression of Kindlin 2 promotes tumor formation. (a′) The level of Kindlin 2 was detected by WB in MCF7–Flag-Kindlin 2 versus MCF7–Flag cells. (b′) The two stable cells were implanted into mammary fat pads of mice bilaterally. After 6 weeks, the tumors were dissected (6 mice per group). A representative mouse in tumor formation experiment was shown. (c′) The average weight of tumors was shown. Data are mean ± S.D. (n = 6); * indicates p < 0.05. (d′, e′) Total RNA was extracted from the tumor xenograft, and then RT-qPCR was performed to detect the levels of EMT markers and miR-200 family. The levels of miR-10a and miR-10b were detected as a control. Data are mean ± S.D. (n = 3). (d) Overexpression of miR-200b inhibits tumor formation of MCF7–Flag-Kindlin 2 stable cells. (a′) 1×104 MCF7–Flag-Kindlin 2 stable cells were implanted into fat pad of mice laterally. After 2 weeks, Agomir-control or Agomir-200b was injected into tail vein of mice every three days for 4 weeks. The tumors were dissected and shown (3 mice per group, from two independent experiments). (b′) The average weight of tumors was quantified. Data are mean ± S.D. (n = 6); * indicates p < 0.05. (e) Agomir-control or -200b was separately transfected into MCF7–Flag or MCF7–Flag-Kindlin 2 cells. After 24 hr, 500 cells were seeded in soft agar and after two weeks cell colonies (larger than 50 μm in diameter) were counted. Data are mean ± S.D. from three independent assays,* indicates p < 0.05.

Download figure to PowerPoint

Since the regulation of Kindlin 2 on EMT has been identified in breast cancer cell lines,[15] it raises a question as to whether Kindlin 2 influences the breast tumor formation through activating EMT process. To this end, MCF7 stable cells overexpressing Flag or Flag-Kindlin 2 were established (Fig. 1c-a′) and the levels of a panel of EMT markers were determined in MCF7–Flag-Kindlin 2 cells versus MCF7–Flag cells (Supporting Information Fig. 1b). After these two stable cells were implanted into mammary fat pads of NOD/SCID mice for 6 weeks, tumors were formed in the side implanted with MCF7–Flag-Kindlin 2 stable cells, whereas no tumors were found in the control side (Figs. 1c-b′ and 1c-c′). Further, total RNA was extracted from the tumor xenograft of MCF7–Flag-Kindlin 2 and the levels of key EMT markers were determined. Results showed that EMT process was obviously activated in Kindlin 2-induced tumors (Fig. 1c-d′). Due to the role of miR-200 family in EMT process, the levels of miR-200 family were also detected in Kindlin 2-induced tumors. As shown in Figure 1c, the expression of miRNA-200 family was significantly repressed in Kindlin 2-induced tumors (Fig. 1c-e′). However, it is unclear whether the downregulation of miR-200 family is required for Kindlin 2-promoted tumor formation. To this end, Agomir-control or Agomir-200b was injected into the tail vein of NOD/SCID mice that were implanted with MCF7–Flag-Kindlin 2 stable cells. As shown in Figure 1d, the size and weight of the tumors were significantly reduced in the Agomir-200b-treated mice, compared with the mice of Agomir-control (Figs. 1d-a′ and 1d-b′). Moreover, the levels of EMT markers and miR-200 family were also measured in the tumors with or without Agomir-200b treatment (Supporting Information Figs. 1c and d). These data indicated that downregulation of miR-200b is required for Kindlin 2-induced tumor formation in NOD/SCID mice. To further support this in vivo data, colony formation assay was performed in vitro. As shown in Figure 1e, overexpression of Kindlin 2 significantly promoted the anchorage-independent growth of MCF7 cells. However, upon the treatment of Agomir-200b, the numbers of colony formed by MCF7–Flag-Kindlin 2 cells were markedly decreased, again indicating that Kindlin 2 promotes cell growth dependent on the repression of miR-200b. Taken together, we established the role of miR-200 repression in the Kindlin 2-induced tumor formation.

Kindlin 2 represses the expression of miR-200 family

To uncover the interplay between Kindlin 2 and miR-200 family in the regulation of EMT process, as a start we analyzed the levels of Kindlin 2 and miR-200 family in a variety of human breast cancer cell lines, including a panel of cell lines with epithelial phenotype that highly express E-cadherin and a panel of cell lines with mesenchymal phenotype that highly express Vimentin and ZEB1. Our results showed that the epithelial cells expressed low levels of Kindlin 2 and high levels of miR-200 family, whereas an opposite expression profile for Kindlin 2 and miR-200 family was observed in mesenchymal cells (Fig. 2a). These data suggested that a regulatory mechanism may exist between Kindlin 2 and miR-200 in breast cancer cells. To examine the potential regulatory role of Kindlin 2 on miR-200 family, Kindlin 2-overexpressed or depleted stable cells were established (Figs. 1c-a′ and 2c-a′). As shown in Figures 2b and 2c, overexpression of Kindlin 2 downregulated the miR-200 family, whereas depletion of endogenous Kindlin 2 led to the increase of miR-200 family. Importantly, when Kindlin 2 was re-expressed in Kindlin 2-depleted cells, the increase of miR-200 family induced by Kindlin 2 knockdown was partially reversed (Fig. 2d). These results demonstrated that Kindlin 2 negatively regulates the levels of miR-200 family in breast cancer cells.

image

Figure 2. Kindlin 2 downregulates the expression of miR-200 family. (a) Total RNA was extracted from the indicated cell lines, and then RT-qPCR assays were performed to detect the levels of E-cadherin (a′), Vimentin (b′), ZEB1 (c′), Kindlin 2 (d′) and miR-200 family (e′). Data are mean ± S.D. (b) The levels of miR-200 family were detected in MCF7–Flag or MCF7–Flag-Kindlin 2 cells by qPCR. Data are mean ± S.D. (c) Kindlin 2 was knocked down in Hs578T cells using shRNA, and the level of Kindlin 2 was detected by WB (a′). Then the levels of miR-200 family were detected by qPCR. Data are mean ± S.D. (b′). (d) Flag or Flag-Kindlin 2 was transfected into Hs578T-Kindlin 2 shRNA stable cells, the level of Kindlin 2 was detected by WB (a′). Then total RNA was extracted and RT-qPCR was performed to detect the levels of miR-200 family. Data are mean ± S.D. (b′).

Download figure to PowerPoint

Kindlin 2 induces the CpG methylation of miR-200 family

Aberrant epigenetic modifications are involved in the downregulation of tumor suppressors in tumor development.[20] To scrutinize whether Kindlin 2 is involved in the epigenetic regulation of miR-200 family, the methylation of miR-200 family was examined upon Kindlin 2 overexpression or knockdown. MiR-200 family is composed of two polycistronic clusters: miR-200b/200a/429 and miR-200c/141 (Figs. 3a and 3b). By bisulfite sequencing analyses, the increase of CpG methylation rate in miR-200b/200a/429 promoter was induced by Kindlin 2 (Fig. 3c). Conversely, knockdown of Kindlin 2 led to partial demethylation of miR-200b/200a/429 promoter (Fig. 3d). Similarly, the effect of Kindlin 2 on CpG methylation of miR-200c/141 was also explored. As shown in Figure 3e, the CpG island of miR-200c/141 was markedly methylated in MCF7–Flag-Kindlin 2 cells versus MCF7 control cells (Fig. 3e), whereas knockdown of Kindlin 2 induced the demethylation of CpG island of miR-200c/141 promoter (Fig. 3f). Collectively, these data indicated that Kindlin 2 promotes CpG methylation of miR-200 family promoters, which in turn results in silencing of miR-200 family.

image

Figure 3. Kindlin 2 induces the CpG methylation of miR-200 family. (a, b) Schematic diagrams of the fragments for bisulfite sequencing in the CpG islands of miR-200b/200a/429 and miR-200c/141. (c–f) Bisulfite sequencing analyses of DNA methylation in the miR-200b/200a/429 and miR-200c/141 promoters were performed in the indicated stable cells. Methylated CG (filled circles) and unmethylated CG (open circles) are represented. The methylation rate in each region (as a percentage) is shown under each panel. Ten separate subclones per group were sequenced.

Download figure to PowerPoint

Kindlin 2 interacts with DNMT3A on the promoter of miR-200b

In mammals, three DNMTs (DNA methyltransferase) including DNMT1, DNMT3A and DNMT3B are known to play important roles in the generation and maintenance of DNA methylation. DNMT3A and DNMT3B, also known as de novo methyltransferases, are in charge of the generation of new DNA methylations, whereas DNMT1 is required to maintain the existing DNA methylations. Given the roles of DNMTs in DNA methylation,[21, 22] it is of interest to explore whether Kindlin 2 may increase the expression of DNMTs or enhance their enzymatic activity. To this end, we examined the expression of DNMTs as well as the total activity of DNMTs in MCF7–Flag-Kindlin 2 cells versus MCF7 control cells. Results showed that Kindlin 2 not only upregulated the expression of DNMTs, but also enhanced the methyltransferase activity of DNMTs (Figs. 4a and 4b). Further, to clarify whether DNMTs were involved in the regulation of Kindlin 2 on miR-200 family, DNMTs were knocked down in Kindlin 2-overexpressed stable cells and the level of miR-200b was then determined. As shown in Figure 4c, the increases of miR-200b expression were observed in both DNMT1 and DNMT3A siRNA-treated cells as well as in 5-aza-2′-deoxycytidine (5-aza-CdR, an inhibitor of DNA methyltransferase)-treated cells (Fig. 4c). Interestingly, DNMT3A apparently played a more important role than DNMT1 in the regulation of miR-200b. Moreover, the efficacy of DNMT siRNAs was determined by qPCR (Fig. 4d). These results revealed that both DNMT1 and DNMT3A were involved in the regulation of Kindlin 2 on miR-200b expression.

image

Figure 4. Kindlin 2 interacts with DNMT3A on miR-200b promoter. (a) The levels of DNMT1, DNMT3A and DNMT3B were detected by WB in indicated stable cells. (b) The nuclear protein was extracted from indicated stable cells, and the total activity of DNMT was detected using EpiQuik™ DNA Methyltransferase Activity/Inhibition Assay Kit. Data are mean ± S.D. from three independent assays; * indicates p < 0.05. (c) The levels of miR-200b were detected by qPCR in indicated siRNAs or 5-aza-CdR (5 μm, 72 hr)-treated stable cells. Data are mean ± S.D. from three transfection experiments; * indicates p < 0.05. (d) The efficacy of DNMT siRNAs was detected by qPCR. Data are mean ± S.D. * indicates p < 0.05. (e) Flag or Flag-Kindlin 2 was transfected into HEK293T cells. After 48 hr, lysates were extracted for ChIP assay using antibodies as indicated. Anti-acetyl-Histone H3 (AcH3) antibody was used as a positive control. Anti-Myc antibody was used as a negative control. Then qPCR was used to quantify the ChIP-enriched DNAs. Data are mean ± S.D. * indicates p < 0.05. (f) Flag-Kindlin 2 and Myc-DNMT3A were co-transfected into HEK293T cells. ChIP and Re-ChIP assays were performed using anti-Myc and anti-Flag antibodies. Anti-AcH3 antibody or IgG was separately used as a positive or negative control both in the 1st ChIP and 2nd ChIP. No-Antibody (no Ab) was used as a negative control in the 2nd ChIP. Anti-Myc antibody was used as a positive control in the 2nd ChIP. Data are mean ± S.D. (g, h) HEK 293T cells were transfected with Flag-Kindlin 2 and Myc-DNMT3A for 48 hr, later protein was extracted for Co-IP using Flag or Myc antibody, followed by WB analysis using indicated antibodies. (i) Lysates from MCF7–Flag-Kindlin 2 stable cells were extracted, and then anti- Flag antibodies were used for Co-IP. (j) Lysates from Hs578T cells were extracted, and then anti-DNMT3A antibodies were used for Co-IP. (k) Fusion protein His-MBP-Kindlin 2 and GST-DNMT3A were expressed and purified in vitro. GST-DNMT3A was incubated with or without His-MBP-Kindlin 2 for GST pull-down assay. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

To uncover how Kindlin 2 cooperates with DNMT1 or DNMT3A to inhibit the expression of miR-200b, the occupancy of DNMT1 or DNMT3A on miR-200b promoter was analyzed using ChIP assay in cells with or without Kindlin 2 overexpression. Results showed that the occupancies of both Kindlin 2 and DNMT3A at the miR-200b promoter were increased in Kindlin 2-overexpressing cells, compared with the control cells (Fig. 4e and Supporting Information Fig. 2). The antibody of anti-acetyl-Histone H3 (AcH3) or anti-Myc was separately used as a positive control or negative control. To further investigate whether Kindlin 2 and DNMT3A bind together at the miR-200b promoter, a ChIP-reChIP assay was performed in HEK293T cells co-expressing Flag-Kindlin 2 and Myc-DNMT3A. Results demonstrated that Kindlin 2 and DNMT3A co-occupied the promoter of miR-200b (Fig. 4f), suggesting that a physical association between Kindlin 2 and DNMT3A may exist. To this end, co-immunoprecipitation (Co-IP) assays were performed to identify whether Kindlin 2 interacts with DNMT3A. Results showed that Kindlin 2 did form a complex with DNMT3A in HEK293T, MCF-7–Flag-Kindlin-2 and Hs578T cells (Figs 4g4j). Further, both full-length GST-DNMT3A and His-MBP-Kindlin-2 were expressed and purified from E.coli, and a direct interaction between Kindlin-2 and DNMT3A was identified in vitro by a GST pull-down assay (Fig. 4k). These results demonstrated that Kindlin 2 interacts with DNMT3A both in vivo and in vitro.

To determine where the two molecules are complexed together, immunofluorescence staining was performed in Hs578T cells and results clearly showed that Kindlin 2 and DNMT3A co-localized in the nucleus (Fig. 5a). Moreover, both cytoplasmic and nuclear extracts were prepared from MCF-7–Flag-Kindlin-2 stable cells for Co-IP assays using an anti-Flag antibody. Our results showed that only nuclear DNMT3A could be co-immunoprecipitated by Kindlin-2 (Fig. 5b), again demonstrating that the complex of Kindlin 2 and DNMT3A resided in the nucleus.

image

Figure 5. Kindlin 2 and DNMT3A co-localized in the nucleus. (a) Immunofluorescence (IF) staining was performed in Hs578T cells using anti-Kindlin-2 antibody (Alexa Flour 488, green) and anti-DNMT3A antibody (Alexa Flour 568, red). Nuclei was stained by DAPI (blue). Images were captured with a confocal microscopy. Bars (upper), 50 μm; bars (lower), 25 μm. (b) Both cytoplasmic and nuclear lysates were extracted from MCF-7–Flag-Kindlin-2 stable cells for performing Co-IP assays using Flag antibody. Both lysates were evaluated by WB. The absence of α-tubulin demonstrates that the fraction was from the nucleus. The absence of YY1 indicates that the fraction was from the cytoplasm.

Download figure to PowerPoint

MiR-200b inhibits the Kindlin 2-promoted breast cancer cell invasion

The miR-200b is a key regulator of EMT process by directly targeting ZEB1,[10, 23] a transcriptional repressor of E-cadherin. Thus, loss of miR-200b expression leads to the accumulation of ZEB1, in turn triggers EMT. As shown in Figure 6a, microRNA precursors (Pre-miRs) of miR-200b were cotransfected with RL-control or RL-ZEB1 reporter construct into the mesenchymal-like MDA-MB-231 cells, and a marked repressive effect on RL-ZEB1 was observed. Further, qPCR was performed to detect the expression of ZEB1. Data indicated that ZEB1 was significantly inhibited in pre-miR-200b-treated cells (Fig. 6b). Moreover, a decrease of invasive capacity was observed in pre-miR-200b-treated cells (Fig. 6c). These results demonstrated a critical role of miR-200b in regulating EMT and breast cancer cell invasion, which confirmed the previously published observations.[24]

image

Figure 6. MiR-200b inhibits the Kindlin 2-promoted breast cancer cell invasion. (a) MDA-MB-231 were transfected with pre-miR-control or pre-miR-200b plus the RL-control or RL-ZEB1 for 48 hr. (b) The level of ZEB1 was detected by qPCR in pre-miR-control or pre-miR-200b-treated MDA-MB-231. (c) Pre-miR-control or pre-miR-200b was transfected into MDA-MB-231. Invasion assays were then performed. (d–f) Pre-miR-control or pre-miR-200b was transfected into the indicated stable cells for 48 hr. Then the activity of RL-ZEB1 (d) and the expression of ZEB1 (e) were detected, and invasion assays (f) were performed. (g–i) Control or DNMT3A siRNA was transfected into the indicated stable cells for 48 hr. Data are mean ± S.D. from three independent experiments; * indicates p < 0.05.

Download figure to PowerPoint

Given that Kindlin 2 could induce the epigenetic repression of miR-200 family, an important question may be raised concerning whether loss of miR-200b is required for the regulation of Kindlin 2 on EMT. To this end, pre-miR-200b was transfected into the Kindlin 2-overexpressed stable cells. Luciferase reporter assay and qPCR were performed to examine the reporter activity of RL-ZEB1 as well as the mRNA level of ZEB1. Results showed that forced miR-200b expression counteracted the upregulation of ZEB1 induced by overexpressing Kindlin 2 (Figs. 6d and 6e), and abolished the invasive ability of Kindlin 2-overexpressed stable cells (Fig. 6f). Further, we found that the decrease of ZEB1 expression as well as the invasive ability of Kindlin 2-overexpressed stable cells can be achieved by knocking down DNMT3A using siRNA (Figs. 6g6i). These results indicated that disruption of the interaction between Kindlin 2 and DNMT3A would restore the expression of tumor suppressor miR-200b, and thus inhibit ZEB1 expression and activation. The above findings established a molecular link that holds Kindlin 2 and miR-200b together by demonstrating that DNMT3A functions in between.

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Kindlin 2 has been reported to associate with integrin, migfilin, integrin-linked kinase (ILK) and β-catenin for regulating cell adhesion, migration and invasion.[15, 25-27] Here, we reported that DNMT3A is a newly identified molecule of potential importance in associating with Kindlin 2 (Figs. 4g4k). This complex of Kindlin 2 and DNMT3A co-occupied the miR-200b promoter (Fig. 4f) and resulted in the repression of miR-200b expression. Thus, knocking down endogenous DNMT3A in Kindlin 2 overexpressing cells rescued the expression of miR-200b (Fig. 4c). However, knockdown of DNMT3A or treatment of 5-aza-CdR only partially reversed the expression of miR-200b. In this scenario, we do not exclude other players that may also involve in the regulation on miR-200b expression. In fact, DNA methylation and histone modification often collaborate to lead to tumor suppressor gene silencing. For example, some histone methyltransferases, such as EZH2 and SETDB1, interact with DNMT3A to form a complex, co-occupying the CpG islands of tumor suppressors and regulating their expression.[28, 29] Therefore, EZH2, SETDB1 or other histone methyltransferases may be recruited by complex of Kindlin 2 and DNMT3A at the promoter of miR-200b, together silencing the expression of miR-200b. These possibilities deserve further investigation. Besides DNMT3A, knocking down endogenous DNMT1 in Kindlin 2 overexpressing cells also partially rescued miR-200b expression (Fig. 4c). However, the absence of DNMT1 binding on miR-200b promoter (Fig. 4e) suggested that DNMT1 may indirectly mediate the regulation of Kindlin 2 on miR-200b through other factors.

Since the mRNA level of DNMT3A was constant in Kindlin 2-overexpressed cells compared with the control cells (data not shown), the interaction between Kindlin 2 and DNMT3A may be beneficial for maintaining the stability of DNMT3A. It is reported that SUMOylation, a post-translational modification, mediates the stability of DNMT3.[30] Cbx4 (chromobox 4), a member of Polycomb group proteins, functions as an E3 ligase to promote the SUMOylation of DNMT3A.[31] Our data from a PCR array indicated that Cbx4 was downregulated by approximately 2.6-fold in Kindlin 2-overexpressed stable cells, compared with the control cells (data not shown). This suggests that Kindlin 2 may be involved in the Cbx4-mediated SUMOylation of DNMT3A and overexpression of Kindlin 2 may inhibit the degradation of DNMT3A dependent on the SUMOylation, which leaves room for future investigation.

Previously findings have revealed the CpG methylation of both clusters of miR-200 family. Davalos et al. found that MCF7 cells were unmethylated in miR-200b/200a/429 promoter,[12] whereas Wee et al. indicated that MCF7 cells were methylated in miR-200b promoter.[29] In this study, bisulfite sequencing analyses showed that the CpG methylation rate of miR-200b is 56 % in MCF7 cells (Fig. 3c). To clarify this ambiguity, methylation-specific PCR analyses were performed and both methylated and unmethylated bands were observed in MCF7 cells (Supporting Information Fig. 3). When Kindlin 2 was overexpressed in MCF7 cells, the unmethylated band was totally disappeared, and only the methylated band was observed. Considering the inconsistency in CpG methylation of miR-200b in MCF7, it is possible that different CpG dinucleotides were detected. Thus, multiple CpG regions in miR-200b promoter remain to be investigated in future research. In addition, we observed that MCF7 is totally unmethylated in miR-200c/141 promoter, whereas Hs578T is highly methylated in the same region, which is consistent with the observation of Vrba et al.[32]

Since knockdown of Kindlin 2 could not markedly remove the methylation of miR-200b/200a/429 promoter in Hs578T cells (Fig. 3d), we considered the following possible reasons. Previous publication has indicated that some miRNAs have multiple promoters.[33, 34] For example, two independent promoters (P1 and P2) have been confirmed in the miR-200b/200a/429 cluster.[29] As the significant re-expression of miR-200b/200a/429 was observed in Kindlin 2 shRNA-treated Hs578T cells compared with control cells (Fig. 2c); therefore, we do not exclude other promoters rather than the one we detected may involve in the epigenetic regulation of miR-200b/200a/429 in Hs578T cells. In addition, other mesenchymal phenotypes of breast cancer cell lines, such as MDA-MB-231 and BT549 might be used for future investigation to uncover the methylation of miR-200b/200a/429 promoter.

We and other researchers have identified that Kindlin 2 plays a positive role in promoting the invasion of breast cancer cells.[17] However, an opposite role of Kindlin 2 was indicated in mesenchymal cancer cell invasion.[35] These evidences suggested that the biological function of Kindlin 2 was complex in different types of tumors, which may provide an explanation for the different expression levels of Kindlin 2 in different tumors.[13, 14, 17]

It is reported that miR-200b inhibits the invasion of tumor cells by directly targeting transcriptional factor ZEB1.[10] Consistent with this, our results demonstrated that Kindlin 2 induces the downregulation of miR-200b (Fig. 2b), which leads to the accumulation of ZEB1, in turn triggering EMT and promoting cell invasion (Fig. 6). In addition to ZEB1, miR-200 also inhibits tumor cell invasion by targeting VEGFR1.[36] Our data from a PCR array indicated that VEGFR1 was increased by approximately 10-fold in Kindlin 2 overexpressing cells, compared with the control cells (data not shown). This suggests that the regulation of Kindlin 2 on the targets of miR-200b may be not only restricted to ZEB1, but also includes VEGFR1.

In conclusion, this study identified that Kindlin 2 is a negative regulator for miR-200 family both in vitro and in vivo. We demonstrated that Kindlin 2 interacts with DNMT3A in the nucleus, and Kindlin 2 and DNMT3A co-occupies the promoter of miR-200b that leads to the epigenetic repression of miR-200b. Finally, disruption of the interaction between Kindlin 2 and DNMT3A may hold promise for blocking breast cancer invasion.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This work was supported by grants from the Ministry of Science and Technology of China 2010CB912203 and 2010CB529402, the National Natural Science Foundation of China 30830048, 81230051, 31170711 and 81101495, the 111 Project of the Ministry of Education, Beijing Natural Science Foundation 7120002 and a Peking University grant BMU20120314 and Leading Academic Discipline Project of Beijing Education Bureau to H.Z.

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
ijc28151-sup-0001-suppinfoFigure01.tif8711KSFig 1 (A) Protein was extracted from the tumors induced by MDA-MB-231-miR-control or MDA-MB-231-miR-200b cells for Western blotting assay using the indicated antibodies (3 tumor bodies / group). (B) Total RNA was extracted from MCF7–Flag and MCF7–Flag-Kindlin-2 stable cells, and then qPCR was performed. (C) 1×104 MCF7–Flag-Kindlin 2 stable cells were implanted into lateral fat pad of mice. After 2 weeks, Agomir-control or Agomir-200b was injected into tail vein of mice for 4 weeks. The tumors were dissected and total RNA was extracted for qPCR analysis of EMT markers and miRNAs (3 tumor bodies / group). Data are mean ± S. D.
ijc28151-sup-0002-suppinfoFigure02.tif1898KSFig 2 Lysates from MCF7–Flag and MCF7–Flag-Kindlin-2 stable cells were used for chromatin immunoprecipitation (ChIP) with antibodies as indicated. Anti-Myc antibody or IgG was as a negative control. qPCR was used to quantify the ChIP-enriched DNAs. Data are mean ± S. D. * indicates p<0.05.
ijc28151-sup-0003-suppinfoFigure03.tif2046KSFig 3 DNA methylation status of miR-200b promoter in MCF7–Flag and MCF7–Flag-Kindlin-2 stable cells was detected by MS-PCR. M indicates methylated DNA, and U indicates unmethylated DNA. H2O was as a negative control. Normal human DNA was treated with methylase SssI as a positive control.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.