Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 8, 15870

Simultaneous Overactivation of Wnt/β-catenin and TGFβ Signalling by miR-128-3p Confers Chemoresistance-Associated Metastasis in NSCLC

Affiliations

Simultaneous Overactivation of Wnt/β-catenin and TGFβ Signalling by miR-128-3p Confers Chemoresistance-Associated Metastasis in NSCLC

Junchao Cai et al. Nat Commun.

Erratum in

Abstract

Cancer chemoresistance and metastasis are tightly associated features. However, whether they share common molecular mechanisms and thus can be targeted with one common strategy remain unclear in non-small cell lung cancer (NSCLC). Here, we report that high levels of microRNA-128-3p (miR-128-3p) is key to concomitant development of chemoresistance and metastasis in residual NSCLC cells having survived repeated chemotherapy and correlates with chemoresistance, aggressiveness and poor prognosis in NSCLC patients. Mechanistically, miR-128-3p induces mesenchymal and stemness-like properties through downregulating multiple inhibitors of Wnt/β-catenin and TGF-β pathways, leading to their overactivation. Importantly, antagonism of miR-128-3p potently reverses metastasis and chemoresistance of highly malignant NSCLC cells, which could be completely reversed by restoring Wnt/β-catenin and TGF-β activities. Notably, correlations among miR-128-3p levels, activated β-catenin and TGF-β signalling, and pro-epithelial-to-mesenchymal transition/pro-metastatic protein levels are validated in NSCLC patient specimens. These findings suggest that miR-128-3p might be a potential target against both metastasis and chemoresistance in NSCLC.

Conflict of interest statement

The authors declare no competing financial interests.

Figures

Figure 1
Figure 1. miR-128-3p overexpression in NSCLC specimens correlates with poor patient survival, disease progression and chemotherapy failure.
(a) Schematic representation for the establishment of a chemoresistant subcutaneous tumour model. Mice bearing A549-luc xenografts were intraperitoneally injected with the first round of CDDP treatment (CDDP-1st) or control treatment (Ctrl-1st), and cells were isolated from the resultantly remaining tumours, cultured and re-transplanted, followed by a second round of CDDP treatment (CDDP-2nd) or control treatment (Ctrl-2nd) and so ons. (b) Tumour volume of subcutaneous tumours at the indicated rounds of CDDP or Ctrl treatment. (c) Lung tissues obtained from mice bearing NSCLC xenografts at the fourth round of Ctrl and CDDP treatment were histologically examined. Scale bar, 100 μm. (d) Tumour-derived cultured cells from Ctrl-4th and CDDP-4th treatment were subjected to miRNA array analysis, revealing the top 10 upregulated miRNAs in chemoresistant, metastatic A549-luc-CDDP-4th cells. Intensity values representing miRNA expression levels were log10 transformed. (e) Expression of miR-128-3p in human lung cancer clinical specimens using the TCGA microRNA Hiseq expression array data. (f) Absolute real-time PCR using a standard curve of miR-128-3p expression in adjacent normal lung tissues and the indicated NSCLC specimens. Error bars represent mean±s.d. derived from three independent experiments. A two-tailed Student’s t-test was used for statistical analysis (*P<0.05). (g) Kaplan–Meier analysis of OS of a cohort of 153 NSCLC patients at all stages and late stages III–IV. Each subgroup was divided into the low- (below or equal to the median value) and high-miR-128-3p expression groups (above the median value). (h) Kaplan–Meier analysis of OS and PFS of a cohort of 234 NSCLC patients (stage I, n=27; stage II, n=65; stage III, n=52; stage IV, n=90) receiving CDDP-based first-line treatment who were divided into high (>median, n=117) and low (<median, n=117) miR-128-3p expression groups. (i) Expression levels of miR-128-3p in NSCLC patients subgrouped by curative responsiveness to chemotherapy, including PR, SD and PD, according to the RECIST. Ordinate utilizes the logarithmic scale (log10). *P<0.05, **P<0.01. (j) Correlation of miR-128-3p expression level with curative responsiveness of NSCLC patients to chemotherapy. RECIST, Response Evaluation Criteria in Solid Tumours.
Figure 2
Figure 2. miR-128-3p induces a phenotype of EMT and CSC in NSCLC cells.
(a) Phase-contrast images show the morphology of indicated NSCLC cells overexpressing miR-128-3p or control vector. Scale bar, 50 μm. (b) WB analysis was used to examine expression levels of pro-EMT markers (decreased epithelial markers such as E-cadherin and γ-catenin and increased mesenchymal markers such as N-cadherin and Vimentin) in the indicated cells. (c) qRT-PCR was used to measure relative expression of stemness-related genes, including Sox2, Myc, CD133 and ABCG2, in the indicated cells. (d) Representative images and average cell number of invading or migratory cells. Scale bar, 50 μm. (e) Representative images and quantification of non-adherent tumour spheres during three consecutive passages seeded by the indicated cells. Scale bar: 100 μm. (f) Cell viability assay of the indicated cells treated with CDDP, gemcitabine (GEM) and paclitaxel (PAC), respectively, at various concentrations to measure their respective IC50 values. (g) Quantification of colonies counted in anchorage-independent growth ability assay in the presence of CDDP (3 μg ml−1), gemcitabine (GEM, 0.5 μg ml−1), paclitaxel (PAC, 0.5 μg ml−1) or control solvent (Ctrl) treatment. (h) Quantification of TUNEL staining in indicated cells after CDDP (3 μg ml−1) gemcitabine (GEM, 0.5 μg ml−1), and paclitaxel (PAC, 0.5 μg ml−1) treatment, respectively. The numbers of TUNEL-positive cells were counted from 10 random fields and presented as percentages of total cell numbers. (i) WB analysis for proteolytic cleavage of Caspase-3 and PARP in indicated cells after indicated treatment. (j) Quantification of the percentages of TUNEL-positive cells in indicated cells transfected with negative control siRNAs (NC), CTR2 siRNAs or ABCG2 siRNAs after CDDP, gemcitabine (GEM), and paclitaxel (PAC) treatment, respectively. Error bars represent mean±s.d. derived from three independent experiments. A two-tailed Student’s t-test was used for statistical analysis (*P<0.05, **P<0.01).
Figure 3
Figure 3. miR-128-3p overexpression promotes tumorigenesis and spontaneous/systemic metastasis of NSCLC xenografts in vivo.
(a) A549-luc-Vector and A549-luc-miR-128-3p cells of the indicated dosages were, respectively, implanted in the inguinal folds of different nude mice. Representative bioluminescent images of subcutaneous tumour outgrowth are shown. (b) Subcutaneous tumours formed by the indicated cells were dissected and imaged. H&E histologically confirmed tumour cells. Scale bar, 25 μm. (c and d) For the spontaneous metastasis model, bioluminescent images of subcutaneous tumours of the indicated cells, distant metastasis signals (with tumours shielded), and ex vivo organ metastases are shown. (e) For the experimental metastasis model, bioluminescent images of systemic metastases and ex vivo organ metastases including those in the lungs, liver, spleen, kidney, colon, heart, stomach, bones and brain, are shown. (f) Immunostaining for the lung adenocarcinoma marker mucin 1 (MUC1) and lung squamous cell carcinoma marker cytokeratin 5 (CK5), respectively, in spontaneous and experimental lung metastatic lesions developed by subcutaneous inoculation (s.c.) and intravenous injection (i.v.) of the indicated cells. Scale bar, 25 μm. (g) Immunostaining of two key EMT biomarkers, E-cadherin and Vimentin, in primary subcutaneous tumour tissues and lung metastatic lesions. Scale bar, 25 μm. H&E, haematoxylin and eosin.
Figure 4
Figure 4. Inhibiting miR-128-3p suppresses tumour growth and distant metastasis and reverses resistance of highly malignant NSCLC cells to chemotherapeutic drugs.
(a) A549-luc-control-sponge and A549-luc-miR-128-3p-sponge cells were, respectively, implanted in the inguinal folds of separate nude mice. On day 45, tumours were dissected and imaged as shown. A549-luc-CDDP-4th cells were subcutaneously implanted, followed by intravenous administration of control (NC) or miR-128-3p antagomirs (Anta), respectively. On day 30, tumours were dissected and imaged as shown. Bioluminescent images of subcutaneous tumours or/and spontaneous metastasis are shown. (b) A group of mice subcutaneously inoculated with vector-control A549-luc-CDDP4th cells were killed at day 30 after the initial inoculation, and a parallel group of mice were inoculated with A549-luc-CDDP4th cells stably silenced with miR-128-3p by the miR-128-3p sponge and killed for further analysis at day 60 after the initial inoculation. Bioluminescent images of subcutaneous tumours or/and spontaneous metastases are shown (left). Bright-field imaging and fluorescent visualization confirmed efficient inhibition of miR-128-3p by miRNA sponge strategy (right panel). Scale bar, 50 μm. (c) Bioluminescent images of subcutaneous tumours and spontaneous metastasis of A549-luc-miR-128-3p or LL/2-luc-M38 cells in response to CDDP treatment or its combination (Comb) with miR-128-3p antagomir (Anta). Corresponding residual tumours were dissected and imaged as shown. (d and e) In response to the indicated treatments, bioluminescent images of experimental distant metastasis of the indicated cells are shown. Lung metastases were histologically confirmed by H&E staining. Scale bar, 200 μm. (f) OS of mice in the experimental metastasis model receiving the indicated treatments. *P<0.05. (g) Cell viability assays determined the resistance or sensitivity to chemotherapeutic drugs including CDDP, gemcitabine (GEM) and paclitaxel (PAC), which administered, respectively, at a final concentration of 3, 0.5 and 1 μg ml−1, were separately added to the indicated cells pretreated with 100 nM control (NC) or miR-128-3p antagomir (Anta). H&E, haematoxylin and eosin.
Figure 5
Figure 5. miR-128-3p activates Wnt/β-catenin and TGF-β signalling.
(a) Effect of miR-128-3p overexpression on activation of significant pathways by microarray analysis. (b) Relative luciferase activities of the TOPflash/FOPflash, TGF-β, STAT3, Hedgehog and NF-κB signalling reporters in the indicated cells. (c) Effects of silencing the TOP/FOP, TGF-β, STAT3, Hedgehog and NF-κB signalling using the indicated treatments on the migration and tumour sphere growth of miR-128-3p-overexpressing A549 and Calu-3 cells. (d) Immunofluorescent staining shows subcellular SMAD2/3 or β-catenin localization in the indicated cells. Original magnification, × 630. (e) Heat map shows the qRT-PCR results of the upregulated downstream target genes of either β-catenin or TGF-β signalling in the indicated cells. Pseudo-colour scale values were log2 transformed. Error bars represent mean±s.d. derived from three independent experiments. A two-tailed Student’s t-test was used for statistical analysis (*P<0.05, **P<0.01). qRT-PCR, quantitative PCR with reverse transcipode.
Figure 6
Figure 6. β-catenin and TGF-β signalling overactivation mediates miR-128-3p-induced aggressiveness in NSCLC cells.
(a and b) Effects of introducing constitutively active β-catenin (Δβ-catenin) and SMAD3 (CA-SMAD3) mutants, alone or in combination, on cell migration and tumour cell sphere growth in A549-luc-CDDP-4th (A) and A549 (B) cells transfected with NC or miR-128-3p inhibitor (Anti-miR). (c) Cell viability assays determining the resistance or sensitivity of A549-luc-CDDP-4th cells treated as indicated to CDDP. (d and e) Mice bearing tumour xenografts by subcutaneous inoculation (d) or experimental metastasis by intravenous injection (e) of A549-luc-CDDP-4th cells transduced with the constitutively active β-catenin and SMAD3 mutants, alone or in combination, were treated by intravenous administration of miR-128-3p antagomir, or by intraperitoneal injection of with ICG-001 and LY2157299, alone or in combination (Comb). Bioluminescent images of subcutaneous tumours and spontaneous/experimental metastasis are shown. (f) Tumour growth curves of A549-luc-CDDP-4th cells with the indicated pretreatments in mice intravenously administered miR-128-3p antagomir and CDDP in combination. (g) Tumour growth curve of A549-luc-CDDP-4th cells in mice intraperitoneally injected with the indicated inhibitors. (h) Immunostaining of β-catenin, SMAD3, E-cadherin, Vimentin and CD34 in tumour tissues of the indicated A549-luc-CDDP-4th cell xenografts. Scale bar: 25 μm. Error bars represent mean±s.d. derived from three independent experiments. A two-tailed Student’s t-test was used for statistical analysis (*P<0.05).
Figure 7
Figure 7. miR-128-3p targets Axin1, SFRP2, WIF1, SMURF2 and PP1c in NSCLC cells.
(a) Targetscan tool showing schematic representation of putative binding sites for miR-128-3p in 3′-UTRs of Axin1, SFRP2, WIF1, SMURFS and PP1c. (b) WB analysis of the protein levels of Axin1, SFRP2, WIF1, SMURF2 and PP1c in the indicated cells. (c) By immunoprecipitation against Ago1, RIP analysis reveals the interaction of miR-128-3p with the 3′-UTRs of Axin1, SFRP2, WIF1, SMURF2 or PP1c mRNA to form miRNP complexes. IgG immunoprecipitation, as well as the interaction of miR-128-3p with GAPDH and 5s rRNA, were used as negative controls. (d) Luciferase assay of pGL3-Axin1-3′-UTR, pGL3-SFRP2-3′-UTR, pGL3-SMURF2-3′-UTR, pGL3-PP1c-3′-UTR or pGL3-WIF1-3′-UTR reporters in the indicated cells, co-transfected with increasing amounts (20 and 50 nM) of the indicated oligonucleotides. The sequence of the miR-128-3p mutant is shown. (e) Effects of restored expression of Axin1, SFRP2, WIF1, SMURFS and PP1c in miR-128-3p-overexpressing cells on luciferase activities of the TOP/FOP reporter and TGF-β reporter. (f) Effects of restored expression of miR-128-3p target genes on cell migration and self-renewal measured by Transwell migration assay and sphere formation assay, respectively, in the indicated NSCLC cells. Error bars represent mean±s.d. derived from three independent experiments. A two-tailed Student’s t-test was used for statistical analysis (*P<0.05, **P<0.01).
Figure 8
Figure 8. Clinical relevance of miR-128-3p with activation of β-catenin and TGF-β signalling, and expression levels of its target genes and pro-EMT/pro-metastatic markers.
(a) miR-128-3p expression levels correlate with localization of β-catenin and SMAD3, as well as expression of its target genes, including Axin1, SFRP2, WIF1, SMURF2 and PP1c, two key EMT markers (E-cadherin and Vimentin) and the endothelial marker CD34. Two representative cases (Low and High miR-128-3p) are shown. Scale bar, 25 μm. (b) Percentage of specimens showing cytoplasmic/nuclear or membrane β-catenin, nuclear or cytoplasmic SMAD3, low- or high expression of E-cadherin, Vimentin, Axin1, SFRP2, WIF1, SMURF2 or PP1c and CD34 intensity in patient specimens, respectively, with low and high miR-128-3p expression. (c) Model for miR-128-3p-mediated tumorigenesis, metastasis and chemoresistance in NSCLC.

Similar articles

See all similar articles

Cited by 20 PubMed Central articles

See all "Cited by" articles

References

    1. Chaffer C. L. & Weinberg R. A. A perspective on cancer cell metastasis. Science 331, 1559–1564 (2011). - PubMed
    1. Lovly C. M. & Carbone D. P. Lung cancer in 2010: one size does not fit all. Nat. Rev. Clin. Oncol. 8, 68–70 (2011). - PubMed
    1. Gonzalez-Angulo A. M., Morales-Vasquez F. & Hortobagyi G. N. Overview of resistance to systemic therapy in patients with breast cancer. Adv. Exp. Med. Biol. 608, 1–22 (2007). - PubMed
    1. Villanueva M. T. Cell signalling: stuck in the middle of chemoresistance and metastasis. Nat. Rev. Clin. Oncol. 9, 490 (2012). - PubMed
    1. Hu G. et al. . MTDH activation by 8q22 genomic gain promotes chemoresistance and metastasis of poor-prognosis breast cancer. Cancer Cell 15, 9–20 (2009). - PMC - PubMed

Publication types

MeSH terms

Feedback