HMGA2 functions as a competing endogenous RNA to promote lung cancer progression - PubMed (original) (raw)

HMGA2 functions as a competing endogenous RNA to promote lung cancer progression

Madhu S Kumar et al. Nature. 2014.

Retraction in

Abstract

Non-small-cell lung cancer (NSCLC) is the most prevalent histological cancer subtype worldwide. As the majority of patients present with invasive, metastatic disease, it is vital to understand the basis for lung cancer progression. Hmga2 is highly expressed in metastatic lung adenocarcinoma, in which it contributes to cancer progression and metastasis. Here we show that Hmga2 promotes lung cancer progression in mouse and human cells by operating as a competing endogenous RNA (ceRNA) for the let-7 microRNA (miRNA) family. Hmga2 can promote the transformation of lung cancer cells independent of protein-coding function but dependent upon the presence of let-7 sites; this occurs without changes in the levels of let-7 isoforms, suggesting that Hmga2 affects let-7 activity by altering miRNA targeting. These effects are also observed in vivo, where Hmga2 ceRNA activity drives lung cancer growth, invasion and dissemination. Integrated analysis of miRNA target prediction algorithms and metastatic lung cancer gene expression data reveals the TGF-β co-receptor Tgfbr3 (ref. 12) as a putative target of Hmga2 ceRNA function. Tgfbr3 expression is regulated by the Hmga2 ceRNA through differential recruitment to Argonaute 2 (Ago2), and TGF-β signalling driven by Tgfbr3 is important for Hmga2 to promote lung cancer progression. Finally, analysis of NSCLC-patient gene-expression data reveals that HMGA2 and TGFBR3 are coordinately regulated in NSCLC-patient material, a vital corollary to ceRNA function. Taken together, these results suggest that Hmga2 promotes lung carcinogenesis both as a protein-coding gene and as a non-coding RNA; such dual-function regulation of gene-expression networks reflects a novel means by which oncogenes promote disease progression.

PubMed Disclaimer

Figures

Figure 1

Figure 1. Hmga2 promotes lung cancer cell transformation in a protein-coding independent but let-7 site dependent manner

a, Diagram of Hmga2 allelic series: expression constructs containing the entire Hmga2 cDNA (“wt”); the cDNA with all seven let-7 sites in the 3′ UTR mutated (“m7”); the cDNA with the start codon mutated (“ATG wt”); and the cDNA with both the start codon and let-7 sites mutated (“ATG m7”). b, Hmga2 wt and m7 induce Hmga2 expression in non-metastatic lung cancer cells (368T1) and restores expression cells in metastatic lung cancer cells (482N1) depleted for endogenous Hmga2 (shHmga2). Two distinct HMGA2 antibodies are used: one recognizes the N-terminus of HMGA2 (HMGA2-CST) and the other recognizes the central AT-hook region of HMGA2 (HMGA2-Narita). c, Hmga2 RNA is comparably expressed by the wt, m7, ATG wt, and ATG m7 in both 368T1 and 482N1 cells. Hmga2 expression is normalized to Gapdh. 368T1 values are normalized to empty and 482N1 values are normalized to shluc empty. Values are technical triplicates, have been performed independently three times, and represent mean +/− standard deviation (s.d.) with propagated error. d, Hmga2 wt and ATG wt promote substantial anchorage-independent growth in both 368T1 and 482N1 cells. Values are technical triplicates, have been performed independently three times, and represent mean +/− s.d. e, Representative images of soft agar colonies. Magnification is 10X. ***: p<0.0005; **: p<0.005; *: p<0.05.

Figure 2

Figure 2. Hmga2 ceRNA activity enhances lung cancer progression in vivo

a, Hmga2 wt and ATG wt restore lung tumour growth in response to endogenous Hmga2 knockdown. B6129SF1/Tac males were intravenously injected with 482N1 cells expressing either a control shRNA and empty vector (shluc empty) or shHmga2 with the Hmga2 allelic series. Three weeks afterwards, animals were scanned by micro-CT and representative transverse images are shown. The heart is demarcated (labelled ‘H’) and white arrows identify lung tumours. b, Representative histological images of lungs transplanted with 482N1 cells from the series described in a. Magnification is 1X. c, Lung surface tumour counts were taken from animals transplanted with 482N1 cells from the series described in a (n=3 animals per group). Values are technical triplicates represent mean +/− s.e.m. d, Hmga2 wt and ATG wt substantially reduce survival of animals transplanted with 482N1 cells expressing the shRNA targeting Hmga2. Animals were intravenously transplanted with cells as in a. Animals were subsequently aged for survival and a Kaplan-Meier analysis was performed (n=9 animals per group). Median survival was 34 days for shluc empty/shHmga2 wt transplants; 37 days for shHmga2 ATG wt transplants; 43 days for shHmga2 m7 transplants; and 50 days for shHmga2 empty/ATG m7 transplants. Statistical significance was assessed by log-rank tests compared to shHmga2 empty. ****: p<0.00005; ***: p<0.0005; **: p<0.005; *: p<0.05; n.s.: not significant.

Figure 3

Figure 3. Hmga2 ceRNA activity enhances TGF-β signalling through over-expression of Tgfbr3

a, Hmga2 wt and ATG wt substantially induce both Tgfbr3 protein expression and phosphorylation of Smad2 (pSmad2) in both 368T1 and 482N1 cells. b, Hmga2 wt and ATG wt significantly promote expression of Tgfbr3 mRNA in both 368T1 and 482N1 cells. Tgfbr3 expression is normalized to Gapdh. 368T1 values are normalized to empty and 482N1 values are normalized to shluc empty. Values are technical triplicates, have been performed independently three times, and represent mean +/− s.d. with propagated error. c, Hmga2 wt and ATG wt specifically induce expression of a luciferase Tgfbr3 3′ UTR reporter in a let-7 site-dependent manner in both 368T1 and 482N1 cells. Cells were transfected with Renilla constructs of the control siCXCR4 multimer and either the Tgfbr3 wild type or let-7 mutant 3′ UTR reporter. Values are normalized to co-transfected pGL3 plasmid. 368T1 values are normalized to empty and 482N1 values are normalized to shluc empty. Values are technical triplicates, have been performed independently three times, and represent mean +/− s.d. with propagated error. d, Hmga2 wt and ATG wt displace Tgfbr3 from Argonaute-2 (Ago2) based RNA-induced silencing complexes. Lysates from 368T1 and 482N1 cells of the Hmga2 allelic series underwent either control immunoprecipitation (IgG) or immunoprecipitation for Ago2. RNA was purified and qRT-PCR was performed for Hmga2 and Tgfbr3 on both the immunoprecipitated and input RNAs. Values are depicted as the percentage of input RNA, are technical triplicates, have been performed independently twice, and represent mean +/− s.d. e, multiple shRNAs elicit substantial knockdown of Tgfbr3 mRNA in both 368T1 and 482N1 cells. 482N1 cells were infected with control shRNA (shluc) or a set of shRNAs targeting Tgfbr3 (shTgfbr3.1-3.5), while 368T1 wt and ATG wt cells were infected with shluc or shTgfbr3.1, 3.2, 3.4, and 3.5. RNA was purified and qRT-PCR was performed. Tgfbr3 expression is normalized to Gapdh and 368T1 wt and ATG wt and 482N1 values are normalized to shluc. Values are technical triplicates, have been performed independently three times, and represent mean +/− s.d. with propagated error. f, multiple shRNAs induce knockdown of Tgbr3 and suppress TGF-β pathway activity in 368T1 and 482N1 cells. Cells were infected with shRNAs as in e and Western analysis was performed for Tgfbr3, pSmad2 and total Smad2 (Smad2). g, Tgfbr3 depletion reduces anchorage-independent growth of 368T1 wt and ATG wt and 482N1 cells. Cells were infected with the listed shRNAs and plated for anchorage-independent growth and colonies were counted as above. Values are technical triplicates, have been performed independently three times, and represent mean +/− s.d. ***: p<0.0005; **: p<0.005; *: p<0.05.

Figure 4

Figure 4. HMGA2 and TGFBR3 are reciprocally and co-ordinately upregulated in NSCLC patients

a, The Cancer Genome Atlas (TCGA) NSCLC gene expression data set was sorted on HMGA2 expression. The top and bottom quartiles (HMGA2 low and high, respectively) were selected (45 patients per group) and HMGA2 expression was compared by box and whisker plot. b, The TCGA data set was sorted into top and bottom quartiles of HMGA2 expression as in a, and TGFBR3 expression was compared by box and whisker plot. c, The TCGA data set was sorted into top and bottom quartiles of TGFBR3 expression (TGFBR3 low and high, respectively) as in a, and TGFBR3 expression was compared by box and whisker plot. d, The TCGA data set was sorted into top and bottom quartiles of TGFBR3 expression as in c, and HMGA2 expression was compared by box and whisker plot. In all box and whisker plots, values are presented on a log2 scale. Significance was assessed by the Mann-Whitney test with a Bonferroni correction for multiple hypothesis testing. ***: p<0.0005. e, Model for Hmga2 acting as a competing endogenous RNA for Tgfbr3. In nonmetastatic NSCLC, Hmga2 expression is low, leading to suppressed Tgfbr3 expression by let-7. In metastatic NSCLC, Hmga2 expression is elevated, titrating away let-7 from Tgfbr3 and allowing for its over-expression. This titration occurs without changes in let-7 expression, reflecting competition for microRNA occupancy by target transcripts.

Comment in

Similar articles

Cited by

References

    1. Ferlay J, et al. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893–2917. - PubMed
    1. Herbst RS, Heymach JV, Lippman SM. Lung cancer. N Engl J Med. 2008;359:1367–1380. - PMC - PubMed
    1. Di Cello F, et al. HMGA2 participates in transformation in human lung cancer. Mol Cancer Res. 2008;6:743–750. - PMC - PubMed
    1. Meyer B, et al. HMGA2 overexpression in non-small cell lung cancer. Mol Carcinog. 2007;46:503–511. - PubMed
    1. Sarhadi VK, et al. Increased expression of high mobility group A proteins in lung cancer. J Pathol. 2006;209:206–212. - PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources