Abstract Autophagy is a bulk degradation mechanism mainly involved in cell survival responses to a variety of stress stimuli. Its role in tumorigenesis is still ambiguous. Low levels of basal autophagy are crucial in order to sustain cellular homeostasis thereby also contributing to tumor prevention, whereas autophagy induction upon a variety of anti-cancer treatments is often involved in resistance mechanism to therapy. Src tyrosine kinase inhibitors (TKIs) inhibit cell migration and invasion in non-small cell lung cancer (NSCLC) cells. In clinical trials, however, they show modest activity in combination with chemotherapeutic agents. We hypothesized that cytoprotective autophagy plays a role in resistance mechanisms against Src TKIs. We found a marked induction of autophagic activity in NSCLC cell lines (A549, H460, H1299) treated with two different Src TKIs (saracatinib, dasatinib) as measured by increased LC3-I to -II conversion, increased GFP-LC3 dot formation, and decreased p62/SQSTM1 protein expression. Most importantly, the addition of pharmacological autophagy inhibitors (chloroquine, bafilomycin) in combination with Src TKI inhibitors resulted in cell death as compared to only decreased migration by Src TKI treatment alone. Next, we identified ULK1 as a key player in Src TKI-induced autophagy using a set of lentiviral vectors expressing shRNAs targeting ATG genes. Src TKIs significantly induced ULK1 mRNA and protein expression and knocking down ULK1 significantly attenuated Src TKI-induced autophagy. Furthermore, to exclude Src inhibition-independent effects of saracatinib and dasatinib on ULK1 expression, we targeted Src by RNAi and found similar induction of ULK1 as seen with the Src TKI treatment. We further investigated if micro(mi)RNAs are involved in ULK1 regulation during Src inhibitor treatment. Using miRNA target identification software and 3′-UTR luciferase reporter assays we identified miR-106a as ULK1-targeting miRNA. Incubating NSCLC cells with Src inhibitors resulted in a dose-dependent decrease of miR-106a paralleled by an increase in ULK1 expression. Moreover, ectopic expression of miR-106a and anti-miR-106a in NSCLC cell lines caused decreased and increased ULK1 expression upon Src inhibition, respectively. Importantly, similar to the pharmacological autophagy inhibition and or knocking down ULK1, expression of miR-106a enhanced the cytotoxicity of Src TKIs. Lastly, we found significantly higher miR-106a levels in human lung adenocarcinoma compared to matched normal lung tissue (n=23), whereas ULK1 mRNA expression levels were significantly lower (p<0.0002) in tumor tissue than in normal lung tissue. In conclusion, Src inhibitors induce protective autophagy in NSCLC cells that is mediated via miR-106 and its target ULK1. Combining Src and autophagy inhibitors may represent attractive treatment option for certain NSCLC. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2263. doi:1538-7445.AM2012-2263
Lineage commitment of haematopoietic cells is tightly regulated through an intricate network of molecular pathways and cascades of genes including microRNAs (miRNAs). Moreover, recent studies show that abnormal expression of miRNAs directly contributes to haematological malignancies, such as acute myeloid leukaemia (AML), chronic myeloid leukaemia (CML) and myelodysplastic syndromes (MDS) (Fabbri et al, 2008; Starczynowski et al, 2011). miRNAs are 19–25 nucleotide long, non-protein coding RNAs that regulate gene expression by binding to partially complementary sequences in the 3′-untranslated regions (3′-UTR) of their target mRNAs. Acute promyelocytic leukaemia (APL) is characterized by the t(15;17) translocation resulting in the oncogenic PML-RARA fusion gene. Pharmacologial doses of all-trans retinoic acid (ATRA) are used to directly target PML-RARA for degradation and, in combination with chemotherapy, have been successfully used in the treatment of APL patients (de The & Chen, 2010). A discrete number of miRNAs modulated upon ATRA treatment of APL cells in vitro have been published (Garzon et al, 2007; Saumet et al, 2009). Among them MIR29B, a key miRNA induced during neutrophil differentiation of APL cells. The tumour suppressor MIR29B plays a role in targeting leukaemic oncogenes such as DNA methyltransferase 3A/B (DNMT3A/DNMT3B) in AML cells (Garzon et al, 2009). In addition, high MIR29B levels were associated with a response to the demethylating agent decitabine in AML (Blum et al, 2010). The present study aimed to confirm a role for MIR29B in the neutrophil differentiation of APL cells and to analyse the transcriptional regulation of MIR29B during this process. We first measured MIR29B expression during ATRA-induced neutrophil differentiation of NB4 and HT93 APL cell lines. MIR29B was clearly induced in both APL cell lines (Fig 1A). Furthermore, MIR29B expression was markedly induced in an APL patient undergoing ATRA therapy (Fig 1B). We then measured MIR29B in blast cells of primary AML patients of different French-American-British (FAB) subtypes (M0-M4) as well as in mature neutrophils from healthy donors. We observed significantly lower MIR29B expression levels in AML patient samples as compared to normal granulocytes from healthy individuals (Fig 1C). Lastly, to further define the role of MIR29B in neutrophil differentiation, we inhibited its expression by antisense MIR29B (anti-MIR29B) in HT93 APL cells. Neutrophil differentiation in anti-MIR29B expressing HT93 cells was significantly reduced compared to control cells as evidenced by lower CD11b levels (Fig 1D,E). As a control for the functionality of our antisense MIR29B, we analysed the protein expression of the reported MIR29B target gene DNMT3B. We found markedly elevated DNMT3B protein levels in anti-MIR29B expressing AML cells indicating successful inhibition of MIR29B (Fig 1F). In general, our results indicated that MIR29B expression was not only suppressed in APL but generally in AML patients, indicating that the low expression is associated with an immature myeloid phenotype. In summary, these data suggest that MIR29B is associated with granulocytic differentiation and further support a tumour suppressor function of MIR29B in AML. Expression of MIR29B during APL differentiation and aberrant neutrophil differentiation upon MIR29B inhibition. (A) Real-time quantitative reverse transcription polymerase chain reaction (qPCR) analysis of miRNAs extracted from NB4 and HT93 APL cells differentiated in vitro for 6 d using 1 μmol/l ATRA. Five hundred nanograms of total RNA was used for reverse transcription. miRNA expression was determined using the miScript SYBR Green PCR Kits and miScript Primer Assay Hs_MIR29B (MS00006566) according to the manufacturer's instructions (Qiagen, Hombrechtikon, Switzerland). Relative miRNA expression levels were normalized to endogenous Hs_SNORA73A_1 (MS00014021) levels. Results for miRNA expression represent the average and standard deviation from three independent experiments. (B) qPCR analysis of MIR29B extracted from an APL patient at diagnosis and at day 6 of ATRA treatment. Relative miRNA expression levels were normalized to the SNORA73A housekeeping gene and are shown as n-fold regulation compared to day 0. (C) qPCR analysis of MIR29B in granulocytes from healthy donors (Granulo, n = 10) and leukaemic blast cells from AML patients (FAB M0-M4, n = 14) was assessed. Values are the differences in Ct-values between the MIR29B and the levels of the 'house keeping' miRNA SNORA73A. ***Mann–Whitney-U; P < 0·001. (D) Neutrophil differentiation was assessed by CD11b flow cytometry analysis in control and anti-MIR29B expressing HT93 APL cells upon ATRA treatment. Lentiviral vectors expressing anti-sense MIR29B as well as a non-targeting control vector were purchased from SBI (System Biosciences, Mountain View, CA, USA). All vectors contain a puromycin antibiotic resistance gene for selection of transduced mammalian cells. Lentivirus was produced in 293T cells and cells were transduced using 350 μl of viral supernatant per 0·4 × 106 AML cells and 4 μg/ml polybrene. Data from one representative experiment are shown. (E) Flow cytometry analysis of CD11b surface expression of control and anti-MIR29B expressing HT93 APL cells upon ATRA-treatment. Data are shown as mean ± SD from four experiments. *Mann–Whitney-U; P < 0·05. (F) Inhibition of MIR29B allows for higher expression of DNMT3B in HL60 AML cells. Protein expression analysis of cells infected with scramble control (Ctrl) or anti-MIR29B expressing lentiviral constructs. Total protein was analysed by Western blotting. Actin was used as loading control. To assess transcriptonal regulation of MIR29B, we analysed its promoter region for myeloid transcription factor binding sites. We identified two putative responsive elements of the myeloid master regulator PU.1 (SPI1) upstream of the MIR29B2/MIR29C (MIR29B2/C) gene (Fig 2A). First, to test if PU.1 activates MIR29B2/C transcription, we cloned a 1·4-kb putative MIR29B2/C promoter carrying the two PU.1 response elements and a shorter 0·7-kb construct encompassing only one PU.1 binding element into luciferase reporter vectors. PU.1 specifically increased promoter activity of both constructs to the same levels in a dose-dependent manner (Fig 2B). Our experiments indicate that the PU.1 binding site between the MIR29B2 and MIR29C coding regions is the major PU.1 responsive element. Next, we showed direct binding of PU.1 to the MIR29B2/C promoter in vivo using chromatin immunoprecipitation (ChIP) assays (Fig 2C). Transcriptional regulation of MIR29B expression. (A) Schematic representation of two potential PU.1 (SPI1) response elements, A and B (open circles), in the MIR29B2/C promoter. The MYC binding site is indicated as a black circle. Lower panel: Position of two MIR29B2/C promoter fragments, 765 and 1471 bp, encompassing either one or both potential PU.1 binding sites, respectively, that were cloned into pGL4.10 luciferase reporter plasmids. (B) MIR29B2/C promoter reporter plasmids were cotransfected into 293T cells together with pcDNA3.1 (−) or increasing amounts of PU.1 expression vector. Protein lysates were harvested after 24 h to assess luciferase activity. Results are expressed relative to a value of 1·0 for cells transfected with empty vector and are the means of triplicate experiments. (C) Chromatin immunoprecipitation (ChIP) was performed in NB4 cells using the ChIP-IT Express Chromatin Immunoprecipitation Kit (ChIP-IT Express, Active Motif, Rixensart, Belgium) according to the manufacturer's recommendation. Upper panel: Chromatin was immunoprecipitated with an anti-PU.1 antibody (#2268, Cell Signaling, Allschwil, Switzerland) and the recovered DNA was subjected to PCR amplifying a genomic region corresponding to the putative PU.1 binding site (primers: forward 5′-GTT CTT CCC TGG ACT TCT CG-3′ and reverse 5′-AAG CTG GTT TCA CAT GGT GG-3′). Anti-IgG and anti-acetylhistone H3 (AhH3) antibodies were used as negative and positive control, respectively. Lower panel: An unrelated sequence in the GAPDH gene was used as a negative control. (D) HT93 APL cells were transduced with a scrambled control lentiviral vector (SHC002) or lentiviral constructs expressing two independent shRNAs against PU.1 (shPU.1_1 and shPU.1_2). MIR29B expression levels upon ATRA treatment were examined by qPCR as described in Fig 1A. PU.1 knockdown was confirmed by qPCR. PU.1 expression is shown relative to the mRNA levels in control cells, which were set to one. Expression of the housekeeping gene HMBS was used for normalization. Data are shown as mean ± SD of triplicate experiments. (E) ATRA-treated HT93 APL cells expressing non-(SHC002) or PU.1-targeting (shPU.1_1 and shPU.1_2) shRNA were tested for PU.1 knockdown by Western blot analysis. (F) Chromatin was immunoprecipitated with an anti-MYC antibody (#9402, Cell Signaling, Allschwil, Switzerland) as described in C. Anti-IgG and anti-acetylhistone H3 (AhH3) antibodies were used as negative and positive control, respectively. Upper panel: The recovered DNA was amplified by primers encompassing the MYC binding site in the MIR29B2/C promoter. Lower panel: An unrelated sequence in the GAPDH gene was used as a negative control as in C. (G) Ectopic MYC expression by lentiviral vectors in NB4 cells results in impaired neutrophil differentiation. Flow cytometry analysis of CD11b surface expression of NB4 control (NB4) and MYC vector transduced NB4 (NB4 MYC) cells upon ATRA-treatment. (H) Control (NB4), MYC transduced (NB4 MYC) and ATRA-resistant (NB4-R2) cells were treated for 96 h with 1 μmol/l ATRA and MIR29 levels were assessed by qPCR. (I) Total protein of ATRA-treated parental and MYC transduced NB4 cells were analysed by MYC Western blotting. Actin served as loading control. To analyse PU.1-dependent MIR29B induction during neutrophil differentiation, we assessed its regulation in HT93 PU.1 knockdown cells upon ATRA treatment. Indeed, inhibiting PU.1 in HT93 cells resulted in 50% reduction of ATRA-induced MIR29B expression (Fig 2D, left panel). Successful PU.1 knockdown was confirmed by real-time quantitative polymerase chain reaction and Western blotting in HT93 shPU.1_1 and _2 knockdown cells (Figs 2D, right panel and 2E). PU.1 induced MIR29B by directly interacting with its promoter and we confirmed an essential role for PU.1 in MIR29B2/C regulation upon myeloid differentiation by ChIP, luciferase reporter and PU.1 knockdown experiments. Chang et al (2008) placed the transcriptional start site of MIR29B2/C within a conserved regulatory region located 20 kb upstream of the MIR29B2/C cluster on chromosome 1. However, we identified a functional PU.1 binding site in the proximal MIR29B2/C promoter, indicating additional regulatory elements for MIR29B2/C transcription. Low MIR29B2/C expression in APL cells might be due to the low PU.1 levels in these cells caused by PML-RARA (Mueller et al, 2006). Interestingly, a report by Saumet et al (2009) predicted a PML-RARA binding site in the proximal MIR29B2/C promoter indicating that PML-RARA may be directly involved in repressing MIR29B2/C. In summary, MIR29B2/C emerged as a novel, direct transcriptional target of PU.1. Previous studies have reported that the oncogenic transcription factor MYC inhibits MIR29B2/C expression by direct binding to its promoter (Chang et al, 2008). We confirmed binding of MYC to the predicted binding site in the MIR29B2/C promoter in NB4 cells by ChIP experiments (Fig 2F). To further show that MYC protein downregulation during granulocytic differentiation is of significance for MIR29 upregulation, we overexpressed MYC in NB4 cells. Expression of MYC markedly impaired ATRA-induced neutrophil differentiation as measured by decreased CD11b levels (Fig 2G). Importantly, MYC expression abolished ATRA-induced MIR29B upregulation and MIR29B was not induced in ATRA-resistant NB4-R2 cells (Fig 2H). Regulation of MYC protein in control and MYC transduced cells upon ATRA treatment was shown by Western blotting (Fig 2I). Our results demonstrate that MYC directly represses MIR29B expression in APL cells and suggest that degradation of MYC is required for MIR29B upregulation during ATRA-induced differentiation. In summary, we observed significantly lower MIR29B levels in AML as compared to normal neutrophils. Inhibition of MIR29B impaired neutrophil differentiation of APL cells. We further showed that PU.1 and MYC are transcriptional regulators of the MIR29B2/C locus in APL cells. We propose that low expression of MIR29B in APL is due to aberrant expression of PU.1 and MYC transcription factors. Deborah Shan and Gustav Arvidsson are gratefully acknowledged for excellent technical support. This work was supported by grants from the Bernese Cancer League (to MPT), the Swiss National Science Foundation 3100A0-118276 (to MPT), the Werner and Hedy Berger-Janser Foundation of Cancer Research (to MFF and MPT), the Bernese Foundation of Cancer Research, the Marlies-Schwegler Foundation, the Ursula-Hecht-Foundation for Leukaemia Research (to MFF), and the NIH R01HL091219 (to BET). The authors have no competing interests. JB and EB performed the research and drafted the article, EAF and MJ performed ChIP analyses, AT and MFF instigated the experimental design and revised the drafted article, BET contributed essential reagents and revised the manuscript, MPT designed the project, wrote the paper and gave final approval of the submitted manuscript.
