Skip to main content
NCBI home page
Autophagy logoLink to Autophagy
. 2019 Jun 28;16(4):659–671. doi: 10.1080/15548627.2019.1634945

Circular RNA circHIPK3 modulates autophagy via MIR124-3p-STAT3-PRKAA/AMPKα signaling in STK11 mutant lung cancer

Xiuyuan Chen a,b, Rui Mao c, Wenmei Su d, Xia Yang e, Qianqian Geng e, Chunfang Guo b, Zhuwen Wang b, Jun Wang a,✉, Laura A Kresty b, David G Beer b, Andrew C Chang b, Guoan Chen f,✉
PMCID: PMC7138221  PMID: 31232177

ABSTRACT

The role of circular RNA in cancer is emerging. A newly reported circular RNA HIPK3 (circHIPK3) is critical in cell proliferation of various cancer types, although its role in non-small cell lung cancer (NSCLC), has yet to be elucidated. Our results provided evidence that silencing of circHIPK3 significantly impaired cell proliferation, migration, invasion and induced macroautophagy/autophagy. Mechanistically, we uncovered that autophagy was induced upon loss of circHIPK3 via the MIR124-3p-STAT3-PRKAA/AMPKa axis in STK11 mutant lung cancer cell lines (A549 and H838). STAT3 abrogation as well as transfection with a MIR124-3p mimic, recapitulated the induction of autophagy. We also demonstrated antagonistic regulation on autophagy between circHIPK3 and linear HIPK3 (linHIPK3). We therefore propose that the ratio between circHIPK3 and linHIPK3 (C:L ratio) may reflect autophagy levels in cancer cells. We observed that a high C:L ratio (>0.49) was an indicator of poor survival, especially in advanced-stage NSCLC patients. These results support that circHIPK3 is a key autophagy regulator in a subset of lung cancer and has potential clinical use as a prognostic factor. The circular RNA HIPK3 (circHIPK3) functions as an oncogene and autophagy regulator may potential use as a prognostic marker and therapeutic target in lung cancer.

Abbreviations 3-MA: 3-methyladenine; AMPK: AMP-activated protein kinase; ATG7: autophagy related 7; Baf-A: bafilomycin A1; BECN1: beclin 1; circHIPK3: circular HIPK3; CQ: chloroquine; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; GFP: green fluorescent protein; HIPK3: homeodomain interacting protein kinase 3; IL6R: interleukin 6 receptor; MAP1LC3B/LC3B: microtubule associated protein 1 light chain 3 beta; NSCLC: non-small cell lung cancer; RFP: red fluorescent protein; RPS6KB1/S6K: ribosomal protein S6 kinase B1; SQSTM1/p62: sequestosome 1; STAT3: signal transducer and activator of transcription 3; STK11: serine/threonine kinase 11

KEYWORDS: Cell death, ceRNA, circRNA, HIPK3, prognosis, STK11

Introduction

Circular RNAs are covalently-closed, single-stranded transcripts derived from pre-mRNAs [1]. Discovered in the 1970s [2–4], they were then regarded as a rare product arising from mRNA splicing errors. Their abundance in human cells has recently been clarified by RNA-seq analysis [5–7], and they appear to have a significant role in tumorigenesis and development [8].

Circular RNAs can be divided into 4 different categories based on their origin [9,10]. CircRNAs derived from exons are the most widely studied, with other types including ciRNAs derived from introns, ElciRNAs derived from a combination of exons and introns, as well as group I and II ribozymes. Multiple hypotheses have been proposed regarding the biogenesis of circRNAs, among which alternative splicing, reverse complementary intronic sequence paring, or RNA binding protein regulation being the most accepted ones [9,11]. Understanding of circRNA function however remains incomplete but it is well accepted that circRNAs can act as microRNA sponges to regulate gene expression and may have diagnostic or therapeutic potential [12,13].

Recent studies have focused on the relationship between circular RNA and cancer [8,14–16]. Among all cancers, NSCLC (non-small cell lung cancer) carries the greatest mortality worldwide with over 150,000 deaths estimated in 2018 in the United States [17]. Although circular RNAs are involved in the development of various cancers [13,16,18], yet their function in lung cancer remains to be characterized.

Circular RNA circHIPK3 is a known oncogene derived from chromosomal region 11p13 and originating from the second exon of HIPK3 [16]. It appears to regulate cell growth in liver, colon and cervical cancer cell lines by sponging multiple miRNAs [16]. Clinically circHIPK3 has been reported to be negatively correlated with bladder cancer grade, invasion as well as lymph node metastasis [19]. The function and clinical potential of circHIPK3 in NSCLC is unknown. In this study, we investigated the function and mechanism of action of circHIPK3 in NSCLC cell lines and examined its clinical significance in patients with primary NSCLC.

Results

CircHIPK3 is abundantly expressed in a wide spectrum of lung cancer cell lines

Although circHIPK3 was reported to play divergent roles in different cancer types and may have different biological functions [16,19], the role of circHIPK3 in lung cancer remains unknown. We systematically investigated the expression of circHIPK3 among multiple lung cancer cell lines carrying different driver gene mutations (Table S1). Derived from the second exon of HIPK3 (Figure S1A), circHIPK3 was expressed in all tested cell lines (Figures 1A and S1B). It was prone to RNase R digestion and was undetectable in genomic DNA (Figure S1C,D). Although circHIPK3 levels were relatively lower than the housekeeping mRNA GAPDH, it was as abundant as the poly(A) isoform linear HIPK3 (linHIPK3) (Figure 1A). Considering that the ratio between most circular RNAs and their linear counterpart is only one percent [7], we suspected that circHIPK3 is highly functional in lung cancer cell lines given its prominent expression in comparison with linHIPK3.

Figure 1.

Figure 1.

Open in a new tab

The expression and cellular distribution of circHIPK3. (A) RT-PCR analyses of the expression of circHIPK, linHIPK3, and GAPDH cDNA in various NSCLC cell lines. Y-axis is the raw Ct value. (B, C) RT-PCR result of the cellular distribution of circHIPK3 and linHIPK3 mRNA in A549 and H1299. (D, E) Cells were treated with si-Ctrl, si-circ-1, si-lin, si-both and si-circ-2 respectively for 48 h prior to collecting RNA. RT-PCR showed over 95% knockdown effect of circHIPK3 upon si-circ-1, si-both and si-circ-2 treatment, and over 80% knockdown of linHIPK3 upon si-lin and si-both treatments. (F, G) RT-PCR result indicated the change in cellular distribution of circHIPK3 and linHIPK3 upon si-circ-1 and si-lin treatment, respectively. All data are presented as the means ± SD of at least 3 independent experiments. * P < 0.05, ** P < 0.01.

To determine the cellular distribution of circHIPK3 and linHIPK3, we performed RT-PCR using RNA samples collected exclusively from either the cytoplasm or nuclear fractions of lung cancer cell line lysates. We found that circHIPK3 was distributed predominantly in the cytoplasm, while linHIPK3 was more abundant in the nucleus in these cell lines (Figures 1B,C and S1E,F).

To test whether circHIPK3 was functional in lung cancer, we generated small interfering RNAs (siRNAs) for gene abrogation experiments. We utilized 3 siRNAs from the publication by Zheng et al. [16]: si-circ-1 targeting circHIPK3, si-lin targeting linHIPK3, and si-both targeting both isoforms (circHIPK3 and linHIPK3) (Figure S2A). We generated an additional siRNA si-circ-2 to target circHIPK3 using CircInteractome Database [20]. On average, si-circ-1, si-circ-2 and si-both treatment resulted in less than 5% of the residual circHIPK3 level seen with si-Ctrl; while si-lin and si-both treatment resulted in approximately 80% reduction of linHIPK3 expression (Figures 1D,E and S2B,C). We observed an approximately 2-fold increase of linHIPK3 in whole cell lysates upon silencing circHIPK3 in all 3 cell lines (Figures 1E and S2B). The increased linHIPK3 arose mostly from the nucleus in H1299 cells (Figure 1G) indicating a possible feedback mechanism in response to loss of circHIPK3. We also noticed that upon linHIPK3 abrogation, circHIPK3 was distributed more in nuclear fractions (Figure 1F,G), although it remained steadily expressed in whole cell lysates.

Previous research revealed that HIPK3 is a frequent methylated target in multiple diseases and disorders [21,22]. We investigated whether circHIPK3 was regulated by epigenetic modification. In A549, by inhibiting deacetylation using TSA or demethylation using 5-azacytidine (5-AZA), we observed a significant decrease in circHIPK3 but an increase in linHIPK3. Although the change is statistically significant, the fold change is not profound at a 1.5-fold max. We consider this more of a general rather than specific effect of epigenetic regulation (Figure S2D) in lung cancer. We did not find DNA amplification or loss in circHIPK3 region by SNP array analysis (Figure S2E).

Silencing of circHIPK3 inhibits cancer cell proliferation, migration and invasion in vitro

To investigate the function of circHIPK3 at the cellular level, cell proliferation assays were measured using water-soluble tetrazolium 1 (WST-1). We evaluated the loss of functional impact on cell viability by circHIPK3 and/or linHIPK3 in multiple lung cancer cell lines including A549, H838, H1299, PC-9, H1975, H1650 and HCC827. Our analysis revealed that loss of circHIPK3 conferred a time-dependent reduction in cell viability (Figures 2A–C and S3A). Exclusively silencing of linHIPK3 minimally affected cell viability in most of the cell lines, and even resulted in a minor increased viability in A549 and H838 at 96 h (Figures 2A–C and S3A,B). We also showed that treatment with si-both (loss of both linHIPK3 and circHIPK3) did not reverse the growth inhibition caused by circHIPK3 abrogation alone. Corroborating these results, colony formation capacity was impaired by silencing of circHIPK3 (with or without linHIPK3) (Figure 2D,E). On the other hand, loss of linHIPK3 alone exhibited only a minor effect (Figure 2D,E). These results indicated that circHIPK3 was the dominant isoform in modulating cell proliferation in lung cancer as compared to linHIPK3.

Figure 2.

Figure 2.

Open in a new tab

Silencing of circHIPK3 suppressed cell viability, colony formation, invasion and migration. (A–C) A549, H838 and H1299 were transfected with si-Ctrl, si-circ-1, si-lin, si-both and si-circ-2 respectively for up to 120 h. Supplementary siRNA was added every 48 h after initial treatment. Cell viability was measured using WST-1 assay every 24 h. (D, E) A549, H838 and H1299 were transfected with 5 siRNAs respectively for 48 h followed by siRNA free culture in normal medium for 11 d to form colonies. (F–H) Boyden chamber matrigel invasion and migration assays were implemented after 48 h of 5 siRNAs treatment respectively. All data are presented as the means ± SD of at least 3 independent experiments. Scale bar: 20 μm. * P < 0.05, ** P < 0.01.

In addition to cell proliferation, we also questioned whether circHIPK3 modulates cell migration and invasion. We selected H1299 for Boyden chamber Matrigel invasion and migration assays since these cells were derived from lymph node metastasis and have been reported to exhibit high invasive potential [23]. Consistent with cell viability assays, loss of circHIPK3 resulted in significant reduction in cell invasion and migration capacity (Figures 2F–H and S3C). Exclusive loss of linHIPK3 slightly decreased invasion capacity and had no influence on migration. Collectively, these cell function analyses demonstrate that circHIPK3, independent of its linear counterpart, was an important modulator of lung cancer cell growth, invasion and migration. As the cell proliferation assays showed that A549, H838 and H1299 had a modest response to circHIPK3 abrogation, we therefore selected these 3 cell lines for further mechanism-based studies.

Abrogation of circHIPK3 induces autophagy in a subset of lung cancer cell lines

A role for HIPK3 protein as an autophagy regulator has been demonstrated in a recent publication in Huntington disease [24,25]. As circHIPK3 is derived from the same pre-mRNA as linHIPK3, we therefore investigated whether circHIPK3 affected lung cancer cell viability by modulating this same mechanism.

LC3B is a ubiquitin-like protein that has a non-lipidated form, LC3B-I, and a lipidated form, LC3B-II. The accumulation of LC3B-II and the conversion from LC3B-I to LC3B-II can serve as sensitive marker for autophagy induction [26,27]. Upon circHIPK3 knockdown, increased LC3B-II expression, upregulated LC3B-I/II conversion along with SQSTM1/p62 degradation were observed in both A549 and H838, indicating enhanced autophagosome synthesis and the onset of LC3B-mediated protein degradation (Figure 3A). In H1299, however, LC3B-II expression and LC3B-I/II conversion was suppressed, the degradation of SQSTM1 was also not observed (Figure 3A), suggesting the induction of autophagy via circHIPK3 abrogation appeared only in a subset of cell lines (further referred as autophagy-induced cell lines).

Figure 3.

Figure 3.

Open in a new tab

Loss of circHIPK3 induced autophagy. (A) In A549 and H838, transfection A549 and H838 with si-circ-1 and si-circ-2 increased LC3B-II accumulation, transfection with si-lin decreased LC3B-I to II conversion, while transfection with si-both partially reversed the induction of autophagy by circHIPK3 abrogation. In H1299, however, loss of circHIPK3 leaded to decreased autophagy. The fold change comparing to si-Ctrl normalized by GAPDH was listed beneath each band. For LC3B, the first line is the quantitative value for LC3B-I and the second line for LC3B-II. (B-D) Representative images of autophagic flow. All 3 cell lines were treated with 5 siRNAs for 48 h and infected with Premo Autophagy Tandem Sensor RFP-GFP-LC3B for 24 h. Cells were visualized alive with fluorescence microscope. Scale bar: 5 µm. (E-G) Autophagosomes and autolysosomes in each 200X field were counted, at least 100 cells were counted for each siRNA treatment per cell line. Autophagic flow was increased upon circHIPK3 silencing and decreased upon linHIPK3 abrogation. Si-both treatments resulted in normal autophagic flux. All data are presented as the means ± SD of at least 3 independent experiments. (H–J). Cell lines were pretreated with 3-MA, Baf-A or normal medium for 1 h before siRNA treatment. Five siRNAs were transfected respectively for 96 h, supplementary siRNAs were added every 48 h. WST-1 assay showed the impairment of cell viability by silencing circHIPK3 was reversed by autophagy inhibitor 3-MA and Baf-A in A549 and H838, but not H1299. All data are presented as the means ± SD of at least 3 independent experiments. * P < 0.05, ** P < 0.01.

To further confirm that circHIPK3 silencing can induce autophagy, we measured the autophagic flux using the Premo Autophagy Tandem Sensor RFP-GFP-LC3B on both autophagy-induced cell lines (A549 and H838) and non-autophagy-induced cell line (H1299). With RFP-GFP-LC3B assay, GFP and RFP are both fluorescence-emitting markers in autophagosomes, while in autolysosomes, the GFP signal is lost due to the low pH environment, allowing only RFP to fluoresce. This assay allows us to inspect the completion of autophagy flux that is deemed to be more sensitive and quantitative than western blot markers [26]. In A549 and H838, we found that silencing circHIPK3 increased the formation of total autophagosomes and autolysosomes especially in H838 cells (Figures 3B,C,E,F and S4A,B). Loss of linHIPK3 in contrast, decreased the autophagic flux by 10–50% in these 2 cell lines (Figure 3B,C,E,F). Furthermore, si-both treatment that silenced both circHIPK3 and linHIPK3 reversed the increase in the autophagic flux caused by c-circ-1 and c-circ-2 in both cell lines (Figure 3B,C,E,F). These results suggested that the regulation of autophagy by circHIPK3 (si-circ-1) and linHIPK3 (si-lin) in autophagy-induced cell lines may be opposing. In the non-autophagy-induced cell line (H1299), both circHIPK3 and linHIPK3 silencing impaired the formation of autophagosome and autolysosomes (Figures 3D,G and S4C).

To further investigate the impact of autophagy on cell viability, we treated 3 cell lines with autophagy inhibitor 3-methyladenine (3-MA) or bafilomycin A1 (Baf-A) followed by circHIPK3 abrogation. We found that the cell growth inhibition caused by circHIPK3 knockdown can be reversed by both autophagy inhibitors in A549 and H838, while in H1299, no reverse effect was achieved (Figure 3H–J). Taken together, these results indicated that cell growth inhibition caused by circHIPK3 silencing may be partially related with autophagy inducing in a subset of lung cancer cell lines.

CircHIPK3 regulates autophagy viaMIR124-3p-STAT3 axis in autophagy-induced cell lines (A549 and H838)

Recent research indicates that circHIPK3 functions as a ceRNA to sequester multiple miRNAs [16,19,28], among which MIR124-3p is a well-known tumor suppressor and its regulation role in autophagy was recently being emphasized [29]. It has been reported that IL6R is regulated by circHIPK3-MIR124-3p axis [16]. STAT3, a main downstream molecule of IL6R, is also regulated by MIR124-3p [30] and was reported as a highly functional autophagy suppressor [25,31]. We therefore hypothesized that STAT3 may be a critical molecule that bridges circHIPK3 and autophagy in lung cancer.

We first validate the circHIPK3-MIR124-3p-IL6R axis using in silico analysis. Targets of the miRNA were predicted with miRmap and TargetScan (Release 6.2) [32,33]. The 3ʹ-UTR of both IL6R and STAT3 were predicted to be MIR124-3p targets (Figure S5A,B). In line with the prediction, RT-PCR showed 2- to 4-fold increase MIR124-3p level and 30% downregulation of IL6R upon circHIPK3 knockdown (Figures 4A,B and S5C), indicating the sponging effect between circHIPK3 and MIR124-3p and the subsequent targeting effect between MIR124-3p and IL6R. Treatment with MIR124-3p mimic caused the accumulation of LC3B-II in both autophagy-induced cell lines (A549 and H838) (Figure S6A) indicating the induction of autophagy. Corroborating the previous publication [30], we found that downregulation of STAT3 also occurred upon MIR124-3p mimic treatment (Figure 4C,D), indicating that STAT3 is regulated via circHIPK3-MIR124-3p axis (Figure S6A–C).

Figure 4.

Figure 4.

Open in a new tab

Autophagy was induced via the MIR124-3p-STAT3 axis. (A) A549 and H838 were treated with both circHIPK3 siRNAs for 48 h. Upregulation of MIR124-3p was observed. (B) IL6R was downregulated on RNA level upon si-circ-1 and si-circ-2 treatment for 48 h in A549 and H838. (C) H838, H1299 and A549 were treated with MIR124-3p mimic or inhibitor for 72 h to collect protein. Upon MIR124-3p mimic treatment, the total STAT3, phospho-STAT3 and SQSTM1 was decreased, while phospho-PRKAA/AMPKa was increased. (D) A549 and H838 were transfected with MIR124-3p mimic (mmc) for 48 h, which downregulated STAT3 on RNA level. (E) A549, H838 and H1299 were treated with 5 siRNAs respectively for 72 h before collecting whole cell protein extracts. Silencing circHIPK3 leaded to downregulation of phospho-STAT3 in all 3 cell lines. Silencing linHIPK3 leaded to upregulation of phospho-STAT3. (F) A549 and H1299, which responded differently upon circHIPK3 silencing, were treated with si-Ctrl and si-STAT3 respectively for 72h. LC3B-II accumulation was observed upon silencing STAT3. (G, H) Three cell lines were treated with si-Ctrl and si-STAT3 respectively for 48 h and infected with Premo Autophagy Tandem Sensor RFP-GFP-LC3B for 24 h. Autophagosomes and autolysosomes in each 200X field were counted, at least 100 cells were counted for each siRNA treatment per cell line. Scale bar: 5 µm. All fold of changes in western blot were listed beneath each band. All data are presented as the means ± SD of at least 3 independent experiments. * P < 0.05, ** P < 0.01.

To pinpoint the role of STAT3 in circHIPK3 abrogation induced autophagy, we first screened the STAT3 expression upon circHIPK3-linHIPK3 abrogation. Western blot results demonstrated that phospho-STAT3 (Y705) was decreased in all 3 cell lines upon circHIPK3 knockdown (Figure 4E). The decrease of total STAT3 was obvious in treatment with si-both and si-circ-2, while sometimes inconsistent with si-circ-1 in A549 and H1299 cell lines (Figures 4E and S6D), which may be caused by off-target effect or is treatment time-dependent (e.g. si-circ-2 has stronger effected on decreasing p-STAT3 than si-circ-1(Figure 4E); upon 24-h treatment, both si-circ-1 and si-circ-2 downregulated t-STAT3 in A549). Loss of linHIPK3, conversely, increased total and phospho-STAT3 (Figure 4E). Rescue experiment by treatment with si-both reversed downregulation of phospho-STAT3 (Figure 4E) as well as autophagy induction (Figure 3A,E,F) caused by circHIPK3 abrogation alone. To further confirm STAT3 plays a key role in circHIPK3-mediated autophagy regulation, a loss-of-function study was carried out. Silencing of STAT3 using siRNA profoundly increased LC3B-II accumulation, LC3B-I to II conversion (Figure 4F), and autophagic flux in all 3 cell lines (Figures 4G,H and S6E). We found that nuclear phosphorylated STAT3 (p-STAT3) was decreased upon si-circ-1 or 2 treatment, while cytoplasmic p-STAT3 was not changed significantly, but both nuclear and cytoplasmic p-STAT3 were increased upon si-lin treatment (Figure S6F). Taken together, these results indicated that STAT3 plays a key role in circHIPK3 abrogation induced autophagy.

CircHIPK3 abrogation induced autophagy via STAT3-PRKAA and is STK11-dependent

The fact that STAT3 silencing and inhibition caused autophagy in not only autophagy-induced cell lines (A549 and H838) but also non-autophagy-induced cell line (H1299) indicated that STAT3 is not the only autophagy regulator in circHIPK3 silencing induced autophagy. In order to determine the underlying mechanism, we focused on differentially expressed autophagy-related proteins between H1299 and 2 other cell lines (A549 and H838). We discovered that phospho-PRKAA/AMPKα was upregulated in A549 and H838 while being downregulated in H1299 after circHIPK3 silencing (Figure 5A). Notably, circHIPK3 and linHIPK3 regulate PRKAA also in an antagonistic pattern as they regulate STAT3, and phospho-PRKAA level is negatively correlated with phospho-STAT3 in A549 and H838. In rescue experiments, silencing PRKAA caused accumulation of LC3B-I and reversed the increased LC3B-I/II conversion by si-circ-1 in both A549 and H838 (Figure 5B) and significantly decreased the formation of autolysosome and blocked autophagic flux (Figures 5C,D and S7A). By investigating the interaction between STAT3 and PRKAA, we noticed that the regulation of STAT3 by PRKAA is cell type specific to A549 only, while the regulation of PRKAA by STAT3 is universal in all 3 cell lines (Figure 5B,E). These findings demonstrated that circHIPK3 abrogation induced autophagy through STAT3/PRKAA pathway in autophagy-induced cell lines (A549 and H838).

Figure 5.

Figure 5.

Open in a new tab

Autophagy induction is STK11 dependent. (A) A549, H838 and H1299 were treated with 5 siRNAs respectively for 72 h before collecting whole cell protein extracts. Silencing circHIPK3 leaded to upregulation of phospho-PRKAA in A549 and H838, but downregulation of phospho-PRKAA in H1299. STK11 was not detectable in A549 and H838 but downregulated upon circHIPK3 abrogation in H1299. Silencing linHIPK3 leaded to upregulation of phospho-PRKAA. (B) Cell lines were treated with si-Ctrl, si-PRKAA, c-si1 and c-si1 + si-PRKAA respectively. Decreased LC3B-I/II conversion was observed in both cell lines and could be reversed upon circHIPK3 abrogation. Downregulation of STAT3 by PRKAA silencing was observed only in A549. All fold of changes in western blot were listed beneath each band. Two exposures for LC3B-I/II were chosen to reveal the change in each band. The upper image with longer exposure time was labeled with LC3B-I quantification, the lower image with shorter exposure time was labeled with LC3B-II quantification. (C, D). A549 was treated by si-Ctrl, si-PRKAA, c-si1 and c-si1+ si-PRKAA respectively for 48h. Premo Autophagy Tandem Sensor RFP-GFP-LC3B was infected 24h prior to fluorescent measurement. Cells were visualized alive with fluorescence microscope. Autophagosomes and autolysosomes in each 200X field were counted, at least 100 cells were counted for each siRNA treatment per cell line. Scale bar: 5 µm. The formation of autolysosome is significantly suppressed upon PRKAA silencing. (E) In all 3 cell lines, silencing of STAT3 leaded to upregulation of phospho-PRKAA. (F) In non-autophagy-induced cell line H1299, silencing of STK11 decreased phospho-PRKAA level and LC3B-II accumulation. Co-treatment with si-STK11 and si-circ-1 and 2 revealed the autophagy inducing effect of circHIPK3 silencing comparing to si-STK11 alone. All data are presented as the mean ± SD of at least 3 independent experiments. * P < 0.05, ** P < 0.01.

In H1299, however, phospho-PRKAA was decreased upon circHIPK3 silencing while upregulated upon loss of STAT3 (Figure 5A,E). The contradictive result indicated that PRKAA is regulated by an additional factor which could bypass the effect mediated by STAT3 in H1299.

To discern the underlying molecules that regulate PRKAA in the circHIPK3 pathway in H1299 cells, we analyzed CCLE database for expression level of PRKAA and its upstream molecules in tested cell lines [34]. We discovered that STK11/LKB1 is highly expressed in H1299, while in A549 and H838, the expression of STK11 was relatively lower (Table S2). A549 and H838 are also reported to carry different STK11 mutations (p.Q37* and p.T212fs respectively) while H1299 retains wild-type STK11 [34–38]. Similar results were observed by western blot, where total STK11 was downregulated upon loss of circHIPK3 in H1299, which resulted in decreased phospho-PRKAA; in A549 and H838, the expression of STK11 was not detected (Figure 5A).

STK11 is a serine/threonine kinase that phosphorylates PRKAA in STK11-PRKAA pathway [39]. In H1299, silencing STK11 using siRNA resulted in decreased phospho-PRKAA and decreased LC3B-II accumulation (Figure 5F). Interestingly, in STK11-silenced H1299, knock down of circHIPK3 using si-circ-1 or si-circ-2 partially increased the LC3B-II accumulation, although p-PRKAA was even decreased, as compared to STK11 silencing only (Figure 5F), indicating that the autophagy effect by circHIPK3 silencing in H1299 was not only upon STK11/p-PRKAA, but may also upon other molecules such as BECN1 or ATG7 (BECN1 and ATG7 were decreased upon circHIPK3 silencing, Figure S7B).

Taken together, the induction of autophagy upon circHIPK3 silencing appears via STAT3-PRKAA pathway in STK11 mutant cells (A549 and H838). Whereas, in STK11 wild-type H1299 cells, circHIPK3 regulated PRKAA is through both STAT3 and STK11. The fact that PRKAA was decreased upon circHIPK3 knockdown indicated the regulation by STK11 was dominant.

To confirm the effects of circHIPK3 on protein expression, we designed 2 new siRNAs that specifically targets circHIPK, e.g. si-circ-4 and si-circ-5. Western blot analysis revealed similar results regarding protein changes of STAT3, p-PRKAA and SQSTM1 in A549, H299 and H838 cell lines using these 2 new siRNAs (Figure S7C).

To strengthen the conclusions regarding the circ-HIPK3/STAT3/PRKAA relationship in STK11-deficient/mutated cells, we utilized 4 additional STK11-deficient/mutated lung cancer cell lines (H1437, H1993, H2122 and A-427) and found similar results regarding the protein changes of STAT3, p-PRKAA and SQSTM1 upon circHIPK3 knockdown (Figure S8A,B). We found STAT3 and SQSTM1 were decreased, and p-PRKAA increased after circHIPK3 knockdown by si-circ-1 and 2. Further autophagic flux assays are planned to confirm the autophagy event.

MTOR and RPS6KB1/p70S6K, the traditional PRKAA downstream molecules, were decreased in A549 cells, but not in H838 and H1299 cells (Figure S7B), indicating a different role of these proteins in the circHIPK3 network. BECN1 is also an autophagy regulator [25], but we found BECN1 was upregulated by circHIPK3 siRNA only in A549, while decreased in H1299 (Figure S7B), indicating that the STAT3-BECN1 pathway may not be universal in lung cancer cell lines. Another autophagy related protein ATG7 was found downregulated upon loss of circHIPK3 in H1299 (Figure S7B). It was neither regulated by STAT3 nor PRKAA silencing (Figure S8C,D). We suspect that this may be another reason why there was no autophagy induction upon circHIPK3 knockdown in H1299 cells.

The circHIPK3:linHIPK3 ratio is upregulated in lung adenocarcinomas (ADC) and is a prognosis factor for patient survival

We next analyzed the clinical impact of circHIPK3 in lung cancer. As we discovered that circHIPK3 promotes NSCLC cell line proliferation, we hypothesized that the expression of circHIPK3 in lung tumors would be higher than in normal tissue. We tested this hypothesis using RT-PCR and 54 paired tumor and normal tissue samples derived from 27 lung ADC patients. The alteration of circHIPK3 was subtle, while linHIPK3 was significantly downregulated (p < 0.01) in tumor tissue compared to its adjacent normal lung tissue (Figure 6A). Although upregulation of circHIPK3 in tumor tissue was not observed, the expression ratio between circHIPK3 and linHIPK3 (C:L ratio) was significantly higher in tumor comparing to normal tissue (43% vs 36%, P = 0.02) (Figure 6B). As circHIPK3 and linHIPK3 function as a pair of antagonistic modulators of autophagy in lung cancer cell lines, this result indicates an anti-autophagic environment in NSCLC tissue.

Figure 6.

Figure 6.

Open in a new tab

Differential expression of circHIPK3 and its impact on NSCLC patient survival. (A, B) The differential expression of circHIPK3 and linHIPK3 in 54 paired tumor and normal tissue derived from 27 patients. The expression of linHIPK3 was significantly higher in normal tissue while C:L ratio is higher in tumor comparing with normal tissue. (C) Five-year overall survival of 76 NSCLC patients. The number of patients in high ratio group was 25, in low ratio group was 51. The log-rank test P value is 0.05. (D) Five-year overall survival of 37 stage II, III and IV patients (excluded 39 stage I from the original cohort). Eight patients in high ratio group and 29 in low ratio group. (E) Five-year overall survival of 39 stage I patients. Seventeen patients in high ratio group and 22 in low ratio group. Log-rank test showed P value is less than 0.01. All data are presented as the means ± SD.

Autophagy is known to exert a paradoxical influence on cancer cells [40]. It can either act as a tumor suppressive effector through oncogenic protein degradation, or be tumor promoting via increasing tolerance to metabolic stress [41]. Since our previous results indicated that circHIPK3 and linHIPK3 regulated autophagy antagonistically and low C:L ratio induced autophagy, we investigated whether it influenced patient survival. To reveal the clinical impact of C:L ratio, we expanded the sample size to 76 tumor tissues from 76 lung ADC patients to investigate the correlation between C:L ratio and 5-year overall survival. Patients were divided into high and low ratio groups based on the cut-off point of 0.48 (upper 1/3 quantile of all patients). Baseline data (Table S3) demonstrated that the high-ratio group tended to have more stage I and less stage IV patients as comparing to the low-ratio group (p value not significant). We hypothesized that with higher tumor burden, stage IV NSCLC may suffer heavier metastatic stress, resulting in a lower C:L ratio. A trend of poor prognosis in the high ratio group was observed (p = 0.05, log-rank test) (Figure 6C). When stratifying patients by pathological stage, both high and low C:L groups had similar 5-year OS in stage I patients. In contrast, among stage II, III and IV patients, the high C:L ratio group had significantly lower survival rates than the low ratio group (p < 0.01) (Figure 6D,E). Survival analysis using circHIPK3 or linHIPK3 alone did not demonstrate significant differences (Figure S9A–D). These results indicated that the C:L ratio might be an effective prognosis factor for non-stage I NSCLC patients. Interestingly, by analysis of autophagy-related genes expressed in different stages of 442 lung cancers (Shedden etal, Nature Med 2008), we found that autophagy promoting genes such as PRKAA1/AMPKa1, CAMKK2 and STK11, as well as autophagy marker MAP1LC3B/LC3B were upregulated in stage 3 as compared to stage 1 tumors (Figure S9E). Meanwhile, autophagy inhibiting genes such as STAT3, MTOR and BCL2, as well as HIPK3 were downregulated in stage 3 tumors (Figure S9F), indicating that late stage tumors may have a higher autophagy activation as compared to early stage, and also indicating a poor patient survival.

Discussion

In this study, we provided further insight into the function, molecular mechanism and clinical significance of circHIPK3 in lung cancer. We showed that a decreased expression of circHIPK3 induced autophagy in selected lung cancer cell lines. The decrease of its linear counterpart linHIPK3 resulted in an antagonistic effect on autophagy which was achieved by an opposing regulation of STAT3. Although si-circ-1, one of the circHIPK3 targeting si-RNAs, did not exhibit the same effect on total STAT3 (may be caused by off-target of time dependent effect), decreasing phospho-STAT3 was identical in all experiments. Silencing STAT3 using siRNAs both recapitulated the effect of circHIPK3 abrogation. We further demonstrated that a circHIPK3 sponging miRNA MIR124-3p mediated the regulation of STAT3 by targeting either STAT3 or IL6R mRNA. These findings provided the first evidence of a correlation between a circular RNA and its parental gene.

STAT3 is generally accepted as an autophagy regulator of several autophagy-related genes [25,31]. STAT3-BECN1 pathway is well known for its role in modulating autophagy induced by various means including MIR506 in pancreatic cancer [25]. In our study, however, the correlation of STAT3 inhibition and BECN1 upregulation was not evident. Instead, phospho-PRKAA was upregulated upon circHIPK3 abrogation and STAT3 inhibition, which further induced autophagy. In normal cell lines, the correlation between PRKAA and STAT3 was previously reported by various groups [34,42,43]. Nerstedt et al, reported that PRKAA directly inhibits the activation of STAT3 in HepG2 cell lines [34]. Vasamsetti et al, discovered that under PMA-stimulated conditions, the relationship between PRKAA and STAT3 phosphorylation appeared to be casually reciprocal in monocytes, and PRKAA was the upstream regulator of the MTOR-STAT3 pathway [43]. In our study, using lung cancer cell lines, STAT3, instead of a traditional downstream molecule of PRKAA, was the upstream regulator of phospho-PRKAA. We also pinpointed the lack of a normal functioning STK11 as the key to circHIPK3 inhibition-induced autophagy through the STAT3-PRKAA pathway. We have not yet fully revealed the direct regulatory mechanism between STAT3 and PRKAA, although our results indicate that MTOR is not responsible for this regulation.

The role of autophagy in cancer is controversial [40,41]. The landmark study by White et al, indicated that acute autophagy ablation in mice with pre-existing NSCLC blocked tumor growth and promoted tumor cell death [44]. However, clinical trials using various autophagy inhibitors exhibit inconsistent responses. White et al. [41] showed that prior autophagy ablation did not alter the efficiency of NSCLC initiation by activation of oncogenic KRASG12D and deletion of the TP53 tumor suppressor. Given existing publications, the role of the basal autophagy rate in lung cancer remains poorly understood. Our study indicates that the C:L HIPK3 ratio antagonistically regulates autophagy. We further hypothesized that C:L ratio may also represent the autophagy level in NSCLC patients. A high C:L ratio represents a low autophagic flow and vice versa. Corroborating our cell line result that a low C:L ratio leads to low proliferation, our patient survival data provided evidence that sustaining a high autophagic flow (indicated by low C:L ratio) may be pro-survival in NSCLC. We discovered that upregulation of the C:L ratio may have potential for treatment by either acute inhibition of autophagy in preexisting cancers, or by sustaining a high autophagic flow in NSCLC patients.

Taken together, our results provide clinical significance for the oncogenic circular RNA circHIPK3 and revealed its role in autophagy regulation. Autophagy is induced upon circHIPK3 silencing via MIR124-3p-STAT3-PRKAA pathway and is dependent on STK11 status (Figure 7). Upregulation of the ratio between circHIPK3 and linHIPK3 may hold treatment potential in NSCLC patients.

Figure 7.

Figure 7.

Open in a new tab

Schematic hypothesis of the autophagy regulation of circHIPK3. (A) For A549 and H838, which carried STK11 mutation and lower STK11 copy number and protein expression, silencing circHIPK3 induced autophagy mainly via decreased p STAT3 and increased pPRKAA signaling. (B) For H1299, which carried wild-type STK11, silencing circHIPK3 inhibited autophagy mainly via decreased STK11-pPRKAA.

Materials and methods

Patient-derived samples

Seventy-six patient-derived lung cancer tissue samples were collected from patients who underwent curative-intent lung cancer surgery from 1991 to 2014 at a single tertiary referral center. Twenty-seven of these samples were paired with adjacent non-tumoral lung tissue. Induction therapies had not been implemented prior to surgery. Written consent was provided by all enrolled patients, and the project was approved by the University of Michigan Institutional Review Board and Ethics Committee. The median follow-up time was 10.12 years. All resected specimens were immediately frozen in liquid nitrogen and then stored at −80°C until use.

Cell lines

All human cancer cell lines were purchased from the The American Type Culture Collection and genotyped for identity at the University of Michigan Sequencing Core. Lung cancer cell lines H1299 (ATCC, CRL-5803), H838 (ATCC, CRL-5844), PC-9 (Sigma-Aldrich, 90071810), HCC-827 (ATCC, CRL-2868), H-1650 (ATCC, CRL-5883) and H1975 (ATCC, CRL-5908) were cultured in RPMI 1640 (Gibco, 22400–089) supplemented with 10% fetal bovine serum (Atlanta biologicals, S11150). A549 (ATCC, CCL-185) was maintained in DMEM supplemented with 10% fetal bovine serum. Culture temperature of 37°C with CO2 concentration at 5%, and mycoplasma contamination assessment was routinely tested.

RNA isolation, RNase R pre-treatment and quantitative RT-PCR

For all total RNA samples derived from patient tissues and cell lines, RNA isolation, cDNA synthesis, and qRT-PCR were performed as described previously [45]. For samples requiring linear RNA depletion, total RNA of 10 μg was incubated 45 min at 37ºC with 2 U/μg RNase R (Epicentre Technologies, RNR07250) following RNA isolation. Products were subsequently purified using RNeasy MinElute Cleaning Kit (Qiagen, 74204). All of the purified products became substrate for cDNA synthesis, and qRT-PCR was performed in the same protocol as other samples. For microRNA analysis, isolated RNA samples were reverse transcribed using miScript II RT Kit (Qiagen, 218161) with miScript HiFlex Buffer (Qiagen, 218161). Quantitative RT-PCR was carried out using miScript SYBR Green PCR Kit (Qiagen, 218073) with 3-step cycling protocol. All procedures were implemented according to the manufacturers’ instructions.

Western blot

Cell lysis buffer (Cell Signaling Technology, 9803S) with 100 nM OmniPur Phenylmethyl Sulfonyl Fluoride (Calbiochem, 329-98-6) and 1:100 Halt Phosphatase Inhibitor Cocktail (Thermo Scientific, 78420) were utilized. Protein quantification was performed using the DC Protein Assay (Bio-Rad, 500–0116). Protein samples were prepared to 5 μg per well and denatured in 95ºC for 5 min before loading. Separation gels were made using polyacrylamide with concentration varying from 6% to 15%. Electrophoresis started at 80 volts for 10 min and then continued using 120 volts for 1.5 h followed by 25 volts transfer for 2 h. After blocking for 1 h with 5% non-fat milk, the membranes were incubated overnight with primary antibodies at 4ºC in a cold room. After incubation with HRP-conjugated secondary antibody (Cell Signaling Technology, 7074,7076) for 1 h at room temperature, the membranes were developed using Luminata Crescendo Western HRP Substrate (Millipore Sigma, WBLUR0500) and exposed to X-ray film or The ChemiDoc MP Imaging System (BIO-RAD). The band intensity was analyzed by Image Gel software. The intensity of bands was showed in the liner range although some bands were overexposed.

Primer, sirNA design and sirNA-mediated gene silencing

Primers for circHIPK3 were designed according to a previous publication [16]. Primers for linHIPK3 were designed to target the junction of exon 10 and exon 11 of the HIPK3 gene to demonstrate the expression level of linear HIPK3 alone. Four siRNAs were used specifically in this study, among which si-circ-1, si-lin and si-both were derived from a previous publication [16], si-circ-2 was designed to exclude off-target effects of si-circ-1. Two siRNAs, si-circ-1 and si-circ-2, target the junction of 5ʹ and 3ʹ of exon 2 of the HIPK3 gene to silence circHIPK3 alone, si-lin targets exon 6 to silence solely linear HIPK3, and si-both targets normal sequence of exon 2 which will cause abrogation of both circHIPK3 and HIPK3 (Figure S1G). Lung cancer cell lines were first plated at the desired number (80,000–120,000 cells/well for a 6-well plate, 800–1200 cells per well for a 96-well plate) for 24 h. Target siRNAs and non-targeting controls (Dharmacon, ON-TARGET plus Non-targeting siRNA number 1) were subsequently added to the medium in combination with lipofectamine RNAiMax Reagent (Invitrogen, 13778–150) and Opti-MEM (Gibco, 31985–062) medium for transfection. Working concentration of all siRNAs was 10 nM. Knockdown efficiency was measured using qRT-PCR.

Reagents and antibodies

Treatment utilized in this study include: 3-Methyladenine (3-MA; Millipore Sigma, 189490), bafilomycin A1 (Baf-A; Cell Signaling Technology, 54645S). Antibodies used in this study are: t-EGFR (Cell Signaling Technology, 54359), p-MET (Cell Signaling Technology, 3077), t-MET (Cell Signaling Technology, 8198), p-STAT3 (Y705; R&D Systems, AF4607), t-STAT3 (R&D Systems, MAB1799), t-STK11 (Cell Signaling Technology, 3050), p-STK11 (Cell Signaling Technology, 3482), p-PRKAA/AMPKα (Cell Signaling Technology, 2535), t-PRKAA1/AMPKα1 (Cell Signaling Technology, 5832), p-S6K (Cell Signaling Technology, 9234), BECN1 (Cell Signaling Technology, 3495), BCL2 (Cell Signaling Technology, 2872), SQSTM1/p62 (Cell Signaling Technology, 5114), LC3 (Cell Signaling Technology, 2775), ATG7 (Cell Signaling Technology, 8558), CASP3 (Cell Signaling Technology, 9662), CASP8 (Cell Signaling Technology, 9746), PARP1 (Cell Signaling Technology, 9532), t-YBX1 (Santa Cruz Biotechnology, sc-398340), NF-kB (Cell Signaling Technology, 3987), p-CREB (R&D Systems, AF2510), c-MYC (Cell Signaling Technology, 5605), p-AKT (Cell Signaling Technology, 4056), t-AKT (Cell Signaling Technology, 4691), p-MAPK/ERK (Cell Signaling Technology, 4370), t-ERK (Cell Signaling Technology, 9102), CDKN1B/p27 (Cell Signaling Technology, 3686), CDKN1A/p21 (Cell Signaling Technology, 2947), CCNE1 (Cell Signaling Technology, 4129), NOTCH1 (Cell Signaling Technology, 4380), p-WNK1 (R&D Systems, AF4720), GAPDH (Millipore Sigma, AB2302).

Cell proliferation, invasion and migration assays

Cell viability was measured 48–120 h after the initial siRNA treatment using WST-1 (Water-soluble Tetrazolium 1 assay; Roche, 11644807001). Each well was supplemented with 10 μl of WST-1 reagent, and the absorbance was measured at wavelengths of 450 nm and 630 nm after 30 and 60 min of incubation according to the manufacturer’s instructions. Colony formation assay and Boyden chamber Matrigel invasion assay were implemented after 48 h of initial siRNA treatment and sequential trypsinization. For colony formation assay, lung cancer cell lines were seeded into 6-well plates at the concentration of 200 cells per plate. Crystal violet staining was performed after 10–14 d of culture for colony counting. For invasion and migration assay, cells were seeded into Boyden chambers (8 mm pore size, BD) present in the insert of a 24-well culture plate. Matrigel (BD, 356234) was prepared at the bottom of Boyden chambers in advance for the invasion assays. After incubation of 24 h, Diff-QuickTM Stain Set (SIEMENS, 10736310) was used for staining of the invasive cells located on the lower side of the chamber. Cells in random 5 100X fields were counted to calculate the invasion and migration cell number. Invasion index [46] (ratio of invasion cells over migration cells) was calculated for reference.

Autophagic flux measurement

Premo Autophagy Tandem Sensor RFP-GFP-LC3 (Life Technologies, P36239) was used to monitor autophagic flux. After initial treatment with siRNA for 24 h, reagent was added to each well at a concentration of 10–50 particles per cell according to the manufacturer’s instructions. Fluorescence images were captured 24 and 48 h after incubation using an Olympus IX71 microscope, and autophagosomes (yellow dots in fusion images) and autolysosomes (red dots in fusion images) were counted in at least 100 cells in a 200X field.

Statistical analysis for experimental studies

All data were analyzed using Excel (Microsoft) and plotted using GraphPad Prism 7 software (GraphPad). Data were examined in the format of mean ± SD. Student’s T test or paired t-test were used to indicate significant differences between 2 data groups. For more than 2 data groups, ANOVA test with Dunnett’s multiple comparisons test was used. For survival analysis, Kaplan-Meier was implemented with log-rank test. A P value < 0.05 was considered statistically significant.

Funding Statement

This work was supported in part by the National Institutes of Health (grant R01CA154365 to D.G.B. and A.M.C.; grant U01CA157715 to D.G.B.; grant R21CA205414 to G.C.), the University of Michigan’s Cancer Center Support Grant (P30 CA46592), University of Michigan’s Cancer Center Thoracic Oncology Program Research Grant (G.C.), University of Michigan Department of Surgery RAC grant (G.C.). China Scholarship Council (CSC) funding (to X. C.)

Acknowledgments

We thank Nithya Ramnath, Khaled Hassan, Katherine Weh and Derek Nancarrow for discussions and constructive comments. We thank Jules Lin, Rishindra M. Reddy, Philip W. Carrott, Jr., William R. Lynch, and Mark B. Orringer for providing clinical samples.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary materials

Supplementary material for this article can be accessed here.

Supplemental Material

References

  • [1].Chen LL. The biogenesis and emerging roles of circular RNAs. Nat Rev Mol Cell Biol. 2016;17(4):205–211. [DOI] [PubMed] [Google Scholar]
  • [2].Arnberg AC, Van Ommen GJ, Grivell LA, et al. Some yeast mitochondrial RNAs are circular. Cell. 1980;19(2):313–319. [DOI] [PubMed] [Google Scholar]
  • [3].Kos A, Dijkema R, Arnberg AC, et al. The hepatitis delta (delta) virus possesses a circular RNA. Nature. 1986;323(6088):558–560. [DOI] [PubMed] [Google Scholar]
  • [4].Sanger HL, Klotz G, Riesner D, et al. Viroids are single-stranded covalently closed circular RNA molecules existing as highly base-paired rod-like structures. Proc Natl Acad Sci U S A. 1976;73(11):3852–3856. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [5].Memczak S, Papavasileiou P, Peters O, et al. Identification and characterization of circular RNAs as a new class of putative biomarkers in human blood. PloS One. 2015;10(10):e0141214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Rybak-Wolf A, Stottmeister C, Glazar P, et al. Circular RNAs in the mammalian brain are highly abundant, conserved, and dynamically expressed. Mol Cell. 2015;58(5):870–885. [DOI] [PubMed] [Google Scholar]
  • [7].Salzman J, Chen RE, Olsen MN, et al. Cell-type specific features of circular RNA expression. PLoS Genet. 2013;9(9):e1003777. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Rong D, Tang W, Li Z, et al. Novel insights into circular RNAs in clinical application of carcinomas. Onco Targets Ther. 2017;10:2183–2188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Chen LL, Yang L.. Regulation of circRNA biogenesis. RNA Biol. 2015;12(4):381–388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Vicens Q, Westhof E. Biogenesis of circular RNAs. Cell. 2014;159(1):13–14. [DOI] [PubMed] [Google Scholar]
  • [11].Jeck WR, Sorrentino JA, Wang K, et al. Circular RNAs are abundant, conserved, and associated with ALU repeats. Rna. 2013;19(2):141–157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].Chen X, Han P, Zhou T, et al. circRNADb: A comprehensive database for human circular RNAs with protein-coding annotations. Sci Rep. 2016;6:34985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Guarnerio J, Bezzi M, Jeong JC, et al. Oncogenic role of fusion-circRNAs derived from cancer-associated chromosomal translocations. Cell. 2016;166(4):1055–1056. [DOI] [PubMed] [Google Scholar]
  • [14].Wang F, Nazarali AJ, Ji S. Circular RNAs as potential biomarkers for cancer diagnosis and therapy. Am J Cancer Res. 2016;6(6):1167–1176. [PMC free article] [PubMed] [Google Scholar]
  • [15].Xia W, Qiu M, Chen R, et al. Circular RNA has_circ_0067934 is upregulated in esophageal squamous cell carcinoma and promoted proliferation. Sci Rep. 2016;6:35576. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Zheng Q, Bao C, Guo W, et al. Circular RNA profiling reveals an abundant circHIPK3 that regulates cell growth by sponging multiple miRNAs. Nat Commun. 2016;7:11215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin. 2018;68(1):7–30. [DOI] [PubMed] [Google Scholar]
  • [18].Dong Y, He D, Peng Z, et al. Circular RNAs in cancer: an emerging key player. J Hematol Oncol. 2017;10(1):2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Li Y, Zheng F, Xiao X, et al. CircHIPK3 sponges miR-558 to suppress heparanase expression in bladder cancer cells. EMBO Rep. 2017;18(9):1646–1659. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Dudekula DB, Panda AC, Grammatikakis I, et al. CircInteractome: A web tool for exploring circular RNAs and their interacting proteins and microRNAs. RNA Biol. 2016;13(1):34–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Moleres A, Campion J, Milagro FI, et al. Differential DNA methylation patterns between high and low responders to a weight loss intervention in overweight or obese adolescents: the EVASYON study. Faseb J. 2013;27(6):2504–2512. [DOI] [PubMed] [Google Scholar]
  • [22].Wan ES, Qiu W, Baccarelli A, et al. Systemic steroid exposure is associated with differential methylation in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2012;186(12):1248–1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Wang L, He Y, Liu W, et al. Non-coding RNA LINC00857 is predictive of poor patient survival and promotes tumor progression via cell cycle regulation in lung cancer. Oncotarget. 2016;7(10):11487–11499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Fu Y, Sun X, Lu B. HIPK3 modulates autophagy and HTT protein levels in neuronal and mouse models of Huntington disease. Autophagy. 2018;14(1):169–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Sun L, Hu L, Cogdell D, et al. MIR506 induces autophagy-related cell death in pancreatic cancer cells by targeting the STAT3 pathway. Autophagy. 2017;13(4):703–714. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [26].Klionsky DJ, Abdelmohsen K, Abe A, et al. Guidelines for the use and interpretation of assays for monitoring autophagy. 3rd edition.Autophagy. 2016;12(1):1–222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [27].Fu Y, Sun X, Lu B. HIPK3 modulates autophagy and HTT protein levels in neuronal and mouse models of Huntington disease. Autophagy. 2018;14(1):169–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [28].Shan K, Liu C, Liu BH, et al. Circular noncoding RNA HIPK3 mediates retinal vascular dysfunction in diabetes mellitus. Circulation. 2017;136(17):1629–1642. [DOI] [PubMed] [Google Scholar]
  • [29].Mehta AK, Hua K, Whipple W, et al. Regulation of autophagy, NF-kappaB signaling, and cell viability by miR-124 in KRAS mutant mesenchymal-like NSCLC cells. Sci Signal. 2017;10:496. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [30].Koukos G, Polytarchou C, Kaplan JL, et al. MicroRNA-124 regulates STAT3 expression and is down-regulated in colon tissues of pediatric patients with ulcerative colitis. Gastroenterology. 2013;145(4):842–852 e842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].You L, Wang Z, Li H, et al. The role of STAT3 in autophagy. Autophagy. 2015;11(5):729–739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [32].Lewis BP, Burge CB, Bartel DP. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell. 2005;120(1):15–20. [DOI] [PubMed] [Google Scholar]
  • [33].Vejnar CE, Zdobnov EM. MiRmap: comprehensive prediction of microRNA target repression strength. Nucleic Acids Res. 2012;40(22):11673–11683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Barretina J, Caponigro G, Stransky N, et al. The cancer cell line encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature. 2012;483(7391):603–607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Forbes SA, Bindal N, Bamford S, et al. COSMIC: mining complete cancer genomes in the catalogue of somatic mutations in cancer. Nucleic Acids Res. 2011;39(Database issue):D945–950. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [36].Kaufman JM, Amann JM, Park K, et al. LKB1 Loss induces characteristic patterns of gene expression in human tumors associated with NRF2 activation and attenuation of PI3K-AKT. J Thorac Oncol. 2014;9(6):794–804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Li D, Liu X, Zhou J, et al. Long noncoding RNA HULC modulates the phosphorylation of YB-1 through serving as a scaffold of extracellular signal-regulated kinase and YB-1 to enhance hepatocarcinogenesis. Hepatology. 2017;65(5):1612–1627. [DOI] [PubMed] [Google Scholar]
  • [38].Mehra R, Shi Y, Udager AM, et al. A novel RNA in situ hybridization assay for the long noncoding RNA SChLAP1 predicts poor clinical outcome after radical prostatectomy in clinically localized prostate cancer. Neoplasia. 2014;16(12):1121–1127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Shackelford DB, Shaw RJ. The LKB1-AMPK pathway: metabolism and growth control in tumour suppression. Nat Rev Cancer. 2009;9(8):563–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Choi AM, Ryter SW, Levine B. Autophagy in human health and disease. N Engl J Med. 2013;368(19):1845–1846. [DOI] [PubMed] [Google Scholar]
  • [41].White E. Deconvoluting the context-dependent role for autophagy in cancer. Nat Rev Cancer. 2012;12(6):401–410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Su M, Mei Y, Sinha S. Role of the crosstalk between autophagy and apoptosis in cancer. J Oncol. 2013;2013:102735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [43].Rochat-Steiner V, Becker K, Micheau O, et al. FIST/HIPK3: a Fas/FADD-interacting serine/threonine kinase that induces FADD phosphorylation and inhibits fas-mediated Jun NH(2)-terminal kinase activation. J Exp Med. 2000;192(8):1165–1174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [44].Chen K, Yang F, Jiang G, et al. Development and validation of a clinical prediction model for N2 lymph node metastasis in non-small cell lung cancer. Ann Thorac Surg. 2013;96(5):1761–1768. [DOI] [PubMed] [Google Scholar]
  • [45].Zeng Y, Du WW, Wu Y, et al. A circular RNA binds to and activates AKT phosphorylation and nuclear localization reducing apoptosis and enhancing cardiac repair. Theranostics. 2017;7(16):3842–3855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Kramer N, Walzl A, Unger C, et al. In vitro cell migration and invasion assays. Mutat Res. 2013;752(1):10–24. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
kaup-16-04-1634945-s001.zip (2.5MB, zip)

Articles from Autophagy are provided here courtesy of Taylor & Francis

RESOURCES