Circular RNA HIPK3 induces cell proliferation and inhibits apoptosis in non-small cell lung cancer through sponging miR-149

ABSTRACT

Circular RNAs (circRNAs) are a class of endogenous noncoding RNAs that are demonstrated to be potent regulators in the development of various types of human cancers, including non-small cell lung cancer (NSCLC). In the present study, the level of circRNA-HIPK3 were measured by Taq-man based quantitative real-time PCR (qRT-PCR) analysis in both NSCLC patient specimens and cells, which showed that circRNA-HIPK3 was upregulated in both NSCLC tissues and cell lines. Cell counting kit-8 (CCK-8), migration and flow-cytometry assays indicated that circRNA-HIPK3 participated in the regulation of the proliferation, migration, invasion and apoptosis of NSCLC cells. MiR-193a expression was increased by circHIPK3 silencing. We then showed that miR-149 interacts with FOXM1 by binding to the 3ʹ-untranslated region (UTR). Further, ectopic overexpression of miR-149 by transfecting miR-149 mimics significantly inhibited growth, migration and invasion of HSCLCs, which was found to be mediated through FOXM1. Moreover, miR-149 overexpression decreases the viability and proliferation of HSCLCs. Therefore, our data suggest that circHIPK3 regulates the function of NSCLCs through miR-149-mediated FOXM1 expression regulation, potentially providing a novel insight into the pathogenesis of NSCLC.

Introduction

Lung cancer remains the most common cause of cancer-related death worldwide, despite tremendous efforts in the development of chemotherapy drugs. This is due to the inherent or acquired chemoresistance of lung cancer, particularly non-small cell lung cancer (NSCLC), a subtype of lung cancer, which is responsible for poor survival of a large population of lung cancer patients.¹ With increasing insight of the molecular basis that drives lung cancer progression, targeted therapies have been devised to counteract oncogenic changes associated with lung cancer tumorigenesis and resistance. To this end, gene therapy, i.e., the delivery of nucleic acids to correct dysregulated expression of genes, has gained increasing attention and deemed to possess to great potential for the treatment of lung cancer.²

One class of gene targets in lung cancer are non-coding RNAs, which constitute a large portion of the human genome, without the ability of protein-coding. Among non-coding RNAs, circRNA is peculiar group of long non-coding RNAs with the length of at least a few hundred nucleotides and the shape of a continuous loop with jarless framework.³ A putative mechanism of the regulation of circRNA is the through sponging miRNAs, thereby regulating gene expression in a transcriptional and post-transcriptional manner. Recent evidences have implicated that abnormal circRNA expression is associated with a number of human cancers, including gastric cancer,⁴ colon cancer,⁵ ovarian cancer,⁶ etc., whereby circRNAs act as potent regulators of oncogenic signaling pathways and correction of these circRNAs exerted pronounced anti-tumor effects. For example, in lung cancer, the upregulation of a circRNA, circ0016760, was shown to be associated with unfavorable prognosis in NSCLC.⁷ circHIPK3 is a circRNA previously been reported to regulate gallbladder cancer cell growth through sponging a number of miRNAs, including miR-124, and miR-588.^8,9 CircHIPK3 also promotes colorectoal cancer growth and metasis by sponging miR-7.¹⁰ In lung cancer, circHIPK3 was shown to exert oncogenic properties by suppressing miR-124.¹¹ One miRNA, miR-149, has been previously shown as a suppressor of NSCLC¹² and miR-149 has been reported to be regulated by a circRNA, CFH.¹³ However, it is unknown an interaction exists between miR-149 and circHIPK3.

The mammalian Forkhead Box (Fox) transcription factor (FOXM1) is a putative master regulator of tumor metastasis,¹⁴ and an important biomarker of lung cancer.¹⁵ Previous studies have shown that FOXM1 overexpression is associated with invasive/metastatic potential of NSCLC and poor prognosis.¹⁵ miR-149 was shown to inhibit NSCLC epithelial-to-mesenchymal transition (EMT) by targeting FOXM1.¹² However, the molecular mechanism involved in FOXM1 overexpression in NSCLC remains to be fully elucidated.

Herein, the goal of the study was to explore the role of HIPK3 as a diagnostic and therapeutic target in NSCLC. Human NSCLC tissues and cells were assessed for their HIPK3 levels, and manipulation of HIPK3, either by transfection of small-interfering RNA (siRNA) specific for HIPK3, or overexpression of HIPK3 by transfection of overexpression vector was performed in cells to clarify its role in lung cancer cell proliferation, migration and in vivo xenograft tumor growth. The interaction between HIPK3, miR-149, and FOXM1 was also investigated to elucidate the mechanism of HIPK3 regulation in NSCLC.

Materials and methods

Clinical samples

All studies that involve patient samples were approved by the Ethics Committee of The first affiliated hospital of Zhengzhou University. Tumor tissue specimens were collected from 25 patients pathologically confirmed with NSCLC. Samples were collected before the start of chemotherapy. Staging of the tissue specimens were performed according to WHO grade criteria and the tumor-node-metastasis (TNM) classification systems. After being obtained during operation, the tissue specimens were immediately frozen at −80°C. Written informed consents obtained from all patients. Specimens were also acquired from twenty healthy subjects, who were randomly selected from a cohort of participants undergoing periodic health examination in the hospital during the same time. No cancer was present in the healthy controls and they were individually matched to cases by age (±5 years) and gender.

Cell culture

One normal human bronchial epithelial cell line, 16HBE and four NSCLC adenocarcinoma cell lines, including SPC-A1, A549, NCI-H1299, and NCI-H1650, were all acuired from the Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China). RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 mg/ml streptomycin (Thermo Fisher Scientific, Inc., Waltham, MA, USA) was used for cell culture at 37°C/5% CO₂.

Trypan blue exclusion assay

Cells collected by trypsinization and centrifugation were stained with 0.4% trypan 107 blue (Sigma Aldrich, MO, EUA) for 5 minutes. The number of dye-excluding (live) cells and positively stained (dead) cells was determined using a Neubauer’s chamber.

qRT–PCR

We used NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific) to extract cytoplasmic and nuclear fractions. RNA from the nuclear and cytoplasmic fractions and whole-cell lysates were extracted using the TRIzol agent (Life Technologies, Carlsbad, CA). Two milligrams of total RNA was incubated 20 min at 37°C with or without 3U RNase R (Epicenter Technologies, Madison, WI), followed by purification using an RNeasy MinElute cleaning Kit (Qiagen). For miRNA quantification, TaqMan MicroRNA assays (Life Technologies) were used, with small nuclear U6B (RNU6B) RNA as the house-keeping gene. cDNA synthesis was conducted with the PrimeScript RT Master Mix (Takara, Dalian, China) using 500 ng RNA. Real-time PCR was performed using SYBR Premix Ex Taq II (Takara). Particularly, the divergent primers annealing at the distal ends of circRNA were used to measure circRNA content. To the quantity of RNA, the purified PCR product that corresponds to the circHIPK3 sequence was serially diluted for plotting a standard curve.

Vector construction

The DNA region coding circHIPK3 with its flanking introns was amplified using PrimerSTAR Max DNA Polymerase Mix (Takara) to recapitulate circRNA, which was inserted into pcDNA3.0 vector. circHIPK3 fragment 3ʹ-UTR was inserted to the region downstream a cytomegalovirus promoter-driven firefly luciferase (FL) cassette in the pCDNA3.0 vector. Mutagenesis was performed within each miRNA-binding sites in circHIPK3 sequence using Mut Express II Fast Mutagenesis Kit (Vazyme, NanJing, China). All constructs were verified by sequencing.

Oligonucleotide transfection

siRNA and miRNA mimics were synthesized by Ribobio (Guangzhou, China). Lipofectamine RNAiMax (LifeTechnologies) was used for cell transfection.

CCK-8 assay and transwell assays

CCK-8 kit (Doindo, Japan) was used to evaluate the proliferation of Huh-7, HCT-116, HeLa cells. Transfected cells (3.5 × 10³ in 100 μL culture medium) were incubated in three separate wells in 96-well plates. The CCK-8 reagent (10 μL) was added to each well at 0, 24, 48, 72 and 96 h, followed by incubation at 37°C for 2 h. The absorbance at 450 nm was measured using a microplate reader (Synergy4; BioTek, Winooski, VT, USA). Costar Transwell plates (Coring, NY) were used to evaluate cell migration and invasion.

Biotin-coupled miRNA capture

HEK-293 T cells were transfected with 3ʹ-biotinylated miR-149 mimic or control RNA (RiboBio) at a final concentration of 20 nM for 1 day. The biotin-conjugated RNA complex was pulled down after incubation of the cell lysates with streptavidin-coated magnetic beads (Life Technologies). The level of circHIPK3 in bound fractions was evaluated by qRT–PCR analysis.

Transfection of miRNA mimics, inhibitors or plasmid vectors

MiR-149 inhibitor (anti-miR-149) or mimics and the negative control oligonucleotides (miR-NC mimics or anti-miR-NC) were synthesized by Ambion Inc. (Austin, TX, USA). The plasmid encompassing the small hairpin RNA (shRNA) specific to FOXM1 (pSil/shFOXM1) and control vector (pSil/shcontrol) were constructed. PCR was used to generate the open reading frame of FOXM1, which was then inserted into the pEGFP-C1 expression vector, yielding pEGFP/FOXM1. The successful construction of the recombinant vector was confirmed by enzyme digestion and DNA sequencing. Lipofectamine™ 2000 (Invitrogen, USA) was used for transfection in accordance to the instructions from the manufacturer. The G418 selection marker of the transfected plasmids to cells were used for selection, by incubating at medium containing 400 mg/mL G418 for 4 weeks. Single clones were obtained and maintained in 100 mg/mL G418.

Tumorigenesis and metastasis experiments

BALB/c mice (male, 6 weeks old) were subcutaneously injected A549 cells of 5 × 10⁶, which were transfected with oligonucleotides, suspended in PBS. Tumor sizes were injected twice a week. Three mice from each group were euthanized every two weeks, for a duration of 6 weeks. The harvested tumors were embedded in paraffin, followed by Ki67 immunohistostaining using standard procedures. To construct the metastasis model, 2 × 10⁶ A549 cells were injected into each mouse via tail vein, and Ctrl-A549, NC-549, si-circHIPK3-A549, or circ-HIPK3-A549 cells were injected every week via tail vein injection. Three mice from each group were euthanized at 2 or 3 weeks after cell injection to monitor the appearance of metastases. All remaining mice were euthanized at 4 weeks after the injection of A549 cells. To clearly observe the metastasis nodules, the lungs were fixed in Bouin’s solution (Sigma). The observable metastatic foci on lung surface were counted

Statistical analysis

All data were shown as mean ± SD, based on three independent experiments. Statistical significance was analyzed by one-way ANOVA or two-tailed Student’s t-test, followed by post-hoc Bonferroni’s test using SPSS 16.0 (SPSS Inc., USA). P < .05 were considered statistically significant.

Results

NSCLC is characterized by circRNA-HIPK3 upregulation

qRT-PCR was performed to detect the circRNAs expression in NSCLC tissues and cells, which showed that circRNA-HIPK3 was significantly upregulated in NSCLC tissues and NSCLC cells compared to normal tissue (Figure 1(a), p < .05) and human normal bronchial cell line 16HBE respectively (Figure 1(b), p < .05).

Figure 1.

circRNA-HIPK3 is upregulated in NSCLC. miR-149 is downregulated in NSCLC (a). circRNA-HIPK3 was upregulated in primary NSCLC tissues when compare with normal tissues. (b). TaqMan-based qRT-PCR analysis showing circHIPK3 upregulation in most NSCLC cell lines compared with human normal branch epithelial cell line 16HBE. *: p < .05. Analysis of miR-149 level in primary (c) NSCLC tissues and (d) NSCLC cell lines. E. qRT–PCR analysis of circHIPK3 level in the captured fractions of the HEK-293 T-cell lysates after transfection with 3ʹ-biotinylated miR-149 or control RNA (NC). F. Reversal of linear isomer HIPK3 overexpression by co-transfection of miR-149 mimics or si-FOXM1. *p < .05, n = 5.

Further, detection of the expression level of miR-149 in NSCLC tissue and cell lines suggested that miR-149 expression was significantly decreased in primary NSCLC tissues and cells (Figures 1(c,d)). Using a biotin-conjugated miR-149 mimic, an over five-fold enrichment of circHIPK3 levels in the miR-149-captured fraction was observed compared with the negative control (Figure 1(e)). As shown in Figure 1(f), circRNA-FHIPK3 induced the upregulation linear isomer HIPK3, which however could be partially reversed by co-transfection of miR-149 mimics or specific HIPK3 siRNAs.

Circhipk3 silencing inhibits human cell proliferation

The upregulation of circHIPK3 in NSCLC prompted us to investigate the effects of circHIPK3 on cell phenotypic changes. First, the expression of both circHIPK3 and HIPK3 mRNA were silenced by RNA interference, using small interfering RNAs (siRNA) to target sequence in a circularized exon shared by both circular and linear species. As a control, a nonspecific siRNA was also employed. Expectedly, the siRNA targeted to exonic sequences shared by both the circular and linear species effectively inhibited the expression of both transcripts (Figure 2(a)). The full-length cDNA segment of circRNA-HIPK3 from A549 cells was amplified by PCR and further cloned into the expression vector (Figure 2(b)). qRT-PCR showed that circRNA-HIPK3 vector transfection significantly increased the level of circRNA-HIPK3 in A549 (p < .05). Cell proliferation assay indicated that the downregulation and upregulation of circHIPK3 significantly suppressed and enhanced A549 cell proliferation, respectively (Figure 2(c), p < .05). The trypan blue exclusion assay (Figure 2(d)) and transwell assay (Figure 2(d,e)) also revealed that the proliferation and migration of A549 cells was impaired after circHIPK3 knockdown.

Figure 2.

Regulation of proliferation by circHIPK3. A. qRT–PCR analysis of circHIPK3 and HIPK3 mRNA in A549 cells treated with siRNA. n = 3. B. qRT-PCR analysis of circHIPK3 expression transfection of circRNA vector. C. CCK-8 proliferation assay of A548 cells. D-E. Transwell assay of NSCLCs. *p < .05, n = 5.

Circhipk3 promotes in vivo tumorigenesis and metastasis

To evaluate the effect of circHIPK3 in vivo, A549 tumor xenografts were constructed in nude mice, using A549 cells transfected with control RNA (Ctrl-A549), non-coding RNA (NC-549), si-circHIPK3 (si-circHIPK3-A549) or circHIPK3 overexpression vector (circ-HIPK3-A549). Two weeks after tumor inoculation, the tumor sizes were monitored twice a week for a duration of 6 weeks. Tumor mass and tumor volume were significantly inhibited by circ-HIPK3 siRNA compared with the control groups (Figure 3(a,b), p < .05). Immunostaining of Ki67, a proliferation marker, indicated that following circ-HIPK3 siRNA transfection, tumor demonstrated a reduced cell proliferation activity (Figure 3(c)).

Figure 3.

The circ-HIPK3 promotes A549 tumor formation by inducing cell proliferation. (a) Tumor size measurement of A549 cells with mock, circ-HIPK3 or si-circHIPK3 transfection (n = 5 per group). (b) Macroscopic photograph of tumors harvested at 6 weeks after tumor inoculation. (c) Ki67 immunohistological staining in the tumors. p < .05, n = 5.

Circrna-hipk3 promoted cell proliferation and invasion through sponging mir-149 in NSCLC cells

Given the prominent regulatory role of cic-HIPK3, we then proceeded to determine its functional mechanism in NSCLC cells. To investigate the role of miR-149 during circHIPK3 mediated anti-tumor effects, we generated miR-149 mimics (Figure 4(a)), and then analyzed cell proliferation, migration, and invasion after co-transfection of miR-155 mimics and miR-149. Our results showed that cell proliferation and migration, which were promoted by circRNA-HIPK3 overexpression, was markedly reversed by co-transfection of miR-149 mimics, as revealed by trypan blue exclusion assay (Figure 4(b)) and transwell assay (Figure 4(c,d)).

Figure 4.

circRNA-HIPK3 promoted cell proliferation and invasion through sponging miR-149 in NSCLC cells. A. miR-149 expression level was detected by RT-qPCR in A549 cells transfected with miR-negative control (NC) or miR-149 mimic. B. Proliferation assessed using a CCK-8 kit in cells transfected with circHIPK3 or miR-149 (10 nM) as indicated for 4 days. C. HSCLCs proliferation was detected by calculating the percentage of EdU positive cells. *p < .05, n = 5.

FOXM1 is a target of mir-149 in NSCLC cells

We next investigated the targets of miR-149 that function in NSCLC cells. FOXM1 was investigated as a target of miR-149. The wild-type (3ʹUTR-wt) or mutant (3ʹ-UTR-mut) 3ʹ-UTR of FOXM1 was cloned into a luciferase reporter vector pLUC. The luciferase reporter assay was conducted to analyze whether miR-149 can directly regulate FOXM1 expression in A549 cells. A pronounced reduction in the luciferase activity of pLUC/FOXM1-3ʹ-UTR-WT was observed after transfection of miR-149 mimics, but not miR-NC mimics (p < .01)(Figure 5(a)). In comparison to anti-miR-NC, the transfection of anti-miR-149 into A549 cells resulted in a significant increase in luciferase activity (p < .05, Figure 5(b)). Whereas, the luciferase activity of pLUC/FOXM1-3ʹ-UTR-mut vector was not affected by co-transfection with miR-149 mimics or inhibitor (p > .05). Subsequently, changes of FOXM1 expression in NSCLC cells after ectopic overexpression or silencing of miR-149 was analyzed, and we showed that upregulation of miR-149 could significantly attenuate the expression of FOXM1 expression in mRNA and protein levels in A549 cells (p < .01; Figure 5(c)). In the meantime, miR-149 knockdown could lead to the FOXM1 upregulation in mRNA and protein levels in LoVo cells (p < .01; Figure 5(c)). These data implicated that miR-149 targets FOXM1 in A549 cells.

Figure 5.

Validation of FOXM1 was identified as a direct target of miR-149 in NSCLC cells. A. Luciferase assay of A549 cells co-transfected with pLUC/FOXM1-3ʹ-UTR-wt or pLUC/FOXM1-3ʹ-UTR-mut vector and miR-149 mimics (or miR-NC mimics) or miR-149 inhibitor (or miR-NC inhibitor). Luciferase activity was normalized to Renilla luciferase activity. B. and C. qRT-PCR and Western blot analyses of FOXM1 expression level. *p < .05, n = 5.

Silencing of FOXM1 significantly inhibits NSCLC migration and

To further explore the role of FOXM1 in NSCLC migration and invasion, A549 cells were stably transfected with pSil/shFOXM1 to induce FOXM1 knockdown in NSCLC cells. qRT-PCR was used to confirm the efficient knockdown of FOXM1 (Figure 6(a)). Next, CCK8 assay (Figure 6(b)) and trypan blue exclusion assay (Figure 6(c)) indicated that FOXM1 silencing resulted in decreased A549 cell proliferation. In addition, transwell assay suggested that FOXM1 silencing could considerably reduce the migration and invasion of NSCLC cells (Figure 6(d,e)). A549 cells were then transfected with miR-149 inhibitor, and downregulation of FOXM1 was shown to attenuate the proliferation enhancement induced by miR-149 inhibitor in A549 cells (Figure 6(f-i)).

Figure 6.

Inhibits migration and invasion of NSCLC cells induced by FOXM1 silencing. (a). qRT-PCR analysis of FOXM1 mRNA expression in A549 cells with pSil/shFOXM1 or FOXM1 overexpression vector transfection. (b). CCK-8 assay of cells transfected with miR-149 inhibitor (10 nM) and shFOXM1 for 4 days. (c). CCK-8 assay of cells transfected with FOXM1 plasmid for 4 days. (d). NSCLC proliferation was detected by transwell assay in in cells transfected with miR-149 inhibitor(10 nM) and shFOXM1 as indicated. (e). HSCLCs proliferation was detected by transwell assay as indicated. *p < .05, n = 5.

Discussions

Here we report that upregulation of circHIPK3 is a characteristic of NSCLC, as verified both in tissues and cell levels. This data implicated that high levels of circHIPK3 is presumably associated with the aggressiveness of NSCLC. This data implicates the use of circHIPK3 as a novel biomarker of NSCLC, potentially providing another circRNA as a diagnostic tool of this highly lethal disease.^16,17 In line with this, knockdown and overexpression of HIPK3 inhibited and promoted the proliferation and migration of NSCLC cells, respectively. In vivo, HIPK3 silencing also retarded tumor growth. Together, these evidences suggested that HIPK3 silencing could be an effective strategy to impede the progression of NSCLC. It is worth noting that here we did not investigate whether manipulation of circHIPK3 expression led to sensitization of NSCLC to chemotherapy drugs, such as cisplatin, oxaliplatin and carboplatin. While studies on the roles of circRNA on chemotherapy sensitivity is relatively rare, some studies have suggested the potential regulation of cancer chemotherapy sensitivity by circRNAs.¹⁸ Therefore, studies on whether our circHIPK3 silencing strategy can reverse NSCLC resistance are warranted.

Indeed, in gallbladder cancer, it was shown that HIPK3 silencing could potently exert anti-tumor effects.^8,9 Our results are also consistent with a previous study showing that HIPK3 is a therapeutic target in lung cancer,¹¹ which however, did not investigate the therapeutic potential of circHIPK3 silencing in vivo. Therefore, our study for the first time provided evidences that of circHIPK3 silencing, a gene therapy approach, is effective in vivo to suppress tumor growth. However, it should be noted that further improvement can be made in regard to the delivery method of circHIPK3 siRNA, as the in vivo tumor model used in our study is established by inoculating cells with HIPK3 siRNA transfection, which does not represent the clinical condition, where genes should be delivered after tumor was found. The use of more sophisticated delivery systems may be required.

We also showed that the effects of HIPK3 are attributable to the mediation by miR-149. In NSCLC, contrary to the upregulation of HIPK3, miR-149 was downregulated. In the biotin-coupled miRNA capture assay, the direct interaction between HIPK3 and miR-194 was verified. While upregulating HIPK3 promoted NSCLC cell proliferation, co-transfection with both HIPK3 and miR-194 abrogated the tumor-promoting effects of HIPK3. Altogether, these data corroborated that HIPK3 and miR-149 exert opposing roles in NSCLC and HIPK3 sponges miR-149 to promote NSCLC. This observation echoes previous findings that circRNAs are potent cancer regulators by sponging a large variety of miRNAs,^19,20 including miR-149.¹³ Other miRNAs that were shown to be sponged by HIPK3 include miR-124 and miR-558.^8,9 It is conceivable that these miRNAs may also be altered by HIPK3.

Further, our results indicate that FOXM1 is a target of miR-149 in NSCLC. Previous studies confirmed that there is a miR-149 binding site on the 3ʹ-UTR of FOXM1 and the interaction between miR-149 and FOXM1 is accountable for the tumor suppressive role of miR-149 in cancer.¹² Hence, the mechanism of circ-HIPK3 in regulating NSCLC is mediated by miR-194 and eventually transduced to FOXM1. Given the important role of FOXM1 in lung cancer, our study also provided a valuable tool for regulating FOXM1 expression in vivo.

Conclusion

In conclusion, our study confirmed that circHIPK3 upregulation is a potential biomarker for NSCLC and circHIPK3 silencing is of great therapeutic value in vivo. Further testing of this biomarker is needed to translate this new circRNA in clinics to improve the survival of NSCLC patients.

Consent for publication

All authors have read and approved the final manuscript.

Competing interests

The authors declare that they have no competing interests.

Availability of data and materials

The analyzed data sets generated during the study are available from the corresponding author on reasonable request.