IRF8 induces senescence of lung cancer cells to exert its tumor suppressive function

ABSTRACT

Lung cancer is the leading cause of cancer-related deaths worldwide. However, tumor suppressor genes remain to be systemically determined for lung cancer. Here we report interferon regulatory factor 8 (IRF8), a member of the IRF family of transcription factors, as a potent lung tumor suppressor gene. Expression of IRF8 is frequently diminished in lung tumoral tissues and is associated with prognosis of non-small cell lung cancer (NSCLC) patients. Ectopic expression of IRF8 suppresses the NSCLC cells proliferation in vitro and tumorigenic potential in vivo. More importantly, forced expression of IRF8 through infection of recombinant virus inhibits lung tumorigenesis in genetically engineered mouse model (GEMM). Mechanistically, IRF8 inhibits AKT signaling and promotes accumulation of P27 protein, which results in senescence of lung cancer cells. Ectopic expression of IRF8 in tumor cells leads to regression of lung cancer tumor nodules in a xenograft tumor model. Our data, therefore, solidly shows IRF8 to be a lung cancer suppressor gene and may denote an opportunity for therapeutic intervention of NSCLC.

Introduction

Lung cancer is the leading cause of cancer-related mortality worldwide [1]. Non-small cell lung cancer (NSCLC) is the most frequently diagnosed pathological type of lung cancer, accounting for approximately 85% of all lung cancer cases [2]. Despite of advances in multi-modality therapies for lung cancer, the prognosis of NSCLC remains disappointingly low, with an overall 5-y survival rate of only around 17% [3–5].

Targeting therapy has been 1^st-line therapy in lung cancer clinic. Unfortunately, a significant portion of patients positive for mutation of an oncogene fail to respond to the corresponding targeting therapy [6]. Increasing lines of evidences suggest that tumor suppressor gene (TSG) inactivation plays an important role in lung tumorigenesis [7] and negatively impacts on the therapeutic effect of targeting therapies [8]. Elucidation of molecular mechanism underlying the tumor suppressive roles of TSGs is expected to provide new insight into NSCLC pathogenesis and may shed light on new therapeutic strategies.

p27^Kip1 (p27) protein belongs to the cyclin-dependent kinase inhibitors, functioning to inhibit G1-to-S phase transition of the cell cycle and induce cellular senescence [9,10]. Diminished expression of p27 negatively correlated with prognosis of patients of several human cancers, including NSCLCs [11,12].

Phosphorylation of serine/threonine kinase AKT at S473 in the carboxy-terminal hydrophobic motif leads to AKT activity [13,14]. AKT is responsible for phosphorylating various protein targets important for cell survival, proliferation and motility [15,16]. For example, the activated AKT pathway results in the activation of nuclear transcription factor kappa B (NF-κB), which mediates oncogenic transformation by Akt in some cases [17–22]. Consistently, NF-κB hyperactivation is detected in NSCLC tissues and not in paired normal tissues [23], which is associated with unfavorable prognosis of the patients [24–26].

Interferon regulatory factor 8 (IRF8), also named as interferon consensus sequence-binding protein (ICSBP), is a member of the IRF family of transcription factors induced by interferon gamma [27]. Numerous studies have demonstrated that IRF8 is implicated as a TSG in certain hematopoietic cancers and solid tumor [28,29]. Recent research indicated that IRF8 expression was frequently silenced by aberrant CpG methylation of promoter region in NSCLC [30]. However, its role and underlying mechanism of IRF8 function in NSCLC remains unclear.

In our current study, we found that IRF8 was a clinically relevant TSG in NSCLC. Ectopic expression of IRF8 inhibited NSCLC cells growth in vitro through senescence induction. We further showed that ectopic expression of IRF8 inhibited tumor growth in vivo. Overexpression of IRF8 down-regulated the phosphorylation of AKT and elevated p27 expression levels in NSCLC cells. Our data, therefore, solidly showed IRF8 to be a lung cancer suppressor gene and may denote an opportunity for therapeutic intervention of NSCLC.

Results

Identification of IRF8 as an essential and clinically relevant TSG in lung cancer

IRF8 has been reported as a TSG in various cancer types. We checked mRNA level of IRF8 in lung cancer tissue in open database (http://xena.ucsc.edu/compare-tissue/) and found lower IRF8 expression in lung tumor tissues than in paired para-tumoral tissues (Figure 1(a)). Using clinic samples, we confirmed a lower expression of IRF8 at both protein and mRNA level in NSCLC tumors than their paired para-tumoral tissues. (Figure 1(b,c)). Analysis of TCGA data revealed that IRF8 expression level was significantly correlated with prognosis of NSCLC patients of all stages (Figure 1(d), left; HR = 0.73, P = 1.3e-06). Of note, the same trend was likewise highly significant in stage I NSCLC patients (Figure 1(d), left; HR = 0.35, p = 3.2e-11), suggesting a critical role of IRF8 in early stages of NSCLC. Taken together, these data strongly suggested that IRF8 was a clinically relevant TSG.

Figure 1.

IRF8 as an essential and clinically relevant tumor-suppressor gene in NSCLC patients.

(a) IRF8 mRNA expression in TCGA lung cancer tissue and GTEX lung tissue. (b) Western blot analysis of IRF8 expression in the indicated lung. T: tumor, P-T: para-tumor. (c) mRNA expression of IRF8 in the indicated lung. (d) K-M survival analysis in NSCLC patients (http://kmplot.com/analysis/).

Ectopic expression of IRF8 inhibited NSCLC cell proliferation in vitro

We then asked whether IRF8 exert its tumor suppressive function through a cell-autonomous mechanism. To this end, we generated lung cancer cell lines for doxycycline (DOX) inducible expression of IRF8. Western analysis and quantitative reverse transcriptase PCR (qRT-PCR) revealed very low IRF8 expression in A549 and H460 cell lines (Figure 2(a,b)). Strong DOX-inducible expression of IRF8 was detected in both cell lines when stably infected with recombinant lentivirus derived from TetOne system (Figure 2(a,b)). We found that overexpression of IRF8 significantly inhibited cell growth of A549 and H460 through CCK8 assay (Figure 2(c,d)) and colony forming in 2D-plate (Figure 2(e,f)). The cell growth inhibitory effect was highly striking as exemplified by delayed cells growth of lung cancer cells in response to DOX treatment in 10-cm plates (Figure 2(g,h)). Taken together, our data solidly showed that IRF8 was a potent TSG in NSCLC and functioned through a cell-autonomous manner.

Figure 2.

Ectopic expression of IRF8 inhibited NSCLC cell proliferation in vitro.

(a & b) Generation of A549-Teton-IRF8 and A549-Teton-IRF8 stable cell line. A549 and H460 were infected with PLVX-Teton-IRF8 lentivirus for 24 h, followed by cultured in complete medium containing puromycin (1 µg/mL) for another 3 days. Then, the cells after puromycin selection were passed in six well plate (2x10⁶), left untreated or treated with DOX for 48 h before western blot analysis (a) and QPCR analysis (b). (C&D) Cell viability was obviously inhibited after IRF8 ectopic expression. The A549-Tet-IRF8 (c) and H460-Tet-IRF8 (d) (1x10³) were treated without or with DOX (1 µg/mL) for 48 h, then cell viability was detected by CCK-8 assay kit for indicated time points. (E&F) IRF8 inhibited colony forming of A549 (e) and H460 (f) cells. The indicated IRF8-inducible cell lines were treated with or without DOX for about 10 days before colony forming assay. (g & h) overexpression of IRF8 repressed A549 (g) and H460 (h) growth. IRF8-inducible cell lines were left untreated or treated with DOX for 4 days, then cells were subjected to microscopy. Data are means ± SD of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 (student’s t-test).

IRF8 is a potent TSG in vivo

We then went on to validate the tumor suppressive function of IRF8 in vivo. We inoculated A549-Teton-IRF8 and H460-Teton-IRF8 cell lines subcutaneously in nude mice, respectively, and randomized these mice for treatment with DOX-containing diet or control diet when tumors reached a volume of around 80 mm³ (Figure 3(a,b)). Tumors in control-diet-fed mice continued growing. In stark contrast, obvious tumor shrinkage was seen in DOX-treated A549 group (Figure 3(c,d)) or H460 group (Figure 3(e,f)). Of note, DOX treatment was not toxic and the weight of mice remained constant (data not shown). After 15 d (A549-Tet-IRF8) or 8 d (H460-Tet-IRF8) treatment, the average weight of tumor in DOX-treated group was around 0.29 g (A549-Tet-IRF8) or 0.72 g (H460-Tet-IRF8) versus an average of around 0.12 g (A549-Tet-IRF8) or 0.29 g (H460-Tet-IRF8) in the control diet group.

Figure 3.

IRF8 inhibited lung tumor growth.

NSCLC cells (2 × 10⁶) were implanted subcutaneously into the flank of 6-wk-old female BALB/c nude mice. When the tumors reached a volume of around 80 mm³, animals were randomized into two groups (n = 5 each): with DOX food group (+DOX), normal food group (-DOX) for around 14 days (A549-tumor xenograft) or 7 days (H460-tumor xenograft) before sacrificed. (a & b) IRF8 had no effects on mice growth. (c & d) The growth of DOX-treated tumors was strongly suppressed compared with control tumors. Tumor growth was recorded every 3 days (c) or 2 days (d) by measuring its diameter with Vernier caliper. Tumor volume was calculated by tumor volume (cm³) = D× d²/2, where D is the longest and d is the shortest diameter, respectively. (e & f) DOX-treated tumors growth was drastically inhibited relative to control tumors. (g) Ectopic expression of IRF8 slowed down the lung tumor initiation in TetO-Kras^G12D/CC10rtTA mouse model. TetO-Kras^G12D/CC10rtTA mice were infected with lenti-Puro (control group) or lenti-IRF8 (IRF8 group) by intranasal, then mice were fed with DOX food for approximately 2 months before CT scan. MRI image of lung of control group (up panel, Control) and IRF8 group (lower panel, IRF8) of lung cancer bearing mice; (h) Representative images of Hematoxylin and eosin (H&E) staining of the lung tissues obtained from lenti-Puro and lenti-IRF8 treated Kras^G12D/CC10rtTA mice. Data are means ± SD of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 (student’s t-test)

To further validate the tumor suppressive function of IRF8 in vivo, we checked the tumor formation after overexpress IRF8 in lung of transgenic mouse model. Earlier, we reported a lung cancer mouse model with DOX inducible Kras^G12D in lung epithelial compartment [31], namely Kras^G12D/CC10rtTA bi-transgenic mice (referred to hereafter as KC). KC mice developed lung cancers after feeding with a DOX diet for 2 to 3 months but remained lung cancer free if fed with normal diet. We then delivered Teton-IRF8 DNA elements into the lung epithelial cells of KC mice through intra-nasal instillation of recombinant lentivirus as reported earlier [32], such that virus-infected mice started to express Kras^G12D and IRF8 in lung epithelial cells when fed with DOX diet.

KC mice received Teton-Puro lenti-virus (serving as negative control) began panting and exhibited hunched posture 3 months after DOX diet treatment, indicative of severe lung disease. Computed tomography (CT) imaging revealed heavy tumor burdens in both lungs of these mice at this stage (Figure 3(g)). Pathological analysis revealed poorly differentiated lung adenocarcinomas with features of diffused bronchial adenocarcinomas. We found around 5% of these tumors are stage IV adenocarcinoma (Figure 3(h)). In Stark contrast, Teton-IRF8 infected KC mice (referred to as KCI) looked largely normal at this stage. Consistently, these mice harbored significantly lower burden of lung cancers as evaluated by CT imaging (Figure 3(g), lower panel). Strikingly, we found that KCI mice had dramatically lower tumor numbers on lung (Figure 3(h), lower panel, left). Pathological analysis revealed a significantly lower percentage of tumor areas (Figure 3(h), lower panel, middle) in these mice. We also found that at this stage, no stage IV lung adenocarcinomas were detected in these mice (Figure 3(h), lower panel, right). Taken together, our data showed potent tumor-inhibiting function of IRF8 in vivo.

Ectopic expression of IRF8 leads to cellular senescence of NSCLC cell lines

We then asked by which means IRF8 exerted its tumor suppressive function. Cancer stem cells play a critical role in tumorigenesis, relapse and metastasis of lung cancer [33]. We performed sphere assay using A549-Teton-IRF8 and H460-Teton-IRF8 cell lines and found that overexpression of IRF8 did not influence sphere-forming ability of A549 and H460 (Figure 4(a)). Likewise, ectopic expression of IRF8 had no inhibitory effects on the 3D colony formation in soft agar experiment (Figure 4(b)). These data indicated that IRF8 did not affect NSCLC cell line on stemness.

Figure 4.

Ectopic expression of IRF8 does not affect stemness and apoptosis of lung cancer cell.

(a) Effects of IRF8 on sphere colonies forming ability. Representative images of sphere assay of A549-Tet-IRF8 and H460-Tet-IRF8 cells (up) and statistics of sphere formation (down). (b) Effects of IRF8 on colonies forming in soft agar. Representative images of sphere assay of A549-Tet-IRF8 and H460-Tet-IRF8 cells (up) and statistics of sphere formation (down). (c) IRF8 did not affect Caspase-9 and Caspase-3 cleavage. The indicated cells (1x10⁵) were treated with or without DOX (1 µg/mL) for 48 h before western blot analysis. Cell lysates were analyzed by immunoblots with the indicated antibodies.

Tumor suppressor genes (TSGs) inhibit tumor formation mainly through the induction of cell-cycle arrest, apoptosis and/or senescence [34]. To this end, we assessed the apoptosis of NSCLC cell in response to IRF8 overexpression. Western blot analysis revealed no obvious cleavage of Caspase-3 and Caspase-9 before or after DOX treatment (Figure 4(c)). Consistently, FACS analysis of annexin V and propidium iodide (PI) stained cells revealed no obvious change of death rate before and after IRF8 expression in A549 and H460 (data not shown). These data suggested that apoptosis was not involved in the process when IRF8 exerted its tumor suppressor function in lung cancer cells.

Interestingly, we found that IRF8 expression resulted in obvious increase of G0/G1 cells and concomitant decrease in S phase in both A549 and H460 (Figure 5(a,b)). Given that senescence demarcates a cell of stable exit of cell cycle and that senescence induction is an important mechanism of TSG functioning, we then checked whether A549 and H460 underwent senescence in response to IRF8 overexpression. β-Galactosidase staining assay revealed significantly enhanced the senescence signals of A549 and H460 cells (Figure 5(c)). In line with this, A549 and H460 exhibited enlarged, flattened and irregular shape after DOX treatment, a typical feature of senescent cell (Figure 5(c) highlighted with arrow head). EdU incorporation assay revealed that IRF8 expression almost completely eliminated DNA replicating events in lung cancer cell lines (Figure 5(d)). This stop of DNA replicating or cell proliferation was not due to DNA damage, as these cells did not show a higher level of Ƴ-H2AX (Figure 5(f)). To explore the possible molecular events associated with IRF8-induced senescence, we checked cyclin-dependent kinase (CDK) inhibitors (CKI) in these cell lines. qRT-PCR analysis revealed that IRF8 expression induced expression of P16, P21 and P53 (Figure 5(e)). Of note, while induction of P21 could be verified at protein levels, P53 remains constant in both cell lines in response to IRF8 expression (Figure 5(f)). Taken together, these results suggested that IRF8 exert its tumor suppressive function through induction of senescence of NSCLC cells.

Figure 5.

Overexpression of IRF8 leading to cellular senescence of NSCLC cells.

(a & b) IRF8 induces G0/G1 cell cycle arrest in NSCLC cell lines. Cell population under different stages of G0/G1, S, G2/M phase. A549-Tet-IRF8 and H460-Tet-IRF8 cells were treated without or with DOX (1 µg/mL) for 3 days, then FACS analysis was used to analysis cell cycle distribution (left) and statistics of cell cycle distribution (right). (c) Overexpression of IRF8-induced senescence of NSCLC cells. A549 cells and H460 cells were treated without or with DOX (1 µg/mL) for 48 h, then senescent cells were determined by senescence-associated β-galactosidase activity analysis (up) and statistics of senescence β-galactosidase staining positive cells (down). (d) IRF8 inhibited the rate of DNA synthesis of A549 and H460 cells. A549 cells and H460 cells were treated without or with DOX (1 µg/mL) for 72 h, EdU staining was used for cell proliferation. Green: EdU stained nuclei of proliferating cells. (e) IRF8-induced transcription of P16 and P21. RNA was extracted from A549-Teton-IRF8 treated with DOX (1 µg/mL) for indicated time points. Expression of the indicated genes was quantified through qPCR. (f) IRF8 promoted P21 expression in A549. A549-Teton-IRF8 were left untreated or treated with DOX for 72 h. Whole cell lysates were analyzed by immunoblots with the indicated antibodies.

Akt-P27 played a critical role in mediating IRF8-induced senescence

We went on to study the molecular mechanism underlying IRF8-induced senescence. To this end, we examine the changes of gene expression at transcriptomic level induced by IRF8 expression in A549 cell through RNA‐sequencing analyzes. We found that IRF8 significantly altered the expression of 154 genes in lung cancer cell lines (Figure 6(a)). Kyoto encyclopedia of genes and genomes (KEGG) analysis revealed the alteration of several pathways (Figure 6(b)). However, typical senescence pathway was not highlighted by KEGG analysis.

Figure 6.

Akt-P27 played a critical role in mediating IRF8-induced senescence.

(a) Heat map of mRNA expression of A549-Tet-IRF8 cells with or without DOX treatment. A549-Tet-IRF8 cells (2x10⁶) were left untreated or treated with DOX (1 µg/mL) for 48 h, then total RNAs were extracted for RNA-sequencing analysis. (b) Kyoto encyclopedia of genes and genomes (KEGG) pathway analyses. (c) Ectopic expression of IRF8 repressed phosphorylation of AKT and prompted an expression of P27. A549-Tet-IRF8 and H460-Tet-IRF8 cells (2x10⁶) were left untreated or infected with DOX for 3 days. Cell lysates were analyzed by immunoblots with the indicated antibodies. (d) IRF8 does not affect transcription of P27. RNA was extracted from A549-Teton-IRF8 and H460-Teton-IRF8 which were treated with DOX (1 µg/mL) for 48 h. Expression of the p27 was quantified through qPCR. (f) P27 shRNA inhibited IRF8-induced cellular senescence of A549. A549-Teton-IRF8 cells were infected with DOX-inducible P27 shRNA, after zeocin selection for 2 wk, cells were treated without or with DOX (1 µg/mL) for another 48 h, then senescent cells were determined by senescence-associated β-galactosidase activity analysis (left panel), statistics of senescence β-galactosidase staining positive cells (middle panel) and P27 mRNA expression (right panel).

Given that CKIs including P16 and p21 was upregulated by IRF8 (Figure 5(e)), and that AKT has been reported to regulate the activity of CKIs [35], we went on to check the impact of IRF8 on AKT activity. Interestingly, we found IRF8 expression drastically inhibited AKT phosphorylation (Figure 6(c)). P27 is a potent CKI, and underwent proteasomal degradation in response to phosphorylation by AKT [36]. Consistent with this, we saw P27 accumulation in lung cancer cells expressing IRF8 (Figure 6(c)). Importantly, this accumulation was not due to higher mRNA expression (Figure 6(d)). Critically, knockdown of P27 largely eliminated the ability of IRF8 to induce senescence in lung cancer cells (Figure 6(e)). Taken together, our data showed that IRF8 expression inhibited Akt activity to induce P27 accumulation and thus, the ensuing senescence of lung cancer cells.

Discussion

Targeting therapy and immunotherapy are 1^st-line therapies in lung cancer clinics. However, the response rate varies among ethnic groups. Functional status of tumor suppressor genes has been reported to impact on the therapeutic effect of both these therapies. Tumor suppressor genes remain to be systemically characterized. Earlier, we have reported genome-wide screening of tumor suppressor gene for lung cancer gene [37]. TGF beta pathway are the frequent target of tumor suppressor gene that we characterized. Interestingly, IRF8 functions through a distinct mode: it negatively regulates the Akt phosphorylation and activation.

In our Kras^G12D/CC10rtTA transgenic mouse model, we found that overexpression of IRF8 potently inhibited lung cancer development, strongly suggesting that IRF8 plays a critical role in the tumorigenesis. Using lung cancer cell lines established from late-stage lung cancer patients, such as A549 and H460 cells, we found that manipulating the expression level of IRF8 potently affected tumor growth, suggesting its role in progression of late-stage lung cancers. These data are in line with our finding that IRF8 functional status significantly correlated with prognosis of stage I lung cancer patients and all stage patients. It is therefore safe to conclude that IRF8 is a critical tumor suppressor for lung cancer.

Senescence could form a feed-forward loop: a senescent cell can release senescence-associated secretory phenotype (SASP) to induce senescence of neighboring cell; the SASP cytokines can then reinforce senescence of the original cell. Ectopic expression of IRF8 leads to senescence of lung cancer through inhibiting Akt activity. Akt can be positively regulated by upstream kinases such as Phosphoinositide 3-kinase, or post-translation modifications, such as ubiquilation [38], and negatively regulated by post-transcription modifications, such as hydroxylation [39]. Whether IRF8 inhibits Akt activity through any of these mechanism remains an interesting question. Yet, our research gives hint that IRF8 deficient lung cancer cells may have higher Akt activity, such that Akt inhibitors may help to treat such lung cancers.

Materials and methods

Constructs, reagents and antibodies

The pLVX-TetOne-Puro plasmid and pLVX-TetOne-Puro-IRF8 were purchased from Clontech; psPAX2 and pMD2.G were purchased from Addgene; Doxycycline hyclate (DOX, Sigma); Cell Counting Kit-8 (CCK8, Dojindo Molecular Technologies); Western blotting substrate (Millipore); Cell Signaling Senescence β-Galactosidase Staining Kit (CST); fetal bovine serum, DMEM/F12 medium, B27, RPMI medium 1640 and DMEM (Gibco); fibroblast growth factor (FGF) and epidermal growth factor (EGF) (PeproTech); RIPA lysis buffer (Santa Cruz); protease and phosphatase inhibitor cocktail (Roche); EdU cell proliferation Kit and cell cycle kit (Beyotime); Lipofectamine 3000 (Invitrogen); antibodies against IRF8, P21 (abcam); Phospho-Histone H2A.X (Beyotime); P27, P53, Caspase-9, Caspase-3, Phos-AKT and AKT (CST) were purchased from indicated manufacturers. A549, H460 and HEK293 were obtained from ATCC; BALB/c nude mice (Beijing Vital River Laboratory Animal Technology Co., Ltd.); Human lung cancer samples were from Sun Yat-sen University Affiliated Cancer Hospital.

Generation of A549-Tet-IRF8 and H460-Tet-IRF8 cell lines

The HEK293 cells were transfected with psPAX2 and pMD2.G together with PLVX-TetOne-Puro plasmid or pLVX-TetOne-Puro-IRF8 lentiviral plasmid, respectively. Twenty-four hours later, cells were incubated with new medium for another 24 h. The recombinant virus-containing medium was filtered and then added to indicated cells in the presence of polybrene (8 µg/mL). The infected cells were selected with puromycin (1 μg/mL) for 7 d before additional experiments.

Cell proliferation assay

For cell proliferation assays, 1 × 10³ cells were seeded in each well of 96-well plates and cultured overnight. RPMI medium 1640 containing 10% FBS medium was supplemented with or without 1 µg/mL DOX for 1, 3, 5 and 7 d. Proliferation activity was then determined using CCK8 cell counting kit following the manufacturer’s protocol.

Colony formation assay

1x10³ cells were plated into six well plate containing culture medium with 10% FBS. Cells were grown for 14 d with or without DOX treatment, then fixed with 4% formaldehyde in 1x PBS and stained with crystal violet. Colonies with the diameter larger than 200 µm were counted.

Sphere formation assays

1x10³ cells were seeded in Ultra-Low attachment six well plate in serum-free conditioned medium containing DMEM/F12 medium, 20 μL/mL B27 supplement, 20 ng/mL basic fibroblast growth factor (FGF) and 20 ng/mL epidermal growth factor (EGF). After about 2 wk, tumor spheres with the diameter over 100 µm were counted.

Senescence-associated β-galactosidase staining assay

β-galactosidase staining was performed using the Cell Signaling Senescence β-Galactosidase Staining Kit. Briefly, 2 × 10⁵ cells were seeded in each well of six well plates which received different treatments before β-galactosidase staining. For quantification of β-galactosidase staining positive cells, the blue positive cells in at least three randomly selected fields were counted at 200 × magnification under microscope.

Protein extraction and immunoblotting

Whole cell lysates were extracted by using the RIPA lysis buffer, protein concentrations were determined by the Bradford assay. Soluble proteins (30~40 μg) were subjected to SDS-polyacrylamide gel electrophoresis. Immunoblot analysis were performed as previously described [40].

In vivo xenograft model

All mice were housed in a pathogen-free environment in Jinan University. All experimental protocols were approved by the Institutional Committee for Animal Care and Use at Jinan University. All animal work was performed in strict accordance with the approved protocol. NSCLC cells (2 × 10⁶) were suspended in a 100 μL mixture of equal volumes of PBS and Matrigel were implanted subcutaneously into the flank of 6-wk-old female BALB/c nude mice. When tumors reached volume of around 80 mm³, the mice were randomly grouped and received DOX food or normal food every day. The volume of tumor was monitored as indicated in Figure 7 after 1–2 wk up to the end of the experiment. At the end of treatment, mice were sacrificed and tumors were picked, photographed and weighed.

Figure 7.

Schematics for the DOX treatments of nude mice.

2 × 106 A549-Teton-IRF8 or H460-Teton-IRF8 cells were inoculated into subcutaneously into the flank of 6-wk-old female BALB/c nude mice. Mice were fed with DOX containing diet or control diet when the tumor volume reached around 80 mm³. Tumor volume was monitored as indicated in Figure 3(c,d).

In vivo Kras^G12D/CC10rtTA mouse model

All the littermates of mice were randomly divided into two groups of three mice. Lenti-IRF8 or Lenti-puro viruses were delivered to mouse lung via nostril inhalation. Seven days after, all mice were fed with Dox containing food until the day of CT scan and sacrifice. Mouse were scanned by Super Nova CT (SNC-100, PINSENO HEALTHCARE) according to manufactures’ protocol. Mouse lungs were collected for hematoxylin and eosin (H&E) staining. For quantification of tumor burden in the Kras^G12D/CC10rtTA mice, we calculated the total size (mm²), relative tumor area and tumor numbers of all the tumor regions in H&E sections under a microscope.

Cell cycle analysis

The cells were collected at 72 h after DOX treatment and fixed with 70% ethanol at −20°C overnight. Cell cycle analysis was performed using the cell cycle kit according to the manufacturer’s specifications, and cell cycle distribution was analyzed using a BD Accuri™ C6 flow cytometer.

EdU staining assay

EdU staining assay were performed using BeyoClick™ EdU-488 cell proliferation Kit according to the manufacturer’s instructions. Briefly, A549-Teton-IRF8 and H460-Teton-IRF8 treated with or without DOX for 72 h, then EdU-488 staining solution was added to cells for 2 h under growth conditions. Cells were fixed and permeabilized before fluorescence microscopy analysis.

RNA sequencing

A549-Teton-IRF8 cells were treated with or without DOX for 72 h, then total RNA was isolated from cells using Trizol reagent (TAKARA, Japan). RNA libraries and RNA sequencing were performed by The Beijing Genomics Institute (BGI).

Realtime PCR

Total RNA was isolated for realtime PCR analysis to measure mRNA levels of the indicated genes. Data shown are the relative abundance of the indicated mRNA normalized to that of GAPDH. Primer sequences for IRF8, P16, P21, P53, P27 and GAPDH are as follows:

IRF8-Forward: AGGTCTTCGACACCAGCCAGTT

IRF8-Reverse: GCACGAGAATGAGTTTGGAGCG

P16-Forward: CTCGTGCTGATGCTACTGAGGA

P16-Reverse: GGTCGGCGCAGTTGGGCTCC

P21-Forward: AGGTGGACCTGGAGACTCTCAG

P21-Reverse: TCCTCTTGGAGAAGATCAGCCG

P53-Forward: CCTCAGCATCTTATCCGAGTGG

P53-Reverse: TGGATGGTGGTACAGTCAGAGC

P27-Forward: ATAAGGAAGCGACCTGCAACCG

P27-Reverse: TTCTTGGGCGTCTGCTCCACAG

GAPDH-Forward: TGTGGGCATCAATGGATTTGG

GAPDH-Reverse: ACACCATGTATTCCGGGTCAAT

RNA interference

Double-strand oligonucleotides corresponding to P27 were cloned into the PLko-Teton-shRNA-Zeocin Plasmid. The following sequences were targeted for P27 mRNA. shP27-AGCAATGCGCAGGAATAAGG.

shP27-transduced stable cells

The 293 cells were transfected with two packaging plasmids (pMD2.G and psPAX2) together with a control or Plenti-Tet-shP27-Zeocin plasmid. Twenty-four hours later, cells were incubated with new medium without antibiotics for another 48 h. The recombinant virus-containing medium was filtered and then added to A549-Teton-IRF8 cells in the presence of polybrene (8 µg/mL). The infected cells were selected with zeocin for 2 wk before additional experiments.

Statistical analysis

Statistics were performed using GraphPad Prism 7.04, student t-test was used to compare differences between two experimental groups. Data are presented as means ± SD and P < 0.05 was considered statistically significant.

Funding Statement

This work was financially supported by National Key R&D Program of China (Grant 2016YFC0905501), the National Natural Science Foundation of China (31800723) and the Fundamental Research Funds for the Central Universities (21618326), NSFC (81672309; 81972778), Guangzhou Research Project Of Science And Technology For Citizen Health (201803010124), the Fundamental Research Funds for the Central Universities (21619101).

Disclosure statement

No potential conflict of interest was reported by the authors.