FTO-mediated m6A modification of QPCT promotes tumorigenesis in lung adenocarcinoma by inducing macrophage chemotaxis and M2 polarization

Abstract

Lung adenocarcinoma (LUAD), the most common histologic subtype of lung cancer, is characterized by malignant and high infiltrating. Glutaminyl-peptide cyclotransferase (QPCT) promotes cancer progression by modifying the N-terminus of chemokine C-C motif ligand 2 (CCL2) to a pyroglutamate residue and stabilizing the protein. The role of QPCT in LUAD is still unknown. QPCT mRNA and protein expression were up-regulated in clinical LUAD specimens. By generating stable HCC44 cells with QPCT overexpression and stable A549 cells with QPCT knockdown, we found that QPCT knockdown notably inhibited LUAD cell proliferation. Additionally, QPCT deletion reduced the CCL2 contents in LUAD cell supernatants and inhibited phorbol 12-myristate 13-acetate-induced THP-1 macrophage chemotaxis toward tumor cells or tumor cell conditioned medium. The CD68⁺/CD206⁺ cell ratio was reduced by QPCT deletion in vitro. Nude mice inoculated with parental A549 or cells with stable QPCT knockdown were used to explore QPCT functions. The results were consistent with in vitro experiments. QPCT is predicted to be modified by N6-methyladenosine (m6A), and we performed methylated RNA immunoprecipitation PCR to confirm this result in A549 cells. The m6A demethylase fat mass and obesity-associated protein (FTO) mRNA expression positively correlated with QPCT mRNA in LUAD samples. FTO bound to QPCT mRNA and FTO knockdown affected QPCT mRNA stability. FTO deletion in HCC44 cells abrogated the macrophage recruitment and macrophage M2 polarization induced by QPCT overexpression. In conclusion, QPCT promotes tumorigenesis in LUAD by increasing macrophage recruitment and M2 macrophage proportion. This may be due to FTO-mediated demethylation increasing the QPCT mRNA stability.

Introduction

Lung cancer is one of the most common cancers in the world, and non-small cell lung cancer (NSCLC) accounts for more than 85% of human lung cancer cases [1]. Lung adenocarcinoma (LUAD), a major subtype of NSCLC, is found to be an aggressive cancer with the characteristics of insidious onset and high infiltration rate [2,3]. Despite recent advaces in therapy, patients with LUAD continue to have a high morbidity and mortality rate [4]. Therefore, it is crucial to study the molecular mechanisms underlying the development of LUAD.

The tumor microenvironment consists of a variety of non-malignant cells including leukocytes, fibroblasts, and endothelial cells, and non-cellular components including cytokines and chemokines. Tumor-associated macrophages (TAMs) are the most abundant infiltrating inflammatory cells in the tumor microenvironment, which are closely associated with the occurrence and development of various tumors [5]. TAMs are usually induced to M2-phenotype macrophages by cytokines secreted by tumor cells, such as IL-10, IL-4, and TGF-β. These M2 macrophages exert immunosuppressive effects in tumors, promoting tumorigenesis and progression [6]. TAMs have been considered potential targets for some tumor therapies, but their mechanism of action is not fully understood.

Our investigation originated from the seminal work of Logtenberg et al. [7], who identified glutaminyl-peptide cyclotransferase-like (QPCTL) as a critical enzyme for pyroglutamate modification of CD47 - a ligand of the myeloid checkpoint SIRPα. Their finding that QPCTL inhibition enhances neutrophil-mediated tumour killing [7] prompted us to explore whether other glutaminyl cyclases (QCs), particularly QPCT and QPCTL, might similarly regulate immune evasion or oncogenic pathways. Indeed, prior studies implicated both QCs in cancer progression: QPCTL through CD47-SIRPα axis modulation [7,8], and QPCT via CCL2-mediated mechanisms [9]. To prioritise between these paralogs, we noted that Kehlen et al. [10] reported QPCT (but not QPCTL) upregulation in thyroid carcinomas, suggesting its tissue-agnostic oncogenic potential. Extending this observation to lung adenocarcinoma (LUAD), we analysed RNA-seq data from three independent cohorts (GSE30219, GSE40791, GSE75037; total n = 307). Using stringent thresholds (FDR < 0.05, |log2FC| > 1), we consistently found QPCT - but not QPCTL - significantly upregulated in LUAD tumours versus normal tissues (P < 0.001, t-test), establishing QPCT as a LUAD-specific candidate. Therefore, we decided to focus on QPCT in LUAD. QPCT encodes human pituitary glutaminyl cyclase, which modifies the peptides and proteins by catalyzing N-terminal l-glutamine residues to pyroglutamic acid [11]. The post-translational modification protects peptides and proteins from proteolytic degradation by stabilizing them and is therefore important for their biological activity. In recent years, QPCT was reported to be differentially expressed in breast cancer [12], thyroid cancer [13], and acute myeloid leukemia [14] and played important roles. QPCT expression was upregulated in tissues from thyroid cancer patients, and the suppression of CCL2 signaling by QPCT knockdown inhibited thyroid cancer cell growth [9]. CCL2 is a chemokine known to recruit monocytes/macrophages to sites of inflammation. The elevated expression of CCL2 correlates with the accumulation of TAMs [15]. Whether the abnormal upregulation of QPCT expression can influence the development of LUAD and what relationships it has with TAMs were what we would like to explore.

N-6-methyladenosine (m6A) is the most prevalent, abundant, and conserved internal co-transcriptional modification in eukaryotic RNA [16]. Reports in recent years have shown that m6A RNA modifications play a crucial role in the occurrence and development of multiple human cancers [16], including LUAD [17-20]. The m6A demethylase fat mass and obesity-associated protein (FTO) played a pro-tumorigenic role in LUAD [21-23]. Whether m6A modification is involved in the translation and stabilization of QPCT remained largely unknown.

In this study, we explored the expression of QPCT in LUAD and its correlation between high/low expression and clinicopathologic characteristics in clinical LUAD and paracancerous samples. The human LUAD cell lines HCC44 and A549 with stable overexpression or knockdown of QPCT were established and used to investigate the effects of QPCT on LUAD and its mechanism in vitro and in vivo.

Materials and methods

Gene expression analysis through public databases

Gene expression data from the LUAD-related datasets GSE30219, GSE40791, and GSE75037 were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/). QPCT expression in LUAD and normal tissues was extracted from the Cancer Genome Atlas (TCGA) database and genotype tissue expression (GTEx) using the GEPIA website (http://gepia2.cancer-pku.cn/#index).

Human sample specimens

The use of human LUAD and paracancerous specimens was approved by the institutional review board of Harbin Medical University Cancer Hospital. We collected 40 pairs of frozen LUAD and paracancerous tissues for real-time quantitative polymerase chain reaction (Rt-qPCR) and western blot (WB) assays and 97 paraffin-embedded LUAD tissues for immunohistochemistry (IHC). Informed consent was obtained from patients involved in the study.

Cell culture and treatment

Human monocytic cell line THP-1 and the human LUAD cell lines HCC44 and A549 were obtained from iCell Bioscience (Shanghai, China). THP-1 cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium supplemented with 0.05 mM β-mercaptoethanol and 10% fetal bovine serum (FBS), HCC44 cells were maintained in RPMI-1640 medium supplemented with 10% FBS, and A549 cells were cultured in Ham’s F-12K medium supplemented with 10% FBS. All cells were cultured in a humidified incubator containing 5% CO₂ at 37°C. THP-1 monocytes were incubated with phorbol 12-myristate 13-acetate (PMA, 100 μM) for 24 h to obtain THP-1-derived M0 macrophages (THP-1-M0). To construct stably transfected cell lines, a vector control (LV-VE) or QPCT-overexpressing (LV-QPCT) lentivirus was used to infect HCC44 cells, while a negative control (LV-shNC) or QPCT shRNA (LV-shQPCT-1/2) lentivirus was used to infect A549 cells. Next, the puromycin was added to the medium to select the stably infected cells.

Animal study

Six-week-old male BALB/c nude mice were housed in a constant environment with a temperature of 22 ± 1°C and humidity of 55 ± 5% in a 12h-light/12h-dark cycle. They had free access to water and food. After one week of adaptive feeding, parental or QPCT-stable knockdown A549 cells (5 × 10⁵ cells) in the logarithmic growth phase were subcutaneously injected into the nude mice. The long diameter and short diameter of the tumor were recorded every three days when the tumors became visible to the naked eye, and tumor volume was calculated using the formula: short diameter² × long diameter × 0.5. The animals were sacrificed 28 days after injection, and the tumors were collected to take pictures, measure the tumor weight, and kept frozen or fixed with 4% paraformaldehyde for following testing. All animal experiments were approved by the institutional review board of Harbin Medical University Cancer Hospital. For euthanasia, the nude mouse was placed into a euthanasia chamber, then carbon dioxide was infused into the box at a rate of 50% of the replacement volume of the chamber per minute. When it was confirmed that the nude mouse was immobile, not breathing and with dilated pupils, the carbon dioxide was turned off, followed by observation for two min to confirm that the nude mouse was dead.

Rt-qPCR assay

The total RNA from paired LUAD and paracancerous specimens or cell samples was extracted using a TRIpure reagent (BioTeke, Beijing, China). cDNA was synthesized using an All-in-One First-Strand SuperMix reagent (Magen Biotechnology, Guangzhou, China). The real-time fluorescence quantitative PCR instrument (Bioneer, Daejeon, Korea) was used for Rt-qPCR analysis using 2× Taq PCR MasterMix (Solarbio, Beijing, China) and SYBR Green (Solarbio, Beijing, China). The housekeeping gene GAPDH was used as an internal reference gene, and a standard ΔΔCt method (2^-ΔΔCt) was used to calculate relative gene expressions. For clinical samples, the 2^-ΔCt method was used. The primer sequences are listed below: QPCT, F: GGCTGACTGGGTCTTGG, R: ATAGTGGCAGGCGAGGA. FTO, F: GAACACCAGGCTCTTTACG, R: ATGAACCCATCCCAACC. GAPDH, F: GAAGGTGAAGGTCGGAGTCA, R: GAGGTCAATGAAGGGGTCAT.

WB assay

Standard WB was carried out using protein lysates from clinical/nude mice LUAD specimens and HCC44/A549 cells, and the following antibodies. Primary antibodies against QPCT (A6711, Abclonal, Wuhan, China), CD206 (A11192, Abclonal, Wuhan, China), Arg1 (A22410, Abclonal, Wuhan, China), FTO (A1438, Abclonal, Wuhan, China), hCCL2 (ab214819, Abcam, Cambridge, UK), mArg1 (ab233548, Abcam, Cambridge, UK), GFP-tag (AE011, Abclonal, Wuhan, China), and GAPDH (60004-1-Ig, Proteintech Group, IL, USA) and secondary antibodies (anti-rabbit IgG, SE134, Solarbio, Beijing, China; or anti-mouse IgG, SE131, Solarbio, Beijing, China) were used.

Cell proliferation assay

HCC44 parental cells/cells stably overexpressing QPCT or expressing empty expression vector and A549 parental cells/cells stably transfected with QPCT/NC shRNA were employed to perform cell proliferation assays. Cell proliferation was evaluated by Cell Counting Kit-8 (CCK-8) assay and 5-ethynyl-2’-deoxyuridine (EdU) staining assay. For the CCK-8 assay, cells were seeded into a 96-well plate at a density of 5 × 10³ cells/well and cultured in the cell culture incubator for 0, 24, 48, and 72 h, respectively. Then, 10 μL CCK-8 solution (KeyGEN, Nanjing, China) was added into each well to incubate cells for 2 h, followed by measuring the absorbance at 450 nm using a microplate reader (BioTek, VT, USA). For EdU staining, a final concentration of 10 μM EdU solution (KeyGEN, Nanjing, China) was added to cells 24 h after replating. Next, staining was carried out according to the instructions of the manufacturer.

Cell co-culture

To test the effects of QPCT on macrophage chemotaxis, co-cultivation of THP-1-M0 and HCC44/A549 cells was performed in transwell chambers. THP-1-M0 cells were plated in the upper chamber, while stably transfected HCC44/A549 cells were seeded in the lower chamber or the lower chamber was filled with conditioned medium derived from the stably transfected HCC44/A549 cells mixed with medium at a 1:2 ratio. Cells were then co-cultured for 24 h. Next, cells on the transwell chamber were fixed with 4% paraformaldehyde, stained with crystal violet for 5 min, and finally, the cells on the lower membrane of the transwell chamber were counted under an inverted microscope. To test the effects of QPCT on macrophage polarization, HCC44/A549 cells were plated in the upper chamber, and THP-1-M0 cells were seeded in the lower chamber and cultured with medium supplemented with 20 μg/mL IL-4 and 20 μg/mL IL-13. THP-1-M0 cells maintained in the medium containing HCC44/A549 conditioned medium, IL-4 (20 μg/mL), and IL-13 (20 μg/mL) were assayed. At the endpoint of the experiment, cells were incubated with corresponding antibodies and subjected to flow cytometry to detect the ratio of CD68 and CD206 double-positive macrophages. Anti-human CD68-PE (PE-65187, Proteintech Group, IL, USA) and anti-human CD206-APC (E-AB-F1161E, Elabscience, Wuhan, China) were used.

Enzyme-Linked Immunosorbent Assay (ELISA)

Concentrations of CCL2 cell culture supernatants or conditioned medium were measured by Human CCL2 ELISA Kit (EK187) according to the manufacturer’s instructions (LIANKE Biotech., Hangzhou, China).

Immunohistochemistry (IHC) and immunofluorescence (IF)

IHC and IF were performed as previously described [24]. Briefly, 5-μm tumor sections from nude mice or LUAD patients were stained with Ki-67 antibody (27309-1-AP, Proteintech Group, IL, USA), QPCT antibody (PA5-52554, ThermoFisher Scientific, PA, USA), or CD68 antibody (28058-1-AP, Proteintech Group, IL, USA). The secondary antibodies HRP-labeled goat anti-rabbit IgG (#31460, ThermoFisher Scientific, PA, USA) and Cy3-conjugated goat anti-rabbit IgG (ab6939, Abcam, Cambridge, UK) were used. IHC scoring standards for QPCT staining intensity were performed as previously described [25].

Methylated RNA immunoprecipitation PCR/qPCR (MeRIP-PCR/qPCR) and RNA immunoprecipitation PCR (RIP-PCR)

MeRIP was performed using the riboMeRIP m6A Transcriptome Profiling Kit according to the recommended instructions from the manufacturer (Ribobio, Guangzhou, China). Next, the immunoprecipitated RNA was analyzed by PCR or RT-qPCR. RIP assays were performed using the EZ-Magna RIP Kit (Millipore, MA, USA) as previously described [26]. Briefly, A549 cells were harvested and lysed with RIP lysis buffer and immunoprecipitated using anti-FTO (A1438, Abclonal, Wuhan, China), and the retrieved RNA was analyzed by PCR. The primer sequences used are shown below: QPCT, F: AGAGGGACCGTGTAAAGA, R: GTTTTCCAACCGCTATTT.

Statistical analysis

Scatter plot with bar and line plot are presented as mean ± standard deviation (SD), while box plot presents median, quartiles, range and individual data points. One-way ANOVA was used to test the statistical significance among multiple groups. Student’s t-test was used to determine the statistical significance of comparisons between the two groups. Two-way ANOVA was used to compare multiple groups at different time points. The Chi-square test or chi-square test with continuity correction were used to analyze the correlation between QPCT expression and clinicopathological characteristics. Correlation analysis was performed using scatter plots and Pearson’s correlation analysis. A p value < 0.05 was considered to be statistically significant. GraphPad Prism 9.5 was utilized to perform statistical analyses and generate graphs.

Results

QPCT gene expression is upregulated in LUAD patient tissues

Two isoenzymes with glutaminyl cyclase activity, QPCT and glutaminyl-peptide cyclotransferase like (QPCTL) have been identified as key effectors in multiple cancers by participating in tumor immunity regulation [9,27]. To investigate the expression changes of QPCT/QPCTL in LUAD, we downloaded multiple LUAD-associated gene expression data from the GEO database and found that QPCT, but not QPCTL, was highly expressed in LUAD (Figure 1A). The high expression of QPCT was also validated by metadata obtained from Gene expression profiling interactive analysis (GEPIA, http://gepia.cancer-pku.cn/) (Figure 1B). Pan-cancer gene expression analysis revealed that QPCT was differentially expressed in several cancers including LUAD (Supplementary Figure 1). As illustrated in Figure 1C and 1D, the mRNA and protein expression of QPCT was upregulated in clinical LUAD specimens. Differential QPCT expression in LUAD led us to evaluate whether QPCT expression in LUAD correlated with tumor characteristics in humans. A tumor histology array containing 97 human LUAD specimens was used to perform IHC staining for QPCT (Figure 1E). The relationship between QPCT expression and tumor characteristics is shown in Table 1. High QPCT expression correlated significantly with the TNM stage (χ² = 23.74, p value < 0.0001) and metastasis (χ² = 12.82, p value = 0.0003).

Figure 1.

QPCT gene expression is upregulated in lung adenocarcinoma (LUAD) patient tissues. A. Heatmap showing the gene expression of QPCT and QPCTL in LUAD from the Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov/geo/; access number: GSE30219, GSE40791, and GSE75037). B. QPCT expression in LUAD/normal tissues was extracted from the Cancer Genome Atlas (TCGA) database and genotype tissue expression (GTEx) using the GEPIA website (http://gepia2.cancer-pku.cn/#index). T means tumor tissues, and N means normal tissues. C. QPCT mRNA expression in clinical LUAD and paracancerous specimens. D. QPCT protein expression in four pairs of LUAD and paired paracancerous lung tissues. P means paracancerous specimens. E. Representative immunohistochemistry images of high and low QPCT expression in clinical tissue samples of LUAD. 200× scale bar = 100 μm. 400× scale bar = 50 μm. **p value < 0.01. Scatter plot with bar is presented as mean ± standard deviation (SD), while box plot presents median, quartiles, range and individual data points. QPCT, glutaminyl-peptide cyclotransferase; QPCTL, glutaminyl-peptide cyclotransferase like.

Table 1.

Correlation between QPCT expression in tumor tissues and clinicopathologic characteristics of lung adenocarcinoma patients

Variable	Category	No. of cases		χ2	p value
		QPCT low expression	QPCT high expression
		(n = 42)	(n = 55)
Age	< 60	14	18	0.0040	0.9498
	≥ 60	28	37
Sex	Male	13	22	0.8453	0.3579
	Female	29	33
Smoking status	Never/former smoker	39	55	2.021	0.1551
	Current smoker	3	0
Tumor differentiation	Well	6	5	1.254	0.5343
	Moderate	27	41
	Poor	9	9
TNM stage	I-II	40	27	23.74	< 0.0001*
	III-IV	2	28
Metastasis	Present	1	17	12.82	0.0003*
	Absent	41	38

^*indicates significance, P value < 0.05.

The knockdown of QPCT inhibits the proliferation of LUAD cells

HCC44 cells stably overexpressing QPCT and A549 cells stably transfected with QPCT shRNA were employed to explore the effect of QPCT on LUAD cell proliferation. Efficiency of overexpression and knockdown of QPCT expression levels were verified (Figure 2A and 2D). The overexpression of QPCT significantly promoted HCC44 cell proliferation starting after 48 h of culture (Figure 2B). Similarly, EdU staining also showed that cells with QPCT overexpression showed more cell growth (Figure 2C). The CCK-8 (Figure 2E) and EdU staining (Figure 2F) results showed that the knockdown of QPCT inhibited the A549 cell proliferation.

Figure 2.

The knockdown of QPCT inhibits the proliferation of lung adenocarcinoma (LUAD) cells. The HCC44 and A549 cell lines with stably overexpressed or knocked down QPCT were established (upper panel). (A) The efficiency of QPCT overexpression in HCC44 cells was verified through western blot analysis. The effect of QPCT overexpression on cell proliferation using Cell counting kit-8 (CCK-8) (B) and EdU assays (C). *p value < 0.05, **p value < 0.01. (D) Knockdown of QPCT in A549 cells was verified by western blot analysis. **p value < 0.01, ##p value < 0.01, compared to LV-shNC group. CCK-8 assay (E) and EdU staining (F) were performed to determine the proliferation of A549 cells. *p value < 0.05, **p value < 0.01; LV-shQPCT-1 compared to LV-shNC group. #p value < 0.05, ##p value < 0.01; LV-shQPCT-2 compared to LV-shNC group. Data are expressed as mean ± standard deviation (SD). LV-VE, vector control lentivirus; LV-QPCT, QPCT-overexpressing lentivirus; LV-shNC, negative control lentivirus; LV-shQPCT-1/2, QPCT shRNA lentivirus; QPCT, glutaminyl-peptide cyclotransferase.

QPCT silencing in LUAD cells inhibits the macrophage chemotaxis and M2 macrophage polarization

As detailed in Figure 3A and 3B, the CCL2 concentration in the supernatant of QPCT-overexpressing HCC44 cells was remarkably increased, while it was significantly decreased in A549 cells stably transfected with QPCT shRNA (shQPCT). The overexpression of QPCT notably increased the number of migrated THP-1-M0 cells to the lower chamber (Figure 3C), whereas QPCT knockdown had the opposite effect (Figure 3D). To test how changes in QPCT expression affect M2 macrophage polarization, we added IL-4 and IL-13 to the lower transwell chamber to differentiate THP-1-M0 [8]. CD68⁺ and CD206⁺ are markers for different macrophage identities, and the ratio of CD68⁺/CD206⁺ cells to total CD68⁺ cells can be used to measure population of M2 macrophages [28]. We found that the percentage of CD68⁺/CD206⁺ cells and the protein expression of Arg1 and CD206 were increased in THP-1-M0 cells co-cultured with QPCT-overexpressing HCC44 cells (Figure 3E-G), indicating that QPCT overexpression induced M2 macrophage polarization. In contrast, QPCT inhibition led to reduced CD68/CD206 double-positive cells, as well as Arg1 and CD206 protein expression in THP-1-M0 cells (Figure 3E, 3F and 3H). As detailed in Figure 3I-P, we got similar results when THP-1-M0 cells were co-cultured with the conditional medium from HCC44/A549 cells. These results revealed that QPCT overexpression is responsible for increased CCL2 contents, macrophage migration to LUAD cells, and M2 macrophage differentiation, which is prevented by silencing QPCT.

Figure 3.

QPCT silencing in lung adenocarcinoma (LUAD) cells inhibits the macrophage chemotaxis and M2 macrophage polarization. CCL2 levels in culture supernatants from HCC44 cells (A) and A549 cells (B). (C, D) The effect of QPCT in LUAD cell lines on macrophage chemotaxis was determined by crystal violet staining and counting cells in the lower transwell membrane. 200× scale bar = 100 μm. (E) The co-culture transwell systems were established to study the influence of LUAD cells on the M2 polarization of THP-1-derived M0 macrophages (THP-1-M0). (F) Flow cytometry analysis of the proportion of CD68/CD206 double-positive cells. (G, H) Quantification of flow cytometry results depicted in (F) (upper panel) and western blot analysis of Arg1 and CD206 in THP-1-M0 (lower panel). (I) Transwell migration assay used THP-1-M0 cells in the upper chamber and the conditioned medium of transfected HCC44 and A549 cells in the lower chamber. (J-L) Cells migrated to the lower side of the membrane in the transwell chamber were measured by crystal violet staining and quantification. 200× scale bar = 100 μm. (M) THP-1-M0 were cultured in HCC44 or A549 conditioned medium supplemented with IL-4 and IL-13 to study the conditioned medium of transfected LUAD cells on M2 polarization of THP-1-M0 cells. (N) The percentage of CD68/CD206-positive cells was examined using flow cytometry. (O, P) The quantification of flow cytometry results (upper panel) and Arg1 and CD206 protein expression in THP-1-M0 cells detected using western blot assay (lower panel). For HCC44 cells, ** indicated p value < 0.01, compared to LV-VE group. For A549 cells, **/## indicated p value < 0.01, compared to LV-shNC group. Data are expressed as mean ± standard deviation (SD). LV-VE, vector control lentivirus; LV-QPCT, QPCT-overexpressing lentivirus; LV-shNC, negative control lentivirus; LV-shQPCT-1/2, QPCT shRNA lentivirus; PMA, phorbol 12-myristate 13-acetate; QPCT, glutaminyl-peptide cyclotransferase; CCL2, chemokine C-C motif ligand 2.

QPCT knockdown suppresses tumor growth of human LUAD cells in nude mice by inhibiting macrophage chemotaxis and M2 macrophage polarization

We assessed the effect of QPCT knockdown on the xenograft tumor growth in nude mice and found that tumors induced by A549 cells stably transfected with shQPCT showed significantly smaller tumor sizes than tumors induced by parental or A549 cells expressing shNC (Figure 4A-C). The tumor volume of nude mice in the LV-shQPCT-1 group was remarkably smaller than that of nude mice in the LV-shNC group from the 10th day after cell injection, while nude mice in the LV-shQPCT-2 group had notably smaller tumor volume from the 13 days (Figure 4B). The knockdown of QPCT by LV-shQPCT-1 significantly reduced the tumor weight rather than by LV-shQPCT-2 (Figure 4C). The deletion of QPCT was verified by WB assay (Figure 4D). The IHC staining of Ki-67 (Figure 4E) further illustrated that the QPCT silencing inhibited proliferation marker Ki-67 expression in LUAD. These results indicated that QPCT knockdown significantly inhibited the tumorigenic ability of LUAD cells. As presented in Figure 4F, the human CCL2 protein expression was inhibited by the knockdown of QPCT in LUAD tumors. Furthermore, the deletion of QPCT decreased mCD68-positive cells in tumor tissues (Figure 4G). The WB results revealed that the QPCT knockdown remarkedly inhibited mouse Arg1 protein expression in tumors of nude mice (Figure 4H). These results indicated that the knockdown of QPCT inhibited the xenograft tumor growth by decreasing the CD68⁺ cells recruited to the tumors and macrophage M2 polarization under the regulation of CCL2 signaling.

Figure 4.

QPCT knockdown suppresses tumor growth of human lung adenocarcinoma (LUAD) cells in nude mice by inhibiting macrophage chemotaxis and M2 macrophage polarization. A. The flow chart of the animal experiment and photos of dissected tumor masses on day 28. B. Tumor volume. *p value < 0.05, **p value < 0.01; LV-shQPCT-1 compared to LV-shNC group. ##p value < 0.01; LV-shQPCT-2 compared to LV-shNC group. C. Tumor weight. D. The efficiency of the knockdown of QPCT was verified by western blot assay. E. Immunohistochemistry of Ki-67 in LUAD tumors. 400× scale bar = 50 μm. F. The hCCL2 protein expression in LUAD tumors. G. Immunofluorescence staining of mCD68 (red) and DAPI (blue) in LUAD tumors. 400× scale bar = 50 μm. H. The mArg1 protein expression in LUAD tumors. **p value < 0.01, #p value < 0.05, ##p value < 0.01, ns = no significant, compared to LV-shNC group. Data are expressed as mean ± standard deviation (SD). LV-shNC, negative control lentivirus; LV-shQPCT-1/2, QPCT shRNA lentivirus; QPCT, glutaminyl-peptide cyclotransferase; CCL2, chemokine C-C motif ligand 2.

Dysregulated modification of m6A contributes to the upregulation of QPCT mRNA in LUAD

According to bioinformatics analysis, there was an m6A modification site on QPCT (Supplementary Figure 2). MeRIP-PCR results illustrated that QPCT mRNA was m6A modified in A549 cells (Figure 5A). The m6A demethylase FTO has been reported to play roles in promoting tumorigenesis in LUAD [21-23]. To investigate whether FTO correlates with QPCT expression levels, the mRNA expression of FTO in clinical LUAD samples was examined. The FTO mRNA expression was notably upregulated in human LUAD tissues compared to the paracancerous tissues (Figure 5B). We found a significant positive relationship between QPCT and FTO mRNA expression in clinical LUAD samples (Figure 5C). As detailed in Figure 5D, RIP-PCR results showed that FTO bound to QPCT mRNA. Two specific siRNAs were used to target FTO to explore the effect of m6A modification on QPCT mRNA stability. The knockdown of FTO was verified at the protein level (Figure 5E). The siRNA with higher knockdown efficiency siFTO-2 was used for subsequent experiments. Furthermore, MeRIP-qPCR assays demonstrated that FTO knockdown significantly increased the m6A levels on mRNA transcripts of QPCT (Figure 5F). In addition, FTO knockdown reduced the mRNA expression level of QPCT in A549 cells (Figure 5G). To further verify the effect of FTO on QPCT RNA stability, we treated A549 cells with the transcriptional inhibitor actinomycin D after FTO knockdown. We found that the FTO deletion inhibited the QPCT mRNA degradation (Figure 5H). These results suggest that m6A modification of QPCT affects its mRNA stability in A549 cells.

Figure 5.

Dysregulated modification of m6A contributes to the upregulation of QPCT mRNA in lung adenocarcinoma (LUAD). A. Methylated RNA immunoprecipitation PCR (MeRIP-PCR) assay was used to detect the QPCT m6A modification level in A549 cells. B. FTO mRNA expression in clinical lung adenocarcinoma (LUAD) and paracancerous specimens. C. The correlation between FTO and QPCT mRNA expression in clinical LUAD specimens. Data are presented as scatter plots with a linear fit. D. RNA immunoprecipitation PCR (RIP-PCR) assay was used to verify the FTO binding to QPCT mRNA. E. The knockdown of FTO was verified by measuring the FTO protein expression in A549 cells. **/##p value < 0.01, compared to siNC group. ns = no significant. F. MeRIP-qPCR assay was used to demonstrate FTO-mediated QPCT m6A modification in A549 cells. **p value < 0.01, compared to siNC group; ##p value < 0.01, compared to siFTO-2 group. ^^p value < 0.01. G. QPCT mRNA expression in A549 cells detected using real-time quantitative PCR (Rt-qPCR). **p value < 0.01. H. QPCT mRNA expression in A549 cells treated with actinomycin D (10 μg/mL) at different times. **p value < 0.01. Scatter plot with bar is presented as mean ± standard deviation (SD). m6A, N6-methyladenosine; siNC, negative control siRNA; si-FTO, FTO siRNA; QPCT, glutaminyl-peptide cyclotransferase; FTO, fat mass and obesity-associated protein.

FTO regulates the effect of QPCT on macrophage chemotaxis and M2 macrophage polarization

The siRNA targeting FTO was used to silence the FTO expression in parental or HCC44 cells stably overexpressing QPCT (Figure 6A). As shown in Figure 6B, overexpression of QPCT significantly increased the content of CCL2 in the conditioned medium of HCC44 cells in cells where FTO was knocked down. The deletion of FTO remarkedly reduced the number of THP-1-M0 cells migrated to HCC44 conditioned medium (Figure 6C and 6D) and the percentage of CD68/CD206 double-positive cells (Figure 6E and 6F) in HCC44 cells, while QPCT overexpression reversed this trend. These results revealed that the effects of QPCT on TAMs chemotaxis and macrophage M2 polarization were regulated by FTO in LUAD cells.

Figure 6.

FTO regulates the effect of QPCT on macrophage chemotaxis and M2 macrophage polarization. (A) The simplified process of conditioned medium preparation. (B) The CCL2 concentrations in the conditioned medium. The representative violet crystal staining pictures showed THP-1-M0 cells in the lower membrane of the transwell chamber (C) and the quantification analysis (D). 200× scale bar = 100 μm. (E) Flow cytometry assay was performed to detect the percentage of CD68⁺/CD206⁺ cells in HCC44 conditioned medium supplemented with IL-4 and IL-13. (F) The quantification of flow cytometry results in (E). **p value < 0.01, *p value < 0.05. Data are expressed as mean ± standard deviation (SD). LV-VE, vector control lentivirus; LV-QPCT, QPCT-overexpressing lentivirus; siNC, negative control siRNA; si-FTO, FTO siRNA; QPCT, glutaminyl-peptide cyclotransferase; FTO, fat mass and obesity-associated protein; CCL2, chemokine C-C motif ligand 2.

Discussion

Glutaminyl cyclases including QPCT and QPCTL result in the progression of cancers through regulating CCL2 or the immune checkpoint CD47-SIRPα [7,9]. The two enzymes catalyze the intramolecular cyclization of peptides and proteins N-terminal l-glutamine residues into pyroglutamic acid, which can stabilize the peptides and proteins and be of great importance for their biological activity [11]. QPCTL has been found to be highly expressed in glioma tumor tissues, and its high expression was closely associated with high-grade malignancy and advanced tumor stage [27]. For QPCT, it was reported to be upregulated in breast cancer [12] and thyroid carcinomas [9,13]. In the study, we investigated the expression of differentially expressed genes in the LUAD-related datasets and found that QPCT but not QPCTL was significantly upregulated in LUAD. Therefore, we decided to focus on QPCT in the following study. As expected, we confirmed the high mRNA and protein expression of QPCT in the clinical LUAD specimens, and high QPCT expression was significantly correlated with TNM stage and metastasis in LUAD. These evidences suggested that QPCT may play an important role during LUAD development.

The overexpression of QPCT promoted the proliferation of breast cancer cells [12]. Similarly, QPCT overexpression promoted the proliferation of LUAD cells in our study, while QPCT knockdown inhibited this trend. It has been reported that the N-terminus of CCL2 was modified to a pyroglutamate-residue by QPCT, and thus stabilized CCL2 [9]. In thyroid carcinomas, the QPCT knockdown inhibited thyroid cancer cell growth by reducing CCL2 signaling [9]. Consistent with this result, the overexpression of QPCT increased CCL2 levels in LUAD cell supernatant, whereas CCL2 concentrations were significantly reduced in QPCT-knockdown LUAD cells. Therefore, we hypothesized that QPCT may influence the proliferation of LUAD cells through CCL2 signaling. CCL2 is known to be a chemokine that recruits monocytes/macrophages to tumor sites [15]. Elevated CCL2 expression is associated with the accumulation of TAMs [15]. Our results showed that the LUAD cell line that stably overexpressed QPCT or its conditioned medium promoted the chemotaxis of THP-1-M0 cells toward tumor cells. This may result from elevated levels of CCL2 in the supernatants of QPCT-overexpressing LUAD cells. Invariably, the knockdown of QPCT showed the opposite effect. The M2-type TAMs exert immunosuppressive effects at the tumor site and promote tumorigenesis and progression [6]. We found that the knockdown of QPCT reduced the proportion of THP-1-M0 cells induced into M2-type by IL-4/IL-13. These results indicated that QPCT deletion in the LUAD cell line can inhibit the proliferation of LUAD cells by suppressing THP-1-M0 macrophage chemotaxis toward tumor cells and reducing the proportion of M2-type in THP-1-M0 macrophages.

In in vivo experiments, QPCT knockdown inhibited the growth of xenograft tumors in nude mice. Importantly, the protein expression of hCCL2 in tumor tissues from nude mice transplanted with QPCT-knockdown A549 cells was reduced, and the number of monocytes/macrophages recruited to the tumor tissues was decreased, which was consistent with the results of the in vitro experiments. In addition, QPCT silencing reduced the proportion of M2 macrophages in tumor tissues. These results also supported our conclusions drawn from in vitro experiments. Furthermore, previous reports showed that the N-terminus of CCL2 is modified to a pyroglutamate (pE)-residue by both glutaminyl cyclases QPCT and QPCTL [9,10], we believe that manipulations in QPCTL may have a similar effect on LUAD tumor cell proliferation as QPCT. Further studies are needed for clarifying the involvement of both the glutaminyl cyclases, QPCT or QPCTL, in CCL2 formation of LUAD.

Many reports focused on the DNA methylation of QPCT on poor prognosis or risk factors in cancers [29,30]. The m6A modification of QPCT has not been investigated in LUAD. There was a predicted m6A modification site on QPCT, and we confirmed this by MeRIP-PCR experiments. FTO, a major m6A demethylase [31], has been reported to promote tumorigenesis in LUAD [21-23]. In our study, FTO mRNA expression was significantly positively correlated with QPCT mRNA expression in clinical LUAD samples. Additionally, FTO is bound to the mRNA of QPCT in LUAD cells. These findings led us to explore the effect of m6A modification on the mRNA stability of QPCT. By investigating the effect of the transcriptional repressor actinomycin D on QPCT mRNA decay experiments, we found that the knockdown of m6A demethylase FTO promoted the degradation of QPCT mRNA. We therefore concluded that FTO affected the mRNA stability of QPCT. To study whether the effects of QPCT on TAMs chemotaxis and the macrophage M2 polarization were under FTO regulation, we performed rescue experiments in HCC44 cells in vitro. The knockdown of FTO reversed the increase in CCL2 levels in cell supernatant, TAMs recruited to the conditioned medium, and the percentage of M2 macrophages induced by QPCT overexpression in HCC44 cells.

In conclusion, we found that QPCT is highly expressed in clinical LUAD samples, and its high expression is associated with malignant phenotypes in LADC patients. QPCT knockdown inhibits the LUAD cell proliferation both in vivo and in vitro by suppressing macrophage M2 polarization and CCL2-induced chemotaxis of TAMs. Mechanistic studies show that FTO-mediated demethylation increased the stability of QPCT mRNA in LUAD cells, which affected the effect of QPCT on chemotaxis and M2 polarization of TAMs.

Disclosure of conflict of interest

None.