Somatic long interspersed element-1 (LINE-1) retrotransposition is a genomic process that relates to gene disruption and tumor occurrence. However, the expression and function of LINE-1 retrotransposition in lung squamous cell carcinoma (LUSC) remain unclear. We analyzed the transcriptomes of LUSC samples in The Cancer Genome Atlas and observed LINE-1 retrotransposition in 90% of tumor samples. Thirteen LINE-1 retrotranspositions of high occurrence were identified and further validated from an independent Chinese LUSC cohort. Among them, LINE-1-FGGY (L1-FGGY) was identified as the most frequent LINE-1 retrotransposition in the Chinese cohort and significantly correlated with poor clinical outcome. L1-FGGY occurred with smoke-induced hypomethylation of the LINE-1 promoter and contributed to the development of local immune evasion and dysfunctional metabolism. Overexpression of L1-FGGY or knockdown of FGGY promoted cell proliferation and invasion in vitro, facilitated tumorigenesis in vivo, and dysregulated cell energy metabolism and cytokine/chemotaxin transcription. Importantly, specific reverse transcription inhibitors, nevirapine and efavirenz, dramatically countered L1-FGGY abundance, inhibited tumor growth, recovered metabolism dysfunction, and improved the local immune evasion. In conclusion, hypomethylation-induced L1-FGGY expression is a frequent genomic event that promotes the development and progression of LUSC and represents a promising predictive biomarker and therapeutic target in LUSC.

Significance:

LINE-1-FGGY is a prognosis predictive biomarker and potential therapeutic target to overcome local immune evasion in lung squamous cell carcinoma.

Lung cancer is one of the most common human malignancies with high incidence and mortality worldwide (1, 2). The 5-year survival rate of lung cancer is only about 18.1%, which is directly related to the late diagnosis of advanced cancer (2). Non–small cell lung cancer (NSCLC) accounts for the majority (>85%) of all lung cancers, including lung adenocarcinoma, lung squamous cell carcinoma (LUSC), and large cell lung cancer (3, 4). Among them, LUSC is strongly associated with smoking (3), and about 30% of NSCLC patients in China have LUSC histology, similar to that of the United States (5). However, the actual patient number is higher due to the proportionality of the larger Chinese population. Therefore, there is an urgent need to identify optimal treatments for LUSC within the Chinese population and aboard. To date, LUSC patients have little benefit from targeted therapies unlike lung adenocarcinoma patients, which may be due to the higher mutation rate but lack of actionable driver gene mutations in LUSC patients (6). Therefore, as the standard treatments, platinum-based chemotherapy and tyrosine kinase inhibitor–based targeted therapy only achieve limited efficacy on LUSC (7). However, recently reported trials indicated that immunotherapy might be a promising regimen for LUSC.

The checkpoint inhibitors–based immunotherapy has been proved to improve overall survival (OS) and progression-free survival of LUSC patients (8, 9), which completely transformed the therapeutic landscape of LUSC with impressive therapeutic outcomes, but lack of predictive biomarkers limited the efficacy of immunotherapy agents. It is reported that tumor mutation burden (TMB), a parameter to assess tumor genomic instability, is a feasible predictive biomarker, because patients with higher TMB benefited more from immunotherapy in NSCLC (10, 11). However, TMB is not sufficient to accurately predict the clinical efficacy of immunotherapy (12), which indicated that more genomic biomarkers than genetic mutations will be valuable in predicting patients’ survival more precisely in LUSC.

Retrotransposition is one mechanism of chromosomal rearrangements with the specific insertion of transposons into genome via reverse transcription of the transposon (13). Somatic retrotransposition has been identified as a frequent genomic event that plays fundamental roles in early development and genome evolution (14, 15), as well as carcinogenesis (16). Among the retrotransposition elements, long interspersed element-1 (LINE-1) is the only currently known active autonomous transposon in humans, which occupies approximately 17% of the entire human genome (17). LINE-1 propagates itself through an RNA intermediate, which has the potential to disrupt the coding sequence of endogenous genes and alter gene expression via insertion into a genomic region (18). In addition, increased LINE-1 copy numbers provide more chances for recombination events to occur between retrotransposons, which can lead to chromosomal breaks and rearrangements (19). LINE-1–triggered genomic instabilities provide fuel to drive tumor occurrence because 47% of human neoplasms were reported immunoreactive to LINE-1, including invasive breast carcinomas, high-grade ovarian carcinomas, and pancreatic ductal adenocarcinomas (20).

The mechanism of how somatic LINE-1 retrotransposition alters the expression of neighboring genes remains controversial (18, 19) because somatic LINE-1 insertions lead to cancer by either activating proto-oncogenes or inhibiting tumor-suppressor genes. In breast cancer, the LINE-1 insertion was found to cause the rearrangement and amplification of the MYC gene, resulting in the initiation of breast ductal adenocarcinoma (21). But in colon cancer, the LINE-1 insertion disturbed the last coding exon of APC gene and inhibited its function to initiate colon cancer (22). Although it was reported that in lung cancer patients, LINE-1 retrotranspositions in circulating DNA were hypomethylated (23), and even LINE-1 hypomethylation was associated to specific clinicopathologic features, including histology, p53 immunoreactivity, and smoking habit (24); however, none of the study focusing on the function and mechanism of somatic LINE-1 retrotransposition in lung cancer, especially in LUSC, has ever been reported.

In this study, we analyzed the transcriptomes of LUSC samples in The Cancer Genome Atlas (TCGA) to characterize the most common somatic LINE-1 retrotranspositions in LUSC and to further validate them in an independent Chinese LUSC cohort. We found that LINE-1-FGGY (L1-FGGY) was the most frequent LINE-1 retrotransposition significantly correlating with poor clinical outcome, which occurred with smoke-induced hypomethylation and contributed to the development of local immune evasion. Further study indicated the oncogenic roles of L1-FGGY by disrupting the expression and function of tumor-suppressor gene FGGY via promoting cell proliferation and invasion in vitro, facilitating tumorigenesis in vivo, and inducing the dysregulation of cell energy metabolism and cytokine/chemotaxin transcription, which could be fully recovered by specific reverse transcription inhibitors. Therefore, L1-FGGY is not only a prognosis predictive biomarker, but also a potential therapeutic target to overcome local immune evasion in order to achieve more clinical benefits from immunotherapy in LUSC.

Cell lines

NCI-H520, SK-MES-1, and BEAS-2B were purchased from Cellcook Co., Ltd., with cell authentication via short tandem repeat multiamplification method. A549, H1299, NCI-H460, NCI-H446, and HEK293T were obtained from Chinese Academy of Medical Sciences tumor cell libraries. Mycoplasma for the cells cultured in our laboratory was tested using the Mycoplasma Detection Kit according to the manufacturer's protocol in every 3 months.

Mice

Female NOD-SCID mice, which were 7-week-old and weighed about 17 to 18 g, were obtained from the Beijing Vital River Laboratory Animal Technology Co., Ltd. All mice were housed in a specific pathogen-free animal facility.

Patient information

This study selected 109 cases of LUSC patients from a well-informed cohort of patients who were treated with partial lung resection surgery at the Department of Lung Cancer of Tianjin Medical University Cancer Institute and Hospital from October 2004 to October 2006 (Table 1; ref. 25). Among these 109 cases, 52 cases of LUSC samples were coupled with matched paracarcinoma tissues. All paracarcinoma lung tissues sectioned at least 5 cm from the tumor's boundary to avoid any potential of tumor cell infiltration. All smoking information was collected and recorded when the patient was hospitalized for the first time. And we defined the patients who never smoked after he reached adulthood as “negative smokers,” whereas the patients who ever smoked after adulthood as “positive smokers.” No prior treatments, including chemotherapy or radiotherapy, were conducted before lung resection surgery was performed. Postoperative follow-up time was 67 to 96 months. Written-informed consents were obtained from the patients, and this project was approved by the Ethics Committee of Tianjin Medical University. All experiments were performed in accordance with the principles of the Declaration of Helsinki.

Table 1.

The basic clinicopathologic information of all patients

Clinicopathologic parametersNumber of patients
Total 109 
Gender  
Male 89 
Female 20 
Age  
<60 years 46 
≥60 years 63 
Stage  
I–II 64 
III–IV 45 
T stage  
1–3 98 
11 
N stage  
66 
14 
29 
M stage  
97 
12 
Metastatic site  
Negative 97 
Lung 
Bone 
Brain 
Others 
Location  
Central 58 
Periphery 51 
Smoking  
Negative 14 
Positive 95 
KPS  
≤60 15 
>60 94 
Clinicopathologic parametersNumber of patients
Total 109 
Gender  
Male 89 
Female 20 
Age  
<60 years 46 
≥60 years 63 
Stage  
I–II 64 
III–IV 45 
T stage  
1–3 98 
11 
N stage  
66 
14 
29 
M stage  
97 
12 
Metastatic site  
Negative 97 
Lung 
Bone 
Brain 
Others 
Location  
Central 58 
Periphery 51 
Smoking  
Negative 14 
Positive 95 
KPS  
≤60 15 
>60 94 

Lentivirus construction

For L1-FGGY insertion lentivirus construction, the L1-FGGY fragment was amplified by PCR using the complementary DNA (cDNA) of NCI-H520 cells as template. Then the amplified L1-FGGY fragment was inserted into pHBLV-CMV-MCS-EF1-ZsGreen-T2A-Puro lentiviral vectors (Hanbio Co., Ltd.), and the constructed positive plasmid was confirmed by DNA sequencing. The recombinant lentivirus with L1-FGGY sequence was generated by cotransfection in NCI-H520 cells as previously described (26), and the empty lentivector lenti-puromycin was used as negative control. For FGGY knockdown lentivirus construction, specific shRNA sequence based on the sequence of FGGY (Gene ID: 55277, on NCBI) was designed. The synthesized FGGY shRNA and control shRNA were inserted into plvx-U6-CMV-RFP-P2A-BSD lentiviral vectors respectively. The lentivirus was generated as described above.

Cell culture and cell treatment

NCI-H520, A549, H1299, NCI-H460, and NCI-H446 were cultured in RPMI1640 (Gibco BRL). SK-MES-1 was cultured in Eagle Minimum Essential Medium. HEK293T and BEAS-2B cells were cultured in DMEM. All medium contained 10% FBS and 1% penicillin/streptomycin. All of the above cells were cultured at 37°C, under 5% CO2. The general length of time between collection/thawing and use in our laboratory was no more than 6 months. For reverse transcriptase inhibitors’ treatment experiments, NVR and EFV (TargetMol) were dissolved in dimethyl sulfoxide (DMSO, Sigma Aldrich) to make stock regent respectively. Five hours after cells seeded, NVR was diluted to 350 μmol/L, and EFV was diluted to 15 μmol/L, followed by replacing the cell medium. The same DMSO volume (0.2% final concentration) was added to control cells. Fresh NVR- or EFV-containing medium was changed every 48 hours.

Detection of cell proliferation, cell apoptosis analysis, wound-healing assay, and transwell invasion assay

All these detections of cellular functions were performed according to the manufacturer's protocol as previously described (27). And the experiments were repeated at least 3 times.

In vivo tumorigenicity study

After the mice construction, the tumor sizes of each NOD-SCID mouse were monitored every 2 days. Each group contains 5 mice. And the experiments were repeated at least 3 times. The tumor volume (V) was calculated by the formula: V = 3.14 × L × W × H/6 (L: length, W: width, H: height). For drug treatment experiments, animals were then subjected to treatment with either NVR (50 mg/kg/day) or EFV (20 mg/kg/day) every day. Simultaneously, the mice with no treatment and the mice with DMSO treatment were cultivated as controls. The animal protocol used in this study was approved by the Ethics Committee for Animal Experiments of the Tianjin Medical University Cancer Hospital and Institute, and was approved by the Wistar Institutional Animal Care and Use Committee (IACUC). The Wistar IACUC guideline was followed in determining the time for ending the survival experiments (tumor burden exceeds 10% of body weight).

RNA extraction, retrotransposition-PCR, and qPCR analysis for gene expression

RNA extraction, cDNA synthesis, retrotransposition-PCR, and regular qPCR analysis were performed following the manufacturer's protocol as previously described (27). The sequences of primers are shown in Supplementary Table S1. To confirm that the bands detected in the PCR assay were the genes as predicted, we purified and sequenced the PCR products (Invitrogen). High-throughput qPCR analysis was performed on Smartchip (Differential Gene Technology Co., Ltd.) following the manufacturer's protocols. All experiments were performed in triplicates and were calculated for ΔCt. Relative expression quantity of mRNA was calculated as 2−ΔCt (ΔCt = Cttarget gene – Ctreference gene).

Quantitative methylation-specific PCR analysis for the methylation levels of LINE-1 retrotranspositions

Genomic DNA was obtained and purified from frozen LUSC tissues in quantities sufficient for bisulfite treatment. Bisulfite conversion was carried out on 500 ng genomic DNA using the EpiTect Bisulfite Kit (Qiagen), according to the manufacturer's protocol as described above (28). The experiments were repeated at least 3 times.

IHC

All procedures were performed as described above (29). The antibodies we used here are as follows: anti-CD3 (Abcam), anti-CD68 (Santa Cruz Biotechnology), anti-CD33 (Abcam), anti-FGGY (Bioss), anti-ki67 (Cell Signaling Technology), anti–N-cadherin (Zsbio), anti–β-catenin (Zsbio), anti–PD-L1 (BioSS), anti-CD11b (Abcam), and a biotinylated secondary goat anti-mouse IgG antibody (Santa Cruz Biotechnology), labeled with streptavidin-horseradish peroxidase using a DAB staining kit (Maixin Biotechnology) according to the manufacturer's instructions. For negative controls, IgG1 was used to substitute for each primary antibody. Positively stained cells were counted in 5 fields at 200× magnification, and the sum of the cells was calculated as positive cell counts.

Detection of somatic LINE-1 retrotranspositions in TCGA LUSC datasets

We downloaded paired-end RNA-seq data of LUSC samples (tumor and paired adjacent normal tissue) from TCGA upon approval of TCGA commission. The somatic insertion of LINE-1 retrotranspositions into a gene was detected by using deFuse (http://shahlab.ca/projects/defuse/; ref. 30). Briefly, the detection of read pairs that discordantly map to two distinct genes generates a first set of gene insertion candidates. Subsequently, the exact insertion junction is determined for each candidate by searching for reads spanning the breakpoint, i.e., reads that partially map to both genes. The results produced by deFuse were further filtered to reduce the number of false positives: predictions had to be supported by at least eight reads spanning an insertion breakpoint and five reads split by an insertion breakpoint. Pegasus tool (http://sourceforge.net/p/pegasus-fus; ref. 31) was then used for the functional characterization of RNA-seq gene insertion candidates with one end mapped to LINE-1 retrotransposition consensus sequences from Repbase database (32) by BLAST and quantification of their oncogenic potential. After annotation by Pegasus tool, we reserved those LINE-1 retrotranspositions that occurred only in the cancer samples, but not in any normal samples.

RNA library preparation and sequencing

Library preparation and sequencing steps were commissioned to Novogene Co., Ltd. The Novogene pipeline included the production of strand-specific mRNA libraries and quality control. The libraries were sequenced on Illumina (NEB) following the manufacturer's recommendations. The RNA sequencing data have been uploaded to GEO database (accession number: GSE124625).

Differential expression analysis and KEGG enrichment analysis

Prior to differential gene expression analysis, the read counts were adjusted by edgeR program package through one scaling normalized factor. Differential expression analysis of two conditions was performed using the edgeR R package (3.18.1). The P values were adjusted using the Benjamini and Hochberg method. Corrected P value of 0.05 and absolute fold change of 2 were set as the threshold for significantly differential expression. Here, we used clusterProfiler R package to test the statistical enrichment of differential expression genes in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.

Gene set enrichment analysis to detect immune markers

We collected 12 immune cell datasets, including 6 single cell datasets from 10x Genomics (33), 1 single cell dataset from Todd and colleagues (34), and 5 sorted pure cell datasets (35). For each cell type, we compared its expression values against each other cell and selected genes with log fold change > 0 and P value < 0.01 as markers for each dataset. For markers that appeared more than 2 times in those datasets, we selected markers with meta P value less than 0.01 and mean log fold changes larger than 2 as meta markers. We got 967 metamarkers for 8 cell types. We then performed gene set enrichment analysis (GSEA) based on these metamarkers for each immune cell type.

Statistical analysis

Data were analyzed using SPSS 17.0 and GraphPad Prism 5.0 software. Measurement data were presented as median (interquartile range) and compared through χ2 test. Quantitative data were presented as mean ± SD and compared through ANOVA and LSD tests. The Spearman rank order test and linear regression analysis were performed to assess the correlations between expression levels detected by qPCR. Univariate Cox regression and multivariate Cox regression analyses were used to identify common genes associated with OS. Cumulative survival was determined via the Kaplan–Meier method. Univariate survival analysis between the different LINE-1 retrotranspositions and the OS of LUSC patients was conducted through the two-sided log-rank test. Statistical significance was set at P < 0.05.

High occurrence of somatic LINE-1 retrotranspositions in LUSC tissues

In order to identify somatic LINE-1 retrotranspositions in LUSC, we obtained data from 50 individual cases of paired-end RNA-sequencing data of paired LUSC samples from TCGA. Discovery of the LINE-1 retrotranspositions is described by the procedure depicted in the schematic diagram (Fig. 1A). Briefly, the deFUSE program (30) was first used to identify fusion events from RNA transcripts compared with the reference genome followed by annotation with the Pegasus tool (31). The somatic LINE-1 retrotranspositions that occurred only in tumor were discovered by aligning the sequences of the fusion partners against the LINE-1 consensus sequences from the Repbase database (32). Figure 1B illustrates the L1-FGGY with 47 reads capturing transcripts containing LINE-1 insertions within the FGGY locus. The 5′ untranslated region of L1-HS (human-specific L1; a LINE-1 subfamily) was inserted into the beginning of exon 13 of gene FGGY (NM_001113411). We observed that 90% (45/50) of LUSC samples possessed somatic LINE-1 retrotranspositions with various rates of insertion ranging from 1 to 24 insertions per individual (Fig. 1C). We had identified a total of 39 LINE-1 retrotranspositions in 50 paired LUSC samples, which occurred only in the cancer samples, but not in any normal samples, and the three most frequent LINE-1 retrotranspositions include L1-SMYD3 (56%), L1-CBWD2 (42%), and L1-FGGY (38%; Fig. 1D). We then analyzed these 13 LINE-1 retrotranspositions in additional 398 unpaired LUSC samples and detected the top 3 LINE-1 retrotranspositions also displayed comparably high frequency (Fig. 1E). The retrotransposition events occurred most frequently on chromosomes 1, 2, 3, and 12 (Fig. 1F), and the majority of retrotransposition events were found within the coding regions (Fig. 1G).

Figure 1.

High occurrence of somatic LINE-1 retrotranspositions in LUSC tissues. A, Bioinfomatic workflow to identify LINE-1 retrotranspositions in LUSC. B,L1-FGGY with 47 supporting reads in one sample. C, Somatic LINE-1 retrotransposition counts across LUSC samples. D, Thirteen LINE-1 retrotranspositions in paired LUSC samples. E, Thirteen LINE-1 retrotranspositions in unpaired LUSC samples. F, Genome location distribution of LINE-1 retrotranspositions. G, Gene location distribution of LINE-1 retrotranspositions. H, The qPCR results of the 13 LINE-1 retrotranspositions in an independent Chinese cohort.

Figure 1.

High occurrence of somatic LINE-1 retrotranspositions in LUSC tissues. A, Bioinfomatic workflow to identify LINE-1 retrotranspositions in LUSC. B,L1-FGGY with 47 supporting reads in one sample. C, Somatic LINE-1 retrotransposition counts across LUSC samples. D, Thirteen LINE-1 retrotranspositions in paired LUSC samples. E, Thirteen LINE-1 retrotranspositions in unpaired LUSC samples. F, Genome location distribution of LINE-1 retrotranspositions. G, Gene location distribution of LINE-1 retrotranspositions. H, The qPCR results of the 13 LINE-1 retrotranspositions in an independent Chinese cohort.

Close modal

Next, we performed qPCR to validate the transcripts of the top 13 LINE-1 retrotranspositions in an independent Chinese cohort of 52 pairs of LUSC tumors and matched normal adjacent tissues collected from the Tianjin Medical University Cancer Institute and Hospital. The results showed that out of the 13 LINE-1 retrotranspositions, the expression of 12 insertions was significantly higher in LUSC tissues than the matched adjacent normal tissues (Fig. 1H; Table 1), which confirmed the findings of the retrotransposition events detected in the TCGA LUSC samples.

Somatic LINE-1 retrotranspositions correlated with specific clinicopathologic features in LUSC patients

We further checked these 13 LINE-1 retrotranspositions in several lung cancer cell lines (lung adenocarcinoma, A549 and H1299; LUSC, NCI-H520; large cell, NCI-H460; and small, NCI-H446) and two normal cell lines (human epithelial HEK293T and human lung epithelial cell BEAS-2B). Both qPCR and retrotransposition-PCR results revealed that 11 of 13 LINE-1 retrotranspositions were also detected in the cancer cell lines tested but not in the normal cell lines (Supplementary Fig. S1A–S1B). PCR products of these LINE-1 retrotranspositions were sequenced, and 8 products matched the expected LINE-1 retrotranspositions (Supplementary Fig. S1C).

We then explored association of somatic LINE-1 retrotranspositions, performed with the patients’ OS in 109 cases of LUSC samples. The patients were stratified into two groups by the expression level of LINE-1 retrotranspositions by qPCR and survival analysis with the Kaplan–Meier method to reveal that high expression of the three highly recurrent LINE-1 retrotranspositions correlated with poor survival outcomes (Fig. 2A), whereas other LINE-1 insertions did not (Supplementary Fig. S2). Then we explored the association of the 3 somatic LINE-1 retrotranspositions with the stage I–II patients’ OS, which gives a more homogenous population. The results consistently showed high expression of them was correlated with poor survival outcomes (Fig. 2B).

Figure 2.

Somatic LINE-1 retrotranspositions correlated with specific clinicopathologic features in LUSC patients. A, The OS was compared between L1-FGGY+ and L1-FGGY, L1-ATP8B1+ and L1-ATP8B1, and L1-SVEP1+ and L1-SVEP1 patients, respectively. B, The association of three LINE-1 retrotranspositions with the stage I–II patients’ OS was compared. C, Tumor T stages, tumor locations, and smoking histories were compared. D–F, CD3+ T cell (D), CD68+ macrophage (E), and CD33+ myeloid-derived cell (F) infiltrated number were compared.

Figure 2.

Somatic LINE-1 retrotranspositions correlated with specific clinicopathologic features in LUSC patients. A, The OS was compared between L1-FGGY+ and L1-FGGY, L1-ATP8B1+ and L1-ATP8B1, and L1-SVEP1+ and L1-SVEP1 patients, respectively. B, The association of three LINE-1 retrotranspositions with the stage I–II patients’ OS was compared. C, Tumor T stages, tumor locations, and smoking histories were compared. D–F, CD3+ T cell (D), CD68+ macrophage (E), and CD33+ myeloid-derived cell (F) infiltrated number were compared.

Close modal

We further analyzed the association between each LINE-1 retrotransposition and other clinicopathology (Supplementary Table S2). We found that high expression of the 3 LINE-1 retrotranspositions was correlated with larger tumor size, 5 LINE-1 retrotranspositions with central-type primary tumors, and 5 LINE-1 retrotranspositions with smoking history (Fig. 2C, Supplementary Table S2). We further detected that smoking status rather than smoking dose significantly affected the expression levels of 5 LINE-1 retrotranspositions (Supplementary Fig. S3A). Collectively, these findings reveal that LINE-1 retrotranspositions strongly correspond with large central-type tumors and smoking history, but not smoking dose.

Furthermore, we correlated the expression of the three most common LINE-1 retrotranspositions with immunocyte content in tumor tissues in situ. The IHC staining analysis showed that less CD3+ T cells were detected in L1-FGGY+ and L1-SVEP1+ (Fig. 2D) tissues, and more CD68+ macrophages and CD33+ myeloid-derived cells (Fig. 2E and F) were detected only in L1-FGGY+ tissues. We further found that smoking status rather than smoking dose significantly affected the distribution of immunocytes (Supplementary Fig. S3B). This finding provides evidence that local immune evasion was associated with certain LINE-1 retrotranspositions.

Among these 3 LINE-1 retrotranspositions, L1-FGGY corresponded the most strongly with smoking history, large tumor size, central tumor location, local immune evasion, as well as poorer prognosis, suggesting L1-FGGY in directing LUSC. Because of these inherent features of L1-FGGY, it was selected for further testing for the underlying function in epigenetic regulation and oncogenic role.

L1-FGGY corresponded with smoke-induced LINE-1 promoter hypomethylation, lipid metabolism dysregulation, and immune microenvironment alteration

In order to explore which pathways were affected by L1-FGGY, we performed RNA sequencing on 20 LUSC tumor samples (10 L1-FGGY+ vs. 10 L1-FGGY). We identified 1,529 (826 up and 703 down) dysregulated genes in L1-FGGY+ tissues at an adjusted P value of 0.05. The enrichment analysis indicated that many signaling pathways were upregulated in L1-FGGY+ tissues (Fig. 3A), most of which were metabolic pathways, especially lipid-related metabolism. We also performed GSEA to assess the dysregulation of immune cells in tumor tissues by L1-FGGY using immune cell-specific markers. The results showed CD4+ T cells were significantly downregulated in L1-FGGY+ tissues (Fig. 3B), suggesting immune evasion by L1-FGGY in tumors. Then, we further checked association of the transcription of cytokines with L1-FGGY. We observed that IFNL4, TNFRSF11A, TNFSF12, IL17RD, IL34, and IL27RA were downregulated in L1-FGGY+ tissues (Fig. 3C), which were mostly reported to promote the functions of T cells and activate tumor cell immune response. Although IL1A and IL6R were upregulated in L1-FGGY+ tissues (Fig. 3C), they could promote the enrichment of myeloid-derived suppressor cells (MDSC) and induce immune evasion.

Figure 3.

L1-FGGY corresponded with smoke-induced LINE-1 promoter hypomethylation, lipid metabolism dysregulation, and immune microenvironment alteration. A, The KEGG analysis between L1-FGGY+ and L1-FGGY tissues (N = 10). B, GSEA showed the distribution of immune cells. C, The expression of cytokines was compared. D, Upregulated gene validation implicated in lipid-related metabolism (N = 30). E, IHC staining results of CD3+ T cells and PD-L1+ tumor cells. F, Validation of altered cytokines. G, Spearman rank correlation between LINE-1 methylation level and L1-FGGY expression. H,LINE-1 methylation was compared between different groups. I,L1-FGGY expression was compared. *, P < 0.05; **, P < 0.01; N.S., nonsignificant.

Figure 3.

L1-FGGY corresponded with smoke-induced LINE-1 promoter hypomethylation, lipid metabolism dysregulation, and immune microenvironment alteration. A, The KEGG analysis between L1-FGGY+ and L1-FGGY tissues (N = 10). B, GSEA showed the distribution of immune cells. C, The expression of cytokines was compared. D, Upregulated gene validation implicated in lipid-related metabolism (N = 30). E, IHC staining results of CD3+ T cells and PD-L1+ tumor cells. F, Validation of altered cytokines. G, Spearman rank correlation between LINE-1 methylation level and L1-FGGY expression. H,LINE-1 methylation was compared between different groups. I,L1-FGGY expression was compared. *, P < 0.05; **, P < 0.01; N.S., nonsignificant.

Close modal

To confirm the alteration of lipid-related metabolic pathways based on the RNA sequencing data, we examined the top upregulated genes involved in these metabolic pathways, in 60 LUSC tissues (30 L1-FGGY+ vs. 30 L1-FGGY). The results showed the genes involved in cytochrome P450, arachidonic acid (AA) metabolism, and glycerolipid metabolism were upregulated in L1-FGGY+ tissues (Fig. 3D). We then further performed IHC staining, and less CD3+ T cells were detected, and the expression of PD-L1 was increased in L1-FGGY+ tissues (Fig. 3E). Then, we also validated the abnormal transcription of cytokines related with the immunosuppressive micromilieu. We found that T-cell activation–related cytokines, including IFNγ, IL17, and IL27, were downregulated, whereas immune suppression–related cytokines, including IL1α, IL6, and IL34, were upregulated in L1-FGGY+ tissues (Fig. 3F).

Then, we tested the correlation between the expression of L1-FGGY and the methylation ratio of LINE-1 promoter of the 60 tumor samples mentioned above by quantitative methylation-specific PCR. Spearman rank correlation showed a significant negative correlation between LINE-1 methylation levels and the abundance of L1-FGGY expression (Fig. 3G). Then we assessed the association between LINE-1 methylation and clinicopathology parameters. We found the reduced LINE-1 methylation was significantly associated with clinical stage of disease, lymph node activation/metastasis, and smoking history (Fig. 3H). And further analysis showed high L1-FGGY expression was significantly associated with clinical stage of disease and smoking history (Fig. 3I), implicating smoke-induced hypomethylation of LINE-1 promoter led to L1-FGGY upregulation and more activity in tumors.

Collectively, these results indicated that L1-FGGY corresponds with smoke-induced LINE-1 promoter hypomethylation, as well as with lipid metabolism upregulation and immune evasion.

L1-FGGY inhibited the transcription of FGGY gene

Insertion of LINE-1 retrotransposition into a gene via a target-primed reverse transcription mechanism (36) may interrupt the gene structure and alter its expression (37). To test whether the L1-FGGY influenced FGGY expression, we examined the expression correlation between L1-FGGY and FGGY in 52 pairs of LUSC tissues and matched normal tissues. The results showed higher expression of FGGY over L1-FGGY in normal tissues; however, higher expression of L1-FGGY over FGGY (Fig. 4A) in LUSC tissues. Furthermore, Spearman rank correlation showed a significant negative correlation between L1-FGGY and FGGY expression in the 109 LUSC tumor samples (Fig. 4B). In normal lung tissues, we detected FGGY expression, but could not detect L1-FGGY expression (Fig. 4C). Furthermore, we found relatively high expression of FGGY among normal cell lines HEK293T and BEAS-2B, and lung cancer cell lines with undetectable L1-FGGY in A549, H1299, and NCI-H446. Conversely, relatively low expression of FGGY was detected in lung cancer cell lines with abundance for L1-FGGY in NCI-H520 and NCI-H460, which indicated the reversed correlation of L1-FGGY and FGGY in multiple cell lines (Fig. 4D).

Figure 4.

L1-FGGY inhibited the transcription of FGGY gene. A, The relative expression of FGGY and L1-FGGY detected in the adjacent normal tissues and LUSC tissues. B, Spearman rank correlation between L1-FGGY and FGGY. C, The relative expression of L1-FGGY and FGGY in normal lung tissues. D, The reversed correlation between L1-FGGY and FGGY in cell lines. E, Normalized expression of FGGY between normal lung tissues and LUSC tissues (N = 50) from TCGA data. F, The OS was compared between FGGY+ and FGGY patients in our cohort. G, The different tumor T stages were compared.

Figure 4.

L1-FGGY inhibited the transcription of FGGY gene. A, The relative expression of FGGY and L1-FGGY detected in the adjacent normal tissues and LUSC tissues. B, Spearman rank correlation between L1-FGGY and FGGY. C, The relative expression of L1-FGGY and FGGY in normal lung tissues. D, The reversed correlation between L1-FGGY and FGGY in cell lines. E, Normalized expression of FGGY between normal lung tissues and LUSC tissues (N = 50) from TCGA data. F, The OS was compared between FGGY+ and FGGY patients in our cohort. G, The different tumor T stages were compared.

Close modal

We further validated the lower transcription levels of FGGY in LUSC tissues compared with matched normal tissues (Fig. 4E) in TCGA dataset. Lastly, we observed that, in contrast to L1-FGGY, low expression of FGGY was associated with poor OS (Fig. 4F) and the large tumor size (Fig. 4G; Supplementary Table S3) in our cohort.

Collectively, these data suggest the expression of FGGY is suppressed by LINE-1 retrotransposition, which could lead to the tumor progression and poor outcomes.

Overexpression of L1-FGGY or knockdown of FGGY exhibited an oncogenic role in vitro and in vivo

In order to investigate the underlying biological role of L1-FGGY and FGGY in carcinogenesis and cell proliferation in LUSC NCI-H520 cell line, we overexpressed L1-FGGY (H520OV-L1-FGGY) by a recombinant lentivirus carrying synthetic L1-FGGY core sequences and suppressed FGGY expression with shRNA (H520sh-FGGY). First, we confirmed overexpression of L1-FGGY in H520OV-L1-FGGY cells compared with the control cell line infected with empty vector (H520OV-CTRL; Fig. 5A), which suppressed the transcription of FGGY (Fig. 5B, left) by insertion into endogenous FGGY locus through homologous recombination (Supplementary Fig. S4), similar to the results in H520sh-FGGY compared with H520sh-CTRL (Fig. 5B, right).

Figure 5.

Overexpression of L1-FGGY or knockdown of FGGY exhibited an oncogenic role in vitro and in vivo. A,L1-FGGY expression in H520OV-CTRL and H520OV-L1-FGGY cells. B,FGGY expression in different cells. C, The expression of lipid metabolism–related genes. The results for H520OV-L1-FGGY were relative expression values compared with H520OV-CTRL, and the results for H520sh-FGGY were relative expression values compared with H520sh-CTRL. D, Cell proliferation results. E, Cell apoptosis results. F, Representative images in wound-healing assays. G, Representative images in transwell invasion assays. H, The expression of EMT marker genes. I, The expression of cytokines. J, Representative image of the forming tumors and the size of them at various time points upon injection. *, P < 0.05; **, P < 0.01; N.S., nonsignificant.

Figure 5.

Overexpression of L1-FGGY or knockdown of FGGY exhibited an oncogenic role in vitro and in vivo. A,L1-FGGY expression in H520OV-CTRL and H520OV-L1-FGGY cells. B,FGGY expression in different cells. C, The expression of lipid metabolism–related genes. The results for H520OV-L1-FGGY were relative expression values compared with H520OV-CTRL, and the results for H520sh-FGGY were relative expression values compared with H520sh-CTRL. D, Cell proliferation results. E, Cell apoptosis results. F, Representative images in wound-healing assays. G, Representative images in transwell invasion assays. H, The expression of EMT marker genes. I, The expression of cytokines. J, Representative image of the forming tumors and the size of them at various time points upon injection. *, P < 0.05; **, P < 0.01; N.S., nonsignificant.

Close modal

We next validated upregulation of lipid metabolism–related genes in H520OV-L1-FGGY and H520sh-FGGY cells by qPCR assay. We first validated the upregulation of the genes in AA metabolism and cytochrome P450 metabolism (Fig. 5C). Then, we detected the genes related to glucose and lipid metabolism, which mediates ATP generating in cells. The genes involved in fatty acid oxidation were upregulated, and genes involved in glucose aerobic oxidation and glycolysis were altered slightly (Fig. 5C).

Then, we further explored the role of FGGY in carcinogenesis. We found a greater proliferation rate in H520OV-L1-FGGY and H520sh-FGGY (Fig. 5D) than the control cell lines via the Cell Counting Kit 8 (CCK8) proliferation assay. Consistently, a reduced apoptotic rate was found in H520OV-L1-FGGY and H520sh-FGGY (Fig. 5E) using the Annexin-V apoptosis assay. These results implied that LINE-1 inserted into FGGY significantly stimulated cell proliferation and reduced cell apoptosis through suppressing FGGY.

The migration and invasion capacities of H520OV-L1-FGGY and H520sh-FGGY were evaluated via the wound-healing assays and by the transwell invasion assay preformed. We found that the wound closure rate of H520OV-L1-FGGY and H520sh-FGGY (Fig. 5F) were significantly higher than corresponding control cell lines. Consistently, more of those cells expressing H520OV-L1-FGGY and H520sh-FGGY (Fig. 5G) migrate across the matrigel layer after 48 hours. Furthermore, the enriched transcription of epithelial–mesenchymal transition (EMT)–related biomarker was detected using qPCR assay. The conventionally used epithelium cell biomarker of E-cadherin was reduced, but other well-described mesenchymal cell biomarkers of N-cadherin and β-catenin and EMT-related transcription factors (snail, slug, zeb1, and Twist) were increased in those cells expressing H520OV-L1-FGGY and H520sh-FGGY (Fig. 5H). The qPCR results also showed that immune suppression–related cytokines, including IL1α, IL6, IL33, and IL34, were upregulated, whereas T-cell activation–related cytokines, including IFNγ, IFNλ, IL17, and IL27, were downregulated in H520OV-L1-FGGY and H520sh-FGGY cells (Fig. 5I). These results imply that LINE-1 inserted into FGGY promotes cell invasion and migration by stimulating the EMT phenotype, accompanied by influencing the transcription of cytokines in tumor cells.

Next, we tested if L1-FGGY would affect carcinogenesis in vivo. Cells transduced with the H520OV-L1-FGGY and H520sh-FGGY lentiviral constructs and their control cells were injected into NOD-SCID mice subcutaneously as xenografts. After 24 days, the average volumes of tumors generated by engrafted tumor cells from the H520OV-L1-FGGY and H520sh-FGGY groups (Fig. 5J) were at least 2-fold greater volumes when compared with the control groups. Consistently, the growth rates of tumors in H520OV-L1-FGGY mice were much greater than those in H520OV-CTRL mice (Fig. 5J). To further validate our observations in NCI-H520, we repeated the experiments in another LUSC cell line, SK-MES-1, to further validate our observations in NCI-H520. We found consistent results in SK-MES-1, in which forced overexpression of L1-FGGY or knockdown of FGGY promoted cell proliferation and migration, reduced cell apoptosis, as well as implicated with dysregulation of lipid metabolism and cytokine transcription (Supplementary Fig. S5).

Taken together, these findings implied that L1-FGGY could significantly suppress FGGY expression, stimulate cell proliferation, inhibit cell apoptosis, promote cell invasion and EMT, and thereby predicted to facilitate carcinogenesis. Besides, L1-FGGY also seemed to be implicated in cell energy metabolism and cytokine transcription.

Reverse transcriptase inhibitors inhibited the proliferation of LUSC cells and impaired the growth of LUSC xenografts in vitro and in vivo via recovering FGGY

Reverse transcriptase inhibitors such as nevirapine (NVR) and efavirenz (EFV) block the enzymatic activity of endogenous reverse transcriptase and regarded as potential specific inhibitors for LINE-1 retrotranspositions (38). We tested the antitumor effort of NVR and EFV in LUSC. In the CCK8 experiment, we show that both inhibitors could effectively reduce the proliferation of NCI-H520 cells (Fig. 6A, top). In order to discover the toxicity of NVR and EFV in the normal lung cell line, we then performed CCK8 experiment in BEAS-2B cell. The results showed that NVR did not influence the proliferation rate, though EFV showed some inhibition on the proliferation of BEAS-2B (Fig. 6A, bottom), which indicated NVR is safer and has less toxicity than EFV. And both NVR and EFV further increase cell apoptosis (Fig. 6B). Furthermore, both NVR and EFV decrease the invasive potential of NCI-H520 cells (Fig. 6C). Upon NVR and EFV treatment, the RNA level of both β-catenin and slug is reduced and accompanied by increases in the RNA level of E-cadherin (Fig. 6D). Collectively, our data indicate that the reverse transcriptase inhibitors attenuate cell proliferation and invasion by inhibiting L1-FGGY, leading to upregulation of FGGY expression.

Figure 6.

Reverse transcriptase inhibitors inhibited the proliferation of LUSC cells in vitro and in vivo via recovering FGGY. A, NCI-H520 cells were treated with reverse transcriptase inhibitors NVR or EFV, followed by CCK8 detection. B, Cell apoptosis results. C, Representative images in transwell invasion assays. D, The expression of EMT marker genes. All the results for NVR- and EFV-treated NCI-H520 cells were relative expression values compared with DMSO-treated cells. E, NCI-H520 cells were inoculated subcutaneously in the NOD-SCID mice, which were subjected to either NVR or EFV treatment. Representative image of the forming tumors and the size of them at various time points upon injection. F, The expression of EMT marker genes. All the results for mice subjected to DMSO, NVR, and EFV were relative expression values compared with the mice with no treatment. G,L1-FGGY expression. H,FGGY expression. I, The expression of lipid metabolism–related gene. J, The expression of cytokines. K, IHC staining results. The data are shown as mean ± SD. *, P < 0.05; **, P < 0.01, between the groups as indicated.

Figure 6.

Reverse transcriptase inhibitors inhibited the proliferation of LUSC cells in vitro and in vivo via recovering FGGY. A, NCI-H520 cells were treated with reverse transcriptase inhibitors NVR or EFV, followed by CCK8 detection. B, Cell apoptosis results. C, Representative images in transwell invasion assays. D, The expression of EMT marker genes. All the results for NVR- and EFV-treated NCI-H520 cells were relative expression values compared with DMSO-treated cells. E, NCI-H520 cells were inoculated subcutaneously in the NOD-SCID mice, which were subjected to either NVR or EFV treatment. Representative image of the forming tumors and the size of them at various time points upon injection. F, The expression of EMT marker genes. All the results for mice subjected to DMSO, NVR, and EFV were relative expression values compared with the mice with no treatment. G,L1-FGGY expression. H,FGGY expression. I, The expression of lipid metabolism–related gene. J, The expression of cytokines. K, IHC staining results. The data are shown as mean ± SD. *, P < 0.05; **, P < 0.01, between the groups as indicated.

Close modal

NCI-H520 cells were engrafted subcutaneously in the NOD-SCID mice, which were subjected to either NVR (50 mg/kg/day) or EFV (20 mg/kg/day) treatment. After 22 days, the average volume of tumors in inhibitor-treated mice was characterized as being much smaller than those in either untreated or DMSO-treated mouse groups (Fig. 6E, left). The tumor growth curves indicate that both inhibitors markedly reduce the growth of the xenografts shown (Fig. 6E, right). We determine that NVR inhibits tumor growth more effectively than EFV. Moreover, both inhibitors display comparable drug safety because neither significant body weight loss nor mortality occurred during the treatment.

Furthermore, transcription of multiple EMT gene markers in xenografts was examined. As expected, the qPCR analysis demonstrated that the epithelial cell marker E-cadherin was elevated after the treatment, whereas the mesenchymal cell markers N-cadherin, β-catenin, snail, slug, zeb1, and Twist1 were reduced (Fig. 6F). Similarly, the mRNA level of L1-FGGY was reduced (Fig. 6G), whereas the mRNA level of FGGY increased dramatically after both inhibitors’ treatment (Fig. 6H), consistent with increased protein expression by IHC analysis. Transcription of the lipid-related genes and fatty acid oxidation–related genes in xenografts was reduced after the treatment (Fig. 6I), which was opposite to increased transcription in H520OV-L1-FGGY and H520sh-FGGY cells. Similarly, the mRNA level of IFNγ, IFNλ, IL17, and IL27 was increased after the treatment, whereas the RNA level of IL1α, IL6, IL33, and IL34 was reduced (Fig. 6J).

Further staining by IHC reveals that less Ki67+, N-cadherin+, and β-catenin+ cells (Fig. 6K) were detected with both inhibitors from the two groups of treated mice. Meanwhile, we noticed that NVR and EFV could also decrease the expression of PD-L1 on tumor cells. And even CD11b+ MDSCs were reduced in the mice treated with the two inhibitors (Fig. 6K). Collectively, these data clearly suggest that the reverse transcriptase inhibitors reduce tumor growth and appear to reflect the local immune evasion in vivo by suppressing L1-FGGY, leading to increased FGGY expression.

Somatic LINE-1 retrotransposition has been detected in multiple types of tumors (39–41). However, the expression and function of LINE-1 retrotransposition in LUSC were unclear. Considering most of LINE-1 studies were exclusively based on DNA-seq (39, 42), in this study, we carried out a new approach to analyze somatic LINE-1 retrotransposition at transcriptomic level, which enabled us to explore their functions and mechanisms in a quantitative manner.

Based on the RNA sequencing data from TCGA and Chinese cohort, we identified L1-FGGY as being the most frequent LINE-1 retrotransposition in Chinese LUSC patients that significantly correlated with poor clinical outcome. We observed the significant correlation between smoking history and high occurrence of L1-FGGY in LUSC. Tobacco exposure is the leading cause of cancers (43). Multiple studies suggest the relationship between tobacco exposure and the occurrence and even worse outcome of lung cancer (44, 45). Furthermore, we confirmed a correlation between smoking history and LINE-1 hypomethylation. Therefore, we proposed a hypothesis that tobacco smoking induces hypomethylation of CpG islands at LINE-1 promoter region in lung epithelium cells and, in turn, activates LINE-1 insertion into tumor-suppressor genes, such as FGGY, to facilitate carcinogenesis.

Recently, a study by Jung and colleagues had discovered multiple immune pathways were significantly negatively correlated with LINE-1 insertion counts, including immunoregulatory interactions between a lymphoid and a nonlymphoid cell, Toll-like receptors’ cascade, STAT6-mediated induction of chemokines, and IFN signaling (46). Their study uncovered the correlation between LINE-1 retrotranspositions and suppressive immune signatures in gastrointestinal cancers, which is consistent with our findings in LUSC. FGGY is a novel tumor-suppressor gene that was known to encode a protein that phosphorylates carbohydrates and associates with obesity and sporadic amyotrophic lateral sclerosis (47). But less information about FGGY in carcinogenesis has been provided yet. Here, we found the LINE-1 insertion could inhibit the transcription of FGGY gene probably via homologous recombination.

We firstly compared the energy metabolism because tumor cells might enhance energy intake through increasing glucose uptake and aerobic glycolysis (48). But no alteration of glucose glycolysis in L1-FGGY+ tumors was detected. Meanwhile, a number of lipid metabolism–related genes changed dramatically. It was reported that FGGY suppression caused lipid metabolism disturbance and diet-induced obesity in mice (49). Therefore, we compared the genes involved in fatty acid oxidation, which significantly increased in H520OV-L1-FGGY cells and H520sh-FGGY cells, which are consistent with previous report (50, 51). Hence, we hypothesize that L1-FGGY+ cells utilized fatty acid oxidation to provide energy for tumor cell growth and invasion.

Secondly, we studied local immune evasion. L1-FGGY inhibited the infiltration of T cells, promoted the recruitment of immunosuppressive cells, and enhanced the expression of PD-L1 on tumor cells. Recent studies have shown that dysfunctional AA metabolism affected the immune system significantly. It has been reported that the cancer cell–derived AA metabolite LTB4 suppressed antitumorigenic cytotoxic CD8+ T cells and recruited multiple immunosuppressive cells (52). We found enhanced AA metabolism-related signaling pathways and multiple inflammatory cytokines/chemotaxins in L1-FGGY+ tumors. Among them, type I IFN was regulated by AA in a dose-dependent manner (53) and considered as a critical molecule to induce PD-L1 upregulation in tumors (54). IL6 and IL1α were preferentially attracted and activated tumor-associated macrophages and MDSCs, which subsequently suppress cytotoxic CD8+ T-cell activity (55, 56).

Therefore, we summarized the findings and complemented our hypothesis in Supplementary Fig. S6. Tobacco smoking induced more L1-FGGY in lung epithelium cells, which disrupted the transcription and expression of FGGY, as well as FGGY-related adipose metabolism, thus generated more energy via enhancing fatty acid oxidation, to enhance the cell proliferation and invasion. On the other hand, metabolic disorder caused tumor microenvironment abnormality, leading to alterations in cytokine/chemotaxin profiles, followed by T-cell suppression, immunosuppressive cell recruitment, and PD-L1 upregulation.

In order to evaluate the therapeutic value of somatic LINE-1 retrotransposition, the inhibitors targeting the reverse transcription enzyme during the retrotransposition process as FDA-approved antiretroviral drugs for HIV infection treatments (57) were applied. NVR and EFV share common chemical properties and biochemical effects by binding the hydrophobic pocket in the subunit of many reverse transcriptases (58). These inhibitors were reported to reduce cell proliferation (59). Here, we found that NVR and EFV efficiently inhibited cell proliferation and invasion in vitro, as well as impaired tumor growth in vivo. Furthermore, NVR and EFV interfered with the expression pattern of cytokines, decreased the expression of PD-L1, and inhibited the infiltration of immunosuppressive cells. Considering higher antitumor efficiency and less toxicity of NVR, we proposed NVR as a more promising candidate drug, which could even convert the immunosuppressive tissue microenvironment into immunoactive. Therefore, combining treatment of reverse transcriptase inhibitors probably improves the therapeutic efficacy of checkpoint inhibitors in LINE-1–positive LUSC patients.

Collectively, we established a novel strategy to analyze retrotranspositions at the transcriptomic level and identified somatic LINE-1 retrotranspositions, especially L1-FGGY, associated with the development and progression of LUSC. It is the first time that L1-FGGY was proposed as a candidate biomarker that not only enhanced the proliferation and invasion potential of tumor cells, but also affected local immune homeostasis via disrupting cell energy metabolism. Finally, we proposed a practical therapeutic strategy to overcome L1-FGGY–mediated carcinogenesis using FDA-approved antiretroviral drugs, NVR and EFV. Although more extensive experiments need to be conducted to elucidate the concrete mechanisms, L1-FGGY is feasible as a promising predictive biomarker and therapeutic target in LUSC.

No potential conflicts of interest were disclosed.

Conception and design: R. Zhang, W. Zhang, J. Yu

Development of methodology: R. Zhang, F. Zhang, Z. Sun, P. Liu, X. Zhang, W. Zhang, J. Yu

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R. Zhang, P. Liu, X. Zhang, Y. Ye

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Zhang, F. Zhang, Z. Sun, P. Liu, B. Cai

Writing, review, and/or revision of the manuscript: R. Zhang, F. Zhang, M.J. Walsh, W. Zhang, J. Yu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): X. Ren, X. Hao, J. Yu

Study supervision: X. Ren, X. Hao, W. Zhang, J. Yu

We thank Cancer Biobank of Tianjin Medical University Cancer Institute and Hospital for their help.

This work was supported by National Natural Science Foundation of China (Grant Nos. 81702280, 81472473, and 81872143), National Science and Technology support Program of China (Grant Nos. 2015BAI12B15 and 2018ZX09201015), National Key Research and Development program of China: The Net construction of human genetic resource Bio-bank in North China (2016YFC1201703), Projects of Science and Technology of Tianjin (Grant Nos. 13ZCZCSY20300 and 18JCQNJC82700), Key project of Tianjin Health and Family Planning Commission (Grant No. 16KG126), and Tianjin Medical University Cancer Institute and Hospital Research Program (No. B1618).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Rafiemanesh
H
,
Mehtarpour
M
,
Khani
F
,
Hesami
SM
,
Shamlou
R
,
Towhidi
F
, et al
Epidemiology, incidence and mortality of lung cancer and their relationship with the development index in the world
.
J Thoracic Disease
2016
;
8
.
2.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2018
.
Ca-a Cancer J Clinicians
2018
;
68
:
7
30
.
3.
Thomas
A
,
Liu
SV
,
Subramaniam
DS
,
Giaccone
G.
Refining the treatment of NSCLC according to histological and molecular subtypes
.
Nat Rev Clin Oncol
2015
;
12
:
511
26
.
4.
Gridelli
C
,
Rossi
A
,
Carbone
DP
,
Guarize
J
,
Karachaliou
N
,
Mok
T
, et al
Non-small-cell lung cancer
.
Nat Rev Dis Primers
2015
;
1
:
15009
.
doi: 10.1038/nrdp.2015.9
.
5.
Lu
S
. 
Development of treatment options for Chinese patients with advanced squamous cell lung cancer: focus on afatinib
.
Onco Targets Ther
2019
;
12
:
1521
38
.
6.
Koutsoukos
K
,
Mountzios
G
. 
Novel therapies for advanced squamous cell carcinoma of the lung
.
Future Oncology
2016
;
12
:
659
67
.
7.
Planchard
D
,
Popat
S
,
Kerr
K
,
Novello
S
,
Smit
EF
,
Faivre-Finn
C
, et al
Metastatic non-small cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up
.
Ann Oncol
2018
;
29
:
192
237
.
8.
Gettinger
SN
,
Horn
L
,
Gandhi
L
,
Spigel
DR
,
Antonia
SJ
,
Rizvi
NA
, et al
Overall survival and long-term safety of nivolumab (Anti-Programmed Death 1 Antibody, BMS-936558, ONO-4538) in patients with previously treated advanced non-small-cell lung cancer
.
J Clin Oncol
2015
;
33
:
2004
U2032
.
9.
Stinchcombe
TE
. 
Unmet needs in squamous cell carcinoma of the lung: potential role for immunotherapy
.
Med Oncol
2014
;
31
.
10.
Goodman
AM
,
Kato
S
,
Bazhenova
L
,
Patel
SP
,
Frampton
GM
,
Miller
V
, et al
Tumor mutational burden as an independent predictor of response to immunotherapy in diverse cancers
.
Mol Cancer Ther
2017
;
16
:
2598
608
.
11.
Yaghmour
G
,
Pandey
M
,
Ireland
C
,
Patel
K
,
Nunnery
S
,
Powell
D
, et al
Role of genomic instability in immunotherapy with checkpoint inhibitors
.
Anticancer Res
2016
;
36
:
4033
8
.
12.
Brahmer
JR
,
Tykodi
SS
,
Chow
LQM
,
Hwu
WJ
,
Topalian
SL
,
Hwu
P
, et al
Safety and activity of anti-PD-L1 antibody in patients with advanced cancer
.
N Engl J Med
2012
;
366
:
2455
65
.
13.
Hancks
DC
,
Kazazian
HH
. 
Roles for retrotransposon insertions in human disease
.
Mobile DNA
2016
;
7
.
14.
Percharde
M
,
Lin
CJ
,
Yin
Y
,
Guan
J
,
Peixoto
GA
,
Bulut-Karslioglu
A
, et al
A LINE1-nucleolin partnership regulates early development and ESC identity
.
Cell
2018
;
174
:
391
405
e319
.
15.
Rodic
N
,
Steranka
JP
,
Makohon-Moore
A
,
Moyer
A
,
Shen
P
,
Sharma
R
, et al
Retrotransposon insertions in the clonal evolution of pancreatic ductal adenocarcinoma
.
Nat Med
2015
;
21
:
1060
4
.
16.
Burns
KH
. 
Transposable elements in cancer
.
Nat Rev Cancer
2017
;
17
:
415
24
.
17.
Tubio
JMC
,
Li
YL
,
Ju
YS
,
Martincorena
I
,
Cooke
SL
,
Tojo
M
, et al
Extensive transduction of nonrepetitive DNA mediated by L1 retrotransposition in cancer genomes
.
Science
2014
;
345
:1251343.
18.
Beck
CR
,
Garcia-Perez
JL
,
Badge
RM
,
Moran
JV
. 
L1 elements in structural variation and disease
.
Annu Rev Genomics Hum Genet
2011
;
12
:
187
215
.
19.
Ewing
AD
,
Ballinger
TJ
,
Earl
D
,
Harris
CC
,
Ding
L
,
Wilson
RK
, et al
Retrotransposition of gene transcripts leads to structural variation in mammalian genomes
.
Genome Biol
2013
;
14
.
20.
Kemp
JR
,
Longworth
MS
. 
Crossing the LINE toward genomic instability: LINE-1 retrotransposition in cancer
.
Frontiers in Chemistry
2015
;
3
.
21.
Morse
B
,
Rotherg
PG
,
South
VJ
,
Spandorfer
JM
,
Astrin
SM
. 
Insertional mutagenesis of the myc locus by a L1 sequence in a human breast carcinoma
.
Nature
1988
;
333
:
87
90
.
22.
Miki
Y
,
Nishisho
I
,
Horii
A
,
Miyoshi
Y
,
Utsunomiya
J
,
Kinzler
KW
, et al
Disruption of the APC gene by a retrotransposal insertion of L1 sequence in a colon cancer
.
Cancer Res
1992
;
52
:
643
5
.
23.
Gainetdinov
IV
,
Kapitskaya
KY
,
Rykova
EY
,
Ponomaryova
AA
,
Cherdyntseva
NV
,
Vlassov
VV
, et al
Hypomethylation of human-specific family of LINE-1 retrotransposons in circulating DNA of lung cancer patients
.
Lung Cancer
2016
;
99
:
127
30
.
24.
Imperatori
A
,
Sahnane
N
,
Rotolo
N
,
Franzi
F
,
Nardecchia
E
,
Libera
L
, et al
LINE-1 hypomethylation is associated to specific clinico-pathological features in Stage I non-small cell lung cancer
.
Lung Cancer
2017
;
108
:
83
9
.
25.
Liu
L
,
Zhang
L
,
Yang
L
,
Li
H
,
Li
RM
,
Yu
JP
, et al
Anti-CD47 antibody as a targeted therapeutic agent for human lung cancer and cancer stem cells
.
Front Immunol
2017
;
8
.
26.
Li
YH
,
Ning
QY
,
Shi
JL
,
Chen
Y
,
Jiang
MM
,
Gao
L
, et al
A novel epigenetic AML1-ETO/THAP10/miR-383 minicircuitry contributes to t(8;21) leukaemogenesis
.
Embo Molecular Medicine
2017
;
9
:
933
49
.
27.
Liu
PP
,
Zhang
R
,
Yu
WW
,
Ye
YN
,
Cheng
YN
,
Han
L
, et al
FGF1 and IGF1-conditioned 3D culture system promoted the amplification and cancer sternness of lung cancer cells
.
Biomaterials
2017
;
149
:
63
76
.
28.
Zhu
J
,
Ling
Y
,
Xu
Y
,
Lu
MZ
,
Liu
YP
,
Zhang
CS
, et al
Elevated expression of MDR1 associated with LINE-1 hypomethylation in esophageal squamous cell carcinoma
.
Int J Clin Exp Pathol
2015
;
8
:
14392
400
.
29.
Zhang
WW
,
Jiang
MM
,
Chen
JY
,
Zhang
R
,
Ye
YN
,
Liu
PP
, et al
SOCS3 suppression promoted the recruitment of CD11b(+)Gr-1(-)F4/80(-)MHCII(-) early-stage myeloid-derived suppressor cells and accelerated interleukin-6-related tumor invasion via affecting myeloid differentiation in breast cancer
.
Front Immunol
2018
;
9
.
30.
McPherson
A
,
Hormozdiari
F
,
Zayed
A
,
Giuliany
R
,
Ha
G
,
Sun
MG
, et al
deFuse: an algorithm for gene fusion discovery in tumor RNA-Seq data
.
Plos Comput Biol
2011
;
7
:
e1001138
.
31.
Deelman
E
,
Singh
G
,
Su
MH
,
Blythe
J
,
Gil
Y
,
Kesselman
C
, et al
Pegasus: a framework for mapping complex scientific workflows onto distributed systems
.
Scientific Programming
2005
;
13
:
219
37
.
32.
Jurka
J
,
Kapitonov
VV
,
Pavlicek
A
,
Klonowski
P
,
Kohany
O
,
Walichiewicz
J
, et al
Repbase update, a database of eukaryotic repetitive elements
.
Cytogenet Genome Res
2005
;
110
:
462
7
.
33.
Zheng
GXY
,
Terry
JM
,
Belgrader
P
,
Ryvkin
P
,
Bent
ZW
,
Wilson
R
, et al
Massively parallel digital transcriptional profiling of single cells
.
Nat Commun
2017
;
8
.
34.
Gierahn
TM
,
Wadsworth
MH
,
Hughes
TK
,
Bryson
BD
,
Butler
A
,
Satija
R
, et al
Seq-Well: portable, low-cost RNA sequencing of single cells at high throughput
.
Nat Methods
2017
;
14
:
395
.
35.
Newman
AM
,
Liu
CL
,
Green
MR
,
Gentles
AJ
,
Feng
WG
,
Xu
Y
, et al
Robust enumeration of cell subsets from tissue expression profiles
.
Nat Methods
2015
;
12
:
453
.
36.
Cost
GJ
,
Feng
QH
,
Jacquier
A
,
Boeke
JD
. 
Human L1 element target-primed reverse transcription in vitro
.
EMBO J
2002
;
21
:
5899
910
.
37.
Elbarbary
RA
,
Lucas
BA
,
Maquat
LE
. 
Retrotransposons as regulators of gene expression
.
Science
2016
;
351
.
38.
Patnala
R
,
Lee
SH
,
Dahlstrom
JE
,
Ohms
S
,
Chen
L
,
Dheen
ST
, et al
Inhibition of LINE-1 retrotransposon-encoded reverse transcriptase modulates the expression of cell differentiation genes in breast cancer cells
.
Breast Cancer Res Tr
2014
;
143
:
239
53
.
39.
Lee
E
,
Iskow
R
,
Yang
LX
,
Gokcumen
O
,
Haseley
P
,
Luquette
LJ
, et al
Landscape of somatic retrotransposition in human cancers
.
Science
2012
;
337
:
967
71
.
40.
Solyom
S
,
Ewing
AD
,
Rahrmann
EP
,
Doucet
T
,
Nelson
HH
,
Burns
MB
, et al
Extensive somatic L1 retrotransposition in colorectal tumors
.
Genome Res
2012
;
22
:
2328
38
.
41.
Shukla
R
,
Upton
KR
,
Munoz-Lopez
M
,
Gerhardt
DJ
,
Fisher
ME
,
Nguyen
T
, et al
Endogenous retrotransposition activates oncogenic pathways in hepatocellular carcinoma
.
Cell
2013
;
153
:
101
11
.
42.
Helman
E
,
Lawrence
MS
,
Stewart
C
,
Sougnez
C
,
Getz
G
,
Meyerson
M
, et al
Somatic retrotransposition in human cancer revealed by whole-genome and exome sequencing
.
Genome Res
2014
;
24
:
1053
63
.
43.
Moolgavkar
SH
,
Holford
TR
,
Levy
DT
,
Kong
CY
,
Foy
M
,
Clarke
L
, et al
Impact of reduced tobacco smoking on lung cancer mortality in the United States during 1975–2000
.
J Natl Cancer I
2012
;
104
:
541
8
.
44.
Gutierrez
M
,
Wozniak
A
,
Langer
C
,
Fang
B
,
Suero-Abreu
G
,
Norden
A
, et al
Impact of tobacco smoking on outcomes in patients with metastatic non-small cell lung cancer in the era of targeted therapy
.
J Thoracic Oncology
2018
;
13
:
S680
S1
.
45.
Gibbons
DL
,
Byers
LA
,
Kurie
JM
. 
Smoking, p53 mutation, and lung cancer
.
Mol Cancer Res
2014
;
12
:
3
13
.
46.
Jung
H
,
Choi
JK
,
Lee
EA
. 
Immune signatures correlate with L1 retrotransposition in gastrointestinal cancers
.
Genome Res
2018
;
28
:
1136
46
.
47.
Casas
BD
,
Waddell
D
. 
FGGY carbohydrate kinase domain containing is upregulated during neurogenic skeletal muscle atrophy
.
Faseb J
2016
;
30
.
48.
Hanahan
D
,
Weinberg
RA
. 
Hallmarks of cancer: the next generation
.
Cell
2011
;
144
:
646
74
.
49.
Taylor
JA
,
Shioda
K
,
Mitsunaga
S
,
Yawata
S
,
Angle
BM
,
Nagel
SC
, et al
Prenatal exposure to Bisphenol A disrupts naturally occurring bimodal DNA methylation at proximal promoter of FGGY, an obesity-relevant gene encoding a carbohydrate kinase, in gonadal white adipose tissues of CD-1 mice
.
Endocrinology
2018
;
159
:
779
94
.
50.
Zha
S
,
Ferdinandusse
S
,
Hicks
JL
,
Denis
S
,
Dunn
TA
,
Wanders
RJ
, et al
Peroxisomal branched chain fatty acid beta-oxidation pathway is upregulated in prostate cancer
.
Prostate
2005
;
63
:
316
23
.
51.
Currie
E
,
Schulze
A
,
Zechner
R
,
Walther
TC
,
Farese
RV
. 
Cellular fatty acid metabolism and cancer
.
Cell Metabolism
2013
;
18
:
153
61
.
52.
Moore
GY
,
Pidgeon
GP
. 
Cross-talk between cancer cells and the tumour microenvironment: the role of the 5-lipoxygenase pathway
.
Int J Mol Sci
2017
;
18
.
53.
Song
W
,
Jiang
R
,
Zhao
CM
. 
Regulation of arachidonic acid in esophageal adenocarcinoma cells and tumor-infiltrating lymphocytes
.
Oncol Letters
2013
;
5
:
1897
902
.
54.
Ribas
A
. 
Adaptive Immune resistance: how cancer protects from immune attack
.
Cancer Discov
2015
;
5
:
915
9
.
55.
Ostuni
R
,
Kratochvill
F
,
Murray
PJ
,
Natoli
G
. 
Macrophages and cancer: from mechanisms to therapeutic implications
.
Trends Immunol
2015
;
36
:
229
39
.
56.
Parker
KH
,
Beury
DW
,
Ostrand-Rosenberg
S
. 
Myeloid-derived suppressor cells: critical cells driving immune suppression in the tumor microenvironment
.
Adv Cancer Res
2015
;
128
:
95
139
.
57.
Holec
AD
,
Mandal
S
,
Prathipati
PK
,
Destache
CJ
. 
Nucleotide reverse transcriptase inhibitors: a thorough review, present status and future perspective as HIV therapeutics
.
Curr HIV Res
2017
;
15
:
411
21
.
58.
Ren
J
,
Nichols
C
,
Bird
L
,
Chamberlain
P
,
Weaver
K
,
Short
S
, et al
Structural mechanisms of drug resistance for mutations at codons 181 and 188 in HIV-1 reverse transcriptase and the improved resilience of second generation non-nucleoside inhibitors
.
J Mol Biol
2001
;
312
:
795
805
.
59.
Bruning
A
,
Burger
P
,
Gingelmaier
A
,
Mylonas
I
. 
The HIV reverse transcriptase inhibitor tenofovir induces cell cycle arrest in human cancer cells
.
Invest New Drugs
2012
;
30
:
1389
95
.