The inherited variations of a p53-responsive enhancer in 13q12.12 confer lung cancer risk by attenuating TNFRSF19 expression
Inherited factors contribute to lung cancer risk, but the mechanism is not well understood. Defining the biological consequence of GWAS hits in cancers is a promising strategy to elucidate the inherited mechanisms of cancers. The tag-SNP rs753955 (A>G) in 13q12.12 is highly associated with lung cancer risk in the Chinese population. Here, we systematically investigate the biological significance and the underlying mechanism behind 13q12.12 risk locus in vitro and in vivo.
We characterize a novel p53-responsive enhancer with lung tissue cell specificity in a 49-kb high linkage disequilibrium block of rs753955. This enhancer harbors 3 highly linked common inherited variations (rs17336602, rs4770489, and rs34354770) and six p53 binding sequences either close to or located between the variations. The enhancer effectively protects normal lung cell lines against pulmonary carcinogen NNK-induced DNA damages and malignant transformation by upregulating TNFRSF19 through chromatin looping. These variations significantly weaken the enhancer activity by affecting its p53 response, especially when cells are exposed to NNK. The effect of the mutant enhancer alleles on TNFRSF19 target gene in vivo is supported by expression quantitative trait loci analysis of 117 Chinese NSCLC samples and GTEx data. Differentiated expression of TNFRSF19 and its statistical significant correlation with tumor TNM staging and patient survival indicate a suppressor role of TNFRSF19 in lung cancer.
This study provides evidence of how the inherited variations in 13q12.12 contribute to lung cancer risk, highlighting the protective roles of the p53-responsive enhancer-mediated TNFRSF19 activation in lung cells under carcinogen stress.
KeywordsLung cancer Risk SNP Enhancer TNFRSF19 13q12.12 risk locus
Lung cancer is the most common type of cancer and shows complex pathogenesis and high heterogeneity. Variations in the predisposition to the disease and disease progression in different ethnic groups imply important roles for germline genetic factors in lung cancer pathogenesis . Although genomic alterations discovered in NSCLC have provided valuable clues for understanding the molecular pathogenesis and genetic susceptibility associated with this disease [2, 3, 4, 5, 6], the inherited mechanism of lung cancer has not been well understood yet.
The most common genetic variation in humans is single nucleotide polymorphism (SNP). Genome-wide association study (GWAS) has led to an explosion in the identification of SNPs associated with a variety of complex diseases, including breast, colon, lung, and pancreatic cancers [7, 8, 9, 10, 11, 12]. Most of these risk SNPs are located in non-coding regions of the genome. At present, more than a dozen lung cancer risk non-coding SNPs have been identified in various ethnic populations and regions of the world [13, 14, 15]. Nevertheless, how these specific risk variants contribute functionally to lung cancer susceptibility and pathogenesis remains unclear.
One challenge to understand the biological functions of specific non-coding DNA risk variants is that these variants do not alter the amino acid composition of a protein. In addition, GWAS attempts to “tag” the approximate locations of disease variants and can identify disease-associated alleles with strong linkage disequilibrium (LD) to tagged SNPs, rather than identifying disease-causative SNPs. Many identified disease-associated SNPs may therefore function as genetic markers or indicators, which increase the complexity of elucidating their biological significance.
The ENCODE Project Consortium has revealed that arrays of long-range regulatory elements are interspersed throughout the whole genome, and most are enhancers [16, 17]. Notably, thousands of GWAS variants have been localized to enhancer elements identified through epigenomic profiling studies. GWAS SNPs are usually correlated with enhancer elements marked with H3K4me1, H3K27ac, and H3K4me3 [18, 19, 20, 21]. By contrast, only 10–15% are in LD with a protein-coding variant [22, 23]. Moreover, mutations in single or multiple enhancers are responsible for numerous human diseases, including preaxial polydactyly and Hirschsprung’s disease; breast, prostate, and colon cancers; and human autoimmune traits [24, 25, 26]. For example, the new studies reported by Gao et al. and Hua et al. showed that a risk SNP resided in an enhancer directly impact on PCTA19 and CEACAM21 gene expression and prostate cancer prognosis [27, 28]. These studies all support a hypothesis that non-coding causal GWAS variants can contribute to common diseases by perturbing enhancer regulatory activity and consequently interfering with the target gene expression.
The newly identified 13q12.12 locus is highly associated with lung cancer risk in the Han Chinese population. The well-replicated GWAS risk SNP rs753955 (A>G) within this locus is situated in the gene desert region, about 150 kb away from the nearest upstream gene, TNFRSF19. The underlying biological effects of the 13q12.12 lung cancer risk locus are unknown.
In this study, we used an integrative strategy of bioinformatics, laboratory experiments, and clinical analyses to investigate the causative mechanism underlying lung cancer susceptibility associated with the 13q12.12 locus. Our in vitro and in vivo data provided evidence that three inherited causal variations rs17336602 (G>C), rs4770489 (A>G), and rs34354770 (A>C) in 13q12.12 contributed to the lung cancer risk by attenuating the p53-responsive enhancer-mediated TNFRSF19 activation. Our findings provided new insight into the understanding of the lung cancer inherited mechanisms.
Identification of the active enhancer within the 13q12.12 locus in high linkage disequilibrium with the risk rs753955
Subsequently, we tested the regulatory activity and cell type specificity of the enhancer element by cloning the element into pGL3-promoter vectors for luciferase activity tests in different cancer and normal cell lines. Figure 1c showed dramatic enhancer activity displayed by the 13q-Enh element in three normal lung tissue cell lines, Beas-2B human bronchial epithelial cell line, HFL1, and MRC-5 human fetal lung fibroblast cell lines, with 3 to 6 times higher activity than the control, and significantly higher than in other normal tissue cell lines and cancer cell lines. ChIP assays using anti-H3K4me1 and H3K27ac antibodies confirmed the enrichment of H3K4me1 and H3K27ac, the histone marks for active enhancers, on the 13q-Enh (Additional file 1: Figure S1). The tissue specificity of the 13q-Enh enhancer was further evaluated using the available H3K27ac ChIP-seq data for lung and 11 non-lung tissues released by ENCODE database. The 13q-Enh enhancer was rich in H3K27ac in lung tissue. In contrast, the 13q-Enh enhancer was seldom rich in H3K27ac in non-lung tissues except for kidney and breast tissue (Fig. 1d). These in vivo and in vitro data indicated that the 13q-Enh was an active enhancer with high lung tissue specificity.
Deletion of the 13q-Enh enhancer promoted NNK-induced malignant transformation of Beas-2B bronchial epithelial cells
Deletion of the 13q-Enh enhancer reduced DNA repair efficiency and apoptosis responses of Beas-2B human bronchial epithelial cells
Increased DNA damage, together with abnormal apoptosis responses, are well known to contribute to carcinogenesis. We further explored the functions of the 13q-Enh enhancer in cell transformation by using γ-H2AX assays to compare the DNA damage levels in NNK-treated 13q-Enh−/− clones and NNK-treated wild-type Beas-2B cells. Figure 2c and d show significantly increased γ-H2AX signals in the C5-NNK and C23-NNK clones when compared with the NNK-treated wild-type cells, indicating that deletion of the 13q-Enh enhancer increased the sensitivity to NNK-induced DNA breakage. We then used host-cell reactivation (HCR) assays to compare the DNA repair capability of 13q-Enh−/− clones and wild-type Beas-2B cells. The HCR assay is a well-established method to determine the DNA repair capacity of cells. A diagram to explain the principle of the HCR assays was displayed in Additional file 1: Figure S3. Briefly, the pGL3-promoter luciferase plasmids were treated by H2O2 (v/v) at room temperature for 1 h to induce DNA breakages. The cells were transfected by the H2O2-treated plasmids, cultured and harvested at different time points. The plasmids were then purified by ethanol precipitation for luciferase assays. The DNA repair capacity of cells is reflected by the fluorescence curve. The stronger the DNA repair ability of cells, the less the damaged plasmid DNA left in the cells at a certain time point and the stronger the fluorescence value of the reporter plasmids. As shown in Fig. 2e, knockout of the enhancer significantly impaired the DNA damage repair of the cells. The efficiency of DNA damage repair was more than 50% lower in the two 13q-Enh−/− clones compared with the wild-type Beas-2B cells. This clearly implicated that the 13q-Enh enhancer functioned as an important element for maintaining normal DNA repair capability.
We also tested the effects of deletion of 13q-Enh on cell apoptosis. The C5 and C23 clones and the wild-type Beas-2B cells were treated with H2O2 for 4 h, and subsequent flow cytometry confirmed attenuation of apoptosis in the 13q-Enh−/− clones. The proportion of apoptotic cells decreased from the corrected value of 5.42% in the control to 0.42% and 1.85% in the C5 and C23 clones, respectively (Fig. 2f, g), indicating an involvement of the 13q-Enh enhancer in the regulation of apoptosis.
The 13q-Enh physically interacted with the TNFRSF19 promoter to upregulate gene expression
Subsequently, we used chromosome conformation capture (3C) assays to precisely assess the regulation of the 13q-Enh enhancer on TNFRSF19 gene in wild-type Beas-2B cells and 13q-Enh−/− clones. Our 3C assays detected the ligation-dependent PCR products in the wild-type Beas-2B cells, but not in the 13q-Enh−/− clones under the same experimental conditions (Fig. 3b). DNA sequencing further confirmed that the PCR products were derived from the ligation of the 13q-Enh and the TNFRSF19 promoter (Additional file 1: Figure S5), indicating that the 13q-Enh specifically and physically linked with the TNFRSF19 promoter via chromatin looping.
Taken together, these data provided strong evidence that the 13q-Enh enhancer directly targeted TNFRSF19 gene expression by chromatin looping over a span of 180 kb.
Restoration of TNFRSF19 expression significantly suppressed the NNK-induced cell transformation and DNA damages in 13q-Enh−/− cells
We used rescue experiments to determine whether 13q-Enh executed its biological functions by targeting the TNFRSF19 gene. The 2 13q-Enh−/− clones treated by NNK for 15 generations (C5-NNK and C23-NNK) were transfected with the TNFRSF19-expressing lentivirus or the empty vector control, and the DNA repair efficiency and NNK-induced transformation were examined. As expected, restoring the TNFRSF19 expression significantly suppressed the anchorage-independent growth ability of NNK-treated 13q-Enh−/− clones (Fig. 3c), and the colony formation was reduced by 50% compared with the control (Fig. 3d). Furthermore, γ-H2AX assays confirmed that restoration of TNFRSF19 expression reduced NNK-induced DNA damage in the 13q-Enh−/−clones (Fig. 3e, f). These rescue experiments provided evidence that the 13q-Enh enhancer attenuated the risk of carcinogen-induced DNA damage and malignant transformation of bronchial epithelial cells by promoting TNFRSF19 expression.
The germline genetic variations at rs17336602, rs4770489, and rs34354770 significantly attenuated the enhancer activity by impairing its p53 response
The effect of these variations on p53 bindings was subsequently evaluated by electrophoretic mobility shift assay (EMSA) experiments using three wild-type sequences, namely ES1, ES2, and ES3, and the three paired mutated probes, namely MES1, MES2, and MES3, as probes (Fig. 5a). Each pair of the probes contained the potential p53 binding site and the single SNP site (wild type vs. mutant). Several bands were formed when the Beas-2B nuclear extract was incubated with each biotin-labeled wild-type probe. In EMSA using probe ES1, the first band was successfully completed by a 100-fold molar excess of the unlabeled cold probes, but not by cold SP1 consensus sequences, confirming the specific binding of the protein complex to the ES1 sequence (Fig. 5c, left panel, lanes 2–4). Furthermore, the first band was clearly weakened by pre-incubating the nuclear extract with an increasing amount of anti-p53 antibodies, indicating the p53 binding to the ES1 region (Fig. 5c, left panel, lanes 5–6). The genetic variation rs17336602 (G>C) affected the p53 binding, since the band was clearly weakened when Beas-2B nuclear extract was incubated with the biotin-labeled mutant MES1 probe containing this variation under the same experiment conditions (Fig. 5c, left panel, lane 9). The EMSA experiments using the other two pairs of probes, ES2/MES2 and ES3/MES3, also proved that p53 specifically bound to the ES2 and ES3 regions (Fig. 5c, the middle panel, lanes 2–8; right panel, lanes 2–8). The genetic variations rs4770489 (A>G) and rs34354770 (A>C) affected the p53 binding to these two regions as demonstrated by incubating Beas-2B nuclear extract with biotin-labeled MES2 and EMS3 mutant probes containing these two variations, respectively (Fig. 5c, middle panel, lane 9; right panel, lane 9). Our ChIP and EMSA experimental data strongly suggested that the three germline genetic variations impacted the regulatory function of the enhancer by affecting p53 binding to the 13q-Enh enhancer.
The germline genetic variations at rs17336602, rs4770489, and rs34354770 significantly reduced TNFRSF19 expression in vivo
The impacts of deregulated TNFRSF19 expression in vivo and in vitro
We also evaluated the relationship between the TNFRSF19 expression levels and the TNM stages in the 117 lung cancer patients. The TNFRSF19 expression levels were inversely correlated with the tumor staging, with significantly high expression in the tumors at stage I and low expression in the tumors at stages II/III/V (Fig. 7d). More specifically, tumors at the T1 stage expressed significantly higher levels of TNFRSF19 than those at the T2/3/4 stages (Fig. 7e). The same was true for lymph node metastasis (Fig. 7f). However, the follow-up time of these 117 patients was not sufficient to determine a relationship between the survival time and the TNFRSF19 expression, so we turned to a Kaplan-Meier plotter database for the information. The survival analysis of 1145 lung cancer patients for more than 200 months showed the survival time was significantly longer for patients with high expression of TNFRSF19 than with low expression  (p = 1.8e−09, log-rank test; Fig. 7g). The observations that the TNFRSF19 expression was inversely correlated with tumor staging and positively correlated with patient survival time strongly suggest TNFRSF19 functions as a lung cancer suppressor.
We further experimentally tested the tumor suppressor function of TNFRSF19 by transfecting A549 lung cancer cells that express a low level of this protein with TNFRSF19-expressing lentivirus vectors. As expected, the introduction of TNFRSF19 into A549 lung cancer cells significantly suppressed malignant phenotypes of the cells, including colony formation and invasive ability, which is consistent with the clinic observations (Fig. 7h, i) and supported the notion that TNFRSF19 functions as a lung cancer suppressor.
Our present study systematically explored the biological significance and underlying molecular mechanism behind the 13q12.12 locus that is highly associated with lung cancer risk in the Chinese population. We characterized a novel p53-responsive enhancer with normal lung tissue cell specificity in a 49-kb high linkage disequilibrium block of rs753955. It suppressed carcinogen-induced DNA damage and malignant transformation via directly regulating the TNFRSF19 gene over long distance by chromatin looping. The inherited variations at rs17336602 (G>C), rs4770489 (A>G), and rs34354770 (A>C) that were highly linked with the tag-SNP rs753955 (A>G) significantly weakened the enhancer activity by impairing its p53 responsiveness, empowering the potential lung cancer suppressive TNFRSF19 to decrease its eQTL gene expression. In the clinical, the rs17336602-C, rs4770489-G, and rs34354770-C impact lung cancer progression.
One challenge in revealing causative mechanisms of disease risk SNPs in genomic non-coding regions is to connect a risk allele to a target gene. In this study, we provided strong evidence that the TNFRSF19 was a key target gene of the 13q-Enh p53-responsive enhancer and that the inherited variations at rs17336602 (G>C), rs4770489 (A>G), and rs34354770 (A>C) in 13q12.12 significantly weakened the p53-dependent activity of the enhancer. Therefore, it is reasonable to infer that these SNPs can attenuate the 13q-Enh enhancer-mediated TNFRSF19 activation. The inference is further supported by our eQTL analysis with 117 Chinese NSCLC samples. The eQTL data showed a significant association between these SNPs and the TNFRSF19 gene expression, strongly suggesting the involvement of these SNPs in the regulation of TNFRSF19 in vivo. Further test of this inference by introducing the mutant haplotype in the endogenous locus using CRISPR/Cas9 technology will provide more direct evidence for the inference.
The TNFRSF19 protein (also known as Troy) encoded by the TNFRSF19 gene is a member of the TNF-receptor superfamily. The function of TNFRSF19 has not been well understood. Previous reports have suggested complex and pleiotropic roles of this protein in various cellular contexts [36, 37, 38]. TNFRSF19 has been reported to have a function of an oncogene in nasopharyngeal carcinoma and colorectal cancers cells [39, 40], but little is known about the biological roles of the TNFRSF19 in lung tissues. The evidence provided by our present study clearly points to TNFRSF19 as a lung cancer suppressor. First, our NNK-induced transformation experiments demonstrated that homologous deletion of the 13q-Enh enhancer resulted in a dramatic decrease in TNFRSF19 expression and increase in double-strand DNA breaks and malignant transformation, while restoration of the TNFRSF19 expression significantly reversed these phenotypes. Second, the introduction of TNFRSF19 into A549 lung cancer cells dramatically suppressed malignant phenotypes of the cells, including soft agar colony formation and invasive ability. Third, TNFRSF19 expression levels were significantly reduced in lung cancer tissues when compared with the para-cancer tissues. Fourth, the TNFRSF19 expression levels were inversely correlated with the tumor staging and survival time, as revealed by our clinical sample analysis of 117 lung cancer patients and Kaplan-Meier analysis of 1145 lung cancer cases. Furthermore, the TNFRSF19 expression patterns in diverse normal human tissues of 262 adult individuals revealed exclusively high expressions of TNFRSF19 in the lung tissue cells and skin cells (Additional file 1: Figure S7), which strongly implies that this protein is indispensable for the normal phenotypes of human lung tissues after birth . In view of the potential clinical significance of TNFRSF19, the lung cancer suppressor function of TNFRSF19 and the underlying mechanisms are worthy of further investigation in vitro and in vivo. The related study is underway in our laboratory.
The tumor suppressor function of TNFRSF19 observed in the present study is not consistent with the previously published papers [39, 40]. However, considering the high tissue-specific expression pattern of TNFRSF19 in adult tissues, it is not difficult to understand such inconsistency. Probably, whether TNFRSF19 functions as an oncogene or a tumor suppressor may depend on when and where it is expressed.
Although the TNFRSF19 was proved to be a critical target gene of the 13q-Enh and there existed no other potential target gene within the 2-Mbp window of the risk SNP rs753955, the present study has not yet ruled out a possibility that 13q-Enh regulates genes other than TNFRSF19 completely. Further analysis using 4C-seq  and TNFRSF19 knockout cells to test this possibility is required and helpful.
The p53 tumor suppressor pathway plays a central role in tumor suppression, and the p53 gene is the most frequently mutated gene in human cancer. Alterations in the p53 pathway also attenuate the function of wild-type (WT) p53 in tumors . Our present study suggests a novel route for attenuation of p53 function, namely, by enhancer mutations that disturb p53-dependent enhancer activity. We confirmed that the 13q-Enh harbored six known p53 motifs situated between or near the three causal SNPs and one atypical p53 binding sequence overlapping with the SNP rs4770489. ChIP assays proved that p53 bound to these sequences in vivo. Importantly, the p53 bindings could be significantly affected by the three SNPs as confirmed by our EMSA experiments. Interestingly, in addition to binding to the four sequences with typical p53 motifs, p53 specifically bound to S3 where there does not appear to be a p53 motif (Fig. 5a, b). This may suggest that the lack of a typical p53 binding motif does not rule out the possibility of p53 binding. There might be other unknown motifs that are beneficial to p53 specific bindings. Consistent with the mutant effects of the three causal SNPs on the p53 bindings, the enhancer displayed allele-dependent differential responses to p53, with a significantly lower response of the mutant C-G-C allele compared with the WT allele. This differential response was strengthened when cells were exposed to NNK. The p53-dependent regulation of the enhancer on the target gene was further confirmed by analyzing endogenous gene response. We show that the endogenous TNFRSF19 significantly responds to NNK treatment in wild-type Beas-2B cells, but not in the clones with deletion of the enhancer. These data support the notion that the three inherited variations are involved in the regulatory role of the 13q-Enh by affecting the p53 binding.
In addition to affecting p53 binding, the possibility that the three causal genetic variants affect other TF binding could not be ruled out. Bioinformatics analysis shows that each of these SNPs overlaps with one or more other predicted binding sites of transcriptional factors as usually observed in the cases of genetic variation sites. Whether or not these variants actually affect the other transcription factor binding remains to be explored. Even so, the mutational effects of the three variants on the p53 binding to the 13q-Enh enhancer are significant without a doubt.
Melo et al. have consistently described p53 binding regions located distantly from any known p53 target gene . Many of these p53-bound enhancer regions show enhancer activity and interact intrachromosomally with multiple neighboring genes to convey long-distance p53-dependent transcription regulation. Thus, p53-responsive enhancers may be distributed widely in the genome, lending credence to investigations that explore the extent to which germline genetic variations impair p53-dependent enhancer functions and consequently confer risk of lung cancer and other human cancers. Investigations of this type of genetic variations are especially meaningful for tumors with wild-type p53.
The observation that allele-dependent differential response to p53 is strengthened by NNK indicates that the causal effects of the risk SNPs could be amplified under a given environmental stress. The three tightly linked causal SNPs described in this study are common, as they all occur with the frequency of 0.41 in the Asian population. Therefore, their contributions to lung cancer risk could be significant in individuals exposed to ambient carcinogens or related environmental stresses. The dynamic penetrance of the causal effects of specific germline genetic variations could also partly explain why cancer-risk SNPs can maintain high frequencies in populations.
At any given GWAS locus, multiple SNPs are often found in LD with the GWAS risk SNP. Fine mapping of the location of the causal SNPs and determination of their causal effects, both as individuals or in synergies, are required to identify the causal mechanisms. In the present study, we mapped three causal risk SNPs within the 13q-Enh enhancer that were in high LD with the GWAS lung cancer risk SNP rs753955. Luciferase reporter gene assays showed that the impact of single variation was relatively moderate compared with that of all three variations together, although the mutant alleles including the single variation had a significantly lower activity when compared with the wild-type allele (Additional file 1: Figure S8). The impact on the enhancer activity was distinctively strengthened when all three variations were included, indicating the three highly linked variations acted cooperatively. Furthermore, the mutational effect of the three combined variations on enhancer activity was significantly enhanced in response to p53 or in the case of NNK exposure compared to the control.
Our present study reveals the biological significance of 13q12.12 lung cancer risk locus. We provide evidence that the three inherited variations at rs17336602 (G>C), rs4770489 (A>G), and rs34354770 (A>C) in 13q12.12 contribute to the lung cancer risk and development by declining the p53-responsive enhancer-mediated TNFRSF19 activation. Our study gives a novel insight into the understanding of the inherited mechanism of lung cancer. It also implies the 13q-Enh and TNFRSF19 as potential biomarkers for lung cancer risk screening and clinical prognosis.
Materials and methods
The Beas-2B human bronchial epithelial cell line was kindly supplied by Professor Chaojun Li (Nanjing University, Nanjing, China). The MRC-5 and HFL1 human fetal lung fibroblast cell lines were kindly supplied by Professor Luo Gu (Nanjing Medical University, Nanjing, China) and Wen Ning (Nankai University, China), respectively. The A549 and H1299 human non-small cell lung cancer cell lines were kindly supplied by Professor Lin Xu (Jiangsu Cancer Hospital, Nanjing, China). HeLa human cervical cancer cell line was supplied by professor Wei De (Nanjing Medical University, Nanjing, China). The PANC-1 human pancreatic cancer cell line, HEK293 human embryonic kidney cell line, MCF-7 human breast cancer cell line, and MCF-10A human breast epithelial cell line were purchased from the American Type Culture Collection (Manassas, VA, USA). The cell lines used in this study were authenticated by STR profiling.
A total of 117 human NSCLC tissues and their adjacent non-cancerous tissues were collected for DNA and RNA isolation from patients with NSCLC from Tumor Hospital of Yunnan province from January 2014 to June 2016. None of the patients had undergone radiotherapy, chemotherapy, or other anticancer treatment before the surgery. The histological features of all specimens were evaluated by pathologists according to the standard criteria, and the clinicopathological characteristics of 117 NSCLC patients are listed in Additional file 2: Table S1. We also obtained mRNA data for NSCLC tissues from The Cancer Genome Atlas (TCGA) on July 8, 2014. The normalized expectation-maximization (RSEM) read counts were available for 107 paired samples (tumors with adjacent normal tissues). The paired sample t test was used to examine the differences in gene expression between the tumors and adjacent normal tissues. Expression quantitative trait loci (eQTL) analysis was first performed in the Chinese samples with a linear regression model, and the results were further replicated using data from the Genotype-Tissue Expression Project (GTEx v7).
Construction of plasmids
The sequence of candidate enhancers designated as 13q-Enh was cloned into the pGL3-promoter vector (Promega Corporation, Madison, WI, USA) between the 5′-MluI-XhoI-3′ restriction sites upstream of the SV40 promoter. The mutant enhancer 13q-Enh was generated at the rs17336602, rs4770489, and rs34354770 sites by site-directed mutagenesis. The wild-type enhancer allele was G-A-A, while the mutant enhancer allele was C-G-C.
The TP53 and TNFRSF19 CDS sequences were cloned into pcDNA3.0 and p-EGFP-C1 vector, respectively, to generate ectopic overexpression plasmids. All primer sequences are listed in Additional file 2: Table S2. All plasmids were confirmed by sequencing (BGI and Invitrogen).
RNA extraction and quantitative real-time PCR assay
Total RNA was extracted using Trizol reagent (Invitrogen) according to the manufacturer’s protocol. Five hundred nanograms RNA was reverse-transcribed to prepare cDNA according to Roche manufacturer’s instructions. Quantitative real-time PCR was carried out using the SYBR Green for the detection of PCR products: denaturation, 95 °C for 10 min, followed by 40 cycles of denaturation at 95 °C 15 s, annealing at 60 °C for 15 s, and extension at 72 °C for 15 s. The mRNA level of TNFRSF19 and MIPEP was normalized to β-actin. All the primers are listed in Additional file 2: Table S2.
Beas-2B cells were transfected with siRNA using Lipofectamine 3000 (Invitrogen, USA) to knockdown p53, and the cells were incubated for 48 h before harvesting for luciferase reporter assays. The siRNA sequences are listed in Additional file 2: Table S3.
Luciferase reporter assay
Two days after transfection, luciferase assays were performed according to the manufacturer’s instructions (Dual-Luciferase System, Promega), and independent triplicate experiments were run for each plasmid. The pGL3-promoter vector was used as a negative control.
Chromatin immunoprecipitation assay
ChIP assay was conducted according to the manufacturer (Upstate) of ChIP assay kit. Detailed experimental procedures are described in our previous article .
Electrophoretic mobility shift assays
EMSA were performed essentially as previously describe . Nuclear proteins were extracted from Beas-2B cells using NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (78833, Thermo Fisher Scientific), and then measured using TaKaRa BCA Protein Assay Kit (T9300A, TaKaRa) according to the manufacturer’s instructions. Double-stranded oligonucleotides with or without 5′ biotin-labeled were synthesized (Shanghai Generay Biotech Company, Shanghai, China). The sequences of probes are listed in Additional file 2: Table S3. Electrophoretic mobility shift assay (EMSA) was performed with LightShift™ Chemiluminescent EMSA Kit (20148, Thermo Fisher Scientific) and the anti-p53 antibody (ab1101, Abcam). Briefly, nuclear proteins were pre-incubated with unlabeled probe or anti-p53 antibody in a binding mixture for 10–20 min at room temperature, then incubated with labeled wt-probe or mt-probe for 20 min. The mixtures were electrophoresed in 6% non-denatured polyacrylamide gel, transferred to a nylon membrane (INYC00010, Millipore), and detected biotin-labeled DNA by chemiluminescence.
CRISPR/Cas-9-mediated genome editing
The 13q-Enh enhancer was deleted by co-transfection of sgRNA plasmids and the Cas9 overexpression plasmid into Beas-2B cells. After confirming the deletion efficiency using primers flanking the 13q-Enh region, the cells were sorted into individual 96-well plates with the Aria II cell sorter (BD Biosciences) and subsequently expanded for further analyses.
The 13q-Enh+/+ mixture clone and 13q-Enh−/− clones were exposed to NNK (DMSO as solvent control) at 100 μg/ml (450 μM) continuously for 30 days to induce malignant transformation. Cells were treated with 10 μM NNK for 24 h or 48 h and harvested for luciferase assay and mRNA expression detection.
Soft agar assays
For soft agar assays, 2 × 103 cells were seeded in 2 ml 0.2%. The agarose stock was diluted with 2× DMEM; medium was overlaid on a 0.25% agarose base in 6-well culture plates. Colonies were stained after 22–28 days with 5% MTT at 37 °C for 4 h and visually counted.
Cells cultured on chamber slides were fixed with 4.0% paraformaldehyde (PFA) at room temperature for 15 min. Slides were blocked with 2% BSA in PBST (PBS + 0.25% Triton X-100), incubated with antibodies (γH2AX, 1:200) overnight at 4 °C, and then incubated with secondary antibody (1:500) at room temperature for 1 h. After three washes with PBST, the slides were incubated with Hoechst (Invitrogen) at room temperature for 20 min. Images were obtained with a FV1000 confocal microscope (Olympus, Center Valley, PA). The fluorescence intensity was determined using ImageJ software.
The HCR assay was used to measure the DNA repair capacity . The pGL3-promoter luciferase vector was exposed to H2O2 (v/v) at room temperature for 1 h. After stimulation, the damaged plasmids were purified by ethanol precipitation. The luciferase assay was then used to measure the DNA repair capacity of 13q-Enh+/+ and 13q-Enh−/− cells after transfection. The stronger the DNA repair ability of cells, the less the damaged plasmid DNA left in the cells at a certain time point and the stronger the fluorescence value of the report plasmids.
Cell apoptosis analysis
Collected cells were stained with 5 μl Annexin V-FITC and 5 μl PI for 15 min at room temperature, and apoptosis was detected by flow cytometry. All the procedures were conducted according to the manufacturer’s instructions (KeyGenBioTech).
Chromosome conformation capture assay
The 3C protocol was described previously . Briefly, in a total of 106 wild-type or 13q-Enh−/− Beas-2B cells were harvested and crosslinked in formaldehyde at room temperature. After quenching with glycine, the cells were lysed for nucleus isolation. The isolated nuclei were then digested with ASEI at 37 °C. After inactivation of the enzyme, the samples were diluted in 700 μl of ligation reaction system with 100 U T4 ligase and incubated at 16 °C overnight. The ligated chromatin was digested with proteinase K and purified by phenol-chloroform extraction. The interactions between TNFRSF19 and 13q-Enh were detected by specific primers and confirmed by DNA sequencing.
Construction of lentivirus vector for overexpression of TNFRSF19
The expression vector encoding full-length open reading frame of human TNFRSF19 was constructed by synthesis and PCR amplification. Briefly, the synthetic oligonucleotides were spliced into the complete sequence by PCR and validated by sequencing. Then, it was sub-cloned into the Plvx-EF1A-puro expression vector and validated by sequencing. Plvx-EF1A-puro-TNFRSF19, pSPSV-2 packing vector, and pMD2G envelope vector were transferred to 293T cells. After 72 h transfection, the virus supernatant was filtered and stored at − 80 °C for using.
Western blot analysis
Protein extracts were boiled in SDS loading buffer and then subjected to 8% SDS-polyacrylamide gel electrophoresis and transferred to PVDF membranes (Bio-Rad). Membranes were blocked in 5% milk-TBST (Tris-buffered saline Tween) for 1 h and then incubated overnight with mouse p53 antibody (Santa Cruz Biotechnology) and mouse actin/GAPDH antibody (Bioworld Technology). The membranes were then washed with TBST and incubated with the appropriate secondary antibody. After washing with TBST, the membranes were developed with an ECL detection system.
Transwell invasion assays
Matrigel (BD) was diluted with serum-free RPMI 1640 to a final concentration of 3 mg/mL and polymerized in Transwell inserts at 37 °C for at least 4 h. 7 × 104 cells were seeded onto the Matrigel in 10% FBS medium while the bottom chambers contain 500 μl of 20% FBS medium. Cells were allowed to invade for 48 h at 37 °C in 5% CO2. Three independent experiments were performed, and at least ten random fields were counted per experiment.
Each experiment was performed at least in triplicate. Results are shown as mean value ± standard deviation (SD). Statistical analysis was performed using unpaired Student’s t test. A p value less than 0.05 was considered statistically significant.
We thank professor Wei Gao and professor Xingyin Liu for their kind help in the manuscript revision.
This study was supported by grants from the National Natural Science Foundation of China (Grant No. 81874045; Grant No.81572789,) and the National Natural Science Foundation of China and Yunnan Joint Foundation (No.1502222).
Availability of data and materials
The data of histone modifications H3K4me1 (GSM733649), H3K4me3 (GSM733723), and H3K27ac (GSM733646) in NHLF can be found in the UCSC database (http://genome.ucsc.edu/). The data of H3K27ac in lung and non-lung tissues are available at the ENCODE database (https://www.encodeproject.org/), including the data from tissues of lung (ENCFF566ZDJ), pancreas (GSM1013129), kidney (GSM1112806), breast (GSE100978) , spleen (GSM906398), adrenal gland (GSM1013126), small intestine (GSM1127172), heart (GSE101345) , esophagus (GSM906393), liver (GSM1127173), ovary (GSM956009), and stomach (ENCFF299PTM). The data of p53 binding sites were downloaded from online Genomatix software (http://www.genomatix.de/). The data of TNFRSF19 expression in lung tissues and tumors are available at TCGA (http://www.cbioportal.org/) [30, 31], Oncomine (https://www.oncomine.org/, GSE32867) , and GEPIA (http://gepia.cancer-pku.cn/) . The data of TNFRSF19 expression pattern in normal human tissues are available at BioGPS (http://biogps.org/, 223827_at) . Patient’s survival time data was downloaded from Kaplan-Meier plotter (http://kmplot.com/analysis/) . Lung tissue eQTL data are available at the GTEx datasets (gtexportal.org/home/).
YJS and HBS conceived the original concept and led the entire team during the course of this study. YJS, LPS, and XLZ wrote the paper and analyzed all the data involved in this study. YJS, LPS, XLZ, and YY designed most of the experiments. XS provided human lung cancer tissue samples and technical supports for DNA and RNA isolation from these tissue samples and organized the clinic data analyses. RLL, ZL, and QFF were involved in the isolation of DNA and RNA from the tissue samples and clinical data analysis. GFJ was involved in the bioinformatics analysis of the study. LPS, XLZ, and XCW performed and analyzed the luciferase experiments. XLZ performed the ChIP assays for confirming histone modifications of the 13q-Enh and the p53 binding. BS and JYW were responsible for the deletion of 13q-Enh using CRISPR/Cas9 technique. YY performed the chromosome conformation capture in detecting the interaction between 13q-Enh and TNFRSF19 promoter. DWY performed the experiments involved in the construction of 13q-Enh knockout cell lines. XLZ and YC were involved in the phenotype analysis of 13q-Enh-deleted Beas-2B cells. YC performed the functional analysis of TNFRSF19. LPS and YZ performed the real-time qPCR assay. LS and NY were involved in the γ-H2AX assays and EMSA experiments. ZBH and XH were involved in the experimental data analysis. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The study has been approved by the Ethics Committee of Nanjing Medical University. Written informed patient consent was obtained from all patients. All experimental methods abided by the Helsinki Declaration.
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 25.Corradin O, Saiakhova A, Akhtar-Zaidi B, Myeroff L, Willis J, Cowper-Sal lari R, Lupien M, Markowitz S, Scacheri PC. Combinatorial effects of multiple enhancer variants in linkage disequilibrium dictate levels of gene expression to confer susceptibility to common traits. Genome Res. 2014;24:1–13.CrossRefGoogle Scholar
- 32.Selamat SA, Chung BS, Girard L, Zhang W, Zhang Y, Campan M, Siegmund KD, Koss MN, Hagen JA, Lam WL, et al. Genome-scale analysis of DNA methylation in lung adenocarcinoma and integration with mRNA expression. Genome Res. 2012;22:1197–211 https://genome.cshlp.org/content/22/7/1197.full?sid=8c58f9ca-177e-44d5-b462-e1e74e1345f3.CrossRefGoogle Scholar
- 33.Tang Z, Li C, Kang B, Gao G, Li C, Zhang Z. GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses. Nucleic Acids Res. 2017; https://academic.oup.com/nar/article/45/W1/W98/3605636.
- 35.Gyorffy B, Surowiak P, Budczies J, Lanczky A. Online survival analysis software to assess the prognostic value of biomarkers using transcriptomic data in non-small-cell lung cancer. PLoS One. 2013;8:e82241 https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0082241.CrossRefGoogle Scholar
- 44.Gong F, Sun L, Wang Z, Shi J, Li W, Wang S, Han X, Sun Y. The BCL2 gene is regulated by a special AT-rich sequence binding protein 1-mediated long range chromosomal interaction between the promoter and the distal element located within the 3′-UTR. Nucleic Acids Res. 2011;39:4640–52.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.