Genome-wide association studies (GWAS) have implicated the 1q22 gastric cancer risk locus in disease, but little is known about its underlying oncogenic functions. This study represents a systematic investigation of the biological significance and potential mechanism associated with the gastric cancer risk of SNP rs2075570(C>T) in 1q22. We identified two functional germline variations (rs2049805-C and rs2974931-G) in an active enhancer in a 64.8 kb high-linkage disequilibrium block of rs2075570. The enhancer upregulated ubiquitin associated protein 2 like (UBAP2L) gene expression over a 960 kb distance by chromatin looping. Gastric cancer tissues expressed significantly higher levels of UBAP2L than was observed in the matched noncancerous tissues, and the UBAP2L expression was negatively correlated with patient survival. Downregulation of UBAP2L inhibited the proliferation and invasion of human gastric cancer cells in vitro and in a xenograft mouse model. Notably, the two mutant variations significantly enforced the enhancer activity and UBAP2L expression. In conclusion, this study revealed two causal variations in the 1q22 region using tag-SNP rs2075570 as a genetic marker. These variations may affect the occurrence and progression of gastric cancer by reinforcing the expression of the 1q22-Enh enhancer-regulated UBAP2L target gene.

Implications:

Our study provides an important clue of how noncoding germline variations contribute to gastric cancer, which gives a novel insight into understanding the genetic mechanism of gastric cancer.

Gastric cancer remains the fifth most frequently diagnosed cancer and the third leading cause of cancer-related death (1). Genetic factors play an important role in gastric carcinogenesis and development, as indicated by the variations in disease predisposition and progression in different ethnic backgrounds and by data from genome-wide association studies (GWAS; refs. 2, 3). However, little is known about the mechanism of inheritance of gastric cancer, although genomic alterations discovered in gastric cancer have provided valuable clues regarding the molecular pathogenesis and genetic susceptibility associated with gastric cancer.

From a genetic perspective, complex human diseases like gastric cancer are mainly driven by noncoding genetic variants (4). In particular, enhancers account for the vast majority of regulatory elements in the genome and play a central role in the regulation of transcription (5, 6). Enhancers generally execute their regulatory functions from a distance through interactions that involve the formation of a chromatin loop with the promoter regions of their target genes (7). The most common genetic variations in humans are the SNPs. GWAS indicate that SNPs are generally correlated with enhancer elements that are marked with high enrichment in H3K4me1 and H3K27ac and low enrichment of H3K4me3 (8–10).

Studies on various cancers, including lung, prostate, breast, and colon cancer, have shown that the genetic variations contained in an enhancer can influence the risk of cancer and its development by altering the regulatory function of that enhancer (11, 12). In recent years, dozens of SNPs have been associated with gastric cancer risk (13, 14). Among them, the SNP rs2075570(C>T), located at the noncoding region of chromosome 1q22, has been viewed as one of the most important SNP biomarkers for gastric cancer, and the T allele is associated with an increased risk of diffuse gastric cancer (13, 15–17). However, the biological function of this 1q22 gastric cancer risk region is unclear.

In this study, we systematically investigated the biological function of the 1q22 gastric cancer risk region by integrating bioinformatics, laboratory experiments, animal modeling, and clinical analyses. Our data indicated that two inherited variations, rs2049805 (T>C) and rs2974931 (C>G), are highly linked with the risk SNP rs2075570 and contribute to the gastric cancer risk by enhancing the 1q22-Enh enhancer-mediated upregulation of the UBAP2L target gene. This work reveals a biological function for the 1q22 gastric cancer risk region and provides a useful experimental model for investigating the inheritance mechanisms of gastric cancer.

Expression quantitative trait loci (eQTL) analysis

The eQTL of all genes within the 2 Mbp window of the risk SNP rs2075570 were calculated directly on the platform of the GTEx eQTL datasets (RRID:SCR_001618, https://www.gtexportal.org/home/testyourown; ref. 18). This revealed the relationship between the associated SNPs in the LD region of rs2075570 and the UBAP2L target gene.

Cell culture

The AGS (RRID: CVCL_0139) and HGC-27 (RRID: CVCL_1279) gastric cancer cell lines were kindly supplied by Professor Yun Chen (Nanjing Medical University) in November 2019, and were confirmed by short tandem repeat (STR) analysis in December 2019. The SNU668 cell line (RRID: CVCL_5081) was purchased from Nanjing Cobioer Biotechnology Co., Ltd., and was confirmed by STR analysis in May 2020. The 293T cell line (RRID: CVCL_0063) was kindly supplied by Professor Wei Gao (Nanjing Medical University) in July 2018, and was confirmed by STR analysis in August 2018. The AGS, HGC-27, and SNU668 cells were cultured in RPMI1640 medium (Gibco), the 293T cells were cultured in DMEM medium (Gibco) containing 10% FBS (Gibco) and 100 U/mL penicillin/streptomycin (Invitrogen). All cells were incubated at 37°C in an environment of 5% CO2. All cells were tested for mycoplasma to exclude mycoplasma contamination after resuscitation using GMyc-PCR Mycoplasma Test Kit (Yeasen). The culture time between thawing and use in the described experiments was less than 20 passages.

UBAP2L expression analysis

The UBAP2L mRNA expression data of 408 gastric cancer tissues and 36 normal gastric tissues were downloaded from The Cancer Genome Atlas (TCGA) database by cBioPortal (cBioPortal, RRID: SCR_014555, http://www.cbioportal.org/). ANOVA was performed to identify the expression differences in the genes associated with rs2075570. A P value less than 0.01 and a value of log2FC [log2FC = log2(expression in cancer tumor) − log2(expression in normal tissues)] greater than 1 or less than −1 were considered statistically significant. The analysis was verified on the GEPIA website (Gene Expression Profiling Interactive Analysis, RRID: SCR_018294, http://gepia.cancer-pku.cn/).

Patient tissue specimens

A total of 44 gastric cancer tissue samples and matched noncancerous gastric mucosa epithelial tissues were collected from the First Affiliated Hospital of Nanjing Medical University. Written informed consent was obtained from all subjects, and the study approved by the Institutional Ethics Committee of the First Affiliated Hospital of Nanjing Medical University. The collected gastric tissue included the full-thickness structure of the mucosa and excluded the lamina propria tissues. Prior to surgery, no patient had undergone chemotherapy, radiotherapy, or any other adjuvant therapy. Subsequently, all patients were confirmed as having primary gastric cancer by the Department of Pathology. The clinicopathologic characteristics of the 44 patients are listed in Supplementary Table S1.

RNA extraction and qRT-PCR assay

Total RNA was extracted with RNAiso Plus reagent (Takara) following the manufacturer's protocol. The Transcriptor First Strand cDNA Synthesis Kit (Roche) was used to reverse-transcribe 500 ng total RNA to cDNA. The level of UBAP2L was determined by qRT-PCR using SYBR Green (Roche), following the manufacturer's protocol. The mRNA level of UBAP2L was normalized to β-actin. The primers were purchased from Shanghai Generay Biotech Company, and the sequences were listed in Supplementary Table S2.

Plasmid construction and lentivirus production

The short hairpin RNA targeting human UBAP2L (sh-U) and nontarget shRNA (scrambled) were designed and synthesis by RiboBio Co., Ltd. The sequences of shRNA were listed in Supplementary Table S3. Both shRNAs were ligated into the pLKO.1-puro vector (RRID: Addgene_8453). The shRNA vector, together with lentiviral packaging plasmid psPAX2 (RRID: Addgene_12260) and envelope expressing plasmid pMD2.G (RRID: Addgene_12259), were subsequently transfected into 293T cells using PEI (Invitrogen), according to the manufacturer's instructions. Supernatants containing lentiviruses were harvested at 72 hours after transfection and then purified by passage through 0.45 μm filters. The collected lentiviruses were named as sh-U and SCR, respectively.

The UBAP2L CDS sequence (NM_014847.4) was cloned into pcDNA3.1 expression vector (V790–20, Invitrogen) between the 5′-Hind III-Xho I-3′ restriction sites to construct UBAP2L expression plasmid named as ex-U.

The putative enhancer region (Chr1:155224910–155226126, NCBI, hg38) containing the rs2049805 and rs2974931 sites was confirmed on the basis of the H3K4me1, H3K4me3, and H3K27ac ChIP-seq data of normal gastric tissues from the ENCODE database (Encode, RRID: SCR_015482). The region was amplified from the genomic DNA of patients. The obtained sequence was inserted into the pGL3-promoter vector (E1761, Promega) between the 5′-Kpn I-Xho I-3′ restriction sites upstream of the SV40 promoter. The variant enhancer allele contained the variants rs2049805 (T>C) and rs2974931(C>G). We obtained the mutant enhancer allele directly, and the wild type was obtained by site-directed mutagenesis. All plasmids were confirmed by sequencing. The primer sequences were listed in Supplementary Tables S4 and S5.

Cell infection and transfection

The AGS, HGC-27, or SNU668 cells were infected with polybrene (Santa Cruz Biotechnology) and lentivirus-containing medium at a multiplicity of infection (MOI) of 10. The efficiency of infection was confirmed by Western blotting after 72 hours. The infected cells were passaged and cultured for no longer than 1 month.

The UBAP2L expression plasmid ex-U or empty pcDNA3.1 plasmid were transiently transfected into sh-U infected cells using Lipofectamine 3000 (Invitrogen). The efficiency of transfection was confirmed by Western blotting after 48 hours.

Western blotting

UBAP2L protein expression was assessed by Western blotting, and the samples were normalized to GAPDH. Protein concentrations were routinely measured by BCA Protein Assay Kit (Takara). Protein samples (20 μg) were boiled with SDS loading buffer, loaded onto denatured SDS-PAGE gels, separated by electrophoresis, and transferred to PVDF membranes (Millipore). Membranes were blocked in 5% milk-TBST (Tris-buffered saline plus Tween) for 1 hour, and incubated overnight at 4°C with primary antibodies for UBAP2L (1:5000, Abcam, Catalog No. ab70319, RRID: AB_1271381) or GAPDH (1:1000, Santa Cruz Biotechnology, Catalog No. sc-47724, RRID: AB_627678). The membranes were then washed three times in TBST and incubated with the relevant secondary antibody (1:10,000, Jackson ImmunoResearch Labs, Catalog No. 111–035–003, RRID: AB_2313567; Catalog No. 115–035–003, RRID: AB_10015289) at room temperature for 1 hour. After washing with TBST, the proteins were detected with the ECL detection system (Thermo Fisher Scientific), and gray scale analysis was performed by ImageJ (ImageJ, RRID:SCR_003070).

Transwell migration assay

Cells were collected and counted 24 hours after transient transfection, then diluted with serum-free RPMI1640 and inoculated into a blank transwell upper chamber at a density of 2 × 104 cells/well. The bottom chambers contained 500 μL of cell culture medium containing 10% FBS. Cells were cultured in a 5% CO2 incubator at 37°C for 24 to 48 hours. Subsequently, cells on the top side of the bottom were washed off, migrating cells were stained with crystal violet and counted under a microscope. Three independent experiments were performed, and at least three to five random fields were counted per experiment.

Transwell invasion assay

Matrigel (BD Biosciences) was diluted to 3 mg/mL with serum-free RPMI1640 and coated onto the membrane of a transwell chamber. Cells were collected and counted 24 hours after transient transfection, each sample of 5 × 104 cells in 200 μL of RPMI1640 containing 10% FBS was loaded onto the Matrigel, and 500 μL of culture medium containing 20% FBS was loaded into the lower chamber. The cells were incubated for 24 to 48 hours. Next, Matrigel coat was cleaned off and migrating cells were stained with crystal violet and counted. Three independent experiments were performed, and at least three to five random fields were counted per experiment.

Clonogenic formation assay

The transiently transfected cells at 24 hours were collected and counted, inoculated into 6-well plates at a density of 300 cells/well (HGC-27, AGS) or 700 cells/well (SNU668) and cultured for a further 7 days (HGC-27, AGS) or 14 days (SNU668). The medium was then removed, and the colonies were washed twice with PBS and fixed with 4% paraformaldehyde for 30 minutes. The colonies were then stained with crystal violet for 15 minutes, and photographed with a digital camera (Sony). The experiment was repeated three times. The rate of colony formation was calculated as the colony number/inoculated cell number × 100%.

In vivo xenograft mouse model

Eight-week-old male and female BALB/c nude mice were used for the HGC-27 xenograft model. The study was approved by the Animal Ethics Committee of Nanjing Medical University. All mice were bred in specific-pathogen-free (SPF) condition in the Animal Center of Nanjing Medical University. Animal studies were carried out accordance to the NIH Guide for the Care and Use of Laboratory Animals. The mice were divided randomly into an experimental and a control group (five males and five females per group). A sample of 1 × 107 HGC-27 SCR or sh-U cells was subcutaneously injected into the armpit of the right forelimb of each mouse. The resulting tumors were measured every third day, and the formula V = a2 × b/2 was used to calculate the tumor volumes, where a is the shortest diameter and b is perpendicular to a. Mice were euthanized 6 weeks after inoculation, and the tumors were excised, weighed, and subjected to pathologic examination by hematoxylin and eosin (H&E) staining.

Luciferase reporter assays

AGS, HGC-27, or SNU668 cells were seeded in 24-well plates (2.5 × 105/well). Reporter plasmids and the pRL-SV40 luciferase control vector (E2231, Promega) were cotransfected into the three cell lines using Lipofectamine 3000 (Invitrogen), according to manufacturer's protocol. The pGL3-promoter vector was used as a negative control. Cells were collected for luciferase assays 48 hours after transfection, and luciferase activity was assayed with the Dual-Luciferase Reporter Assay System (E1910, Promega). The luminescent signals were normalized to negative controls. The experiment was repeated in triplicate.

Clone fragments containing the germline genetic variations at rs2075570, rs2049805, and rs2974931

Genomic DNA was extracted from the AGS, HGC-27, or SNU668 cell lines using a Multisource Genomic DNA Miniprep Kit (Axygen). PCR with specific primers was performed to clone two fragments. One fragment included rs2075570, and the other included rs2049805 and rs2974931. The alleles were confirmed by DNA sequencing. The primer sequences were listed in Supplementary Table S6.

Chromosome conformation capture assay

Chromosome conformation capture (3C) was performed as described previously (19). Briefly, 1 × 106 AGS, HGC-27, or SNU668 cells were crosslinked in formaldehyde for 10 minutes at room temperature, followed by quenching with glycine and lysing with NP-40 (Sigma) for nucleus isolation. The isolated nuclei were digested with Hind III (TAKARA, 500 units, 22 hours, 37°C), and then the samples were ligated by T4 DNA ligase (TAKARA, 2100 units, 17 hours, 16°C). The DNA fragments were digested with proteinase K and cleaned by phenol–chloroform extraction. PCR with specific primers was then performed to detect the interactions between the three gastric cell lines and 1q22-Enh. Gene loops were confirmed by DNA sequencing. All the primers were listed in Supplementary Table S7.

Statistical analysis

All experiments were performed in triplicate. Results are presented as mean value ± SD. The data for UBAP2L expression in 44 paired tumor and para-tumor tissues were calculated using a paired Student t test. All other results were analyzed using an unpaired Student t test. P < 0.05 was considered statistically significant.

Availability of data

The data for UBAP2L expression in normal gastric tissues and gastric tumors were downloaded from TCGA (http://www.cbioportal.org/; ref. 20). Patient survival time data were downloaded from GEPIA (http://gepia.cancer-pku.cn/; ref. 21), and Kaplan–Meier plotter (http://kmplot.com/analysis/; ref. 22). The data for the histone modifications H3K4me1 (ENCSR903QBX), H3K4me3 (ENCSR357ROS), and H3K27ac (ENCSR133NBJ) in gastric tissues were downloaded from the ENCODE database (https://www.encodeproject.org/experiments/ENCSR539HNK/; ref. 23). The LD information and linked SNP can be found in the Broad Institute database (http://pubs.broadinstitute.org/mammals/haploreg/haploreg.php).

Identification of the target UBAP2L gene associated with risk SNP rs2075570

We used the GTEx database for gastric tissues to perform an eQTL analysis to identify the target gene of the risk SNP rs2075570. Among all the genes within the 2 Mbp window of the risk SNP, 5 genes were involved with rs2075570: three protein genes (GBA, THBS3, and UBAP2L), one RNA gene (RUSC1-AS1), and one pseudogene (GBAP1; Fig. 1A). The expression of UBAP2L, RUSC1-AS1, and THBS3 was suppressed by the C allele of rs2075570, whereas GBA expression was enhanced (Fig. 1B; Supplementary Fig. S1A).

We then sought the key gene that participates in the progression of gastric cancer. We used the TCGA database to compare the mRNA expression levels of the 5 genes in human gastric cancer tissues versus normal gastric tissues. We found that only UBAP2L mRNA expression levels showed a statistically significant increase in gastric cancer tissues (P = 1.97 × 10E-12, log2FC = 1.067; Fig. 1C; Supplementary Fig. S1B). The data from the GEPIA (21) and Kmplot (22) databases further indicated shorter survival in patients with high UBAP2L expression than with low expression (Fig. 1E; Supplementary Fig. S1C). The mRNA expression level of RUSC1-AS1 was very close to the log2FC cutoff (P = 8.87 × 10−8, log2FC = 0.974; Supplementary Fig. S1B), but no correlation was found between RUSC1-AS1 and overall survival (OS) in patients with gastric cancer (GEPIA database, Supplementary Fig. S1D). So we were focused on UBAP2L gene in this study.

We then used qRT-PCR to measure UBAP2L mRNA expression in tumor tissues from 44 patients with gastric cancer and matching adjacent noncancerous tissues. Comparison with the noncancerous tissues revealed a significant upregulation of UBAP2L mRNA expression in tumor tissues (P = 0.0281; Fig. 1D). These findings suggested that UBAP2L may act as the target gene of rs2075570 and that increased expression of UBAP2L could be involved in the susceptibility to gastric cancer and its progression.

Downregulation of UBAP2L inhibited malignant phenotypes of gastric cancer cells in vitro and in vivo

We explored the biological significance of UBAP2L in gastric cancer cells by lentivirus-mediated knockdown of UBAP2L and rescue experiments in the AGS, HGC-27, and SNU668 gastric cancer cell lines. The sh-U group was the cells infected with sh-U to knockdown UBAP2L protein expression, SCR was the control of sh-U group. Rescue group was the sh-U cells transiently transfected with ex-U plasmids, and rescue-ctrl was sh-U cells transiently transfected with pcDNA3.1 plasmids. The efficiency of UBAlP2L expression at 48 hours was confirmed by Western blotting (Fig. 2A).

Clonogenic formation assays revealed that suppression of UBAP2L markedly repressed the proliferation of AGS, HGC-27, and SNU668 cells, whereas the suppressive effect on proliferation was significantly reversed in the rescue experiments (Fig. 2B). These results suggested that downregulation of UBAP2L significantly suppressed the proliferation of gastric cancer cells. We also used transwell assays to evaluate the influence of UBAP2L knockdown on in vitro gastric cancer cell migration and invasiveness. UBAP2L silencing suppressed the ability of cells to transfer to the lower chamber, whereas the invasion and migration abilities were restored by transfection with the ex-U plasmids (Fig. 3A and B). These results demonstrated that the depression of UBAP2L expression significantly decreased the motility and invasion of gastric cancer cells.

A nude mouse xenograft model was also constructed to verify the effect of UBAP2L silencing on in vivo tumor formation by gastric cancer cells. The average tumor volume in SCR group were markedly larger than in the sh-U group at the 45th day after inoculation (Fig. 4A and B), indicating that inhibition of UBAP2L significantly decreased the xenograft tumor growth. Furthermore, tumor cells of the sh-U group showed a high level of degenerative change indicated by eosinophilic cytoplasm, and most of tumor cells lacked nuclei and showed necrosis (Fig. 4C). Taken together, these results indicated that UBAP2L acted as a proto-oncogene that promoted the progression of gastric cancer.

Identification of the possible active enhancer within the 1q22 locus that supports communication between the risk SNP rs2075570 and the target UBAP2L gene

We investigated the underlying mechanism by which the risk SNP rs2075570 regulated the expression of the target UBAP2L gene first by identifying all the associated SNPs of the risk SNP. We used HaploView software to explore the linkage disequilibrium (LD) region of rs2075570, which spans 64.8 kb (r2 > 0.8, D′ > 0.9) and a total of 35 highly linked SNPs. We then used gastric CHIP-seq data from the ENCODE database to identify the putative active enhancers in the LD region, based on the criteria of specific histone marks, including high enrichment of H3K4me1 and H3K27ac and low enrichment of H3K4me3 (13–15). The GTEx database (v8) was used to perform eQTL analysis of every SNP individually to screen out potential alleles that may make contact with UBAP2L, and this yielded eight related SNPs. Finally, we overlapped the locations of the selected SNPs and enhancers, focusing on one putative enhancer element, 1q22-Enh. This 1q22-Enh element was identified within the 64.8 kb LD block of rs2075570 and contained two UBAP2L-associated genetic variations, rs2049805 (T>C) and rs2974931 (C>G; Fig. 5AD).

We confirmed the enhancer activity and explored the influence of these two genetic variations by inserting the wild type and variant 1q22-Enh sequences into pGL3-promoter plasmids. The enhancer activity and effects of the two genetic variations on the enhancer activity were examined using luciferase reporter gene assays. The results confirmed that 1q22-Enh was an active enhancer in three gastric cancer cell lines, AGS, SNU668, and HGC-27. Notably, luciferase activity of variant 1q22-Enh allele was significantly increased compared with the wild-type 1q22-Enh allele in AGS and SNU668 cells (Fig. 5E), indicating the genetic variations enhanced the 1q22-Enh activity. No significant difference was detected in the HGC-27 cells, suggesting influence of different cell context on the enhancer activity in addition to enhancer genetic variations.

The genotypes at rs2075570, rs2049805, and rs2974931 of the three cell lines were also determined by sequencing. The results showed that at rs2075570 AGS cell line was a risk allele homozygote (T/T), HGC-27 and SNU668 cell lines were heterozygotes (C/T). At rs2049805 and rs2974931, all three cell lines were risk allele homozygotes (C/C and G/G, respectively; Fig. 5F). Taken together, these results support the hypothesis that rs2049805 and rs2974931 are directly related to the risk of gastric cancer.

The 1q22-Enh sequence physically interacted with the UBAP2L promoter to upregulate gene expression

Enhancers generally execute their regulatory functions from a distance through interaction with the promoter regions of their target genes via chromatin looping. The physical interaction of 1q22-Enh with the UBAP2L promoter was confirmed by chromosome conformation capture (3C) assays in the AGS, HGC27, and SNU668 cells. For 3C analysis, formaldehyde was used to crosslink protein–DNA in intact nuclei. The crosslinked chromatin was then digested by a restriction enzyme, followed by ligation. If there is apposition between a remote regulatory sequence and a promoter, new hybrid fragments containing these two elements are generated; and carefully designed PCR reactions can be used to detect these new combined fragments. In our experiments, Hind III was used to digest nuclei to evaluate the interaction between 1q22-Enh and the UBAP2L promoter (Fig. 6A). 3C assays showed ligation-dependent PCR products in all three gastric cancer cell lines (line 2 in Fig. 6BD), and DNA sequencing further confirmed that the PCR products were derived from the ligation of the 1q22-Enh and the UBAP2L promoter region (Fig. 6EG). Negative control samples either from non-crosslinked chromatin or from cross-linked chromatin that was digested without subsequent ligation did not generate any PCR products (lines 3, 4, 5 in Fig. 6BD).

Taken together, these results supported the presence of interactions between 1q22-Enh and the active UBAP2L gene and that such interaction may be widespread in gastric cancer cells.

In this study, we explored the biological significance and underlying regulatory mechanism of the 1q22 locus, a locus that is highly associated with a diffuse gastric cancer risk in the Eastern Asian population. Our bioinformatics analysis and laboratory experiments demonstrate that UBAP2L is a target gene of the tag SNP rs2075570, and we confirm that UBAP2L acts as a proto-oncogene in the progression of gastric cancer. We identify an active enhancer, 1q22-Enh, which harbors two inherited variations (rs2049805 T>C, rs2974931 C>G) in a 64.8 kb high-linkage disequilibrium block of rs2075570. The enhancer regulates UBAP2L expression by interacting with its promoter through chromatin looping, and the two inherited variation mutations in the enhancer significantly enhance its activity.

The challenge in exploring the role of cancer disease-related risk SNPs in genomic noncoding regions is connecting the risk allele with its target gene. In this study, we used 3C technology to prove that the 1q22-Enh enhancer physically interacts with the promoter region of the UBAP2L gene, whereas the inherited variations at rs2049805 (T>C) and rs2974931 (C>G) significantly enhance the activity of the enhancer. Therefore, a reasonable inference is that these two inherited variations could enhance 1q22-Enh-mediated UBAP2L expression. This notion was supported by our eQTL analysis using the GTEx database, as we found a significant association between these SNPs and the level of UBAP2L gene expression. The tissue samples carrying the variants expressed significantly higher levels of UBAP2L when compared with the wild-type sample tissues. Further tests of the variant effect by introducing the variants haplotype into the endogenous genomics locus by CRISPR/Cas9-mediated single base replacement technology will provide more direct evidence for this inference in the future.

UBAP2L, as a member of the ubiquitin-related protein family, is a highly conserved protein with an ubiquitin-associated (UBA) domain and multiple RGG/RG repeat regions in its N-terminus (24). UBAP2L is a BMI-1 binding protein and has a reported involvement in the maintenance of the activity of hematopoietic stem cells and progenitor cells (25). The RGG/RG repeat region of UBAP2L suggests that it is essential for the correct kinetochore–microtubule attachment and for accurate distribution of chromosomes (26). UBAP2L is also an important stress granule (SG) nucleator, as it represents the most common pathway of SG nucleation (27). Research has confirmed a statistically higher expression of UBAP2L in various malignant tumors than in normal tissues (28–30), while silencing of the UBAP2L gene inhibits the growth and migration of several tumors, including colon, breast, and uterine cervical cancers (28–30). However, a role for UBAP2L in gastric cancer has not been elucidated.

Our results showed a significant upregulation of UBAP2L in 44 cases of gastric cancer, as well as a positive correlation between UBAP2L expression levels and poor prognosis in patients with gastric cancer. Downregulation and rescue experiments further supported the enhancing effects of UBAP2L on the carcinogenesis, proliferation, invasion, and migration of human gastric cancer cells. Also, in our zenograft mouse experiment the tumor size was drastically reduced, although shRNA knocked down about 50% of the gene in HGC-27 cells. All These results suggest that UBAP2L functions as a strong proto-oncogene in development and progress of gastric cancer. Further investigating the oncogenic role of the UBAP2L gene in gastric cancer will be significant for understanding the underlying pathology and related mechanisms. GWAS have successfully identified hundreds of loci associated with nearly all the common malignant diseases, and most of these variations are located in noncoding genome regions (31, 32). Further GWAS studies have confirmed the tissue-specific regulation of cancer risk by SNPs (19, 33, 34). For example, Zhang and colleagues (35) found that a prostate cancer risk SNP rs1859962 is located within a 130 kb LD region at 17q24.3. A prostate cancer-specific enhancer in this region inhibits the expression of SOX9, which is about 1 Mb away from the regulatory element. Currently more than a dozen non-coding SNPs associated with gastric cancer have been identified. The molecular mechanisms underlying their genetic variations could be complicated, since each GWAS SNP tags tens to hundreds of causal SNPs in its LD region. In this study, we combined the information from the GTEX, ENCODE, and TCGA databases to predict the target gene and causal SNPs, and we bridged them functionally though the active enhancer. We were able to confirm our assumptions through both in vivo and in vitro laboratory experiments, thereby providing a quick and dependable solution for the functional study of known risk SNPs.

Our findings do not rule out the possibility that the 1q22-Enh enhancer regulates genes other than UBAP2L, although UBAP2L was clearly an important target gene and no other potential target gene was identified within the 2 Mbp window of the risk SNP rs2075570. In future studies, we intend to combine 4C, eQTL, and the TCGA database to determine whether other target genes are under the control of the 1q22-Enh enhancer and are influenced by the two identified genetic variations.

The results presented here identify the biological significance of the 1q22 gastric cancer risk locus. We provide evidence that the two inherited variations at rs2049805 T>C and rs2974931 C>G in 1q22 contribute to the risk and development of gastric cancer by enhancing 1q22-Enh-mediated UBAP2L activation. Our research provides a useful experimental model for understanding the mechanism of inheritance of gastric cancer and reveals the potential for using UBAP2L as a biomarker for gastric cancer risk screening and clinical prognosis.

No disclosures were reported.

W. Guan: Resources, data curation, formal analysis, validation, investigation, visualization, writing–original draft. N. Yang: Resources, data curation, formal analysis, validation, investigation, visualization, writing–original draft. X. Zuo: Formal analysis, validation, investigation. X. Wang: Formal analysis, investigation, visualization. P. Cao: Investigation, methodology. Y. Chu: Software, formal analysis. Z. Qin: Investigation, methodology. H. Cheng: Investigation, methodology. X. Shi: Investigation. T. Ma: Investigation. Z. Xu: Resources, supervision, funding acquisition. Y. Sun: Conceptualization, supervision, funding acquisition, project administration, writing–review and editing.

This study was supported by grants from the National Natural Science Foundation of China (Grant Nos. 81572789, 81874045, 81871946). We thank Prof. Wei Gao and Prof. Yun Chen for their generous help in the experiments.

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.
Bray
F
,
Ferlay
J
,
Soerjomataram
I
,
Siegel
RL
,
Torre
LA
,
Jemal
A
. 
Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries
.
CA Cancer J Clin
2018
;
68
:
394
424
.
2.
Gallagher
MD
,
Chen-Plotkin
AS
. 
The post-GWAS era: from association to function
.
Am J Hum Genet
2018
;
102
:
717
30
.
3.
Gurdasani
D
,
Barroso
I
,
Zeggini
E
,
Sandhu
MS
. 
Genomics of disease risk in globally diverse populations
.
Nat Rev Genet
2019
;
20
:
520
35
.
4.
Zhang
F
,
Lupski
JR
. 
Non-coding genetic variants in human disease
.
Hum Mol Genet
2015
;
24
:
R102
10
.
5.
Karnuta
JM
,
Scacheri
PC
. 
Enhancers: bridging the gap between gene control and human disease
.
Hum Mol Genet
2018
;
27
:
R219
27
.
6.
Corradin
O
,
Scacheri
PC
. 
Enhancer variants: evaluating functions in common disease
.
Genome Med
2014
;
6
:
85
.
7.
Mora
A
,
Sandve
GK
,
Gabrielsen
OS
,
Eskeland
R
. 
In the loop: promoter-enhancer interactions and bioinformatics
.
Brief Bioinform
2016
;
17
:
980
95
.
8.
Ernst
J
,
Kheradpour
P
,
Mikkelsen
TS
,
Shoresh
N
,
Ward
LD
,
Epstein
CB
, et al
Mapping and analysis of chromatin state dynamics in nine human cell types
.
Nature
2011
;
473
:
43
9
.
9.
Calo
E
,
Wysocka
J
. 
Modification of enhancer chromatin: what, how, and why?
Mol Cell
2013
;
49
:
825
37
.
10.
Trynka
G
,
Sandor
C
,
Han
B
,
Xu
H
,
Stranger
BE
,
Liu
XS
, et al
Chromatin marks identify critical cell types for fine mapping complex trait variants
.
Nat Genet
2013
;
45
:
124
30
.
11.
Visel
A
,
Rubin
EM
,
Pennacchio
LA
. 
Genomic views of distant-acting enhancers
.
Nature
2009
;
461
:
199
205
.
12.
Kurdistani
SK
. 
Enhancer dysfunction: how the main regulators of gene expression contribute to cancer
.
Genome Biol
2012
;
13
:
156
.
13.
Mocellin
S
,
Verdi
D
,
Pooley
KA
,
Nitti
D
. 
Genetic variation and gastric cancer risk: a field synopsis and meta-analysis
.
Gut
2015
;
64
:
1209
19
.
14.
Yan
C
,
Zhu
M
,
Ding
Y
,
Yang
M
,
Wang
M
,
Li
G
, et al
Meta-analysis of genome-wide association studies and functional assays decipher susceptibility genes for gastric cancer in Chinese populations
.
Gut
2020
;
69
:
641
51
.
15.
Saeki
N
,
Saito
A
,
Choi
IJ
,
Matsuo
K
,
Ohnami
S
,
Totsuka
H
, et al
A functional single nucleotide polymorphism in mucin 1, at chromosome 1q22, determines susceptibility to diffuse-type gastric cancer
.
Gastroenterology
2011
;
140
:
892
902
.
16.
Sung
H
,
Yang
HH
,
Hu
N
,
Su
H
,
Taylor
PR
,
Hyland
PL
. 
Functional annotation of high-quality SNP biomarkers of gastric cancer susceptibility: the Yin Yang of PSCA rs2294008
.
Gut
2016
;
65
:
361
4
.
17.
Sung
H
,
Hu
N
,
Yang
HH
,
Giffen
CA
,
Zhu
B
,
Song
L
, et al
Association of high-evidence gastric cancer susceptibility loci and somatic gene expression levels with survival
.
Carcinogenesis
2017
;
38
:
1119
28
.
18.
eGTEx Project
. 
Enhancing GTEx by bridging the gaps between genotype, gene expression, and disease
.
Nat Genet
2017
;
49
:
1664
70
.
19.
Shao
L
,
Zuo
X
,
Yang
Y
,
Zhang
Y
,
Yang
N
,
Shen
B
, et al
The inherited variations of a p53-responsive enhancer in 13q12.12 confer lung cancer risk by attenuating TNFRSF19 expression
.
Genome Biol
2019
;
20
:
103
.
20.
Cancer Genome Atlas Research Network
. 
Comprehensive molecular characterization of gastric adenocarcinoma
.
Nature
2014
;
513
:
202
9
.
21.
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
;
45
:
W98
102
.
22.
Szász
AM
,
Lánczky
A
,
Nagy
Á
,
Förster
S
,
Hark
K
,
Green
JE
, et al
Cross-validation of survival associated biomarkers in gastric cancer using transcriptomic data of 1,065 patients
.
Oncotarget
2016
;
7
:
49322
33
.
23.
ENCODE Project Consortium
. 
An integrated encyclopedia of DNA elements in the human genome
.
Nature
2012
;
489
:
57
74
.
24.
Maeda
M
,
Hasegawa
H
,
Sugiyama
M
,
Hyodo
T
,
Ito
S
,
Chen
D
, et al
Arginine methylation of ubiquitin-associated protein 2-like is required for the accurate distribution of chromosomes
.
FASEB J
2016
;
30
:
312
23
.
25.
Bordeleau
ME
,
Aucagne
R
,
Chagraoui
J
,
Girard
S
,
Mayotte
N
,
Bonneil
E
, et al
UBAP2L is a novel BMI1-interacting protein essential for hematopoietic stem cell activity
.
Blood
2014
;
124
:
2362
9
.
26.
Hindorff
LA
,
Sethupathy
P
,
Junkins
HA
,
Ramos
EM
,
Mehta
JP
,
Collins
FS
, et al
Potential etiologic and functional implications of genome-wide association loci for human diseases and traits
.
Proc Natl Acad Sci U S A
2009
;
106
:
9362
7
.
27.
Huang
C
,
Chen
Y
,
Dai
H
,
Zhang
H
,
Xie
M
,
Zhang
H
, et al
UBAP2L arginine methylation by PRMT1 modulates stress granule assembly
.
Cell Death Differ
2020
;
27
:
227
41
.
28.
Aucagne
R
,
Girard
S
,
Mayotte
N
,
Lehnertz
B
,
Lopes-Paciencia
S
,
Gendron
P
, et al
UBAP2L is amplified in a large subset of human lung adenocarcinoma and is critical for epithelial lung cell identity and tumor metastasis
.
FASEB J
2017
;
31
:
5012
8
.
29.
He
J
,
Chen
Y
,
Cai
L
,
Li
Z
,
Guo
X
. 
UBAP2L silencing inhibits cell proliferation and G2/M phase transition in breast cancer
.
Breast Cancer
2018
;
25
:
224
32
.
30.
Yoshida
K
,
Kajiyama
H
,
Inami
E
,
Tamauchi
S
,
Ikeda
Y
,
Yoshikawa
N
, et al
Clinical significance of ubiquitin-associated protein 2-like in patients with uterine cervical cancer
.
In Vivo
2020
;
34
:
109
16
.
31.
di Iulio
J
,
Bartha
I
,
Wong
EHM
,
Yu
HC
,
Lavrenko
V
,
Yang
D
, et al
The human noncoding genome defined by genetic diversity
.
Nat Genet
2018
;
50
:
333
7
.
32.
Sud
A
,
Kinnersley
B
,
Houlston
RS
. 
Genome-wide association studies of cancer: current insights and future perspectives
.
Nat Rev Cancer
2017
;
17
:
692
704
.
33.
Gao
P
,
Xia
JH
,
Sipeky
C
,
Dong
XM
,
Zhang
Q
,
Yang
Y
, et al
Biology and clinical implications of the 19q13 aggressive prostate cancer susceptibility locus
.
Cell
2018
;
174
:
576
89
.
34.
Hua
JT
,
Ahmed
M
,
Guo
H
,
Zhang
Y
,
Chen
S
,
Soares
F
, et al
Risk SNP-mediated promoter-enhancer switching drives prostate cancer through lncRNA PCAT19
.
Cell
2018
;
174
:
564
75
.
35.
Zhang
X
,
Cowper-Sal lari
R
,
Bailey
SD
,
Moore
JH
,
Lupien
M
. 
Integrative functional genomics identifies an enhancer looping to the SOX9 gene disrupted by the 17q24.3 prostate cancer risk locus
.
Genome Res
2012
;
22
:
1437
46
.