LINC00673 rs11655237 C>T confers neuroblastoma susceptibility in Chinese population

Neuroblastoma, which accounts for approximately 10% of all pediatric cancer-related deaths, has become a therapeutic challenge and global burden attributed to poor outcomes and mortality rates of its high-risk form. Previous genome-wide association studies (GWASs) identified the LINC00673 rs11655237 C>T polymorphism to be associated with the susceptibility of several malignant tumors. However, the association between this polymorphism and neuroblastoma susceptibility is not clear. We genotyped LINC00673 rs11655237 C>T in 393 neuroblastoma patients in comparison with 812 age-, gender-, and ethnicity-matched healthy controls. We found a significant association between the LINC00673 rs11655237 C>T polymorphism and neuroblastoma risk (TT compared with CC: adjusted odds ratio (OR) =1.80, 95% confidence interval (CI) =1.06–3.06, P=0.029; TT/CT compared with CC: adjusted OR =1.31, 95% CI =1.02–1.67, P=0.033; and T compared with C: adjusted OR =1.29, 95% CI =1.06–1.58, P=0.013). Furthermore, stratified analysis indicated that the rs11655237 T allele carriers were associated with increased neuroblastoma risk for patients with tumor originating from the adrenal gland (adjusted OR =1.51, 95% CI =1.06–2.14, P=0.021) and International Neuroblastoma Staging System (INSS) stage IV disease (adjusted OR =1.60, 95% CI =1.12–2.30, P=0.011). In conclusion, we verified that the LINC00673 rs11655237 C>T polymorphism might be associated with neuroblastoma susceptibility. Prospective studies with a large sample size and different ethnicities are needed to validate our findings.


Introduction
Neuroblastoma, originating from the developing sympathetic nervous system, is the most common extracranial tumor of infancy and childhood. The median age at diagnosis is approximately 18 months [1]. It is a fatal solid cancer accounting for approximately 8-10% of all childhood malignancies [1-3] and contributes to approximately 10-15% of all cancer-related deaths in children [4]. Despite intensive multimodal treatments [5], 5-year event-free survival for high-risk (approximately 50% of all cases) neuroblastoma patients remains less than 50% [6,7], and the 5-year survival rate for all neuroblastoma patients is approximately 50% [8]. Due to the rarity of neuroblastoma, strict associations are difficult to certify, and no particular environmental exposure has been involved in the development of this disease [1]. Thus, neuroblastoma remains a huge therapeutic challenge and global burden.
Previous genome-wide association studies (GWASs) of pancreatic ductal adenocarcinoma have identified a long intergenic non-coding RNA (lincRNA) LINC00673 rs11655237 C>T (also reported as G>A elsewhere) polymorphism at 17q25.1 region to be significantly associated with the risk of pancreatic cancer [9]. It was the most significant polymorphism in the LINC00673 [9]. Further studies have proven that the rs11655237 in exon 4 of LINC00673 forms a target site for miR-1231 binding, which weakens the tumor suppressing function of LINC00673 in an allele-specific manner and thus endows pancreatic cancer with susceptibility [10]. Moreover, LINC00673 has been identified to be associated with non-small-cell lung cancer (NSCLC) [11][12][13], gastric cancer [14], tongue squamous cell carcinoma [15], and breast cancer [16] consecutively. Together, these findings indicated that the LINC00673 rs11655237 appears to be implicated in the development of broad-spectrum tumors and also involved in cell homeostasis maintenance. However, the association between neuroblastoma susceptibility and LINC00673 rs11655237, is not yet reported.
To comprehensively evaluate and provide new insights into the impact of LINC00673 rs11655237 C>T polymorphism on neuroblastoma susceptibility, we performed a hospital-based case-control study using data from a Chinese Han population composed of 393 neuroblastoma cases and 812 age-, gender-, ethnicity-matched cancer-free controls.

Study subjects
The subjects included were described in our previous studies [17][18][19][20]. A total of 393 histopathological cases of primary neuroblastoma and 812 cancer-free controls from Guangdong province (South China) and Henan province (North China) were recruited in the current study (Supplementary Table S1). Briefly, all the 393 cases were newly confirmed and histopathologically diagnosed as neuroblastoma patients without progressive disease and previous treatments before the collection of clinical classification. The cases were genetically unrelated Chinese Han children who received treatments at the Guangzhou Women and Children's Medical Center between February 2010 and March 2017, and at the First Affiliated Hospital of Zhengzhou University between August 2011 and April 2017. Age-, gender-, and ethnicity-matched controls were randomly recruited from children undergoing routine medical examination at the same center during the same period. The parents or guardians of the children provided written informed consent for the children's participation in the present study. The current study was approved by the Institutional Review Board of both centers and performed in accordance with the study protocol.

Genotyping
DNA samples were extracted as previously described [21]. Briefly, DNA samples were diluted to a stock concentration of 10 ng/μl and added to 96-well plates. Genotyping for the LINC00673 rs11655237 C>T polymorphism was carried out in a 384-well plate using TaqMan Real-Time PCR method as described elsewhere [22][23][24]. For quality control and accuracy of genotyping results, approximately 10% of the samples were randomly selected and re-genotyped. The results were 100% concordant.

Statistical analysis
Comparison of the differences in the demographic and genotypic information between neuroblastoma cases and controls was performed by chi-square test. Hardy-Weinberg equilibrium for controls was analyzed by goodness-of-fit chi-square test. To estimate the strength of association between rs11655237 C>T polymorphism and neuroblastoma risk, unconditional univariate and multivariate logistic regression analyses were conducted and adjusted for age and gender, while odds ratios (ORs) and 95% confidence intervals (CIs) were used. Further stratified analysis was performed by age, gender, the International Neuroblastoma Staging System (INSS) clinical stages [25], and tumor sites. P-values <0.05 were considered as statistically significant. All statistical analyses were two-sided and were calculated by SAS software (version 9.1; SAS Institute, Cary, NC, U.S.A.).

Demographic characteristics
The demographic characteristics of the included participants are summarized in Supplementary Table S1. No statistically significant differences were observed between neuroblastoma cases and controls regarding age (P=0.

LINC00673 rs11655237 C>T polymorphism and neuroblastoma susceptibility
The LINC00673 rs11655237 genotype frequencies and their association with the susceptibility of neuroblastoma are presented in Table 1. A total of 391 of the included neuroblastoma cases and all controls (812) were successfully genotyped. We found that the rs11655237 T allele carriers (CT/TT compared with CC: adjusted OR = 1.31, 95% CI = 1.02-1.67, P=0.033; and T compared with C: adjusted OR = 1.29, 95% CI = 1.06-1.58, P=0.013), especially mutated-type homozygote carriers (TT compared with CC: adjusted OR = 1.80, 95% CI = 1.06-3.06, P=0.029), were significantly associated with an increased risk of neuroblastoma when compared with wild-type homozygote carriers.

Stratified analysis
Next, the included participants were stratified by age, gender, tumor sites, and INSS stages. We further estimated the effects of the variant genotype of the rs11655237 C>T polymorphism on the neuroblastoma risk amongst the different strata (Table 2). We found that the rs11655237 CT/TT genotype carriers were significantly associated with an increased risk of a tumor originating in the adrenal gland (adjusted OR = 1.51, 95% CI = 1.06-2.14, P=0.021). Furthermore, we also found that the carriers of the CT/TT genotypes had a significantly increased risk of INSS clinical stage IV disease (adjusted OR = 1.60, 95% CI = 1.12-2.30, P=0.011) when compared with the carriers of the CC genotype.

Discussion
In this hospital-based case-control study, we evaluated the association of the GWAS-identified LINC00673 rs11655237 C>T polymorphism with neuroblastoma susceptibility in 393 patients and 812 cancer-free controls. Our results revealed that the rs11655237 T allele significantly increased the risk of neuroblastoma. In addition, stratified analysis showed that the rs11655237 C>T polymorphism increased the risk of adrenal gland and clinical stage IV neuroblastoma. The results from the present study suggested that the rs11655237 T allele was positively associated with neuroblastoma and could confer neuroblastoma susceptibility.
Described as the largest subclass in the non-coding transcriptome in humans, lincRNAs are non-coding transcripts longer than 200 nts. In the past decade, the roles of lincRNAs in human disorders, including tumors, have attracted a lot of attention [26,27]. lincRNAs are known to play vital regulatory roles in various processes [28], including but not limited to imprinting (H19 [29] and KCNQ1OT1 [30]), metastasis (MALAT1 [31], HOTAIR [32], and COLDAIR [33]), X inactivation (XIST [34]), deregulation of the tumor suppressors [35], and pseudogene pairing (PTENP1 [36]). Therefore, aberrant expression and polymorphisms within lincRNAs have been associated with susceptibility to a range of human diseases, including cancers. Cancer-associated lincRNAs, which demonstrate developmental and tissue-specific expression properties along with aberrant regulation in various malignancies, may indicate new approaches for the diagnosis and treatment of cancer. Systematic identification of the expression patterns and characterization of lincRNAs together with their associated proteins might contribute to the development of lincRNA-targetted therapies for tumors and various human diseases.
LINC00673 is located at 17q24.3, which is a chromosome region recently documented to have a high frequency of loss of heterozygosity [37], and is associated with pancreatic cancer susceptibility in individuals of European ancestry [9]. Further studies have indicated that the LINC00673 rs11655237 variant, a germline C>T transition, can lead to a decrease in the level of LINC00673 in cells, which may trigger SRC-ERK oncogenic signaling, while attenuating STAT1-dependent anti-oncogenic signaling. Consistent with these functional discoveries, researchers have observed significantly decreased pancreatic ductal adenocarcinoma susceptibility in the rs11655237 C allele carriers when compared with the rs116552337 T allele carriers [10,38]. Intriguingly, LINC00673 has been proven to be associated with susceptibility, progression, and outcome of other malignancies as either a tumor suppressor or promoter. Shi et al. [39] found that up-regulation of LINC00673 can promote tumor proliferation through LSD1 interaction and repression of Neurocalcin δ in NSCLC, which implied LINC00673 as an oncogene in NSCLC. For the first time, Ma et al. [40] showed that LINC00673 promoted NSCLC metastasis by binding with EZH2 that resulted in epigenetic silencing of HOXA5. The latter is a tumor suppressor gene that inhibits NSCLC cell metastasis via regulating cytoskeletal remodeling. Lu et al. [11] found a similar situation in which LINC00673 regulated NSCLC proliferation, invasion, migration, and even epithelial-mesenchymal transition by sponging miR-150-5p. A study by Abdul-Rahman et al. [16] showed that LINC00673 levels were modulated by hormone signaling and inversely associated with breast cancer survival. Huang et al. [14] reported that LINC00673 was significantly up-regulated in gastric cancer. LINC00673 overexpression induced cell proliferation and invasion and inhibited cell apoptosis. Research conducted by Yu et al. [15] showed that LINC00673 was highly expressed in a significant proportion of human tongue squamous cell carcinoma samples and associated with poor prognosis.
Taken together, these findings shed light on an important functional interaction between LINC00673 and malignancies. Therefore, it would be interesting to examine the role of LINC00673 in neuroblastoma. With such purpose, we conducted the current study in the Chinese Han population. The current study has several important strengths. To the best of our knowledge, this is the first investigation to validate the association of neuroblastoma susceptibility with GWAS-identified polymorphism within LINC00673. The identification of LINC00673 as an antitumor factor in neuroblastoma was based on discoveries in GWASs in several tumors. Although our findings showed that the association between rs11655237 and neuroblastoma susceptibility were obtained in Chinese Han populations and are consistent with previous reports in other malignancies, further studies on this and other tumors in a different ethnic group would be beneficial. In addition, it has been suggested that rs11655237 might have a potential role in the regulation of gene expression via an enhancer, promoter, or silencer mechanism [9]. Therefore, further exploration of this issue should be addressed. Additionally, it would be intriguing to validate the effects of rs11655237 variation on the development of neuroblastoma using in vivo experimental settings.
Several limitations should be addressed in the present study. First, although it was the first study on the association between LINC00673 and neuroblastoma, we included only 393 neuroblastoma cases and 812 cancer-free controls. The relative small sample size may limit the statistical power. Replication studies from other centers with a larger sample size are encouraged to confirm the association. Second, only one polymorphism of the newly identified LINC00673 was investigated in the current study. More polymorphisms, especially the potentially functional polymorphisms not yet contained in previous GWASs, remain to be explored. Third, due to the design of the retrospective study, information bias and selection bias might be unavoidable. Because of the lack of information on dietary intake, living environmental factors, and parental exposures, only frequency matching of neuroblastoma cases and controls by age and gender could be used to reduce these biases. Fourth, in vitro and in vivo experiments should be performed to interrogate mechanism(s) underlying the association in the future. Finally, although recruited from two centers, including residents in Southern and Northern China, participants were all Han Chinese, so the results should be cautiously extrapolated to other ethnic groups.
In conclusion, our results verified that the LINC00673 rs11655237 C>T polymorphism was significantly associated with neuroblastoma susceptibility in the Chinese Han population, especially for children with neuroblastoma of the adrenal gland region and clinical stage IV. In the future, well-designed prospective studies with larger sample size including different ethnic populations, detailed information (including living environment, dietary intake, and parental exposures) and functional studies should be performed to strengthen our findings.