Association between lncRNA‐H19 polymorphisms and hepatoblastoma risk in an ethic Chinese population

H19 polymorphisms are associated with increased susceptibility to several cancers; however, their role in hepatoblastoma remains unclear. In this study, we investigated the association between three H19 polymorphisms (rs2839698 G>A, rs3024270 C>G, rs217727 G>A) and hepatoblastoma susceptibility in 213 hepatoblastoma patients. The rs2839698 and rs3024270 polymorphisms were associated with significantly increased hepatoblastoma risk, with the GG genotype associated with a higher risk of hepatoblastoma than the CC genotype at the rs3024270 locus. The rs217727 polymorphism was associated with significantly decreased hepatoblastoma risk, with the AG genotype associated with a lower risk of hepatoblastoma than the GG genotype. These findings were confirmed by combined analysis, and stratification analysis revealed that age, gender and clinical stage were associated with increased hepatoblastoma susceptibility. The GGG and AGG haplotypes were significantly associated with increased hepatoblastoma risk compared with the GCA reference (rs2839698, rs3024270, rs217727). The rs2839698 and rs3024270 polymorphisms correlated with decreased MRPL23‐AS1 expression, whereas the rs217727 polymorphism was associated with increased MRPL23‐AS1 expression. Overall, the H19 rs2839698, rs3024270 and rs217727 polymorphisms were associated with hepatoblastoma susceptibility in a Chinese Han population.


| INTRODUC TI ON
Hepatoblastoma is the most common hepatic malignancy in children, occurring in around 1/1 000 000 individuals and accounting for over 90% of primary hepatic malignancies in children <5 years old. 1 Hepatoblastoma usually presented with a large abdominal mass and increased alpha-fetoprotein (AFP) value. 2 Most hepatoblastomas are sporadic, but some are related to genetic aberrations and had a family history of cancers. 3 The aetiology and pathogenesis of hepatoblastoma remain largely unknown. However, increasing evidence suggested that chromosomal abnormalities, genetic aberrations and various adverse factors during pregnancy may contribute to the development of hepatoblastoma. 4 The management of hepatoblastoma is multidisciplinary, consisting of surgery, chemotherapy and liver transplantation, and the 5-year overall survival rate is approximately 80%. 5 Nonetheless, the treatment of high-risk hepatoblastomas remains challenging.
Hepatoblastoma is associated with several genetic syndromes, such as Edwards syndrome, Beckwith-Wiedemann syndrome (BWS) and familial adenomatous polyposis syndrome. 4,6 Evidence suggests that children with BWS have a much higher risk of developing hepatoblastoma than those without. 6 The abnormal methylation of tumour-specific genes in differentially methylated regions (DMRs), such as 11p15.5 and 20q13.3, has been observed prior to hepatoblastoma development in patients with BWS. 6,7 In addition, the DNA hypermethylation of IGF2 biallelic expression has been observed in hepatoblastoma. 8 Indeed, abnormal DNA methylation in imprinted DMRs is thought to be a key mechanism in the malignant transformation of progenitor cells in various tissues, including the liver. 9 Additionally, previous studies have shown that the classical Wnt signalling pathway, which is involved in somatic and germline mutations in various genes, is commonly deregulated in hepatoblastoma. 10 H19 is a 3.0 kb gene located on human chromosome 11, and its transcript is a highly conserved long non-coding RNA (lncRNA), known as lncRNA-H19. 11 H19 plays an important role in tumorigenesis via transcriptional regulation, post-transcriptional regulation and epigenetic regulation. Moreover, H19 is associated with the occurrence of various malignant tumours, such as breast cancer, bladder cancer and gastric cancer. 12 H19 has been shown to promote cell cycle progression in breast cancer, with its deletion arresting breast cancer cells in the pre-S-phase of the cell cycle. 13 Moreover, knockdown of H19 inhibits the growth of hepatocellular carcinoma (HCC) and gastric cancer cells under hypoxia recovery conditions, 14 whereas H19 overexpression partially suppresses p53 activation in gastric cancer cells. 15 In addition, Zhu et al 16 found that silencing H19 can inhibit the proliferation of ovarian cancer SKOV3 cells and is related to the regulation of certain cell cycle-and apoptosis-related proteins, thus exerts an inhibitory effect on ovarian cancer cell growth. A study by Zhang et al found that high H19 expression positively correlates with the lymph node metastasis of cancer cells and tumour size, and can be used as an independent prognostic factor for the overall survival of non-small-cell lung cancer (NSCLC).
Furthermore, the proto-oncogene c-myc can directly regulate H19, with H19 gene silencing significantly inhibiting the proliferation of NSCLC cells. 17 Taken together, these studies indicate that H19 is closely related to the occurrence and development of tumours; however, its role may differ in different types of cancer. For instance, H19 can exhibit proto-oncogene activities that promote tumour development and growth, 18 but can act as a tumour suppressor gene in other tumours by inhibiting tumour proliferation, metastasis and invasion. 19,20 Recent evidence has shown that polymorphic sites in the H19 gene are associated with susceptibility to a variety of malignancies. 21 Lin et al 22 found that rs217727 C>T in H19 may be associated with an increased risk of breast cancer, whereas Xia et al 23 found no significant association between rs3741219 polymorphisms and breast cancer susceptibility in a Chinese population. Others have found that the H19 rs3741219 polymorphism is associated with increased HCC risk 24 and that the rs2839698 polymorphism is associated with an increased risk of various gastrointestinal cancers, with this association being even more prominent in the Asian population. 25 Despite increasing evidence suggesting that H19 polymorphisms are associated with increased susceptibility to many cancers, the association between H19 polymorphisms and hepatoblastoma susceptibility has not yet been investigated. Therefore, we investigated the association between three H19 polymorphisms (rs2839698 G>A, rs3024270 C>G, rs217727 G>A) and hepatoblastoma susceptibility, using the odds ratio and 95% confidence interval to determine the strength of the association.

| Study population
We enrolled 213 patients with hepatoblastoma and 958 cancer-free controls in this study. [26][27][28] The initial diagnosis of hepatoblastoma was made based upon ultrasound detected liver mass and elevated AFP value. Then, tumour tissue was obtained by image-guided liver biopsy. All cases of hepatoblastoma were diagnosed by pathological examination, and no direct blood relationship was found among any of the cases. Patients older than 18 years or with HCC were excluded. Healthy controls were recruited in hospitals. All healthy controls had no history of malignancy and were matched to the hepatoblastoma cases by age (±5 years), gender, ethnicity and geographical region (as shown in Table S1). This study was approved by the Institutional Review Board of Guangzhou Women and Children's Medical Center (approval number 2017120101), with written informed consent provided by the parents or legal guardians of the participants.

| SNP selection and genotyping
We selected H19 gene SNPs of potential functional interest from the dbSNP database (http://www.ncbi.nlm.nih.gov/) and SNPinfo (http://snpin fo.niehs.nih.gov/) for analysis (rs2839698 G>A, rs3024270 C>G and rs217727 G>A). Total genomic DNA was isolated from the peripheral blood leucocytes of each patient using a TIANamp Blood DNA Kit (TianGen Biotech Co., Ltd.). H19 gene SNPs were genotyped by TaqMan real-time PCR, [26][27][28] with eight blank wells in each 384-well plate containing water as negative controls.
We also randomly repeated the genotyping of 10% of the samples, achieving a 100% concordance rate.

| Statistical analysis
To test for HWE in the controls, we used the observed genotype frequencies. A two-sided chi-squared test was used to compare the allele frequencies and demographic variables of two groups.
Associations between genotypes and hepatoblastoma risk were evaluated using the OR and 95% CI calculated by logistic regression analysis. Statistical analyses were performed using SAS software version 9.4 (SAS Institute). P-values of <.05 were considered significant.

| Associations between H19 polymorphisms and hepatoblastoma susceptibility
In this study, we genotyped the frequencies of three H19 singlenucleotide polymorphisms (SNPs) in 213 patients with hepatoblastoma and 957 controls, as shown in Table 1

| Stratified analysis of H19 polymorphisms and hepatoblastoma risk
To further analyse the effects of the rs2839698, rs3024270 and rs217727 genotypes on hepatoblastoma susceptibility, we conducted stratification analyses based on age (<17 months, ≥17 months), gender and clinical stage (I + II, III + IV). As shown in

| Haplotype analysis
We also investigated the effects of H19 haplotypes (

TA B L E 3 Frequency of inferred H19 haplotypes based on their observed genotypes and their association with hepatoblastoma susceptibility
b Obtained using logistic regression models adjusted for age and gender.

| D ISCUSS I ON
LncRNAs play regulatory roles in a variety of biological processes, including tumorigenesis. 29 Emerging evidence has suggested that lncRNA SNPs can affect the function and expression of cancer-associated genes, leading to increased cancer susceptibility. 30 In this case-control study of 213 patients with hepatoblastoma and 958 healthy controls, we found that rs2839698, rs3024270 and rs217727 H19 polymorphisms are significantly associated with hepatoblastoma susceptibility.
To our knowledge, this is the first study to discover that H19 polymorphisms are associated with hepatoblastoma susceptibility.
A variety of chromosomal abnormalities have been identified in patients with hepatoblastoma, such as the amplification or loss of DNA copies in chromosomes 5 and 11p15.5. 31 Moreover, the F I G U R E 1 The rs2839698, rs3024270 and rs217727 genotypes correlated with MRPL23-AS1 expression in whole blood (A, C and E) and transformed fibroblast cells (B, D and F) (data from GTEx Portal) inactivation of the APC gene, located on chromosome 5, has been observed in 67%-89% of sporadic hepatoblastoma cases, while chromosome 11 aberrations may play an important role in hepatoblastoma pathogenesis in patients with BWS. 32 Furthermore, Rumbajan et al 3 found that abnormal DMR methylation occurs before hepatoblastoma development, suggesting that DMR methylation is related to hepatoblastoma occurrence. In this study, we found that lncRNA H19 polymorphisms were significantly associated with hepatoblastoma risk.
H19 is an imprinted oncofetal gene that is located close to the IGFII gene locus, 33 and its role in cancer has been widely studied.
A meta-analysis study showed that the rs2735971 A>G, rs2839698 C>T and rs3024270 G>C H19 polymorphisms were associated with increased overall cancer risk. 34 In addition, a population-based

CO N FLI C T S O F I NTE R E S T
All authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.