LIN28A gene polymorphisms modify neuroblastoma susceptibility: A four‐centre case‐control study

Abstract Neuroblastoma ranks the most common seen solid tumour in childhood. Overexpression of LIN28A gene has been linked to the development of multiple human malignancies, but the relationship between LIN28A single nucleotide polymorphisms (SNPs) and neuroblastoma susceptibility is still under debate. Herein, we evaluated the correlation of four potentially functional LIN28A SNPs (rs3811464 G>A, rs3811463 T>C, rs34787247 G>A, and rs11247957 G>A) and neuroblastoma susceptibility in 505 neuroblastoma patients and 1070 controls from four independent hospitals in China. The correlation strengths were determined by using odds ratios (ORs) and corresponding 95% confidence intervals (CIs). Among these SNPs, rs34787247 G>A exhibited a significant association with increased susceptibility in neuroblastoma (GA vs GG: adjusted OR = 1.30, 95% CI = 1.03‐1.64; AA vs GG: adjusted OR = 2.51, 95% CI = 1.36‐4.64, AA/GA vs GG: adjusted OR = 1.42, 95% CI = 1.12‐1.80, AA vs GG/GA: adjusted OR = 2.39, 95% CI = 1.29‐4.42). Furthermore, the combined analysis of risk genotypes revealed that subjects carrying three risk genotypes (adjusted OR = 1.64, 95% CI = 1.02‐2.63) are more inclined to develop neuroblastoma than those without risk genotype, and so do carriers of 1‐4 risk genotypes (adjusted OR = 1.26, 95% CI = 1.01‐1.56). Stratification analysis further revealed risk effect of rs3811464 G>A, rs34787247 G>A and 1‐4 risk genotypes in some subgroups. Haplotype analysis of these four SNPs yields two haplotypes significantly correlated with increased neuroblastoma susceptibility. Overall, our finding indicated that LIN28A SNPs, especially rs34787247 G>A, may increase neuroblastoma risk.


| INTRODUC TI ON
Neuroblastoma is a solid tumour that predominantly affects infants and young children. 1 It mainly develops from neural crest progenitor cells. 2,3 Neuroblastoma constitutes about 8%-10% of all paediatric cancers, but disproportionally causes 12%-15% cancer death in children. 2,4,5 Neuroblastoma displays considerable clinical heterogeneity, ranging from spontaneous recovery to therapy-refractory progression. 6,7 The distinct difference in survival rate among subgroups was another reflection of such heterogeneity. 8,9 Patients with neuroblastoma can be classified into three risk groups: low risk, intermediate risk and high risk by using some clinical and biological prognostic factors. Patients with a non-high risk (low and intermediate risk) of neuroblastoma have a long-term survival rate up to of 90% or above, while those with high risk of neuroblastoma only achieve a survival rate as low as 40%. 10,11 In recent decades, remarkable advancement has been achieved in comprehending the fundamental aetiology of neuroblastoma. 12,13 Children's and pregnant women's exposures to many environmental factors were reported to predispose to neuroblastoma, yet the causality could not be finally confirmed. 14,15 On the other hand, genetic alterations have been shown to be linked to neuroblastoma susceptibility. Mutations in genes ALK 16,17 and PHOX2B 18 are frequently observed in hereditary neuroblastoma. Moreover, a number of single nucleotide polymorphisms (SNPs) in association with neuroblastoma predisposition have been identified in genes recently, including TP53, 19 LIN28B, 20 HACE1, 20 LMO1, 21 BARD1, 22 NEFL 23 and CDKN1B. 24 Moreover, a fine-mapping analysis of BARD1 locus (2q35) also identified two independent genome-wide neuroblastoma-associated loci. 25 However, the present identified genetic variations could not fully account for the carcinogenesis of neuroblastoma. We are still on the discovery journey of unveiling more causative genetic alterations hidden in the bush.
LIN28 is a conserved RNA-binding protein that plays a significant part in the regulation of cell proliferation, glucose metabolism and pluripotency through interacting with miRNAs. 26,27 The mammalian genome, LIN28 gene, encodes two Lin28 paralogs, Lin28A and Lin28B. 28 LIN28A inhibits the maturation process of let-7 microRNAs and thereby enhances the translation of let-7 target mRNAs. 29,30 Briefly, cytoplasmic LIN28A induces pre-let-7 oligo-uridylation through a TUTase-dependent mechanism. 31 Such poly-uridylation leads to pre-let-7 instability and eventually reduces the amount of mature let-7. 32 let-7 is a tumour suppressor. Its downregulation promotes tumorigenesis and correlates with poor prognosis. 33 By binding to a variety of mRNA targets, LIN28A also has additional functions except for suppressing let-7 maturation. 28,34 Over-activation of the LIN28A gene has been observed in various human cancers. 35,36 The mechanism of the LIN28A-mediated tumorigenesis has been extensively investigated. However, the implications of LIN28A gene SNPs in neuroblastoma risk remain undiscovered. To determine the relationship between LIN28A gene SNPs and neuroblastoma susceptibility, we performed this multi-centre epidemiological study.

| Study subjects
The current case-control study included 505 cases and 1070 healthy non-cancer controls, as noted previously (Table S1). 37 Cases were newly diagnosed and histologically confirmed with neuroblastoma.
Frequency-matched controls on age and sex were recruited from the same residing area as cases. Without no exception, every participant provided his/her necessary written informed consent.
Demographic information was gathered up by trained interviewers.
The complete criterion for selecting participants was addressed in our previous work. 38 This study has gained its approval from the in-

| Polymorphism selection and genotyping
We retrieved four SNPs with potential function in the LIN28A gene from the dbSNP database and SNPinfo software. 39 Selection genotypes (adjusted OR = 1.26, 95% CI = 1.01-1.56). Stratification analysis further revealed risk effect of rs3811464 G>A, rs34787247 G>A and 1-4 risk genotypes in some subgroups. Haplotype analysis of these four SNPs yields two haplotypes significantly correlated with increased neuroblastoma susceptibility. Overall, our finding indicated that LIN28A SNPs, especially rs34787247 G>A, may increase neuroblastoma risk.

K E Y W O R D S
case-control study, LIN28A, neuroblastoma, polymorphism, risk criteria were briefly described below: (a) the minor allele frequency reported in HapMap was >5% for Chinese Han subjects; (b) putative functional potentials SNPs located in the 5′-flanking region, exon, 5′-untranslated region (UTR) and 3′ UTR, which might affect transcription activity or binding capacity of the microRNA binding site; and (c) SNPs in low linkage disequilibrium (LD) with each other (R 2 < 0.8). There was no significant LD (R 2 < 0.8) among these four SNPs of LIN28A (R 2 = 0.183 between rs3811464 and rs3811463, R 2 = 0.009 between rs3811464 and rs34787247, R 2 = 0.054 between rs3811464 and rs11247957; R 2 = 0.03 between rs3811463 and rs34787247, R 2 = 0.052 between rs3811463 and rs11247957; R 2 = 0.002 between rs34787247 and rs11247957) ( Figure S1). The locations of these SNPs in the LIN28A are as below: rs3811464 G>A in the upstream, rs3811463 T>C, rs34787247 G>A, and rs11247957 G>A are all in the 3′ UTR. More detailed selection standards were exhibited at our previous work. 38 DNA was isolated from the blood sample using a TIANamp Blood DNA Kit (TianGen Biotech Co. Ltd.). Then, the DNA was further performed to genotype using the TaqMan methodology instructed by the manufacturers. [40][41][42] Negative controls (water samples) were used to ensure genotyping preciseness. A repeated genotyping of 10% randomly selected sample was also conducted in all plates with concordance rates of 100%.

| Statistical analysis
The χ 2 test was applied to test the difference in the distributions of subject characteristics between the cases and controls.
A goodness-of-fit chi-squared test was adopted to find whether there exists Hardy-Weinberg equilibrium (HWE) among controls.
Logistic regression analysis was applied to detect any association with neuroblastoma risk, with the crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs). The adjusted ORs were adjusted for age and gender. We determined the risk genotypes for each SNP based on its association with neuroblastoma risk. If a genotype of a SNP was shown to increase neuroblastoma risk (OR > 1), the genotype was regarded as a risk genotype. Carriers of 3 risk genotypes represented those carrying three risk genotypes of the four SNPs, while 1-4 risk genotypes represented those carrying 1-4 risk genotypes. 43 The stratification analyses were also performed to identify the associations by age, gender, sites of origins and clinical stages. Moreover, a combination of rs3811464 G>A, rs3811463 T>C, rs34787247 G>A, rs11247957 G>A was regarded as a haplotype. Unphased genotype data were used to identify haplotype frequencies and individual haplotypes.
Logistic regression analyses also help to obtain haplotype frequencies and distinct haplotypes, with the adjustment for gender and age. 44,45 The haplotype of the highest rate was used as the reference group to calculate ORs for haplotype associated with neuroblastoma risk. We set P < .05 as a significant borderline for all tests. We used SAS 9.1 (SAS Institute, Cary, NC) to compute all statistics.

| The relationship between LIN28A SNPs and neuroblastoma susceptibility
The association of all variant genotypes of the four LIN28A SNPs (rs3811464 G>A, rs3811463 T>C, rs34787247 G>A, rs11247957 G>A) with neuroblastoma risk is shown in Table 1 for combined subjects and in Table S2 for divided subjects. All these SNPs in controls were in accordance with HWE (all with an HWE P > .05).
In the single-locus analysis, only one variant, rs34787247 G>A,  Table 2

| LIN28A haplotype analysis
We further determined whether the haplotypes of the four LIN28A SNPs were linked to neuroblastoma risk. As shown in Table 3, the haplotype consisting of wild-type alleles (GTGG) was defined as the reference group. We detected a significant elevated neuroblas-  .646 Abbreviations: AOR, adjusted odds ratio; CI, confidence interval.
Significance of bold values are the P values less than 0.05 or the 95% CIs excluded 1.
a Adjusted for age and gender, omitting the corresponding stratify factor.

| D ISCUSS I ON
At the present, there remain many hidden genetic factors in association with neuroblastoma risk to be discovered to fill up the knowledge gaps. Thus, the identification of more polymorphisms is needed to unearth the full range of neuroblastoma susceptibility variations.
Herein, we undertook a four-centre case-control study to investigate the role of LIN28A polymorphisms on neuroblastoma risk in Chinese children. We are the pioneer in unveiling the association of the rs34787247A allele with an elevated neuroblastoma risk in a Chinese population. Abbreviations: CI, confidence interval; OR, odds ratio.
Significance of bold values are the P values less than 0.05 or the 95% CIs excluded 1.
a The haplotype order was rs3811464, rs3811463, rs34787247 and rs11247957. b Obtained in logistic regression models with adjustment for age and gender. carrying 3 or 1-4 risk genotypes. However, no significant relationship of the rest three variants rs3811464, rs3811463 and rs11247957 was detected.
We carried out a pioneering study on the association between LIN28A gene SNPs and susceptibility to neuroblastoma. Limitations also existed. Firstly, the sample size is not large enough to generate reliable statistics. Some of the results might be merely fortuitous events, particularly the stratification analysis. Secondly, we examined four SNPs in this research. More potential neuroblastoma risk-associated SNPs in the LIN28A gene await to be explored.
Thirdly, although the participants were enrolled from four different cities, findings from the restricted Chinese population could not be extrapolated to other ethnicities directly. Lastly, environmental factors were not considered in this study.
In all, we presented a multi-centre case-control study in Chinese children. For the first time, our findings unveiled a contributing role of LIN28A gene SNPs in neuroblastoma risk. In the future, an integrative analysis, covering more profound and specific factors, with environmental factors, genetic-environmental interaction, should be carried out to unearth the aetiology of neuroblastoma.

ACK N OWLED G EM ENTS
This study was supported by grants from the Pearl River S&T Nova