1. Selection of candidate SNPs
2. Association between candidate SNPs and CRC risk
3. The effect of rs2147578 on miR-128-3p: lnc-LAMC2-1:1 binding
rs2147578位于lnc-LAMC2-1:1 的外显子区，该位点位于miR-128-3p结合lnc-LAMC2-1:1 的区域；
CG/GG与CC基因型相比，miR-128-3p结合lnc-LAMC2-1:1 的能力变弱，因此miR-128-3p可能对lnc-LAMC2-1:1 有下调作用，CG/GG亚型lnc-LAMC2-1:1 表达上调；
The functional prediction of rs2147578 at the 1q25.3 locus. ( A ) The lnc-LAMC2-1:1locates on chromosome 1, overlaps with the protein-coding gene LAMC1 and is close to LAMC2 . ( B ) The rs2147578 is in the first exon of lnc-LAMC2-1:1 and 26th intron of LAMC1. ( C ) The rs2147578 C allele gain the binding of miR-128-3p as predicted by lncRNASNP database. ( D ) Expression correlation between rs2147578 and LAMC2 .
Bar graphs of ( A ) and ( B ) show the relatively luciferase activity of vectors containing the rs2147578G or C allele in SW480 and LoVo cells, respectively. Luciferase expression vectors containing the lnc-LAMC2-1:1 with the rs2147578G or C allele were constructed and co-transfected with miRNA mimics (miR-128-3p) in SW480 and LoVo cells, respectively. Renilla luciferase/Firefly luciferase were calculated and normalized to blank or NC controls as relatively luciferase activity. Six replicates for each group were conducted and all transfection experiments were repeated at least three times. Data are presented as the mean ± SD and asterisk indicates a significant change ( P < 0.001).
lncRNA-SNP-miRNA [SNP影响miRNA结合lncRNA] —— LAMC2癌症相关基因 —— CRC疾病表型
SNP位点rs2147578的不同基因型通过影响miR-128-3p对lnc-LAMC2-1:1 的结合与调控能力，影响lnc-LAMC2-1:1 的表达，可能通过对癌基因LAMC2的表达进行调控参与结直肠癌发生。
1. Selection of candidate SNPs
CRC risk tagSNPs identified by GWASs were downloaded from the NHGRI GWAS Catalog up to December 31, 2013 ( 20 ). We also searched the literature via PubMed using the search terms of ‘GWAS’, ‘Genome-wide association studies’ and ‘colorectal cancer’ to retrieve CRC risk tagSNPs. Then, Haploview software 4.2 ( 33 ) was used to calculate the linkage disequilibrium (LD) blocks of each tagSNP by analyzing the Chinese Han Beijing (CHB) genotype information of ±500kb around the tagSNP (setting R2 ≥ 0.8), and these LD blocks were defined as CRC susceptibility loci. Next, SNPs in these CRC susceptibility loci and lncRNAs were screened out. SNP data were downloaded from dbSNP of NCBI (dbSNP v138), while lncRNA information was obtained from the LNCipedia database ( 34 ). Further to narrow down the potential functional SNPs, we only selected SNPs which may impact miRNA:lncRNA interaction according to the lncRNASNP database ( 35 ) and limited the minimum allele frequency of SNPs in the Chinese population (CHB) ≥ 0.05. Finally, eight SNPs were selected as candidate SNPs.
2. Study populations
4. Copy number of the locus lnc-LAMC2-1:1：
细胞株：LoVo and SW480 cells
All primers used in quantitative PCR are shown as follows, lnc-LAMC2-1:1 primer 5′- GGCCCAAGGAAGAACTAAGG -3′ and 5′- ATCCAAACCAACATCCACCC-3′; β-globin primer: 5′-GGTGAGCCAGGCCATCACTA-3′ and 5′-GGCAACCC TAAGGTGAAGGC-3′. The copy number of lnc-LAMC2-1:1 gene was normalized to that of β-globin (known as two).
5. Plasmid construction
质粒： psiCHECK TM -2 vector
突变：定点突变rs2147578 site (G>C).
The wild-type transcript sequence of lnc-LAMC2-1:1 (524bp) was downloaded from the LNCipedia database and synthesized by Genewiz Company (Suzhou, China). The sequence was cloned into the psiCHECK TM -2 vector (Promega, Madison, WI). The mutation type of lnc-LAMC2-1:1 was generated by site-specific mutagenesis at the rs2147578 site (G>C).
The primer sequences for site-specific mutagenesis were 5′-CATAGTCCCTCAGTGTGGGTCATTTTCATTAG-3′` and 5′-CTAATGAAAAT GACCCACACTGAGGGACTATG-3′. The mimics of miR-128-3p were chemically synthesized by Shanghai GenePharma Company (Shanghai, China).
6. Cell cultures, transient transfections and luciferase reporter assays
7. Cis-eQTL analysis
For the eQTL analysis, we downloaded the germline genotypes, the somatic copy number, methylation and expression profiles of colon adenocarcinoma (COAD) and rectum adenocarcinoma (READ) from ‘The Cancer Genome Atlas’ (TCGA) data portal. Of them, we selected 254 individuals with genotypes that clustered with HapMap CEU controls by the EIGENSTRAT program ( 38 ) and both germline genotypes and expression data were available. The genotypes of TCGA were obtained by Affymetrix 6.0 array. For each interesting SNP, if the SNP was not designed in the Affymetrix 6.0 array, we first imputed its genotypes by impute2 (39) software using the genotypes of all SNPs on the Affymetrix 6.0 array within 2Mb of either side of the SNP. Then, for each SNP, the association between the SNP and nearby genes was evaluated using a linear regression model with the effects of somatic copy number and CpG methylation being deducted. The detail of the algorithms is as previously reported in Li et al.
8. Statistical analysis