Analysis of LOI of LIT1, IGF2 and H19 RT-PCR at LIT1, IGF2 and H19 were further analysed for possible allele-specific expression. One microgram total RNA from heterozygous

normal and tumor samples was reverse transcribed for the first strand cDNA using the AMV-RT-PCR system (Sangon, Shanghai, China) in a 20 μl reaction. This reaction mixture was added to 80 μl of 100 μM dNTP and 2 mM MgCl2, 10% glycerol and 2.5 units Taq polymerase in 1 × PCR buffer. Amplification conditions were carried out as described above. For negative PCR controls, the same primers and reaction conditions with RNA, minus the reverse transcription step were performed. After RsaI digestion of RT-PCR products, informative cases of LIT1 with LOI show biallelic expression of both the 222 and 410 bp, while without LOI, showing 222 PI3K inhibitor or 410 band. For IGF2, the RT-PCR product was analysed on 1.5% agarose gel to verify the 1.12 kb bands, which were smaller than those observed in DNA analysis (1.4 kb) with the inclusion of 280 bp intron. Nested PCR wascontinued with the primer P2 as P3 from this 1.12-kb RT-PCR product, resulting in a 292-bp band. After digesting the 292-bp cDNA product from the above RT-PCR reaction with ApaI and

HinfI, the presence of 256-bp and 231-bp fragments in a tumor sample indicated biallelic expression. The presence of either the 256 bp or 231 bp band was considered as retention of imprinting. RT-PCR products of H19 resulted in an obvious 575 bp band from

cDNA compared to the control of 655 bp fragment from Ro 61-8048 Bay 11-7085 genomic DNA which includes 80 bp intron. Constitutive imprinting yielded either a single see more 575-bp band or 407- and 168-bp bands, LOI resulting in 575-bp, 407- and 168-bp fragments after RsaI digestion. The threshold for scoring LOI was defined as a ratio of less than 3-fold difference in expression between two alleles [29]. Statistical analysis The prevalence of LOI in patients with gastric cancer was described as a proportion. The demographic and clinicopathological characteristics in LOI positive and LOI negative patients were compared and tested using the Chi-Square test. Logistic regression analyses were used to compute the odds ratios (ORs) and 95% confidence interval (CI). Independent sample t-test was used to compare the mean age differences between LOI-positive and -negative patients. All statistical analyses were performed with statistical software with SPSS version 13.0 for windows (SPSS, Inc., Chicago IL). All p-values were two-tailed with 0.05 as statistical significance. Results Loss of imprinting at LIT1, IGF2 and H19 in gastric cancer tissues We examined the status of genomic imprinting of the LIT1, IGF2 and H19 genes in 89 gastric cancers by PCR-restriction fragment length polymorphism (RFLP) analysis (Fig. 1, Fig. 2, Fig. 3). Of the 89 tumours analysed, 22, 40 and 35 cases were heterozygous and thus informative for LIT1, IGF2 and H19 LOI analyses respectively as shown in Table 1.