Negative controls were obtained by omitting the primary antibody [8]. Statistical analysis The criterion for a positive reaction was a single epithelial cell with yellow particles in its plasma membrane and cytoplasm. Immunostaining was assessed in a blinded manner for extent and intensity.

In brief, a sample with no positive epithelial cells was scored as 0, that with less than 25% total positive epithelial cells was scored as 1+, that with positive epithelial cells accounting for more than 25% but less than 50% of the total was scored as 2+, that with more than 50% but less than 75% positive cells was scored as 3+, and that with more than 75% positive cells was scored as 4+. The intensity of immunostaining Etomoxir in vivo was scored semiquantitatively as follows: no obvious yellow particle in epithelial cell plasma membrane or cytoplasm as 0; with light yellow particles as 1+ (weak); with general yellow particles as 2+ (moderate); and with deep yellow particles as 3+ (strong). For each case, an immunoscore was calculated as the product of 2 scores assessed separately. Statistical analysis was performed using SPSS 17 software (SPSS, Inc, Chicago, IL, USA). The differential expression of LCMR1 protein between tumorous tissues and Selleck Batimastat normal tissues was determined by Mann-Whitney U-test. The correlations between LCMR1 expression

and clinicopathologic characteristics were analyzed using Pearson χ2 analysis. The influence of each variable on the expression of LCMR1 was assessed by logistic regression analysis. In survival analysis, Kaplan-Meier curves were drawn, univariate and multivariate analyses in a Cox proportional hazards model were used for LCMR1 scores. All statistical tests were 2-sided, and P values of 0.05 or less were considered statistically significant. Results Cloning and identification Aspartate of a novel gene differentially expressed

in 95C and 95D cell lines using DD-PCR In order to find lung cancer metastasis related genes, the DD-PCR method was used to identify genes differentially expressed in human 95C and 95D cell lines, which have the same genetic backgrounds but different metastatic potential. Several cDNAs were found expressed differentially in these two cells (Figure 1A). These fragments were subcloned into T easy vector, sequenced, and analyzed for nucleotide and amino acid homology in the GenBank database. Of these, a 778 bp cDNA fragment, designated as P9, expressed higher in 95D cells than in 95C cells, did not show a significant homology with any nucleotide/amino acid sequence in the database, but has many supports of EST. After alignment in Genbank Genomic Database, we found this fragment existed in chromosome 11 discontinuously. These suggested that this cDNA might code a novel gene, and thus was selected for further studies. RACE (rapid amplification of cDNA ends) was used to get the complete cDNA.