Whole-transcriptome analysis of tumor tissues of lung adenocarcinoma reveals molecular changes under the influence of IFN-γ

IFN-γ is a critical cytokine which regulates both innate and adaptive immune. Ample evidences have elucidated the mechanisms of IFN-γ mediated antitumor effects, including driving T Helper Type 1 (Th1) polarization, activating cytotoxic lymphocyte (CTL) and natural killer T cells, upregulating major histocompatibility complex class I expression (MHC Class I) in tumor cells, and inhibiting tumor cells proliferation, inducing tumor cells apoptosis. Despite the multiple effects in cancer immune elimination phase (also known as cancer immune surveillance), IFN-γ also effectuates antitumor immunity and protumorigenic activities throughout the cancer immunoediting process including elimination phase and escape phase. One of the most important mechanism elucidating the dual effect of IFN-γ in immunoediting is IFN-γ induced PD-L1 expression in cancer cells. To obtain a detailed molecular appreciation of what was going on in tumor microenvironment under the influence of IFN-g, we performed global gene expression analysis using the Agilent oligonucleotide microarray system. First, we evaluated the IFN-γ expression in a uniform cohort of patients with IIIA lung adenocarcinoma from our patient archive. IFN-g expression was assessed by means of RT-PCR. The median DCt value from tested specimens was set as breakpoint, IFN-g expression below median DCt value was considered as low IFN-g group (IFN-g-) whereas above median DCt value was the high IFN-g group (IFN-g+). Then, we analyzed PD-L1 expression by IHC staining of biopsy specimens in whole slides. PD-L1 expression was quantified by multiplying the staining intensity scores by the percentage of positive cells. Biopsy specimens from 12 lung adenocarcinoma patients were included to perform the Agilent oligonucleotide microarray system. Six of biopsy specimens contain high levels of IFN-g while others expressing IFN-g at low or undetectable levels. Furthermore, we divided IFN-g+ group into two subgroups according to PD-L1 expression, three IFN-g+ specimens expressing PD-L1 (IFN-g+PD-L1+) but others not (IFN-g+PD-L1-).