KIAA1522 is a novel prognostic biomarker in patients with non-small cell lung cancer

Abstract

Nowadays, no robust biomarkers have been applied to clinical practice to provide prognostic evaluation of non-small cell lung cancer (NSCLC). This study aims to identify new potential prognostic biomarkers for NSCLC. In the present work, KIAA1522 is screened out from two independent GEO datasets as aberrantly up-regulated gene in NSCLC tissues. We evaluate KIAA1522 expression immunohistochemically in 583 NSCLC tissue samples and paired non-tumor tissues. KIAA1522 displays stronger staining in NSCLC cases than in adjacent normal lung tissues. Importantly, patients with KIAA1522 overexpression had a significantly shorter overall survival compared to those with low expression (P < 0.00001). Multivariate Cox regression analyses show that KIAA1522 is an independent prognostic indicator, even for early-stage NSCLCs (P = 0.00025, HR = 2.317, 95%CI: 1.477-3.635). We also found that high expression of KIAA1522 is a significant risk factor for decreased overall survival of the patients who received platinum-based chemotherapy. Gene set enrichment analysis (GSEA) and functional studies reveal that KIAA1522 is associated with oncogenic KRAS pathways. Taken together, high expression of KIAA1522 can be used as an independent biomarker for predication of poor survival and platinum-resistance of NSCLC patients, and aberrant KIAA1522 might be a new target for the therapy of the disease.

Figure 1. Identification of KIAA1522 as an overexpressed gene in NSCLC.

(A) The flow chart of the present work. (B) Screen of uncharacterized genes (genes without GO function annotation) in GSE19804 and GSE32863 datasets. For each selected probe, the average value of log₂ (Tumor/Non-tumor) was calculated and shown in the rank order. (C) The average values of log₂ (Tumor/Non-tumor) of the genes presented in both GSE19804 and GSE32863 datasets were shown. (D–F) Transcript levels of KIAA1522 in non-tumor tissues and NSCLC tissues from GSE19804 (D), GSE32863 (E) and GSE19188 (F) datasets. Data are represented as scatter dot plot with mean ± s.d. (by t-test analysis, ***P < 0.001).

Figure 2. Expression status of KIAA1522 protein in NSCLC tissue samples.

(A) The mRNA levels of KIAA1522 in 10 pairs of NSCLC tissue samples relative to their adjacent non-malignant tissues measured by real-time PCR. The expression of GAPDH mRNA was used as loading control in the real-time PCR analysis. (B) Western blotting analysis of KIAA1522 protein levels in NSCLC tissue samples (T) and the matched non-tumor tissues (N). GAPDH was used as loading control in the immunoblotting assays. (C) Representative examples of high (positive) and low (negative) expression of KIAA1522 in SCC or ADC samples and their adjacent non-malignant tissues. (D) Representative immunohistochemical microphotographs of KIAA1522 with negative (1,5), weak (2,6), moderate (3,7) and strong (4,8) expression. The subtypes are SCCs (1–4) and ADCs (5–8). Bar = 100 μm. (E) The expression scores indicating the protein levels of KIAA1522 for NSCLC and non-tumor lung tissues. Boxes represent the upper and lower quartiles and median; whiskers show the data points that are neither lower than the first percentile nor greater than the 99th percentile (by Mann-Whitney test analysis, ***P < 0.001).

Figure 3. Correlation of KIAA1522 mRNA in NSCLC tissues and patients’ survival/survival signatures.

(A) Kaplan-Meier curves showing the survival percentage of patients with high KIAA1522 mRNA and low KIAA1522 mRNA levels in stage I and II patients in the dataset GSE31210. Samples with incomplete resection or adjuvant therapy were excluded for prognosis analysis. P values are shown in the graph, by log-rank test. (B) Survival analysis of the stage II lung cancer patients with KIAA1522 high and low expression using the Kaplan-Meier Plotter website for lung cancer (Version 2015). (C,D) Gene set enrichment analysis in GSE31210 showed that the genes within a good survival signature were observed to enrich in the groups with low KIAA1522 expression (C), while analysis in GSE63074 illustrated the enrichment of poor survival signature genes in the subset of high KIAA1522 expression group (D).

Figure 4. Relationship between expression of KIAA1522 and patient OS.

Kaplan-Meier curves showing the association between expression of KIAA1522 and patients OS in all stages, in the stratified stages and in different histological tumor types in tissue samples (all of the P values are shown in the graph, by log-rank test).

Figure 5. Relationship between expression of KIAA1522 and overall survival of patients with platinum-based chemotherapy.

Kaplan-Meier curves showing that patients with high expression of KIAA1522 had a poorer OS compared to those with low expression of the protein (P = 0.007, by log-rank test).

Figure 6. Correlation of KIAA1522 expression with KRAS pathway signatures.

(A–C) Gene sets enrichment analysis of the NSCLC dataset showing that high KIAA1522 expression was associated with the hyper-activation of KRAS signatures in dataset GSE63074 (A), GSE37745 (B) and GSE 31210 (C). (D) KIAA1522 high expression showed positive correlation with activation of MEK signaling in GSE37745. (E) The expression of KIAA1522 in the A549 and H460 cells that were transfected with control siRNA or two independent KIAA1522 siRNAs for three days. Data are shown as mean ± s.d. (by t-test analysis, *** P < 0.001, n = 3). (F) The relative KRAS mRNA levels in the A549 and H460 cells transfected with control siRNA or two independent KIAA1522 siRNAs for three days were analyzed by real-time PCR. Data shown are mean ± s.d. (by t-test analysis, ***P < 0.001, **P < 0.01, n = 3). (G) Western blotting analysis of the Phospho-ERK (Thr202/Tyr204), ERK, RAS, KIAA1522 and GAPDH (loading control) levels in the A549 and H460 cells transfected with the indicated siRNAs, four days after transfection. (H–I) The H460 cells were treated with DMSO (control) or the indicated concentration of U0126 for 24 hours; the cells were then subjected for real-time PCR (H) and western blotting (I) analysis of KIAA1522 expression. The efficiency of MEK inhibitor U0126 were revealed by the reduced Phospho-ERK (Thr202/Tyr204) levels tested by western blot. Data shown are mean ± s.d. (by t-test analysis, **P < 0.01, ***P < 0.001, n = 3).

Figure 7. Knockdown of KIAA1522 expression inhibits the growth of NSCLC cells.

(A,B) The growth rate of NSCLC cell lines A549 and H460 transfected with the indicated siRNAs were tested by the MTT assays (A) and colony formation assays (B). Data shown are mean ± s.d. (by ANOVA analysis, ***P < 0.001, n = 3).