Identification and validation of roles of lysyl oxidases in the predictions of prognosis, chemotherapy and immunotherapy in glioma.

Abstract

Background: Previous investigations have illustrated that lysyl oxidase family enzymes (LOXs) are contributing factors for tumor progression and remodeling immunomicroenvironment. However, it is scarce regarding comprehensive analysis of LOXs in the predictions of prognosis, chemotherapy and immunotherapy in glioma, the highly invasive brain tumor. Our present work aimed to explore the prognostic value, chemotherapeutic drug sensitivity and immunotherapy according to distinct LOXs expressions in glioma through bioinformatics analysis and experimental verification. Methods: We collected gene expression data and clinical characteristics from the public databases including Chinese Glioma Genome Atlas (CGGA)-325, CGGA-693, the Cancer Genome Atlas (TCGA), IMvigor210 and Van Allen 2015 cohorts. The correlations between the clinicopathological factors and differential LOXs expressions were analyzed. The ROC curve and Kaplan-Meier analysis were conducted to evaluate the prediction ability of prognosis. Chemotherapeutic drug sensitivity via distinct LOXs expression levels was predicted using the pRRophetic package. Immune score, immune cell infiltration and immune checkpoint expression levels were also analyzed through diverse algorithms in R software. Finally, mRNA and protein expressions of LOXs were validated in glioma cells (T98G and A172) by real-time quantitative PCR and Western blot, respectively. Results: Our results demonstrated that high levels of LOXs expressions were positively associated with glioma grades, older age and MGMT unmethylated status while elevations of LOXs were negatively correlated with IDH mutation or 1p/19q co-deletion. Furthermore, the glioma patients with low levels of LOXs also exhibited better prognosis. Also, differential LOXs expressions were associated with at least 12 chemotherapeutic drug sensitivity. Besides, it was also found that glioma patients with high LOXs expressions showed higher enrichment scores for immune cell infiltration and increased levels of immune checkpoints, suggesting the critical role of distinct LOXs expression levels for glioma immunotherapy. The predictive roles of LOXs expression in tumor immunotherapy were also validated in two immunotherapy cohorts including IMvigor 210 and Van Allen 2015. Experimental results revealed that expressions of LOX, LOXL1, LOXL2, and LOXL3 were higher in glioma cell lines at mRNA and protein levels. Conclusion: Our findings altogether indicate that LOXs have potent predictive value for prognosis, chemotherapy and immunotherapy in glioma patients.