30P Bioinformatics analysis for lysyl oxidases as therapeutic targets for breast cancer

LOX (lysyl oxidase) and LOXL 1-4 are amine oxidases, which catalyze the cross-linking of elastin and collagen in the extracellular matrix (ECM). This activity facilitates cell migration and the formation of metastases. Consequently, inhibition of these enzymes, in particular of LOXL2 has been suggested as a therapeutic strategy to prevent breast cancer metastasis. Although medicinal chemistry studies have struggled to specifically inhibit LOXL2, the importance of selectivity in this context is not yet clear. To explore the role of each LOX in breast cancer, a series of bioinformatic-based approaches was followed. The expression profile of LOXs and the impact of this expression on breast cancer patient survival were assessed using the TCGA database with GEPIA 2.0. The correlation between the expression of LOX enzymes and of those with other ECM-related genes was also assessed. The association between expression of LOXs and the breast tumor infiltrates was analyzed using TIMER 2.0. Higher levels of LOXL1, LOXL2, and LOXL3 are generally found in breast cancer tissues when compared with normal tissues, while LOXL4 expression is decreased, particularly in higher stages. The overexpression of LOXL1 has a negative impact on disease free survival (DFS) in all breast cancer subtypes, while LOXL2 negatively impacts the DFS of basal and HER2+ subtypes. The impact of LOXL3 on DFS is cancer subtype dependent. The downregulation of LOXL4 is associated with a reduction of DFS and overall survival in HER2+ patients. The high expression of these enzymes is associated with increased CAFs in all subtypes. The expression of LOX, LOXL1 and LOXL2 is linked with lower infiltration of B, T CD4+, and T CD8+ cells. This was not observed for LOXL3 and LOX4. A strong positive correlation was found between the expression of LOX, LOXL1, and LOXL2 and the expression of ECM-related genes, namely TIMP2, PDGFR-β, MMP-2, MMP-14, FN1, ITGA5 and ZEB1. A correlation between expression of LOX genes was also observed, particularly between LOX/LOXL2 and between LOXL1/LOXL2. Overall, we speculate that while the impact of LOXL3 inhibition may vary with breast cancer subtype, the specific therapeutical inhibition of both LOXL1 and LOXL2 but not of LOXL4 may be beneficial.