New Functions of Intracellular LOXL2: Modulation of RNA-Binding Proteins

Abstract

Lysyl oxidase-like 2 (LOXL2) was initially described as an extracellular enzyme involved in extracellular matrix remodeling. Nevertheless, numerous recent reports have implicated intracellular LOXL2 in a wide variety of processes that impact on gene transcription, development, differentiation, proliferation, migration, cell adhesion, and angiogenesis, suggesting multiple different functions for this protein. In addition, increasing knowledge about LOXL2 points to a role in several types of human cancer. Moreover, LOXL2 is able to induce the epithelial-to-mesenchymal transition (EMT) process-the first step in the metastatic cascade. To uncover the underlying mechanisms of the great variety of functions of intracellular LOXL2, we carried out an analysis of LOXL2's nuclear interactome. This study reveals the interaction of LOXL2 with numerous RNA-binding proteins (RBPs) involved in several aspects of RNA metabolism. Gene expression profile analysis of cells silenced for LOXL2, combined with in silico identification of RBPs' targets, points to six RBPs as candidates to be substrates of LOXL2's action, and that deserve a more mechanistic analysis in the future. The results presented here allow us to hypothesize novel LOXL2 functions that might help to comprehend its multifaceted role in the tumorigenic process.

Keywords: EMT; LOXL2; RNA-binding proteins; intracellular LOXL2; nuclear interactome.

Figure 1

LOXL2’s nuclear interaction partners: (A) Immunoprecipitation of flag-LOXL2 in two independent experiments (IP-1 and IP-2) in HEK293T cells identified 107 proteins, which are listed below the Venn diagram. RBPs are marked in red. (B) Co-immunoprecipitation of endogenous LOXL2 in Hs578T cells and selected RBPs.

Figure 2

Molecular and biological functions of LOXL2’s nuclear interaction partners: Enriched categories were obtained using the DAVID functional annotation tool and ranked by the FDR value (−Log10 FDR) shown on the x-axis. (A) Functional enrichment analysis of molecular function using the GOTERM_CC_DIRECT category. (B) Functional enrichment analysis of biological function using the GOTERM_MF_DIRECT category. (C) Functional enrichment analysis of biological function using the GOTERM_BP_DIRECT category.

Figure 3

LOXL2’s SRCR-1 domain is required for protein interaction: (A) Schematic representation of wild-type LOXL2 (WT) and mutants carrying a deletion of the following SRCR domains: SRCR-1 (Δ-1), SRCR-2 (Δ-2), SRCR-3 (Δ-3), or SRCR-4 (Δ-4). (B) Whole-cell lysates of HEK293T cells transfected with LOXL2 WT or indicated SRCR-deletion mutants were immunoprecipitated with anti-ELAVL1/HuR antibody and analyzed by WB, with anti-LOXL2 antibody and anti-HuR antibodies as controls. (C) HEK293T cells were transfected with LOXL2 (WT) or the mutant carrying the deletion of the SRCR-1 domain (Δ-1); LOXL2 was immunoprecipitated using anti-HA antibody and analyzed by WB with the indicated antibodies. (D) The activity of the E-cadherin promoter in HEK293T cells was measured in the presence of the indicated LOXL2 mutants. The activity was determined as relative luciferase units (RLU) and normalized to the activity detected in the presence of the control pcDNA3 vector. Results represent the mean ± SEM. of at least three independent experiments performed in triplicate (*** p < 0.001, ns, not significant).

Figure 4

Gene expression profile of LOXL2-silenced cells: (A) Volcano plot showing transcriptome changes in LOXL2-depleted cells. The cutoffs were established at the log2 fold change > 1.0 or <−1.0 and the p-value < 0.05. Orange and blue dots represent significantly downregulated and upregulated genes, respectively. (B) Quantitative RT-qPCR confirming regulation of selected genes in LOXL2-ablated cells (dark green) compared to LOXL2 control cells (light green). Data are the mean ± SEM of three independent experiments assayed in triplicated samples in two independent breast cancer cell lines. The p-value was calculated by two-sided unpaired Student’s t-test. (C) GSEA plot of differentially expressed genes (DEGs) showing enrichment of hallmark signatures. FDR values (−log10 FDR) of such enrichments are shown on the x-axis. (* p < 0.05, ** p < 0.01, *** p < 0.001).

Figure 5

RBP candidates to be regulated by LOXL2: (A) List of differentially expressed genes in LOXL2-silenced cells with a described role in EMT. Up- and downregulated genes are marked in red and blue, respectively. (B) Volcano plot showing LOXL2-regulated EMT-related genes. The cutoffs were established at the log2 fold change > 1.0 (red line) or <−1.0 (blue line) and the p-value < 0.05 (grey line). Red and blue dots denote up- and downregulated EMT-related genes, respectively. (C) Enrichment plot of RBPs’ targets in the EMT-related signature relative to the non-EMT-related signature. The p-values were obtained by using Fisher’s exact test to compare the number of targets in both groups of genes. A p-value < 0.05 was considered statistically significant. Boxes marked in green represent significantly enriched RBPs.