Fibroblast growth factor receptors (FGFRs) are integral membrane proteins that transmit signals through the plasma membrane. FGFRs signaling needs to be precisely adjusted as aberrant FGFRs function is associated with development of human cancers or severe metabolic diseases. The subcellular localization, trafficking and function of FGFRs rely on the formation of multiprotein complexes. In this study we revealed galectins, lectin family members implicated in cancer development and progression, as novel FGFR1 binding proteins. We demonstrated that galectin-1 and galectin-3 directly bind to the sugar chains of the glycosylated extracellular part of FGFR1. Although both galectins compete for the same binding sites on FGFR1, these proteins elicit different impact on FGFR1 function and cellular trafficking. Galectin-1 mimics fibroblast growth factor as it efficiently activates FGFR1 and receptor-downstream signaling pathways that result in cell proliferation and apoptotic evasion. In contrast, galectin-3 induces extensive clustering of FGFR1 on the cell surface that inhibits constitutive internalization of FGFR1. Our data point on the interplay between extracellular galectins and FGFRs in the regulation of cell fate.

FGF1 and FGF2 bind to specific cell-surface tyrosine kinase receptors (FGFRs) and activate intracellular signaling that leads to proliferation, migration or differentiation of many cell types. Besides this classical mode of action, under stress conditions, FGF1 and FGF2 are translocated in a receptor-dependent manner via the endosomal membrane into the cytosol and nucleus of the cell. However, despite many years of research, the role of translocated FGF1 and FGF2 inside the cell remains unclear. Here, we reveal an anti-apoptotic activity of intracellular FGF1 and FGF2, which is independent of FGFR activation and downstream signaling. We observed an inhibition of cell apoptosis induced by serum starvation or staurosporine upon treatment with exogenous FGF1 or FGF2, despite the presence of highly potent FGFR inhibitors. Similar results were found when the tyrosine kinase of FGFR1 was completely blocked by a specific mutation. Moreover, the anti-apoptotic effect of the growth factors was abolished by known inhibitors of the translocation of FGF1 and FGF2 from the endosomes to the interior of the cell. Interestingly, FGF2 showed higher anti-apoptotic activity than FGF1. Since FGF2 is not phosphorylated by PKCδ and is present inside the nucleus longer than is FGF1, we speculated that the different activities could reflect their diverse nuclear export kinetics. Indeed, we observed that FGF1 mutations preventing binding to nucleolin and therefore phosphorylation in the nucleus affect the anti-apoptotic activity of FGF1. Taken together, our data indicate that the translocation of FGF1 and FGF2 protects cells against apoptosis and promotes cell survival.

No abstract

Fibroblast growth factor 1 (FGF1) is considered primarily as a ligand for FGF surface receptors (FGFRs) through which it activates a number of cellular responses. In addition to its canonical mode of action, FGF1 can act intracellularly, before secretion or after internalization and translocation from the cell exterior. The role of FGF1 inside the cell is to provide additional protection against apoptosis and promote cell survival. The FGF1 protein contains a specific N-terminal nuclear localization sequence (NLS) that is essential for its efficient transport to the nucleus. Here, we investigated the role of this sequence in the anti-apoptotic response of FGF1. To this end, we produced recombinant FGF1 variants with mutated or deleted NLS and added them to apoptosis-induced cells in which FGFR1 was inactive, either as a result of chemical inhibition or kinase-dead mutation. After internalization, all FGF1 variants were able to protect the differentiated cells from serum starvation-induced apoptosis. To verify the results obtained for NLS mutants, we knocked down LRRC59, a protein that mediates the nuclear transport of FGF1. Upon LRRC59 silencing, we still observed a decrease in caspase 3/7 activity in cells treated exogenously with wild-type FGF1. In the next step, FGF1 variants with mutated or deleted NLS were expressed in U2OS cells, in which apoptosis was then induced by various factors (e.g., starvation, etoposide, staurosporine, anisomycin and actinomycin D). Experiments were performed in the presence of specific FGFR inhibitors to eliminate FGFR-induced signaling, potentially activated by FGF1 proteins released from damaged cells. Again, we found that the presence of NLS in FGF1 is not required for its anti-apoptotic activity. All NLS variants tested were able to act as wild type FGF1, increasing the cell viability and mitochondrial membrane potential and reducing the caspase 3/7 activity and PARP cleavage in cells undergoing apoptosis, both transiently and stably transfected. Our results indicate that the nuclear localization of FGF1 is not required for its intracellular anti-apoptotic activity in differentiated cells and suggest that the mechanism of the stress response differs according to the level of cell differentiation.