Intracellular iron levels alter the protein interactome of PCBP1

Abstract Background: Iron is essential for cellular function. Homeostasis of iron is controlled post-transcriptionally via highly conserved mechanisms. We recently found broad changes in alternative splicing with reduced iron availability, consistent with numerous operant mechanisms of iron sensitive RNA regulation. Poly C binding proteins (PCBPs) are RNA-binding proteins that regulate alternative splicing, translation, and RNA stability. PCBPs have also been found to be critical iron chaperones regulating intracellular iron flux by delivering iron to target proteins, such as ferritin. We have found PCBP1 to have iron-sensitive alternative splicing activity. Specifically, we found PCBP1-mediated splicing events that change with iron chelation and loss of PCBP1 attenuates iron chelation-induced splicing. Moreover, PCBP1 has enhanced association with RNA during iron chelation and mutation of PCBP1 residues associated with iron binding also enhances RNA association. The specific molecular mechanisms by which iron modulates PCBP1 mRNA regulation is not known. Because PCBP1 is a splicing factor that participates in ribonucleoprotein complexes, we hypothesize that low iron levels affect the PCBP1 protein interactome to influence splicing outcomes. Study Aim: Determine how protein-protein interactions contribute to iron-sensitive RNA regulation by PCBP1. Methods: We performed an APEX-proximity labeling assay to assess PCBP1 protein interactors under conditions of iron chelation or iron overload. APEX-PCBP1 fusion protein was expressed in K562 cells, and cells were treated with iron chelation (21H7) or iron overload (ferric ammonium citrate, FAC). Cells were then treated with biotin-phenol and hydrogen peroxide to biotinylate proteins proximal to APEX-PCBP1. Omission of hydrogen peroxide served as background controls. LC-MS was performed on streptavidin pulldown of whole cell lysate to determine enrichment or depletion of biotinylated proteins. Results: Altering iron levels resulted in widespread changes in biotinylated proteins that were proximal to PCBP1. Notably, we found a network of pre-mRNA binding proteins that are preferentially biotinylated under iron chelation compared to iron overload, including LSM5 (Log2 fold change [L2FC]=3.48, 21H7/FAC), LSM1 (L2FC=3.45, 21H7/FAC), and PRPF39 (L2FC =2.62, 21H7/FAC). Notably, ACO1 biotinylation was dramatically increased in chelation compared to overload (L2FC =5.46, 21H7/FAC). We also found various kinases with increased biotinylation under iron chelation compared to iron overload, including MAP kinases such as MAPK14 (L2FC =3.40, 21H7/FAC) and MAP2K2 (L2FC = 1.89, 21H7/FAC). Conclusions: These data support a model wherein low iron levels promote PCBP1 association with splicing factors to enhance alternative splicing regulation. Future work will explore the role of post-translational modifications in iron-sensitive PCBP1 function.