Abstract
A switch in the alternative splicing pattern in protein 4.1R pre-mRNA, involving activation of exon 16 splicing, occurs in late erythropoiesis. Because the inclusion of exon 16 leads to synthesis of protein isoforms with high affinity for spectrin and actin, and mechanical strengthening of the membrane, this switch is critical for normal erythyroid differentiation. Previous studies have shown that Fox-2 protein is a splicing regulator that binds to intronic enhancer sequences downstream of exon 16, and that this binding stimulates the inclusion of exon 16. Fox binding sites are also located downstream of other tissue-specific alternative exons, some of which exhibit variable splicing efficiencies, leading us to investigate whether splicing switch strength is influenced by variations in the sequence of the Fox binding-site motif (UGCAUG) and by the number of these protein binding sites. A series of minigenes was created in which protein 4.1R E16 splicing efficiency was measured in transfection studies. Mutation of the first U residue of the Fox binding site resulted in a significantly weaker, but still Fox dependent activation of splicing, whereas mutation of the terminal G residue dramatically reduced enhancer activity. Insertion of two wild-type UGCAUG elements enhanced splicing substantially in a Fox protein concentration-dependent manner, and four elements gave even stronger inclusion. To test whether the reduced splicing efficiency observed for minigenes containing mutated enhancer elements was due to a reduction in Fox binding, surface plasmon resonance (SPR) was employed. Our binding studies indicate that full-length mouse Fox-2a bound to the UGCAUG sequence with high affinity (KD ∼100nM). Furthermore, in agreement with our functional splicing assays, Fox binding affinity for the hexameric sequence was reduced with mutations in position one, and almost eliminated with mutations in position six. There was an excellent correlation between binding affinity of Fox protein for the enhancer motif, and strength of enhancer activity measured in functional splicing assays. Additional studies demonstrated that Fox enhancer activity can be reduced by the presence of a closely linked binding site for hnRNP A1, a known splicing inhibitor. Taken together, these studies demonstrate that the composition and number of binding sites for Fox can significantly affect the efficiency of splicing for E16. It is likely that through these combinatorial mechanisms, Fox splicing factors are able to modulate the efficiency with which tissue specific exons in the erythroid alternative splicing program can be spliced into mature mRNA.
Author notes
Disclosure: No relevant conflicts of interest to declare.