Evidence favoring a positive feedback loop for physiologic auto upregulation of hnRNP-E1 during prolonged folate deficiency in human placental cells

Background: Previously, we determined that heterogeneous nuclear ribonucleoprotein E1 (hnRNP-E1) functions as an intracellular physiologic sensor of folate deficiency. In this model, L-homocysteine, which accumulates intracellularly in proportion to the extent of folate deficiency, covalently binds to and thereby activates homocysteinylated hnRNP-E1 to interact with folate receptor-a mRNA; this high-affinity interaction triggers the translational upregulation of cell surface folate receptors, which enables cells to optimize folate uptake from the external milieu. However, integral to this model is the need for ongoing generation of hnRNP-E1 to replenish homocysteinylated hnRNP-E1 that is degraded. Objective: We searched for an interrelated physiologic mechanism that could also maintain the steady-state concentration of hnRNP-E1 during prolonged folate deficiency. Methods: A novel RNA-protein interaction was functionally characterized by using molecular and biochemical approaches in vitro and in vivo. Results: L-homocysteine triggered a dose-dependent high-affinity interaction between hnRNP-E1 and a 25-nucleotide cis element within the 5#-untranslated region of hnRNP-E1 mRNA; this led to a proportionate increase in these RNA-protein complexes, and translation of hnRNP-E1 both in vitro and within placental cells. Targeted perturbation of this RNA-protein interaction either by specific 25-nucleotide antisense oligonucleotides or mutation within this cis element or by small interfering RNA to hnRNP-E1 mRNA significantly reduced cellular biosynthesis of hnRNP-E1. Conversely, transfection of hnRNP-E1 mutant proteins that mimicked homocysteinylated hnRNP-E1 stimulated both cellular hnRNP-E1 and folate receptor biosynthesis. In addition, ferrous sulfate heptahydrate [iron(II)], which also binds hnRNP-E1, significantly perturbed this L-homocysteine–triggered RNA-protein interaction in a dose-dependent manner. Finally, folate deficiency induced dual upregulation of hnRNP-E1 and folate receptors in cultured human cells and tumor xenografts, and more selectively in various fetal tissues of folate-deficient dams. Conclusions: This novel positive feedback loop amplifies hnRNP-E1 during prolonged folate deficiency and thereby maximizes upregulation of folate receptors in order to restore folate homeostasis toward normalcy in placental cells. It will also functionally impact several other mRNAs of the nutrition-sensitive, folate-responsive posttranscriptional RNA operon that is orchestrated by homocysteinylated hnRNP-E1.