Abstract
Multiple components of the highly conserved let-7 cascade are developmentally regulated during the fetal-to adult transition of human ontogeny. Previous studies showed the let-7 miRNAs to be correlated with the developmentally regulated expression of fetal hemoglobin (HbF) in humans. Ex vivo manipulation of the main known let-7 regulator (LIN28) and downstream let-7 targets (IGF2BP1, IGF2BP3 and HMGA2) have shown variant increases in HbF levels (ranging from around 15-68% HbF) among cultured adult human erythroid cells. In addition, IGF2BP1 over-expression (OE) in cord blood erythroblasts raised HbF levels to an average of 90% compared to 55% in control transductions. Remarkably, in the fetal environment, the manipulation of the let-7 cascade has profound effects on HbF levels (cord blood control for the OE vectors: 57.4 ± 8.6%; LIN28A-OE: 90.8 ± 1.7%, p = 0.0003; IGF2BP3-OE: 79.0 ± 3.8%, p=0.001; HMGA2-OE: 73.4 ± 3.1, p=0.014; control for the tough decoy (TuD) vector: 51.6 ± 5.7%; let-7a-TuD: 90.3 ± 2.5%, p=0.001). As such, we hypothesized that the variable HbF effects in the adult cells may be due to independent or parallel effects from each component of the let-7 cascade since LIN28, IGF2BP1, IGF2BP3 and HMGA2 are expressed at very low levels or undetectable in the erythroid compartment after the fetal-to-adult transition.
For these studies, lentiviral transduction of human CD34+ cells were investigated in erythropoietin-supplemented serum-free cultures for 21 days and compared to donor-matched control transductions. Based upon its important role in globin gene regulation, mRNA and protein expression patterns of the transcription factor BCL11A were investigated by RT-qPCR and Western blot at day 14 as a mechanism for HbF regulation by the let-7 cascade components. BCL11A mRNA was significantly down-regulated after LIN28A-OE (control: 4.1.E+03 ± 8.2.E+02 copies/ng, LIN28A-OE: 1.7.E+03 ± 1.2.E+03 copies/ng, p=0.04) and let-7a -TuD (control: 1.7E+03 ± 4.5E+02 copies/ng, let-7a-TuD: 4.3E+02 ± 1.8E+02 copies/ng, p=0.003), while no significant differences were observed after IGF2BP1-OE (control: 5.42.E+03 ± 1.77.E+03 copies/ng, IGF2BP1-OE: 4.38.E+03 ± 2.57.E+02 copies/ng, p=0.375), IGF2BP3-OE (control: 5.64.E+02 ± 2.68.E+02 copies/ng, IGF2BP3-OE: 6.70.E+02 ± 3.47.E+02 copies/ng, p=0.694) and HMGA2-OE (control: 4.26E+02 ± 8.18E+01 copies/ng, HMGA2-OE: 2.84E+02 ± 1.48E+02 copies/ng, p=0.104). Interestingly, the protein levels of BCL11A were consistently down-regulated only by LIN28A-OE, let-7a -TuD and IGF2BP1-OE, with minimal changes observed after IGF2BP3-OE and HMGA2-OE.
To investigate which elements of the let-7 cascade are required for the fetal hemoglobin effects in adult cells, the expression levels of each component were compared with HbF effects after lentiviral-mediated transduction of LIN28A-OE, let-7a -TuD, IGF2BP1-OE, IGF2BP3-OE, or HMGA2-OE. RT-qPCR analyses were performed at day 14 and HPLC at day 21. HbF levels for the control transductions were less than 5%. Following LIN28A-OE (HbF: LIN28A-OE: 34.8 ± 2.7%, p=0.028), both IGF2BP1 and IGF2BP3 mRNA levels remained below the detection limits, while HMGA2 was slightly down-regulated. After let-7a -TuD (HbF: let-7a-TuD: 31.0 ± 5.0%, p=0.09), LIN28A and HMGA2 remained at background levels while IGF2BP1 and IGF2BP3 were below the detection limits. Following IGF2BP1-OE (HbF: IGF2BP1-OE: 64.6 ± 3.4%, p=0.02), LIN28A, IGF2BP3, and HMGA2 remained undetectable or at background levels. After IGF2BP3-OE (HbF: IGF2BP3-OE: 22.8 ± 4.4%, p=0.021), LIN28A, IGF2BP1 and HMGA2 also remained unchanged at background or undetectable levels. Finally, after HMGA2-OE (HbF: HMGA2-OE: 13.5 ± 1.5%, p=0.01), LIN28A, IGF2BP1 and IGF2BP3 remained below the detection limits. These results were confirmed by Western analysis, with the exception of HMGA2, which was consistently increased at the protein level following LIN28A-OE, let-7a -TuD, IGF2BP1-OE and IGF2BP3-OE.
In summary, these results demonstrate that several components the let-7 cascade share a downstream mechanism of BCL11A protein suppression in human erythroblasts. However, the variability of BCL11A and HbF effects as well as the ability of LIN28A-OE, let-7a-TuD and HMGA2-OE to increase HbF in the absence of IGF2BP1 and IGF2BP3 suggest both shared and independent mechanisms for hemoglobin regulation by the let-7 cascade during the human fetal-to-adult switch.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.