(A) (i) Sensitivity of BFU-E erythroid progenitors to EPO. Hypersensitive EPO response characterized by the increased in vitro growth of BFU-Es in the presence of low concentrations of EPO (0-30 mU/mL) was found in all patients with JAK2V617F (n = 3), EPORQ434X (n = 2), HIF2AM535V (n = 1), VHLR200W (n = 2), and LNKI257T (n = 1) mutations. In patients with JAK2V617F and EPORQ434X mutations, some BFU-E colonies also grew in the absence of EPO. Dashed line show the response of erythroid progenitors to EPO of tested healthy controls (n = 9). The number of BFU-Es grown in individual concentrations of EPO was expressed as a percentage of maximum vs the concentration of EPO. Results were pooled when n > 1. T bars designate standard deviations. (ii) Relative expression of RUNX1 and NF-E2 in hypersensitive BFU-E colonies. BFU-Es grown in low concentrations of EPO (15-30 mU/mL, ie, EPO-hypersensitive BFU-Es) were harvested and used for expression assay. Expression in BFU-Es was analyzed from patients with JAK2V617F (n = 3), VHLR200W (n = 2), VHLP138L (n = 1), HIF2AM535V (n = 1), LNKI257T (n = 1), EPORQ434X (n = 2), and EPOR5967insT (n = 2) mutations. The RUNX1 (Hs00257856) and NF-E2 (Hs00232351) TaqMan Gene Expression probes were used for quantitative real-time polymerase chain reaction. All samples were investigated in triplicate and normalized to expression of HPRT (4333768F) and GAPDH (4333764F) reference genes. The data were normalized to mRNA levels of healthy controls (black, n = 6), T bars designate SEM; *P < .05. The statistical significance of relative expression changes in target mRNA levels were analyzed for all expression analysis using REST 2009 software.20 (B) (i-ix) Relative expression of RUNX1, NF-E2, and HIF-regulated genes in granulocytes. Expression in granulocytes was analyzed from patients with JAK2V617F (n = 6), EPORQ434X (n = 2), HIF2AM535V (n = 1), HIF2AG537R (n = 1), VHLR200W (n = 2), LNKI257T (n = 1), VHLT124A/L188V (n = 2), and VHLH191D (n = 2) mutations and patients with secondary polycythemia (n = 6). The following TaqMan Gene Expression probes were used for quantitative real-time polymerase chain reaction: transferrin receptor (TFRC; Hs00951083), glucose transporter-1 (SLC2A1; Hs00892681), vascular endothelial growth factor (VEGF; Hs00900055), BNIP3 (Hs00969291), hexokinase-1 (HK1; Hs00175976), pyruvate dehydrogenase kinase, isozyme 1 (PDK1; Hs01561850), RUNX1 (Hs00231079), and NF-E2 (Hs00232351). All samples were investigated in triplicate and normalized to expression of HPRT (4333768F) and GAPDH (4333764F) reference genes. The data represents the mean of 3 independent experiments and were normalized to mRNA levels of healthy controls (black, n = 16); T bars designate SEM; *P < .05 and **P < .01. (x) Relative expression of Runx1 in Hif1α−/− yolk sacs and whole embryos. Expression in samples isolated from Hif1α−/− yolk sacs (n = 7) and whole embryos (n = 6) were analyzed using TaqMan Gene Expression probe for mouse Runx1 gene (Mm0123404). The data were normalized to expression of β-actin (Actb; 4352341E) and to mRNA levels of stage-matched, wild-type yolk sacs (black, n = 7) and whole embryos (black, n = 6); T bars designate SEM; *P < .05.
