The Interplay of GATA1 with Heme Regulates an Erythroid Cell's Differentiation

Abstract

GATA1 promotes the transcription of ALAS2, the first and rate limiting step of heme synthesis, and the transcription of many other erythroid-specific genes. It also increases its own transcription while silencing proliferation genes active in early progenitors and thus assures that erythroid differentiation correctly initiates. Heme then transcriptionally and translationally upregulates globin to guarantee adequate hemoglobin production in each cell as it matures. In mice lacking the heme exporter, FLVCR1, excess heme and ROS accumulate and erythropoiesis fails at the CFU-E/proerythroblast stage, resulting in a severe macrocytic anemia (HGB 4.4±0.97 vs 14.8±0.57 g/dL; MCV 66.9±6.2 vs 48.4±0.65 fL). To determine how excess heme causes ineffective erythropoiesis and whether heme is key to terminating differentiation in normal erythroid cells, we performed RNA sequencing of single early erythroid cells (BFU-E to basophilic erythroblasts) from wildtype control and Flvcr1-deleted mice and linked this transcription data to the total quantity of Ter119 on that cell. Principal component analysis (PCA) identified 4 transcriptionally unique clusters A, B, C, & D, which contained cells with negative, low, intermediate, and high Ter119 levels respectively. α- and β-globin transcription were highly correlated (r=0.975), occurred in all cells, increased as Ter119 expression increased, and upregulated in Flvcr1-deleted cells. Gene set enrichment analysis (GSEA) comparing control cells to Flvcr1-deleted cells revealed excess heme results in significant downregulation of the hallmark heme metabolism pathway genes (heme biosynthesis and erythroid differentiation genes), upregulation of the ribosome pathway genes, and no alteration of the P53 pathway genes. All eight heme biosynthetic enzyme genes were expressed equivalently in cluster A cells from control and Flvcr1-deleted mice; however expression in Flvcr1-deleted cells was significantly reduced in clusters B-D. Of the 181 erythroid differentiation genes in the hallmark heme pathway, Gata1 had the greatest reduction (67%) in Flvcr1-deleted cells. Coupled two-way clustering analysis (CTWC) identified 150 genes co-regulated with Gata1 including 106 known GATA1 target genes which were all poorly upregulated in Flvcr1-deleted cells in clusters B-D. Independent microarray analysis of mRNA from control and Flvcr1-deleted CD71+ erythroid cells confirmed low Gata1 mRNA and low GATA1-dependent gene expression in the Flvcr1-deleted cells. To determine if excess heme was directly responsible for Gata1 downregulation, we treated K562, HEL-R, and primary human erythroid marrow cells with aminolevulinic acid (ALA) and iron to increase endogenous heme synthesis. In the primary cells, GATA1 protein decreased by 30-43% (p=0.03) within 15 minutes and 66% by 90 minutes (similar decreases observed in cell lines), suggesting that heme disrupts GATA1 protein function resulting in the loss of autoregulation and reduced GATA1 mRNA. Of 88 genes in the ribosome pathway, 73 were significantly upregulated in Flvcr1-deleted cells, including 16 of the 17 ribosomal protein genes linked to Diamond-Blackfan anemia (DBA) or del(5q) myelodysplastic syndrome (MDS). When heme synthesis was induced in primary human erythroid marrow cells with ALA and iron, the transcription of ribosome protein genes such as Rps19, Rps14, and Rpl35 increased, further supporting the concept that heme assures sufficient ribosome production for globin protein synthesis. While P53 activation is a key factor in ineffective erythropoiesis caused by ribosomal protein imbalance (i.e., DBA and del(5q) MDS), GSEA did not reveal any increased activation of the P53 pathway in Flvcr1-deleted cells. To confirm that P53 was not involved in the ineffective erythropoiesis caused by excess heme, we generated mice lacking both P53 and FLVCR1. These double mutant mice had severe macrocytic anemia (HGB 2.4±0.70 g/dL; MCV 56.5±4.3 fL) comparable to mice lacking just FLVCR1. Thus, GATA1 turns on heme synthesis and initiates the erythroid differentiation program. GATA1 with heme assure each cell’s appropriate progression. Then heme turns off GATA1 to end differentiation. By linking excess heme to prematurely low GATA1, our data may also explain the ineffective (early termination of) erythropoiesis in DBA and reconcile the observations of Sci Transl Med 8:338ra67, 2016 and Nat Med 20:748, 2014. DisclosuresNo relevant conflicts of interest to declare.