Rap-536 (Murine ACE-536/Luspatercept) Inhibits Smad2/3 Signaling and Promotes Erythroid Differentiation By Restoring GATA-1 Function in Murine b-Thalassemia

We have previously reported that Smad2/3 signaling (of the TGFβ superfamily) is elevated in myelodysplastic syndromes (MDS) and β-thalassemia, diseases that are characterized by ineffective erythropoiesis (Suragani et al. 2014). Smad2/3 pathway inhibition using RAP-536 (murine version of ACE-536/luspatercept), a modified activin receptor type IIB ligand trap, decreased ineffective erythropoiesis (IE) and alleviated disease pathology in a murine model of β-thalassemia. In this study, we investigated the a) potential role of different Smad2/3 ligands that bind to luspatercept in the regulation of erythropoiesis and b) molecular mechanism of RAP-536 therapy in the murine model of β-thalassemia.

Wild-type (WT) mice were treated with neutralizing antibodies against activin B, GDF8 or GDF8/11 (10mg/kg, s.c, twice weekly for 2- weeks, N=5/group) either as a single agent or in combination, and compared with RAP-536 (10 mg/kg, s.c) treatment. β-thalassemic mice (Hbb^th3/+) were administered a single bolus of vehicle (VEH) or RAP-536 (30 mg/kg, i.p) (N=2/group). At 16 hours following administration the splenic basophilic erythroblasts (CD71⁺ Ter119⁺ FSC^high) were sorted by flow cytometry and RNA was isolated and subjected to genome wide transcriptome profiling using RNA sequencing analysis.

a) Surface plasmon analysis revealed that ACE-536 binds Smad2/3 signaling ligands GDF11 and GDF8 with high affinity and activin B with lower affinity. There was minimal binding detected with Activin A, TGFβ1 or TGFβ3 ligands. Wt mice treated with RAP-536 increased RBC (+19%, P<0.001) and Hgb (+15.2%, P<0.001) compared to VEH treated mice. Treatment with anti-GDF8 or anti-activin B antibodies marginally affected RBC parameters (~2-4%, N.S) where as anti-GDF8/11 treatment alone increased RBC (+6.1%, P<0.05) and Hgb (6.9%, P<0.05) compared to VEH treatment. A combination treatment of anti-GDF8/11 and activin B antibodies synergistically increased RBC (10.7%, P<0.001) and Hgb (11%, P<0.001) compared to VEH treated mice. These data suggests that in addition to GDF11 and activin B, other TGFβ superfamily ligands are possibly involved in the stimulation of erythropoiesis by luspatercept.

b) Transcriptome analysis of β-thalassemic erythroblasts revealed a total of 74 genes that were differentially expressed (absolute fold change >1.5, false discovery rate adjusted P value <0.05) in RAP-536 treated samples compared to VEH treatment. To identify molecular mechanisms, we performed gene set enrichment analysis (GSEA) (Subramanian et al., 2005) on data from RAP-536 and VEH treated samples. The analysis depicted significant upregulation of target genes of multiple transcriptional regulators including GATA-1 (erythroid differentiation), NFE2 and heat shock factor (involved in globin expression and protein quality-control). Previously, multiple studies established GATA-1 as a master transcriptional regulator of terminal erythroid differentiation. The individual gene symbols based comparative analysis revealed up-regulation of 53/478 GATA-1 activators and down regulation of 9/342 GATA-1 repressors. The GATA-1 target genes that were up regulated by RAP-536 treatment are involved in heme biosynthesis (such as Ppox, Fech, Alas2 and Abcb10) and erythroid differentiation (such as Klf1, Nfe2, Gypa, Bcl2l, Bnip3l, Bach1, and Ank1). Further GSEA of GATA-1 activator and repressor signatures against RAP-536 treatment data revealed a significant up-regulation of 158/328 activated genes (Normalized Enrichment Score=2.7, P=0) involved in heme biosynthesis, and cell cycle regulation whereas there was no statistically significant down regulation of GATA-1 repressed genes. Consistent with this data, our preliminary results in differentiating mouse erythroleukemic (MEL) cells showed increased Smad2/3 phosphorylation that is correlated with reduced GATA-1 protein levels suggesting that pSmad2/3 may negatively regulate terminal erythroid differentiation by decreasing GATA-1 availability.

These data provide a potential mechanistic role for luspatercept treatment in β-thalassemia, by transcriptionally upregulating genes that promote erythroid differentiation and processing of unpaired α-globins. By inhibiting SMAD2/3 signaling, luspatercept relieves the block of terminal erythroid maturation and decreases ineffective erythropoiesis in diseases such as β-thalassemia and MDS.