GDF11 is not the target of luspatercept

Clara Camaschella

In this issue of Blood, Guerra et al have excluded that growth differentiation factor 11 (GDF11) is the target of luspatercept, a novel drug that promotes differentiation of terminal erythropoiesis and is under evaluation for treatment of β-thalassemia and low-risk myelodysplastic syndromes (MDSs).1

Ineffective erythropoiesis is the pathogenic hallmark of β-thalassemia. Correcting ineffective erythropoiesis might improve anemia and attenuate iron overload and related complications, such as hepatic and cardiac disease, bone deformities, thrombosis, and pulmonary hypertension.2 Current experimental treatments of thalassemia include gene replacement and editing by gene therapy, erythropoiesis iron restriction by hepcidin manipulation, and pharmacologic fetal hemoglobin (Hb) induction. In MDS, options to treat ineffective erythropoiesis are limited to erythropoietin.

Luspatercept (ACE-536) is the first pharmacologic compound that, in phase 2/3 clinical trials, improves erythropoiesis, increases Hb levels, and reduces red blood cell (RBC) transfusion requirements in both β-thalassemia3,4 and in patients with low-risk MDS.5 ACE-536 is the modified extracellular domain of activin receptor type IIB (ACVR2B) fused to the Fc fragment of human immunoglobulin G1 (IgG1).6 This artificial trap, as well as the analogous protein sotatercept (the activin receptor IIA extracellular domain fused to the Fc of IgG1) competes with the physiological receptor of erythroid cells for binding members of the transforming growth factor β (TGF-β) superfamily.7 As for the mechanism, studies in thalassemia mice (Hbbth1/th1) showed that the murine version of the trap (RAP-536) targets GDF11, identifying this TGF-β family member as a key inhibitor of late-stage erythroid differentiation through increased oxidative stress.6,7 GDF11 sequestration by RAP-536 restores erythroid maturation and allows more effective and erythropoietin-independent RBC production in β-thalassemia mice.

By using different genetic tools, Guerra et al clearly excluded GDF11 as the main target of RAP-536 in murine erythropoiesis. First, they examined and disproved the proposed inhibitory effect of GDF11 on terminal erythropoiesis. Indeed, specific deletions of Gdf11, induced either in erythroid cells by Cre recombinase expressed under the control of EpoR promoter or in all bone marrow lineages, with Cre recombinase placed under the Vav promoter control did not improve anemia in β-thalassemic (Hbbth3/+) mice. If GDF11 were the key inhibitor of late erythroid maturation, an increase in Hb would have been observed in these transgenic mice.

To rule out the possibility of a persistent synthesis of GDF11 by stromal or other cells after bone marrow or erythroid Gdf11 deletion, Guerra et al generated thalassemic mice and controls with Gdf11 deletion in the whole mouse. To overcome the embryonic lethality of the germinal deletion, they developed tamoxifen-inducible RosaCre thalassemic and wild-type models. Gdf11 deletion induced by tamoxifen treatment did not improve erythropoiesis or anemia in transgenic animals followed for up to 6 months. Similarly, no benefits were obtained by deleting Gdf11 in the bone marrow of an MDS murine model.6 The results of these studies cast doubt on the proposed inhibitory role played by GDF11 in late erythroid maturation.

Next, the authors addressed the efficacy of RAP-536 treatment in wild-type and thalassemic models transgenic for Gdf11 and observed improved Hb, RBC, and hematocrit in both thalassemic and wild-type mice in the absence of GDF11 in bone marrow lineages. This clearly excludes GDF11 as the relevant target of RAP-536 and as a major player in the process of ineffective erythropoiesis.

Finally, the expression levels of GDF11 and its receptor ACVR2B were low in erythroid liquid cultures derived from CD34+ cells of healthy donors. The same low expression was observed in Ter119+ cells from spleens of both wild-type and β-thalassemic mice also after RAP-536 treatment. Altogether, the results of Guerra et al exclude that RAP-536 acts by trapping GDF11 and that the latter is the only (or the main) effector of TGF-β inhibition of late erythropoiesis in mice.

Unfortunately the absence of reliable anti-GDF11 antibodies prevents protein analysis, and the results of the study are based on Gdf11 expression levels. Although valuable, the study does not identify the real targets of RAP-536, so the race to discover the mechanism of erythropoiesis correction is still on. It is possible that identifying the specific mechanism will prove complex, considering the pleiotropic functions and the promiscuity of the ligand-receptor interactions of the TGF-β family. Identifying the culprit of this phenomenon would not only advance our knowledge of erythropoiesis control, but might also lead to the discovery of more novel therapeutic targets.

Luspatercept partially corrects anemia in patients who are not transfusion dependent3,5 and, although it has no direct effect on iron metabolism, over time the correction of ineffective erythropoiesis increases hepcidin levels and reduces iron burden.8 The ability of luspatercept to decrease transfusion requirements in both β-thalassemia and MDS4,5 reflects its effect on residual endogenous abnormal erythropoiesis. These observations might extend the potential of such treatment to all iron-loading anemias that are a result of ineffective erythropoiesis,8 including those in transfusion-dependent patients. In addition, because luspatercept works independently of erythropoietin, a potential combination therapy might be envisaged. It is also worth noting that normal women treated with sotatercept to improve postmenopausal osteoporosis show increased Hb levels.9 Hence, it is likely that activin receptor ligand traps interfere with the physiological mechanisms that normally limit the number of mature erythroblasts; this would extend their potential use to other anemias that are the result of defective RBC production. As often occurs in medical practice, luspatercept might provide therapeutic benefits to patients even before its mode of action is fully elucidated.


  • Conflict-of-interest disclosure: C.C. has received honoraria from Vifor Pharma and is a member of the advisory boards for Celgene and Novartis.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.