Characteristic Macrocytic Anemia, Ineffective Erythropoiesis, and Iron Overload in 5q- Mice and Effect of Exogenous Transferrin

 

Low/intermediate risk myelodysplastic syndromes (MDS) is associated with relatively longer survival and high transfusion requirements, resulting in secondary iron overload. Iron overload is an independent variable associated with poor prognosis, impacting survival, and iron chelation therapy is associated with prolonged survival in transfusion-dependent MDS patients. We have previously shown that exogenous transferrin (Tf) results in more circulating red blood cells (RBCs), increased hemoglobin (Hb), reversal of splenomegaly, and improvement in ineffective erythropoiesis in β-thalassemic mice (Li Nat Med 2010) and recently demonstrate that erythroferrone (ERFE), a newly described physiologic erythroid regulator of hepcidin, is normalized in Tf-treated β-thalassemic mice (Chen, manuscript in review). We postulate that ineffective erythropoiesis in β-thalassemia shares many characteristics with that of low risk MDS, such as defective erythroid differentiation, anemia, splenomegaly, and systemic iron overload, suggesting that similar effects of exogenous Tf may be relevant also in MDS. Partial deletion of chromosome 5, 5q- syndrome, is the most common cytogenetic abnormality in low/intermediate risk MDS. We characterize a mouse model of 5q- syndrome, Cd74+/lox Nid67+/lox, Lmo2Cre+ mice (5q- mice) (Barlow Nat Med 2010). Our findings reveal that 5q- mice exhibit macrocytic anemia (Hb 6 vs. 12 g/dL, P<0.0001; MCV 61 vs. 48 fL, P<0.0001), splenomegaly (0.007 vs. 0.003 spleen/body weight, P<0.0001), extramedullary hematopoiesis in the liver (Fig 1A), expanded erythropoiesis in spleen (Fig 1B), and a lower percentage of erythroid precursors in the bone marrow (Fig 1C) despite increased serum erythropoietin (79 vs. 0.4 pg/L, P=0.04). Furthermore, 5q- mice exhibit evidence of iron overload relative to WT mice with increased Tf saturation (70 vs. 39%, P=0.004) and liver iron stores (1.0 vs. 0.3 mg iron/g liver weight, P=0.001). Although no difference in liver hepcidin expression or serum hepcidin concentration is evident (Fig 2A and 2B), bone marrow ERFE expression is increased (14-fold, P=0.002) and a trend toward decreased hepcidin:liver iron is observed in 5q- relative to WT mice (Fig 2C). We hypothesize that the effect of exogenous Tf applies to 5q- as β-thalassemic mice in light of similarities in disease characteristics. Using bone marrow transplantation to generate a cohort of 5q- mice with similar disease severity, 8x10^6 cells from 2 month old 5q- mice were transplanted into a cohort of sub-lethally irradiated 8 week old WT mice. Transplanted mice were allowed to recover and started Tf injections IP (10mg/day or equal volume PBS (200 uL) 2 weeks post-transplant.  After completing 20 days of injection, mice were sacrificed and all erythroid- and iron-related parameters analyzed. Tf-treated 5q- mice exhibit statistically increased RBC count (10 vs. 8 x10^6 cells/uL, P<0.0001) and lower MCV (41 vs. 53 fL, P<0.0001) and MCH (12 vs. 16 pg, P<0.0001) compared to PBS injected 5q- mice. Results demonstrate an increase in reticulocyte count (676 vs. 477 x 10^9 cells/L, P=0.003) and platelet count (549 vs. 364 x 10^3 cells/ul, P=0.02) in the circulation and an increase in orthochromatophilic erythroblasts (19.3 vs. 16.1%, P=0.04) and reticulocytes (32.8 vs. 26.6%, P=0.04) in the bone marrow, without changes in Erfe expression. Although TfR1 mRNA expression is unchanged, membrane TfR1 MFI is increased in late stage erythroid precursors in both bone marrow (11 vs. 7 x 10^3 MFI, P=0.01) and spleen (16 vs. 12 x 10^3 MFI, P<0.05) from Tf-treated relative to PBS injected 5q- mice. Interestingly, intracellular ROS is increased in late stage splenic erythroid precursors from Tf-treated relative to PBS injected 5q- mice. Although increased ROS levels result in cellular damage, emerging evidence suggests that ROS is required for normal hematopoiesis, such that high ROS states are associated with differentiating HSCs. Taken together, our data demonstrates erythropoiesis- and iron-related characteristics of 5q- mice consistent with human disease. Although these preliminary experiments using exogenous Tf reveal no increase in Hb, improved ineffective erythropoiesis, or reversal of iron overload, our data suggests that TfR1 and ROS are intimately involved in the effect of exogenous Tf in erythropoiesis.