Excess activation of EPO receptor (EPOR) signaling leads to erythrocytosis, which is defined as an absolute increase of red cell mass. EPO-independent and erythrocytes-intrinsic EPOR activation is induced by gain-of-function EPOR or JAK2 mutations, resulting in "primary" erythrocytosis. Erythrocytes-extrinsic EPOR activation is called "secondary" erythrocytosis, which is caused by EPOR stimulation with high level of EPO.

In order to investigate primary and secondary erythrocytosis in detail, we generated two unique mouse models. LysM -Cre; Jak2 V617F knock-in heterozygous mice express constitutively active JAK2 in LysM-expressing cells, which include a small portion of HSCs and their downstream progenitors. We also generated a Alb -Cre; Rosa26 LoxP-Stop-LoxP-rtTA (LSL-rtTA);human EPAS1 double-point-mutant (DPM) mice (Alb -Cre;DPM). EPAS1 is a master regulator of EPO gene. The EPAS1-DPM has mutations in P531 and N847 sites, which are critical sites for stability and transcriptional inactivation in normoxia. This mouse expresses the DPM mutant (P531A and N847A) specifically in hepatocytes in a doxycycline-inducible manner, resulting in high EPO production in the liver. We used LysM -Cre; Jak2 V617F as an erythrocytes-intrinsic "primary" erythrocytosis model, and Alb -Cre;DPM as an erythrocytes-extrinsic "secondary" erythrocytosis model. Interestingly, both of them developed comparable degree of erythrocytosis quickly. The plasma EPO level in LysM -Cre; Jak2 V617F was significantly decreased, while that in Alb -Cre/DPM was significantly increased in comparison to the control and also to LysM -Cre; Jak2 V617F. They both exhibited splenomegaly, but spleens of Lys M-Cre; Jak2 V617F mice were significantly larger than those of Alb -Cre;DPM mice. In Alb -Cre;DPM mice, CD45- CD71+ erythroblasts were detected in the liver, whereas there were almost no erythroblasts in the liver of the control mice, suggesting that extremely high level of EPO expression in the liver causes erythropoiesis in the liver of Alb -Cre;DPM mice.

Increased monocyte count has been reported in patients with erythrocytosis. Consistent with the observation in the patients, total CD11b+ CD115+ monocytes were significantly increased in PB from both LysM -Cre; Jak2 V617F and Alb -Cre;DPM mice compared to those in PB from the wild-type mice. It is also well known that macrophages play a crucial role in erythropoiesis, supplying the iron and engulfing the protruded nuclei from erythroblasts in the terminal maturation process. The complex of erythroblasts and macrophages directly interacting each other is called as "erythroblastic island (EBI)". Thus, we next analyzed macrophages in bone marrow (BM) and spleen (SP) in our erythrocytosis models. Macrophages in SP were dramatically increased in the both models, while those in BM were not significantly altered. The percentage of EBI (multiplet F4/80+ Ter119+) in BM and SP were increased in both models.

EPOR expression on the cell surface is required for EPO-EPOR signaling. To confirm the expression of EPOR by genetic approach, we combined EPOR -Cre knock-in allele and Rosa26LSL-GFP reporter allele to generate EPOR -Cre;GFP reporter mice. In this model, pro-erythroblasts and basophilic erythroblasts in BM expressed GFP as expected. We did not find GFP expression on the Macrophage and dendritic cell precursors (MDPs), common monocyte progenitors (cMoPs), mature monocytes in BM and PB monocytes. In a single-cell-gated macrophage population, we found a small fraction expressing GFP. However, FACS sorting revealed that those GFP+ cells are not macrophages alone, but macrophages with erythroblasts adhered to them. Even after adding EDTA, some of those macrophages still retain bound erythroblasts. It is possible that they could acquire activated EPOR signaling from erythroblasts within EBI through direct interaction. We are currently investigating the mechanisms for expansion of monocytes/macrophages in erythrocytosis.

Collectively, here we demonstrate that both Jak2V617F and high EPO could activate EPOR signaling and expand erythroblasts and monocytes/macrophages, which might contribute to erythrocytosis phenotype in our primary and secondary erythrocytosis models. Further exploration of the roles of monocytes/macrophages in erythrocytosis could lead to development of novel therapeutic strategy for erythrocytosis by modulating their functions in EBI.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.