In vivo regulation of erythropoiesis by chemically inducible dimerization of the erythropoietin receptor intracellular domain

Abstract

Erythropoietin (Epo) and its receptor (EpoR) are required for the regulation of erythropoiesis. Epo binds to the EpoR homodimer on the surface of erythroid progenitors and erythroblasts, and positions the intracellular domains of the homodimer to be in close proximity with each other. This conformational change is sufficient for the initiation of Epo-EpoR signal transduction. Here, we established a system of chemically regulated erythropoiesis in transgenic mice expressing a modified EpoR intracellular domain (amino acids 247-406) in which dimerization is induced using a specific compound (chemical inducer of dimerization, CID). Erythropoiesis is reversibly induced by oral administration of the CID to the transgenic mice. Because transgene expression is limited to hematopoietic cells by the Gata1 gene regulatory region, the effect of the CID is limited to erythropoiesis without adverse effects. Additionally, we show that the 160 amino acid sequence is the minimal essential domain of EpoR for intracellular signaling of chemically inducible erythropoiesis in vivo. We propose that the CID-dependent dimerization system combined with the EpoR intracellular domain and the Gata1 gene regulatory region generates a novel peroral strategy for the treatment of anemia.

Fig 1. Establishment of G1HRD-idEpoRic Transgenic Mouse Lines.

(A) Schematic structure of the transgene product, idEpoRic (inducible dimerization of EpoR intracellular domain). The myristoylation signal sequence (Myr) is anchored to the amino terminal of idEpoRic for membrane association with myristate. The EpoR intracellular domain (ic, amino acids 247–406), which contains 2 tyrosine residues (Y), was fused with GFP to monitor the transgene expression. Two copies of the CID binding domain (DmrB) were inserted. The terms ec and tm indicate the extracellular domain and transmembrane domain of EpoR, respectively. (B) Sequence alignment of the mouse (amino acids 247–406) and human EpoR sequences (amino acids 248–407). Identical amino acids between the mouse (M) and human (H) EpoRs are indicated by asterisks (*). Important motifs for EpoR signaling (JM, juxtamembrane domain; Box1 and Box2, highly conserved sequence among cytokine receptors) and the phosphorylated tyrosine residues (Y343 and Y401) on mouse EpoR are illustrated. (C) Semi-quantitative RT-PCR analysis of the transgene expression in the spleens of 3 independent transgenic mouse lines. Expression of the G1HRD-idEpoRic transgene was detected by 27, 30, 33, 36, and 39 PCR cycles (left to right). HPRT was used as an internal control. (D) Membrane-associated localization of idEpoRic was observed in the spleen section of the line A transgenic mouse using laser confocal microscopy (left). A spleen section of a G1HRD-GFP transgenic mouse, which expresses GFP throughout the whole cell, is also shown as a control (right). The scale bar indicates 10 μm.

Fig 2. CID Dose- and Transgene Expression Level-dependent Colony Formation of G1HRD-idEpoRic Bone Marrow Cells.

(A) Increased CFU-E—derived colony formation in response to incremental doses of CID or rHuEPO. Bone marrow MNCs from line A transgenic mice were used. (B) The bone marrow MNCs of 3 independent transgenic mouse lines were cultured in methylcellulose medium supplemented with CID (1.0 nmol/mL) or rHuEPO (2.0 U/mL) for 3 days, and the numbers of CFU-E—derived colonies were counted. Note that the numbers of CFU-E—derived colonies correlate tightly with the expression levels of the transgene (line A > line B > line C, see Fig. 1C). These assays were performed in triplicate and repeated 3 times. The results are shown as the mean ± standard deviations.

Fig 3. Changes in idEpoRic Localization and STAT5 Phosphorylation After CID Stimulation in Fetal Liver Hematopoietic Cells.

(A) Flowcytometry of the fetal liver cells from wild-type (WT) or line A (Tg) mouse embryos at E12.5. Ter119 and c-Kit expressions levels were analyzed by GFP (idEpoRic) expression, and the percentages of cells in each quadrangle are shown (blue letters in dot plot data). The GFP⁺ fraction was analyzed using an anti-Ter119 antibody (green) or an isotype control antibody (as a negative control [NC], black line) in the top histogram. Cells positive for Ter119 or c-Kit have been analyzed by GFP expression, and black-line histograms show the data from wild-type littermates (middle and bottom histograms). The percentages of cells in the gated regions are shown in each panel. (B) The subcellular localization of idEpoRic in the E12.5 fetal liver was examined using laser confocal microscopy. Single cell suspensions from wild-type (WT) and transgenic (Tg) fetal livers at E12.5 were incubated with (bottom) or without (top and middle) 1 nmol/mL CID for 15 minutes, and the cells were observed after nuclear staining with Hoechst33342 (blue). Note the membrane-associated localization of GFP (green) in the transgenic cells (middle) and the internalization of the GFP⁺ clusters after CID stimulation (arrowheads). A GFP-negative cell (asterisk) is found in a preparation of the transgenic cells. The scale bar indicates 10 μm. (C) Intracellular phosphorylated STAT5 (p-STAT5) in E12.5 fetal liver cells was detected using PE-conjugated anti-phosphorylated STAT5(Y694)-specific antibodies. Cells from line A fetal liver were exposed to rHuEPO (2.0 U/mL) or CID (1.0 nmol/mL) for 0 (Pre), 15, or 180 minutes. The left panels show basal levels of p-STAT5 in the transgenic (black lines) and wild-type (WT, blue) cells compared with the isotype antibody controls (NC, gray).

Fig 4. Dose-dependent, Reversible, and Erythroid Lineage-Specific Effect of CID on In Vivo Erythropoiesis.

(A) Changes in the reticulocyte counts in the peripheral blood of the line A transgenic mice by intraperitoneal injection of CID (10.0, 1.0, or 0.1 mg/kg) or vehicle every day for 9 days (arrowheads). Reversible and dose-dependent induction of the reticulocyte counts by CID administration was observed. (B) The levels of reticulocytes, hematocrit, platelets, and white blood cells (WBC) in the peripheral blood of line A transgenic mice were counted before (Day0) and after (Day10) daily intraperitoneal injections of CID (10 mg/kg), rHuEPO (300 U/kg), or vehicle for 9 days (Days 1–9). These assays were performed in triplicate (3 mice for each group) and repeated 3 times. The results are shown as the mean ± standard deviations. *p < 0.01 compared with Day0 (hatched bars).

Fig 5. Enlargement of the Spleen Red Pulp by the CID-idEpoRic System.

(A) Gross appearance of enlarged spleens after the injection of the vehicle, CID (10 mg/kg), or rHuEPO (300 U/kg) for 9 consecutive days. (B–G) Hematoxylin and eosin staining of the spleen sections from vehicle- (B, E), CID- (C, F), or rHuEPO-treated (D, G) line A transgenic mice. E, F, and G show higher magnification views. The enucleated red blood cells were observed in the spleen red pulps of CID- and rHuEPO-treated mice (F, G). (H–J) The spleen sections from vehicle- (H), CID- (I), or rHuEPO-treated (J) mice were stained with an anti-ß-globin antibody. ß-globin-positive (brown signals) erythroid cells were expanded in the spleens of CID- (I) or rHuEPO-treated (J) transgenic mice. The scale bars indicate 1.0 cm (A), 200 μm (B–D, H–J), and 80 μm (E–G).

Fig 6. CID Stimulates the Growth of Erythroid Progenitors and Erythroblasts in the Spleens of Transgenic Mice.

(A) CD71 and Ter119 were analyzed in splenic MNCs of line A transgenic mice by flowcytometry (upper panels) after the injection of vehicle, CID (10 mg/kg), or rHuEPO (300 U/kg) for 9 consecutive days. The erythroblastic fraction was separated into 4 maturation stages based on CD71 and Ter119 expression, and the percentages of cells in each region are represented in the graph. (B) GFP expression in splenic MNCs after the administration of CID or rHuEPO by the same strategy described for (A). CID exclusively induced GFP⁺ transgene-expressing cells, whereas a specific increase in GFP⁺ cells was not observed following rHuEPO administration. (C) The numbers of CFU-E and BFU-E erythroid progenitors were examined in the splenic MNCs of line A transgenic mice after the administration of vehicle (open bars), CID (gray bars), or rHuEPO (solid bars) by the same strategy described for (A). The results are shown as the mean ± standard deviations. **p < 0.05, *p < 0.01 compared with the vehicle-treated mice.

Fig 7. Oral Administration of CID Stimulates Erythropoiesis.

(A) Reticulocytes in the peripheral blood of wild-type (WT) and transgenic (Tg) mice were counted before (Day0) and after (Day11) daily oral administration of CID (100 mg/kg) or vehicle for 10 days (Day1–Day10). (B) The numbers of CFU-E and BFU-E erythroid progenitors were also examined in splenic MNCs on Day11. Open bars indicate CID-treated wild-type mice (WT-CID). Gray and solid bars indicate CID- and vehicle-treated transgenic mice, respectively (Tg-CID and Tg-Veh). These assays were performed in triplicate with 4 mice for each group. The results are shown as the mean ± standard deviations.