Ineffective erythropoiesis with reduced red blood cell survival in serotonin-deficient mice

Abstract

Serotonin (5-HT) has long been recognized as a neurotransmitter in the central nervous system, where it modulates a variety of behavioral functions. Availability of 5-HT depends on the expression of the enzyme tryptophan hydroxylase (TPH), and the recent discovery of a dual system for 5-HT synthesis in the brain (TPH2) and periphery (TPH1) has renewed interest in studying the potential functions played by 5-HT in nonnervous tissues. Moreover, characterization of the TPH1 knockout mouse model (TPH1(-/-)) led to the identification of unsuspected roles for peripheral 5-HT, revealing the importance of this monoamine in regulating key physiological functions outside the brain. Here, we present in vivo data showing that mice deficient in peripheral 5-HT display morphological and cellular features of ineffective erythropoiesis. The central event occurs in the bone marrow where the absence of 5-HT hampers progression of erythroid precursors expressing 5-HT(2A) and 5-HT(2B) receptors toward terminal differentiation. In addition, red blood cells from 5-HT-deficient mice are more sensitive to macrophage phagocytosis and have a shortened in vivo half-life. The combination of these two defects causes TPH1(-/-) animals to develop a phenotype of macrocytic anemia. Direct evidence for a 5-HT effect on erythroid precursors is provided by supplementation of the culture medium with 5-HT that increases the proliferative capacity of both 5-HT-deficient and normal cells. Our thorough analysis of TPH1(-/-) mice provides a unique model of morphological and functional aberrations of erythropoiesis and identifies 5-HT as a key factor for red blood cell production and survival.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.

TPH1−/− mice display a phenotype of aregenerative macrocytic anemia. (A–C) A significant decrease in the number of RBCs (7.5 ± 0.2 vs. 8.8 ± 0.1 × 106 cells/μL), hematocrit (41.5 ± 0.8 vs. 45.0 ± 0.7%), and hemoglobin levels (15.1 ± 0.3 vs. 16.6 ± 0.2 g/dL) is seen in TPH1−/− vs. WT mice and indicates an anemic phenotype. (D and E) A significant increase in mean corpuscular volume (MCV) (55.0 ± 0.3 vs. 51.2 ± 0.2 fl), associated with a general increase in RBC cell size, is observed in TPH1−/− vs. WT mice and is consistent with macrocytosis. (F) Mean corpuscular hemoglobin (MCH) is increased significantly (20.2 ± 0.2 vs. 18.8 ± 0.1 pg), but (G) no difference is seen in mean corpuscular hemoglobin concentration (MCHC) (36.7 ± 0.2 vs. 36.8 ± 0.3 g/dL) in TPH1−/− vs. WT mice. (H) A significant increase in EPO level is observed in plasma of TPH1−/− mice (541 ± 81 pg/mL, n = 30) compared with WT mice (229 ± 49 pg/mL, n = 17) and indicates a response to the anemia. (I) No reticulocytosis is observed in TPH1−/− mice (184,000 ± 20,000 reticulocytes/mL, n = 23) compared with WT mice (164,000 ± 18,000 reticulocytes/mL, n = 17) and points to aregenerative anemia. P was calculated using an unpaired t test in G and I, an unpaired t test with Welch's correction in B and H, and a Mann–Whitney test in A, C, D, and F. P < 0.05 was considered significant. In A–D, F, and G each dot represents a single blood count (WT, n = 42; TPH1−/−, n = 56; **P < 0.01; ***P < 0.001).

Fig. 2.

Decreased RBC survival and increased splenic iron deposits in TPH1−/− mice. (A) Flow cytometry analysis shows that a higher proportion of TPH1−/− RBCs binds annexin V and thus exposes PS at the surface. (B) TPH1−/− RBCs also have an increased sensitivity to osmotic shock because incubation in a hypertonic solution exacerbates the difference in annexin V binding between WT and TPH1−/− RBCs. (C) TPH1−/− RBCs show a marked susceptibility to in vitro phagocytosis compatible with previous studies reporting the enhanced clearance of PS-exposing cells by macrophages. (D) RBCs obtained from TPH1−/− mice show a significantly reduced half-life (10.3 ± 0.6 d, n = 12) compared with WT controls (14.4 ± 0.4 d, n = 12). (E) Decreased half-life of TPH1−/− RBCs could result from their ingestion by splenic macrophages because iron deposits are visualized by Perls’ staining. (F) Howell–Jolly bodies (arrowheads) also are seen on RBC smears of TPH1−/− mice, suggesting improper enucleation. P was calculated using an unpaired t test in B and C and an unpaired t test with Welch's correction in A. *P < 0.05; **P < 0.01.

Fig. 3.

Decreased number of erythroid precursors in the BM of TPH1−/− mice. (A) (Left and Center) Comparative analysis of femoral histology sections shows that TPH1−/− mice have a decreased hematopoietic cellularity in the BM. (Right) Total cell counts of BM cells confirm this observation, given that femur and tibia of TPH1−/− mice contain significantly fewer nucleated cells (13.2 ± 0.9 × 106 cells, n = 21) than those of WT mice (18.6 ± 0.9 × 106 cells, n = 16). (B) The frequency of the erythroid progenitor BFU-E is increased significantly in 5-HT–deficient mice (Left), but the total number of CFU-E is maintained (Right), suggesting a compensatory response to the anemia. (WT, n = 6; TPH1−/− mice, n = 6.) (C) No difference in the profile of Ter119+CD71+ cells (erythroblast population) is observed between TPH1−/− and WT cells isolated from the BM (Left and Center), but the total number of Ter119+CD71+ cells is significantly decreased in TPH1−/− compared with WT mice (Right) (3.5 ± 0.2 × 106 cells, n = 24, vs. 5.3 ± 0.3 × 106 cells, n = 20). P was calculated using an unpaired t test in A and B and a Mann–Whitney test in C. P < 0.05 was considered significant. **P < 0.01; ***P < 0.001.

Fig. 4.

Abnormal proliferation of erythroid precursors in TPH1−/− mice. WT and TPH1−/− mice were injected with BrdU, and Ter119+CD71+ cells were isolated from BM and analyzed by flow cytometry. (A) A typical flow cytometry profile reveals no difference in the DNA content, but (B) the percentage of cells incorporating BrdU is reduced significantly in 5-HT–deficient vs. WT mice (19.3 ± 1.4%, n = 11, vs. 24.7 ± 1.4%, n = 11). (C) BM aspirate smear from a TPH1−/− mouse showing binucleated erythroblasts (arrow) indicative of a cell-division defect. In B, P was calculated using a Mann–Whitney test. *P < 0.05.

Fig. 5.

Proliferation defect in TPH1−/− erythroid precursors rescued by 5-HT supplementation. Proerythroblasts obtained from WT (open square) or TPH1−/− (open circle) BM were induced to differentiate in vitro. Proliferative capacity was measured at 24, 48, and 72 h following the induction of proerythroblasts into a differentiation medium containing EPO and transferrin only (dotted line) or supplemented with 5-HT (solid line). (A) The significant decrease in proliferative capacity observed in 5-HT–deficient erythroid precursors was rescued by 5-HT supplementation (filled circles). Supplementation with 5-HT also increased the proliferative capacity of WT precursors (filled squares). (B) The significant increase in proliferative capacity when 5-HT is added to the differentiation medium (filled squares) also is observed when PNU 22394, a 5-HT2 agonist, is added to the medium (filled diamonds). Results are representative of two independent experiments (n = 6).