Negative autoregulation by Fas stabilizes adult erythropoiesis and accelerates its stress response

Abstract

Erythropoiesis maintains a stable hematocrit and tissue oxygenation in the basal state, while mounting a stress response that accelerates red cell production in anemia, blood loss or high altitude. Thus, tissue hypoxia increases secretion of the hormone erythropoietin (Epo), stimulating an increase in erythroid progenitors and erythropoietic rate. Several cell divisions must elapse, however, before Epo-responsive progenitors mature into red cells. This inherent delay is expected to reduce the stability of erythropoiesis and to slow its response to stress. Here we identify a mechanism that helps to offset these effects. We recently showed that splenic early erythroblasts, 'EryA', negatively regulate their own survival by co-expressing the death receptor Fas, and its ligand, FasL. Here we studied mice mutant for either Fas or FasL, bred onto an immune-deficient background, in order to avoid an autoimmune syndrome associated with Fas deficiency. Mutant mice had a higher hematocrit, lower serum Epo, and an increased number of splenic erythroid progenitors, suggesting that Fas negatively regulates erythropoiesis at the level of the whole animal. In addition, Fas-mediated autoregulation stabilizes the size of the splenic early erythroblast pool, since mutant mice had a significantly more variable EryA pool than matched control mice. Unexpectedly, in spite of the loss of a negative regulator, the expansion of EryA and ProE progenitors in response to high Epo in vivo, as well as the increase in erythropoietic rate in mice injected with Epo or placed in a hypoxic environment, lagged significantly in the mutant mice. This suggests that Fas-mediated autoregulation accelerates the erythropoietic response to stress. Therefore, Fas-mediated negative autoregulation within splenic erythropoietic tissue optimizes key dynamic features in the operation of the erythropoietic network as a whole, helping to maintain erythroid homeostasis in the basal state, while accelerating the stress response.

Figure 1. Epo regulation of erythropoiesis through Fas-mediated apoptosis.

(A) Epo-dependent erythroblastic island precursors CFU-e, ProE and EryA (in blue) co-express Fas and FasL, and mature into Epo-independent EryB, EryC and red blood cells (RBC, in red). ‘F’  = Fas expressing cells, shown undergoing cell death as a result of interaction with FasL-expressing cells within the Epo- dependent (blue) compartment (black flat-headed arrow). A negative feedback loop driven by tissue pO₂ regulates Epo levels in blood, which in turn enhance erythroblast survival, by either suppressing Fas and FasL expression, or by non-Fas dependent pathways. HSC  = hematopoietic stem cells. (B) Flow-cytometric identification of Epo-dependent ProE and EryA subsets (in blue) and Epo-independent EryB and EryC (in red), in adult Balb/C mouse spleen, in basal conditions (top panels) or 48 hours following Epo injection (300 U/ mouse, lower panels). ProE are defined as Ter119^medCD71^high cells; Ter119^high cells are further subdivided based on forward scatter (FSC) and CD71 expression into EryA (CD71^highTer119^highFSC^high), EryB (CD71^highTer119^highFSC^low) and EryC (CD71^lowTer119^highFSC^low). (C) The erythropoietic response of mice to a hypoxic environment. Mice (Balb/C) were examined either in the basal state (‘a’, 21% atmospheric oxygen), when housed in 11% oxygen for 8 days (assay times ‘b’ and ‘c’ at 13 hours and 3 days, respectively), and when placed back in normoxia (21%; assay times ‘d’ and ‘e’, at 1 and 2 days post-hypoxia). Top panels show representative flow-cytometric histograms of Fas expression and Annexin V binding at the indicated assay times. Gates refer to the Fas⁺ and Annexin V⁺ populations, determined with reference to staining controls in which either the anti-Fas antibody (left panel) or Annexin V (right panel) were omitted. The fraction of cells positive for Fas or Annexin V at each time point is noted. Middle panel shows a summary of similar data, 2 to 9 mice per time point. Lower panels show corresponding serum Epo levels and EryA cell number in spleen (expressed as total EryA cells /gram body weight). *p<0.05, **p<0.002, ***p<0.0001 (two-tailed t test, unequal variance).

Figure 2. Fas and FasL-mediated negative autoregulation of the EryA pool.

(A) Wild-type Balb/C mice were injected with a single Epo injection subcutaneously, of 1, 3, 6, 10, 30, 100 or 300 U/ 25 g body weight. Spleen EryA were examined on day 3 post injection. Left panel shows EryA cell frequency relative to basal frequency, plotted against the number of EryA cells that express Fas (Fas⁺ EryA, expressed as a ratio to basal levels). Data points represent individual mice. Blue  = mice in the basal state (n = 15), red  = mice injected with Epo (n = 38). For clarity only mice injected with 30 U (which maximally suppress Fas) or less are included in the left panel. Data is fitted with a curve derived from the mathematical model described in panel (B) and in the Text S2. Right panel shows the dependence of Fas⁺ EryA on the dose of injected Epo, in the same dataset as in the left panel; all mice injected with a given Epo dose were pooled into one data point, mean ± sem. (B) Schematic of the factors that regulate the size of the EryA pool, ‘A’, measured as the fraction of all Ter119⁺ cells that are EryA. F  = fraction of EryA cells that express Fas. ß  = input into the EryA pool from earlier progenitor stages. αA  = output from the EryA pool into later erythroid subsets (EryB). A²F  = Fas-mediated cell loss. See mathematical model described in Text S2. In panel (A) of this figure, A is plotted against F, expressed as a ratio to the A and F values in the basal state, respectively.

Figure 3. Increased erythropoiesis in mice deficient in the Fas pathway.

Legend in A also applies to panels B, D. lpr-Rag1^−/− mice are on the C57BL/6 background (in blue), and are compared with control Rag1^−/− mice on the C57BL/6 background. gld-Rag1^−/− mice are on the Balb/C background (in red), and are compared with control Rag1^−/− mice on the Balb/C background. (A) Hematocrit ( = fraction of the blood volume that is due to red cells) and Plasma Epo of lpr-Rag1^−/−, gld-Rag1^−/− and Rag1^−/− age and strain-matched control mice. M =  males. F =  females. Box and whiskers delineate the central 50% and 90% of readings, respectively. Median is indicated with a horizontal line; arithmetic mean with a ‘+’. Data points correspond to individual mice. Between 11 and 40 mice examined per genotype. *p<0.05, **p<.005, ***p<0005 (ANOVA). (B) Hematocrit vs. plasma Epo in the subset of mice where both values were measured, in the basal state, for lpr-Rag1^−/− and matched Rag1^−/− control mice (left panel), and for gld-Rag1^−/− and matched Rag1^−/− controls (right panel). Data are mean ± sem of ≥16 mice *p≤0.001 (two-tailed t test, unequal variance). (C) Flow cytometric measurement of reticulocyte number. Top: whole blood stained with either DRAQ5 (detects DNA) or thiazole orange (TO, detects both DNA and RNA). Reticulocytes lack a nucleus but retain RNA. They therefore form a DRAQ5-negative, TO-positive population. Bottom panel shows analysis in wild-type (WT) mice either in the basal state or following Epo injection; and in gld-Rag1^−/− and control Rag1^−/− mice. (D) Reticulocyte in lpr-Rag1^−/−, gld-Rag1^−/− and matched Rag1^−/− controls, measured by flow-cytometry ***p<0.0001, two-tailed t-test, unequal variance; ff =  p<0.001, F test.

Figure 4. Increased frequency of spleen, but not bone-marrow, erythroid progenitors and precursors in mice deficient in the Fas pathway.

The Legend in panel B also applies to C,D. (A) Representative flow-cytometric analysis of spleen erythroid subsets in gld-Rag1^−/− and matched Rag1^−/− controls, showing increased frequency of ProE and EryA within Ter119⁺ cells. (B) Frequency of erythroblast subsets in spleen erythropoietic tissue, measured as in A, expressed as fraction of all spleen Ter119⁺ cells. F = female M = male. Box and whiskers delineate the central 50% and 95% of readings, respectively, with the median indicated with a horizontal line and arithmetic mean with a ‘+’. Data points are individual mice (11–40 mice per sex/genotype combination). Data was pooled from several independent experiments. (C) CFU-e progenitors in spleen and bone-marrow. Data pooled from two independent experiments for females, and one experiment for males, for each genotype. (D) Frequency of erythroblast subsets in bone-marrow expressed as fraction of all spleen Ter119⁺ cells. For all panels, *p<0.05 **p<0.02, ***p<0.002 (ANOVA, for difference in means). f = p<0.05 ,ff = p<0.02, fff = p<0.002 (F test, for difference in variance).

Figure 5. Loss of Fas function results in a larger and more variable basal spleen erythroid progenitor pool.

(A) Spleen erythroid subsets ProE, EryA or EryB, expressed as absolute number of cells per gram body weight, in gld-Rag1^−/−, lpr-Rag1^−/− and matched Rag1^−/− controls, shown separately for male (M) and female (F) mice. Data corresponds to the same mouse dataset as in Fig 4. Box and whiskers delineate the central 50% and 90% of readings, respectively, with the median indicated with a horizontal line and arithmetic mean with a ‘+’. Data points correspond to individual mice. Between 11 and 40 mice examined per genotype; data pooled from several independent experiments. *p<0.05, **p<0.005, ***p<0.0005 (ANOVA, for difference in means); f = p<0.05, ff = p<0.005, fff = p<0.0005 (F test, for difference in variance). (B) Frequency distribution histograms for EryA, in male and female lpr-Rag1^−/− and matched Rag1^−/− controls. The coefficient of variation for each group is shown. Purple line is the fitted normal distribution curve. Same data set as in panel (A).

Figure 6. Delayed response to Epo-induced stress in mice deficient in the Fas pathway.

(A–D) gld-Rag1^−/−, lpr-Rag1^−/− and matched Rag1^−/− control mice were injected with Epo (300 U/ 25 g body weight) subcutaneously at t = 0. The erythropoietic response was followed for 6 days. Data is mean ± sem for 3 to 18 mice per time point per genotype, pooled from up to 3 experiments per time point. Data at t = 0 is the basal state data shown in Figures 3A and 5A, pooled for males and females. (A) Hematocrit measurements (B) Spleen EryA (cells per gram body weight) in the same mouse set as in the top panel. Data points are individual mice, with the mean ±sem for each day marked as a horizontal line. (D) The difference in mean EryA number (shown in panel B) between the mutant gld-Rag1^−/− or lpr-Rag1^−/− and their matched Rag1^−/− controls, for each day. The size of the corresponding control (Rag1^−/−) EryA pool is marked with a black bar. For all panels: *p<0.05, **p<0.005, ***p<0.0005 (t test, unequal variance). (D) The rate of change in hematocrit between days 2 and 3 following Epo injection. The same dataset as in panel (A), showing the differences in hematocrit measured in multiple independent experiments on days 2 and 3. Altogether 5 independent comparisons are shown. For all panels: *p<0.05, **p<0.005, ***p<0.0005 (t test, unequal variance).

Figure 7. Delayed response to Epo-induced stress in mice deficient in the Fas pathway.

(A–C) Analysis of the ProE response to Epo injection. The same experiment and dataset as in Figure 6. (A) Spleen ProE (cells per gram body weight). Data points are individual mice, with the mean ±sem for each day marked as a horizontal line. Data pooled from 1 to 3 experiments with 3 to 18 mice per genotype. (B) The rate of change in spleen ProE between days 1 and 2 post-Epo injection in the mutant gld-Rag1^−/− (indicated in red) or lpr-Rag1^−/− (blue) and their matched Rag1^−/− controls. Dataset as in panel (A). Each point represents the mean difference in ProE between independent experiments done on days 2 and 3, in 5 independent comparisons. (C) The rate of change in spleen ProE throughout the first 3 days of response to Epo, computed as in panel B. The rate of change between days 0 and 1, days 1 and 2, and days 2 and 3, were plotted on days 0.5, 1.5 and 2.5, respectively. Data points represent pooled mutant (lpr/gld) or control differences (mean ±sem). For all panels: *p<0.05, **p<0.005, ***p<0.0005 (t test, unequal variance).

Figure 8. Delayed response to hypoxia-induced stress in mice deficient in the Fas pathway.

(A–D) gld-Rag1^−/− or control mice were transferred to a hypoxia chamber with ambient 11% oxygen for 1 or 3 days. Data is mean ±sem for 4 to 9 age and gender-matched mice per time point per genotype. Data at t = 0 is the basal state data shown in Figures 3A and 5A, pooled for males and females. (A) Hematocrit measurements, performed via CritSpin microcentrifugation. (B) Reticulocyte measurements, performed flow cytometrically, as in Figure 3C. (C) Plasma Epo, measured by ELISA. (D) Spleen erythroid subsets ProE, EryA, EryB and EryC, expressed as absolute number of cells per gram body weight in gld-Rag1^−/− and matched Rag1^−/− controls. For all panels: *p<0.05, **p<0.005, ***p<0.0005 (t test, unequal variance).

Figure 9. Absence of a Fas-regulated EryA reserve delays the response to stress.

EryA cells are continuously formed from earlier precursors (‘input’). In the basal state, when Epo concentrations are low, only a small fraction of these cells survive, forming the ‘basal EryA pool’ (in purple). The remaining EryA undergo apoptosis, either through Fas (‘Fas-regulated reserve’, green) or alternative mechanisms (‘Alternative reserve’, blue). Together, the EryA reserve pools are 30 to 60 fold the size of the basal pool (see Figure 6). During the initial response to stress, high Epo levels rescue the EryA reserve pools from apoptosis, resulting in an immediate increase in the size of the surviving EryA pool and an increase in erythropoietic rate (solid colors indicate surviving cells, dashed lines indicate cells that underwent apoptosis). We suggest that lpr and gld mice partially compensate for the absence of the Fas-regulated reserve by generating fewer EryA cells (a smaller input). In this way, the absence of Fas –mediated apoptosis does not excessively increase the basal EryA pool (which does increase 1.5–4 fold, see Figure 5; this increase is much smaller than the stress-induced increased and is not shown). During stress, the absence of the Fas-regulated reserve in lpr and gld mice reduces the number of EryA that may be immediately recruited into the surviving EryA pool and consequently delays the stress response.