Canonical Notch signaling is dispensable for adult steady-state and stress myelo-erythropoiesis

Abstract

Although an essential role for canonical Notch signaling in generation of hematopoietic stem cells in the embryo and in thymic T-cell development is well established, its role in adult bone marrow (BM) myelopoiesis remains unclear. Some studies, analyzing myeloid progenitors in adult mice with inhibited Notch signaling, implicated distinct roles of canonical Notch signaling in regulation of progenitors for the megakaryocyte, erythroid, and granulocyte-macrophage cell lineages. However, these studies might also have targeted other pathways. Therefore, we specifically deleted, in adult BM, the transcription factor recombination signal-binding protein J κ (Rbpj), through which canonical signaling from all Notch receptors converges. Notably, detailed progenitor staging established that canonical Notch signaling is fully dispensable for all investigated stages of megakaryocyte, erythroid, and myeloid progenitors in steady state unperturbed hematopoiesis, after competitive BM transplantation, and in stress-induced erythropoiesis. Moreover, expression of key regulators of these hematopoietic lineages and Notch target genes were unaffected by Rbpj deficiency in BM progenitor cells.

Figure 1.

Unperturbed myelo-E progenitor hierarchies in steady-state BM of Rbpj-deficient mice. (A-D) Analysis of hematopoietic stem and myelo-E progenitor cells in 6- to 10-week-old Rbpj^fl/fl Vav^Cre/+ (N = 8) and Rbpj^fl/fl Vav^+/+ (N = 4) and Rbpj^fll+ Vav^Cre/+ (N = 2) littermate controls. (A) Mean (standard deviation [SD]) BM cellularity per 2 femurs and 2 tibias. (B-D) Representative FACS profiles and mean (SD) frequencies of (B) lineage⁻Sca-1⁺c-Kit⁺CD150⁺Flt3⁻ HSCs and (C-D) myelo-E progenitor subsets. (E) Rbpj expression in HSCs and myelo-E progenitor subsets purified from 8-week-old Rbpj^fl/fl Vav^Cre/+ (N = 5) and Rbpj^fl/fl Vav^+/+ or Rbpj^fll+ Vav^Cre/+ littermate controls (N = 1 and 2, respectively). Mean (standard error of the mean [SEM]) expression normalized to hypoxanthine phosphoribosyltransferase 1 (Hprt). Samples in which the mean value of replicates was ≤0.001 (relative to Hprt) were considered below cutoff value (#). (F-G) In vitro colony assays using total BM from Rbpj^fl/fl Mx1^Cre/+ (N = 5-6) and age-matched (10-14-week-old) Rbpj^fl/fl Mx1^+/+ (N = 2) or Rbpj^fl/+ Mx1^Cre/+ (N = 6) control mice treated with Poly(I:C) 4 to 5 weeks before analysis. (F) Mean (SD) in vitro CFU-GM and burst-forming units-E (BFU-E) colonies. (G) Pure (CFU-Mk) and mixed-lineage (Mk-mix) Mk-containing colonies scored after staining with acetylthiocholiniodide. Mean (SD) values from 2 to 3 experiments. (H-I) Mean (SEM [H] or SD [I]) of circulating platelet and red blood cell (RBC) counts in 6- to 10-week-old Rbpj^fl/fl Vav^Cre/+ (N = 18-12) and age-matched controls (Rbpj^fl/fl Vav^+/+ [N = 8-11] or Rbpj^fl/+ Vav^Cre/+ [N = 2-5]). For all data sets (A-I), statistical significance was investigated between homozygously Rbpj-deleted and control mice. *P < .05; **P < .01; ***P < .001. MegE, Mk-E progenitor; MkP, Mk progenitor; PreGM, pregranulocyte/macrophage progenitor; ProEry, pro-erythroblasts. See also supplemental Figure 1A-G.

Figure 2.

Canonical Notch signaling is dispensable for stem and myelo-E progenitor cell replenishment in competitive bone marrow chimeras. (A-C) One million BM cells were harvested from 10- to 14-week-old Rbpj^fl/fl Mx1^Cre/+ and control (Rbpj^fl/fl Mx1^+/+ and Rbpj^fl/+ Mx1^Cre/+) mice (CD45.2) 4 to 5 weeks after poly(I:C) treatment and transplanted into lethally irradiated wild-type (WT; CD45.1 or CD45.1/2) recipients together with 1 million competitor WT (CD45.1 or CD45.1/2) adult BM cells. Reconstitution of HSC (N = 2 of Rbpj^fl/fl Mx1^Cre/+ and N = 1 and N = 3 of Rbpj^fl/fl Mx1^+/+ and Rbpj^fl/+ Mx1^Cre/+, respectively) and myeloid progenitor subsets (N = 5 of Rbpj^fl/fl Mx1^Cre/+, N = 2 of Rbpj^fl/fl Mx1^+/+, N = 3 of Rbpj^fl/+ Mx1^Cre/+) in the BM of engrafted mice was assessed 7 to 9 weeks after transplantation. Percentage donor (CD45.2)-derived reconstitution of (A) HSC (mean ± SD) and (B) myeloid progenitor subsets (mean ±SEM) relative to total BM cells. (C) Mean (SEM) CD45.2 contribution of Rbpj^fl/fl Mx1^Cre/+ (N = 6) and control (N = 2 and N = 6 of Rbpj^fl/fl Mx1^+/+ and Rbpj^fl/+ Mx1^Cre/+, respectively) BM cells to total NK1.1⁻Mac1⁺ myeloid, NK1.1⁻Mac1⁻CD19⁺ B and NK1.1⁻Mac1⁻CD4/CD8⁺ T cells. (D) Reconstitution of CD45.2-derived blood platelets (CD41⁺CD150⁺eGFP⁻) of total platelets in Vwf-eGFP (CD45.1/2) recipients 4 to 5 weeks after transplantation with 1 million 10- to 11-week-old CD45.2 Rbpj^fl/fl Mx1^Cre/+ (N = 4) or control Rbpj^fl/+ Mx1^Cre/+ (N = 6) BM cells and 1 million Vwf-eGFP (CD45.1/2) competitor BM cells. (Left) Representative FACS profiles of platelet reconstitution in engrafted mice. (Right) Mean (SEM) percentage of CD150⁺CD41⁺ platelets derived from transplanted CD45.2 BM cells. For all data sets (A-D), statistical significance was investigated between Rbpj-deleted and control mice. ***P < .001.

Figure 3.

Expression of myelo-E lineage programs and Notch-target genes in Rbpj-deficient E and Mk progenitors. (A-D) E and Mk progenitor subsets were purified from individual 10- to 12-week-old Rbpj^fl/fl Mx1^Cre/+ (N = 6) and age-matched Rbpj^fl/fl Mx1^+/+ (N = 2) and Rbpj^fl/+ Mx1^Cre/+ (N = 2) poly(I:C)-treated control mice and subjected to quantitative gene expression analysis for (A,C) erythroid and (B-C) megakaryocytic-related genes and (D) Notch1 and Notch2 receptors. Mean (SEM) values normalized to Hprt. No differences between Rbpj^fl/fl Mx1^Cre/+ and control mice reached statistical significance. (E) BM HSCs, Mk, and E progenitor cells (100 cells per replicate) were purified from individual 8-week-old RbpJ^fl/fl Vav^Cre/+ (N = 5) and age-matched RbpJ^fl/fl Vav^+/+ or RbpJ^fl/+ Vav^Cre/+ controls (N = 1 and 2, respectively) and analyzed for expression of the Notch target genes Hes1, Hes5, Nrarp, and Gata3. Mean (SEM) values. Samples in which the mean value of replicates was ≤0.001 (relative to Hprt expression) were considered below cutoff value (#). For all data sets (A-E), statistical significance was investigated between Rbpj-deleted and control mice. *P < .05; **P < .01. See also supplemental Figure 1H.

Figure 4.

Canonical Notch signaling is dispensable for phenylhydrazine-induced stress erythropoiesis. Seven- to 8-week-old RbpJ^fl/fl Vav^Cre/+ (N = 3-6) and age-matched RbpJ^fl/fl Vav^+/+ or RbpJ^fl/+ Vav^Cre/+ controls (N = 2-3 and 2-3, respectively) were injected with PHZ to induce acute hemolytic anemia. (A) Circulating red blood cell counts in steady-state (day 0) and 2 and 4 days after last PHZ injection. Mean (SD) values are shown. (B-C) Mean (SEM) absolute numbers of E progenitors (PreCFU-E and CFU-E), 4 days after PHZ-induced anemia in BM (B) and spleen (C). *P < .05.