Canonical notch signaling is dispensable for the maintenance of adult hematopoietic stem cells

Abstract

Gain-of-function experiments have demonstrated the potential of Notch signals to expand primitive hematopoietic progenitors, but whether Notch physiologically regulates hematopoietic stem cell (HSC) homeostasis in vivo is unclear. To answer this question, we evaluated the effect of global deficiencies of canonical Notch signaling in rigorous HSC assays. Hematopoietic progenitors expressing dominant-negative Mastermind-like1 (DNMAML), a potent inhibitor of Notch-mediated transcriptional activation, achieved stable long-term reconstitution of irradiated hosts and showed a normal frequency of progenitor fractions enriched for long-term HSCs. Similar results were observed with cells lacking CSL/RBPJ, a DNA-binding factor that is required for canonical Notch signaling. Notch-deprived progenitors provided normal long-term reconstitution after secondary competitive transplantation. Furthermore, Notch target genes were expressed at low levels in primitive hematopoietic progenitors. Taken together, these results rule out an essential physiological role for cell-autonomous canonical Notch signals in HSC maintenance.

Figure 1. Stable engraftment of DNMAML-transduced progenitors

Lethally irradiated mice were reconstituted with 5FU-treated BM cells after transduction with the GFP-expressing retrovirus MigR1 or with a retrovirus expressing the GFP-tagged Notch inhibitor DNMAML. (A) Percentage of GFP⁺ cells among Gr1⁺CD11b⁺ myeloid cells in the peripheral blood 4 and 20 weeks after transplantation. The percentages are means ± SDs (n=4 in each group). The experiment is representative of 10 similar experiments. (B) Relative percentage of GFP⁺ myeloid cells at various time points after bone marrow transplantation (BMT), with the percentage observed at week 4 normalized to 1. The data represent the normalized means ± SDs (n=4). White triangles, MigR1; black squares, DNMAML. (C) Contribution of CD45.2⁺ cells to blood Gr1⁺CD11b⁺ myeloid cells after competitive transplantation of CD45.1⁺ BM and poly(I:C)-induced MxCre⁻ ROSA^DNMAML/+ BM (white triangles) or MxCre⁺ ROSA^DNMAML/+ BM (black triangles) (1:1 ratio). Data shown are the means ± SDs for 9 control and 8 DNMAML recipients.

Figure 2. Normal frequency of populations enriched for long-term hematopoietic stem cells in Notch-deprived bone marrow

(A) Frequency of Lin^−/loSca-1^hic-Kit^hi (LSK) progenitors in the control GFP⁺ fraction of MigR1 BM (n=7) or in DNMAML-transduced BM (n=8) 16 weeks after reconstitution. One representative example is shown. (B) Frequency of LSK progenitors among GFP⁻and GFP⁺ BM cells in mice reconstituted with MigR1 (n=7) or DNMAML-transduced BM (n=8). The data are shown as mean ± 2 SEM. (C) Analysis of BM LSK progenitors in Mx-Cre⁻ ROSA^DNMAML/+ or Mx-Cre⁺ ROSA^DNMAML/+ mice, after induction of Cre expression with poly(I:C). More than 98% of the cells in the LSK population had achieved DNMAML activation (Suppl. Fig. 1). One representative example is shown (n=5). (D) CD150 and CD48 expression in LSK cells of Mx-Cre⁻ ROSA^DNMAML/+ (n=6) or Mx-Cre⁺ ROSA^DNMAML/+ (n=4) mice after poly(I:C) induction. One representative example is shown. (E) Flt3 and Thy1.1 expression in GFP⁺ BM LSK cells of mice reconstituted with MigR1 or DNMAML-transduced B6-Thy1.1 BM. One representative example is shown (n=5). (F) Side population (SP) analysis after labeling BM cells with Hoechst 33342 and gating on GFP⁺ BM. One representative example is shown (n=5).

Figure 3. Competitive repopulation potential of Notch-deprived LSK progenitors in secondary recipients

(A) Experimental design. MigR1 or DNMAML-transduced progenitors were used to repopulate primary lethally irradiated recipients. After 12 weeks, purified GFP⁺ LSK cells were used in a secondary transplantation together with host-type competitor BM cells (500 and 2 × 10⁵ cells, respectively). (B) Blood of the secondary recipients analyzed 16 weeks after transplantation for donor-derived myeloid (Gr1⁺CD11b⁺), B (CD19⁺) and T (Thy1.2⁺) lineage cells (MigR1 LSK, n=8; DNMAML LSK, n=10). There was no significant difference between the ability of control and DNMAML-transduced LSK cells to repopulate the myeloid and B cell lineages (p=0.42 and p=0.26, respectively, Student t test). DNMAML LSK cells were deficient at repopulating the T cell compartment (p<0.001).

Figure 4. Long-term HSC numbers and function in the absence of CSL/RBPJ

(A) Stable CD45.2⁺ chimerism in lethally irradiated recipients of BM cells from poly(I:C)-treated Mx-Cre⁺ Rbpj^f/f (black triangles) or control Rbpj^f/f CD45.2⁺ mice (white triangles) (n=5 in each group). Chimerism was assessed by analysis of blood Gr1⁺CD11b⁺ myeloid cells. Data are the means ± SEMs. (B) Mixed BM recipients analyzed 18 weeks after BMT for the contribution of CD45.2⁺ cells to BM LSK progenitors (Lin^−/loSca-1^hic-Kit^hi), myeloid cells (Gr1⁺CD11b⁺), pro- and pre-B cells (B220⁺BP-1⁺sIgM⁻), double positive (DP) thymocytes (CD4⁺CD8⁺) and spleen marginal zone B cells (B220⁺CD21^hiCD23^lo). Data are means ± SEMs (n=5 in each group). ** indicates statistical significance (Student t test, p<0.01). (C) Similar frequency of LSK progenitors and CD34⁻Flt3⁻ LSK cells (enriched for LT-HSC) in engrafted control and CSL/RBPJ-deficient BM 18 weeks after competitive transplantation. Plots show a representative example; bar graphs indicate means ± SDs.

Figure 5. LSK progenitors are physiologically exposed to a low intensity of Notch signaling

(A) LSK progenitors were assessed by real-time RT-PCR in comparison to early T lineage progenitors (ETP), DN3 and CD4⁺CD8⁺ DP thymocytes. The amount of the Notch target gene transcripts Hes1 and Dtx1 (normalized to Hprt) was very low in LSK cells and DP thymocytes, but high in ETP and DN3 cells. Results are given as the mean ± SEM. ND, not detectable. (B) Hes1, Dtx1, c-myc Notch1, Notch2 and Runx1 transcripts in control and DNMAML-expressing LSK progenitors (normalized to Hprt). The measurements were performed in triplicates for 4-5 independent samples. Results are given as the mean ± SEM. ND, not detectable.

Figure 6. DNMAML inhibits the rapid induction of Notch signaling after exposure of LSK progenitors to Notch ligands in vitro

Freshly purified LSK progenitors from control or DNMAML-expressing BM were cultured with stromal monolayers of OP9 cells or OP9-DL1 cells expressing the Notch ligand Delta-like1. The amount of Hes1 and Dtx1 mRNA (normalized to Hprt) was assessed by real-time RT-PCR in LSK progenitors after 6 hours (A) or 48 hours (B) of co-culture. The data are the means ± SEMs. ND, not detectable.