Conditional deletion of STAT5 in adult mouse hematopoietic stem cells causes loss of quiescence and permits efficient nonablative stem cell replacement

Abstract

Currently, there is a major need in hematopoietic stem cell (HSC) transplantation to develop reduced-intensity regimens that do not cause DNA damage and associated toxicities and that allow a wider range of patients to receive therapy. Cytokine receptor signals through c-Kit and c-Mpl can modulate HSC quiescence and engraftment, but the intracellular signals and transcription factors that mediate these effects during transplantation have not been defined. Here we show that loss of one allele of signal transducer and activator of transcription 5 (STAT5) in nonablated adult mutant mice permitted engraftment with wild-type HSC. Conditional deletion of STAT5 using Mx1-Cre caused maximal reduction in STAT5 mRNA (> 97%) and rapidly decreased quiescence-associated c-Mpl downstream targets (Tie-2, p57), increased HSC cycling, and gradually reduced survival and depleted the long-term HSC pool. Host deletion of STAT5 was persistent and permitted efficient donor long-term HSC engraftment in primary and secondary hosts in the absence of ablative conditioning. Overall, these studies establish proof of principle for targeting of STAT5 as novel transplantation conditioning and demonstrate, for the first time, that STAT5, a mitogenic factor in most cell types, including hematopoietic progenitors, is a key transcriptional regulator that maintains quiescence of HSC during steady-state hematopoiesis.

Figure 1

Host STAT5 dosage in adult mice controls HSC engraftment during nonablative transplantation. Recipient mice (CD45.2) were transplanted with either 5 × 10⁶ GFP-transgenic or CD45.1 BM cells under nonablative conditions. (A) Percentage of donor chimerism in each recipient mouse 16 weeks after transplantation. From 2 separate injection dates, wild-type (n = 10), STAT5ab^+/ΔN (n = 8), STAT5ab^+/null (n = 6), and from 10 separate injection dates (29 mice total), STAT5ab^ΔN/ΔN (n = 9 surviving mice). (B) Percentage of donor-derived overall and Gr-1⁺, B220⁺, or CD4⁺ cells in each recipient mouse 16 to 24 weeks after transplantation. From 4 separate injection dates, wild-type (n = 4), STAT5ab^ΔN/null (n = 4). (C) Peripheral blood hematology of each STAT5ab^ΔN/null mouse before and 16 to 24 weeks after injection with donor BM cells. (D) Percentage of donor-derived Gr-1, B220, Ter119, or CD4 cells obtained in lethally irradiated secondary recipients. Two wild-type and 2 STAT5ab^ΔN/null engrafted mice from panel B were used as donors for the secondary transplantation. Each donor was transplanted into 5 recipients. The representative dot plot from one secondary recipient is shown. Mean plus or minus SD values for all recipients are indicated above each plot. (E) E14.5 fetal liver cells from wild-type, STAT5ab^ΔN/ΔN, or STAT5ab^null/null CD45.2⁺ donors were transplanted into lethally irradiated CD45.1 recipients. Sixteen weeks later, the transplanted BM chimeras were challenged with 5 × 10⁶ GFP-transgenic BM cells. The percentage of donor-derived (GFP⁺) overall and Gr-1, B220, and CD4 cells in each mouse was determined by flow cytometry 16 weeks later. The number of chimeras challenged from 2 separate injection dates were wild-type (n = 9), STAT5ab^ΔN/ΔN (n = 7), and STAT5ab^null/null (n = 5). For t tests relative to wild-type, P < .001.

Figure 2

Improved engraftment in STAT5ab^+/null host mice by Gab2 deletion during nonablative transplantation. Recipient mice (CD45.2) were transplanted with 5 × 10⁶ CD45.1⁺ BM cells. (A) Percentage of donor chimerism (% CD45.1⁺ cells) in each recipient mouse 16 weeks after BM injection. From 2 separate injection dates, wild-type (n = 9), STAT5ab^+/null (n = 8), Gab2^−/− (n = 9), and from 5 separate injection dates STAT5ab^+/nullGab2^−/− (n = 14). The Wilcoxon 2 sample test was used for statistical analysis. Numbers 1 to 3 are mice that were further analyzed in panel B. (B) Representative dot plots of primary recipient mice as well as secondary transplanted recipients 16 weeks after BM injection and the percentage of donor-derived Gr-1, B220, Ter119, or CD4 cells are shown. In panel A, numbers 2 and 3 denote specific mice that were killed and BM analyzed by secondary transplantation (one donor into 5 recipients for Gab2^−/−STAT5ab^+/null), and average plus or minus SD values are indicated above each plot.

Figure 3

Improved engraftment of STAT5ab^+/null recipients on either the BoyJ (CD45.1) or F1 (CD45.1/CD45.2) background. (A,B) Wild-type or STAT5ab^+/null mice on the CD45.1 background were injected with 5 × 10⁶ CD45.2 BM cells. The numbers of mice injected from 3 separate injection dates were wild type (n = 8) and STAT5ab^+/null (n = 8). The percentage of donor chimerism (% CD45.2⁺ cells) in each recipient mouse was determined 4, 8, and 18 weeks after transplantation. Multilineage analysis in Gr-1, B220, and CD4 cells was determined 18 weeks after transplantation on 4 of the wild-type and all of the mutant recipients. (C) A representative dot plot from the secondary transplantation is shown (one donor into 5 recipients). Data are from a pool of 3 wild-type or 3 STAT5ab^+/null engrafted mice used as donors. (D,E) Recipient mice on the F1 background were transplanted with 5 × 10⁶ CD45.1 BM cells on a single injection date. The percentage of donor chimerism (% CD45.1⁺CD45.2⁻ cells) in each recipient mouse was determined 7, 19, and 25 weeks after transplantation for wild-type (n = 8) and STAT5ab^+/null (n = 7) mice. One mouse died between 19 and 25 weeks. For multilineage analysis, the percentage of CD45.1⁺CD45.2⁻ cells was determined for Gr-1, B220, or CD4 cells in each recipient mouse 25 weeks after transplantation. (F) Representative dot plot (gated first on Gr-1, B220, Ter119, or CD4) after secondary transplantation with a pool of 3 wild-type and 3 STAT5ab^+/null engrafted mice (one donor into 5 recipients). Above each plot in panels C and F are mean plus or minus SD values of the percentage of donor chimerism from 5 recipients.

Figure 4

Conditional deletion of STAT5 in adult host HSCs, but not stroma, permits efficient stem cell replacement. (A) The percentage of deletion of STAT5 in Gr-1⁺ cells 7 days (D7) or 4 months (4M) after pI:pC treatment (3 doses, 16 mg/kg, every other day) was determined by real-time quantitative PCR (n = 5) normalized to wild-type STAT5 (■). The control is STAT5ab^flox/null mice, which should yield 50% the amount of wild-type STAT5. Deletion in the KLS pool on day 7 is also shown (□). (B) Percentage of CD45.1-derived overall and Gr-1, B220, Ter119, or CD4 cells in the peripheral blood of each recipient 22 to 40 weeks after transplantation into wild-type (n = 7) or STAT5 KO (n = 10) from 2 separate injection dates. (C) Percentage of CD45.1-derived Gr-1, B220, or c-Kit⁺Lin⁻Sca-1⁺ cells in BM from each recipient (1-5 in Figure 4B) 40 weeks after transplantation. (D) A representative dot plot from each secondary recipient mouse (nos. 1-5) 16 weeks after transplantation with one primary donor equivalent per 5 secondary recipients. Each engrafted mouse in panel C was transplanted into 3 lethally irradiated CD45.2 recipient, and the numbers shown are mean plus or minus SD values of the percentage of donor chimerism from all 3 recipients. (E) Experimental outline, representative fluorescence-activated cell sorting, and average donor chimerism (mean ± SD) from chimeric mice before and after transplantation with 5 × 10⁶ GFP-transgenic BM cells.

Figure 5

Conditional deletion of STAT5 increases host HSC apoptosis and decreases HSC pool size. BM from STAT5 KO and littermates was assayed by multiparameter flow cytometry to quantitate the number of primitive BM fractions. BM was stained with antibodies against lineage markers, c-Kit, Sca-1, and Flk2 (A) or lineage markers, c-Kit, Sca-1, CD34, CD16/32, and IL-7R (B,C). Three separate experiments were performed with 3 to 5 mice per genotype compared. (A) The absolute number of primitive HSC populations in BM cells from both hind limbs. KLS cells were defined as c-Kit⁺Lin⁻Sca-1⁺ cells. ST-HSCs are identified as Flk2⁺KLS cells and LT-HSCs as identified as Flk2⁻KLS cells. (B) The absolute number of common myeloid progenitor (CD34^+/lowCD16/32^intLin⁻c-Kit⁺Sca-1⁻), granulocyte-macrophage progenitor (CD34⁺CD16/32⁺Lin⁻c-Kit⁺Sca-1⁻), and megakaryocyte-erythroid progenitor (CD34⁻CD16/32⁻Lin⁻c-Kit⁺Sca-1⁻) cells per both hind limbs are shown. (C) The absolute number of CLP was defined by IL-7R⁺Lin⁻Sca-1^lowc-Kit^low phenotype and is shown per both hind limbs. (D) The proportion of annexin V–positive (DAPI-negative) cells within the ST-HSC and LT-HSC fractions is shown (n = 3). (E) Three to 4 months after pI:pC treatment (7 doses), the absolute number of LT-HSC defined both as CD34⁻ and Flk2⁻ KLS were analyzed from both hind limbs (n = 7). (F) One or 5 months after pI:pC treatment (7 doses), the percentage of annexin V–positive/DAPI-negative LT-HSCs (CD34⁻ KLS) was analyzed for wild-type (n = 6) and KO (n = 7) mice.

Figure 6

Conditional deletion of STAT5 causes rapid and sustained loss of HSC quiescence. (A) Representative flow cytometry analysis of Pyronin Y and Hoechst 33342 staining on KLS cells. (B) The percentage of KLS cells in G₀, G₁, and S/G₂/M phase is shown as mean plus or minus SD values from wild-type (n = 9) and KO (n = 6) analyses. (C) Wild-type or KO mice were injected intraperitoneally with a single dose of BrdU (as described in “BrdU staining and cell-cycle analysis of HSC”) and killed 3 days later. Representative flow cytometric analysis is shown for BrdU incorporation into gated KLS cells. (D) Bars represent the percentage of BrdU⁺ cells per KLS cells expressed as mean plus or minus SD values for n = 3 per group. P values are indicated when significant differences were observed between WT and KO groups. (E) The 7-dose pI:pC treatment regimen with treatments on alternating days for 2 weeks. These mice were analyzed 4 months later in LT-HSC subsets defined as CD34⁻ KLS. (F) Mice that received 7 doses of pI:pC were analyzed for cell-cycle status by Hoechst/Pyronin Y staining. Shown are values for G₀, G₁, and S/G₂/M fractions for both wild-type and KO mice (n = 3).

Figure 7

Conditional deletion of STAT5 causes rapid and sustained reduction in expression of quiescence-associated genes in HSCs. (A) BM was isolated from 4 independent pairs of wild-type and STAT5 KO mice, and KLS cells were sorted before mRNA isolation. Deletion of STAT5b is shown as a control. The relative mRNA expression levels of Tie-2, p57, p21, p27, Cyclin D1, Gfi-1, Bmi-1, CXCR4, CD44, and Myc were evaluated by quantitative real-time RT-PCR. Means from 3 or 4 independent experiments are shown with SD. (B) BM was isolated 1 month after 7 doses of pI:pC, and means from 3 independent experiments are shown for analysis of STAT5b, Tie-2, p57, and p21. From panels A and B, each batch of sorted KLS cells from WT and KO mice was analyzed by quantitative real-time PCR performed with the gene of interest and endogenous control GAPDH in triplicate.