Prostaglandin E2 enhances hematopoietic stem cell homing, survival, and proliferation

Abstract

Adult hematopoietic stem cells (HSCs) are routinely used to reconstitute hematopoiesis after myeloablation; however, transplantation efficacy and multilineage reconstitution can be limited by inadequate HSC number, or poor homing, engraftment, or self-renewal. Here we report that mouse and human HSCs express prostaglandin E2 (PGE2) receptors, and that short-term ex vivo exposure of HSCs to PGE2 enhances their homing, survival, and proliferation, resulting in increased long-term repopulating cell (LTRC) and competitive repopulating unit (CRU) frequency. HSCs pulsed with PGE2 are more competitive, as determined by head-to-head comparison in a competitive transplantation model. Enhanced HSC frequency and competitive advantage is stable and maintained upon serial transplantation, with full multilineage reconstitution. PGE2 increases HSC CXCR4 mRNA and surface expression, enhances their migration to SDF-1 in vitro and homing to bone marrow in vivo, and stimulates HSC entry into and progression through cell cycle. In addition, PGE2 enhances HSC survival, associated with an increase in Survivin mRNA and protein expression and reduction in intracellular active caspase-3. Our results define novel mechanisms of action whereby PGE2 enhances HSC function and supports a strategy to use PGE2 to facilitate hematopoietic transplantation.

Figure 1

PGE₂ enhances hematopoietic stem cell engraftment. (A) Test bone marrow from CD45.1 or CD45.2 mice was treated with vehicle or dmPGE₂, respectively. CD45.1/CD45.2 hybrid marrow cells were used as competitors. Limiting dilutions were transplanted into lethally irradiated (1100 cGy, split dose) CD45.1/CD45.2 hybrid mice and chimerism in PB was analyzed for 20 weeks. A representative flow plot detecting each cell population is shown. (B) Frequency analysis (top) for vehicle (red)– or dmPGE₂ (blue)–pulsed cells, determined by Poisson statistics, at 12 weeks; P₀ = 85 560 (vehicle) and P₀ = 23 911 (dmPGE₂ treated). Chimerism in PB and CRU analysis is shown at 12 weeks (mean ± SEM). Data represent 2 pooled experiments; n = 5 mice/group/experiment, each assayed individually. (*P < .05 compared with vehicle control.) (C) HSC frequency analysis in recipients of vehicle- or dmPGE₂-treated bone marrow over 20 weeks. Fold change indicates increase in frequency of engraftment of dmPGE₂-pulsed cells compared with vehicle. (D) Representative FACS plots of multilineage reconstitution (M indicates myeloid; B, B lymphoid; and T, T lymphoid). Multilineage analysis for primary transplantation (32 weeks) and a cohort of 4 mice that received transplants from mice that underwent primary transplantation at 20 weeks, with analysis 12 weeks later. For mice that underwent primary transplantation at 32 weeks, vehicle-treated cells were (mean ± SEM) 14.1% plus or minus 3.5% M, 70.8% plus or minus 1.1% B, and 17.8% plus or minus 1.4% T, versus dmPGE₂-treated cells, which were 15.7% plus or minus 2.5% M, 76.9% plus or minus 3.4% B, and 7.5% plus or minus 1.2% T. For mice that underwent secondary transplantation at 12 weeks, vehicle-treated cells were 15.7% plus or minus 5.3% M, 60.3% plus or minus 4.8% B, and 22.1% plus or minus 3.6% T, versus dmPGE₂-treated cells, which were 37.0% plus or minus 6.5% M, 52.3% plus or minus 5.4% B, and 9.0% plus or minus 1.4% T. (*P < .05 vs vehicle control.) Increased chimerism of dmPGE₂-treated cells versus vehicle is shown for primary transplantation at 20 weeks (time of secondary transplantation) and in a subcohort at 32 weeks (time of 12-week analysis of secondary transplantation), for secondary transplantation at 12 weeks and 24 weeks. Data for 20-week primary transplantation were from 2 pooled experiments; n = 5 mice/group/experiment, each assayed individually. Data for secondary transplantations were from n = 5 mice/group, each assayed individually.

Figure 2

PGE₂ receptors are expressed on murine and human HSPCs. (A) Representative FACS gating of Lin^neg murine bone marrow showing c-kit⁺ and Sca-1⁺ gates and SLAM (CD150⁺, CD48⁻) and CD34 gating of SKL cells. EP1-EP4 surface receptor expression on murine KL, SKL, SLAM SKL, and CD34^neg SKL cells is shown. (B) Representative FACS gating of human CD34⁺ and CD34⁺, CD38⁻ UCB cells. EP surface receptor expression on CD34⁺ and CD34⁺, CD38⁻ cells is shown.

Figure 3

Amplification plots for mRNA for PGE₂ receptors. (A) Primers designed to specifically detect murine EP1, EP2, EP3, or EP4 were used for QRT-PCR (with SYBR green) and plots with an activation step of 50°C for 2 minutes, denaturation at 95°C for 2 minutes, and amplification for 45 cycles at 95°C for 15 seconds, 50°C for 30 seconds, and 72°C for 30 seconds are shown. Plots corresponding to specific EP receptors are indicated in each amplification plot, where the legend key on the right shows the relative order of transcripts top to bottom. *Denotes the presence of at least 2 dissociation peaks indicating the presence of splice variants. (B) EP receptor amplification on human UCB CD34⁺ and CD34⁺ CD38⁻ cells with the same QRT-PCR procedure as described for panel A.

Figure 4

PGE₂ increases homing efficiency of HSPCs. (A) Test murine bone marrow cells were labeled with CFSE and treated with vehicle (red) or dmPGE₂ (blue) and 2 × 10⁷ labeled and treated WBM cells were transplanted into lethally irradiated mice. Sixteen hours later, bone marrow was analyzed by FACS for homed events. Data are expressed as mean plus or minus SEM; n = 6 mice/group, each assayed individually. (B) Test bone marrow cells from CD45.1 mice were treated with PBS, vehicle, or dmPGE₂, and 2 × 10⁷ treated WBM cells were transplanted into lethally irradiated CD45.2 mice. Sixteen hours later bone marrow was analyzed for homed SKL cells. The left panel shows representative data from 1 experiment; n = 3 mice/group, each assayed individually. The right panel shows the combined increase in homing efficiency of SKL cells after dmPGE₂ treatment for 3 experiments (n = 6 mice/group per experiment, each assayed individually). (C) SKL cells from CD45.1 and CD45.2 mice were isolated by FACS sorting and treated with either dmPGE₂ or vehicle. Five lethally irradiated CD45.1/CD45.2 hybrid mice received 3 × 10⁴ vehicle-treated CD45.1-sorted SKL plus 3 × 10⁴ dmPGE₂-treated CD45.2 SKL cells (top panel). Five mice received a similar transplant with treatment groups switched between strains (bottom panel). Representative flow gating of marrow 16 hours after transplantation and combined data for the homing efficiency of dmPGE₂- or vehicle-treated, sorted SKL cells (mean ± SEM; n = 10 mice, each assayed individually) are shown. (D) Low-density UCB mononuclear cells (LDMCs) were isolated and treated with either dmPGE₂ or vehicle. Five sublethally irradiated NS2 mice received dmPGE₂-treated LDMCs and 5 received vehicle-treated LDMCs. Bone marrow was analyzed 16 hours later and the number of CD34⁺ cells determined and homing efficiency calculated. Data are mean plus or minus SEM for n = 5 mice, each assayed individually.

Figure 5

CXCR4 receptor expression is increased on murine and human HSPCs after dmPGE₂ treatment. (A) CXCR4 expression (mean ± SEM; n = 3) on murine KL and SKL cells, and human CD34⁺ UCB cells 24 hours after treatment with dmPGE₂. Data are expressed as percentage change in mean fluorescence intensity (MFI) of CXCR4 due to treatment with dmPGE₂ or vehicle. (B) Freshly isolated Lin^neg cells were pulsed with dmPGE₂ or vehicle for 2 hours, washed, and resuspended in media with 10% HI-FCS and cultured at 37°C for 16 hours. After incubation, cells were washed, resuspended in RPMI/0.5% BSA, and allowed to migrate to rmSDF-1α for 4 hours. Total cell migration was quantitated by flow cytometry. Data are the mean plus or minus SEM percentage migration for 3 experiments. (†P < .05 for dmPGE₂-treated cells compared with cells treated with vehicle.) (Top inset) Percentage migration of gated SKL cells to positive (100 ng/mL SDF-1 in bottom chamber), negative (100 ng/mL SDF1 in upper chamber), or neutral (100 ng/mL SDF-1 in both upper and bottom chambers) gradients. Data are the mean ± SEM percentage migration for 3 experiments. (†P < .05 for dmPGE₂-treated cells compared with cells treated with vehicle.) (Bottom inset) Percentage migration of sorted SKL cells to 100 ng/mL SDF-1. Data are the mean plus or minus SEM percentage migration for 3 experiments. (†P < .05 for dmPGE₂-treated cells compared with cells treated with vehicle.) (C) Freshly isolated UCB CD34⁺ cells were pulsed with dmPGE₂ or vehicle for 2 hours, washed, and resuspended in media with 10% HI-FCS and cultured at 37°C for 16 hours. After incubation, cells were washed and resuspended in RPMI/0.5% BSA, and migration to rhSDF-1 was quantitated by flow cytometry. To block the CXCR4 receptor, replicate cells were incubated with 5 μg/mL AMD3100 for 30 minutes prior to the migration assay. Data are the mean plus or minus SEM percentage migration for 3 experiments. (†P < .05 for dmPGE₂-treated cells compared with cells treated with vehicle.) (D) Homing efficiency of vehicle- and dmPGE₂-treated cells to bone marrow in the absence and presence of 10 μM AMD3100. Cells were incubated with AMD3100 for 30 minutes prior to the homing assay. Data are expressed as mean plus or minus SEM; n = 3 mice/group, each assayed individually.

Figure 6

PGE₂ decreases apoptosis, increases Survivin expression, and decreases active caspase-3 in SKL cells. (A) Lin^neg bone marrow cells were treated with dmPGE₂ or vehicle and cultured in media supplemented with 2% FBS without growth factors for 24 hours to induce apoptosis. Cultured cells were stained for SKL and SLAM and PE-anti–active caspase-3 and/or FITC–annexin-V and the proportion of SKL or SLAM SKL cells undergoing apoptosis was determined by FACS. (Inset) dose-response analysis of the effects of dmPGE₂ on SKL cell apoptosis. (B) Fold increase in intracellular Survivin levels (mean fluorescence intensity [MFI]) in control and dmPGE₂-pulsed murine SKL and human CD34⁺ cells 24 hours after treatment; data are mean plus or minus SEM from 3 experiments; n = 3 mice/group, each assayed individually, or 3 separate cord blood samples. (C) Evaluation of intracellular Survivin and active caspase-3 levels (MFI) in SKL cells 24, 48, and 72 hours after treatment with dmPGE_2. Data are expressed as mean plus or minus SEM for 3 experiments; n = 3-6 mice/group, each assayed individually (*P < .05).

Figure 7

PGE₂ increases the proliferation of SKL cells in vitro and in vivo. (A) Lin^neg cells were treated with either vehicle or 1 μM dmPGE₂ for 2 hours, washed, and cultured in media with rmSCF, rhFlt3, and rhTpo. After 20 hours, cells were stained for SKL and Hoechst-33342 and Pyronin-Y. The proportion of SKL cells in cell cycle was quantitated by FACS. Representative flow plot showing cell-cycle distribution of gated SKL cells and combined data for fold increase in cell cycle for dmPGE₂-treated cells compared with vehicle control from 3 experiments; mean plus or minus SEM; n = 9 mice, each assayed individually. (B) CD45.1 Lin^neg bone marrow cells were treated with dmPGE₂ or vehicle and transplanted into lethally irradiated CD45.2 mice. Immediately after transplantation, BrdU was provided in drinking water and administered by intraperitoneal injection. Bone marrow was analyzed 16 hours later and the proportion of CD45.1⁺, SKL cells that was BrdU⁺ was analyzed by FACS analysis. Data are mean plus or minus SEM; n = 5 per mice/group, each assayed individually.