CXCR4/CXCL12 axis counteracts hematopoietic stem cell exhaustion through selective protection against oxidative stress

Abstract

Hematopoietic stem cells (HSCs) undergo self-renewal to maintain hematopoietic homeostasis for lifetime, which is regulated by the bone marrow (BM) microenvironment. The chemokine receptor CXCR4 and its ligand CXCL12 are critical factors supporting quiescence and BM retention of HSCs. Here, we report an unknown function of CXCR4/CXCL12 axis in the protection of HSCs against oxidative stress. Disruption of CXCR4 receptor in mice leads to increased endogenous production of reactive oxygen species (ROS), resulting in p38 MAPK activation, increased DNA double-strand breaks and apoptosis leading to marked reduction in HSC repopulating potential. Increased ROS levels are directly responsible for exhaustion of the HSC pool and are not linked to loss of quiescence of CXCR4-deficient HSCs. Furthermore, we report that CXCL12 has a direct rescue effect on oxidative stress-induced HSC damage at the mitochondrial level. These data highlight the importance of CXCR4/CXCL12 axis in the regulation of lifespan of HSCs by limiting ROS generation and genotoxic stress.

Figure 1. Defective hematopoiesis in BM of CXCR4^−/− chimeras.

(a) BM sections from CXCR4^+/+ and CXCR4^−/− chimeras were stained with hematoxylin/eosin/safran. Chimeric mice were analyzed 3, 8 and 64 weeks after engraftment of eight-week-old, lethally irradiated mice with 5 × 10⁶ CXCR4^+/+ or CXCR4^−/− FL cells. (b) Total donor-derived LSK-SLAM cell numbers in BM. Mean ± SEM from 3 independent experiments (5–7 mice per group). (c) Restored hematopoietic reconstitution potential by retroviral expression of CXCR4 in CXCR4^−/− HSCs. Mean ± SEM of the percentages of donor-derived BM cells 14 and 24 weeks post engraftment. Data represent a pool from 8 mice over two independent experiments.

Figure 2. Increased endogenous ROS production in BM of CXCR4^−/− chimeras.

(a) Whole genome microarray analysis of CXCR4^−/− and CXCR4^+/+ LK and LSK cells. Venn diagrams show number of differentially expressed genes that are at least 1.5-fold changed. (b,c) GSEA of CXCR4^−/− and CXCR4^+/+ LSK cells showing significant enrichment of Hallmark Oxidative phosphorylation (b) and Hallmark DNA Repair (c) gene sets in CXCR4^−/− cells. (d) Full gating strategy with actual staining plots for immunophenotypic analysis of HSCs from CXCR4^+/+ and CXCR4^−/− chimeras performed 10 weeks after FL cell transplantation. (e) Mean Fluorescence Intensity (MFI) of DCF on BM progenitor cells from CXCR4^+/+ and CXCR4^−/− chimeras according to the gating strategy described in (d). DCF staining showed increased intracellular ROS in CXCR4^−/− LT-HSCs and HSPCs. Data are representative of 6 independent experiments. (f) DCF MFI of donor-derived LSK-SLAM (SLAM), LSK, LK and BM mononuclear (BM MNC) cells. Mean ± SEM from 6 mice per group, pooled from 2 experiments. (g) DCF MFI of LSK-SLAM cells from CXCR4^+/+ and CXCR4^−/− FL cells. Data from 3 independent experiments are shown. (h) Increased intracellular ROS in ten-week-old WT mice treated with CXCR4 antagonist TN140. Mean DCF MFI ± SEM from 6 mice per experiment in 2 independent experiments. (i) Increased phosphorylation of p38 MAPK in CXCR4^−/− HSPCs. Mean phospho-p38 MFI ± SEM in donor-derived LSK-SLAM, LSK, LK and BM mononuclear (BM-MNC) fractions. Data from 8 mice per group over 2 independent experiments are shown. CXCR4^+/+ and CXCR4^−/− cells were obtained from chimeric mice 10 weeks after FL cell transplantation.

Figure 3. Defects in CXCR4^−/− HSPC function are counteracted by the antioxidant NAC.

(a–d) DCF MFIs of donor-derived BM subsets of CXCR4^+/+ and CXCR4^−/− chimeras treated with NAC or vehicle. Mean ± SEM from 6 mice per group pooled from three independent experiments. (e–g) Phospho-p38 MFIs in different BM subsets. Mean ± SEM from 6 mice per group pooled from 3 independent experiments. (h,i) Percentages of Annexin-V⁺ cells in BM LSK and LK cells. Mean ± SEM from 8 mice per group pooled from 2 independent experiments. (j) Representative γH2AX staining (red) in BM LSK cells of CXCR4^+/+ and CXCR4^−/− chimeras. Nuclei were stained with DAPI (blue). (k) Frequencies of γH2AX-positive BM LSK cells. More than 100 cells were counted per group in 3 independent experiments. Mean ± SEM from 3 independent experiments with 3 to 5 mice per group.

Figure 4. In vivo treatment with NAC ameliorates the functional deficit of CXCR4^−/− HSPCs without influencing their proliferative status.

(a,b) CXCR4^+/+ and CXCR4^−/− chimeras were treated with NAC or vehicle, BrdU⁺ cell frequencies in BM LSK-SLAM and LSK subsets. Data represent mean ± SEM from 6 mice per group pooled from 2 independent experiments. (c) Total number of BM LSK-SLAM cells. Data represent mean ± SEM from 8 mice per group, pooled from 2 independent experiments. (d–f) Restoration of total CFC numbers after NAC-treatment in BM, spleen and peripheral blood (PB). Mean ± SEM of colonies scored for 6 samples per group of 2 independent experiments performed in duplicate.

Figure 5. CXCR4/CXCL12 axis prevents BSO-induced oxidative stress in HSCs.

(a) DCF MFIs of sorted WT LSK cells cultured for 48 hr with vehicle or BSO, in the absence or presence of 100 ng/mL CXCL12 or 2 mM NAC. Mean ± SEM, n = 6. (b) Annexin-V staining of LSK cells. Mean ± SEM, n = 3. (c) γ-H2AX-staining of LSK cells cultured for 48 hr in the absence (vehicle) or presence of BSO and CXCL12. Representative pictures are shown. (d) Percentages of γ-H2AX-positive LSK cells. More than 100 cells were counted per group in 3 independent experiments. (e) Effect of BSO-treatment on LSK proliferation. Mean ± SEM, n = 4. (f) Colony numbers (CFC) from LSK cells. Mean ± SEM of colonies from 5 independent experiments performed in duplicate. (g) CXCL12 rescues compromised hematopoietic long-term reconstitution with BSO-treated LSK cells. Before transplantation into lethally irradiated WT CD45.2⁺ recipients, CD45.1⁺ LSK cells were treated with BSO for 48 hr in the absence or presence of CXC12. Mean ± SEM of percentages of CD45.1⁺ donor-derived cells 6 months after transplantation from the pool (10 mice) independent experiments.

Figure 6. CXCR4^−/− HSPCs exhibit enhanced sensitivity to oxidative stress.

Donor-derived BM cells from CXCR4^+/+ and CXCR4^−/− chimeras 6 weeks post transplantation were sorted and cultured for 24 hr in the presence of a low BSO concentration (10 μM), with or without CXCL12. (a) Colony formation. Mean ± SEM of 3 independent experiments performed in duplicate. (b–d) DCF MFIs on CXCR4^−/− BM LSK-SLAM (b), LK (c) and BM-MNC cells (d). (e–g) DCF MFIs on CXCR4^+/+ BM LSK-SLAM (e), LK (f) and BM-MNC cells (g). Mean ± SEM of 3 independent experiments, with 3 mice per group. (h) GSH/GSSG ratios in sorted LSK and LK cells from CXCR4^+/+ and CXCR4^−/− chimeras, 6 weeks after transplantation. Mean ± SEM of 3 independent experiments performed in triplicate, with 4 mice per group. (i,j) CFC frequencies in sorted CD45.2+ BM and in peripheral blood cells from CXCR4^+/+ and CXCR4^−/− chimeras determined 3 months after transplantation. Sorted cells were plated with vehicle, NAC (2 mM) or p38 MAPK inhibitor SB203580 (25 μM). Mean ± SEM of 3 independent experiments performed in duplicate.

Figure 7. Mitochondria are targets for CXCL12 in oxidative stress control.

CXCL12 or the mitochondrial ROS scavenger Mito Tempo reduces Mitosox Red MFI in sorted LSK cells. Cells were treated with rotenone (5 μM) or vehicle for 1 hour in absence or presence of CXCL12 (10 ng/ml) or Mito Tempo (2 μM). Mean ± SEM of 3–7 independent experiments.