Remodeling of Bone Marrow Hematopoietic Stem Cell Niches Promotes Myeloid Cell Expansion during Premature or Physiological Aging

Abstract

Hematopoietic stem cells (HSCs) residing in the bone marrow (BM) accumulate during aging but are functionally impaired. However, the role of HSC-intrinsic and -extrinsic aging mechanisms remains debated. Megakaryocytes promote quiescence of neighboring HSCs. Nonetheless, whether megakaryocyte-HSC interactions change during pathological/natural aging is unclear. Premature aging in Hutchinson-Gilford progeria syndrome recapitulates physiological aging features, but whether these arise from altered stem or niche cells is unknown. Here, we show that the BM microenvironment promotes myelopoiesis in premature/physiological aging. During physiological aging, HSC-supporting niches decrease near bone but expand further from bone. Increased BM noradrenergic innervation promotes β₂-adrenergic-receptor(AR)-interleukin-6-dependent megakaryopoiesis. Reduced β₃-AR-Nos1 activity correlates with decreased endosteal niches and megakaryocyte apposition to sinusoids. However, chronic treatment of progeroid mice with β₃-AR agonist decreases premature myeloid and HSC expansion and restores the proximal association of HSCs to megakaryocytes. Therefore, normal/premature aging of BM niches promotes myeloid expansion and can be improved by targeting the microenvironment.

Keywords: Hutchinson-Gilford progeria; aging; hematopoietic stem cell; lymphoid; microenvironment; myeloid; niche.

Graphical abstract

Figure 1

Reduction of Endosteal Niches and Expansion of Non-endosteal Niches during Aging (A–B′ and H–I′) Representative whole-mount immunofluorescent staining of thick femoral sections for CD31 (A and B, green; H and I, red) and EMCN (A and B, red; H, I, blue) of young (8–30 weeks) and old (70–100 weeks) Nes-gfp mice with genetically labeled nestin⁺ cells (H and I, green). Arrowheads in insets (A′, B′, H′, and I′) depict CD31^hiEMCN⁻ capillaries and their coverage by Nes-GFP⁺ cells. (C–G) Quantification of (C) CD31^hiEMCN^hi transition zone vessels, (D) CD31^loEMCN^lo sinusoids, (E) CD31^hiEMCN⁻ arterioles with ≥6 μm diameter, and (F) CD31^hiEMCN⁻ capillaries with <6 μm diameter. Scale bar, 200 μm (A, B, H, and I), 100 μm (A′, B′, H′, and I′). (G) Frequency of endosteal and non-endosteal BM Nes-GFP⁺ cells from young adult (10–20 weeks, n = 11) and old mice (>66 weeks, n = 8). (J–N) Concentration of (J) IL-1β, (K) IL-6, (L) IL-1α, (M) IL-3, and (N) IFNγ in endosteal BM extracellular fluid (BMECF) from young WT mice (n = 5) and old WT mice (n = 4). Data are means ± SEM. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001. (C–F and J–N) Unpaired two-tailed t test. (G) One-way ANOVA and Bonferroni pairwise comparisons.

Figure 2

Increased Sympathetic Nerve Fibers during Aging (A, B, E, and F) Immunofluorescence of tyrosine hydroxylase (TH)⁺ sympathetic noradrenergic nerve fibers (white), CD31⁺ endothelial cells (red), and GFP⁺ cells (green) in the skull (A and B) and tibial (E and F) BM of young (A and E) and old (B and F) Nes-gfp mice. Scale bar, 100 μm. (C and D) Area covered by TH⁺ fibers in the (C) skull or (D) tibia of young (n = 12) and old (n = 8) Nes-gfp mice. Young mice were analyzed between 8–30 weeks of age, and old mice were 66–120 weeks old. Data are means ± SEM. ^∗p < 0.05; ^∗∗p < 0.01 (unpaired two-tailed t test).

Figure 3

β-Adrenergic Signals Promote Megakaryopoiesis during Aging (A and B) Representative flow chart (A) and quantification (B) of the frequency of CD41⁺ myeloid or megakaryocyte progenitors within lin⁻sca-1⁺c-kit⁺ (LSK) cells in endosteal BM from young WT mice (n = 7), young Adrb2^−/−Adrb3^−/− (double knockout [DKO]) mice (n = 7), old WT mice (n = 5), and old Adrb2^−/−Adrb3^−/− mice (n = 4). (C–F) Representative immunofluorescence staining for CD41 (red) and CD42 (green) in femoral BM sections of young (C) or old (D) WT mice and young (E) or old (F) Adrb2^−/−Adrb3^−/− mice. Arrowheads depict megakaryocytes with protrusions (CD41⁺CD42⁺ cells with cell body extensions). Scale bar, 250 μm. (G and H) Number of CD41⁺CD42⁺ megakaryocytes (G) forming protrusions (H) per BM area (n = 4 young WT; n = 5 young DKO; n = 3 old mice). (I–L) Representative immunofluorescence staining for CD41 (red) and EMCN (green) in femoral BM sections of young (I) or old (J) WT mice and young (K) or old (L) Adrb2^−/−Adrb3^−/− mice, depicting CD41⁺ megakaryocytes adjacent (arrowheads) or nonadjacent (asterisks) to EMCN⁺ vasculature. Scale bar, 100 μm. (M) Frequency of CD41⁺ cells in contact with EMCN⁺ vasculature (n = 4 young WT; n = 4 young DKO; n = 4 old WT; n = 3 old DKO). (N) Circulating platelets in young (n = 11) and old (n = 6) WT mice, compared with young (n = 9) and old (n = 5) Adrb2^−/−Adrb3^−/− mice. Young mice were analyzed between 8–30 weeks of age, and old mice were 66–120 weeks old. Data are means ± SEM. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001 (one-way ANOVA followed by Bonferroni pairwise comparisons).

Figure 4

β₂-AR Signaling in the Microenvironment Promotes Megakaryocyte Differentiation through IL-6 (A) Circulating platelets in young WT (n = 11), young Adrb2^−/− (n = 3), or old Adrb2^−/− (n = 3) mice. (B–C″) Frequency of (B, B′, and B″) CD41⁺ myeloid or megakaryocyte progenitors within lin⁻sca-1⁺c-kit⁺ (LSK) cells or (C, C′, and C″) CD150⁺CD41⁺ megakaryocyte progenitors (MkPs) within lin⁻c-kit⁺ (LK) cells in endosteal BM cells of the following mice: (B and C) young WT (B, n = 6; C, n = 3), young Adrb2^−/− (B, n = 6; C, n = 8) or old Adrb2^−/− (n = 4) mice; (B′ and C′) lethally irradiated WT recipients of WT (n = 5) or Adrb2^−/− (n = 4) BM cells; (B″ and C″) lethally irradiated WT (n = 5) or Adrb2^−/− (n = 4) recipients of WT BM cells. (D) Scheme of human umbilical-cord-blood-derived CD34⁺ HSPCs cocultured with MS-5 stromal cells. (E–G) Representative immunofluorescence (E and F) and number (G) of CD61⁺ (red) human megakaryocytes in cocultures treated with (E) vehicle or (F) β₂-AR agonist (clenbuterol, 10 μM) for 10 days (n = 3). Scale bar, 250 μm. (F′) Inset of (F). (H) IL6 concentration in endosteal (e) or non-endosteal (n-e) BM supernatant from adult WT (n = 6), Adrb2^−/− (n = 4), or Adrb3^−/− (n = 7) mice. (I) Il6 mRNA expression (fold change) in MS-5 stromal cells treated with β₂-AR agonist (clenbuterol, 10 μM), PKA inhibitor (H-89, 5 μM), or vehicle for 2 days (n = 3). (J–M) Quantification (J) and representative immunofluorescence (K–M) of CD41⁺ (red) CD42⁺ (green) megakaryocytes (yellow) in adult WT (n = 5), Adrb2^−/− (n = 3), or Il6^−/− (n = 5) mice. Scale bar, 250 μm. (N) Myeloid or megakaryocyte progenitors (CD41⁺ LSK cells) in primary BM culture from WT or Il6^−/− mice treated with β₂-AR agonist (clenbuterol, 10 μM) or vehicle for 4 days (n = 4). Data are means ± SEM. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001. (B′, B″, C′, C″, and G) Unpaired two-tailed t test. (A, B, C, H, I, J, and N) one-way ANOVA and Bonferroni pairwise comparisons.

Figure 5

Lack of β₃-AR and Nitric Oxide Synthase 1 (Nos1) Signaling in the Microenvironment Accelerates Aging (A, A′, and A″) Frequency of CD150^lo/−CD41⁻ lymphoid-biased HSCs in endosteal BM CD34⁻LSK cells from the following mice: (A) young WT (n = 7), young Adrb3^−/− (n = 7), or old Adrb3^−/− (n = 5) mice; (A′) lethally irradiated WT recipients of WT (n = 5) or Adrb3^−/− (n = 5) BM cells; and (A″) lethally irradiated WT (n = 4) or Adrb3^−/− (n = 3) recipients of WT BM cells. (B) Nitrate concentration in BM extracellular fluid (BMECF) of young WT (n = 6) or Adrb3^−/− (n = 7) mice. (C) Number of lin⁻sca-1⁺c-kit⁺CD34⁻CD48⁻CD150⁺Vwf-eGFP⁻ lymphoid-biased HSCs in primary BM culture from Vwf-eGFP mice treated with β₃-AR agonist (BRL37344, 10 μM), Nos1 inhibitor (L-VINO, 100 μM), or vehicle for 4 days (n = 8). (D–G) Representative immunofluorescence (D–F) and quantification (G) of c-kit⁺ (green) CD41⁺ (red) myeloid or megakaryocyte progenitors (arrowheads) in endosteal BM of WT (n = 10), Adrb3^−/− (n = 8), or Nos1^−/− (n = 4) mice. Endosteal BM is considered as regions within one-fifth marrow width from the bone surface. Scale bar, 300 μm. (H–K) Representative immunofluorescence (H-J) and quantification (K) of CD41⁺ (red) megakaryocytes adjacent (arrowheads) or nonadjacent (asterisks) to EMCN⁺ (green) BM vasculature of WT (n = 9), Adrb3^−/− (n = 6), or Nos1^−/− (n = 4) mice. Scale bar, 250 μm. (L–N′) Representative whole-mount immunofluorescence of CD31 (green) and EMCN (red) in WT (n = 7), Adrb3^−/− (n = 5), or Nos1^−/− (n = 4) BM. Arrowheads in insets (L′), (M′), and (N′) depict CD31^hiEMCN⁻ capillaries. (O and P) Quantification of (O) CD31^hiEMCN^hi transition zone vessels and (P) CD31^hiEMCN⁻ capillaries with <6 μm diameter. Scale bar, 250 μm. Data are means ± SEM. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001. (A′ and A″) Unpaired two-tailed t test. (A–C, G, K, O, and P) One-way ANOVA and Bonferroni pairwise comparisons.

Figure 6

Premature Hematopoietic Aging in HGPS Is Not HSC-Autonomous (A–D) Peripheral blood counts in adult (A and B) WT (Nos1^+/+; n = 15) or Nos1^−/− (n = 16), and (C and D) WT (Lmna^+/+; n = 22) or Lmna^G609G/G609G (n = 14) mice. (A, C, and K) Frequency of white blood cells (WBC); LYM, lymphocytes; MON, monocytes; NEU, neutrophils. (B, D, and L) Concentration of platelets. (E–G) Frequency of (E) CD11b⁺ myeloid cells, (F) B220⁺ B cells, and (G) CD3⁺ T cells in circulating leukocytes. (H–J) Frequency of (H) CD11b⁺ myeloid cells, (I) B220⁺ B cells, and (J) CD3⁺ T cells among donor-derived leukocytes 120 days after transplantation into WT mice (n = 11). (K and L) Peripheral white blood cell (K) and platelet (L) counts in adult CD45.1 C57BL/6J mice 4 months after lethal irradiation and transplantation with CD45.2 BM cells from WT (n = 8) or Lmna^G609G/G609G (n = 6) mice. (M–Q) Concentration of (M) IL-1β, (N) IL-6, (O) IL-1α, (P) IL-3, and (Q) IFNγ in the BM extracellular fluid (BMECF) of adult WT (Lmna^+/+; n = 9) and Lmna^G609G/G609G (n = 9) male mice. Data are means ± SEM. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001 (unpaired two-tailed t test).

Figure 7

β₃-AR Agonist Improves Lineage Skewing, HSC Number, and Localization near Megakaryocytes in HGPS (A and B) Frequency of (A) granulocytes and (B) lymphocytes in white blood cells (WBCs) of WT mice (n = 8) or Lmna^G609G/G609G mice (n = 6) treated with β₃-AR agonist (BRL37344, 2 mg/kg/day, intraperitoneally [i.p.]) or vehicle for 8 weeks. (C–F) Frequency of BM (C) Ly6G⁺CD11b⁺ neutrophils, (D) B220⁺ B cells, endosteal (E), and non-endosteal (F) LT-HSCs in these mice (n = 6). (G–J) Representative immunofluorescence (G–I) and quantification (J) of CD41⁺ megakaryocytes adjacent (asterisks) or nonadjacent (arrowheads) to c-kit⁺ (green) HSPCs in BM of WT mice (n = 3) and Lmna^G609G/G609G mice (n = 5) treated with β₃-AR agonist or vehicle. Scale bar, 50 μm. (K–N) Representative immunofluorescence (K–M) and distribution (N) of BM CD150⁺ (red) HSCs (negative for mature hematopoietic lineage markers, blue) adjacent (asterisks), or nonadjacent (arrows) to CD42⁺ (green) megakaryocytes. Scale bar, 50 μm. Data are means ± SEM. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001. (A–F and J) One-way ANOVA and Bonferroni pairwise comparisons. (N) Two-way ANOVA and Bonferroni pairwise comparisons.