FOXO3A directs a protective autophagy program in haematopoietic stem cells

Abstract

Blood production is ensured by rare, self-renewing haematopoietic stem cells (HSCs). How HSCs accommodate the diverse cellular stresses associated with their life-long activity remains elusive. Here we identify autophagy as an essential mechanism protecting HSCs from metabolic stress. We show that mouse HSCs, in contrast to their short-lived myeloid progeny, robustly induce autophagy after ex vivo cytokine withdrawal and in vivo calorie restriction. We demonstrate that FOXO3A is critical to maintain a gene expression program that poises HSCs for rapid induction of autophagy upon starvation. Notably, we find that old HSCs retain an intact FOXO3A-driven pro-autophagy gene program, and that ongoing autophagy is needed to mitigate an energy crisis and allow their survival. Our results demonstrate that autophagy is essential for the life-long maintenance of the HSC compartment and for supporting an old, failing blood system.

Figure 1. HSCs induce autophagy following metabolic stress

a, b, Fluorescent microscopy of cultured Gfp-Lc3 GMPs (a) and HSCs (b). c, Electron microscopy images of WT HSCs. Arrowheads indicate autophagic vesicles. d, GFP-LC3 loss in cultured Gfp-Lc3 HSCs (n = 3). Results are expressed as percentage of GFP-LC3 MFI in -BafA compared to +BafA conditions. e, GFP-LC3 levels in HSCs of fed and starved Gfp-Lc3 mice (n = 3). Results are expressed as percent of GFP-LC3 MFI in fed mice. Data are means ± s.d. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Figure 2. Autophagy protects HSCs from starvation-induced apoptosis

a, Apoptosis levels in cultured WT HSCs (n = 3–5). Results are expressed as percent caspase activation compared to +cytokines conditions. b, Apoptosis levels in cultured control (Cnt) and Atg12^cKO HSCs and GMPs (n = 3). Results are expressed as percent caspase activation compared to +cytokines Cnt HSCs (100%). c, Absolute number of HSCs in fed and starved Cnt and Atg12^cKO mice (n = 4–8 mice per group). Results are expressed as percent of HSCs in fed mice. Data are means ± s.d. Hatching indicates starved conditions. *P ≤ 0.05, **P ≤ 0.01.

Figure 3. FoxO3a poises HSCs for rapid autophagy induction

a, Status of the autophagy machinery in WT HSCs (n = 3). Results are expressed as log2 fold expression compared to WT GMPs (set to 0). b, c, qRT-PCR analyses of pro-autophagic genes in WT (b) and FoxO3a^−/− (c) HSCs (n = 3). Results are expressed as log2 fold expression compared to WT GMPs (b) or FoxO3a^+/+ HSCs (c) (set to 0). d, Autophagy flux in cultured FoxO3a^−/−::Gfp-Lc3 HSCs (n = 3). Results are expressed as percent GFP-LC3 MFI in -BafA vs. +BafA conditions, and normalized to FoxO3a^+/+−::Gfp-Lc3 HSCs. Data are means ± s.d. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Figure 4. Ongoing autophagy in old HSCs

Electron microscopy images of young and old HSCs. Arrowheads indicate autophagic vesicles.

Figure 5. Ongoing autophagy is essential for the continued survival of old HSCs

a, qRT-PCR analyses of pro-autophagic genes in old HSCs (n = 3–5). Results are expressed as log2 fold expression compared to young HSCs (set to 0). b, Apoptosis levels in cultured young and old HSCs (n = 3). Results are expressed as percent caspase activation compared to +cytokines young HSCs. Hatching indicates –cytokines conditions. c, 2-NBD glucose uptake in cultured young and old HSCs (n = 3). d, Percent colony formation in young and old WT HSCs cultured for 32h ± BafA and methylpyruvate (M-Pyr) (n = 3). Colonies were counted at day 10 and normalized to -BafA conditions. Data are means ± s.d. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.