Prolonged fasting reduces IGF-1/PKA to promote hematopoietic-stem-cell-based regeneration and reverse immunosuppression

Abstract

Immune system defects are at the center of aging and a range of diseases. Here, we show that prolonged fasting reduces circulating IGF-1 levels and PKA activity in various cell populations, leading to signal transduction changes in long-term hematopoietic stem cells (LT-HSCs) and niche cells that promote stress resistance, self-renewal, and lineage-balanced regeneration. Multiple cycles of fasting abated the immunosuppression and mortality caused by chemotherapy and reversed age-dependent myeloid-bias in mice, in agreement with preliminary data on the protection of lymphocytes from chemotoxicity in fasting patients. The proregenerative effects of fasting on stem cells were recapitulated by deficiencies in either IGF-1 or PKA and blunted by exogenous IGF-1. These findings link the reduced levels of IGF-1 caused by fasting to PKA signaling and establish their crucial role in regulating hematopoietic stem cell protection, self-renewal, and regeneration.

Figure 1. Prolonged fasting cycles protect the hematopoietic system and reverse the chemotherapy-induced hematopoietic suppression

(A) Diagrammatic representation of the experimental procedure to analyze the effects of prolonged fasting (PF, 48hr) during 6 cycles of cyclophosphamide chemotherapy (CP, 200mg/kg, i.p.). (B) Survival curve with vertical dashed lines indicating the pre-chemo starvation period; p<0.01, Log-rank (Mantel-Cox) test; n=20 (10 male and 10 female). (C) DNA damage measurement (olive tail moment) in bone marrow (BM)cells (day 81, 6th recovery phase). (D) Apoptosis measurement (TUNEL assay) in HSCs and MPP (day 81, 6th recovery phase). (E) Hematological profile of mice. Total white blood cell (WBC), lymphocyte counts and lymphoid/myeloid ratio (L/M) in mice treated with 6 cycles of CP (200mg/kg, i.p.). Each point represents the mean ± s.e.m; horizontal dashed lines indicate the ranges of baseline values; * p<0.05, Two-way ANOVA, comparing CP vs. PF+CP during the recovery phase, n=12 (6 male and 6 female); L/M ratio of peripheral blood (PB) is defined as number of lymphocytes divided by number of myeloid cells (i.e. granulocytes and monocytes). See also supplementary Figure S1F and S1G. (F) Hematological profile of human subjects. Lymphocyte counts and lymphoid/myeloid ratio (L/M) in patients undergoing two cycles (C1 and C2) of platinum-based doublet chemotherapy in combination with either 24hr or 72hr (48 before and 24 hours after chemo) prolonged fasting; D1 and D8 indicate the 1 day (before chemo) and 8 day of each chemotherapy cycle; Each point represents the mean ± s.e.m; ** p<0.01, Two-way ANOVA; Sample size is indicated in parentheses. (G) FACS analysis of hematopoietic stem and progenitor cells (day 84, end of 6^th cycle); horizontal dashed lines indicate the baseline value. (H) Proportion of the lymphoid-biased (Ly-HSC), balanced (Bala-HSC) and of the myeloid-biased (My-HSC) hematopoietic stem cells. The markers used are lower side population of LSK (lower-SP^LSK) for My-HSC, middle-SP^LSK for Bala-HSC and upper-SP^LSK for Ly-HSC. The lower panels show a magnification of the SP population in the upper panels.*p<0.05, One-way ANOVA comparing to AL. (I) BM cells collected from mice treated with either CP or PF+CP were transplanted into the recipient mice. The chimerism of donor-derived cells in PB and that in BM was determined 16 weeks after primary BM transplantation. The ratio of lymphocytes to myeloid cells (L/M) in the reconstituted blood was also measured. For (G and I), n= 6 to 10 per group, * p<0.05, ** p<0.01, t-test comparing the PF with the non-fasted control group both in combination with cyclophosphamide treatment.

Figure 2. Prolonged fasting cycles promote a chemotherapy-independent hematopoietic regeneration

Mice in the control group were fed ad libitum and those in the PF group were fasted for one or two cycles as indicated. n=4 to 12 female mice per group. (A) BrdU incorporation assay for LSK cells. Mice undergoing 24+48hr prolonged fasting were injected (i.p.) with BrdU (0.1mg/g, twice a day, for 2 days, starting after 24hr of fasting. (B) Number of long-term hematopoietic stem cells (LT-HSC), short-term hematopoietic stem cells (ST-HSC) and multipotent progenitors (MPP). (C) Number of common lymphoid progenitors (CLP) and myeloid progenitors (MP) (D) Cell cycle analysis for BM cells using Ki67 and Hoechst33342. (E) Apoptosis analysis for BM cells using TUNEL assay. For (A) to (E), * p<0.05, ** p<0.01, *** p<0.005, t-test comparing the AL-fed controls. (F) Proportion of the lymphoid-biased (Ly-HSC), balanced (Bala-HSC) and the myeloid-biased (My-HSC) hematopoietic stem cells. The markers used are lower side population of LSK (lower-SP^LSK) for My-HSC, middle-SP^LSK for Bala-HSC and upper-SP^LSK for Ly-HSC. (G) Number of lymphocytes and myeloid cells in young (6 months, 48hrs fasting) and old (18 months, 8 cycles of fasting) mice. For (F) and (G), * p<0.05, **p<0.01 and *** p<0.005, one-way ANOVA.

Figure 3. Deficiency in GHR/IGF-1 signaling promotes hematopoietic regeneration in both chemo-treated and untreated mice

Measurements were performed in GHRKO and their age matched littermates, with or without treatment with 6 cycles of CP (200mg/kg, i.p.). n=4 to 8 female mice per group. (A) BM IGF-1 level in GHRKO mice and PF mice compared to wild type mice fed ad libitum (WT-AL), * p<0.05, ** p<0.01, one-way ANOVA. (B) DNA damage measurement (olive tail moment) in BM cells and mononuclear peripheral blood cells (PB) from GHRKO and their littermates (WT) (day 81, 6th recovery phase). (C) Apoptosis measurement (TUNEL assay) in hematopoietic stem and progenitor cells (day 81, 6th recovery phase). (D) Number of hematopoietic stem and progenitor cells (day 84, end of 6^th cycle); horizontal dashed lines indicate the chemo-free baseline value. (E) Total white blood cell (WBC) and lymphocyte counts in PB of GHRKO mice and their littermates (WT); each point represents the mean ± s.e.m; vertical dashed lines indicate CP treatments; horizontal dashed lines indicate baseline value; * p<0.05, Two-way ANOVA for recovery phases; lymphoid/Myeloid ratio (L/M) after 6 cycles of CP treatments. PB L/M ratio is defined as the number of lymphocytes divided by the number of myeloid cells (i.e. granulocytes and monocytes). (F) Number of long-term hematopoietic stem cells (LT-HSC) and short-term hematopoietic stem cells (ST-HSC). (G) Cell cycle analysis using Ki67 and Hoechst33342. (H) Number of lymphocytes and myeloid cells in young (age 6 months) and old (age 18 months) mice. For (B to D) and (F to I) *p<0.05, **p<0.01; t-test comparing to the wild-type control.

Figure 4. Prolonged fasting promotes IGF-1/PKA dependent hematopoietic regeneration

(A) PKA-dependent phosphorylation of CREB visualized by ICC in mouse embryonic fibroblast (MEFs) devoid of endogenous IGF-1R (R- cells) or overexpressing human IGF1R (R⁺ cells). R+ cells were treated with IGF-1 and compared to cells transfected with PKACα siRNA. (B) Prolonged fasting (PF) reduces both circulating IGF-1 levels and PKA activity in BM cells in mice. IGF-1 injection blunted the PF-induced (C) reduction of PKA/pCREB,(D) increase in hematopoietic stem cells, (E-H) The chimerism of donor-derived cells in PB and that in the BM was determined 16 weeks after primary and secondary BM transplantation.. n=4 to 8 female mice per group, * p<0.05, **p<0.01 and *** p<0.005, one-way ANOVA.

Figure 5. The role of stromal niche in PF-induced HSC self-renewal

(A) Levels of the indicated proteins in BM stromal niche cells (Lin-CD45-). (B) Diagrammatic representation of the co-culture experiment. (C) Number of CD45+ progenies generated by the purified LT-HSCs exposed to the indicated niche cells. * p<0.05, **p<0.01 and ***p<0.005, t-test for (A) and one-way ANOVA for (C).

Figure 6. Reduction of IGF-1-PKA signaling promotes hematopoietic stem cell self-renewal

(A) Yeast cells (DBY746 background) overexpressing BCY1 (BCY1oe), which reduces PKA activity, or cells carrying mutations that activate PKA activity (bcy1CA1 and bcy1CA2) were grown in SDC for 3 days and treated with H₂O₂ (50 or 100mM) for 30min at 30°C. Cells were serially diluted and plated onto YPD plates. (B) PKA-regulated self-renewal pathways in PF mice. The levels of phosphorylation or expression of intracellular proteins in the indicated cellular populations and expression of indicated genes in total BM cells. BM cells were collected from mice with or without 48hr starvation (AL and PF). n=4 female mice per group, **p<0.01, *p<0.05, t-test. (C) Number of hematopoietic stem cells (per 5×10⁵ total BM) and progenitor cells (LT-HSC, ST-HSC and MPP) under the indicated treatments. * p<0.05, **p<0.01 and *** p<0.005, one-way ANOVA. See also Figure S7F and S7G. BM cells treated with PKA siRNA, IGF-1R siRNA or IGF-1 (versus non-treated cells) were transplanted into immuno-compromised recipient mice. (D) The engraftment in PB was measured at indicated time point after primary transplantation and the engraftment in BM was measured at the end of the 16 weeks after primary transplantation. n=4 to 8 female mice per group, *p<0.05 **p<0.01 and *** p<0.005, one-way ANOVA.

Figure 7. PF reduce IGF-1/PKA to promote lineage-balanced hematopoietic regeneration

(A) A simplified model for a partially conserved nutrient signaling PKA pathway in yeast and mammalian cells. Arrows show promotion actions and horizontal bars indicate inhibitory actions. GH, growth hormone; AC, adenylate cyclase; PKA, protein kinase A; CREB, cAMP response element-binding protein; Foxo1, Forkhead box protein O1; G9a, H3 Lys-9 methyltransferase. (B) A simplified model for PF-induced effects on WBC and HSCs. Fasting causes a major reduction in WBCs followed by their replenishment after re-feeding, based on effects on HSCs self-renewal resulting in increased progenitor and immune cells. These effects of PF can result in reversal of chemotherapy-based immunosuppression but also in the rejuvenation of the immune cell profile in old mice.