Abstract
Age-related declines in hematopoietic stem cell (HSC) function may contribute to anemia, poor response to vaccination, and tumorigenesis. Here, we show that mammalian target of rapamycin (mTOR) activity is increased in HSCs from old mice compared to those from young mice. mTOR activation through conditional deletion of Tsc1 in the HSCs of young mice mimicked the phenotype of HSCs from aged mice in various ways. These included increased abundance of the messenger RNA encoding the CDK inhibitors p16Ink4a, p19Arf, and p21Cip1; a relative decrease in lymphopoiesis; and impaired capacity to reconstitute the hematopoietic system. In old mice, rapamycin increased life span, restored the self-renewal and hematopoiesis of HSCs, and enabled effective vaccination against a lethal challenge with influenza virus. Together, our data implicate mTOR signaling in HSC aging and show the potential of mTOR inhibitors for restoring hematopoiesis in the elderly.
Introduction
Hematopoietic stem cells (HSCs) show decreased function with age; these functional deficits include reduced self-renewal, hematopoiesis, and differentiation into lymphocytes (1–5). The ensuing decrease in lymphopoiesis likely contributes to the weakened adaptive immune response characteristic of the elderly (5). Both cell-intrinsic and extrinsic mechanisms may contribute to these age-related changes in HSC function (1, 6–9); however, the underlying molecular pathways have not been elucidated.
The mammalian target of rapamycin (mTOR) pathway integrates multiple signals from nutrients, growth factors, and oxygen to regulate cell growth, proliferation, and survival (10–12). Here, we describe an increase in mTOR signaling in HSCs from aged mice and show that inhibition of mTOR signaling with rapamycin restores HSC function and enhances the immune response to influenza virus in old mice. Moreover, mTOR signaling has been shown to regulate the longevity of yeast (13), Caenorhabditis elegans (14), and Drosophila (15). The data herein and in a recent study (16) show increased life span in rapamycin-treated mice. Thus, involvement of mTOR activation in aging may represent a mechanism that is conserved in yeast, C. elegans, Drosophila, and mammals.
Results
The mTOR pathway in HSCs from aged mice is dysregulated
We isolated bone marrow (BM) cells from young (2 months old) and old (26 months old) C57BL/6 mice and analyzed them for surface markers to identify HSCs and for intracellular staining of phosphorylated mTOR (p-mTOR). Using flow cytometry, we found that the amount of p-mTOR was significantly increased in both HSC-enriched Lin−Sca-1+c-Kit+ (LSK) and the Flk2− lin−Sca-1+c-kit+ CD150+CD48−CD34− (FLSKCD150/48/34) HSCs from old mice compared to that in HSCs (Fig. 1A) from young mice. Consistent with the increase in phosphorylated mTOR, the abundance of the phosphorylated form of the mTOR complex 1 (mTORC1) substrate S6K and of the S6K substrate S6 was significantly increased in HSCs from old mice compared to that in HSCs from young mice (Fig. 1, B and C). These data indicate that the overall activity of mTOR in HSCs from old mice is greater than that in HSCs from young ones. To see whether this increase in mTOR phosphorylation was secondary to increased activity in the phosphoinositide 3-kinase (PI3K)–AKT signaling pathway, we evaluated AKT activation by measuring the abundance of AKT phosphorylated on Ser residue 473 (p-AKT) by flow cytometry. We found that the amount of p-AKT in HSCs from young and old mice was indistinguishable (Fig. 1D).
mTOR activity in young and aged HSCs. Fresh BM cells were isolated from either 2-month (young)– or 26-month (old)–old wild-type mice and stained with antibodies specific for surface markers to identify the Flk2− lin−Sca-1+c-kit+ CD150+CD48−CD34− (FLSKCD150/48/34) HSCs or the HSC-enriched lin−Sca-1+c-kit+ (LSK) cells, followed by intracellular staining with antibodies against p-mTOR (A), p-S6K (B), p-S6 (C), and p-AKT (D). The left panels show histograms of HSCs that are representative of five independent experiments each involving one mouse per group. Gray lines are negative controls [secondary antibodies only for (A), (B), and (D); isotype control for (C)], with filled gray areas depicting cells from old mice and dotted lines depicting cells from young mice; green lines depict profiles of stained BM cells from young mice and red lines depict profiles of BM cells from old mice. The mean fluorescence intensity (MFI) ± SD of all experiments is shown in the right panels. *P < 0.05; **P < 0.01. n.s., not significant.
Tsc1 deletion is sufficient to induce premature aging of HSCs
To determine whether increased mTOR signaling could explain the functional deficits of HSCs from old mice, we deleted the Tsc1 gene in the HSCs of young adult mice. Deletion of Tsc1, which encodes tuberous sclerosis complex (TSC) protein 1, leads to constitutive activation of mTOR in HSCs (17). The abundance of the messenger RNAs (mRNAs) encoding p16Ink4a, p19Arf, and p21Cip1 were all significantly increased in Tsc1−/− HSCs (Fig. 2A). To test the effect of mTOR signaling on the hematopoietic capacity of HSCs, we used the Mx1-Cre transgene for conditional deletion of the Tsc1flox/flox gene in HSCs after polyinosine–polycytidylic acid (pIpC) treatments. We transplanted, into lethally irradiated B6Ly5.2 recipients, 2 × 105 recipient-type (B6Ly5.1) BM cells in conjunction with 50 HSCs (FLSKCD150/48/34) isolated 10 days after pIpC treatment from either Tsc1flox/flox;Mx1-cre+ or Tsc1flox/flox;Mx1-cre− mice (Fig. 2B). Four weeks after transplantation with cells from wild-type donors, 30% of leukocytes in the peripheral blood of the recipients were derived from donor HSCs. The ratio of leukocytes derived from wild-type donor HSCs to recipient-type leukocytes increased to ~50% at 8 weeks. In contrast, Tsc1-deficient HSCs gave rise to only about 8% of the leukocytes present at 4 weeks, and by 16 weeks their contribution was barely detectable (Fig. 2C). Furthermore, leukocytes derived from the Tsc1-deficient HSCs showed markedly reduced ratios of B lymphocytes (B220+) to myeloid cells (Mac-1+ or Ter119+ or both) (Fig. 2D). The reductions in overall hematopoiesis and in lymphopoiesis are HSC-intrinsic because hematopoiesis from recipient-type cells was unaffected (Fig. 2E). Although Tsc1 deletion causes an immediate increase in HSC proliferation (17), self-renewal of the transplanted Tsc1−/− HSCs decreased over time (Fig. 3). The reduced 5-bromo-2′-deoxyuridine (BrdU) incorporation of the Tsc1−/− HSCs seen 280 days after transplantation recapitulates a feature of HSCs from old mice (18) (see also “Rapamycin restores HSC function in old mice” below).
Tsc1 deficiency and premature aging of HSCs from young mice. (A) Six-week-old Tsc1flox/flox (WT) and Tsc1flox/flox; Mx1-cre+ (KO) mice were treated seven times with pIpC every other day for 14 days and killed 10 days after the last treatment. The abundance of p16Ink4a, p19Arf, and p21Cip1 mRNAs in FACS-sorted HSCs was measured by real-time PCR and normalized to hypoxanthine-guanine phosphoribosyltransferase (HPRT). n = 5. (B to E) Fifty HSCs were mixed with 2 × 105 recipient-type total BM cells (WBM) and transplanted into lethally irradiated B6Ly5.2 recipients. The donor and recipient-type cells were identified by congenic CD45 markers. (B) Diagram of the experimental design for (C) and (D). (C) Hematopoiesis by WT and Tsc1−/− HSCs. The left panel shows representative dot plots of recipient peripheral blood 16 weeks after transplantation with WT (top) or Tsc1−/− cells (bottom). Data on the right summarize two independent experiments with a total of 10 recipients for each group. (D) Percentage of donor HSC-derived and B220+ (B) and Mac-1+ (M) cells in recipient peripheral blood 4 weeks after transplantation. (E) Percentage of B and M cells in total white blood cells. *P < 0.05; **P < 0.01; ***P < 0.001.
Targeted mutation of Tsc1 in BM cells resulted in transient increase but long-term reduction in HSC self-renewal. (A) Diagram of experimental design. (B) BrdU incorporation of HSCs from WT or KO donors after 24 hours of labeling. n = 4. (C) HSC numbers derived from WT or KO donors in the recipients after pIpC treatment. n = 4. *P < 0.05; ***P < 0.001; n.s., not significant.
Rapamycin restores HSC function in old mice
Because mTOR signaling is increased in HSCs from aged mice, we wondered whether inhibition of mTOR could reverse the functional deficits characteristic of HSCs from old mice. We treated old mice (22 months old) with either vehicle or the mTOR inhibitor rapamycin at a dose of 4 mg per kilogram of body weight every other day for 6 weeks and monitored their life span. We found that rapamycin treatment significantly extended the life span of aged mice (Fig. 4). Moreover, without affecting the overall cellularity of BM (Fig. 5A), rapamycin treatment reduced the percentage and the absolute number of HSCs (Fig. 5B).
Effect of rapamycin on life span of old mice. Age-matched 22- to 24-month-old male B6 mice were treated with vehicle or rapamycin (4 mg/kg body weight ip) every other day for 6 weeks. Survival of mice after the first rapamycin treatment was monitored daily. Death ratio refers to fraction of mice that were either dead or moribund, n = 10.
Rapamycin rejuvenates HSCs from old mice. (A) Bone marrow cellularity of tibiae and femurs of young (Y, 2 months old) mice, or age-matched, 22- to 24-month-old mice treated with either vehicle (O-veh) or rapamycin (O-rapa). n = 5. (B) The percentage of LSK and FLSKCD150/48/34 HSC populations in BM of Y, O-veh, and O-rapa mice. n = 5. (C) BrdU incorporation. Left panels show representative profiles of BrdU incorporation in HSCs, and right panels show mean ± SD of BrdU+ cells (n = 4). (D) HSCs were sorted by FACS, and p16Ink4a and p19Arf mRNA abundance was measured by real-time PCR and normalized to Hprt abundance; values shown are mean ± SD (n = 3). (E) Diagram of experiments. (F) Hematopoietic function of HSCs as measured by CD45 congenic markers on leukocytes of recipient peripheral blood at 4 weeks and 28 weeks after transplantation. The left panels show the percent of donor-type cells among leukocytes (CD45+) and the right panels show the reconstitution rates of B220+ (B), CD3+ (T), and Mac-1+ (M) lineages. Data shown were from three independent experiments with 15 recipients per group. *P < 0.05; **P < 0.01; ***P < 0.001.
These data suggest that the rapamycin-treated HSCs from old mice might have enhanced regenerative capacity. To test this notion, we pulsed young mice, vehicle-treated old mice, or rapamycin-treated old mice with BrdU for 52 hours starting 2 days after the last rapamycin treatment. We determined the proportion of BrdU+ cells among HSCs by flow cytometry. Only 5% of the HSCs from vehicle-treated old mice were BrdU+; however, pretreatment with rapamycin caused a significant increase in the proportion of BrdU-positive HSCs (16%) (Fig. 5C). Rapamycin pretreatment also decreased the expression of HSC p16Ink4a and p19Arf (Fig. 5D), which have been reported to contribute to aging of both HSCs and neuronal stem cells (18, 19).
To test the functional effects of rapamycin treatment on HSCs from old mice more directly, we transplanted 50 HSCs from old mice treated with vehicle or rapamycin into lethally irradiated young recipients, in each case in conjunction with 2 × 105 recipient-type BM cells. The efficacy of hematopoietic reconstitution by transplanted HSCs was assessed by the percentage of donor-type peripheral blood cells in the recipients (Fig. 5E). Consistent with previous studies, hematopoietic reconstitution by HSCs from old mice was decreased compared to that by HSCs from young mice (1, 20). However, both 4 weeks and 28 weeks after transplantation, the fraction of circulating leukocytes derived from rapamycin-treated HSCs from old mice was significantly increased compared to that derived from vehicle-treated HSCs (Fig. 5E). Thus, inhibition of mTOR activity by rapamycin enhances the in vivo regenerative capacity of HSCs from old mice and this enhancement is cell-autonomous.
Short-term rapamycin treatment boosts immunity of old mice
Old mice have impaired B cell generation; in particular, they show decreased numbers of pre–B cells (21, 22). We noted enhanced production of B lymphocytes and decreased myelogenesis in rapamycin-treated old mice (Fig. 6, A to C). Remarkably, rapamycin treatment doubled the percentage and the number of B220+ B cells in the BM (Fig. 6C). Rapamycin treatment also quadrupled the percentage of B220lowIgM− pre–B cells, which was very low in the vehicle-treated group (Fig. 6, A and B).
Vaccination-induced protection in old mice after vehicle or rapamycin treatment. (A to C) B6 mice (22 to 24 months old) were treated with vehicle or rapamycin (4 mg/kg body weight ip) every other day for 6 weeks. (A) Representative FACS files of BM cells from young (Y) and old mice treated with either vehicle (O-veh) or rapamycin (O-rapa). n = 5. (B) The percentage of immature B progenitor cells (B220lowIgM−) in whole BM, gated as in (A). (C) The percentage (left panel) and absolute number (right panel) of B220+ B lineage cells and Mac-1+ myeloid lineage cells in whole BM. (D to F) B6 mice (26 to 28 months old) were treated with vehicle or rapamycin (4 mg/kg body weight ip) every other day for 6 weeks. Two weeks later, young and vehicle- or rapamycin-treated old mice were immunized (Im) with 200 HAU of A/PR/8/34 virus by intraperitoneal injection. (D) The diagram of experimental design for (E) to (F). (E). Ten days after immunization, influenza-specific antibodies in the plasma of peripheral blood were measured by ELISA. (F) Two-weeks after immunization, mice were challenged with 400 HAU of A/PR/8/34 virus by intranasal delivery. Data shown are Kaplan-Meier survival analysis. n = 12. The P value is derived from a log-rank test. *P < 0.05; **P < 0.01; ***P < 0.001.
Because rapamycin treatment enhanced the generation of B cells in old mice, we hypothesized that it might also improve their immune response. To test this idea, we treated 26-month-old mice for 6 weeks with either vehicle or rapamycin. After pausing for 2 weeks to avoid the possible suppression of the immune response by rapamycin (23), we immunized the mice with 200 hemagglutinin units (HAU) of influenza virus intraperitoneally (ip) (Fig. 6D). Naïve and immunized young mice were used as controls. Ten days after immunization, we assessed the immune response by measuring the abundance of influenza virus–specific antibodies in the peripheral blood. We focused on immunoglobulin G (IgG) because it is more effective against influenza virus than IgM and IgA (24). As shown in Fig. 6E, pretreatment with rapamycin increased the amount of antigen-specific IgG in old mice by ~10 times for all isotypes of virus-specific IgG. Rapamycin treatment did not increase the percentage of T cells (fig. S1), and the percentage of influenza-specific CD8 T cells was higher in old mice than in young mice (fig. S2) regardless of rapamycin treatment. To test whether rapamycin pretreatment affects the response to vaccination, we challenged the vaccinated mice with the same strain of influenza virus at a dose that is lethal to unimmunized young mice. As shown in Fig. 6F, 9 out of 12 (75%) vehicle-treated old mice succumbed to infection even after immunization, whereas all mice that had been pretreated with rapamycin were protected by the vaccination.
Discussion
Together, our data show that the activity of the mTOR pathway is increased in HSCs from aged mice and that increasing mTOR signaling is sufficient to cause premature aging of HSCs in young mice. These data demonstrate a fundamental role for mTOR signaling in HSC aging and thus reveal a functional conservation in mechanisms that contribute to aging from yeast (13), C. elegans (14), and Drosophila (15) to mammals. Moreover, pharmaceutical inhibition of mTOR improved the regenerative capacity of HSCs from aged mice, showing that the functional capacity of these HSCs can be restored.
The underlying cause for overactivation of mTOR in the HSCs from the aging mice remains to be resolved. PTEN (phosphatase and tensin homolog deleted from chromosome 10) dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-trisphosphate (25), indirectly repressing AKT phosphorylation and thereby decreasing mTOR activity (26), and previous studies have shown that HSCs lacking PTEN show mTOR-dependent functional defects (27, 28). However, our data show that the amount of phosphorylated AKT in HSCs from young mice and that from old mice are substantially the same. Therefore, it is unlikely that overactivation of AKT is responsible for mTOR-dependent HSC aging.
The elderly mount a poorer immune response to influenza vaccination than do the young (29), and this may contribute to the significant morbidity and mortality caused by influenza virus in the elderly, even among vaccinated individuals (30). Here we report that mTOR inhibition reinitiated B cell development and enabled aged mice to mount a robust antibody response capable of protecting them against lethal influenza infection. Our data suggest that rapamycin, or possibly other inhibitors of mTOR, can be used to rejuvenate HSCs for better immune protection. We also found that 6 weeks of rapamycin treatment extended mouse life span, consistent with a recent study that showed that continuously feeding old mice with rapamycin extended their life span (16). Although rapamycin has been used as an immune suppressant in transplantation (23), treatment with rapamycin during infection can paradoxically induce a more effective memory T cell response (31).
Materials and Methods
Mice
Old C57BL/6 wild-type mice were obtained from National Institute of Aging. Tsc1fl/fl mice (32) were provided by D.J. Kwiatkowski (Brigham and Women’s Hospital, Boston, MA) and were backcrossed for at least six generations onto the C57BL/6 background. Mx1-cre mice were purchased from the Jackson Laboratory. Recipients in reconstitution assays were 8-week-old B6ly5.2 mice from National Cancer Institute. All the mice were kept in the Unit of Laboratory Animal Facility (ULAM) at University of Michigan, Ann Arbor. All procedures involving experimental animals were approved by the University Committee on the Use and Care of Animals at the University of Michigan.
pIpC and rapamycin treatment
pIpC (Sigma-Aldrich) was resuspended in 1× PBS at 1 mg/ml. Mice received pIpC (15 mg/kg ip) every other day seven times. Rapamycin (LC Laboratories) was reconstituted in ethanol at 10 mg/ml and then diluted in 5% Tween-80 (Sigma-Aldrich) and 5% PEG (polyethylene glycol) 400 (Hampton Research). Mice received rapamycin (4 mg/kg ip) every other day for 6 weeks (17).
Transplantation assays
Eight-week-old congenic recipient mice were lethally irradiated with a cesium-137 x-ray source delivering 0.97 Gy/min (1 Gy = 100 rads). A combined dose of 11.50 Gy was delivered in two installments 4 hours apart. Given numbers of donor BM cells were transplanted into recipients through the retro-orbital venous sinus within 24 hours of irradiation. For HSC transplant, FLSKCD150/48/34 HSCs were sorted by FACSAria (BD Biosciences) and mixed with the indicated number of recipient-type BM cells. Reconstitutions were measured by flow cytometry of peripheral blood from the recipient tail vein at the time points indicated.
Flow cytometry
BM cells were flushed out from the long bones (tibiae and femurs) by a 25-gauge needle with 1× Hanks’ balanced saline solution without calcium or magnesium (Invitrogen), supplemented with 2% heat-inactivated fetal bovine serum. Peripheral blood was obtained from the tail veins of recipients at the time points indicated in the figures, and red blood cells were lysed by ammonium chloride–potassium bicarbonate buffer before staining. For flow cytometry and purification of HSCs, the immunophenotype FLSKCD150/48/34 was used for long-term HSCs as indicated. Lineage markers included B220 (B cells), CD3 (T cells), Gr-1 (granulocyte), Mac-1 (myeloid cells), and Ter119 (erythrocytes). CD150 was from BioLegend, Inc. All other antibodies were obtained from BD Biosciences. For intracellular staining, cells were first stained with the indicated surface markers and then fixed with Fix buffer (BD Biosciences) for 2 hours at 4°C, followed by incubation with BD cytoperm+ buffer for 10 min at room temperature and refix for 10 min. p-mTOR (Ser2448), p-S6K (Thr389), and AKT (Ser473) antibodies (Cell Signaling Technology) were diluted at 1:100 and Alexa Fluor 488–conjugated pS6 (pSer235/236) antibodies (Cell Signaling Technology) were diluted at 1:10 and incubated overnight at 4°C. Fluorescein isothiocyanate–conjugated secondary antibodies (Jackson Immunoresearch) were diluted at 1:100 and incubated for 2 hours at 4°C. Flow cytometry analysis was performed on an LSR II (BD Biosciences).
BrdU incorporation
BrdU (Sigma-Aldrich) was injected intraperitoneally into adult mice at 100 mg/kg body weight. Mice were then given BrdU water (1 mg/ml) for 24 (Fig. 3) or 52 (Fig. 5) hours as indicated before being killed for analysis. BrdU staining kit (BD Biosciences) was used according to the producer’s manual.
Real-time polymerase chain reaction
HSCs were purified by fluorescence-activated cell sorting (FACS) and RNA was isolated with Trizol (Invitrogen) and further purified by RNeasy kit (Invitrogen). Complementary DNA was made from purified RNA with Superscript III (Inivitrogen). Real-time polymerase chain reaction (PCR) was performed on a 7500 Real Time PCR system (Applied Biosystems).
Immunization and ELISA
Mice were immunized with 200 HAU influenza A/PR/8/34(H1N1) (Charles River Laboratories) by intraperitoneal injection. Ten days after immunization, peripheral blood was collected and plasma was used for enzyme-linked immunosorbent assay (ELISA). To test the relative concentrations of influenza-specific antibodies, 96-well ELISA plates were coated with inactivated and purified influenza A/PR/8/34(H1N1) virus (0.4 μg/ml; Charles River Laboratories). Plasma was serially diluted by 1:100, 1:1000, and 1:10000. Secondary antibodies were horseradish peroxidase–conjugated goat anti-mouse IgG1, IgG2a, IgG2b, and IgG3 (Southern Biotechnology Associates), respectively, diluted at 1:1000. Color was developed by BD OptEIA (BD Biosciences) for 15 min for IgG2a and IgG2b, 30 min for IgG1 and IgG3 and terminated by acid. Two weeks after immunization, mice were challenged with 400 HAU H1N1 influenza virus through intranasal delivery.
Flow cytometric measurement of influenza-specific CD8 T cells
B6 mice (26 to 28 months old) were pretreated with vehicle or rapamycin (4 mg/kg body weight ip) every other day for 6 weeks. Two weeks later, mice were immunized with 200 HAU PR/8/34 influenza virus by intraperitoneal injection. Two weeks after immunization, peripheral blood leukocytes were harvested to determine the percentage of CD8 T cells that reacted with influenza nucleoprotein peptide (Ala-Ser-Asn-Glu-Asn-Met-Glu-Thr-Met)–loaded Db. An unrelated peptide (Leu-Pro-Tyr-Leu-Gly-Trp-Leu-Val-Phe)-loaded H-2Ld was used as negative control. The tetramers were prepared by the tetramer facility of the National Institutes of Health.
Statistics
The function of HSCs in the rescue of lethal irradiation was compared by a Kaplan-Meier survival analysis, and the P value of the log-rank tests are provided either in the figures or in the figure legends. Student’s t tests were used for all other analyses.
Acknowledgments
We thank R. Miller for critical reading of the manuscript. This study is supported by grants from American Cancer Society (RSG-06-072-01-TBE), the US Department of Defense (W81XWH-07-1-0169 and W81XWH-08-1-0036), and NIH (CA112001 and AG024824). The aged animals were provided through the Pepper Center’s Core Facility for Aged Rodents. The authors have no financial conflict of interest.
Supplementary Materials
www.sciencesignaling.org/cgi/content/full/2/98/ra75/DC1
Fig. S1. Percent of T cells in the peripheral blood of old mice two weeks after vehicle- or rapamycin-treatment.
Fig. S2. Clonal expansion of influenza-specific CD8 T cells after immunization with influenza virus, as measured by % of CD8+tetramer+ T cells in the peripheral blood.
