nature aging

Letter https://doi.org/10.1038/s43587-023-00473-3

Induction of mitochondrial recycling reverts

age-associated decline of the hematopoietic
and immune systems

Received: 8 September 2021
Accepted: 24 July 2023
Published online: 31 August 2023
Mukul Girotra1,8, Yi-Hsuan Chiang 1,8, Melanie Charmoy2, Pierpaolo Ginefra 1
,
Helen Carrasco Hope 1, Charles Bataclan 3, Yi-Ru Yu 1, Frederica Schyrr 3
,
Fabien Franco1, Hartmut Geiger4, Stephane Cherix5, Ping-Chih Ho1,
Olaia Naveiras3,6, Johan Auwerx 7, Werner Held2 & Nicola Vannini 1

Check for updates

Aging compromises hematopoietic and immune system functions, making
older adults especially susceptible to hematopoietic failure, infections and
tumor development, and thus representing an important medical target for
a broad range of diseases. During aging, hematopoietic stem cells (HSCs)
lose their blood reconstitution capability and commit preferentially toward
the myeloid lineage (myeloid bias)1,2. These processes are accompanied by
an aberrant accumulation of mitochondria in HSCs3. The administration of
the mitochondrial modulator urolithin A corrects mitochondrial function
in HSCs and completely restores the blood reconstitution capability of
‘old’ HSCs. Moreover, urolithin A-supplemented food restores lymphoid
compartments, boosts HSC function and improves the immune response
against viral infection in old mice. Altogether our results demonstrate
that boosting mitochondrial recycling reverts the aging phenotype in the
hematopoietic and immune systems.

Hematopoietic stem cells (HSCs) generate all the blood and immune
cells throughout the entire life of an organism and they ensure the cor-
rect homeostasis between lymphoid and myeloid lineages. However,
aging dramatically reduces the blood reconstitution capability of HSCs
and skews their fate toward myeloid lineages to the detriment of lym-
phoid cell production1,2. In fact, specific age-related dysfunctions of
the immune system stem from the deteriorated HSC pool4,5. Thus, it has
been proposed that reversing HSC aging would improve both hemato­
poietic and immune functions6. HSC aging is coupled with increased
DNA damage due to an impairment of DNA repair enzymes7,8 and epi-
genetic changes9,10. Interestingly, aging has also been directly linked
to metabolic alterations in HSCs11. These studies led to the emerging
concept that cellular metabolism operates as a potent regulator of the
HSC pool12–14, and more broadly, of stem cell function and fate11,15,16.
Defective autophagy and consequent accumulation of mitochondria
in HSCs has been associated with aging processes3, providing evidence
that cellular quality control processes are crucial for the maintenance
of a functional HSC pool both in homeostasis and in stress condi-
tions17–19. Even though HSCs are refractory to systemic rejuvenating
approaches20, nutritional interventions, such as prolonged fasting and
nicotinamide adenine dinucleotide (NAD) boosting strategies, have
been demonstrated to partially restore HSC function in old mice21,22.
In particular, NAD precursors have been shown to exert their effect by
directly modifying HSC mitochondrial function18,22.

Department of Oncology, Ludwig Institute for Cancer Research, University of Lausanne, Epalinges, Switzerland. 2Department of Oncology, University of
1

Lausanne, Epalinges, Switzerland. 3Laboratory of Regenerative Hematopoiesis, Department of Biomedical Sciences, University of Lausanne and ISREC,
School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland. 4Institute of Molecular Medicine, Ulm University, Ulm,
Germany. 5Orthopedic and Traumatology Service, Lausanne University Hospital, Lausanne, Switzerland. 6Hematology Service, Department of Oncology,
Lausanne University Hospital, Lausanne, Switzerland. 7Laboratory of Integrative and Systems Physiology, Institute of Bioengineering, Ecole Polytechnique
Fédérale de Lausanne, Lausanne, Switzerland. 8These authors contributed equally: Mukul Girotra, Yi-Hsuan Chiang. e-mail: nicola.vannini@unil.ch

Nature Aging | Volume 3 | September 2023 | 1057–1066 1057

Letter https://doi.org/10.1038/s43587-023-00473-3

a Primary transplant Secondary transplant

Spleen Spleen
Competitor 0 4 . . 24 analysis
Ex vivo culture 0 4 . . 24 analysis
± UA
BM
Recipient BM Secondary
Old mouse analysis
LKSCD150+CD48− analysis recipient
(>100 weeks)

b Primary transplant c
Total Lymphoid Myeloid BM Spleen
80 60
***
100 ******* 125 NS 125 NS
*** *** *** **
*** *** **** *** **

Donor chimerism
**
Donor chimerism

60 *** *** *** 80 *** 100 100

*
40
60 75 75 Young control
40 Old control
40 50 50
20 Old UA
20 Young control 25 25
20 Old control
Old UA 0 0
0 0 0
0 4 8 12 16 20 24 0 4 8 12 16 20 24 0 4 8 12 16 20 24
Weeks Weeks Weeks

d Secondary transplant e Spleen BM

Total Lymphoid Myeloid Lymphoid Myeloid LKS
100 80 0.06 100 125 NS 125 125 NS
0.07
* * ** *
** *
NS
0.05 NS NS
** *

Donor chimerism
Donor chimerism

80 * * * 0.06 60
80 NS 100 100
** *
100

60 60 75 75 75
40
40 40 50 50 50
20 Young control 25 25 25
20 20
Old control
0 0 0 Old UA 0 0 0
0 4 8 12 16 20 0 4 8 12 16 20 0 4 8 12 16 20
Weeks Weeks Weeks

Fig. 1 | UA in vitro treatment reverses age-associated defects in HSCs.
a, HSCs isolated from old mouse (100 weeks) BM were cultured for 3 days in the
presence of 20 μM UA. After culture, 2,000 cells were transplanted in lethally
irradiated primary recipient mice together with BM derived from competitor
mice. At the endpoint, BM from primary recipient mice was transplanted into
lethally irradiated secondary recipient mice. b, Blood donor chimerism analyses
of primary transplant at the indicated time points (total: 4 weeks P = 0.0001;
8 weeks P = 0.0002; 12 weeks P = 0.0001; 16 weeks P = 0.0001; 24 weeks P = 0.0001;
lymphoid: 8 weeks P = 0.0023; 12 weeks P = 0.0009; 16 weeks P = 0.0006;
24 weeks P = 0.0006; myeloid: 4 weeks P < 0.0001; 8 weeks P = 0.0001; 12 weeks
P = 0.0011; 16 weeks P = 0.0007; 24 weeks P = 0.0007). c, BM and spleen donor
chimerism at the endpoint (BM: young control versus old control P < 0.0001;
old control versus old UA P = 0.0003; spleen: young control versus old control
P < 0.0001; old control versus old UA P = 0.0003). d, Blood donor chimerism of
secondary transplant at the indicated time points (total: 4 weeks P = 0.0181;
8 weeks P = 0.0352; 12 weeks P = 0.0460; 16 weeks P = 0.0584; 20 weeks P = 0.0504;
lymphoid: 4 weeks P = 0.006; 8 weeks P = 0.0455; 12 weeks P = 0.0669; 16 weeks
P = 0.0611; 20 weeks P = 0.0732; myeloid: 4 weeks P = 0.0414; 8 weeks P = 0.0352;
12 weeks P = 0.1019; 16 weeks P = 0.1133; 20 weeks P = 0.0872). e, Spleen analyses
of donor chimerism at the endpoint and analysis of donor-derived stem and MPP
population LKS (spleen lymphoid: young control versus old control P = 0.0034;
old control versus old UA P = 0.0231; spleen myeloid: young control versus old
control P = 0.0058; old control versus old UA P = 0.0404; LKS: young control
versus old control P < 0.0055; old control versus old UA P = 0.0156). n = 10; values
are the mean ± s.e.m., two-tailed Student’s t-test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001,
****P ≤ 0.0001. NS, not significant.

In this context, urolithin A (UA), a natural compound belonging secondary recipient mice, and donor chimerism analysis was followed
to the ellagitannin family23, has been shown to improve metabolic over a period of 20 weeks (Fig. 1a). The UA effect was maintained dur-
fitness in humans and rodents, and prolong the life span in worms ing secondary transplantation, where UA-treated old HSCs sustained
by sustaining mitophagy and mitochondrial function24–27. Because an identical blood reconstitution capability of young HSCs (Fig. 1d), pro-
HSCs have defective autophagy and accumulate mitochondria during duced a comparable donor-derived chimerism in primary and secon­
aging3,28 (Extended Data Fig. 1), we tested the capacity of UA to recover dary lymphoid tissues (Fig. 1e and Extended Data Fig. 2b), restored
the functions of HSCs purified from mice older than 18 months (‘old multipotent stem and progenitor population (Lin−c-Kit+Sca1+ (LKS))
HSCs’). Purified old HSCs (Lin−Sca1+c-Kit+CD48−CD150+) cultured for (Fig. 1e) and pure HSC pool (LKS CD150+CD48−) in the BM (Extended
3 days in the presence of 20 μM UA were transplanted into lethally irradi­ Data Fig. 2d). Similarly, UA treatment improved the hematopoietic
ated recipient mice and blood chimerism was assessed over a period of stem and progenitor activities of CD34+CD38−CD45RA− human HSCs
24 weeks by exploiting the CD45.1 and CD45.2 double congenic allelic purified from old adults in a colony-forming assay (Extended Data
system (Fig. 1a). Comparison of the blood reconstitution capabilities Fig. 3a), as evidenced by improved colony-forming capacity both in
revealed that exposure to UA restored the reconstitution function of primary (Extended Data Fig. 3b) and long-term secondary colony-
old HSCs to levels seen in HSCs purified from young mice (8–12 weeks forming assay, as measured by colony-forming units (CFUs) (Extended
old; ‘young HSCs’) (Fig. 1b) without significantly affecting the lym- Data Fig. 3c). Interestingly, UA treatment enhances the reconstitution
phoid and myeloid proportion during blood recovery (Extended Data capability of young HSCs even though to a minor extent compared to
Fig. 2a). The functional restoration of old HSCs was further confirmed by old HSCs (Extended Data Fig. 2e). Taken together, these data highlight
an increased donor chimerism in primary and secondary lymphoid tis- that a short in vitro exposure (3 days) to UA restores the long-term
sues, bone marrow (BM) and spleen, respectively (Fig. 1c and Extended blood reconstitution capability of old HSCs.
Data Fig. 2b). To ascertain the long-term HSC gain of function, BM To study whether UA recovers hematopoietic and immune
from primary recipient mice was transplanted into lethally irradiated functions when delivered systemically in vivo, we fed mice with

Nature Aging | Volume 3 | September 2023 | 1057–1066 1058

Letter https://doi.org/10.1038/s43587-023-00473-3

a Control diet
Competitor
UA diet
Weeks 0 16 Spleen Weeks 0 4 .. 20
analysis
BM Recipient Secondary
analysis recipient
Old

b Cellularity HSCs MPP CLPs

60 0.8 0.8 0.3
*
No. of cells (106)
50
*
0.6 0.6

Percentage
Percentage

Percentage
40 0.2

30 0.4 0.4

20 0.1
0.2 0.2
10
0 0 0 0
BM

CMP GMP MEPs

0.8 1.6 1.5

0.6
*

Percentage
Percentage

1.4 1.0

0.4
Old control
1.2 0.5 Old UA
0.2

0 1.0 0

c
Cellularity CD3+ 1.5 CD4+ 0.5 CD8+
30
2.0
No. of cells (106)

No. of cells (106)

0.4
No. of cells (106)

No. of cells (106)

20 1.5 1.0
Spleen

0.3
1.0
0.2
10 0.5
0.5 0.1

0 0 0

d Total Lymphoid Myeloid

100 *** 100 125
** ** *** **** ** ** *******
** ** *** *** ***
Donor chimerism
Donor chimerism
Donor chimerism

80 80 100
**
60 60 75

40 40 50

20 Old control 20 Old control 25 Old control

Old UA Old UA Old UA
0 0 0
0 4 8 12 16 20 0 4 8 12 16 20 0 4 8 12 16 20
Weeks Weeks Weeks
e Control diet
Competitor
UA diet
Weeks 0 16
Weeks 0 4 .. 16

Recipient
200 HSCs
Old LKS CD150+CD48–

f g Proportion of lymphoid and myeloid cells

50 ** 0.059
* * 0.068 0.14
Donor chimerism (%)

0.068
Percentage of cells

40 100

30

20 50 Myeloid
Young control
Lymphoid
10 Old control
Old UA
0 0
0 4 8 12 16
Young
Old
Old + UA

Young
Old
Old + UA

Young
Old
Old + UA

Young
Old
Old + UA

Weeks

4 weeks 8 weeks 12 weeks 16 weeks

a UA-enriched diet (50 mg kg day−1) for a period of 4 months, char- which was monitored monthly, was not different between UA-treated
acterized hematopoietic stem and progenitor pools and analyzed and untreated groups (Extended Data Fig. 4). However, analyses of
immune compartments (Fig. 2a). The cellular composition of blood, BM hematopoietic stem and progenitor populations (Extended

Nature Aging | Volume 3 | September 2023 | 1057–1066 1059

Letter
Fig. 2 | UA oral supplementation rejuvenates the hematopoietic system of old
mice. a, Old mice (81–84 weeks old) were fed with a control or UA-supplemented
diet (50 mg kg day−1) for 16 weeks. Mice were euthanized to analyze the spleen
and BM. A part of the BM was transplanted into lethally irradiated primary
and secondary recipient mice. b, Analysis of different stem and progenitor
populations of the BM using flow cytometry (HSCs: P = 0.0147; CLPs: P = 0.0492;
P = 0.0298). n = 5; values are the mean ± s.e.m.; two-sided Student’s t-test;
*P ≤ 0.05. c, Analysis of different T cell compartments in the spleen using flow
cytometry (n = 5; values are the mean ± s.e.m.). d, Blood donor chimerism
analyses of the primary transplant at the indicated time points (total: 4 weeks
P = 0.0011; 8 weeks P = 0.0017; 12 weeks P = 0.0008; 16 weeks P = 0.0001;
20 weeks P < 0.0001; lymphoid: 4 weeks P = 0.0059; 8 weeks P = 0.0018;
Data Fig. 5a) revealed an expansion of HSCs and common lymphoid
progenitor (CLP) cells (Lin−CD127+Sca1lowc-Kitlow) and a reduction of
megakaryocyte erythroid progenitor (MEP) cells (Lin−c-Kit+Sca1−
CD16/32−CD34−) (Fig. 2b), while no changes were observed in common
myeloid progenitor (CMP) cells (Lin−c-Kit+Sca1−CD16/32−CD34+),
granulocyte monocyte progenitor (GMP) cells (Lin−c-Kit+Sca1−CD16/32+
CD34−) and multipotent progenitor (MPP) cells (Lin−Sca1+c-Kit+
CD48−CD150−) (Fig. 2b). Indeed, during aging CLPs are reduced while
megakaryocyte progenitors expand29; thus, UA supplementation
reverted the aging-associated features of the hematopoietic system.
Interestingly, CLP expansion correlated with a partial expansion of lym-
phocytes in the spleen (Fig. 2c). To test the BM functionality of old mice
supplemented with UA, we transplanted whole BM into lethally irradi-
ated mice (Fig. 2a). Donor chimerism analyses revealed that BM derived
from UA-treated old mice had a higher reconstitution capacity com-
pared to the untreated group (Fig. 2d) and no significant changes in the
lymphoid and myeloid proportion during blood recovery (Extended
Data Fig. 5b). Importantly, the improved reconstitution capability
of HSCs was maintained on secondary transplantation (Extended
Data Fig. 5c). These data indicate that UA dietary supplementation
boosts hematopoietic stem and progenitor functions and restores
the CLP population in old mice. Moreover, to assess the cell-intrinsic
effect of UA feeding on HSC function, a fixed number of HSCs purified
from UA-fed mice were transplanted in lethally irradiated mice and
blood chimerism was monitored over time (Fig. 2e). HSCs purified
from UA-supplemented old mice had better blood reconstitution
capability, with an efficiency comparable to HSCs purified from young
mice (Fig. 2f) and restored lymphoid lineage production (Fig. 2g). Taken
together, these results indicate that UA supplementation improves the
cell-intrinsic function of old HSCs.
To evaluate whether the boost in hematopoietic capacity
improves immune function, old mice supplemented for 8 weeks with a
UA-enriched diet were infected with lymphocytic choriomeningitis
Fig. 3 | UA supplementation improves the CD8+ T cell response to viral
infection in old mice. a, Old mice (84 weeks old) were fed with a control or a
UA-supplemented diet (50 mg kg day−1) for 8 weeks before infection with LCMV
(Armstrong strain). Mice were euthanized 8 days later and spleen cells were
analyzed. b, Gating strategy for virus-specific CD8+ T cells and their immune
profile. FSC-A, forward scatter-area. c, Quantification of CD8+ T cells specific
for the gp33 (gp33+) or np307 (np396+) epitope (gp33+: young control versus
old control P = 0.049; young control versus old UA P = 0.0007; np396+: young
control versus old control P < 0.0001; young control versus old UA P = 0.0002;
old control versus old UA P = 0.0222). n = 10; values are the mean ± s.e.m.;
one-way analysis of variance (ANOVA); **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.
d, Gated gp33+ and np396+ CD44+CD8+ T cells were analyzed for KLRG1
expression (gp33+: young control versus old control P < 0.0001; young control
versus old UA P = 0.0002; old control versus old UA P = 0.0007; np396+: young
control versus old control P < 0.0001; young control versus old UA P = 0.0002;
old control versus old UA P = 0.0222). n = 10; values are the mean ± s.e.m.;
one-way ANOVA; **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. e, Analysis of GzmB
expression in virus-specific CD8+ T cells (gp33+: young control versus old control
Nature Aging | Volume 3 | September 2023 | 1057–1066
https://doi.org/10.1038/s43587-023-00473-3
12 weeks P = 0.0010; 16 weeks P = 0.0002; 20 weeks P < 0.0001; myeloid: 4 weeks
P = 0.0021; 8 weeks P = 0.0094; 12 weeks P = 0.0009; 16 weeks P = 0.0003;
20 weeks P = 0.0002). n = 5; values are the mean ± s.e.m.; two-sided Student’s
t-test; **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. e, Two hundred HSCs purified
from young, old and UA-fed old mice were transplanted into lethally irradiated
recipient mice. f, Blood donor chimerism analyses of the primary transplant
at the indicated time points (4 weeks P = 0.0023; 8 weeks P = 0.068; 12 weeks
P = 0.0599; 16 weeks P = 0.0455). n = 5; values are the mean ± s.e.m.; one-tailed
Student’s t-tests; *P ≤ 0.05, **P ≤ 0.01. g, Analyses of donor lymphoid versus
myeloid compartments of primary transplants at the indicated time points
(8 weeks P = 0.0259; 12 weeks P = 0.068; 16 weeks P = 0.14). n = 5; values are the
mean ± s.e.m.; one-tailed Student’s t-tests; *P ≤ 0.05.
virus (LCMV) Armstrong strain, which resulted in an acute resolved
infection mediated by a CD8+ T cell response. The specific immune
response against the virus was assessed at day 8 after infection (Fig. 3a,b).
Aging strongly reduced the abundance of CD8+ T cells specific for the
gp33 and np396 viral epitopes (Fig. 3a,b)30,31. UA supplementation
increased the presence of np396-specific CD8+ T cells (Fig. 3c). Impor-
tantly, UA treatment improved the effector differentiation of old CD8+
T cells as indicated by an increase in KLRG1+ cells (Fig. 3d) and granzyme
B (GzmB) expression by gp33- and np396-specific cells and increased
interferon-γ (INFγ) expression by np396 cells (Fig. 3e,f) (Extended Data
Fig. 6a). Moreover, UA-treated, virus-specific CD8+ T cells expressed
reduced levels of the inhibitory programmed cell death protein 1 (PD-1)
and lymphocyte-activation gene 3 (LAG-3) receptors (Extended Data
Fig. 6b). Consistent with increased effector differentiation and func-
tion, UA-supplemented old mice showed better virus control (Fig. 3g).
Next, we addressed whether the effects of UA feeding on the CD8+
T cell response were directly due to the recovery of HSC function. To
this end, mice were transplanted with a fixed number of HSCs derived
from UA-fed mice (Fig.3h). Sixteen weeks after transplantation, mice
were infected with LCMV (Armstrong strain), and donor-derived CD8+
T cells were monitored by exploiting the double congenic allelic system
(Fig. 3h). Despite the presence of virus-specific CD8+ T cells (Extended
Data Fig. 7a), the proportion of KLRG1+ cells were not different
(Extended Data Fig. 7b), while virus-specific CD8+ T cells derived from
UA-treated HSCs showed improved GzmB expression (Fig. 3i), which
was more evident in GzmB-producing IFNγ+ cells (Fig. 3j). These results
indicate that HSCs contribute directly to the improved functionality of
CD8+ T cells. Importantly, UA in vitro treatment of splenocytes derived
from old mice (Extended Data Fig. 8a) drove a mild modulation of
mitochondrial membrane potential without affecting mitochondrial
mass (Extended Data Fig. 8b) and enhanced functionality, as shown by
increased cytokine production (Extended Data Fig. 8c) in both CD8+
and CD4+ T cells. Taken together, these data suggest that the UA-driven
P = 0.0003; old control versus old UA P = 0.0005; np396+: young control versus
old control P < 0.0001; old control versus old UA P = 0.0002). n = 10; values are
the mean ± s.e.m.; one-way ANOVA; ***P ≤ 0.001, ****P ≤ 0.0001. f, INFγ+CD8+ T
cells (np396+: young control versus old control P = 0.0044; old control versus old
UA P = 0.0259). n = 10; values are the mean ± s.e.m.; one-way ANOVA; *P ≤ 0.05,
**P ≤ 0.01. g, Quantification of viral titers in the spleen (young control versus old
control P = 0.0092; old control versus old UA P = 0.0151). n = 10; values are the
mean ± s.e.m.; one-way ANOVA; *P ≤ 0.05, **P ≤ 0.01. h, Mice stably reconstituted
with 200 HSCs derived from old mice fed with a control or a UA-supplemented
diet (50 mg kg day−1) were infected with LCMV (Armstrong strain) and euthanized
8 days after infection to analyze the spleen. i, GzmB expression level in donor-
derived, virus-specific CD8+ T (gp33+: old control versus old UA P = 0.07; np396+:
old control versus old UA P = 0.0354). n = 10; values are the mean ± s.e.m.;
one-way ANOVA; *P ≤ 0.05. j, INFγ+ donor-derived, virus-specific CD8+ T cells
(gp33+: old control versus old UA P = 0.0249; np396+: young control versus
old UA P = 0.0149; old control versus old UA P = 0.0030). n = 5; values are the
mean ± s.e.m.; one-way ANOVA; *P ≤ 0.05, **P ≤ 0.01. NS, not significant.
1060
Letter https://doi.org/10.1038/s43587-023-00473-3

amelioration of immune function in old mice is due both to a direct
modulation of HSC and T cell functions.
Το assess whether the UA effect was exerted through mitochon-
drial recycling, we first tested the capacity of UA to clear mitochon-
dria in HSCs. HSCs purified from mito-QC mitophagy reporter mice
were cultured for 3 days in the presence of 20 μM UA, and mitophagy
induction was measured using flow cytometry. Mito-QC mice express
a tandem mCherry-green fluorescent protein (GFP) targeted to the
mitochondrial outer membrane. When mitochondria are degraded
in lysosomes, the acidic environment quenches GFP but not mCherry
fluorescence; consequently mitochondrial clearance can be measured
based on the shift in the GFP signal18,32,33. Indeed, UA treatment induced
mitochondrial clearance in HSCs, as indicated by the increased propor-
tion of cells with a high mitophagy rate (Fig. 4a) and by the decreased
GFP to mCherry ratio on the whole HSC population (Extended Data
Fig. 9a). These results were further corroborated based on the increased
expression of key genes involved in autophagy and mitophagy (Atg5,
Park2 and p62) (Fig. 4b).
Importantly, induction of mitochondrial clearance did not change
cellular basal and maximal respiration (Fig. 4c–e) of c-Kit+ hematopoi-
etic progenitor cells, but was accompanied by an improved spare res-
piratory capacity (% SRC) (Fig. 4f), indicating improved mitochondrial

a LCMV b
(Armstrong strain)
–
CD44+ Tet CD44+ Tet+ KLRG1+
Control diet
CD8+

FSC-A

KLRG1
CD44
UA diet Day 8
after infection
Weeks 0 8
–
CD44

4 Days CD8 Tetramer CD127

washout (virus-specific)
Old

c d KLRG1+ KLRG1+
+ + (gp33+) (np396+)
gp33 np396
** ***
*** 80 80 NS Young control
4
***
6
**** Young control ****
Number of cells (106)

*** ****
Percentage of cells

Old control Old control

3
Old UA
60 60 ** Old UA
4
*
2 NS 40 40

2
1 20 20

0 0 0 0

e GzmB GzmB f GzmB GzmB g Viral titer

(gp33+) (np396+) (gp33+ IFNγ+) (np396+ IFNγ+) NS
NS NS
NS ** ** * Young control
200 150 **** 250
NS
NS 250
NS
*
107
Old control
*** *** 10 6
200 200
Normalized MFI

150 *** 105

Old UA
PFU/spleen

100
150 150 104
100
100 100 103
50 102
50 50 50
101
0 0 0 0 0

LCMV
(Armstrong strain)
h
Control diet
Competitor
CD45.1 and CD45.2
UA diet
Day 8
Weeks 0 16 after infection
Weeks 0 4 .. 16

Recipient
200 HSCs CD45.2
Old LKS CD150+CD48–
CD45.1

i GzmB GzmB j GzmB GzmB

(gp33+) (np396+) (gp33+ IFNγ+) (np396+ IFNγ+)
NS NS NS
5,000 10,000 15,000
P = 0.07 6,000
* *
*
4,000 8,000 NS **
Young control 10,000
NS 4,000 6,000 Young control
3,000 NS NS
Old control
MFI

Old control
MFI

2,000 4,000
2,000
Old UA 5,000 Old UA
1,000 2,000

0 0 0
0

Nature Aging | Volume 3 | September 2023 | 1057–1066 1061

Letter https://doi.org/10.1038/s43587-023-00473-3

a b 2.0
Old Control **
8 Old UA ** **
Mitophagy Mitophagy ** 1.5

Percentage of cells

Fold expression
6.7 6 NS
2.6
*
1.0
4
mCherry

2 0.5
Control
GFP UA
0 0
Pik3c3 Becn1 Atg5 Park2 p62

c d e f g
70 30 60
P = 0.05
Old Control NS NS 250 10
60 UA **
OCR (pmol min−1)

SRC (% of baseline)
OCR (pmol min−1)

OCR (pmol min−1)

200 8
50
20 40

MFI (103)
40 150 6
30
100 4
20 10 20
Old Control
10 50 2 Old UA

0 0 0 0 0
0 20 40 60 80
Time (min)

h 8 i 10 j 8
*
4 **
* high **
CD150 8
6 3 6

MFI (103)

MFI (103)
MFI (103)

MFI ratio

6
CD150

** 4
4 2
4
low
Old Control CD150 Old Control Old Control
2 1 Old Control 2
Old UA 2 Old UA Old UA
Old UA

0 0 FCS 0 0

gh

w
lo
hi

50
50

D1
D1

C
C

k l m n o
* 3
80 250
Control 50
NS 150 *
70 UA
SRC (% of baseline)
OCR (pmol min−1)

200
OCR (pmol min−1)

OCR (pmol min−1)

60 40 NS
2
50
30
100
150 MFI (103)
40
30 20 100 1
50
20
10 50
10
0 0 0
0 0
0 20 40 60 80
Time (min)

Fig. 4 | UA induces mitochondrial recycling and restores mitochondrial methyl ester (TMRM)/MitoTracker Green FM (MTG) (right) using flow cytometry
fitness. a, Analysis of mitochondrial clearance in HSCs purified from of cultured HSCs in the presence of UA (P = 0.0324); values are expressed as mean
mitophagy reporter mice (mito-QC) on a 3-day UA treatment in vitro measured fluorescence intensity (MFI). n = 3; values are the mean ± s.e.m.; two-tailed paired
as percentage of cells in the mitophagy gate (left) (P = 0.0037). n = 5; values Student’s t-test; *P ≤ 0.05. i, Measurement of mitochondrial mass by MitoTracker
are the mean ± s.e.m.; two-tailed paired Student’s t-test; *P ≤ 0.05, **P ≤ 0.01. Green staining in CD150high (P = 0.0028) and CD150low (P = 0.0048) HSCs cultured
b, Gene expression analyses by quantitative PCR of key genes involved in the for 3 days in the presence of UA; data are expressed as the MFI. n = 3; values are
autophagy and mitophagy processes in HSCs cultured for 3 days in the presence the mean ± s.e.m.; two-way ANOVA; **P ≤ 0.01. j, Analysis of PGC1α expression in
of UA (whole progeny) (Becn1 P = 0.035; Atg5 P = 0.0049; Park2 P = 0.0040; p62 HSCs cultured for 3 days in the presence of UA using flow cytometry (P = 0.0032).
P = 0.0058). n = 3; values are the mean ± s.e.m.; two-tailed paired Student’s t-test; n = 4; values are the mean ± s.e.m.; Student’s t-test; **P ≤ 0.01. k–n, Measurement
*P ≤ 0.05, **P ≤ 0.01). c–f, Measurement of OCR (c), basal respiration (d), maximal of OCR (k), basal respiration (l), maximal respiration (m) and SRC (% baseline) (n)
respiration (e) and SRC (% baseline) (f) of 3-day cultured Lin−c-Kit+ hematopoietic of Lin−c-Kit+ hematopoietic stem and progenitor population derived from mice
stem and progenitor population in the presence of UA (P = 0.0526) (n = 3; values fed with UA for 4 months (P = 0.0424). n = 3; values are the mean ± s.e.m.; two-
are the mean ± s.e.m.; paired two-tailed Student’s t-test). g, Analysis of CD150 tailed Student’s t-test; *P ≤ 0.05. o, Analysis of the mitochondrial mass of HSCs
expression (P = 0.0036). n = 3; values are the mean ± s.e.m.; two-tailed paired derived from mice fed with UA for 4 months. n = 3; values are the mean ± s.e.m.;
Student’s t-test; **P ≤ 0.01. h, Mitochondrial mass (MitoTracker) (P = 0.0127) Student’s t-test; *P ≤ 0.05,). NS, not significant.
(left) and mitochondrial activity per mitochondrial mass tetramethylrhodamine,

performance. As HSCs accumulate mitochondria due to autophagy reduced CD150 expression and mitochondrial mass in purified old
defects (Extended Data Fig. 1) and have increased expression of CD150 HSCs (Fig. 4g,h). Interestingly, old CD150high HSCs, which have a marked
during aging (Extended Data Fig. 9b)3,28, we investigated the capacity of myeloid bias and aging phenotype28, have higher mitochondrial con-
UA to revert these aging phenotypes in old HSCs. UA in vitro treatment tent compared to CD150low HSCs. Treatment with UA decreased the

Nature Aging | Volume 3 | September 2023 | 1057–1066 1062

Letter
mitochondrial content in both CD150high and CD150low HSCs compara-
bly (Fig. 4i). Importantly, UA treatment increased the expression of the
master regulator of mitochondrial biogenesis PGC1α in old HSCs
(Fig. 4j). These data indicate that UA treatment reestablishes mito­
chondrial function and fitness by promoting mitochondrial recycling,
which affects both mitochondrial clearance and biogenesis.
We further investigated the effect of dietary UA supplementation
on metabolic features of hematopoietic stem and progenitor popula-
tions. As observed in in vitro treatment, UA dietary supplementation
did not change cellular basal and maximal respiration (Fig. 4k–m) but
improved SRC (Fig. 4k,n) of hematopoietic progenitor cells. Congru-
ently, these metabolic changes were accompanied by a reduction of
mitochondrial mass in the HSC population (Fig. 4o). Overall, our data
indicate that induction of mitochondrial recycling in old HSCs allows
the restoration of mitochondrial function both in vitro and in vivo.
Interestingly, HSCs derived from Mx1-PARK2−/− mice, bearing spe-
cific deletion of the mitophagy gene PARK2 in the hematopoietic line-
age, fed with UA and transplanted in lethally irradiated mice (Extended
Data Fig. 9c), did not exhibit better blood reconstitution capacity
(Extended Data Fig. 9d). Consistently, deletion of PARK2 exacerbated
the aging phenotype in HSCs, leading to a further shrinkage of the
lymphoid lineage (Extended Data Fig. 9e). These data indicate that
mitochondrial clearance is necessary to restore HSC function. In fact,
the maintenance of metabolic fitness and function of HSCs requires a
fine balance between the regeneration of newly functional mitochon-
dria and the clearance of old defective ones.
Interestingly, NAD booster supplementation can partially reju-
venate ‘old’ HSCs through the restoration of the metabolic features of
‘young’ HSCs22. As we have observed on UA treatment, ‘old’ HSCs reduce
their mitochondrial mass on nicotinamide riboside supplementation,
reinforcing the concept that mitochondria are key regulators of the
aging process in HSCs3,22,34. Differently, UA treatment did not require
continuous supplementation to restore ‘old’ HSC function and, most
importantly, we showed that the effect observed in HSCs translated
into amelioration of immune functions22. Our data evidence that UA
supplementation might have a major impact on older adults by rein-
forcing both their hematopoietic and immune systems, therefore
reducing the risk of hematopoietic failure, boosting immune surveil-
lance and improving their response to vaccination. Another category
of patients who could benefit from UA supplementation is patients
undergoing chemotherapy, which induces immunosenescent features,
increases the susceptibility to infections and reduces the response to
vaccination.
The loss of cellular quality control processes during aging leads
to the accumulation of defective cellular machineries and to a rapid
deterioration of cellular functions35. Our results demonstrate that UA
treatment is capable to revert the metabolic defects of old HSCs and
rejuvenates hematopoietic and immune system functions, providing
further evidence on the important interplay between cellular metabo-
lism and aging-related cellular dysfunctions.
Methods
This study complies with all relevant ethical regulations. For human
sample collection the protocol has been approved by Commission
Cantonale d’Éthique de la Recherche sur l’Être Humain CER-VD
(authori­zation no. CER-VD 2017-00846), where human CD34+ cells
were collected from patients who were informed about the research
project according to a written consent. For the animal study the pro-
tocols have been approved by the Service de la Consummation et des
Affaires Vétérinaires (authorization nos. VD3684 and VD3572).
Animals
Experiments were performed on C57BL/6 female mice throughout
the study. Young mice were 8 weeks old while old mice were more than
80 weeks old.
Nature Aging | Volume 3 | September 2023 | 1057–1066
https://doi.org/10.1038/s43587-023-00473-3
BM extraction and cell sorting
Flow cytometry analysis was performed on freshly isolated BM from
C57BL/6 mice (young: 8 weeks old; old more than 80 weeks old). BM was
extracted from crushed femur, tibia and pelvis. The cell suspension was
filtered through a 70-μm cell strainer and erythroid cells were eliminated
by incubation with red blood cell lysis buffer (eBioscences). Isolation
and stains were performed in ice-cold PBS 1 mM EDTA. Lineage-positive
cells were then removed with a magnetic lineage depletion kit (BD Bio-
sciences). The cell suspension was then stained with specific antibodies
for progenitor and stem cell compartments and sorted on a BD FACSAria
III into 1.5-ml Eppendorf tubes. The HSC compartment was identified
and sorted with the following cell surface phenotype LKS CD150+CD48−.
Antibodies
The following antibodies were used in this study: rat monoclonal anti-
bodies (mAbs) against streptavidin− Texas Red, c-Kit-PeCy7 (clone 2B8),
Sca1-allophycocyanin (APC) (clone D7), CD150-PE (clone TC-15-12F12.2),
CD48-PB (clone HM48-1), CD45.2-PB (clone 104), CD45.1-FITC (clone A20),
Gr1-APC (clone RB6-8C5), F4/80-APC (clone BM8), CD19-PE (clone 6D5),
CD3-PE (clone 17A2) and CD16/CD32 (clone 2.4G2). The antibodies were
purchased from BioLegend, eBioscience and BD Biosciences. A mixture
of biotinylated mAbs against CD3, CD11b, CD45R/B220, Ly-6G, Ly-6C
and TER-119 was used as lineage marker (‘lineage cocktail’) and was pur-
chased from BD Biosciences. Human-specific antibodies were: hCD56
(clone NCAM16.2), hCD16 (clone 3G8), hCD45 (clone HI30), hCD19 (clone
HIB19), hCD4 (clone RPA-T4), hCD3 (clone SK7), hCD14 (clone M5E2),
hCD8b (clone SIDI8BEE), hCD34 (clone 8G12) and hCD38 (clone HB-7),
and were either from eBioscience or BD Biosciences; 4,6-diamidino-
2-phenylindole staining was used for LIVE/DEAD cell discrimination.
For the viral infection analyses, the following antibodies were
used: CD44-PB (clone IM781), KLRG1-APC (clone 2F1), CD127-BV786
(clone A7R34), PD-1-PE-Cy7 (clone RMP1-30), LAG3-PercPCy5.5 (clone
eBioC9B7W), GrzB-PB (clone GB11), IFNγ-PercPCy5.5 (clone XMG1.2),
TNFα-PE-Cy7 (clone MP6-XT22) and IL2-APC (clone JES6-5H4).
For the splenocyte culture assays, the following antibodies were
used: CD3ε (clone 145-2C11), CD8α (clone 53.6.7), CD4 (clone RM4-5),
IFNγ (catalog no. 505810, BioLegend) and TNFα (catalog no. 11-7321-82,
Thermo Fisher Scientific).
Mouse HSC, progenitor and human CD34+ cell culture
Murine HSCs were sorted in 1.5-ml Eppendorf tubes. Cells were seeded
in 96-well round bottom plates and cultured in Stemline II (Sigma-
Aldrich) supplemented with 100 ng ml−1 stem cell factor (SCF) (R&D)
and 2 ng ml−1 Flt3 (R&D). UA (Sigma-Aldrich) was added at the indicated
concentrations (dissolved in dimethylsulfoxide (DMSO)); an equal
quantity of DMSO was added to the control wells. All cultures were
maintained at 5% CO2 at 37 °C. Cryopreserved CD34+ cells isolated
from patients with hip replacement orthopedic surgery were thawed
and cultured in StemSpan (STEMCELL Technologies) medium sup-
plemented with human SCF (100 ng ml−1), human FLT3L (100 ng ml−1),
human thrombopoietin (50 ng ml−1) and human low-density lipoprotein
(10 μg ml−1), and different concentrations of UA (dissolved in DMSO)
were added; an equal amount of DMSO was added to the control wells.
All cultures were maintained at 5% CO2 at 37 °C.
Analysis of mitochondrial mass using flow cytometry
Postculture cells were incubated at 37 °C for 45 min with 100 nM
MitoTracker Green. Cells were then washed with fluorescence-acti-
vated cell sorting (FACS) buffer and analyzed by flow cytometry on a
BD LSR II flow cytometer.
DNA extraction and mitochondrial DNA and nuclear DNA
estimation
A progeny of 1,000 cells were collected at the end of the culture period
and DNA was isolated using the DNeasy kit (QIAGEN) according to
1063
Letter
the manufacturer’s instructions. Quantitative PCR was carried out
to estimate the relative values for mitochondrial (COX1) and nuclear
(NDUFA2) DNA to estimate the mitochondrial to nuclear DNA ratio.
Murine BM transplantations
C57BL/6 Ly5.2 adult mice (8–12 weeks old) were lethally irradiated
with a total 8.5 Gy dose in an X-ray irradiator (RS 2000, RAD source)
24 h before transplant. The dose was split in two doses of 4.25 Gy sepa-
rated by a 4–6-h interval. Mice were injected alternatively with 2,000
donor HSC progeny postculture or 200 donor HSCs freshly isolated or
150,000 total BM cells freshly isolated derived from C57BL/6 Ly5.1 mice
(young or old) and 150,000 competitor cells derived from C57BL/6
Ly5.1/5.2 mice, via tail vein injection. Peripheral blood was collected
every 4 weeks to determine the percentage of chimerism by FACS
analysis. Spleen and BM were analyzed at the endpoints. For secondary
transplants, adult mice (8–12 weeks old) were lethally irradiated with
the same dose 24 h before transplant and each mouse was injected with
3 million whole BM cells from a donor mouse.
Colony-forming unit assay
A colony-forming unit (CFU) assay with human HSCs was carried
out using MethoCult H4434 (STEMCELL Technologies) according to
the manufacturer’s instructions. One thousand cells from each well
were plated in duplicate at different dilutions. Colonies were counted
15 days after plating using Stem Vision (STEMCELL Technologies). For
secondary CFU assays, all cells were recovered from the primary CFU
plates and plated at serial dilutions (10–500 K per plate). Plates with
fewer than 150 colonies were scored to determine the normalized
counts per 1,000 HSCs plated in the primary CFU assay.
Food preparation for in vivo feeding
The UA diet was prepared by mixing custom synthesized (by Novalix)
UA dissolved in DMSO with a 2916 powder diet (Charles River Labo-
ratories) and air-dried into pellets under sterile conditions. The UA
dosage in the mix was calculated by considering the average mouse
food intake per day (5 g food per mouse per day) for a calculated intake
of 50 mg UA per mouse per day). An equivalent amount of DMSO was
added in the control food. Weekly food consumption was monitored
in all cages and no difference was found in the feeding behavior of mice
on UA supplementation in the diet.
Viral infections and staining
Mice were infected intraperitoneally with 2 × 105 plaque-forming units
of LCMV 53b Armstrong strain. To determine viral titers, spleens from
LCMV-infected mice were ‘shock-frozen’. Diluted spleen suspensions
were then used to infect Vero cells, and viral titers were determined by
an LCMV focus-forming assay, as described elsewhere36.
Cell suspensions from the spleen were obtained by mashing
through a 40-μM nylon cell strainer, followed by red blood cell lysis
using ammonium-chloride-potassium buffer. Surface staining was
performed with mAbs for 20 min at 4 °C in PBS supplemented with 2%
FCS (FACS buffer).
For tetramer staining, cell suspensions were incubated with anti-
CD16/32 (clone 2.4G2) hybridoma supernatant before staining for
90 min at 4 °C with APC-conjugated major histocompatibility complex
class I tetramers.
For intranuclear staining, cells were surface-stained before fixa-
tion and permeabilization using the Foxp3 Transcription Factor Stain-
ing Kit (catalog no. 00-5523, eBioscience) followed by intranuclear
staining in permeabilization buffer 1× (permeabilization buffer).
For the detection of cytokine production, splenocytes were res-
timulated in vitro with LCMV gp33-41 (gp33) or np396-404 (1 μM)
peptide for 5 h in the presence of brefeldin A (5 μg ml−1) for the last
4.5 h. Cells were then surface-stained before fixation and permeabili-
zation (using the Intracellular Fixation & Permeabilization Buffer Set,
Nature Aging | Volume 3 | September 2023 | 1057–1066
https://doi.org/10.1038/s43587-023-00473-3
catalog no. 88-8824, eBioscience) followed by intracellular staining in
1× permeabilization buffer.
Metabolic assay
The oxygen consumption rate (OCR) was measured using a 96-well
Seahorse Bioanalyzer XF96 according to the manufacturer’s instruction
using a Seahorse Mito Stress Test. Briefly, c-Kit+ hematopoietic progeni-
tor cells were sorted and suspended in Seahorse XF basic medium with
11 mM glucose, 1 mM sodium pyruvate and 2 mM glutamine (pH 7.4, at
37 °C) (100,000 cells per well). The injection ports were loaded with
1 μM oligomycin, 2 μM carbonyl cyanide-p-trifluoromethoxyphenyl-
hydrazone and 0.5 μM rotenone/antimycin. During sensor calibration,
cells were incubated in a 37 °C non-CO2 incubator for 45 min.
Splenocyte culture
Splenocytes from old mice were stimulated in vitro with plate-bound
anti-CD3 (3 μg ml−1, catalog no. 16-0037-85, Thermo Fisher Scientific)
and soluble anti-CD28 (2 μg ml−1, catalog no. 102116, BioLegend) for
3 days and expanded in medium containing 10 ng ml−1 human inter-
leukin-2 (catalog no. 200-02, PeproTech) for 2 days. UA was supple-
mented at 5 μM. To assess cytokine production, cells were restimulated
with plate-bound anti-CD3 (3 μg ml−1) for 4 h in presence of GolgiStop
(catalog no. 554724, BioLegend) and brefeldin A (catalog no. 420601,
BioLegend) according to the manufacturer’s instruction. For mito-
chondrial mass and mitochondrial potential, cells were cultured at
37 °C for 30 min with 25 nM MitoTracker Green and 100 nM TMRM
before flow analysis.
Mice model for mitophagy
Mito-QC mice in the C57BL/6J background provided by P.C. Ho and bred
at the Ludwig Institute for Cancer Research (LICR), Lausanne were used
as reporter mice for mitophagy. Mitophagy induction was analyzed by
flow cytometry based on the shift on the GFP signal. Mx1-Cre mice were
provided by F. Radtke from École Polytechnique Fédérale de Lausanne
(EPFL) while PARK2loxP/loxP mice were provided by P.C. Ho. The cross of
these two strains to achieve PARK2 deletion in the hematopoietic line-
age was performed at LICR.
qPCR
HSCs were FACS-sorted (Lin−c-Kit+Sca1+CD150+CD48−) and cultured in
vitro for 3 days in the presence or absence of UA (20 mM). After incu-
bation, total RNA was extracted from HSCs using the RNA Microprep
Kit (Zymo Research), according to the manufacturer’s instructions.
RNA was retrotranscribed to complementary DNA (cDNA) with the
PrimeScript cDNA Synthesis Kit (Takara Bio). qPCR with reverse tran-
scription reactions was performed using 1.5 μl cDNA, 5 μl Power SYBR
Green Master Mix (Applied Biosystems) and 200 nM of primers, and
analyzed on the 7900HT system (Applied Biosystems). The relative
expression of Actb was used to normalize gene expression across sam-
ples. The following primers were used: Atg5 (forward: GGAGAGAAGAG-
GAGCCAGGT; reverse: GCTGGGGGACAATGCTAATA); Park2 (forward:
CCGAATCACCTGACGGTTCA; reverse: TCTGGCTGCTTCTGAATCCC);
Becn1 (forward: CCGCGGTAGAACGAGCC; reverse: AAGTAATGGAGCT
GTGAGTTCCT); Pik3c3 (forward: GTGAAGTACCCTGACCTGCC; reverse:
AGTCATGCATTCCTTGGCGA); p62 (forward: GCTGAAGGAAGCTGCC
CTAT; reverse: TTGGTCTGTAGGAGCCTGGT); Actb (forward: GAGAC
CTTCAACACCCC; reverse: GTGGTGGTGAAGCTGTAGCC).
Statistic and reproducibility
All experiments were reproduced at least once. Data were analyzed
using two-sided or one-sided Student’s t-tests or a two-way ANOVA or
Mann–Whitney U-test with a 95% confidence interval.
No statistical method was used to predetermine sample size. Sample
sizes were chosen based on previous studies18,37. The investigators were
not blinded to allocation and outcome assessment. Some data points
1064
Letter
were excluded from the analyses based on an outlier Grubbs statistical
test or when an animal developed splenomegaly. Data distribution was
assumed to be normal but this was not formally tested.
Reporting summary
Further information on research design is available in the Nature Port-
folio Reporting Summary linked to this article.
Data availability
All data and materials are available from the corresponding author
to any researcher for the purpose of reproducing or extending the
analyses. Source data are provided with this paper.
References
1. Morrison, S. J., Wandycz, A. M., Akashi, K., Globerson, A. &
Weissman, I. L. The aging of hematopoietic stem cells. Nat. Med.
2, 1011–1016 (1996).
2. Pang, W. W. et al. Human bone marrow hematopoietic stem
cells are increased in frequency and myeloid-biased with age.
Proc. Natl Acad. Sci. USA 108, 20012–20017 (2011).
3. Ho, T. T. et al. Autophagy maintains the metabolism and function
of young and old stem cells. Nature 543, 205–210 (2017).
4. Geiger, H., de Haan, G. & Florian, M. C. The ageing haematopoietic
stem cell compartment. Nat. Rev. Immunol. 13, 376–389 (2013).
5. Chandel, N. S., Jasper, H., Ho, T. T. & Passegué, E. Metabolic
regulation of stem cell function in tissue homeostasis and
organismal ageing. Nat. Cell Biol. 18, 823–832 (2016).
6. Hine, C. & Mitchell, J. R. Saying no to drugs: fasting protects
hematopoietic stem cells from chemotherapy and aging. Cell
Stem Cell 14, 704–705 (2014).
7. Flach, J. et al. Replication stress is a potent driver of functional
decline in ageing haematopoietic stem cells. Nature 512, 198–202
(2014).
8. Beerman, I., Seita, J., Inlay, M. A., Weissman, I. L. & Rossi, D. J.
Quiescent hematopoietic stem cells accumulate DNA damage
during aging that is repaired upon entry into cell cycle. Cell Stem
Cell 15, 37–50 (2014).
9. Beerman, I. & Rossi, D. J. Epigenetic regulation of hematopoietic
stem cell aging. Exp. Cell Res. 329, 192–199 (2014).
10. Chambers, S. M. et al. Aging hematopoietic stem cells decline in
function and exhibit epigenetic dysregulation. PLoS Biol. 5, e201
(2007).
11. Sahin, E. & Depinho, R. A. Linking functional decline of telomeres,
mitochondria and stem cells during ageing. Nature 464, 520–528
(2010).
12. Nakada, D., Saunders, T. L. & Morrison, S. J. Lkb1 regulates cell
cycle and energy metabolism in haematopoietic stem cells.
Nature 468, 653–658 (2010).
13. Gurumurthy, S. et al. The Lkb1 metabolic sensor maintains
haematopoietic stem cell survival. Nature 468, 659–663 (2010).
14. Gan, B. et al. Lkb1 regulates quiescence and metabolic
homeostasis of haematopoietic stem cells. Nature 468, 701–704
(2010).
15. Zhang, J. et al. UCP2 regulates energy metabolism and
differentiation potential of human pluripotent stem cells. EMBO J.
30, 4860–4873 (2011).
16. Suda, T., Takubo, K. & Semenza, G. L. Metabolic regulation of
hematopoietic stem cells in the hypoxic niche. Cell Stem Cell 9,
298–310 (2011).
17. Ito, K. et al. Self-renewal of a purified Tie2+ hematopoietic stem
cell population relies on mitochondrial clearance. Science 354,
1156–1160 (2016).
18. Vannini, N. et al. The NAD-booster nicotinamide riboside potently
stimulates hematopoiesis through increased mitochondrial
clearance. Cell Stem Cell 24, 405–418 (2019).
Nature Aging | Volume 3 | September 2023 | 1057–1066
https://doi.org/10.1038/s43587-023-00473-3
19. Mohrin, M. et al. Stem cell aging. A mitochondrial UPR-mediated
metabolic checkpoint regulates hematopoietic stem cell aging.
Science 347, 1374–1377 (2015).
20. Ho, T. T. et al. Aged hematopoietic stem cells are refractory to
bloodborne systemic rejuvenation interventions. J. Exp. Med. 218,
e20210223 (2021).
21. Cheng, C.-W. et al. Prolonged fasting reduces IGF-1/PKA to
promote hematopoietic-stem-cell-based regeneration and
reverse immunosuppression. Cell Stem Cell 14, 810–823 (2014).
22. Sun, X. et al. Nicotinamide riboside attenuates age-associated
metabolic and functional changes in hematopoietic stem cells.
Nat. Commun. 12, 2665 (2021).
23. Larrosa, M., García-Conesa, M. T., Espín, J. C. & Tomás-Barberán,
F. A. Ellagitannins, ellagic acid and vascular health. Mol. Aspects
Med. 31, 513–539 (2010).
24. Ryu, D. et al. Urolithin A induces mitophagy and prolongs lifespan
in C. elegans and increases muscle function in rodents. Nat. Med.
22, 879–888 (2016).
25. Luan, P. et al. Urolithin A improves muscle function by inducing
mitophagy in muscular dystrophy. Sci. Transl. Med. 13, eabb0319
(2021).
26. Andreux, P. A. et al. The mitophagy activator urolithin A is safe and
induces a molecular signature of improved mitochondrial and
cellular health in humans. Nat. Metab. 1, 595–603 (2019).
27. D’Amico, D. et al. Impact of the natural compound urolithin A on
health, disease, and aging. Trends Mol. Med. 27, 687–699 (2021).
28. Beerman, I. et al. Functionally distinct hematopoietic stem cells
modulate hematopoietic lineage potential during aging by a
mechanism of clonal expansion. Proc. Natl Acad. Sci. USA 107,
5465–5470 (2010).
29. Ho, Y. H. et al. Remodeling of bone marrow hematopoietic stem
cell niches promotes myeloid cell expansion during premature or
physiological aging. Cell Stem Cell 25, 407–418 (2019).
30. Decman, V. et al. Defective CD8 T cell responses in aged mice
are due to quantitative and qualitative changes in virus-specific
precursors. J. Immunol. 188, 1933–1941 (2012).
31. Ucar, D. et al. The chromatin accessibility signature of human
immune aging stems from CD8+ T cells. J. Exp. Med. 214,
3123–3144 (2017).
32. McWilliams, T. G. et al. mito-QC illuminates mitophagy and
mitochondrial architecture in vivo. J. Cell Biol. 214, 333–345 (2016).
33. Klionsky, D. J. et al. Guidelines for the use and interpretation of
assays for monitoring autophagy (4th edition)1. Autophagy 17,
1–382 (2021).
34. Mansell, E. et al. Mitochondrial potentiation ameliorates age-
related heterogeneity in hematopoietic stem cell function.
Cell Stem Cell 28, 241–256 (2021).
35. Jovaisaite, V., Mouchiroud, L. & Auwerx, J. The mitochondrial
unfolded protein response, a conserved stress response pathway
with implications in health and disease. J. Exp. Biol. 217, 137–143
(2014).
36. Battegay, M. et al. Quantification of lymphocytic choriomeningitis
virus with an immunological focus assay in 24- or 96-well plates.
J. Virol. Methods 33, 191–198 (1991).
37. Vannini, N. et al. Specification of haematopoietic stem cell fate
via modulation of mitochondrial activity. Nat. Commun. 7, 13125
(2016).
Acknowledgements
We thank G. Coukos (LICR, University of Lausanne (UNIL), Centre
Hospitalier Universitaire Vaudois) for critical feedback and F. Derouet
(UNIL) for animal care. Flow cytometry analysis and sorting was
performed at the Flow Cytometry Facility at UNIL with the help of
R. Bedel, F. Sala de Oyanguren, K. Blackney and A. Wilson. We thank
Nestlé Health Science for their suggestions about the study.
1065
Letter
M.G. was partially funded by a collaborative Jebsen Foundation
grant (EPFL/UNIL) to N.V. and O.N. The N.V. laboratory is supported
by the Swiss Cancer Research Foundation (KFS-4993-02-2020-R)
and San Salvatore Foundation. The J.A. laboratory is supported by
grants from EPFL, the European Research Council (ERC-AdG-787702),
the Swiss National Science Foundation (SNSF) (31003A_179435)
and a Global Research Laboratory grant of the National Research
Foundation of Korea (2017K1A1A2013124). F.S. is supported by the
SNSF (323530_183986). The O.N. laboratory was supported by SNSF
grant nos. PP00P3_183725 and CRSII5-186271. P.-C.H. is funded in part
by the European Research Council Staring Grant (802773-MitoGuide),
an SNSF project grant (no. 31003A_182470), the Cancer Research
Institute (Lloyd J. Old STAR Award) and LICR. W.H. is funded in part by
a grant from the SNSF (no. 310030_200898). All the funders had no
role in study design, data collection and analysis, decision to publish
or preparation of the manuscript.
Author contributions
M.G., Y-H.C. and N.V. conceived the ideas, designed the experiments
and analyzed the results. M.G., Y-H.C., N.V. and F.S. performed the HSC
experiments. M.C., H.C.H. and W.H. performed and interpreted the
infection model experiments. P.G. evaluated the impact of UA dietary
supplementation on secondary lymphoid tissues and performed and
interpreted the cell respiration analyses. J.A. and O.N. conceived
the ideas and provided key reagents and samples. S.C. provided the
human samples for HSC purification. O.N. and C.B. performed the
human HSC experiments. F.F., Y-R.Y., H.G. and P.-C.H. provided the
animal model and helped with data interpretation. N.V. wrote the
manuscript. All authors edited and reviewed the final manuscript.
Competing interests
J.A. is a scientific advisor to Amazentis, a company that develops
UA as a therapeutic agent. Some elements of this work are part
Nature Aging | Volume 3 | September 2023 | 1057–1066
https://doi.org/10.1038/s43587-023-00473-3
of patent no. 19152437.0 of which N.V. and M.G. are inventors. The
remaining authors declare no competing interests.
Additional information
Extended data is available for this paper at
https://doi.org/10.1038/s43587-023-00473-3.
Supplementary information The online version
contains supplementary material available at
https://doi.org/10.1038/s43587-023-00473-3.
Correspondence and requests for materials should be addressed to
Nicola Vannini.
Peer review information Nature Aging thanks Saghi Ghaffari and the
other, anonymous, reviewer(s) for their contribution to the peer review
of this work.
Reprints and permissions information is available at
www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard
to jurisdictional claims in published maps and institutional
affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds
exclusive rights to this article under a publishing agreement with
the author(s) or other rightsholder(s); author self-archiving of the
accepted manuscript version of this article is solely governed by the
terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to Springer Nature America,
Inc. 2023
1066
Letter https://doi.org/10.1038/s43587-023-00473-3

Extended Data Fig. 1 | Old HSCs have higher mitochondria content. Mitochondrial DNA (mtDNA) quantification by QPCR expressed as fold change compared to
young HSCs. Young values are normalized on their mean value (P-value = 0.0571) (n = 4; values are mean ± s.e.m.; two-tailed Mann-Whitney test; * P ≤ 0.05, ** P ≤ 0.01,
*** P ≤ 0.001).

Nature Aging
Letter
Extended Data Fig. 2 | Donor chimerism analysis in the bone marrow.
a, Lymphoid vs myeloid compartments of primary transplants in peripheral
blood at the indicated timepoints. b, Analysis of donor derived contribution
for lymphoid and myeloid lineages in bone marrow at the endpoint of primary
transplant (Lymphoid: young ctrl vs old ctrl P-value = 0.0003; old ctrl vs old
UA P-value = 0.0011; myeloid: young ctrl vs old ctrl P-value = 0.0002; old ctrl vs
old UA P-value = 0.0006;). c, Analysis of donor derived HSCs at the endpoint
Nature Aging
https://doi.org/10.1038/s43587-023-00473-3
of primary transplant. d, Analysis of donor derived HSCs at the endpoint of
secondary transplant (n = 10; values are mean ± s.e.m.; two-tailed Student’s t test;
* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001). e, HSCs isolated from bone
marrow of young mice were cultured for three days in the presence of 20μM UA
and transplanted in lethally irradiated primary recipient together with bone
marrow derived from competitor mice. Blood donor chimerism analyses of
primary transplant at indicated timepoints (n = 10; values are mean ± s.e.m.).
Letter
Extended Data Fig. 3 | UA improves Colony forming capacity (CFU assay) of
old human HSCs. a, 1000 CD34+CD38−CD45RA- cells derived from the bone
marrow of anonymous old adults (58-77 years old) were treated for 3 days with
20μM UA, and colony formations was measured at 15 days post seeding and
replated for secondary CFUs. b, analysis of primary CFUs (P-value < 0.0001)
Nature Aging
https://doi.org/10.1038/s43587-023-00473-3
(n = 14; values are mean ± s.e.m.; two-tailed Student’s t test; * P ≤ 0.05, ** P ≤ 0.01,
*** P ≤ 0.001). c, Assessment of long-term hematopoietic capacity by secondary
CFUs analysis (P-value = 0.0042) (n = 8; values are mean ± s.e.m.; two-tailed
Student’s t test; * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001).
Letter https://doi.org/10.1038/s43587-023-00473-3

Extended Data Fig. 4 | Peripheral blood analyses of UA fed mice. Peripheral counter and normalized to day 0 (WBC: white blood cell; Lym: Lymphocytes; Gra:
blood cell analyses of old mice fed with ctrl (old ctrl) or UA enriched diet (old Granulocytes; Mon: Monocytes; RBC: red blood cells; HGB: hemoglobin) (n = 8;
UA). Analyses were performed at the indicated time points with blood cell values are mean ± s.e.m.;).

Nature Aging
Letter
Extended Data Fig. 5 | Hematopoietic stem and progenitor analyses of old
mice supplemented with UA. a, Gating strategy to identify hematopoietic
stem and progenitor populations from the BM for endpoint analysis. HSCs:
Lineage−cKit+Sca1+ (LKS) CD150+CD48−; Multipotent progenitors (MPPs):
LKS CD150−, Common lymphoid progenitor (CLP): CD127+ cKitlow Sca1low;
Committed progenitors (cKit+) include common myeloid progenitors (CMPs:
Lineage−cKit+Sca1–(KLS–) FcRlowCD34+), granulocyte–macrophage progenitors
(GMPs: KLS-FcRlowCD34 + ), and megakaryocyte-erythroid progenitors (MEPs:
Nature Aging
https://doi.org/10.1038/s43587-023-00473-3
KLS–FcR–CD34–). b, Analyses of lymphoid vs myeloid compartments of primary
transplants at the indicated timepoints (n = 10; values are mean ± s.e.m). c, Donor
chimerism analysis of secondary recipient mice in peripheral blood (Total:
4 weeks P-value = 0.0013; 8 weeks P-value < 0.0001; 12 weeks P-value = 0.0003;
Lymphoid: 4 weeks P-value < 0.0001; 8 weeks P-value < 0.0001; 12 weeks
P-value = 0.0002; Myeloid: 4 weeks P-value = 0.0089; 8 weeks P-value = 0.0049;
12 weeks P-value = 0.0153) (n = 5; values are mean ± s.e.m.; two-tailed Student’s
t test; * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001).
Letter
Extended Data Fig. 6 | Expression of inhibitory receptors markers and
cytokines by virus-specific CD8 + T cells. a, Expression of IFNγ (gp33 + : young
ctrl vs old UA P-value = 0.0112; np396+: young ctrl vs old UA P-value = 0.0034;
old ctrl vs old UA P-value = 0.0160), TNFα (gp33 + :young ctrl vs old ctrl
P-value < 0.0001; young ctrl vs old UA P-value = 0.0003; np396+:young ctrl
vs old ctrl P-value < 0.0001; young ctrl vs old UA P-value < 0.0001) and IL-2
by CD8+ T cells in response to restimulation with gp33 or np396 peptides.
b, Expression of co- inhibitory receptors PD-1 (gp33 + :young ctrl vs old ctrl
Nature Aging
https://doi.org/10.1038/s43587-023-00473-3
P-value < 0.0001; young ctrl vs old UA P-value = 0.0125; old ctrl vs old UA
P-value = 0.0146 np396+:young ctrl vs old ctrl P-value < 0.0001; young ctrl vs old
UA P-value = 0.0025; old ctrl vs old UA P-value = 0.0159)and Lag3 (gp33 + :young
ctrl vs old ctrl P-value < 0.0001; young ctrl vs old UA P-value = 0.0007; old ctrl
vs old UA P-value = 0.0173; np396+:young ctrl vs old ctrl P-value < 0.0001; young
ctrl vs old UA P-value < 0.0001; old ctrl vs old UA P-value = 0.0301) by gp33+ and
np396 CD44+ CD8+ T cells (n = 10; values are mean ± s.e.m.; two-way ANOVA;
* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001).
Letter https://doi.org/10.1038/s43587-023-00473-3

Extended Data Fig. 7 | Viral response of CD8 + T cells derived from old np396+:young ctrl vs old ctrl P-value = 0.0298). b, Proportion of activated KLRG1+
UA treated HSCs. a, Quantification of donor derived (CD45.1+) virus specific within donor derived virus specific CD8+ T. (n = 5; values are mean ± s.e.m.;
(gp33+ or np396+) CD8+ T cells (gp33 + :young ctrl vs old ctrl P-value = 0.0407; two-way ANOVA; * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001).

Nature Aging
Letter https://doi.org/10.1038/s43587-023-00473-3

Extended Data Fig. 8 | UA in vitro treatment modulates T cell function and Green) in CD8+ and CD4 + T cells measured as mean fluorescence intensity (MFI).
mitochondrial profiles. a, Splenocytes derived from old mice were activated c, Analyses of the proportion of cells INFγ+TNFα+ upon restimulation with CD3/
at day0 and restimulated at day5 with CD3/CD28 beads cultured with 5 μM UA. CD28 beads at day 5 (n = 4; values are mean ± s.e.m.; two-tailed paired Student’s
b, Analyses of mitochondrial activity (TMRM) and mass (MTG, mitotracker t test; * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001).

Nature Aging
Letter
Extended Data Fig. 9 | UA modulates mitochondrial recycling in HSCs.
a, Analysis of mitophagy induction in HSCs purified from mito-QC mice
measured as MFI ratio between GFP and mCherry signals(P-value = 0.0037).
(n = 4; two tailed paired Student’s t test; * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001).
b, Expression of CD150 measured by mean fluorescence intensity (MFI) in young
and old HSCs (P-value = 0.0009)(n = 4; values are mean ± s.e.m.; two-tailed
Student’s t test; * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001). c, HSCs derived from
Nature Aging
https://doi.org/10.1038/s43587-023-00473-3
Mx1-PARK2 mice (CD45.2) were transplanted in lethally irradiated recipient mice
(CD45.1/2) together with total bone marrow competitor cells (CD45.1).d, Blood
donor chimerism of primary transplant at the indicated timepoint. e, Lymphoid
vs myeloid compartments of primary transplants at the indicated timepoints
(Ctrl vs PARK2-/- P-value = 0.0002; Ctrl vs PARK2-/- UA P-value = 0.0006) (n = 10;
values are mean ± s.e.m.; two-way ANOVA; * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001).