TISSUE-SPECIFIC STEM CELLS
Direct Toll-Like Receptor-Mediated Stimulation of Hematopoietic
Stem and Progenitor Cells Occurs In Vivo and Promotes
Differentiation Toward Macrophages
JAVIER MEGÍAS,a ALBERTO YÁÑEZ,b SILVIA MORIANO,b JOSÉ-ENRIQUE O’CONNOR,a DANIEL GOZALBO,b
MARÍA-LUISA GILb
a
Laboratorio de Citomica, Unidad Mixta CIPF-UVEG, Centro de Investigacion ‘‘Prı́ncipe Felipe,’’ Valencia,
Spain; bDepartamento de Microbiologı́a y Ecologı́a, Universitat de València, Burjassot, Spain
Key Words. Toll-like receptors • MyD88 • Hematopoietic stem and progenitor cells • Macrophages
ABSTRACT
As Toll-like receptors (TLRs) are expressed by hematopoietic stem and progenitor cells (HSPCs), they may play a
role in hematopoiesis in response to pathogens during infection. We show here that TLR2, TLR4, and TLR9 agonists
(tripalmitoyl-S-glyceryl-L-Cys-Ser-(Lys)4 [Pam3CSK4],
lipopolysaccharide [LPS], and CpG oligodeoxynucleotide
[ODN]) induce the in vitro differentiation of purified murine lineage negative cells (Lin2) as well as HSPCs (identified as Lin2 c-Kit1 Sca-11 IL-7Ra2 [LKS] cells) toward
macrophages (Mph), through a myeloid differentiation
factor 88 (MyD88)-dependent pathway. In order to investigate the possible direct interaction of soluble microorganism-associated molecular patterns and TLRs on HSPCs in
vivo, we designed a new experimental approach: purified
Lin2 and LKS cells from bone marrow of B6Ly5.1 mice
(CD45.1 alloantigen) were transplanted into TLR22/2,
TLR42/2, or MyD882/2 mice (CD45.2 alloantigen), which
were then injected with soluble TLR ligands (Pam3CSK4,
LPS, or ODN, respectively). As recipient mouse cells do
not recognize the TLR ligands injected, interference by
soluble mediators secreted by recipient cells is negligible.
Transplanted cells were detected in the spleen and bone
marrow of recipient mice, and in response to soluble TLR
ligands, cells differentiated preferentially to Mph. These
results show, for the first time, that HSPCs may be
directly stimulated by TLR agonists in vivo, and that the
engagement of these receptors induces differentiation toward Mph. Therefore, HSPCs may sense pathogen or
pathogen-derived products directly during infection,
inducing a rapid generation of cells of the innate immune
system. STEM CELLS 2012;30:1486–1495
Disclosure of potential conflicts of interest is found at the end of this article.
INTRODUCTION
Toll-like receptors (TLRs) constitute a family of pattern-recognition receptors (PRRs) that recognize molecular signatures of
microbial pathogens and function as sensors for infection.
TLRs induce the activation of innate immune responses as well
as the subsequent development of adaptive immune responses,
as TLR signaling is an important component of dendritic cell
(DC) maturation and activation. TLRs are type I membrane
proteins characterized by an ectodomain responsible for recognition of microorganism-associated molecular patterns
(MAMPs) and a cytoplasmic domain homologous to the cytoplasmic region of the interleukin (IL)-1 receptor (TIR-domain).
Signal transduction starts with the recruitment of a set of intracellular TIR-domain-containing adaptors (myeloid differentiation factor 88 [MyD88], Toll-interleukin 1 receptor (TIR) domain containing adaptor protein [TIRAP], TIR-domaincontaining adapter-inducing interferon-b [TRIF] and TIR do-
main-containing adaptor-inducing IFN-b (TRIF)-related adaptor
molecule [TRAM]) that interact with the cytoplasmic TIR domain of the TLRs. MyD88 is the universal adaptor molecule,
shared by all TLRs except TLR3, that triggers expression of
inflammatory cytokine and chemokine genes [1].
Recent findings suggest that TLRs may play a role in
hematopoiesis during infection [2]. Murine hematopoietic
stem cells (HSCs) and their progeny express TLRs, and upon
in vitro exposure to soluble TLR2 and TLR4 ligands are
stimulated to enter cell cycle and acquire lineage markers [3].
Signaling through TLR7/8 induces the differentiation of
human bone marrow CD34þ progenitor cells along the myeloid lineage [4], and the TLR1/2 agonist tripalmitoyl-S-glyceryl-L-Cys-Ser-(Lys)4 [Pam3CSK4] instructs commitment of
human HSCs to a myeloid cell fate [5]. In addition, this
newly described mechanism, which may represent a potential
means for pathogen products to signal the rapid generation of
innate immune cells, has been explored in some in vivo infectious models; production of DCs from murine lymphoid
Author contributions: J.M.: performed experiments, interpreted data, and wrote the manuscript; A.Y.: designed research and interpreted
data; S.M.: performed experiments; J.E.O.: interpreted data; D.G. and M.L.G.: designed research, interpreted data, and wrote the
manuscript.
Correspondence: Marı́a-Luisa Gil, Ph.D., Departamento de Microbiologı́a y Ecologı́a, Universitat de València, Edificio de
Investigacion, C/ Dr. Moliner, 50, 46100 Burjassot, Valencia Spain. Telephone: 34-96-3543410; Fax: 34-96-3544570; e-mail: m.luisa.
[email protected] Received February 17, 2012; accepted for publication April 3, 2012; first published online in STEM CELLS EXPRESS April 17,
C AlphaMed Press 1066-5099/2012/$30.00/0 doi: 10.1002/stem.1110
2012. V
STEM CELLS 2012;30:1486–1495 www.StemCells.com
Megı́as, Yan~ez, Moriano et al.
1487
precursors during herpes infection is TLR9 dependent [6],
expansion of HSCs during vaccinia virus infection is MyD88
dependent [7], TLR-mediated signals play an essential role in
monocyte expansion during systemic Listeria monocytogenes
infection [8], and hematopoietic stem and progenitor cells
(HSPCs) increase markedly following Mycobacterium tuberculosis infection in a TLR2/MyD88-dependent manner [9].
We have previously demonstrated that the fungal pathogen Candida albicans induces the proliferation of HSPCs and
their differentiation toward the myeloid lineage in vitro. This
response requires signaling through TLR2/MyD88 and gives
rise to functional phagocytes that are able to internalize yeasts
and secrete proinflammatory cytokines [10, 11]. Moreover, in
an experimental model of invasive candidiasis in mice, we
have shown that Lin c-Kitþ Sca-1þ IL-7Ra (LKS) cells are
rapidly expanded and new monocyte-derived DCs (moDCs)
and inflammatory macrophages (Mph) are generated in the
spleen in a TLR2-dependent manner [12].
All these results indicate that TLRs control hematopoiesis
during infection and raise the question of whether this is due
to the direct recognition of pathogens by progenitor cells and/
or to secondary effects due to pathophysiological changes during infection. To deal with this issue, we have performed a
new experimental approach to investigate whether direct stimulation of HSPCs via TLRs occurs in vivo. Differentiation of
transplanted wild-type CD45.1 HSPCs in TLR2/, MyD88/
, or TLR4/ CD45.2 mice was determined following in
vivo challenge with TLR ligands. Our results show, for the
first time, that HSPCs are directly stimulated in vivo via
TLRs. The in vivo differentiation of HSPCs in response to
TLR2, TLR4, or TLR9 agonists is similar to the in vitro
results, as in both cases, differentiation toward Mph is
induced. These results indicate that HSPCs can sense pathogen products directly during infection to rapidly replenish the
innate immune compartment and generate more of the mature
cells needed to deal with the pathogen.
MATERIALS
AND
METHODS
Mice
TLR2/, MyD88/, and TLR4/ mice (C57BL/6 background)
provided by Dr. Shizuo Akira (Osaka University, Osaka, Japan)
were bred and maintained at the animal production service facilities (University of Valencia); wild-type C57BL/6 mice (Harlan
Iberica, Barcelona, Spain; http://www.harlan.com) were used as
controls; wild-type CD45.1-positive allotype mice (B6.SJLPtprcaPepcb/BoyCrl strain) were used as lineage negative (Lin)
progenitor cells and LKS cells transplant donors (Charles River
Laboratories, Wilmington, MA; http://www.criver.com). Mice of
both sexes between 8 and 12 weeks old were used, and all assays
involving mice were approved by the Institutional Animal Care
and Use Committee (A1264596506468, Universitat de València).
Purification of Lin2 and LKS Cells
Lin and LKS cells were purified as previously described [10,
11]. Briefly, murine bone marrow was obtained by flushing the
femurs and tibias; cells were depleted of lineage-positive cells by
immunomagnetic cell sorting using MicroBeads (Miltenyi Biotec,
Madrid, Spain; http://www.miltenyibiotec.com): bone marrow
cells were labeled with a cocktail of antibodies against a panel of
lineage antigens (CD5, CD45R [B220], CD11b, Gr-1 [Ly-6G/C],
7-4, and Ter-119), and then cells were purified by negative selection according to the manufacturer’s instructions (Lin progenitor
cells). Purity of the sorted cells was assessed by labeling with
anti-Lin cocktail and by flow cytometry analysis, and no Linþ
cells were detected. LKS cells were purified from Lin cells by
www.StemCells.com
immunomagnetic positive selection using anti-Sca-1 MicroBeads
(Miltenyi Biotec). Purity of the sorted cells was assessed by
labeling with phycoerythrin (PE)-labeled anti-c-Kit monoclonal
antibody (clone 3C1; Miltenyi Biotec) and PE-Cy7-labeled antiIL-7Ra monoclonal antibody (clone SB/119; BD Pharmingen,
San Jose, CA; http://www.bdbiosciences.com). A homogeneous
LKS cell population was purified.
In Vitro Assays
Purified cells were immediately cultured in complete cell culture
medium (RPMI 1640 medium supplemented with 2 mM L-glutamine, 50 lM 2-mercaptoethanol, 5% heat-inactivated fetal bovine
serum (FBS), and 1% penicillin-streptomycin stock solution;
Gibco, Barcelona, Spain; http://www.invitrogen.com) containing
three cytokines: stem cell factor (SCF, 20 ng/ml), Flt-3 ligand
(FL, 100 ng/ml) (Peprotech, Rocky Hill, NJ; https://www.
peprotech.com), and IL-7 (10 ng/ml) (MBL, Woburn, MA; http://
www.mblintl.com). Lin progenitor cells were cultured in flatbottomed 24-well plates at a density of 400,000 cells per well in
0.5 ml, and LKS cells were cultured in round-bottomed 96-well
plates at a density of 50,000 cells per well in 0.2 ml. Cells were
challenged for 7 days with Pam3CSK4 (1 lg/ml), ultrapure Escherichia coli lipopolysaccharide (LPS) (1 lg/ml) or type A CpG
oligodeoxynucleotide (ODN 1585) (10 lg/ml) (InvivoGen, San
Diego, CA; http://www.invivogen.com). Control cultures were
performed without stimuli. At day 4, the culture medium was
refreshed including both the cytokines and the stimuli. At day 7,
cells were collected and analyzed by flow cytometry (see below).
In Vivo Transplantation of CD45.1 Cells
Murine bone marrow was extracted from CD45.1 mice and Lin
progenitor cells or LKS cells were purified as described above.
Approximately 2.5 106 Lin cells in 100 ll of phosphate buffered saline (PBS) (purified from four mice) or 0.4 106 LKS
cells in 100 ll of PBS (purified from eight mice) were intravenously injected into one CD45.2 TLR2/, one CD45.2 TLR4/,
or one CD45.2 MyD88/ knockout mouse. The transplanted mice
were then injected with TLR ligands for 3 days. Each stimulated
mouse received, by intravenous administration, three doses of TLR
ligand at days 0, 1, and 2 after transplantation. TLR2/ mice
were injected with 100 lg Pam3CSK4 per dose, TLR4/ mice
with 100 lg LPS per dose, and MyD88/ mice with 100 lg
ODN per dose (Fig. 3A). Each assay was performed with six Lin
or LKS transplanted mice per group (TLR2/, TLR4/, or
MyD88/); three mice within each group were challenged with
TLR ligands and three mice were used as controls.
Detection and Characterization of CD45.1
Transplanted Cells
Each transplanted mouse was killed at day 3, and the spleen and
bone marrow from femurs and tibias were removed aseptically.
Total spleen cells were obtained by collagenase D treatment of
the organ as previously described [13, 14]. Erythrocytes were
lysed in ammonium chloride buffer. Fc receptors were blocked
with FcR blocking reagent (Miltenyi Biotec), and cells were
depleted of CD45.2 cells by immunomagnetic cell sorting using
biotinylated anti-CD45.2 antibody and anti-biotin magnetic
MicroBeads (both from Miltenyi Biotec). Recovered cells were
microscopically counted, labeled with various combinations of
antibodies, and analyzed by flow cytometry (see below).
Antibodies and Flow Cytometry Analyses
The following antibodies used in flow cytometry analyses were
purchased from Miltenyi Biotec: cocktail of biotinylated anti-lineage antigens (CD5, CD45R [B220], CD11b, Gr-1 [Ly-6G/C], 7-4,
and Ter-119), fluorescein isothiocyanate (FITC)-labeled anti-Sca1 (clone D7), PE-labeled anti-c-Kit (clone 3C1), FITC-labeled
anti-CD45.1 (clone A20), allophycocyanin (APC)-labeled antiCD11c (clone N418), and APC-labeled anti-mPDCA-1 (clone
JF05-1C2.4.1). The following antibodies were from eBioscience
TLRs Directly Activate HSPCs In Vivo
1488
(San Diego, CA; http://www.ebioscience.com): PE-labeled antiCD11b (clone M1/70) and PE-Cy7-labeled anti-F4/80 (clone
BM8), or from BD Pharmingen: PE-Cy7-labeled anti-IL-7Ra
(clone SB/119) and peridinin chlorophyll protein (PerCP)-Cy5.5labeled anti-Ly6C (clone AL-21). Flow cytometry analyses were
performed on a FACSCanto cytometer (BD Biosciences, http://
www.bdbiosciences.com), and the data were analyzed with
FACSDiva and FlowJo 7.6.5. software.
Statistical Analysis
Statistical differences were determined using one-way analysis of
variance (ANOVA) followed by Dunnett’s test for multiple comparisons and two-tailed Student’s test for dual comparisons. Data are
expressed as mean 6 SD. Significance was accepted at *, p < .05
and **, p < .01 level.
RESULTS
In Vitro Differentiation of Lin2 and LKS Cells Is
Altered by Exposure to TLR Ligands and Biased to
Macrophage Production
We investigated the in vitro effect of Pam3CSK4 (which only
activates TLR2), LPS (which only activates TLR4), and ODN
(which only activates TLR9) on differentiation of Lin progenitor cells. Control unchallenged cells from C57BL/6 mice were
cultured for 7 days in complete cell culture medium containing
SCF, FL, and IL-7 to guarantee the survival of stem cells and
common myeloid and lymphoid progenitors; in these conditions,
progenitors differentiate to a mixed population of cells (Fig.
1A, medium, Fig. 1B, white bars). These differentiated cells did
not express the CD19 lymphoid marker (data not shown), indicating that in these conditions, B cells were not produced. Analysis of the expression of CD11b and CD11c allowed the identification of three different cell populations: (a) roughly 25% of
cells were CD11b CD11cþ and also expressed mPDCA-1
(data not shown), therefore corresponding to plasmacytoid DCs
(pDCs), (b) 30% of cells were CD11bþ CD11cþ containing
pre-DCs or classic DCs (cDCs, 12% of cells CD11bþ CD11cþ
Ly6C F4/80) and moDCs (5% of cells CD11bhigh CD11cþ
Ly6Cþ F4/80þ) and (c) 17% of cells were CD11bþ CD11c
containing 5% Ly6Chigh monocytes (Mc, CD11bþ CD11c
Ly6Cþ F4/80). When Lin progenitor cells were cultured in
the same conditions but in the presence of different TLR ligands,
the differentiation pattern was completely different (Fig. 1A,
Pam3CSK4, LPS, and ODN and Fig. 1B, colored bars). The generation of DCs, both pDCs and cDCs, decreased dramatically,
whereas a new major population of Mph (CD11bþ CD11c
Ly6Cþ F4/80þ) appeared. Although the response to TLR2,
TLR4, and TLR9 ligands was qualitatively quite similar, the generation of Mph in response to Pam3CSK4 was higher than to
LPS or ODN. As expected, Lin progenitor cells from TLR2/
mice did not respond to Pam3CSK4, Lin progenitor cells from
TLR4/ mice did not respond to LPS, and Lin progenitor cells
from MyD88/ mice did not respond to Pam3CSK4 or ODN
(Fig. 1B). Interestingly, Lin progenitor cells from MyD88/
mice did not respond to LPS, indicating that in these progenitor
cells, the response to LPS was completely dependent on MyD88,
although TLR4 can also signal through TRIF. It should be noted
that the differentiation of Lin progenitor cells from knockout
mice in the absence of TLR ligands was similar to the differentiation of cells from control mice (Fig. 1B).
As the assayed Lin population contains a variety of transitional intermediates including HSCs, myeloid, and lymphoid
committed progenitors, we next investigated whether stem cells
and uncommitted progenitors would also be responsive to TLR
ligands. LKS cells were purified and cultured in the same conditions as described above for Lin cells in presence or absence of TLR ligands (Fig. 2). In the absence of TLR ligands
(medium), the LKS population from control (Fig. 2A, 2B) and
knockout mice (Fig. 2B) similarly differentiated to a mixed
population of cells. However, analysis of the expression of different markers on the culture cells showed that LKS cells generated fewer pDCs (10%) and more cDCs (34%) and Mc
(16%) than the Lin population. When LKS cells were cultured
in the presence of TLR ligands, the generation of pDCs and
cDCs decreased and a new population of Mph appeared. Again,
the TLR2 ligand induced the strongest production of Mph, and
all responses were completely dependent on MyD88. Overall,
these results clearly indicate that TLR2, TLR4, and TLR9 signaling via MyD88 biases the in vitro differentiation of Lin
and LKS cells toward the production of Mph.
Transplantation and Distribution of CD45.1
Lin2 and LKS Cells in TLR22/2, TLR42/2, and
MyD882/2 CD45.2 Mice: Effect of TLR Ligands
The observed in vitro effect of TLR signaling on HSPCs may
be of biological relevance in vivo. However, direct in vivo
interaction between microbial pathogens, or their ligands, and
TLRs on the HSPCs is difficult to demonstrate as HSPCs
could respond to other stimuli generated when mature
immune cells detect microbial products via their TLRs.
In order to investigate the possible direct interaction of
soluble MAMPs with TLRs on HSPCs in vivo, we designed a
new experimental approach (Fig. 3A). We purified HSPCs
(Lin or LKS cells) from bone marrow of B6Ly5.1
mice (CD45.1 alloantigen) and transplanted these cells into
TLR2/, TLR4/, or MyD88/ mice (CD45.2 alloantigen).
We then injected the mice with one daily dose of Pam3CSK4,
LPS, or ODN, respectively, for 3 days. Using this experimental
approach, the recipient mouse cells do not recognize the TLR
ligand injected, so there should be no interference by cytokines
or soluble mediators secreted by recipient cells. After 3 days,
bone marrow and spleen cells were enriched for CD45.1 cells
by depletion of CD45.2 cells and analyzed by multicolor fluorescence and flow cytometry. Analysis of the expression of
CD45.1 allowed the detection of transplanted cells (Fig. 3B);
approximately 3.3% of the Lin cells and 1.2% of the LKS
cells detected in the spleen and 0.9% of the Lin cells and
0.7% of the LKS cells detected in the bone marrow were donor
derived. A significant increase in CD45.1 cells was detected in
the spleen of TLR2/ mice transplanted with both Lin and
LKS cells and in the bone marrow of TLR2/ mice transplanted with Lin cells following Pam3CSK4 challenge (Fig.
3C). These results indicate that TLR2 signaling resulting from
direct detection of Pam3CSK4 by the transplanted Lin/LKS
cells induces their proliferation and/or improves their survival
in vivo. Although LPS induced an increase in the percentage
of CD45.1 cells in the spleens of TLR4/ mice transplanted
with either Lin or LKS cells, this increase was not statistically
significant. Similarly, treatment of MyD88/ mice with the
TLR9 ligand ODN did not trigger any increase in the frequency of transplanted CD45.1 cells. Taken together, our data
show that CD45.1 cells can be detected in the spleen and in
the bone marrow 3 days after the transplantation, and that in
these locations they are responsive to TLR2 agonists.
Transplanted CD45.1 Lin2 and LKS Cells Respond
to TLR Ligands and Are Directed to Produce Mph
Next, we analyzed the expression of different markers on the
CD45.1 cells in order to determine whether the transplanted
cells differentiated (Figs. 4, 5). Most of the CD45.1 cells in
Megı́as, Yan~ez, Moriano et al.
1489
Figure 1. In vitro differentiation of Lin cells in response to TLR ligands. (A): Lin progenitor cells from C57BL/6 mice were cultured without stimuli (medium) or with Pam3CSK4 (1 lg/ml), LPS (1 lg/ml), or ODN (10 lg/ml) for 7 days, labeled with antibodies, and analyzed by
flow cytometry. Cells were gated as pDCs (CD11b CD11cþ mPDCAþ), cDCs (pre-cDCs or cDCs: CD11bþ CD11cþ Ly6C F4/80), moDCs
(CD11bhigh CD11cþ Ly6Cþ F4/80þ), Mc (CD11bþ CD11c Ly6Cþ F4/80), and Mph (CD11bþ CD11c Ly6Cþ F4/80þ). The indicated percentages refer to total analyzed cells. Ly6C versus F4/80 plots are subgated from the CD11b versus CD11c plot. At least 20,000 events were analyzed in each sample. Results shown are from one representative of three independent experiments. (B): The figure shows the percentages of
pDCs, cDCs, and Mph generated from Lin progenitor cells from C57BL/6, TLR2/, TLR4/, and MyD88/ mice in the presence or absence
of TLR ligands, analyzed following the same schedule as in (A). Data represent means 6 SD, from three experiments. **, p < .01 with respect
to cells incubated with medium alone (medium) within each mouse type. Abbreviations: cDCs, classic dendritic cells; Lin, lineage negative
cells; LPS, lipopolysaccharide; Mc, monocytes; moDCs, monocyte-derived dendritic cells; Mph, macrophages; MyD88, myeloid differentiation
factor 88; CpG ODN, oligodeoxynucleotide; pDCs, plasmacytoid dendritic cells; TLR, toll-like receptors.
www.StemCells.com
1490
TLRs Directly Activate HSPCs In Vivo
Figure 2. In vitro differentiation of LKS cells in response to TLR ligands. (A): LKS cells from C57BL/6 mice were cultured without stimuli
(medium) or with Pam3CSK4 (1 lg/ml), LPS (1 lg/ml), or ODN (10 lg/ml) for 7 days, labeled with antibodies, and analyzed by flow cytometry.
Cells were gated as pDCs (CD11b CD11cþ mPDCAþ), cDCs (pre-cDCs or cDCs: CD11bþ CD11cþ Ly6C F4/80), moDCs (CD11bhigh
CD11cþ Ly6Cþ F4/80þ), Mc (CD11bþ CD11c Ly6Cþ F4/80), and Mph (CD11bþ CD11c Ly6Cþ F4/80þ). The indicated percentages refer
to total analyzed cells. Ly6C versus F4/80 plots are subgated from the CD11b versus CD11c plot. At least 20,000 events were analyzed in each
sample. Results shown are from one representative of three independent experiments. (B): The figure shows the percentages of pDCs, cDCs, and
Mph generated from LKS cells from C57BL/6, TLR2/, TLR4/, and MyD88/ mice in the presence or absence of TLR ligands, analyzed
following the same schedule as in (A). Data represent means 6 SD, from three experiments. *, p < .05, **, p < .01 with respect to cells incubated with medium alone (medium) within each mouse type. Abbreviations: cDCs, classic dendritic cells; LKS, Lin c-Kitþ Sca-1þ IL-7Ra;
LPS, lipopolysaccharide; Mc, monocytes; moDCs, monocyte-derived dendritic cells; Mph, macrophages; MyD88, myeloid differentiation factor
88; CpGODN, oligodeoxynucleotide; pDCs, plasmacytoid dendritic cells; TLR, toll-like receptors.
unstimulated mice did not express markers of stem cells
(roughly, 1%–2% cells were c-Kitþ Sca-1þ IL-7Ra when
LKS cells were transplanted and less than 1% when Lin
cells were transplanted), indicating that after 3 days in vivo,
both Lin and LKS cells had differentiated toward more
mature cells. The cells did not express CD19 and B220
lymphoid markers either (data not shown), indicating that in
these conditions, lymphopoiesis is not promoted. Finally, we
analyzed the expression of myeloid markers and found that
transplanted cells had differentiated toward this lineage, as
significant percentages of CD11bþ cells were detected. Some
cells expressed markers of mature pDCs, cDCs, or Mc, and
the potential differentiation of transplanted cells was quite
similar in TLR2/, TLR4/, or MyD88/ mice (Figs. 4, 5,
Megı́as, Yan~ez, Moriano et al.
1491
Figure 3. Transplantation of CD45.1 Lin and LKS cells into CD45.2 mice and detection of CD45.1 cells in spleen and bone marrow. (A):
Schematic protocol of cell transplantation (as described in Materials and Methods). (B): Three days after transplantation, donor-derived CD45.1
cells were detected in the spleen and bone marrow of CD45.2 control mice (unchallenged with TLR ligands). Dot plots show FSC against
CD45.1 expression of the purified spleen and bone marrow cells. Percentage of recovered CD45.1 cells is calculated as follows: total number of
recovered cells 100/(total number of transplanted cells), where total number of recovered cells ¼ % of CD45.1 cells determined by flow
cytometry total number of purified cells from spleen or bone marrow/100. Indicated percentages are the mean 6 SD of nine mice. (C): Percentages of recovered CD45.1 cells from the spleen and bone marrow of TLR2/, TLR4/, and MyD88/ knockout mice transplanted with
CD45.1 Lin progenitor cells or LKS cells and stimulated daily with 100 lg/day of Pam3CSK4, LPS, or ODN, respectively, for 3 days. Data
represent means 6 SD, from three experiments. **, p < .01 with respect to CD45.1 cells recovered from transplanted unstimulated control mice.
Abbreviations: FSC, forward scatter; HSPCs, hematopoietic stem and progenitor cells; Lin, lineage negative cells; LKS, Lin c-Kitþ Sca-1þ IL7Ra; LPS, lipopolysaccharide; MyD88, myeloid differentiation factor 88; CpGODN, oligodeoxynucleotide; TLR, toll-like receptors.
unstimulated). However, after the injection of Pam3CSK4,
LPS, and ODN, into the TLR2/, TLR4/, and MyD88/
mice, respectively, we detected the generation of a significant
percentage of CD45.1 Mph (CD11bþ CD11c Ly6Cþ F4/
80þ, 10%–20% CD45.1-positive cells) from both Lin and
LKS cells (Figs. 4, 5). In response to Pam3CSK4 (TLR2/
mice), an increase in total CD11bþ cells, as well as Mph, was
detected, indicating that transplanted cells proliferate and differentiate toward Mph. However, in response to LPS (TLR4/
mice) and to ODN (MyD88/ mice), the number of
CD11b cells was not increased (for LKS transplanted cells,
Fig. 5) and the increase in Mph correlated with a decrease in
Mc (statistically significant in most cases, Figs. 4, 5). This
result suggests that LPS and ODN induce less proliferation
than Pam3CSK4, but that all stimuli induce differentiation to
Mph, resulting in a decrease in the percentage of Mc. It
should be noted that the percentage of pDCs generated from
Lin cells was decreased both in the spleen and in the bone
marrow of TLR2/, in response to Pam3CSK4 (Fig. 4).
Overall, the pattern of differentiation detected in vivo correlates with the in vitro assays, both in the percentages and
the cell types that are generated in response to Pam3CSK4,
LPS, and ODN. These results demonstrate that HSPCs, in the
bone marrow and in the spleen, are directly stimulated by
TLR2, TLR4, and TLR9 agonists, and that the engagement of
these receptors gives rise preferentially to Mph.
www.StemCells.com
DISCUSSION
Under physiological conditions, the process of HSC selfrenewal, as well as their conversion into lineage-committed
progenitors, is tightly controlled to maintain daily blood cell
production. Many cytokines and transcription factors ‘‘finetune’’ the proliferation of HSPCs and their differentiation into
mature myeloid and lymphoid cells [15]. However, hematopoiesis can be dramatically altered during infections, which
influence numbers and types of cells that are produced. During most bacterial, viral, and fungal infections, myelopoiesis
becomes predominant with inhibition of other lineage (lymphoid and erythroid) development, and this is accompanied by
alterations of the cellular composition and/or phenotype of
bone marrow HSPCs [16, 17].
Additional perspective on hematopoiesis during infection
has come from the discovery that murine and human HSPCs
express functional TLRs, and that TLR signals provoke cell
cycle entry and myeloid differentiation in vitro [3–5, 10–12].
HSPCs expansion and alterations in hematopoiesis during
infection have been described in several models of bacterial,
viral, and fungal infection, although the contribution of TLR
signaling to this phenomenon is still a matter of discussion
[16, 17].
1492
TLRs Directly Activate HSPCs In Vivo
Figure 4. In vivo differentiation of CD45.1 Lin progenitor cells in response to TLR ligands. CD45.1 Lin progenitor cells were transplanted
into TLR2/, TLR4/, and MyD88/ knockout mice and stimulated daily with 100 lg Pam3CSK4, LPS, or ODN, respectively, for 3 days.
Afterward, recovered cells from the spleen and bone marrow of mice were labeled with antibodies and analyzed by flow cytometry. The CD45.1
population was gated and cells in this gate were identified as CD11bþ (total CD11bþ cells), pDCs (CD11b CD11cþ mPDCAþ), cDCs (pre-DCs
or cDCs: CD11bþ CD11cþ Ly6C F4/80), Mc (CD11bþ CD11c Ly6Cþ F4/80), and Mph (CD11bþ CD11c Ly6Cþ F4/80þ). Percentages
refer to the total analyzed CD45.1 population. At least 10,000 CD45.1 events were analyzed in each spleen sample and 2,000 in each bone marrow sample. Data represent means 6 SD, from three experiments. *, p < .05; **, p < .01 with respect to CD45.1 cells within each cell type
recovered from transplanted nonstimulated mice. Abbreviations: cDCs, classic dendritic cells; Lin, lineage negative cells; LPS, lipopolysaccharide; Mc, monocytes; Mph, macrophages; MyD88, myeloid differentiation factor 88; CpGODN, oligodeoxynucleotide; pDCs, plasmacytoid dendritic cells; TLR, toll-like receptors.
Although most of the infection models demonstrate that
TLR-mediated signals play an essential role in the control of
HSPC expansion [6–8, 12], other authors [18] have described
that the expansion of HSPCs following bacterial infection occurs
in the absence of TLR signaling. It is important to notice that
the interpretation of in vivo results is difficult as TLR//
MyD88/ mice are more susceptible to most of the infections.
Therefore, comparison of hematopoiesis between control and
knockout mice during infection may reflect different tissue invasion by the microorganism, and consequently, differences in
secretion of cytokines that can regulate hematopoiesis, a
response to the pathogen that is also TLR/MyD88 mediated.
Megı́as, Yan~ez, Moriano et al.
1493
Figure 5. In vivo differentiation of CD45.1 Lin c-Kitþ Sca-1þ IL-7Ra (LKS) cells in response to TLR ligands. CD45.1 LKS cells were
transplanted into TLR2/, TLR4/, and MyD88/ knockout mice and stimulated daily with 100 lg Pam3CSK4, LPS, or ODN, respectively,
for 3 days. Afterward, recovered cells from the spleen and bone marrow of mice were labeled with antibodies and analyzed by flow cytometry.
The CD45.1 population was gated and cells in this gate were identified as CD11bþ (total CD11bþ cells), pDCs (CD11b CD11cþ mPDCAþ),
cDCs (pre-DCs or cDCs: CD11bþ CD11cþ Ly6C F4/80), Mc (CD11bþ CD11c Ly6Cþ F4/80), and Mph (CD11bþ CD11c Ly6Cþ F4/
80þ). Percentages refer to the total analyzed CD45.1 population. At least 2,000 CD45.1 events were analyzed in each spleen sample, and 2,000
in each bone marrow sample. Data represent means 6 SD, from three experiments. *, p < .05; **, p < .01 with respect to CD45.1 cells within
each cell type recovered from transplanted unstimulated mice. Abbreviations: cDCs, classic dendritic cells; LPS, lipopolysaccharide; Mc, monocytes; Mph, macrophages; MyD88, myeloid differentiation factor 88; CpGODN, oligodeoxynucleotide; pDCs, plasmacytoid dendritic cells; TLR,
toll-like receptors.
Thus, in conventional in vivo models, the alterations in
hematopoiesis and in the HSPC populations during infection
can be explained by at least two mechanisms: (a) MAMPs
may directly induce HSPCs proliferation and differentiation,
as suggested by the in vitro results, or alternatively (b) the
alterations could be caused by an indirect effect, due to pathowww.StemCells.com
physiological changes during infection. These possibilities are
not mutually exclusive, and both of them may involve TLR
recognition of the pathogen. In this work, we deal with the
issue of whether interaction of pathogens or the MAMPs
released by them, directly on progenitor cells, can control
hematopoiesis.
TLRs Directly Activate HSPCs In Vivo
1494
Purified Lin cells (containing stem and all the progenitor
cells) or the LKS population (containing stem and uncommitted progenitors) from the bone marrow of B6Ly5.1 mice
(CD45.1 alloantigen) were transplanted into TLR2/,
TLR4/, or MyD88/ mice (CD45.2 alloantigen), which
where then injected with Pam3CSK4, LPS, or ODN, respectively. Using this experimental approach, the cells of the recipient mice do not recognize the TLR agonist injected, and
therefore any effects of cytokines or soluble mediators
secreted by surrounding cells were excluded. Moreover, the
recipient mice were not irradiated, in order to avoid an
inflammatory environment that may generate artifact results.
After 3 days, transplanted cells were detected in the spleen
and in the bone marrow of mice. In these in vivo conditions
and in the absence of TLR challenge, the HSPCs differentiated toward the myeloid lineage, as most of the cells had lost
stem cell markers, did not express lymphoid markers, but
most of them expressed myeloid markers. The differentiated
cells expressed the markers of mature pDCs, cDCs, or Mc.
This result is in accordance with Massberg et al. [19] who
showed that migratory HSPCs give rise to myeloid cells in
peripheral tissues.
However, when each knockout mouse type was injected
with the corresponding TLR agonist, the emergence of a new
population of Mph was found. In this model, the differentiation of CD45.1 cells in response to the injected ligands can
only be due to a direct ligand-TLR interaction on transplanted
Lin or LKS cells and not to an indirect effect of TLR
ligands on cells of the recipient knockout mice. The differentiation to Mph was similar in response to all TLR agonists
assayed, although an increase in the total number of CD45.1positive cells was found only in response to the TLR2 ligand.
In addition, the increase in Mph was not accompanied by a
decrease in Mc in response to the TLR2 ligand, further supporting the idea that Pam3CSK4 induces higher proliferation
of CD45.1 cells than LPS or ODN. The stronger response to
Pam3CSK4, as compared with LPS, is in accordance with the
higher expression of TLR2 than TLR4 on long-term repopulating HSCs and LKS cells [3, 12].
The in vivo results clearly correlate with the in vitro assays,
as TLR2, TLR4, and TLR9 agonists induce the differentiation
of Lin or LKS cells toward Mph. Therefore, signaling through
all the TLRs assayed induces the differentiation of HSPCs toward the same type of cell, although the response was stronger
for TLR2 as compared with TLR4 and TLR9. In a previous
work, we demonstrated that C. albicans drives the differentiation of Lin cells toward moDCs in a TLR2/MyD88 and Dectin-1 (a PRR that recognizes the b-glucan of yeasts)-dependent
manner in vitro [12]. Therefore, the response to C. albicans
that is recognized simultaneously by both receptors [20] is different to the response to a specific TLR2 agonist, indicating
that hematopoietic progenitor cells may integrate signals from
different PRRs, and therefore the response may be pathogen
specific. It should be noted that Dectin-1, in contrast to TLRs,
is not expressed on the most primitive stem cells (side population), although the expression of this receptor is acquired by
REFERENCES
1
Kawai T, Akira S. The role of pattern-recognition receptors in innate
immunity: Update on Toll-like receptors. Nat Immunol 2010;11:
373–384.
2
King KY, Goodell MA. Inflammatory modulation of HSCs: Viewing
the HSC as a foundation for the immune response. Nat Rev Immunol
2011;11:685–692.
progenitors before their differentiation to mature cells, as Lin
cells express significant levels of Dectin-1 [12].
The role of TLR2, TLR4, and TLR9 in the differentiation
of Lin and LKS cells in vitro is in line with several previous
reports showing that TLRs have a role in hematopoiesis during infection [3–12]. However, this is the first report showing
that HSPCs use TLRs to directly sense pathogen-derived
products in vivo and drive their differentiation toward myeloid lineage inducing the production of Mph.
The ability of stem and progenitor cells to sense pathogen
products may be protective, allowing rapid mobilization and
generation of cells of the innate immune system, in a pathogen-specific manner. Moreover, in the case of local infections,
migratory HSPCs that sense microbial danger signals in peripheral tissues can proliferate within that pathogen-challenged
location and contribute to the supply of effector cells [21].
Therefore, TLR ligation directly activates HSPCs in vivo
suggesting a mechanism by which pathogens may modulate
hematopoiesis in real time. This points to a selection of innate
immune populations during the infectious process, an idea
that sits outside the current dogma of innate immune function,
but which is gaining momentum in the literature.
CONCLUSIONS
Recent findings indicate that HSPCs respond to infection via
multiple pathways. In this context, we have presented the first
in vivo demonstration of direct HSPCs stimulation via TLRs.
We showed here that TLR2, TLR4, and TLR9 agonists
(Pam3CSK4, LPS, and ODN) induce the differentiation of
purified murine Lin as well as HSPCs (identified as LKS
cells) toward Mph both in vivo and in vitro. This new mechanism represents a potential means for the host to detect pathogen products and to signal the rapid generation of innate
immune cells within the bone marrow or other tissues containing HSPCs.
ACKNOWLEDGMENTS
This work was supported by Grant SAF2010-18256 (Ministerio
de Economı́a y Competitividad, Spain). Javier Megı́as and
Alberto Ya~nez are recipients of fellowships from Fundacion
Bancaja and Ministerio de Educacion y Ciencia, respectively.
We are grateful to the SCSIE (Servicio Central de Soporte a la
Investigacion Experimental), University of Valencia, for technical assistance.
DISCLOSURE OF POTENTIAL
CONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.
3
Nagai Y, Garrett K, Ohta S et al. Toll-like receptors on hematopoietic
progenitor cells stimulate innate immune system replenishment. Immunity 2006;24:801–812.
4
Sioud M, Fløisand Y, Forfang L et al. Signaling through toll-like receptor 7/8 induces the differentiation of human bone marrow CD34þ
progenitor cells along the myeloid lineage. J Mol Biol 2006;364:
945–954.
5
De Luca K, Frances-Duvert V, Asensio M et al. The TLR1/2 agonist
PAM(3)CSK(4) instructs commitment of human hematopoietic stem
cells to a myeloid cell fate. Leukemia 2009;23:2063–2074.
Megı́as, Yan~ez, Moriano et al.
6
7
8
9
10
11
12
13
Welner R, Pelayo R, Nagai Y et al. Lymphoid precursors are directed
to produce dendritic cells as a result of TLR9 ligation during herpes
infection. Blood 2008;112:3753–3761.
Singh P, Yao Y, Weliver A et al. Vaccinia virus infection modulates
the hematopoietic cell compartments in the bone marrow. Stem Cells
2008;26:1009–1016.
Serbina N, Hohl T, Cherny M et al. Selective expansion of the monocytic lineage directed by bacterial infection. J Immunol 2009;183:
1900–1910.
Choi HH, Kim KK, Kim KD et al. Effects of mycobacterial infection
on proliferation of hematopoietic precursor cells. Microbes Infect
2011;13:1252–1260.
Ya~nez A, Murciano C, O’Connor JE et al. Candida albicans triggers
proliferation and differentiation of hematopoietic stem and progenitor
cells by a MyD88-dependent signaling. Microbes Infect 2009;11:
531–535.
Ya~nez A, Flores A, Murciano C et al. Signalling through TLR2/
MyD88 induces differentiation of murine bone marrow stem and progenitor cells to functional phagocytes in response to Candida albicans.
Cell Microbiol 2010;12:114–128.
Ya~nez A, Megı́as J, O’Connor JE et al. Candida albicans induces
selective development of macrophages and monocyte derived dendritic
cells by a TLR2 dependent signalling. PLoS One 2011;6:e24761.
Murciano C, Villam
on E, Gozalbo D et al. Toll-like receptor 4 defective mice carrying point or null mutations do not show increased sus-
www.StemCells.com
1495
14
15
16
17
18
19
20
21
ceptibility to Candida albicans in a model of hematogenously
disseminated infection. Med Mycol 2006;44:149–157.
Murciano C, Ya~
nez A, Gil ML et al. Both viable and killed Candida
albicans cells induce in vitro production of TNF-alpha and IFNgamma in murine cells through a TLR2-dependent signalling. Eur
Cytokine Netw 2007;18:38–43.
Iwasaki H, Akashi K. Myeloid lineage commitment from the hematopoietic stem cell. Immunity 2007;26:726–740.
Baldridge MT, King KY, Goodell MA. Inflammatory signals regulate
hematopoietic stem cells. Trends Immunol 2011;32:57–65.
Boiko JR, Borghesi L. Hematopoiesis sculpted by pathogens: Toll-like
receptors and inflammatory mediators directly activate stem cells.
Cytokine 2012;57:1–8.
Scumpia PO, Kelly-Scumpia KM, Delano MJ et al. Cutting edge: Bacterial infection induces hematopoietic stem and progenitor cell expansion in the absence of TLR signaling. J Immunol 2010;184:2247–2251.
Massberg S, Schaerli P, Knezevic-Maramica I et al. Immunosurveillance by hematopoietic progenitor cells trafficking through blood,
lymph, and peripheral tissues. Cell 2007;131:994–1008.
Goodridge H, Underhill D. Fungal recognition by TLR2 and Dectin-1.
Handb Exp Pharmacol 2008;183:87–109.
Mazo IB, Massberg S, von Andrian UH. Hematopoietic stem and progenitor cell trafficking. Trends Immunol 2011;32:493–503.