Pathogen-Induced TLR4-TRIF Innate Immune Signaling in ... - Cell Press

110 downloads 0 Views 11MB Size Report
Jul 20, 2017 - Please cite this article in press as: Takizawa et al., Pathogen-Induced TLR4-TRIF Innate Immune Signaling in Hematopoietic Stem Cells ...
Article

Pathogen-Induced TLR4-TRIF Innate Immune Signaling in Hematopoietic Stem Cells Promotes Proliferation but Reduces Competitive Fitness Graphical Abstract

Authors Hitoshi Takizawa, Kristin Fritsch, Larisa V. Kovtonyuk, ..., A´lvaro Gomariz, Ce´sar Nombela-Arrieta, Markus G. Manz

Gram-negative bacteria LPS TLR4

Correspondence

LPS

[email protected] (H.T.), [email protected] (M.G.M.)

Myd88 TRIF TLR4

HSC

Signal inhibition

ROS p38MAPK Proliferative stress

Functional HSC maintenance Impaired HSC fitness

Highlights d

Direct TLR4 activation in HSCs induces HSC cycling and inflammatory responses

d

Sustained TLR4 activation in HSCs impairs their competitive repopulating ability

d

d

LPS and S. Typhimurium cause proliferative stress in HSCs via TLR4-TRIF signals Inhibition of TLR4-TRIF-ROS-p38 signaling prevents LPSinduced HSC dysfunction

Takizawa et al., 2017, Cell Stem Cell 21, 225–240 August 3, 2017 ª 2017 Elsevier Inc. http://dx.doi.org/10.1016/j.stem.2017.06.013

In Brief Takizawa et al. show that self-renewing hematopoietic cells directly sense gramnegative bacterial infection through Tolllike receptor 4 (TLR4) activation, which leads to impaired function via proliferative stress. Genetic and pharmacological blockage of the TLR4TRIF-ROS-p38 axis can maintain HSC function without disrupting emergency granulopoietic responses to infection.

Cell Stem Cell

Article Pathogen-Induced TLR4-TRIF Innate Immune Signaling in Hematopoietic Stem Cells Promotes Proliferation but Reduces Competitive Fitness Hitoshi Takizawa,1,2,* Kristin Fritsch,1 Larisa V. Kovtonyuk,1 Yasuyuki Saito,1 Chakradhar Yakkala,1 Kurt Jacobs,3 Akshay K. Ahuja,3 Massimo Lopes,3 Annika Hausmann,4 Wolf-Dietrich Hardt,4 A´lvaro Gomariz,1 Ce´sar Nombela-Arrieta,1 and Markus G. Manz1,5,* 1Hematology,

University Hospital Zurich and University of Zurich, 8091 Zurich, Switzerland Research Center for Medical Sciences, Kumamoto University, Kumamoto 860-0811, Japan 3Institute of Molecular Cancer Research, University of Zurich, 8057 Zurich, Switzerland 4Institute of Microbiology, Department of Biology, ETH Zurich, 8093 Zurich, Switzerland 5Lead Contact *Correspondence: [email protected] (H.T.), [email protected] (M.G.M.) http://dx.doi.org/10.1016/j.stem.2017.06.013 2International

SUMMARY

Bacterial infection leads to consumption of shortlived innate immune effector cells, which then need to be replenished from hematopoietic stem and progenitor cells (HSPCs). HSPCs express pattern recognition receptors, such as Toll-like receptors (TLRs), and ligation of these receptors induces HSPC mobilization, cytokine production, and myeloid differentiation. The underlying mechanisms involved in pathogen signal transduction in HSCs and the resulting biological consequences remain poorly defined. Here, we show that in vivo lipopolysaccharide (LPS) application induces proliferation of dormant HSCs directly via TLR4 and that sustained LPS exposure impairs HSC self-renewal and competitive repopulation activity. This process is mediated via TLR4TRIF-ROS-p38, but not MyD88 signaling, and can be inhibited pharmacologically without preventing emergency granulopoiesis. Live Salmonella Typhimurium infection similarly induces proliferative stress in HSCs, in part via TLR4-TRIF signals. Thus, while direct TLR4 activation in HSCs might be beneficial for controlling systemic infection, prolonged TLR4 signaling has detrimental effects and may contribute to inflammation-associated HSPC dysfunction. INTRODUCTION Hematopoietic homeostasis is maintained by hematopoietic stem cells (HSCs) that produce all blood cells through intermediate, subsequently lineage-committed progenitors (Kondo et al., 2003). In contrast to highly proliferating progenitors, a significant fraction of HSCs do not contribute to hematopoiesis in steady state and are in quiescence with low metabolic activity, a presumably ‘‘safeguard’’ mechanism to protect HSCs from genotoxic reagents and maintain integrity of HSC function (Busch

et al., 2015; Trumpp et al., 2010). Self-renewal and differentiation of HSCs are tightly regulated by intrinsic and extrinsic factors, and loss of this regulation leads to hematopoietic aplasia or neoplasia, respectively (Orkin and Zon, 2008). HSC function is preserved in a bone marrow (BM) microenvironment, referred to as ‘‘niche,’’ in which various type of hematopoietic and nonhematopoietic cells express vital factors for HSC homeostasis (Morrison and Scadden, 2014). During hematopoietic stress such as infection and inflammation, where short-lived mature hematopoietic immune effector cells are activated and consumed, hematopoietic stem and progenitor cells (HSPCs) in the BM need to tailor their cellular output to meet the enhanced demand of blood production. To this end, they need to integrate respective signals, ultimately stemming via various signaling cascades from the sites of inflammation. Recent findings have proposed two possible, not mutually exclusive regulatory mechanisms of HSPC response during systemic bacterial or viral infections: (1) a cytokine-mediated mechanism in which pro-inflammatory cytokines are secreted at peripheral, immune active sites and/or locally in the BM and induce proliferation of HSPCs via their respective cytokine receptors and the downstream signaling cascades (Baldridge et al., 2010; Burberry et al., 2014; Essers et al., 2009; Manz and Boettcher, 2014; Sato et al., 2009; Takizawa et al., 2012); and (2) a direct sensing mechanism, in which HSPCs themselves recognize pathogenassociated molecular patterns (PAMPs) via the respective pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs), and increase their proliferation, differentiation, and migration (Liu et al., 2015; Massberg et al., 2007; Megı´as et al., 2012; Nagai et al., 2006; Schmid et al., 2011). Interestingly, as a combination of both mechanisms, a recent study revealed that, upon direct TLR activation, HSPCs are capable of producing an array of pro-inflammatory cytokines, which, likely in a paracrine, self-sustaining manner, enhance proliferation and myeloid differentiation under neutropenic conditions in vivo (Zhao et al., 2014). However, it remained a thus-far unresolved fundamental question, if most upstream, bona fide long-term, self-renewing HSCs are able to directly sense PAMPs and, if so, what would be the consequences for HSC function. We here set out to Cell Stem Cell 21, 225–240, August 3, 2017 ª 2017 Elsevier Inc. 225

A 0

WT

C

2x/4x PBS/ LPS

BrdU

2

3

BrdU uptake

Day

BM analysis

-7

-5

-3

0

-1

7

Group A PBS PBS 5FU Group B LPS LPS 5FU Group C LPS LPS 5FU

4 wks

chase

Survival

5FU 5FU 5FU

100

** *

HSC 100

BrdU(-) 4xLPS

2xLPS

100

80

**

*

*

80

*

Survival (%)

B

80

BrdU+ (%)

4xPBS 60

4xPBS

40

60

Group A (n=6) Group B (n=7) Group C (n=5)

20

4xLPS

*

40

40

2xLPS

ns

*

60

0 0

20

5

0

0 102

103

104

0

105

HSC

MPP

LKS

F IR

BrdU

Analysis of BrdU retention in BM HSC

*

4xPBS/ LPS BM analysis

***

100

Tlr4-/-

2

3

BrdU uptake

E

PBS

80

chase

LPS

100 80

PBS-WT PBS-Tlr4-/LPS-WT LPS-Tlr4-/-

60 40 20

10 1

10 2

10 3

10 410

0

10 1

10 2

10 3

40

0

Donor

10 4

BrdU

Tlr4-/-

WT

Recipient

PBS-Tlr4-/-

WT

Tlr4-/-

LPS-Tlr4-/-

WT HSC or MPP 2-3x PBS/ LPS 2

4

HSC Tx (Donor LKS CD34- gated)

BM analysis 6 days

I

MPP Tx (Donor LKS CD34+ gated)

4 3 2 1 0 div.

**

4 3 2 1 0 div.

CD16/32

CD150

**

100

PBS LPS

CFSE

Analysis of divisional history with CFSE

Fraction (%)

0

Tlr4-/-

H

60

20

0 10 0

*

4 wks BrdU+ HSCs (%)

0

-8

20

LK

D WT : Tlr4-/- (1 : 1) BM chimera

15

Days

BrdU

G

10

20

50

0-1x divided cells

0 CFSE

2-5x divided cells

PBS

LPS HSC

PBS

LPS MPP

Figure 1. LPS Increases Proliferation of HSCs via Direct TLR4 Activation (A) Experimental scheme of the BrdU label retention assay for results depicted in (B). WT mice were treated with 9 mg/kg BrdU i.p. once and fed with 800 mg/mL BrdU-containing water for 2 weeks. Mice were then treated with PBS or LPS injections two or four times, and BrdU label retention was determined in BM populations 1 week after the last injection. (B) Representative histogram of BrdU label in HSCs (LKSCD34–CD150+) from animals injected with 4 3 PBS, 2 3 LPS, and 4 3 LPS. Right panel depicts percentage of BrdU+ HSCs (LKSCD34–CD150+), MPPs (LKSCD34+), LKS, and LK cells (4 3 PBS: n = 3; 2 3 LPS: n = 3; 4 3 LPS: n = 4). Each dot represents data from individual animals. (legend continued on next page)

226 Cell Stem Cell 21, 225–240, August 3, 2017

determine this by applying in vivo lipopolysaccharide (LPS), a conserved paradigmatic gram-negative bacterial product in context of signaling deficiencies for LPS through its cognate receptor TLR4, and the respective downstream adaptor molecules MyD88 and TRIF. RESULTS Systemic LPS Application Enhances HSC Proliferation Directly via TLR4 It was shown that HSPCs (Lin–c-Kit+Sca-1+, thereafter called LKS, or LKS CD135–) express TLR4 (Nagai et al., 2006; Zhao et al., 2014), but it remained unclear whether long-term HSCs express TLR4. In agreement with a recent report (Liu et al., 2015), we found that highly enriched HSC populations (LKS CD34–CD150+) that contain long-term, multi-lineage repopulating cells with a 25% purity (Morita et al., 2010), uniformly express the TLR4/MD-2 complex at levels comparable to or higher than multipotent and myeloid committed hematopoietic progenitor cells (Figures S1A and S1B). Moreover, we demonstrate that mRNA of TLR4 and genes such as MYD88 (myeloid differentiation primary response 88) and TRIF (also called TICAM-1), involved in respective TLR4 downstream signaling, are higher expressed in HSCs than in all other cell populations tested, including mature dendritic and myeloid cells (Figure S1C). By applying single-cell divisional tracking and subsequent evaluation of biological HSC readout, i.e., self-renewal and multi-lineage reconstitution, we previously demonstrated that dormant HSCs are recruited to proliferation upon systemic LPS injections (Takizawa et al., 2011). To extend these findings, wild-type (WT) mice were treated for 2 weeks with BrdU (5-bromo-20 -deoxyuridine) to label DNA of dividing cells (Cheshier et al., 1999) and were subsequently intraperitoneally (i.p.) injected with PBS or 35 mg LPS every other day for two or four times to determine BrdU retention or loss upon proliferation (Figure 1A). In PBStreated animals, almost all HSCs (LKS CD34–CD150+) in BM retained the BrdU label over the 2-week chasing period, indicating their quiescence over this period of time. In contrast to this and consistent with our previous findings (Takizawa et al., 2011), two and four-time systemic LPS injections lead to significantly decreased BrdU retention in HSCs, thus demonstrating directly or indirectly LPS-induced proliferation (Figure 1B). This was also the case for multipotent progenitors (MPP; LKS CD34+), LKS,

and myeloid progenitors (Lin–c-Kit+Sca-1–, thereafter called LK) as evidenced by loss of the BrdU label upon LPS application. To further evaluate LPS-induced HSPC proliferation in a functional assay, WT mice were injected with LPS or PBS on day 3 and 1 before application of 150 mg/kg 5-fluorouracil (5-FU) at day 0 and +7 (Figure 1C). Mortality of mice at day +10 and beyond was significantly higher when injected with LPS compared to the PBS control (Figure 1C). As increased proliferation renders HSPCs more sensitive to anti-proliferative chemotherapeutic drugs, these data suggest that LPSenhanced proliferation and subsequent 5-FU-mediated HSPC death (Essers et al., 2009; Kovtonyuk et al., 2016). Interestingly, reduced survival was not observed in animals pre-treated with LPS on days 7 and 5 prior to 5-FU application, indicating that LPS-induced HSPC cycling is likely of short duration. Both hematopoietic and non-hematopoietic cells such as endothelial cells (Boettcher et al., 2014) or mesenchymal stromal cells (Shi et al., 2011) express functional TLR4 (Takizawa et al., 2012, 2011). To test the contribution of hematopoietic cell-expressed TLR4 to HSC proliferation, we generated mixed BM chimeric mice by reconstituting lethally irradiated TLR4-deficient (Tlr4/) mice (CD45.1+/2+) with 1 3 106 WT (CD45.1+) and Tlr4/ (CD45.2+) total BM cells in a 1 to 1 ratio (WT;Tlr4/ BM chimeric Tlr4/ mice). Upon establishment of hematopoietic homeostasis after 2–3 months, mice were treated for 2 weeks with BrdU, followed by four applications of PBS or LPS over 1 week, respectively (Figure 1D). Within the same PBS-treated recipients, WT and Tlr4/ HSCs retained BrdU similarly. In contrast, proliferation was enhanced upon LPS treatment in both WT and Tlr4/ HSCs in the same animal. However, WT HSCs proliferated significantly more (Figures 1E and 1F). Thus, while some proliferation depends on hematopoietic cell secreted factors from WT hematopoiesis in these chimeric animals (the delta between proliferation of TLR4/ HSCs upon PBS and LPS treatment), a significant, about equally sized part of proliferation is driven by direct TLR4 stimulation of HSCs (the delta between proliferation of WT and TLR4/ HSCs upon LPS treatment). In order to further test TLR4-mediated direct HSC activation, we employed CFSE (Carboxyfluorescein diacetate succinimidyl ester)-based, single division-sensitive HSC divisional tracking (Takizawa et al., 2011). Highly purified WT HSCs (LKS CD34CD48–CD41–CD150+) or MPPs (LKS CD34+) were labeled with CFSE and transferred into non-irradiated

(C) Survival of mice treated as indicated with PBS (group A: day 3 and 1; n = 6) or LPS (35 mg) injections (group B: day 3 and 1; n = 7; group C: day 7 and 5; n = 5) followed by i.p. application of 150 mg/kg 5-FU at day 0 and 7. (D) Experimental scheme of BrdU label retention assay for results depicted in (E) and (F). Tlr4/ mice were reconstituted with WT and Tlr4/ BM in a 1:1 ratio (mixed WT;Tlr4/ BM hematopoietic chimeric Tlr4/ mice). Eight weeks after reconstitution, the mixed BM chimeric mice were given BrdU-containing water for 2 weeks and injected with PBS or LPS (50 mg) four times followed by BM analysis 1 week after the last injection. (E) Representative histogram of BrdU label in WT and Tlr4/ HSCs (LKS CD34–CD150+) from PBS- or LPS-treated mixed WT;Tlr4/ BM hematopoietic chimeric Tlr4/ mice. (F) Percentage of BrdU+ HSCs derived from WT or Tlr4/ donors in mixed BM chimeric Tlr4/ mice upon PBS or LPS injections (PBS, n = 7; LPS, n = 8). (G) Experimental scheme of CFSE-based cell divisional tracking. Tlr4/ mice were transplanted with CFSE-labeled WT HSCs (LKS CD34–CD48–CD150+) or MPPs (LKS CD34+) and i.p. injected two or three times with PBS or 100 mg LPS followed by BM analysis at day 6. (H) Representative fluorescence-activated cell sorting (FACS) plots of CFSE dilution versus the respective marker indicated, in donor-derived HSCs (LKS CD34–) or MPP (LKS CD34+) from PBS (blue)- or LPS (red)-treated mice that had been previously transplanted with HSCs (left) or MPP (right). (I) Percentage of zero to one time divided cells and proliferating (two to five times divided) cells within donor-derived HSCs (PBS, n = 8; LPS, n = 9) and MPP (PBS, n = 4; LPS, n = 4). Data are pooled from two to three independent experiments. ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed t test). Error bars represent SEM. See also Figures S1 and S2.

Cell Stem Cell 21, 225–240, August 3, 2017 227

steady-state Tlr4/ recipients. Mice were i.p. treated with PBS or LPS starting at 2 days after cell transfer and analyzed at day 6 to determine divisional history of donor-derived HSC or MPPs based on CFSE dilution (Figure 1G). Almost all of transferred HSCs were found quiescent in the non- or 1 3 divided fraction of PBS-treated animals 6 days after initial transfer. In contrast, LPS injections significantly decreased the relative percentage of zero to one time divided cells and consecutively increased the percentage of two to five times divided cells, indicating direct, LPS-induced HSC proliferation (Figures 1H, 1I, and S2). A similar effect occurred also in MPPs, although less pronounced. Collectively, we conclude that highly purified HSCs can be activated from dormancy to proliferation via direct TLR4 stimulation in vivo and that, in settings of systemic LPS spread, HSC divisional activation occurs with about equal strength through direct TLR4 activation on HSCs and through indirect pathways. LPS Challenge Impairs Competitive Repopulating and Self-Renewing Capacity of HSCs through Direct TLR4 Activation Previous studies have shown that chronic exposure of HSCs to LPS damages HSC function as defined by decreased competitive transplantation and hematopoietic reconstitution (Chen et al., 2010; Esplin et al., 2011; Liu et al., 2015). However, it remained unclear whether LPS-induced HSC dysfunction is mediated by TLR4 expression on non-hematopoietic or hematopoietic cells, or both. To dissect this, we employed Tlr4/ and WT mixed BM chimeric mice (Figure 2A). Three months after transplantation, chimeric mice were i.p. injected with PBS or LPS (35 mg) every other day for eight times over 2 weeks, respectively. The relative chimerism of donor-type cells (ratio of Tlr4/to WT-derived cells) in three major blood lineages (CD19+ B cells, CD3ε+ T cells and Mac-1+Gr-1+ granulocytes) in peripheral blood (PB) was tested monthly for up to 5 months after the last injection. Chimerism of Tlr4/- and WT-derived cells was equal in all the lineages before the injections, resulting in a ratio of 1 and remained unchanged over time upon PBS injections as expected (Figures 2B and 2C). In contrast, a significant increase in the relative chimerism of Tlr4/ cells over WT cells was detected in LPS injected animals, indicating that WT cells were outcompeted by Tlr4/ cells. Terminal BM analysis revealed that, similar as in mature blood cells, HSPC populations (HSC [LKS CD34–], MPP [LKS CD34+], CMP/GMP [LK CD34+], and megakaryocyte erythroid progenitor [MEP] [LK CD34–]) in BM were dominated by Tlr4/ cells (Figure 2D), indicating that hematopoietic dominance of Tlr4/ mature cells was due to higher proliferative activity of most upstream but not downstream cell populations. As both WT and Tlr4/ hematopoiesis coexisted in the same mice and thus were exposed to the same environment, these data demonstrate that LPS activates TLR4 directly on WT HSPCs and impairs their competitive repopulating ability, leading to dominance of Tlr4/ hematopoiesis. To test whether the LPS-induced competitive repopulating disadvantage of WT cells resulted from a functional defect in self-renewing HSC, we transplanted 1–2 3 106 total BM cells of primary recipients into lethally irradiated secondary recipients. Strikingly, we found that the same effect was sustained in PB and BM of secondary recipients over 4 months, the longest time of 228 Cell Stem Cell 21, 225–240, August 3, 2017

follow-up (Figures 2E and 2F). Thus, the LPS-induced deficiency indeed affected HSCs. To further evaluate whether direct TLR4 activation on hematopoietic cells is sufficient for the observed findings, i.e., to test whether TLR4 expression on non-hematopoietic cells is dispensable in this process, we generated WT;Tlr4/ BM chimeric mice on a Tlr4/ background and subsequently challenged these with PBS or LPS. The competitive disadvantage of WT cells compared to Tlr4/ cells upon in vivo LPS treatment was even more pronounced in Tlr4/ compared to WT recipients (Figures 2G and 2H). This further documents that direct TLR4 ligation on HSCs leads to their subsequent relative competitive deficiency. As infection or LPS challenge mobilizes HSPC into the circulation, in part via upregulation of endothelial cell-produced granulocyte colony-stimulating factor (G-CSF) (Boettcher et al., 2014; Burberry et al., 2014), the observed LPS-TLR4-induced competitive HSC dysfunction might be due to differential mobilization and homing of WT versus Tlr4/ HSCs, resulting in the outcompetition of WT HSCs in the BM. However, we were not able to detect differential mobilization of WT versus Tlr4/ HSPCs in chimeric animals (data not shown). If a competitive re-homing disadvantage after LPS challenge would account for the observed out-competition, HSPC potential should be conserved in a non-competitive setting in WT cells. To test this, we injected PBS or LPS into WT mice and isolated 3,000 LKS from BM 4 weeks after the last treatment when LPS-mobilized HSCs had migrated back to BM, as evidenced by a normal HSPC profile in BM and spleen (data not shown). The isolated LKS cells (CD45.2+) were then co-transplanted with an equal number of 3,000 competitor LKS cells (CD45.1+) to lethally irradiated WT mice (CD45.1/2+) (Figure S3A). Subsequent monthly PB and terminal BM analysis showed that WT LKS isolated from LPS-pretreated mice harbored significantly reduced competitive repopulation capacity compared to WT LKS from PBS-pretreated mice (Figures S3B and S3C), demonstrating that LPS challenge damages HSC function, irrespective of possible migration effects in chimeric mice. Taken together, these data directly demonstrate that in vivo repetitive TLR4 activation on HSCs impinges on their competitive self-renewing and repopulating ability. LPS-Induced Functional HSC Impairment Is Dependent on TRIF-Mediated, but Not Myd88-Mediated, Signaling In mature innate immune cells such as macrophages or dendritic cells, TLR4 ligation activates two distinct signaling cascades, mediated by two adaptor proteins carrying TIR (Toll-interleukin receptor) domains, MYD88 (myeloid differentiation primary response 88) and TRIF (also called TICAM-1, toll-like receptor adaptor molecule 1), that both lead, among other effects, to NF-kB activation at an early and later phase, respectively (O’Neill and Bowie, 2007). However, it is not known which TLR4 downstream signaling cascades are active in HSCs. We therefore investigated whether the two adaptor molecules MYD88 and TRIF are involved in the LPS-induced HSC dysfunction. As acute inflammation was shown to enhance Sca-1 expression on HSPC (Essers et al., 2009), we tested whether LPS challenge also lead to upregulated Sca-1 expression and whether genetic deletion of either Myd88 or Trif might abrogate

B

A

20 wks post injections

Before

47.3 47.5

PB analysis (fig.2C,2G)

8x PBS/ LPS

BM analysis (fig.2D,2H) & transplantation

1st Tx -14

WT or Tlr4-/-

-2

0

PBS

PB analysis (fig.2E)

4 8 12 16 20 wks IR

BM analysis (fig.2F)

52.3

52.3 25 48

2nd Tx 4 12 16

WT

33 wks

LPS

CD45.1 (WT)

IR

WT : Tlr4-/- (1 : 1) BM chimera

74.6

52

CD45.2 (Tlr4-/-)

D

PB analysis of 1st WT recipients post LPS CD45+

Ratio: Tlr4-/- / WT

3

***

2

2

**

1

0 -2

4

8

12

16

20

2

***

1

0

0 -2

4

8

12

16

20

10

**

* 2

**

BM analysis of 1st transplant 18-36 wks post LPS injections

3

***

**

**

1

3

3

LPS PBS

Mac-1+Gr-1+

CD3+

CD19+

Ratio: Tlr4-/- / WT

C

1

8 6

4

8

12

16

20

-2

4

8

12

16

ns ns

**

2 0

20

HSC

MPP

Weeks post injections

15

15

LPS PBS

10

5

0

15

10

* * 4

8

10

*

*

5

12

16

0

*

5

* 8

10

50

*

*

*

*

4

*

15

12

16

0

4

8

12

5

0

16

4

8

12

40

PBS LPS

20

**

10 0

LPS PBS

8 6 4

*

*

10

10

8

8

8

6

6

6

4

4

2

2

4

*

2

2 0

10

HSC

MPP

-2

4

8

12

16

0

-2

4

8

12

**

16

0

-2

4

8

12

16

0

*

20

**

4

8

12

16

PBS LPS

ns

15 10

MEP

**

ns

*

5 0

-2

Weeks post injections

CMP/ GMP

BM analysis of 1st transplant 20 wks post LPS injections Ratio: Tlr4-/- / WT

10

Ratio: Tlr4-/- / WT

H Mac-1+Gr-1+

CD3+

CD19+

*

ns

30

16

PB analysis of 1st Tlr4-/- recipients post LPS CD45+

MEP

*

Weeks post secondary transplantation

G

CMP/ GMP

BM analysis of 2nd transplant 33 wks after transplantation

Mac-1+Gr-1+

CD3+

CD19+

Ratio: Tlr4-/- / WT

CD45+

Ratio: Tlr4-/- / WT

F

PB analysis post 2nd Tx

E

*

4

0 -2

PBS LPS

HSC

MPP

CMP/ GMP

MEP

Figure 2. TLR4 Activation Impairs Competitive Self-Renewing and Repopulating Capacity of HSCs (A) Experimental scheme of LPS challenge in mixed WT;Tlr4/ BM chimeric animals for results depicted in (B)–(H). Lethally irradiated WT or Tlr4/ mice were reconstituted with WT (CD45.1+) and Tlr4/ (CD45.2+) cells. Twelve weeks after reconstitution, mice were i.p. treated with PBS or LPS (35 mg) for eight times every other day, followed by monthly PB analysis of hematopoietic chimerism, starting at 1 month after the last injection. Twenty to 36 weeks post-transplantation, total BM cells from primary recipients were serially transplanted into lethally irradiated secondary WT recipients (CD45.1/2+). Secondary recipients were assessed monthly for PB chimerism, and BM was analyzed at termination of the experiment. (B) Representative flow-cytometric profiles of WT (CD45.1+) and Tlr4/ (CD45.2+)-derived granulocytes (Mac-1+Gr-1+) in WT;Tlr4/ BM chimeric WT mice before (left) and 20 weeks after (right) PBS (top) or LPS (bottom) treatment. (C) Ratio of Tlr4/- to WT-derived CD45+ cells, CD19+ B cells, CD3+ T cells, and Mac-1+Gr-1+ granulocytes in PB of primary WT recipients (PBS: n = 6–14; LPS: n = 6–13). (D) Ratio of Tlr4/- to WT-derived HSCs (LKS CD34–), MPPs (LKS CD34+), CMPs+GMPs (LK CD34+), and MEPs (LK CD34+) in BM of primary WT recipients 18–36 weeks post-treatment (PBS: n = 11; LPS: n = 15). (E) Ratio of Tlr4/- to WT-derived CD45+ cells, CD19+ B cells, CD3+ T cells, and Mac-1+Gr-1+ granulocytes in PB of secondary WT recipients (PBS: n = 8; LPS: n = 9). (F) Ratio of Tlr4/- to WT-derived HSCs, MPPs, CMPs+GMPs, and MEPs in BM of secondary WT recipients 33 weeks post-treatment (PBS: n = 5; LPS: n = 5). (G) Ratio of Tlr4/- to WT-derived CD45+ cells, CD19+ B cells, CD3+ T cells, and Mac-1+Gr-1+ granulocytes in PB of primary Tlr4/ recipients (PBS: n = 8–12; LPS: n = 10–13). (H) Ratio of Tlr4/- to WT-derived HSCs, MPPs, CMPs+GMPs, and MEPs in BM of primary Tlr4/ recipients 20 weeks post-treatment (PBS: n = 5; LPS: n = 8). Data are pooled from two to five independent experiments. ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed t test). Error bars represent SEM. See also Figure S3.

Cell Stem Cell 21, 225–240, August 3, 2017 229

A

C PBS/ LPS

3x105

16h Number of LKS

0

WT or KO

BM analysis 16 hours post injection

BM analysis

B WT

Tlr4-/1.2 10.8 8 1.17 7

9.95 5

Myd88-/1.5 8.81 1 1.54 4

Trif-/-

1.0 12.8 12 0.961 1

*** ns

2x105

**

ns

1x105

2.5 0

4.3 13.5 5 4.34 4

1.99 9

1.1 1.95 5 1.13 3

3.9 3.9

Number of LKS CD34-CD48-CD41-CD150+

PBS

2.6

c-Kit

LPS

Sca-1

D

WT

*

2x104

ns

1x104

ns 0

WT

Tlr4-/-

Myd88-/-

Trif-/-

PB analysis

8x PBS/ LPS -2

0

BM analysis of 1st transplant 18-20 wks post injections

4 8 12 16 wks 15

*

6

*

4

*

*

*

PBS-Trif-/LPS-Trif

-/-

Ratio: KO/ WT

PB analysis of 1st WT recipients post LPS Ratio: KO/ WT in Mac-1+Gr-1+

E

-14

***

F

WT : KO (1 : 1) BM chimera

IR

3x104

PBS-Trif-/-

ns

10

LPS-Trif-/-

* 5

ns

PBS-Myd88-/-

*

ns ns

LPS-Myd88-/-

ns

PBS-Myd88-/LPS-Myd88-/-

2

0

HSC

MPP

CMP/ GMP

MEP

0 -2

4

8

12

16

Weeks post injections

Figure 3. LPS-Induced Expansion and Dysfunction of HSCs Is Mediated by TLR4-TRIF Signaling (A) Experimental scheme of acute LPS challenge for results depicted in (B) and (C). WT, Tlr4/, Myd88/, and Trif/ mice were injected with PBS or LPS (35 mg) once, and BM was analyzed 16 hr post-injection for phenotypic HSCs. (B) Representative flow-cytometric profiles of LKS in BM of WT, Tlr4/, Myd88/, and Trif/ mice treated with PBS or LPS (upper and lower panel, respectively). (C) Number of LKS or phenotypic LKS CD34–CD48–CD41–CD150+ cells in BM calculated from three independent experiments (PBS: n = 3–7; LPS: n = 3–6). (D) Experimental scheme of LPS challenge to BM chimeric mice for results depicted in (E) and (F). Lethally irradiated WT (CD45.1/2+) mice were reconstituted with WT (CD45.1+) and Myd88/ (CD45.2+) or WT and Trif/ (CD45.2+) cells in a 1:1 ratio and 3 months later treated with PBS or LPS (35 mg) for eight times over 2 weeks. (E) Monthly analysis of Trif/- or Myd88/- to WT-derived granulocyte (Mac-1+Gr-1+) ratio in PB of mixed BM chimeric mice after the last PBS or LPS injection (Trif/-PBS: n = 4–11; Trif/-LPS: n = 8–13; Myd88/-PBS: n = 9–10; Myd88/-LPS: n = 10). Data are pooled from two to five independent experiments. ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed t test). Error bars represent SEM.

it. A single LPS injection (35 mg) into WT mice induced strong Sca-1 upregulation in BM Lin– cells 16 hr post-treatment, resulting in an increase in the absolute number of phenotypic LKS and LKS CD34–CD48–CD41–CD150+ cells, of which a substantial fraction are HSCs (Figures 3A–3C). Surprisingly, Trif/ mice, but not Myd88/ mice, failed to increase Sca-1 expression on HSPCs, similar as observed in Tlr4/ mice, thus indicating downstream TLR4 signaling via TRIF but not MyD88 in HSPCs. 230 Cell Stem Cell 21, 225–240, August 3, 2017

Consistent with this phenotypic observation, WT mice, reconstituted with WT and Myd88/ or Trif/ cells resulting in hematopoietic chimeras, and subsequently treated with PBS or LPS (Figure 3D), showed a competitive repopulation advantage for Trif/ but not Myd88/ cells over WT cells (Figures 3E and 3F). These data demonstrate that the functional defect of HSCs following direct TLR4 activation is mediated via TRIFbut not MYD88-signaling pathways.

Pharmacological Inhibition of ROS and p38 Activation Prevents LPS-TLR4-TRIF-Induced HSC Dysfunction To further characterize the downstream signaling cascade responsible for LPS-TLR4-TRIF-mediated HSC alteration, we tested selective small inhibitors of TLR4 downstream signaling (Figure S4A). WT mice were treated with the respective inhibitor 24 and 5 hr before LPS injection (Figure S4B). BM analysis 16 hr post-LPS injection showed that the phenotypic expansion of HSCs (LKS CD34–CD150+) upon LPS treatment was significantly suppressed by pretreatment with inhibitors against reactive oxygen species (ROS), NAC (N-acetyl-L-cysteine), and against p38 MAPK (mitogen-activated protein kinase), SB203580, or SB202190 (thereafter called SB) (Ito et al., 2006). This suppression was not achieved by inhibitors of ERK1/2 (MAPK 1 and 2) and against NF-kB (nuclear factor kB light polypeptide gene enhancer in B cells), NaS (sodium salicylate) (Kopp and Ghosh, 1994), or NBD peptide (a cell permeable NEMO-binding domain peptide) (Jimi et al., 2004) (Figures S4C and S4D). The most significant suppression of phenotypic HSC expansion was achieved when mice were treated with both ROS and p38 inhibitors (NAC/SB). Consistent with findings and observations made in innate immune cells (Matsuzawa et al., 2005), flow-cytometric analysis confirmed intracellular ROS production and p38 activation as early as 2 hr after in vivo LPS application in LKS CD34– and LKS CD34+ cells but not in CMP/GMP (LK CD34+) (Figure S4E). Furthermore, in vitro stimulation of HSCs with LPS revealed that ERK1/2 and p38 but not NF-kB were directly activated in WT but not Tlr4/ HSCs, while Trif/ cells showed reduced activation of ERK1/2 and p38, which was more pronounced compared to Myd88/ cells (Figures S4F and S4G), supporting direct TLR4 and downstream TRIF signaling in LPS-stimulated HSCs. Next, we addressed whether ROS/p38 inhibitors can block LPS-induced proliferation of HSCs. WT;Tlr4/ BM chimeric Tlr4/ mice were treated with BrdU for 2 weeks and subsequently with NAC and SB at 24 and 5 hr prior to PBS or LPS injection for four cycles (Figure 4A). BM analysis 1 week after the final injection revealed that combined inhibitor treatment abolished the LPS-induced increase in proliferation of both WT and Tlr4/ HSCs (Figure 4B). To test whether inhibition of the ROS-p38-mediated signaling pathway can prevent the observed LPS-induced competitive disadvantage of HSCs, WT;Tlr4/ BM hematopoietic chimeric mice were treated with eight cycles of sequential inhibitor and LPS applications, and the relative contribution of Tlr4/ and WT cells to hematopoietic chimerism in PB and BM was determined (Figure 4C). Strikingly, combined NAC/SB treatment prior to LPS completely abolished the competitive disadvantage of WT hematopoiesis in PB (Figure 4D). The same effect was observed in frequencies of HSCs, LKS, and LK in BM upon termination of the experiment (Figure 4E). Together these data demonstrate that activation of ROS and p38 through TLR4 ligation is a key step to cause HSC dysfunction and that this dysfunction can be prevented by respective pharmacological blockade. Emergency Granulopoiesis Depends on TLR4-MYD88, but Not TRIF-ROS-p38, Signaling We recently demonstrated that systemic LPS- or bacterial-infection-driven granulopoiesis, i.e., emergency granulopoiesis, is

mediated by TLR4-MYD88 signaling in endothelial cells and subsequent G-CSF release (Boettcher et al., 2014, 2012; Manz and Boettcher, 2014). Given our findings on TLR4-TRIF but not TLR4-MyD88-mediated signaling in HSCs, we determined whether TRIF deficiency would affect LPS-induced emergency granulopoiesis. As expected and previously shown, in vivo LPS challenge of Tlr4/ and Myd88/ mice resulted in absence of enhanced granulopoiesis, whereas Trif/ mice showed an intact emergency granulopoietic response (Figures S5A–S5C). Moreover, none of the inhibitors tested had a suppressive effect on LPS-induced emergency granulopoiesis (Figures S5D and S5E). Together these data demonstrate that immediate emergency granulopoiesis and HSC activation via TLR4 are differentially regulated by Myd88- and TRIF-dependent signals, respectively. TLR4 Stimulation Induces Proliferative Stress in HSCs via TLR4-TRIF-ROS-p38 Signaling Increased levels of ROS are associated with cell-cycle induction (Walter et al., 2015) and attenuated HSC function during serial transplantation (Ito et al., 2006). Also, ROS can cause DNA breaks and chromosomal aberrations in BM cells in a Fanconi anemia mouse model (Zhang et al., 2007). We thus tested whether TLR4-mediated ROS production leads to genotoxic stress in HSCs. Mice were treated with PBS or LPS for eight times, and HSCs (LKS CD34–CD150+ or LKS CD34–CD48–CD41–CD150+) and myeloid progenitors (LK or LK CD34–CD48–CD41–) in BM were analyzed at different time points after the last injection (Figure 5A). Cells were stained with antibodies against phosphorylated H2AX (gH2AX), a surrogate marker for DNA damage response (Blanpain et al., 2011; Ciccia and Elledge, 2010), and gH2AX+ foci formation was visualized and quantified. The proportion of cells with more than three gH2AX+ foci was significantly increased in both HSCs and their progeny upon LPS treatment and was comparable to that observed in mice that were gamma-irradiated (2 Gy) 2 hr before analysis (Figures 5B and 5C). Surprisingly, LPS-induced gH2AX+ foci were maintained for longer than 14 days in vivo, indicating sustained genotoxic stress upon LPS challenge. As it was recently shown that gH2AX+ foci formation results from replication defects in cycling HSCs rather than from DNA double-strand breaks (Flach et al., 2014), we analyzed whether gH2AX+ foci correlate with cell-cycle induction in HSCs. Snapshot cell-cycle analysis showed that a higher fraction of HSCs (LKS CD34–CD150+) were found in G1 at day 3 post-LPS injection but regained normal cell-cycle distribution at day 14 (Figures 5D and 5E), while there was no significant difference in actively cycling myeloid progenitors (LK), consistent with BrdU labeling and dilution experiments shown in Figure 1B. The apoptotic rate was similar in PBS- and LPS-stimulated HSCs at any time point analyzed (data not shown). To further dissect whether LPS induced gH2AX+ foci formation indicate double-strand DNA breaks and to test which downstream signal via TLR4 is critical for foci formation, HSCs from WT, Tlr4–/–, Myd88–/–, and Trif/ mice were analyzed upon LPS application using both gH2AX and 53BP labeling (Figures 5F–5I). LPS induced formation of both gH2AX+ and 53BP1+ foci in WT and Myd88/, but not Tlr4/ and Trif/ cells, whereas sublethal irradiation Cell Stem Cell 21, 225–240, August 3, 2017 231

A WT : Tlr4-/- (1 : 1) BM chimera

IR

Tlr4-/-

PBS/ NAC

DMSO/ SB

PBS/ LPS

-24

-5

0h

BrdU

Analysis of Brd retention in BM HSC

4x cycles

BM analysis

0

-8

B

2

4 wks

chase

ns

**

ns

ns

ns

*

ns

*

80

BrdU+ HSCs (%)

BrdU uptake

3

ns 100

***

ns

ns

PBS + PBS/DMSO PBS + NAC/SB LPS + PBS/DMSO LPS + NAC/SB

60

40

20

C

PBS/ NAC -24

DMSO/ SB

PBS/ LPS

-5

0h

WT : Tlr4-/- (1 : 1) BM chimera

Donor

WT Tlr4

-/-

4 8 12 16 20 wks

*

5

PBS + PBS/DMSO PBS + NAC/SB LPS + PBS/DMSO LPS + NAC/SB

4 3 2

0

* ** **

*

1

-2

4

8

12

16

** ** * Ratio: Tlr4-/- / WT

0

WT Tlr4

-/-

WT Tlr4

-/-

BM analysis 22 wks post 1st Tx

BM analysis

PB analysis of WT;Tlr4-/- BM chimera Ratio: Tlr4-/-/ WT in Mac-1+Gr-1+

D

-2

-/-

E 6

WT or Tlr4-/- -14

WT Tlr4

8x cycles

PB analysis

IR

0

*

*

** * *

4

PBS + PBS/DMSO PBS + NAC/SB LPS + PBS/DMSO LPS + NAC/SB

2

0

HSC

LKS

LK

20

Weeks post injections

Figure 4. Pharmacological Inhibition of ROS-p38 Signaling Prevents HSC Proliferation and Exhaustion upon LPS Challenge (A) Experimental scheme of BrdU label retention with or without NAC and SB for results depicted in (H). 8 weeks after reconstitution of Tlr4/ mice with WT or Tlr4/ cells in a 1:1 ratio, mice were given BrdU-containing water for 2 weeks and treated with four cycles of NAC (24h) and SB (5h) (NAC/SB) prior to each four PBS or LPS (50 mg) injections, followed by BM analysis 1 week after the last injection. (B) Percentage of BrdU+ HSCs (LKS CD34–CD150+) derived from WT or Tlr4/ donors upon PBS or LPS injections with or without NAC and SB (PBS+DMSO: n = 9; PBS+NAC/SB: n = 11; LPS+DMSO: n = 9; LPS+NAC/SB: n = 11). (C) Experimental scheme of PBS or LPS challenge to mixed WT;Tlr4/ BM chimeric mice, with or without NAC and SB for results depicted in (D) and (E). 12 weeks after transplantation of WT or Tlr4/ mice with WT and Tlr4/ cells in a 1:1 ratio, mice were treated with eight cycles of NAC (24h) and SB (5h) prior to PBS or LPS (35 mg) injections followed by monthly PB chimerism analysis and terminal BM analysis. (D) Ratio of Tlr4/- to WT-derived granulocytes (Mac-1+Gr-1+) in PB of the mixed BM chimeras are shown (PBS+DMSO: n = 4; PBS+NAC/SB: n = 4; LPS+DMSO: n = 3; LPS+NAC/SB: n = 4). The ratio from PBS+DMSO, PBS+NAC/SB, and LPS+NAC/SB-treated mice was statistically compared with that from LPS+DMSO-treated mice. (E) Ratio of Tlr4/- to WT-derived HSCs (LKS CD34–), MPPs (LKS CD34+), CMPs+GMPs (LK CD34+), and MEPs (LK CD34+) in BM of mixed BM chimeric mice at terminal analysis 22 weeks post-treatment (PBS+DMSO: n = 4; PBS+NAC/SB: n = 4; LPS+DMSO: n = 3; LPS+NAC/SB: n = 4). Data are pooled from two to five independent experiments. ns, not significant; *p < 0.05, **p < 0.01 (two-tailed t test). Error bars represent SEM. See also Figure S4.

induced gH2AX and 53BP1 double-positive foci, a sign of DNA double-strand breaks. Similar results were obtained in myeloid progenitors (LK CD34–CD48–CD41–) (Figure S6). As only few foci were co-marked with gH2AX and 53BP1 upon LPS treatment, these data suggest that LPS stimulation leads, via TLR4TRIF signaling, to proliferative stress rather than double-strand DNA damage. We then tested whether ROS/p38 inhibition can block gH2AX+ foci formation. When WT mice were treated with NAC and SB prior to LPS injections for eight cycles (Figure 5J), the percentage of HSCs and their progeny carrying gH2AX+ foci at day 3 post-treatment was suppressed significantly but 232 Cell Stem Cell 21, 225–240, August 3, 2017

not completely (Figure 5K), indicating that foci formation is mediated in part by the ROS-p38 pathway. Taken together, the increase of ROS levels and p38 activation upon LPS challenge leads to proliferation and causes sustained proliferative stress in HSCs that can be partially blocked by ROSp38 inactivation. S. Typhimurium Infection Leads to TLR4-TRIF-Mediated HSC Activation To evaluate whether the proliferative stress signal in HSCs also occurs during gram-negative bacterial infection in vivo, we

C

A

B

D

F

E

G

H

J

I

K

(legend on next page)

Cell Stem Cell 21, 225–240, August 3, 2017 233

infected mice systemically with Salmonella Typhimurium (Watson and Holden, 2010). WT, Tlr4/, Myd88/, and Trif/ mice were infected with 1 3 104 colony-forming units (CFUs) of S. enterica serovar Typhimurium (S. Tm) SB300, and, 2 days later, when S. Tm was detectable in both spleen and BM (data not shown), BM was analyzed (Figure 6A). Unlike the case of LPS stimulation alone, S. Tm infection led to strong upregulation of Sca-1 expression in all mice tested, indicating the possible involvement of other signals in this process, such as interferon (IFN) (Baldridge et al., 2010; Essers et al., 2009) (Figure 6B). Of note, the percentage of phenotypically defined HSCs that were sorted for subsequent gH2AX+ and 53BP1+ foci formation appeared to be unchanged after the infection. Interestingly, S. Tm infection induced as many gH2AX+ and 53BP1+ foci in WT and Myd88/ HSCs as observed with LPS application alone, while much less foci were observed in Tlr4/- and Trif/mouse-derived HSCs (Figures 6C and 6D). These findings indicate that systemically spread infections with gram-negative bacteria induce HSC activation and proliferative stress as indicated by gH2AX+ and 53BP1+ foci formation mainly via TLR4TRIF signaling.

alterations were observed, as expected, in Tlr4/ HSC. In contrast, both Myd88/ and Trif/ HSCs showed moderate changes in RNA expression, indicating MYD88- and TRIFdependent transcriptional activation (Figures 7C and 7D). The analysis of differential gene expression patterns among WT, Trif/, and Myd88/ HSCs revealed expression changes that specifically occurred in WT and Myd88/, but not Tlr4/ and Trif/ HSCs (38 up- and 37 downregulated genes; data not shown), indicating that TLR4-TRIF-dependent transcriptional regulation mainly involves cell-cycle, migration, and inflammatory programs (Figures 7E and 7F, and the list of genes shown in Figure S7). Independent qPCR analysis validated that some genes were up- or downregulated in LPS-stimulated WT and Myd88/ HSCs, but not in Trif/ HSCs, as compared to the PBS-stimulated control (Figure 7G). In summary, the transcriptome analysis demonstrates that direct TLR4 activation on HSCs by LPS changes a set of inflammation-related programs and subsequently leads to LPS-induced inflammatory stress responses, which need to be elucidated in future gain- and loss-offunction studies. DISCUSSION

LPS-TLR4-TRIF Signaling Induces a Distinct Transcriptional Signature in HSCs In order to start to address which genetic programs in HSCs (LKS CD34–CD48–CD41–CD150+) are altered by LPS-induced TLR4TRIF signaling, we performed global and selective gene expression analysis (Figure 7A). In agreement with the previous findings (Zhao et al., 2014), gene expression analysis showed that interleukin-6 (IL-6) and IL-1b but not tumor necrosis factor-a (TNF-a) were upregulated in WT-HSCs as quickly as at 6 hr after LPS stimulation and subsequently normalized to steady-state levels at 12 hr (Figure 7B). Thus, we performed HSC RNA sequencing at 6 hr upon LPS or PBS stimulation. While changes in RNA expression were most pronounced in WT HSCs, little

In this study, we provide for the first time definitive in vivo evidence that (1) quiescent HSCs in the BM sense systemically spread LPS directly via HSC-expressed TLR4 and subsequently increase cycling, that (2) TLR4 stimulation on HSCs impairs their competitive repopulation ability and causes proliferative stress via TRIF-mediated, but not MyD88-mediated, ROS-p38 signaling, that (3) HSCs upon LPS stimulation activate proliferation, migration, and inflammation-related gene expression programs, that (4) pharmacologic inhibition of TLR4-TRIF-ROSp38 signaling prevents LPS-induced HSC dysfunction, and that (5) systemic S. Tm infection also leads to proliferative stress via TLR4-TRIF signals.

Figure 5. LPS Induces DNA Stress Responses in HSCs via TLR4-TRIF-ROS-p38 Activation (A) Experimental scheme of sustained LPS challenge for results depicted in (B)–(E). WT mice were treated with PBS or LPS (50 mg) eight or 12 times every other day, and BM were processed for cell-cycle analysis or sorting at different time points (3, 14, R35 days) after the last injection. (B) Representative confocal microscopic images of co-localization of gH2AX+ (red) foci in nuclei (blue) of HSCs (LKS CD34–CD48–CD41–CD150+) and LK CD34–CD48–CD41– cells at day 3 after PBS or LPS treatment or at 2 hr after 2 Gy irradiation. White bars indicate 5 mm. (C) Time-course kinetics of gH2AX+ foci formation in HSCs and LK CD34–CD48–CD41– cells from PBS- (black) or LPS- (red) treated mice, in comparison with those from animals 2 hr after 2 Gy irradiation (blue). The percentage of cells with greater than three gH2AX+ foci is shown. The number of PBS or LPS injections is indicated above each bar. (D) Representative flow-cytometric cell-cycle analysis of HSCs (LKS CD34–CD150+) and LK cells at 3 and 14 days after the last injection. (E) Percentage of each cell-cycle phase (G0, G1, and S/G2/M) of HSCs and LK cells (PBS: n = 4–5; LPS: n = 4–5 at each time point). (F) Experimental scheme of sustained LPS challenge in WT and KO mice (G–I): WT or KO mice were treated with PBS or LPS (100 mg) eight times every other day. HSCs (LKS CD34–CD48–CD41–CD150+) were sorted from treated mice and assessed for foci formation of DNA damage markers at 3 and 14 after the last injection. (G) Representative confocal microscopic images of co-localization of gH2AX+ (red) and 53BP1+ (green) foci in nuclei (blue) of HSCs (LKS CD34–CD48–CD41–CD150+) at day 3 after PBS or LPS treatment or at 2 hr after 4 Gy irradiation. White bars indicate 5 mm. (H and I) Time-course kinetics of gH2AX+ (H) or 53BP1+ (I) foci formation in HSCs and LK CD34–CD48–CD41– cells from PBS- (black) or LPS- (red) treated mice, in comparison with those from animals 2 hr after 4 Gy irradiation (blue). The percentage of cells with greater than three gH2AX+ or 53BP1+ foci is shown. The numbers (n = ) indicated below each bar represent the number of cells analyzed. The time point of analysis (days) after PBS or LPS injections is indicated above each graph. (J) Experimental scheme of DNA damage response upon LPS application in the presence or absence of NAC and SB for results depicted in (K): WT mice were treated with eight cycles of NAC (24h) and SB (5h) in prior to LPS injections. (K) The percentage of cells with greater than three gH2AX+ foci in HSCs and LK CD34–CD48–CD41– cells 3 days after the last LPS injection with or without NAC/SB. Data are pooled from two to four independent experiments. n.d., not determined; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed t test). Error bars represent SEM. See also Figure S6.

234 Cell Stem Cell 21, 225–240, August 3, 2017

A

B

C

D

(legend on next page)

Cell Stem Cell 21, 225–240, August 3, 2017 235

Although not directly addressed in our study, one might speculate that LPS (and possibly other PAMP) sensing by HSCs and subsequent proliferation might enhance hematopoietic cell production in situations of need and thus has been advantageous and selected in evolution. However, by employing BM chimeric mice with one-half of hematopoietic cells expressing TLR4 and the other half not expressing it, our study clearly demonstrates that Tlr4/ HSCs are also induced to divide upon LPS injection, a process that must be driven by secondary, extrinsic signals such as pro-inflammatory cytokines secreted from TLR4-expressing non-hematopoietic (in the case of a TLR4 WT environment) or hematopoietic cells (in the case of a hematopoietic WT;Tlr4/ chimera in a TLR4/ non-hematopoietic environment). One potential cytokine involved in this process could be IL-6, which has been shown to be produced by HSPCs upon TLR4 activation (Zhao et al., 2014) and by non-hematopoi€rch etic BM cells via IFN-g triggering upon viral infections (Schu et al., 2014). As the relative contribution to HSC cycling by direct, HSC-expressed TLR4 ligation and indirect, non-HSCexpressed TLR4 ligation was of about equal strength in the experimental settings in our study, it will be challenging to dissect in which in vivo infectious disease situation the direct TLR4-mediated response will provide a distinct advantage to hematopoietic regulation and subsequent host defense and survival. The observed negative consequence of direct LPS-TLR4mediated HSC activation is a compromised competitive HSC capacity, defined by both decreased hematopoietic output in competition with control-stimulated HSCs upon transplantation, as well as in a relative loss of TLR4-stimulated HSC numbers compared to TLR4-signaling deficient HSC controls. These findings are in line with the observation that LPS application resulted in reduced lymphoid repopulation from HSCs in competitive conditions (Esplin et al., 2011; Zhang et al., 2016). By dissecting TLR4 downstream signaling in HSCs, we found HSC signal and transcriptional changes via the TLR4-associated TIR adaptor protein TRIF, but not MYD88, which is in agreement with recent findings (Zhang et al., 2016). We further investigated LPS-induced HSC dysfunction mechanistically and found that TLR4-TRIF triggers a downstream signaling cascade via ROSp38, which causes cell-cycle progression in HSC and leads to a proliferative stress that is marked by the phosphorylation of H2AX, previously reported as p38 target, but not 53BP1 (Dong et al., 2014). The same HSC stress via TLR4-TRIF can be caused by a highly complex bacteria organism such as S. Tm. We here demonstrate that both genetic and pharmacological inactivation of this pathway can block HSC proliferation, stress response, and dysfunction in response to LPS. Indeed, it has been reported that activation of the ROS-p38 axis upon serial transplantation

results in decreased quiescence and impaired repopulating ability of HSCs (Ito et al., 2006). In addition, we here demonstrate that pretreatment of HSCs with ROS-p38 inhibitors prior to LPS challenge blocks proliferation and maintained HSC function. This is in line with the findings that inhibition of ROS-p38 sustains HSC function in vivo (Liu et al., 2015; Tothova et al., 2007; Yahata et al., 2011) and also supports ex vivo expansion of both mouse and human HSCs (Baudet et al., 2012; Wang et al., 2011). On the contrary, dysregulation of ROS-p38 signaling was proposed to be a hallmark of stem cell aging (Oh et al., 2014). Moreover, it was shown that scavenging ROS restored the quiescence and regenerative potential of different somatic stem cells (Drowley et al., 2010; Ito et al., 2006; Tothova et al., 2007), and blockage of aging-associated p38 activation ameliorated the functional decline in aged muscle stem cells (Bernet et al., 2014). How ROS-p38 activation causes HSC proliferation and dysfunction, and proliferative stress associated with H2AX+ foci formation, as well as possible subsequent genetic alterations, is still controversial, with several studies demonstrating varying results (Beerman et al., 2013; Flach et al., 2014; Sinha et al., 2014). Thus, we do not know from the data presented here whether the observed enhanced cycling and proliferative stress results in functional HSC attrition (Walter et al., 2015) or if, at least in some cases, it also results in genetic lesions in HSCs. However, it is of interest that population-based epidemiological studies showed a significant positive correlation of infectious and inflammatory states and the incidence of neoplastic HSPC diseases, such as acute myeloid leukemia, myelodysplastic syndromes, and myeloproliferative neoplasms (Kristinsson et al., 2010, 2011). Furthermore, a recent study also has shown that TLR4-TRAF6 signaling enhanced by deficiency of TIFAB (TRAF-interacting protein with forkhead-associated domain B), a haplo-insufficient gene in del(5q) myelodysplastic syndrome, causes inefficient hematopoiesis (Varney et al., 2015). Thus, whether direct stimulation of PAMPs on HSPCs can indeed enhance acquisition of genetic events that lead to hematopoietic deficiency, clonal dominance, and finally malignant hematopoietic disease needs to be determined in future studies. TLR4 stimulation induces progenitor cell division and differentiation to myeloid cells via MYD88 (Massberg et al., 2007; Megı´as et al., 2012; Nagai et al., 2006), and TLR4 stimulation also induces MYD88-dependent G-CSF secretion from endothelial cells, leading to subsequent emergency granulopoiesis (Boettcher et al., 2014). In contrast, we show here that TLR4 stimulation in bona fide HSCs leads to TRIF-, but not the MYD88-dependent signaling, revealing differential use of TLR4 downstream pathways in different cells of the hematopoietic hierarchy. We thus tested whether ablation or inhibiting TRIFmediated signaling would negatively affect emergency granulopoiesis. Interestingly, this was not the case. Therefore, it might

Figure 6. DNA Stress Responses in HSCs upon S. Tm Infection (A) Experimental scheme: WT or KO mice were intravenously (i.v.) injected with 1 3 104 CFU S. Tm and 2 days later euthanized for sorting and analysis of HSCs (LKS CD34–CD48–CD41–CD150+). (B) Representative flow-cytometric profiles of HSCs and progenitors in BM of WT or KO mice with (left) or without the infection (right). (C) Representative confocal microscopic images of co-localization of gH2AX+ (red) and 53BP1+ (green) foci in nuclei (blue) of HSCs (LKS CD34–CD48–CD41–CD150+) at day 2 after the infection (PBS, black; S. Tm, red) or at 2 hr after 2 Gy irradiation (blue). White bars indicate 5 mm. (D) Quantification of gH2AX+ (left) and 53BP1+ (right) foci formation. The numbers (n =) indicated below each bar represent the number of cells analyzed. Error bars represent SEM.

236 Cell Stem Cell 21, 225–240, August 3, 2017

A

HSCs LK Granulocytes PBS/ LPS

WT or KOs

0

Cell harvest and qPCR/RNA-seq

6h/12h

B relative expression to Actb %

Il-6

Il-1b

0.8

4

0.6

3

0.4

2

0.2

1

0.0

0

Tnfa 6

HSC

LK

Granulocytes

HSC

LK

-4

Granulocytes

HSC

LK

PB S LP 6h S LP 6 S h 12 h

PB S LP 6h S LP 6 S h 12 h

PB S LP 6h S LP 6 S h 12 h

PB S LP 6h S LP 6 S h 12 h

PB S LP 6h S LP 6 S h 12 h

0

color key

C

PBS LPS

2

PB S LP 6h S LP 6 S h 12 h

PB S LP 6h S LP 6 S h 12 h

PB S LP 6h S LP 6 S h 12 h

PB S LP 6h S LP 6 S h 12 h

4

Granulocytes

D

4

≥2-fold up-regulated genes in LPS-stimulated WT

15000

≥2-fold down-regulated genes in LPS-stimulated WT

Trif-/--PBS

PC 2

10000

Trif-/--LPS

5000 Tlr4-/--LPS 0

Tlr4-/--PBS

-5000 -10000

WT-PBS

Myd88 -PBS Myd88-/--LPS -/-

-20000

0

WT-LPS 20000

40000

PC1

E

PBS LPS PBS LPS PBS LPS PBS LPS WT Tlr4-/MyD88-/Trif-/-

Down-regulated genes in LPS-stimulated WT and Myd88-/- HSCs

Up-regulated genes in LPS-stimulated WT and Myd88-/- HSCs

relative expression to Actb %

Frk

Zpb1

0.010

0.005

0.000

0.8

4

0.6

3

0.4

2

0.2

1

Tlr4-/-

Myd88-/-

Trif-/-

Cd28

Il-6

*

0.15

0.10

PBS LPS

0.05

0.0

WT

3-Way Venn diagram (≥2-fold downregulated genes)

PBS LPS PBS LPS PBS LPS PBS LPS WT Tlr4-/MyD88-/Trif-/-

G

0.015

F

3-Way Venn diagram (≥2-fold upregulated genes)

0.00

0

WT

Tlr4-/-

Myd88-/-

Trif-/-

WT

Tlr4-/-

Myd88-/-

Trif-/-

WT

Tlr4-/-

Myd88-/-

Trif-/-

Figure 7. LPS Stimulation Activates Distinct Molecular Signatures in HSCs via TRIF and MyD88 (A) Experimental scheme: HSCs (LKSCD34–CD48–CD41–CD150+), LK and granulocyte (Mac-1+Gr-1+) were isolated from BM of WT or KO mice and stimulated with PBS or LPS (100 mg/mL) in vitro. (B) After 6 and 12 hr of stimulation, cells were harvested and subjected to qPCR or RNA sequencing. (C) Heatmap of up- (left) or down- (right panel) regulated transcripts in LPS-stimulated WT HSCs as compared to PBS control. Same transcripts from PBS- or LPS-stimulated HSCs from Tlr4/, Myd88/, or Trif/ are shown in parallel. The color code represents the expression level as indicated above the heatmap. (legend continued on next page)

Cell Stem Cell 21, 225–240, August 3, 2017 237

be feasible to prevent HSCs from LPS-induced functional impairment by pharmacologic inhibition of TRIF-dependent downstream signaling, while at the same time preserving indirect HSC cycling induction as well as emergency granulopoiesis in order to augment innate immunity against systemically spread, life-threatening gram-negative infection. In summary, our findings highlight a direct pathogen sensing mechanism in self-renewing HSCs that leads to HSC proliferation and inflammatory responses but decreases their regenerative capacity. Intervention into the TRIF-ROS-p38 signaling axis has the potential to prevent HSC dysfunction, likely without inhibition of short-term protective responses against pathogen insults. Whether inhibition of this pathway as a means of preemptive medicine is able to prevent or delay aging-associated HSC defects and malignant transformation will need to be determined in future studies.

ACKNOWLEDGMENTS We would like to thank S. Akira (Department of Host Defense, Research Institute for Microbial Diseases, Osaka University) for kindly providing Tlr4/, Trif/, and Myd88/ mice. We also thank the Institute of Research in Biomedicine and University Hospital Zurich for the logistical and technical assistance, and the flow cytometry and microscopy core facility at University of Zurich. This work was supported by KAKENHI from the Japanese Society of the Promotion of Science (15H01519), the Kanae Foundation for the Promotion of Medical Science, and the SENSHIN Medical Research Foundation to H.T.; the Swiss National Science Foundation to C.N.-A. (31003A_159597/1); the Swiss National Science Foundation (310030B_166673/1), the Promedica Foundation (Chur, Switzerland), and the Clinical Research Priority Program Human Hemato-Lymphatic Diseases of the University of Zurich to M.G.M. Received: March 20, 2016 Revised: April 22, 2017 Accepted: June 19, 2017 Published: July 20, 2017

STAR+METHODS

REFERENCES

Detailed methods are provided in the online version of this paper and include the following:

Baldridge, M.T., King, K.Y., Boles, N.C., Weksberg, D.C., and Goodell, M.A. (2010). Quiescent haematopoietic stem cells are activated by IFN-g in response to chronic infection. Nature 465, 793–797.

KEY RESOURCES TABLE CONTACT FOR REAGENT AND RESOURCE SHARING EXPERIMENTAL MODEL AND SUBJECT DETAILS B Mice B Generation of BM chimeric mice, LPS challenge and serial transplantation B Salmonella Tm infection METHOD DETAILS B FACS analysis and sorting B BrdU labeling and retention B CFSE labeling and chasing B Kaplan-Meier survival following 5FU administration B In vivo inhibitor treatment B Detection of ROS, NF-kB, Erk1/2 and p38 activation B Immunocytochemistry B Quantitative RT-PCR B RNA sequencing analysis QUANTIFICATION AND STATISTICAL ANALYSIS DATA AND SOFTWARE AVAILABILITY

Baudet, A., Karlsson, C., Safaee Talkhoncheh, M., Galeev, R., Magnusson, M., and Larsson, J. (2012). RNAi screen identifies MAPK14 as a druggable suppressor of human hematopoietic stem cell expansion. Blood 119, 6255–6258.

d d d

d

d d

SUPPLEMENTAL INFORMATION

Beerman, I., Bock, C., Garrison, B.S., Smith, Z.D., Gu, H., Meissner, A., and Rossi, D.J. (2013). Proliferation-dependent alterations of the DNA methylation landscape underlie hematopoietic stem cell aging. Cell Stem Cell 12, 413–425. Bernet, J.D., Doles, J.D., Hall, J.K., Kelly Tanaka, K., Carter, T.A., and Olwin, B.B. (2014). p38 MAPK signaling underlies a cell-autonomous loss of stem cell self-renewal in skeletal muscle of aged mice. Nat. Med. 20, 265–271. Blanpain, C., Mohrin, M., Sotiropoulou, P.A., and Passegue´, E. (2011). DNAdamage response in tissue-specific and cancer stem cells. Cell Stem Cell 8, 16–29. Boettcher, S., Ziegler, P., Schmid, M.A., Takizawa, H., van Rooijen, N., Kopf, M., Heikenwalder, M., and Manz, M.G. (2012). Cutting edge: LPS-induced emergency myelopoiesis depends on TLR4-expressing nonhematopoietic cells. J. Immunol. 188, 5824–5828. Boettcher, S., Gerosa, R.C., Radpour, R., Bauer, J., Ampenberger, F., Heikenwalder, M., Kopf, M., and Manz, M.G. (2014). Endothelial cells translate pathogen signals into G-CSF-driven emergency granulopoiesis. Blood 124, 1393–1403. Burberry, A., Zeng, M.Y., Ding, L., Wicks, I., Inohara, N., Morrison, S.J., and Nu´n˜ez, G. (2014). Infection mobilizes hematopoietic stem cells through cooperative NOD-like receptor and Toll-like receptor signaling. Cell Host Microbe 15, 779–791.

Supplemental Information includes seven figures and can be found with this article online at http://dx.doi.org/10.1016/j.stem.2017.06.013.

Busch, K., Klapproth, K., Barile, M., Flossdorf, M., Holland-Letz, T., Schlenner, S.M., Reth, M., Ho¨fer, T., and Rodewald, H.-R. (2015). Fundamental properties of unperturbed haematopoiesis from stem cells in vivo. Nature 518, 542–546.

AUTHOR CONTRIBUTIONS

Chen, C., Liu, Y., Liu, Y., and Zheng, P. (2010). Mammalian target of rapamycin activation underlies HSC defects in autoimmune disease and inflammation in mice. J. Clin. Invest. 120, 4091–4101.

H.T. designed and performed experiments, analyzed data, and wrote the manuscript. K.F., L.V.K., Y.S., C.Y., K.J., A.K.A., and A.H. performed experiments. M.L., W.-D.H., A.G., and C.N.-A. designed and analyzed data. M.G.M. designed and supervised research and wrote the manuscript.

Cheshier, S.H., Morrison, S.J., Liao, X., and Weissman, I.L. (1999). In vivo proliferation and cell cycle kinetics of long-term self-renewing hematopoietic stem cells. Proc. Natl. Acad. Sci. USA 96, 3120–3125.

(D) Principal component analysis of the expression of a total of 11,678 genes in WT or KO HSCs, stimulated with PBS or LPS (the first component accounts for 81.57% of the total variance and the second for 8.47%). (E and F) Differential expression of LPS-induced genes (Venn diagram). Up- (E) or down- (F) regulated genes in WT, Myd88/, or Trif/ HSCs. (G) qPCR analysis on up- or downregulated genes that are common in LPS-stimulated WT and Myd88/ HSCs. *p < 0.05 (two-tailed t test). Error bars represent SEM. See also Figure S7.

238 Cell Stem Cell 21, 225–240, August 3, 2017

Ciccia, A., and Elledge, S.J. (2010). The DNA damage response: Making it safe to play with knives. Mol. Cell 40, 179–204.

U.H. (2007). Immunosurveillance by hematopoietic progenitor cells trafficking through blood, lymph, and peripheral tissues. Cell 131, 994–1008.

Diard, M., Garcia, V., Maier, L., Remus-Emsermann, M.N., Regoes, R.R., Ackermann, M., and Hardt, W.D. (2013). Stabilization of cooperative virulence by the expression of an avirulent phenotype. Nature 494, 353–356.

Matsuzawa, A., Saegusa, K., Noguchi, T., Sadamitsu, C., Nishitoh, H., Nagai, S., Koyasu, S., Matsumoto, K., Takeda, K., and Ichijo, H. (2005). ROS-dependent activation of the TRAF6-ASK1-p38 pathway is selectively required for TLR4-mediated innate immunity. Nat. Immunol. 6, 587–592.

Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M., and Gingeras, T.R. (2013). STAR: Ultrafast universal RNAseq aligner. Bioinformatics 29, 15–21. Dong, Y., Xiong, M., Duan, L., Liu, Z., Niu, T., Luo, Y., Wu, X., Xu, C., and Lu, C. (2014). H2AX phosphorylation regulated by p38 is involved in Bim expression and apoptosis in chronic myelogenous leukemia cells induced by imatinib. Apoptosis 19, 1281–1292. Drowley, L., Okada, M., Beckman, S., Vella, J., Keller, B., Tobita, K., and Huard, J. (2010). Cellular antioxidant levels influence muscle stem cell therapy. Mol. Ther. 18, 1865–1873. Esplin, B.L., Shimazu, T., Welner, R.S., Garrett, K.P., Nie, L., Zhang, Q., Humphrey, M.B., Yang, Q., Borghesi, L.A., and Kincade, P.W. (2011). Chronic exposure to a TLR ligand injures hematopoietic stem cells. J. Immunol. 186, 5367–5375. Essers, M.A.G., Offner, S., Blanco-Bose, W.E., Waibler, Z., Kalinke, U., Duchosal, M.A., and Trumpp, A. (2009). IFNalpha activates dormant haematopoietic stem cells in vivo. Nature 458, 904–908. Flach, J., Bakker, S.T., Mohrin, M., Conroy, P.C., Pietras, E.M., Reynaud, D., Alvarez, S., Diolaiti, M.E., Ugarte, F., Forsberg, E.C., et al. (2014). Replication stress is a potent driver of functional decline in ageing haematopoietic stem cells. Nature 512, 198–202. Hatakeyama, M., Opitz, L., Russo, G., Qi, W., Schlapbach, R., and Rehrauer, H. (2016). SUSHI: An exquisite recipe for fully documented, reproducible and reusable NGS data analysis. BMC Bioinformatics 17, 228. Ito, K., Hirao, A., Arai, F., Takubo, K., Matsuoka, S., Miyamoto, K., Ohmura, M., Naka, K., Hosokawa, K., Ikeda, Y., and Suda, T. (2006). Reactive oxygen species act through p38 MAPK to limit the lifespan of hematopoietic stem cells. Nat. Med. 12, 446–451. Jimi, E., Aoki, K., Saito, H., D’Acquisto, F., May, M.J., Nakamura, I., Sudo, T., Kojima, T., Okamoto, F., Fukushima, H., et al. (2004). Selective inhibition of NF-k B blocks osteoclastogenesis and prevents inflammatory bone destruction in vivo. Nat. Med. 10, 617–624. Kondo, M., Wagers, A.J., Manz, M.G., Prohaska, S.S., Scherer, D.C., Beilhack, G.F., Shizuru, J.A., and Weissman, I.L. (2003). Biology of hematopoietic stem cells and progenitors: Implications for clinical application. Annu. Rev. Immunol. 21, 759–806. Kopp, E., and Ghosh, S. (1994). Inhibition of NF-kappa B by sodium salicylate and aspirin. Science 265, 956–959. Kovtonyuk, L.V., Manz, M.G., and Takizawa, H. (2016). Enhanced thrombopoietin but not G-CSF receptor stimulation induces self-renewing hematopoietic stem cell divisions in vivo. Blood 127, 3175–3179. Kristinsson, S.Y., Landgren, O., Samuelsson, J., Bjo¨rkholm, M., and Goldin, L.R. (2010). Autoimmunity and the risk of myeloproliferative neoplasms. Haematologica 95, 1216–1220.

Megı´as, J., Ya´n˜ez, A., Moriano, S., O’Connor, J.-E., Gozalbo, D., and Gil, M.-L. (2012). Direct Toll-like receptor-mediated stimulation of hematopoietic stem and progenitor cells occurs in vivo and promotes differentiation toward macrophages. Stem Cells 30, 1486–1495. Morita, Y., Ema, H., and Nakauchi, H. (2010). Heterogeneity and hierarchy within the most primitive hematopoietic stem cell compartment. J. Exp. Med. 207, 1173–1182. Morrison, S.J., and Scadden, D.T. (2014). The bone marrow niche for haematopoietic stem cells. Nature 505, 327–334. Nagai, Y., Garrett, K.P., Ohta, S., Bahrun, U., Kouro, T., Akira, S., Takatsu, K., and Kincade, P.W. (2006). Toll-like receptors on hematopoietic progenitor cells stimulate innate immune system replenishment. Immunity 24, 801–812. O’Neill, L.A.J., and Bowie, A.G. (2007). The family of five: TIR-domain-containing adaptors in Toll-like receptor signalling. Nat. Rev. Immunol. 7, 353–364. Oh, J., Lee, Y.D., and Wagers, A.J. (2014). Stem cell aging: Mechanisms, regulators and therapeutic opportunities. Nat. Med. 20, 870–880. Orkin, S.H., and Zon, L.I. (2008). Hematopoiesis: An evolving paradigm for stem cell biology. Cell 132, 631–644. Pietras, E.M., Mirantes-Barbeito, C., Fong, S., Loeffler, D., Kovtonyuk, L.V., Zhang, S., Lakshminarasimhan, R., Chin, C.P., Techner, J.-M., Will, B., et al. (2016). Chronic interleukin-1 exposure drives haematopoietic stem cells towards precocious myeloid differentiation at the expense of self-renewal. Nat. Cell Biol. 18, 607–618. Robinson, M.D., McCarthy, D.J., and Smyth, G.K. (2010). edgeR: A Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140. Sato, T., Onai, N., Yoshihara, H., Arai, F., Suda, T., and Ohteki, T. (2009). Interferon regulatory factor-2 protects quiescent hematopoietic stem cells from type I interferon-dependent exhaustion. Nat. Med. 15, 696–700. Schmid, M.A., Takizawa, H., Baumjohann, D.R., Saito, Y., and Manz, M.G. (2011). Bone marrow dendritic cell progenitors sense pathogens via Toll-like receptors and subsequently migrate to inflamed lymph nodes. Blood 118, 4829–4840. €rch, C.M., Riether, C., and Ochsenbein, A.F. (2014). Cytotoxic CD8+ Schu T cells stimulate hematopoietic progenitors by promoting cytokine release from bone marrow mesenchymal stromal cells. Cell Stem Cell 14, 460–472. Shi, C., Jia, T., Me´ndez-Ferrer, S., Hohl, T.M., Serbina, N.V., Lipuma, L., Leiner, I., Li, M.O., Frenette, P.S., and Pamer, E.G. (2011). Bone marrow mesenchymal stem and progenitor cells induce monocyte emigration in response to circulating toll-like receptor ligands. Immunity 34, 590–601. Sinha, M., Jang, Y.C., Oh, J., Khong, D., Wu, E.Y., Manohar, R., Miller, C., Regalado, S.G., Loffredo, F.S., Pancoast, J.R., et al. (2014). Restoring systemic GDF11 levels reverses age-related dysfunction in mouse skeletal muscle. Science 344, 649–652.

Kristinsson, S.Y., Bjo¨rkholm, M., Hultcrantz, M., Derolf, A.R., Landgren, O., and Goldin, L.R. (2011). Chronic immune stimulation might act as a trigger for the development of acute myeloid leukemia or myelodysplastic syndromes. J. Clin. Oncol. 29, 2897–2903.

Takizawa, H., Regoes, R.R., Boddupalli, C.S., Bonhoeffer, S., and Manz, M.G. (2011). Dynamic variation in cycling of hematopoietic stem cells in steady state and inflammation. J. Exp. Med. 208, 273–284.

Liao, Y., Smyth, G.K., and Shi, W. (2014). featureCounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930.

Takizawa, H., Boettcher, S., and Manz, M.G. (2012). Demand-adapted regulation of early hematopoiesis in infection and inflammation. Blood 119, 2991–3002.

Liu, A., Wang, Y., Ding, Y., Baez, I., Payne, K.J., and Borghesi, L. (2015). Cutting edge: Hematopoietic stem cell expansion and common lymphoid progenitor depletion require hematopoietic-derived, cell-autonomous TLR4 in a model of chronic endotoxin. J. Immunol. 195, 2524–2528.

Tothova, Z., Kollipara, R., Huntly, B.J., Lee, B.H., Castrillon, D.H., Cullen, D.E., McDowell, E.P., Lazo-Kallanian, S., Williams, I.R., Sears, C., et al. (2007). FoxOs are critical mediators of hematopoietic stem cell resistance to physiologic oxidative stress. Cell 128, 325–339.

Manz, M.G., and Boettcher, S. (2014). Emergency granulopoiesis. Nat. Rev. Immunol. 14, 302–314.

Trumpp, A., Essers, M., and Wilson, A. (2010). Awakening dormant haematopoietic stem cells. Nat. Rev. Immunol. 10, 201–209.

Massberg, S., Schaerli, P., Knezevic-Maramica, I., Ko¨llnberger, M., Tubo, N., Moseman, E.A., Huff, I.V., Junt, T., Wagers, A.J., Mazo, I.B., and von Andrian,

Varney, M.E., Niederkorn, M., Konno, H., Matsumura, T., Gohda, J., Yoshida, N., Akiyama, T., Christie, S., Fang, J., Miller, D., et al. (2015). Loss of Tifab, a

Cell Stem Cell 21, 225–240, August 3, 2017 239

del(5q) MDS gene, alters hematopoiesis through derepression of Toll-like receptor-TRAF6 signaling. J. Exp. Med. 212, 1967–1985.

Accumulation of oxidative DNA damage restricts the self-renewal capacity of human hematopoietic stem cells. Blood 118, 2941–2950.

Walter, D., Lier, A., Geiselhart, A., Thalheimer, F.B., Huntscha, S., Sobotta, M.C., Moehrle, B., Brocks, D., Bayindir, I., Kaschutnig, P., et al. (2015). Exit from dormancy provokes DNA-damage-induced attrition in haematopoietic stem cells. Nature 520, 549–552.

Zhang, X., Sejas, D.P., Qiu, Y., Williams, D.A., and Pang, Q. (2007). Inflammatory ROS promote and cooperate with the Fanconi anemia mutation for hematopoietic senescence. J. Cell Sci. 120, 1572–1583.

Wang, Y., Kellner, J., Liu, L., and Zhou, D. (2011). Inhibition of p38 mitogenactivated protein kinase promotes ex vivo hematopoietic stem cell expansion. Stem Cells Dev. 20, 1143–1152. Watson, K.G., and Holden, D.W. (2010). Dynamics of growth and dissemination of Salmonella in vivo. Cell. Microbiol. 12, 1389–1397. Yahata, T., Takanashi, T., Muguruma, Y., Ibrahim, A.A., Matsuzawa, H., Uno, T., Sheng, Y., Onizuka, M., Ito, M., Kato, S., and Ando, K. (2011).

240 Cell Stem Cell 21, 225–240, August 3, 2017

Zhang, H., Rodriguez, S., Wang, L., Wang, S., Serezani, H., Kapur, R., Cardoso, A.A., and Carlesso, N. (2016). Sepsis induces hematopoietic stem cell exhaustion and myelosuppression through distinct contributions of TRIF and MYD88. Stem Cell Reports 6, 940–956. Zhao, J.L., Ma, C., O’Connell, R.M., Mehta, A., DiLoreto, R., Heath, J.R., and Baltimore, D. (2014). Conversion of danger signals into cytokine signals by hematopoietic stem and progenitor cells for regulation of stress-induced hematopoiesis. Cell Stem Cell 14, 445–459.

STAR+METHODS KEY RESOURCES TABLE

REAGENT or RESOURCE

SOURCE

IDENTIFIER

Biotin anti-CD3e

Biolegend

100304

Biotin anti-IL7Ra (CD127)

Biolegend

121104

Biotin anti-Gr-1

Biolegend

108404

Antibodies

Biotin anti-B220

Biolegend

103204

Biotin anti-CD4

Biolegend

100404

Biotin anti-CD8a

Biolegend

100704

Biotin anti-CD11b

Biolegend

101204

Biotin anti-NK1.1

Biolegend

108704

Biotin anti-Ter119

Biolegend

116204

Pacific blue streptavidin

Thermo Fischer Scientific

S11222

APC-Cy7 Sca-1

Biolegend

108126

PE-Cy7 c-Kit (CD117)

Biolegend

105814

FITC CD41

Biolegend

133904

FITC CD48

Biolegend

103404

APC CD34

Biolegend

128612

PE CD16/32

Biolegend

101307

PE Flt3 (CD135)

Biolegend

135306

FITC B220

Biolegend

103205

PE-Cy5 CD3e

Biolegend

100310

APC Cy7 CD11b

Biolegend

101225

APC Gr-1

Biolegend

108412

Pacific blue CD45.1

Biolegend

110722

PE-Cy7 CD45.2

Biolegend

109830

Biotin TLR4/MD-2

Thermo Fischer Scientific

13-9924-81

Alexa Fluor 488 p38 MAPK (pT180/pY182)

BD Biosciences

612594

Alexa Fluor 647 ERK1/2 (pT202/pY204)

BD Biosciences

612593

Alexa Fluor 488 Mouse anti-NF-kB p65 (pS529)

BD Biosciences

558421

Alexa Fluor 488 Ki67

BD Biosciences

561165

Anti-phospho-Histone H2A.X (Ser139)

Millipore

05-636

Anti-53BP1

Abcam

Ab21083

AlexaFluor 647 goat anti-mouse

Thermo Fischer Scientific

A-21235

AlexaFluor 488 goat anti-rabbit

Thermo Fischer Scientific

A-11008

Bacterial and Virus Strains LPS (derived from Escherichia coli 0111:B4)

Invivogen

LPS-EB Ultrapure

Salmonella enterica ssp. enterica serovar Typhimurium (SL1344)

Diard et al., 2013

N/A

BrdU (5-Bromo-20 -deoxyuridine)

Sigma Aldrich

B5002

SE-CFSE (5-(and-6)-Carboxyfluorescein Diacetate, Succinimidyl Ester)

Thermo Fischer Scientific

C1157

Chemicals, Peptides, and Recombinant Proteins

NAC (N-Acetyl-L-cysteine)

Calbiochem

616-91-1

SB203580

Calbiochem

152121-47-6

SB 202190

Calbiochem

350228-36-3

IKK-NBD peptide

Enzo Life Sciences

BML-P607-0500 (Continued on next page)

Cell Stem Cell 21, 225–240.e1–e5, August 3, 2017 e1

Continued REAGENT or RESOURCE

SOURCE

IDENTIFIER

Sodium Salicylate

Calbiochem

54-21-7

DCF-DA

Thermo Fischer Scientific

C6827

Streptavidin Microbeads

Miltenyi Biotec

130-048-101

APC BrdU Flow Kit

BD Biosciences

552598

PerFix EXPOSE

Beckman Coulter

B26976

Taqman gene expression assay: Tlr4

Thermo Fischer Scientific

Mm00445273_m1

Taqman gene expression assay: Trif

Thermo Fischer Scientific

Mm00844508_s1

Taqman gene expression assay: Myd88

Thermo Fischer Scientific

Hs01573837_g1

Taqman gene expression assay: Mpl

Thermo Fischer Scientific

Mm00440310_m1

Taqman gene expression assay: Actb

Thermo Fischer Scientific

Mm00607939_s1

Taqman gene expression assay: Il-6

Thermo Fischer Scientific

Mm00446190_m1

Taqman gene expression assay: Cd28

Thermo Fischer Scientific

Mm01253994_m1

Taqman gene expression assay: Frk

Thermo Fischer Scientific

Mm00456656_m1

Taqman gene expression assay: Zbp1

Thermo Fischer Scientific

Mm01247052_m1

This paper

GEO: PRJEB20936

Critical Commercial Assays

Deposited Data RNA-seq data Software and Algorithms FACSDiva software

BD Bioscience

N/A

Flowjo

Flowjo

N/A

LAS X software

Leica

N/A

Imaris

Bitplane

N/A

ezRun

R package

https://github.com/uzh/ezRun

SUSHI

Hatakeyama et al., 2016

N/A

Fastqc

N/A

http://www.bioinformatics.babraham.ac.uk/ projects/fastqc/

FastQ Screen

N/A

http://www.bioinformatics.babraham.ac.uk/ projects/fastq_screen/

STAR aligner

Dobin et al., 2013

N/A

featureCounts

Liao et al., 2014

N/A

MATLAB 9.1

The Mathworks

N/A

edgeR package

Robinson et al., 2010

N/A

Shiny

R-Studio

http://shiny.rstudio.com/

MetaCore software (version 6.18)

Thomson Reuters

N/A

CONTACT FOR REAGENT AND RESOURCE SHARING Further information and requests for reagents may be directed to, and will be fulfilled by, the Lead Contact, M.G.M. (markus.manz@ usz.ch). EXPERIMENTAL MODEL AND SUBJECT DETAILS Mice Tlr4/, Trif/, and Myd88/ mice (CD45.2+) were kindly provided by Dr. Shizuo Akira (Osaka University). C57BL/6 (CD45.2+) and B6.SJL (CD45.1+) mice were obtained from Jackson Laboratories. CD45.1/2+ mice were generated by intercrossing C57BL/6 (CD45.2+) with B6.SJL (CD45.1+). Tlr4/ mice were backcrossed on C57BL/6 (CD45.2+) or B6.SJL (CD45.1+) for more than six generations, and CD45.1/2+ Tlr4/ mice were generated by interbreeding of both mice. All mice were maintained at the Institute for Research in Biomedicine and University Hospital Zurich animal facilities according to the guidelines of the Swiss Federal Veterinary €ramt of Office. Experiments were approved by the Dipartimento della Sanita` e Socialita`, Ticino, Switzerland, and by the Veterina Kanton Zurich, Zurich, Switzerland.

e2 Cell Stem Cell 21, 225–240.e1–e5, August 3, 2017

Generation of BM chimeric mice, LPS challenge and serial transplantation Six- to twelve-week old WT or Tlr4/ mice (CD45.1/2+) were lethally irradiated with 9.5 Gy (split dose within a 2-4 hr interval) and intravenously (i.v.) transplanted with equal numbers (1:1 ratio) of 1x106 total BM or 3,000 Lin-c-Kit+Sca-1+ cells from WT and Tlr4/, Trif/, or Myd88/ animals. Two to three months after reconstitution, the BM chimeric mice were i.p. injected with 35-50 mg LPS corresponding to 35-50 EU (Escherichia coli 0111:B4, LPS-EB Ultrapure, Invivogen) according to the indicated schemes. Peripheral blood was harvested from BM chimeric mice and relative donor chimerism was assessed in mature hematopoietic cells. At terminal work-up, BM from chimeric mice was harvested and analyzed. Salmonella Tm infection Mice were infected with SB300, a wild-type Salmonella enterica ssp. enterica serovar Typhimurium (S. Tm) strain derived from SL1344 (Watson and Holden, 2010). S. Tm was grown in LB (0.3 M NaCl) for 12 hr, subcultured for 4 hr and resuspended in PBS. Mice were i.v. injected with 1x104 CFU and were sacrificed 48 hr later. For analysis, spleens were removed aseptically, homogenized in 500 ml ice-cold PBS supplemented with 0.5% Tergitol (Sigma Aldrich) and 0.5% BSA (Sigma Aldrich), and plated in appropriate dilutions on MacConkey agar plates (Oxoid). To determine tissue loads, BM of one tibia per mouse was flushed with 500 ml icecold PBT, homogenized and plated on MacConkey agar plates. METHOD DETAILS FACS analysis and sorting All antibodies used in this study were purchased from Thermo Fisher Scientific (eBiosciences) or Biolegend unless specified otherwise. For analyzing donor chimerism in mature hematopoietic cells, cells from BM, spleen, or peripheral blood (PB) were stained with antibodies against CD45.1 (A20), CD45.2 (104), CD19 (1D3), CD3ε (145-2C11), Gr-1 (RB6-8C5) and CD11b (M1/70). For early hematopoietic cell analysis, cells from BM or spleen were incubated with antibodies against lineage (Lin) markers (NK1.1 (PK136), CD11b, Ter119 (Ter119), Gr-1 (RB6-8C5), CD4 (RM4-5), CD8a (53-6.7), CD3ε (145-2C11), B220 (RA3-6B2), IL-7Ra (A7R34), c-Kit (2B8), Sca-1 (D7), CD34 (RAM34), Flt3 (A2F10), CD150 (TC15-12F12.2), CD48 (HM48-1), CD16/32 (2.4G2). For LKS or HSC (LKSCD34-CD48-CD150+) sorting, total BM cells were stained with biotinylated antibodies against lineage markers, and lineage positive cells were depleted with streptavidin conjugated magnetic beads (Militenyi Biotec GmbH). The lineage depleted cells were stained with antibodies against c-Kit and Sca-1, and LKS or HSC were sorted on a FACS Aria III (BD Biosciences). For TLR4 expression, total BM cells were incubated with a biotinylated antibody against TLR4/MD-2 complex (MTS510, Thermo Fisher Scientific) or an isotype-matched control antibody, together with early hematopoietic cell markers (Lin, c-Kit, Sca-1, CD34, CD150, CD16/32) for more than 90 min, and binding was detected by streptavidin conjugated brilliant violet 421. The samples were analyzed on a FACS Canto II (BD Biosciences). BrdU labeling and retention WT or mixed BM chimeric mice were i.p. injected with 9 mg/kg BrdU (Sigma Aldrich) and subsequently put on drinking water containing 800 mg/ml BrdU and 5% sucrose for two weeks with refreshing the water every 2-3 days. Drinking water was then switched to normal water for another two weeks until analysis. While given normal drinking water, animals were treated with PBS or LPS, together with or without inhibitors as described below. One week after final PBS or LPS injections, BM cells were harvested and subjected to cell surface staining with antibodies against lineage markers above, and to intracellular anti-BrdU staining with an APC conjugated antibody according to the manufacturer’s instructions (BD Bioscience). BrdU positivity was determined by measuring baseline intensity using non-BrdU treated BM, stained with the anti-BrdU antibody. CFSE labeling and chasing WT HSCs (3,000-7,835 LKS CD34-CD48-CD41-CD150+ cells), MPPs (113,250 LKS CD34+ cells) or naive CD4 T (1x106 CD4+CD62L+ cells) were isolated from B6/J mice (CD45.1+), and labeled for 7 min at 37 C with 2 mM CFSE (Thermo Fisher Scientific) in PBS supplemented with 1% FCS as previously described (Takizawa et al., 2011; Kovtonyuk et al., 2016; Pietras et al., 2016). The same volume of ice-cold PBS with 10% FCS was then added to stop the reaction. After wash, the CFSE-labeled cells were i.v. transferred into Tlr4/ mice (CD45.1+) without irradiation. Two or three days after mice were treated with PBS or 100 mg LPS for 2 or 3 times and were analyzed on day 6 for HSCs and MPPs. Kaplan-Meier survival following 5FU administration WT mice were i.p. pretreated with PBS or 35 mg LPS, 1 and 3 days or 5 and 7 days before two i.p. injections of 5-FU (150 mg/kg) in a weekly interval. The survival of animals was followed over one month. In vivo inhibitor treatment For inhibitor treatment, BM chimeric mice were i.p. injected with 100 mg/kg NAC (Calbiochem, CA, USA) for ROS, 15 mg/kg SB203580 or SB 202190 (Calbiochem) for p38, 2 mg/kg of IKK-NBD peptide (Enzo Life Sciences) for NF-kB, 200 mg/kg Sodium Salicylate (Calbiochem) for NF-kB or the respective vehicle, 24 and 5 hr prior to PBS or LPS administration.

Cell Stem Cell 21, 225–240.e1–e5, August 3, 2017 e3

Detection of ROS, NF-kB, Erk1/2 and p38 activation HSCs (LKS CD34-CD150+), MPPs (LKS CD34+), and CMPs/GMPs (LK CD34+) were sorted from BM of WT mice that were treated with PBS or 50 mg LPS 16 hr before, and incubated with 10 mM DCF-DA (Thermo Fisher Scientific) for 30 min at 37 C, followed by fixation with 2% PFA for 10 min. After permeabilization with 0.1% Triton X-100 (Roche)/PBS for 20 min and blocking with 0.1% Triton X-100/ PBS containing 10% goat serum (DAKO) for an hour, the cells were stained with anti-phospho p38 (Cell signaling) and detected by goat anti-rabbit AlexaFluor 488 (Thermo Fisher Scientific). Cells were analyzed on a FACS Canto II, and positivity for each marker was determined based on the intensity of PBS-treated cells that were stained with both markers. For the detection of NF-kB, Erk1/2 and p38 phosphorylation, 3,000-10,000 sorted HSCs, were starved for 1 hr, and stimulated with 50 ng/ml TPO (Peprotech) or 200 mg/ml LPS for 30 min at 37 C in a 5% CO2 incubator. Subsequently, the cells were fixed, permeabilized according to the manufacturer’s instructions (PerFix EXPOSURE, Beckman Coulter), and stained with anti-phospho NF-kB, Erk1/2 and p38 (BD Biosciences), followed by analysis on a FACS LSRFortessa (BD Biosciences). Immunocytochemistry HSC (LKS CD34-CD48-CD41-CD150+) and LK CD34-CD48-CD41- cells were sorted from BM of WT mice at 3 or 14 days after i.p. injection of PBS or 50 mg LPS for 8 or 12 times, and plated on a glass slide, coated with 0.01% poly-L-lysine (Sigma Aldrich), followed by 60 min incubation at R.T (Room Temperature) to let cells settle down on the slide. Cells were then fixed with 4% PFA for 10 min and permeabilized with 0.1% Triton X-100 (Roche)/PBS for 10 min and blocked with 1% Triton X-100/PBS containing 10% goat serum (DAKO) or 1% BSA (Europa Bioproducts) for one hour or overnight. Cells were stained with anti-phospho-histone H2AX (JBW301, Millipore) and/or rabbit anti-53BP1 (Abcam) overnight at 4 C, and subsequently binding was visualized with a goat anti-mouse IgG Alexa Fluor 647 (Thermo Fisher Scientific), goat anti-rabbit IgG Alexa Fluor488 (Thermo Fisher Scientific) for the respective primary antibodies. Nuclei were counterstained with DAPI (Thermo Fisher Scientific) and images were taken on a confocal microscopy with a 63x objective lens (SP5, Leica). The number of phospho-H2AX positive and 53PB1 positive foci was counted manually on image software, Imaris (Bitplane, USA). Quantitative RT-PCR HSC and HPC from BM were enriched for Lin- cells with biotin-conjugated lineage markers as described above. Cells were then stained with the antibodies against CD16/32 (2.4G2), c-Kit (2B8), Sca-1 (D7), Flt3 (A2F10), CD34 (RAM34), CD150 (TC1512F12.2). For mature cell populations, BM and spleen cells were stained with Mac1 (M1/70) and Gr-1, and MHCII (M5/114.15.2), CD11c (N418), respectively, and were directly sorted into lysis buffer containing b-mercaptoethanol. The cell lysate was subjected to RNA isolation (QIAGEN), cDNA synthesis (Applied Biosystems) and qPCR with SYBR green reagent (Applied Biosystems) and the following synthesized gene specific primers on a 7500 Fast Real Time PCR System (Applied Biosystems): Tlr4 (50 -CCA ATG CAT GGA TCA GAA ACT C-30 ; 50 -ATT TCA CAC CTG GAT AAA TCC AGC-30 ), Trif (50 -TAC AGC CAG GTC TGT GCT-30 ; 50 -AGA ATG AAG CCT GGA GCC-30 ), Myd88 (50 -TCG ATG CCT TTA TCT GCT ACT G-30 ; 50 -TCT GTT GGA CAC CTG GAG AC-30 ), Mpl (50 -GAA GCT GTC TCG TCT CAG G-30 ; 50 - TCC AAT TGT CAC TGC ATC TCC-30 ) and Actb (50 -AGA TGA CCC AGA TCA TGT TTG AG-30 ; 50 - GTA CGA CCA GAG GCA TAC AG-30 ). Relative mRNA expression of each gene was normalized against relative expression of beta-actin. For in vitro stimulation, 7,000-8,000 HSCs (LKS CD34-CD48-CD41-CD150+), 20,000 LK from BM and 20,000 Mac-1+Gr-1+ granulocytes from spleen were cultured in conditioned media (1% non-essential amino acid, 1% HEPES, 10% FBS in IMDM) with PBS or 100 mg/ml LPS for 6 or 12 hr, and subjected to RNA isolation, cDNA synthesis and Real Time PCR on a 7500 Fast Real Time PCR System as above with TaqMan primers (Thermo Fisher Scientific) specific to Tlr4, Trif, Myd88, Mpl and Actb. To validate RNA sequencing data, 6,000- 8,000 HSC were isolated from 6-8 mice and tested for TaqMan primers (Thermo Fisher Scientific) specific for Il-6, Cd28, Frk and Zbp1 as described. RNA sequencing analysis 7,000-8,000 HSCs (LKS CD34-CD48-CD41-CD150+) were isolated and cultured in conditioned media (1% non-essential amino acid, 1% HEPES, 1% GlutaMAXTM, 10% FBS in IMDM) with PBS or 100 mg/ml LPS for 6 or 12 hr. RNA was isolated using RNeasy Plus Micro Kit (QIAGEN) from stimulated cells and subjected to cDNA synthesis (High-Capacity cDNA Reverse Transcription Kit, Applied Biosystems) and RNA library preparation as follows: quantity and quality of the isolated RNA was determined with a Qubit (1.0) Fluorometer (Thermo Fisher Scientific, California, USA) and a Bioanalyzer 2100 (Agilent, Waldbronn, Germany). The TruSeq Stranded mRNA Sample Prep Kit (Illumina, Inc, California, USA) was used in the succeeding steps. Briefly, total RNA samples (100 ng) were poly-A selected and then reverse-transcribed into double-stranded cDNA with Actinomycin added during first-strand synthesis. Subsequently, tcDNA samples were fragmented, end-repaired and adenylated before ligation of TruSeq adapters that contain the index for multiplexing. Fragments containing TruSeq adapters on both ends were selectively enriched with PCR, and the quality and quantity were validated using Qubit (1.0) Fluorometer and the Bioanalyzer 2100. The product was a smear with an average fragment size of approximately 360 bp. The libraries were normalized to 10nM in Tris-Cl 10 mM, pH8.5 with 0.1% Tween 20. The TruSeq SR Cluster Kit v4-cBot-HS (Illumina, Inc, California, USA) was used for cluster generation using 8 pM of pooled normalized libraries on the cBOT. Sequencing was performed on the Illumina HiSeq 2500 single end 125 bp using the TruSeq SBS Kit v4-HS (Illumina, Inc, California, USA). Bioinformatic analysis was performed using the R package ezRun (https://github.com/uzh/ezRun) within the data analysis framework SUSHI

e4 Cell Stem Cell 21, 225–240.e1–e5, August 3, 2017

(Hatakeyama et al., 2016). In detail, the raw reads were quality checked using Fastqc (http://www.bioinformatics.babraham.ac. uk/projects/fastqc/) and FastQ Screen (http://www.bioinformatics.babraham.ac.uk/projects/fastq_screen/). Quality controlled reads (adaptor trimmed, first 3 and last 6 bases hard trimmed, minimum average quality Q10, minimum tail quality Q10, minimum read length 20nt) were aligned to the reference genome (Ensembl GRCm38, not patched) using the STAR aligner (Dobin et al., 2013) with the additional options (–outFilterMatchNmin 30–outFilterMismatchNmax 10–outFilterMismatchNoverLmax 0.05–outFilterMultimapNmax 50), which means that we required at least 30 bp matching, and accept at most 10 or 5% mismatches. Read alignments were only reported for reads with less than 50 valid alignments. Expression counts were computed using featureCounts in the Bioconductor package Subread (Liao et al., 2014). Principal Component Analysis (PCA) was performed using MATLAB 9.1 (The Mathworks, Inc) to compare normalized read counts of 11678 protein coding genes in all datasets. Differential expression was computed using the edgeR package (Robinson et al., 2010). Quality checkpoints, such as quality control of the alignment and count results, were implemented in ezRun (https://github.com/uzh/ezRun) and applied throughout the analysis workflow to ensure correct data interpretation. Data analysis was performed using Shiny by R-Studio (http://shiny.rstudio.com/), a customized web interface produced by the functional genomic center Zurich. The differentially expressed genes were identified using a 3-Way-Venn diagram. Gene ontology analysis of the differentially expressed genes was performed using MetaCore version 6.18 (GeneGo, St. Joseph, MI, USA) to generate a network enrichment via GO Processes. QUANTIFICATION AND STATISTICAL ANALYSIS All data are shown as the mean ± s.e.m., unless indicated otherwise. Kaplan-Meier survival was analyzed by Log-rank test on GraphPad Prism 5 (GraphPad sofrware, Inc). All statistical comparisons except Kaplan-Meier survival were evaluated with Student’s t-Test. Kaplan-Meier survival was statistically analyzed by Logrank test. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001 (twotailed t test). DATA AND SOFTWARE AVAILABILITY The accession number for all sequencing data reported in this paper is GEO: PRJEB20936. The list of genes that are differentially expressed among WT and TLR signal-deficient HSCs with or without LPS stimulation are shown in Figure S7.

Cell Stem Cell 21, 225–240.e1–e5, August 3, 2017 e5

Cell Stem Cell, Volume 21

Supplemental Information

Pathogen-Induced TLR4-TRIF Innate Immune Signaling in Hematopoietic Stem Cells Promotes Proliferation but Reduces Competitive Fitness Hitoshi Takizawa, Kristin Fritsch, Larisa V. Kovtonyuk, Yasuyuki Saito, Chakradhar Yakkala, Kurt Jacobs, Akshay K. Ahuja, Massimo Lopes, Annika Hausmann, Wolf-Dietrich Hardt, Álvaro Gomariz, César Nombela-Arrieta, and Markus G. Manz

Takizawa  et.  al.   Running  title:  Direct  TLR4  ligation  on  HSCs  limits  their  fitness    

Supplemental  information      

Pathogen-­induced   TLR4-­TRIF   innate   immune   signaling   in   hematopoietic   stem   cells   promotes   proliferation   but   reduces   competitive  fitness   Hitoshi  Takizawa,  Kristin  Fritsch,  Larisa  V.  Kovtonyuk,  Yasuyuki  Saito,   Chakradhar  Yakkala,  Kurt  Jacobs,  Akshay  K.  Ahuja,  Massimo  Lopes,  Annika   Hausmann,  Wolf-­Dietrich  Hardt,  Álvaro  Gomariz,  César  Nombela-­Arrieta  and   Markus  G.  Manz  

-­‐1-­‐  

Figure S1. Takizawa et. al.

A HSC

MPP GMP

CMP

CD150

c-Kit Sca-1

B

CD16/32

MEP

CD34

MPP

CMP

GMP

MEP

Relative cell number

HSC

CD34

TLR4/MD-2

Relative expression to Actb (log2)

C 8

8

32

8

HSC MPP

1 1

1/8

1

CMP

1

MEP GMP

1/32

CD11c+MHCII+ splenocytes CD11b+Gr-1hi BM cells

Tlr4

1/8

Trif

1/8

Myd88

0

Mpl

Figure S2. Takizawa et. al.

A

Total BM

4 3 2 1 0 div.

45.6

0.0044

CD150

c-Kit

CD45.1

c-Kit

Lineage

CD45.2

CD150

60.0

HSC transplanted

58.8

Sca-1

CFSE

CD34

Donor CD4+CD62L+ Lineage

CD45.2

84.9

12.5

CD4

0.42

CD4

CD62L

CD45.1

CD4+CD62L+ T cell transplanted

B

Sca-1

CD45.2

CFSE

CFSE

Total BM

4 3 2 1 0 div.

14.6

CD150

c-Kit

CD45.2

15.4

Sca-1

CFSE

CD34

Total splenocytes 84.9 12.5

CD4

0.42 CD4

CD4+CD62L+ T cell transplanted

CD45.1

Lineage

0.34

CD62L

c-Kit

CD45.1

34.0

CD16/32

MPP transplanted

CD45.2

CFSE

Figure S3. Takizawa et. al.

A

8xPBSorLPS -2

WT

0

4 wks LKS sorting+Tx with competitor PB analysis

IR WT

B Ratio:donor /competitor

CD45+

4 8 16

22 wks

CD19+

Mac+Gr+

CD3+

5

5

5

5

4

4

4

4

3

3

3

3

2

2

2

2

1 0

**

***

4

8

* 12

16

1 0

**

***

4

8

* 12

16

1 0

1 4

Weeks after transplantation

C

BM analysis 22 wks post 1st Tx Ratio:donor /competitor

3

*

*

ns

ns

2

PBS-WT LPS-WT

1

0

HSC

MPP

CMP/ GMP

MEP

PBS-WT LPS-WT

* 8

12

*

16

0

4

* 8

12

*

16

Figure S4. Takizawa et. al.

A TLR4

B

N-acetyl cystein (NAC)

-24

WT ROS

ERK1/2

IkB

DMSO

16h

0

IKK-NBD

NaS

SB

NAC

SB/NAC

2.8

2.0

2.9

1.8

1.0

1.9

10.7

8.2

7.1

4.0

3.1

2.3

HSC

MMP

CMP/GMP

PBS

p38 SB203580/202190 (SB)

NFkB

-5

BM analysis

LPS

Sodium Salicylate (NaS)

c-Kit

IKK-NBD peptide

C

ASK1

MKK3/6

IKKβ

PBSorLPS

Inhibitors

Sca-1

D §

60,000

Number of HSCs

*

ns

ns

E

**

*

40,000

pp38 §

§§§

§

§§§

20,000

§

LPS 0h LPS 2h LPS 16h

PBS LPS

ROS

LPS 0h LPS 2h LPS 16h

AC SB /N AC

SB

Signal activation

N

aS

BD

N

N K-

IK

D

M

SO

0

F

G

WT

HSCs PBS/ LPS

WT or KOs

0

Cell harvest and pFLOW

30min

pNFKb

Tlr4-/100

100

80

80

80

80

60

60

60

60

40

40

40

40

20

20

20

0 -10

3

0

10

3

10

4

10

5

20

0 -10

3

0

10

3

10

4

10

5

0 -10

3

0

10

3

10

4

10

5

-10

100

100

100

100

80

80

80

80

60

60

60

60

40

40

40

40

20

20

20

0

0 -10

p38

Trif-/-

100

0

pERK1/2

Myd88-/-

100

3

0

10

3

10

4

10

5

3

0

10

3

10

4

10

5

3

0

10

3

10

4

10

5

-10

100

100

100

80

80

80

80

60

60

60

60

40

40

40

40

20

20

20

20

0

0

0

3

0

10

3

10

4

10

5

-10

3

0

10

3

10

4

10

5

10

0

10

0

10

3

10

4

10

5

0 -10

100

-10

0

20

0 -10

3

3

3

10

4

10

5

0 -10

3

0

10

3

10

4

10

5

-10

3

3

10

4

10

5

Figure S5. Takizawa et. al.

PBSorLPS

Analysis

0

16h

WT or KO

B

WT 11.1

40.8

13.5

45.4

13.8

C

Trif-/-

Myd88-/-

Tlr4-/-

34.1 10.7

37.8

PBS

47.5

15.6

24.4

20.8

12.8

41.4 18.7

CD11b

LPS

100

Gr-1

PBSorLPS BM analysis Inhibitors -24

WT

-5

E 150

ns

ns

0

ns

16h

ns

100

PBS LPS

50

AC N

SB

aS N

BD KN IK

D

M

SO

0

*** ***

ns

PBS

ns

50

0

D

Ratio: Gr-1 low/hi in BM (%)

18.9

150

Ratio: Gr-1 low/hi in BM (%)

A

WT

Tlr4-/-

Myd88-/-

LPS

Trif-/-

Figure S6. Takizawa et. al.

A

MyelP sorting/analysis

8x PBS/ LPS

WT or KOs -14

0

3, 14d

B

PBS

LPS

IR

γH2AX

53BP1

Merge

C

3d after LPS

Cells with >3 γH2AX+ foci (%)

100

50

50

0

n= 24 14 35 32 30 39 21 20 14 WT

D

Tlr4

-/-

Myd88

-/-

Trif

-/-

PBS LPS IR

0

n= 41 48 31 24 7 WT

IR

3d after LPS

100 Cells with >3 53BP1+ foci (%)

14d after LPS

100

17 32 32 38

Myd88-/-

Trif-/-

IR

14d after LPS

100

50

Tlr4-/-

50

0

n= 24 14 35 32 30 39 21 20 14 WT

Tlr4

-/-

Myd88

-/-

Trif

-/-

IR

PBS LPS IR

0

n= 41 48 31 24 7 WT

Tlr4-/-

17 32 32 38

Myd88-/-

Trif-/-

IR

Figure S7. Takizawa et. al. A

Up-regulated genes in LPS-stimulated WT and Myd88-/-

Down-regulated genes in LPS-stimulated WT and Myd88-/Log2 Ratio LPS vs PBS

Log2 Ratio LPS vs PBS

B

C

Gene

WT

Myd88-/-

Gene

WT

Myd88-/-

Pou5f2 Apc2 Fap Gm10778 Efcab3 Chst1 Ica1l Mid1 Bhlha15 Tnfrsf11b Ccdc184 Fcgr1 Il1r2 Cd276 Cd300lf Reep6 Bend7 Ddx60 Frk Akap2 Pakap Galnt12 Dlg2 Fam195a Zbp1 Wnt10b Ydjc Gja1 F830016B08Rik Slamf8 Gm4955 Dusp8 Ifi204 Prox1 C1ra Gm5431 Abcc3 Bcat1

3.335 3.146 2.635 2.503 2.355 2.036 1.858 1.838 1.817 1.684 1.677 1.599 1.558 1.554 1.481 1.361 1.267 1.258 1.257 1.255 1.237 1.205 1.195 1.186 1.160 1.154 1.149 1.139 1.112 1.098 1.093 1.081 1.080 1.073 1.027 1.025 1.021 1.016

1.168 1.972 6.491 1.475 6.377 2.131 1.336 1.020 1.651 1.081 1.292 1.080 2.354 1.186 2.842 2.054 1.287 1.327 2.966 1.487 1.487 2.923 1.208 1.028 2.037 2.224 1.807 1.487 1.919 1.624 2.900 1.912 1.512 1.010 1.579 1.236 1.857 2.766

Gpd1 Dysf Ston1 Sfrp2 Tbc1d30 1700024P16Rik Lpar5 Rapgef3 F13a1 Fstl3 Pard6g Sgip1 1700049G17Rik Grhpr Cd28 Hpn Amacr Lhpp Tcea2 Ermard Ap5s1 2310039H08Rik Slit2 Steap3 Dmd Plcb1 Lyplal1 Dusp18 Serpinb10 Slc25a35 Gpr21 Rnf17 9330182L06Rik Ghdc Nudt1 Kcnip3 Bgn

-2.830 -2.504 -2.456 -1.974 -1.886 -1.875 -1.785 -1.711 -1.710 -1.670 -1.668 -1.661 -1.522 -1.489 -1.438 -1.416 -1.415 -1.413 -1.409 -1.402 -1.365 -1.365 -1.332 -1.330 -1.272 -1.266 -1.199 -1.156 -1.130 -1.130 -1.092 -1.089 -1.087 -1.081 -1.076 -1.042 -1.013

-2.438 -3.791 -2.203 -3.488 -1.347 -1.177 -1.369 -2.936 -1.384 -1.044 -3.096 -2.658 -1.358 -1.202 -3.018 -1.033 -1.692 -1.407 -1.217 -1.199 -1.221 -1.286 -1.287 -1.164 -1.228 -1.050 -1.728 -1.068 -1.217 -1.728 -1.471 -1.385 -1.160 -1.850 -1.564 -2.681 -1.182

Development_Neurogenesis in general Proliferation_Negative regulation of cell proliferation Cell adhesion_Cadherins Development_Neurogenesis_Synaptogenesis Signal transduction_WNT signaling Cell cycle_G1-S Development_Ossification and bone remodeling Cardiac development_Wnt_beta-catenin, Notch, VEGF, IP3 and integrin signaling Cell adhesion_Amyloid proteins

Reproduction_GnRH signaling pathway Cell adhesion_Cell junctions Cell adhesion_Platelet aggregation Inflammation_MIF signaling Inflammation_IgE signaling Cell adhesion_Integrin priming Inflammation_Protein C signaling Signal Transduction_Cholecystokinin signaling Neurophysiological process_Long-term potentiation Development_Neurogenesis_Axonal guidance Neurophysiological process_Circadian rhythm Cell adhesion_Leucocyte chemotaxis Signal transduction_WNT signaling Signal Transduction_TGF-beta, GDF and Activin signaling Development_Neuromuscular junction Inflammation_Kallikrein-kinin system

-log(pValue)

F

-log(pValue)

Takizawa  et.  al.   Running  title:  Direct  TLR4  ligation  on  HSCs  limits  their  fitness    

Figure   S1.   Related   to   Figure   1.   HSC   and   progenitors   express   TLR4/MD-­‐2   complex.   (A)   Representative   flow-­‐cytometric   profiles   of   HSCs   and   progenitors   in   BM   of   WT   mice.   Lin-­‐   BM   cells   were   pre-­‐gated   and   developed   by   c-­‐Kit   and   Sca-­‐1,   and   subsequently   CD150   and   CD34,   or   CD16/32   and   CD34,   to   define   the   following   populations:   HSCs   (LKSCD34-­‐CD150+),   MPPs   (LKSCD34+),  

CMPs  

(LKCD16/32lowCD34+),  

GMPs  

(LKCD16/32hiCD34+)  

and  

MEPs  

(LKCD16/32lowCD34-­‐).   (B)   Representative   histogram   of   TLR4/MD-­‐2   expression   in   HSCs,   MPPs,   CMPs,   GMPs   and   MEPs   (red:   anti-­‐TLR4/MD2   antibody;   gray:   isotype-­‐matched   antibody).   (C)   Quantitative   PCR   analysis   on   mRNA   expression   of   the   indicated   genes   in   FACS-­‐sorted   BM   HSCs   (LKS   Flt3−CD34−CD48−CD150+),   MPPs   (LKS   Flt3+CD34+),   CMPs,   MEPs   (LKCD16/32lowCD34-­‐),   GMPs   (LKCD16/32hiCD34+),   dendritic   cells   (CD11c+MHCII+)   from   spleen   and   granulocytes   (Mac1+Gr1+)   from   BM.   Relative   expression   of   the   indicated   genes   to   the   expression   of   Actb   is   shown   in   respective  cell  populations  (mean  ±  SEM).  Data  are  pooled  from  2-­‐4  independent  experiments.     Figure   S2.   Related   to   Figure   1.   Representative   gating   strategy   for   CFSE   dilution   experiment.  (A)   Representative   FACS   plots   of   CD45.2+   BM   transplanted   with   CD45.1+   HSCs   (LKS   CD34-­‐CD48-­‐ CD150+)   and   treated   with   LPS   (red)   or   PBS   (blue)   (upper   panel)   as   depicted   in   Figure   1G-­‐I.   The   non-­‐divided   cell   fraction   was   determined   according   to   CFSE   gating   for   CD45.2+   naïve   T   cells   (CD4+CD62L+)  that  engrafted  in   the  spleen  of  another  CD45.1+   animal   (lower   panel),   and   division   gates   were   set   according   to   CFSE   dilution   (1-­‐4   divisions).   (B)   Representative   FACS   plots   of   CD45.1+   BM   transplanted   with   CD45.2+   MPPs   (LKS   CD34+)   and   treated   with   LPS   (red)   or   PBS   (blue)  (upper  panel).  The  non-­‐divided  cell  fraction  was  determined  according  to  CFSE  gating  for   CD45.2+  naïve  T  cells  (CD4+CD62L+)  that  engrafted  CD45.1+  spleen  (lower  panel).     Figure  S3.  Related  to  Figure  2.  Systemic  LPS  application  limits  competitive  repopulating  ability   of   HSCs,   irrespective   of   LPS-­‐induced   HSC   mobilization.   (A)   Experimental   scheme   of   LPS   treatment   and   transplantation   for   results   depicted   in   (B)   and   (C):   WT   or   Tlr4-­‐/-­‐   mice   were   i.p.   injected   with   PBS   or   LPS   (35µg)   8   times   every   other   day.   Four   weeks   after   the   last   injection,   3,000   LKS   (CD45.2+)   were   isolated,   and   competitively   transplanted   with   3,000   non-­‐stimulated   competitor   LKS   (CD45.1+)   into   lethally   irradiated   WT   mice   (CD45.1/2+)   followed   by   monthly   PB   chimerism  analysis  (B)  and  terminal  BM  analysis  (C).  (B)  Percentage  of  donor  derived  CD45+  cells,   CD19+  B  cells,  CD3+  T  cells  and  Mac-­‐1+Gr-­‐1+  granulocytes  in  PB  of  recipients  (PBS-­‐WT:  n=12;  LPS-­‐ WT:   n=14;   LPS-­‐WT:   n=9).   (C)   Percentage   of   donor   derived   HSCs   (LKSCD34-­‐),   MPPs   (LKSCD34+),   CMPs/GMPs   (LKCD34+)   and   MEPs   (LKCD34+)   in   BM   of   recipients   at   terminal   analysis   22   weeks   after   reconstitution   (PBS-­‐WT:   n=6;   LPS-­‐WT:   n=7).   Data   are   pooled   from   two   independent   experiments.  ns,  not  significant;  *p