Immunity

Article

Gene Deregulation and Chronic Activationin Natural Killer Cells Deficientin the Transcription Factor ETS1Kevin Ramirez,1 Katherine J. Chandler,3 Christina Spaulding,2 Sasan Zandi,4 Mikael Sigvardsson,4 Barbara J. Graves,3

and Barbara L. Kee1,2,*1Committee on Immunology2Department of Pathology, Committee on Cancer Biology and The University of Chicago Comprehensive Cancer CenterThe University of Chicago, Chicago, IL 60615, USA3Department of Oncological Sciences, University of Utah School of Medicine, Huntsman Cancer Center, University of Utah,

Salt Lake City, UT 84112, USA4Department of Clinical and Experimental Medicine, Experimental Hematopoiesis Unit, Faculty for Health Sciences, Linkoping University,58183 Linkoping, Sweden

*Correspondence: bkee@bsd.uchicago.edu

DOI 10.1016/j.immuni.2012.04.006

SUMMARY

Multiple transcription factors guide the developmentof mature functional natural killer (NK) cells, yet littleis known about their function. We used global geneexpression and genome-wide binding analysescombinedwith developmental and functional studiesto unveil three roles for the ETS1 transcription factorin NK cells. ETS1 functions at the earliest stages ofNK cell development to promote expression of crit-ical transcriptional regulators including T-BET andID2, NK cell receptors (NKRs) including NKp46,Ly49H, and Ly49D, and signalingmolecules essentialfor NKR function. As a consequence,Ets1�/�NKcellsfail to degranulate after stimulation through acti-vating NKRs. Nonetheless, these cells are hyperre-sponsive to cytokines and have characteristics ofchronic stimulation including increased expressionof inhibitory NKRs and multiple activation-associ-ated genes. Therefore, ETS1 regulates a broadgene expression program in NK cells that promotestarget cell recognition while limiting cytokine-drivenactivation.

INTRODUCTION

Natural killer (NK) cells are lymphocytes that utilize germline-

encoded activating and inhibitory receptors (NKRs) to recognize

virus-infected, transformed, and stressed cells. NK cells also

contribute to adaptive immune responses through the produc-

tion of inflammatory cytokines and by promoting the maturation

or destruction of immature dendritic cells (Vivier et al., 2011).

NK cells are activated when inhibitory NKRs recognizing clas-

sical or nonclassical major histocompatibility complex (MHC)

antigens fail to be engaged (‘‘missing-self’’ recognition) and/or

when activating NKRs detect their ligands, thereby altering the

balance between activating and inhibitory signals (Lanier,

2005). Themechanisms controlling the threshold for NK cell acti-

vation are not well understood but inhibitory receptor signaling

appears to play a role in ‘‘licensing’’ or ‘‘arming or disarming’’

developing NK cells so that engagement of activating receptors

results in a functional response (Joncker and Raulet, 2008;

Yokoyama and Kim, 2006). NK cell immune deficiency results

in susceptibility to infection and, although rare, NK cell malignan-

cies are aggressive and difficult to treat. Therefore, under-

standing the mechanisms that control the development and

function of NK cells has both basic biological and clinical

significance.

NK cells develop in the bone marrow (BM) from common

lymphoid progenitors (CLPs) through three major stages defined

by expression of CD122 (interleukin 2 [IL-2] and IL-15 receptor-b

chain), NK1.1 (an activating NKR expressed in only some strains

of mice), and DX5 (integrin a2) (Kim et al., 2002; Rosmaraki et al.,

2001). NK progenitors (NKPs) are CD122+ and lack NK1.1 and

DX5 (CD49b). CD27, CD127, and CD244 mark a subset of

NKPs (rNKPs) enriched for NK cell potential as well as pre-

NKP cells, a CD122� intermediate betweenCLP and rNKP (Fath-

man et al., 2011). Acquisition of NK1.1 occurs at the immature

NK (iNK) cell stage where multiple NKRs initiate expression

and the cells become dependent on IL-15 for survival (Vosshen-

rich et al., 2005). Themature NK (mNK) cell stage is defined by an

increase in DX5, IL-15-driven expansion, and licensing or arming

of NK cells (Kim et al., 2002; Rosmaraki et al., 2001). Further

maturation correlates with increased expression of CD11b and

decreased expression of CD27 (Chiossone et al., 2009; Kim

et al., 2002).

Although stages in the NK cell program have been character-

ized, little is known about the transcriptional networks that

establish the NK cell gene program or promote NK cell function.

A few transcription factors have been identified that play a major

role in the generation of mNK cells including T-BET and EOMES

(Gordon et al., 2012; Intlekofer et al., 2005), ETS1 (Barton et al.,

1998), E4BP4 (Gascoyne et al., 2009; Kamizono et al., 2009),

TOX1 (Aliahmad et al., 2010), and ID2 (Boos et al., 2007; Yokota

et al., 1999). Although mNK cells largely fail to develop in

these strains, the mechanisms underlying the observed pheno-

types are not known. Moreover, the transcriptional programs

Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc. 921

Immunity

ETS1 Targets in NK Cells

controlling NKR expression and NK cell maturation or function

remain to be determined, although a few factors such as TCF1,

MEF1, and BLIMP1 play a role (Held et al., 1999; Kallies et al.,

2011; Lacorazza et al., 2002).

ETS1, the founding member of the ETS family of transcription

factors, has been known to be important for development of

mNK cells for nearly 14 years and yet insight into how ETS1 func-

tions is completely lacking (Barton et al., 1998). It is not known

when ETS1 becomes essential and no target genes have been

identified in the NK cell lineage. Here, we demonstrated that

ETS1 functioned as early as the pre-NKP cell stage and that

ETS1 regulated abroad spectrumofNKcell genes including tran-

scription factors, NKRs, and signalingmolecules.We place ETS1

within a transcriptional network specifying the NK cell fate with

direct targets including Tbx21 (T-BET) and Idb2 (ID2). Ets1�/�

mNK cells failed to lyse NK cell targets and we demonstrated

decreased expression or function of multiple activating NKRs.

However, Ets1�/� mNK cells had characteristics of chronic acti-

vation including increased expression of inhibitory NKRs Ly49G2

and Ly49E, increased expression of the IL-15 responsive gene

Nfli3, encoding E4BP4, and increased Ikzf2, encoding HELIOS,

a transcription factor associated with NK cell hyperresponsive-

ness (Narni-Mancinelli et al., 2012). Moreover, Ets1�/�mNK cells

showed an augmented response to IL-15 in vitro. Our data

provide insight into the molecular mechanisms underlying the

requirement for ETS1 in NK cell development and function and

provide a foundation for building the regulatory networks that

control this important innate immune cell lineage.

RESULTS

ETS1 Functions at the Earliest Stages of NK CellDevelopmentEts1�/�mice have a reduced number of mNK cells (Barton et al.,

1998) but it is not known when or how ETS1 functions in the

NK cell lineage. To begin to address this issue, we rigorously

analyzed NK cell development in Ets1�/� mice. As expected, in

the BM and spleen of Ets1�/� mice, mNK cell numbers were

decreased by 90% and 80%, respectively, relative to wild-type

(WT) mNK cells (Figures 1A and 1C; Figure S1 available online;

Barton et al., 1998). There was a decrease in the frequency of

the most mature splenic mNK cells (CD27�CD11b+) but a

similar frequency of these cells expressed KLRG1 (Figure S1;

Chiossone et al., 2009). ETS1 was required for development of

approximately 50% of iNK cells but NKP numbers were similar

to WT (Figures 1A and 1C). However, Ets1�/� rNKPs (Lin�CD27+

CD244+FLT3�CD127+CD122+) were decreased by nearly 50%

and their precursor pre-NKPs (Lin�CD27+CD244+FLT3�

CD127+CD122�) were decreased by 20% (Figures 1B and 1D).

Ets1�/� mice also showed an approximate 50% decrease in

pre-pro-NKb cells (Lin�CD117�Sca1+CD127+FLT3�) (Figure S1;Carotta et al., 2011). These data reveal a function for ETS1 at the

earliest stages of NK cell development.

To determine whether the requirements for ETS1 were cell

autonomous, we created mixed BM chimeras where Ets1�/�

cells developed in competition with WT cells. Both WT

(CD45.2+) and Ets1�/� (CD45.2+) BM gave rise to hematopoietic

stem cells (HSCs), multipotent progenitors (MPPs), lymphoid

primed MPPs (LMPPs), and CLPs that competed well with WT

922 Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc.

(CD45.1) cells (Figures 1E and S1). However, there was

a >80% decline in NK lineage cells by the iNK cell stage (Figures

1E and S1). Therefore, there was a cell-intrinsic requirement for

ETS1 for NK cell development.

ETS1 Regulates the NK Cell TranscriptomeTo gain an understanding of how ETS1 functions in NK cells,

we performed a global analysis of gene expression in mNK cells

isolated from Rag2�/�Ets1+/+ and Rag2�/�Ets1�/� mice.

We used the Rag2�/� background to avoid contamination of

NK cells with activated T lymphocytes (Stewart et al., 2007).

We identified 216 genes that were decreased by 50%

or increased by 2-fold by the absence of ETS1 (Figure 2A;

Table S1). The distribution of these differentially expressed

genes was examined across WT multipotent progenitor cell

populations, pro-B cells, and CD4+ T cells. Nearly 50% of

ETS1-dependent genes were expressed in CD4+ T cells but

not in the other populations (Figure 2A). This finding suggested

that ETS1 regulates a gene program shared with T cells, which

also required ETS1 at multiple stages of development (Hollen-

horst et al., 2009). However, slightly >50% of ETS1-dependent

genes were unique to the NK cell lineage (Figure 2A).

To distinguish direct from possibly indirect targets of ETS1,

we considered a genome-wide analysis of ETS1 binding by

ChIP-sequencing (ChIP-seq). However, such an approach was

hampered by the low abundance of NK cells in vivo and because

ETS1 was downregulated in NK cell lines and primary NK cells

cultured in vitro, limiting the use of in vitro expanded cells (Fig-

ure S2). Given that a set of ETS1-dependent NK cell genes

was expressed in CD4+ T cells, we began by examining the

overlap between these genes and ETS1 binding in a human

CD4+ T cell line as determined by ChIP-seq (Hollenhorst et al.,

2007). Of the 216 ETS1-dependent NK cell genes, 167 had

human orthologs and 106 of these (63.5%) were associated

with ETS1 binding in the T cell line (Figure 2B). Therefore, 106

(49%) of the differentially expressed genes we identified are

high-probability ETS1 target genes (Table S2).

We next determined whether any unique binding motifs were

enriched among the sequences associated with ETS1 binding

at ETS1-dependent NK cell genes via MEME (Bailey et al.,

2009). An ETS binding motif was enriched that was nearly

identical to the motif previously associated with ETS1-specific

binding at distal (enhancer) sites (Figure 2C). These are sites

that failed to be bound by ELF1 and GABPa in the CD4+ T cell

line (Hollenhorst et al., 2009).

ETS1 Is Required for Proper Expression of MultipleNKRs, Signaling Molecules, and Transcription FactorsKEGG pathway analysis of the differentially expressed genes

revealed their involvement in NK cell cytotoxicity and in T cell

receptor-, chemokine-, and Janus kinase-signal transducer

and activator of transcription (Jak-STAT)-signaling pathways

(Figure 2D; Huang et al., 2009). A selected set of NK cell-

associated genes is shown in Figure 3A and among these Ltb,

Tbx21, Itk, Slamf6, Jak1, CD27, Lck, and Lair1 were bound by

ETS1 in the CD4+ T cell line. We demonstrated that ETS1 binds

directly to the Tbx21 andCd122 genes inmNK cells by ChIP (Fig-

ure 3D), confirming that these are direct targets. We confirmed

differential expression of Ncr1 (NKp46), Cd122, Idb2, Ltb,

B

5X

D2

21

DC

Lin

0.91 0.10

0.34

95.9

0.75

0.45

85.1

9.0

NK1.1

Ets1+/+ Ets1-/-A

E

CD122

CD127

CD27

44

2D

C3tl

F3tl

F

Ets1+/+ Ets1-/-

C

100

NKP iNK mNK sNK

102

104

106

103

105

107

101

re

bm

un

lle

cl

ato

T

*

****

8.3

2.8

83

13 5

88

10 2.5

D

100CLP pre-NKP rNKP

102

104

103

105

101

ns

ns

#

Ets1+/+

Ets1-/-

2.5

10.3

Ets1+/+

Ets1-/-

*

Ets1+/+Ets1-/-

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

CS

HP

PM

PP

ML

PLC

PK

NK

NiK

Nm

***

***

re

bm

un

lle

cl

ato

Tn

oitutit

sn

oc

ere

vital

eR

Figure 1. ETS1 Functions at the Earliest Stages of NK Cell Development

(A and B) Flow cytometry analysis of Ets1+/+ and Ets1�/� BM using (A) NK1.1 and DX5 to identify NKP (NK1.1�DX5�), iNK (NK1.1+DX5�), and mNK (NK1.1+DX5+)

cells in CD122+ Lineage (CD19�CD3�TCR-b�CD4�CD8�TER119�)� cells or (B) FLT3 and CD122 to identify CLP (FLT3+CD122�), pre-NKP (FLT3�CD122�), andrNKP (FLT3�CD122+) in Lin�CD27+CD244+CD127+ cells. The frequency of cells in the gated areas is indicated.

(C and D) The total number of (C) BMNKP, iNK, andmNK and splenic mNK (sNK) or (D) CLP, pre-NKP, and rNKP is presented. Each circle represents onemouse,

horizontal bar indicates average. Closed circles, Ets1+/+ mice; open circles, Ets1�/� mice. n = 4–7 for each genotype. *p < 0.05, **p < 0.01, ***p < 0.001 by

a Student’s t test, #p < 0.05 by a paired t test but not significant with a standard t test.

(E) Relative contribution of Ets1+/+ and Ets1�/� CD45.2+ cells in HSCs, MPPs, LMPPs, CLPs, NKP, iNK, and mNK from mixed BM chimeras with WT CD45.1+

competitors. Squares, Ets1+/+; diamonds, Ets1�/�. n = 9 except CLPs where n = 7.

See Figure S1 for analysis of splenic mNK cells, pre-pro-NK cells, and flow cytometery analysis of BM chimeras.

Immunity

ETS1 Targets in NK Cells

Lair1, and Tbx21 mRNA in Lin�CD122+DX5� pro-NK cells

(NKP+iNK) and mNK cells by quantitative polymerase chain

reaction (qPCR) (Figure 3B). Reduced expression of Ltb and

Tbx21 mRNA was also confirmed in Ets1�/� NKPs (Figure 3C).

Therefore, Ets1 deficiency results in decreased expression of

critical regulators of NK cell development including transcription

factors and NKRs.

ETS1 Regulates Idb2 in NK Cells and Their ProgenitorsThe Idb2 gene, which encodes ID2, is required for proper devel-

opment beyond the iNK cell stage (Boos et al., 2007) and its

expression was dependent on ETS1 in mNK cells (Figure 3A).

However, Idb2 was not bound by ETS1 in the CD4+ T cell line,

raising the possibility that Idb2 is not a direct target of ETS1.

To gain insight into the mechanisms controlling Idb2 expression,

we performed mutational analysis of Idb2 promoter-luciferase

reporters in an NKP cell line (Rodewald et al., 1992). We found

that Idb2 reporters containing at least 225 bp of sequence

upstream of the transcription start site (TSS) gave maximal lucif-

erase activity (Figure 4A). In contrast, truncation to 130 bp, which

removes a potential ETS binding site (EBS), decreased luciferase

activity by 80%, indicating that an important cis-regulatory

Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc. 923

49not annotated

167humanorthologs

61no binding

106bound

A B

C

stib 1

0

2

1 2 3 4 5 6 7 8 9 10 1112131415

p = 4.1e-125

CS

H+/+ -/-

PP

MP

PML

PLC B T

PM

G

mNK

KEGG Pathway Analysis of ETS1-dependent NK cell genes

CategoryNatural killer cell mediated cytotoxicity T cell receptor signaling pathway Pathways in cancer Chemokine signaling pathway B cell receptor signaling pathwayJak-STAT signaling pathway

Count282446301522

Percent1.91.73.22.11.01.5

p-value5.6E-7

3.6E-5

9.2E-5

1.7E-4

3.5E-3

7.7E-3

D

Figure 2. Identification of ETS1-Dependent Genes in NK Cells

(A) Microarray analysis of mRNA from Rag2�/�Ets1+/+ and Rag2�/�Ets1�/� mNK cells. Probe sets with differential expression in mNK cells are shown as is their

mRNA expression across multiple hematopoietic cell populations.

(B) Analysis of ETS1 binding to ETS1-dependent NK cell genes in a CD4+ T cell line by ChIP-seq. Genes associatedwith ETS1-bound regions were determined by

the closest TSS. Pie graphs show the percent of ETS1-dependent NK genes that had human orthologs (left) and the number of genes that were bound by ETS1

(right).

(C) MEME analysis of sequences bound by ETS1 in ETS1-dependent NK cell genes.

(D) KEGG analysis for functional pathways regulated by NK cell genes with a 1.53 threshold of differential expression.

See Table S1 for ETS1-dependent NK cell genes.

Immunity

ETS1 Targets in NK Cells

element was deleted (Figure 4A). Mutation of this EBS in the

context of the 670 bp or 225 bp promoter decreased luciferase

activity by 45% and 68%, respectively, demonstrating that an

ETS family protein was important for Idb2 transcription in this

NKP line (Figure 4A).

The putative EBS in the Idb2 promoter was identified previ-

ously as a target of the EWS-FLI1 and EWS-ERG fusion proteins

found in Ewing’s sarcoma (Nishimori et al., 2002). FLI1 and ERG

are members of a different clade of ETS family proteins and they

have a DNA binding preference distinct from ETS1; therefore, it

was not evident that ETS1 should regulate Idb2 through this

EBS. However, the Idb2 EBS fits a consensus motif bound by

multiple ETS family proteins including ETS1, ELF1, and GABPa

(Hollenhorst et al., 2007, 2009). Electrophoretic mobility shift

assays (EMSA) confirmed that both ETS1 and ELF1were present

in the NKP extract and were able to bind the Idb2 promoter EBS

whereas MEF1 (Lacorazza et al., 2002) was not present in the

bound complex (Figure 4B). Importantly, in mNK cells we

detected binding of ETS1 at the Idb2 promoter indicating that

ETS1 could directly regulate Idb2 (Figure 4C). Analysis of

mRNA at defined stages of NK cell differentiation revealed an

earlier onset of expression for Ets1 mRNA as compared to

Idb2 and Ets1 expression peaks in rNKPs just prior to the peak

of Idb2 at the iNK cell stage (Figure 4D). These data are consis-

tent with the hypothesis that Idb2 mRNA is dependent on ETS1

at the initiation of NK cell lineage specification. Although ID2 is

not essential for early NK cell development, its expression is

one of the first indications of NK cell lineage specification, and

decreased expression of ID2 in Ets1�/� mNK cells is predicted

924 Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc.

to have an impact on the differentiation and function of these

cells (Boos et al., 2007).

ETS1 Is Required for Proper Expression and Functionof Multiple Activating NKRsThe few mNK cells present in Ets1�/� mice are defective in their

ability to kill cells lacking MHC class 1 molecules (Barton et al.,

1998). However, the mechanism underlying this loss of cytolytic

function is not known. A failure to express activating receptors

or essential components of the signaling machinery activated

by these receptors would explain this defect. Our microarray

analysis revealed decreased expression of genes encoding

the activating receptor NKp46 and multiple Ly49 receptors

(encoded by the Klra genes) as well as proteins involved in signal

transduction by these receptors. We confirmed the decreased

expression of NKp46, Ly49D, and Ly49H on BM and splenic

mNK cells from Ets1�/� mice by flow cytometry (Figures 5A and

5B). These activating NKRs were also reduced on Ets1�/� mNK

cells isolated from mixed BM chimeras, indicating that this

alteration was cell intrinsic (Figure S3). Therefore, the failure of

Ets1�/�mNK cells to kill NK cell targets may be, in part, a conse-

quence of decreased expression of multiple activating NKRs.

In contrast to NKp46, Ly49D, and Ly49H, the activating recep-

tors NK1.1 and NKG2D were expressed appropriately on BM

and splenic mNK cells (Figure 5C), suggesting that these recep-

tors are available for NK cell target recognition. To determine

whether these receptors were functional, we tested the ability

of NK1.1 or NKG2D crosslinking to induce degranulation, as

measured by surface CD107a. As expected, crosslinking of

BA

0.00.51.01.52.02.53.03.5

0.0

0.5

1.0

1.5

Cd122

0.00.51.01.52.02.53.0 Idb2

Ltb

0.0

0.5

1.0

1.5

2.0 Ncr1

0.0

0.5

1.0

1.5

2.0

C

0.0

0.5

1.0

1.5 Lair1

Tbx21 Tbx21

0.000.250.500.751.001.25 Ltb

NKP NKP

0.00

0.05

0.10

0.15

0.20

0.25

Tbx21 Cd122 SP5 Hbb Ebf1

Rel

ativ

e m

RN

A e

xpre

ssio

n

D

DX5+DX5- DX5+DX5-

)%(tupnI

1stE

-/-

KNm

1stE

+/+

Borp

1stE

+/+

KNm

Itk*

Ets1*Idb2Tbx21*EomesIl2rβLtb*Jak1*Lck*Zap70*

CD27*Hcst

Crtam

Klre1

Klra16Klra3

Lair1*Klra17

Fcgr3*

Klrk1Ncr1

Slamf6*

0.000.250.500.751.001.25

Rel

ativ

e m

RN

A

exp

ress

ion

Figure 3. Deregulation of Multiple NK Cell Genes in the Absence of ETS1

(A) Heat maps for a subset of ETS1-dependent genes in Rag2�/�Ets1+/+ and Rag2�/�Ets1�/� mNK cells (duplicates combined). Asterisk indicates that ETS1 was

bound near this gene by ChIP-seq in CD4 T cells.

(B) qPCR analysis of Tbx21,Ncr1,Cd122, Idb2, Ltb, and Lair1mRNA in Ets1+/+ (black) andEts1�/� (white) pro-NK cells (Lin�CD122+DX5�) andmNK cells. At least

three independent experiments were performed.

(C) qPCR analysis of Tbx21 and Ltb mRNA in NKPs. Error bars indicate SD of triplicate measurements. Two independent experiments were performed.

(D) Anti-ETS1 ChIP for EBS in Tbx21 and Cd122. qPCR SP5, Hbb, and Ebf1 are at regions that lack an EBS.

See Table S2 for ETS1 binding sites in CD4+ T cells.

Immunity

ETS1 Targets in NK Cells

NK1.1 or NKG2D resulted in increased CD107a compared to

IgG on Ets1+/+ mNK cells (Figures 5D and S3). In contrast,

NKG2D stimulation of Ets1�/� mNK cells did not induce

CD107a above that observed with IgG, although CD107a was

higher on IgG-stimulated Ets1�/� mNK cells compared to

Ets+/+ mNK cells (Figures 5D and S3). Crosslinking of NK1.1 on

Ets1�/� mNK cells increased surface CD107a but the frequency

of CD107a+ cells was lower than observed on Ets1+/+ cells. In

contrast, CD107a was efficiently induced by phorbol myristate

acetate (PMA) + ionomycin in both Ets1+/+ and Ets1�/� mNK

cells (Figures 5D and S3). These observations indicated that

Ets1�/� mNK cells were intrinsically defective in their ability to

degranulate in response to activating NKR stimulation. In

contrast, interferon-g (IFN-g) production was not as severely

affected, although crosslinking of NKG2D did not lead to a signif-

icant accumulation of IFN-g at this time point in Ets1+/+ or

Ets1�/� mNK cells (Figures 5E and S3). The reduced expression

of many activating NKRs and the impaired exocytosis function in

Ets1�/�mNK cells is sufficient to explain the decreased cytolytic

function of these cells.

Ets1–/– NK Cells Have Characteristics of ChronicCytokine StimulationIn addition to the ETS1-dependent genes, we noted that multiple

genes associated with NK cell activation were increased in

Ets1�/�mNK cells.Gzmb andPrf1mRNA, encoding the cytolytic

proteins GZMB (GranzymeB) and PRF (Perforin), were increased

as were mRNAs encoding the serine protease inhibitors

SERPINB6A and SERPINB9B (Figure 6A). We confirmed that

GzmbmRNA was increased in Ets1�/� mNK cells by qPCR (Fig-

ure 6B). Nfil3 mRNA, encoding a cytokine-responsive transcrip-

tion factor (Gascoyne et al., 2009; Kamizono et al., 2009), was

also increased in Ets1�/� mNK cells as well as in pro-NK cells

(Figures 6A and 6C). Interestingly, Ikzf2 mRNA, which encodes

HELIOS, whose increased expression contributes to hyperres-

ponsiveness in Noe mice (NKp46-deficient) (Narni-Mancinelli

et al., 2012), was increased in Ets1�/� mNK cells (Figures 6A

and 6D). Ets1�/� mNK cells isolated from mixed BM chimeras

also showed increased granularity and increased expression of

the activation marker CD69 as measured by flow cytometry (Fig-

ure 6E). Taken together, these data indicate that Ets1�/� mNK

cells are in an activated state.

Our observation that at least two IL-15-regulated genes are

increased in Ets1�/� mNK cells led us to question whether these

cells have other characteristics of chronic cytokine stimulation.

Chronic IL-15 stimulation leads to increased expression of the

inhibitory NKRs Ly49G2 and Ly49E (Ly49E is normally not

expressed on adult mNK cells) (Barao et al., 2011; Elpek et al.,

2010; Fraser et al., 2002). We found an increased frequency of

BM mNK cells expressing Ly49G2 and Ly49E but not Ly49A in

Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc. 925

A B

C D

Figure 4. Idb2 Is Regulated by ETS1 in NK Cells and Their Precursors

(A) Promoter-luciferase reporter assay via regions 50 of the Idb2 TSS, including the endogenous EBS site (GGA) or a mutated (m) EBS (GGT). One representative

experiment from at least four is shown.

(B) EMSA of proteins in PTL cell extracts binding to the Idb2 EBS in the presence (+) or absence (�) of cold competitor or a-ETS1, a-MEF1, or a-ELF1.

(C) ChIP was performed on sorted splenic mNK cells via a-ETS1 followed by qPCR with primers flanking the EBS in Idb2 or for sequences lacking an EBS (SP5,

Hbb, and Ebf1).

(D) qPCR analysis for Idb2, Ets1, Elf1, and Gabpa mRNA relative to Hprt mRNA in sorted NK cell and NK cell progenitor populations.

Error bars represent standard error of triplicate measurements and one of at least two to three replicate experiments is shown. See Figure S2 for Ets1 mRNA

expression in NK cell lines and activated NK cells.

Immunity

ETS1 Targets in NK Cells

Ets1�/� as compared toWTmice (Figures 6F and 6G). The inten-

sity of Ly49G2 and Ly49E staining was also increased on Ets1�/�

mNK cells in both the BM and spleen (Figure 6F). These alter-

ations in inhibitory NKR expression were cell intrinsic as shown

by the fact that they were observed on Ets1�/� mNK cells in

mixed BM chimeras (Figure S4). Taken together with the

increased expression of Nfil3, Gzmb, and Prf1 mRNA and

CD69, our findings indicate that Ets1�/� mNK cells resembled

NK cells chronically stimulated by IL-15.

Ets1–/– mNK Cells Are Hyperresponsive to IL-15Given the phenotype of Ets1�/� mNK cells, we questioned how

Ets1�/� NK cells would respond to cytokines. Single-cell anal-

ysis of Ets1�/� and Ets1+/+ DX5� and DX5+ NK cells cultured

in vitro revealed that a comparable frequency of cells could

form colonies in response to IL-2 (Figure 7A). However, Ets1�/�

colonies were larger and the cells were more granular than their

WT counterparts (Figure 7B). To determine whether Ets1�/�

mNK cells were more responsive to IL-15 than WT mNK cells,

we titrated IL-15 in cultures of Ets1�/� and Ets1+/+ mNK cells

and measured induction of GZMB and proliferation by means

of BrdU incorporation. Within 24 hr, Ets1�/� mNK cells showed

an increase in GZMB and BrdU incorporation compared to

Ets1+/+ mNK cells at all concentrations of IL-15 (Figure 7C).

The augmented response of Ets1�/� mNK cells was particularly

evident when IL-15 was present at 50 ng/ml, the concentration

commonly used for expansion of NK cells in vitro (Figures 7C

and 7D). In addition, whereas Ets1+/+ mNK cells showed little

induction of GZMB or BrdU incorporation when cultured in

926 Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc.

1 ng/ml IL-15, Ets1�/� mNK cells showed a 4-fold higher

response (Figures 7C and 7D). These experiments were

performed with mNK cells isolated from mixed BM chimeras,

allowing us to exclude in vivo homeostatic proliferation as a

factor predisposing Ets1�/� mNK cells to an increased cytokine

response. Taken together with the data in Figure S2, showing

thatEts1mRNAdecreasedwhenNKcellswere stimulated in vivo

for 2 days with IL-2 or 1 day with poly(I:C), our findings suggest

a role for ETS1 in limiting NK cell activation in response to

cytokines. Our data support a model in which ETS1 controls

expression of a broad spectrum of NK cell genes including

transcription factors, NKRs, and signaling molecules at the

earliest stages of NK cell development, allowing for appropriate

NK cell activation in pathogenic conditions.

DISCUSSION

In this study, we have revealed at least three major functions for

ETS1 in NK cells. First, ETS1 directly regulates expression of

Idb2 and Tbx21, whose protein products ID2 and T-BET

comprise a part of the transcriptional circuitry necessary for

NK cell differentiation. Second, ETS1 is required for expression

and function of multiple activating NKRs that are necessary for

induction of NK cell-mediated cytolysis. This functional deficit

was revealed primarily as a failure of degranulation rather than

IFN-g production. Thus, the inability of Ets1�/� NK cells to kill

NK cell targets can be explained by their decreased ability

to degranulate in response to activating NKR ligands. Third,

ETS1 sets the threshold for responsiveness to cytokine, and

Ly49D

A 6 44

3056

1453

2150

B

D

BM

Spleen

NKp46

7298

)%(

xa

M

C E

NKG2D

8687

96.5 9097

NK1.1

)%(

xa

M

BM Spleen

9998

100100

-N

FIγ+

)%(

0

50

75

100

25

I/A

MP

1.1

KN

D2G

KN

GgI

a7

01

DC

+)

%(

0

50

75

100

25

**

*

*

4795

Ly49H

I/A

MP

1.1

KN

D2G

KN

GgI

)%(

eviti

so

P

0

25

50

75

100*** ***

BM spleen

NKp46

BM spleen0

20

40

60

80

100

* **

Ly49D

0

20

40

60

80

100

*

**

BM spleen

Ly49H

Figure 5. Ets1–/– mNK Cells Had Decreased Activating NKR Expression and Activating NKR-Triggered Degranulation

(A) Flow cytometry analysis for the activating NKRs NKp46, Ly49D, and Ly49H in BM and spleen of Ets1+/+ (dark line) and Ets1�/� (gray line) mice. Shaded

histogram shows isotype control.

(B) Summary of the percent of mNK cells expressing NKp46, Ly49D, and Ly49H. Ets1+/+ (filled), Ets1�/� (open); horizontal bar indicates average. *p < 0.05,

**p < 0.01, ***p < 0.001; n = 4; ns, p > 0.05.

(C) Flow cytometry analysis for the activating NKRs NK1.1 and NKG2D on Ets1+/+ and Ets1�/� mNK cells. Histograms are as indicated in (A).

(D and E) Summary of percent (D) CD107a+ or (E) IFN-g+ after a 5 hr stimulation with PMA plus ionomycin, a-NK1.1, a-NKG2D, or IgG.

See Figure S2 for activating NKR expression in mixed BM chimeras and representative flow cytometery staining for CD107a and IFN-g.

Immunity

ETS1 Targets in NK Cells

probably other external stimuli, which may prevent expansion

and activation in nonpathogenic conditions. In the absence of

ETS1, mNK cells had hallmarks of chronic IL-15 stimulation

and they had a heightened response to a suboptimal dose of

IL-15. Taken together, our data provide insight into the functions

of this critical transcriptional regulator in NK cells and provide

a foundation on which to build the regulatory circuits driving

NK cell development and function.

The absence of ETS1 resulted in alterations in NK cell progen-

itors at the earliest stages of development, placing ETS1, along

with ID2, TOX1, and E4BP4 (Aliahmad et al., 2010; Kamizono

et al., 2009), as the earliest acting transcriptional regulators

identified in NK cells. We showed that Ets1 mRNA expression

precedes Idb2 mRNA, which was previously the earliest known

marker of NK cell differentiation. Therefore, ETS1 is positioned

to play a key role in NK cell lineage specification. In order for

ETS1 to function in NK cell specification, its expression should

precede NK cell lineage restriction. We previously found that

Ets1 was among the genes primed by E2A in LMPPs (Dias

et al., 2008). During specification of the NK cell lineage, E2A

function is antagonized by ID2 and ID3 (Boos et al., 2007) and

yet Ets1 mRNA increases. Therefore, the transcription factors

controlling Ets1 must evolve as the NK cell fate is specified.

This shift in transcriptional control could occur as a consequence

of the induction of NK cell-associated transcription factors such

as T-BET, or alternatively, ETS1 may autoregulate its own

expression (Hollenhorst et al., 2009; Seth and Papas, 1990).

There are multiple ETS1 binding events near the Ets1 gene in

CD4+ T cells, indicating that ETS1 may control its own expres-

sion (Hollenhorst et al., 2009). Based on these considerations

and our current knowledge of transcriptional networks in B and

T cell development (Nutt and Kee, 2007), we hypothesize that

ETS1 functions in a transcriptional network with re-enforcing

feedback loops to control NK cell lineage specification.

It is important to note that although NKP cell numbers in

Ets1�/� mice were indistinguishable from Ets1+/+ mice, more

highly enriched progenitor populations revealed a requirement

for ETS1. However, identification of pre-NKP, rNKP, and pre-

pro-NK cells relies on expression of surface proteins reported

to be ETS1 targets, raising the possibility that altered gene

expression rather than altered development is responsible for

this decrease. CD127 (IL-7Ra) is critical for identification of these

cells and was reported to be an ETS1 target in peripheral CD8+

T cells (Grenningloh et al., 2011). Importantly, we did not find

decreased CD127 on Ets1�/�CLPs via multiple different staining

strategies or within the larger Lin�CD244+CD27+ population

containing NK cell progenitors. Therefore, ETS1 is not essential

for CD127 expression in multipotent progenitors of NK cells.

Nonetheless, if CD127 were an ETS1 target in the earliest

NK cell progenitors, this would further support our conclusion

that ETS1 controls gene expression at this early stage of NK

cell development.

We have defined a minimal set of high-probability ETS1 target

genes by correlating ETS1-dependent gene expression with

Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc. 927

B

A

C

0

1

2

3

4

5

6

7 Nfil3

+/+ +/+-/- -/-

DX5- DX5+

BM Spleen

7744

SSC

CD69

26 8

D

E

BM spleen0

20

40

60

80

100

F

G

BM spleen BM spleen

Ly49ELy49G2Ly49A

)%(

eviti

so

P

**

**

GzmaSerpinb9bGzmb*Klra5Serpinb6aKlrg1

Prf1*

Klra20GzmK

Nfil3

Ikzf2

1stE

-/-

KN

m

1stE

+/+

Borp

1stE

+/+

KN

m

NSNS

NS

)%(

xa

M

7111

38 7

7745

4225

3616

Ly49A

5541

Ly49G2 Ly49E

BM

spleen

)%(

xa

M

0

2

4

6

8

10 Gzmb

Ets1+/+ Ets1-/-

Re

lative

mR

NA

e

xp

ressio

n

Re

lative

mR

NA

e

xp

ressio

n

0

1

2

3

4

5

6

+/+ +/+-/- -/-

CD11b- CD11b+

Ikzf2

Re

lative

mR

NA

e

xp

ressio

n

Figure 6. Ets1–/– mNK Cells Have Characteristics of Chronic Cytokine Stimulation

(A) Heat maps for a subset of ETS1-repressed genes in Rag2�/�Ets1+/+ and Rag2�/�Ets1�/�mNK cells. Asterisk indicates that ETS1 was bound near this gene in

CD4+ T cells as determined by ChIP-seq.

(B–D) qPCR analysis for (B)Gzmb, (C) Nfil3, and (D) Ikzf2mRNA in Ets1+/+ (black) and Ets1�/� (white) mNK cells. Expression is plotted relative to transcripts from

Hprt. Error bars represent standard error of triplicate measurements and one of at least two to three replicate experiments are shown.

(E and F) Flow cytometry analysis for (E) SSC and CD69 or (F) the inhibitory NKRs Ly49A, Ly49G2, and Ly49E on BM and splenic mNK cells from Ets1+/+ (black)

and Ets1�/� (gray).

(G) Summary of the percent of mNK cells expressing Ly49A, Ly49G2, and Ly49E. Ets1+/+ (filled), Ets1�/� (open), horizontal bar indicates average. *p < 0.05,

**p < 0.01, ***p < 0.001; n = 4; ns, p > 0.05.

Cells in (E) were isolated fromCD45.1 (WT):CD45.2 (WT or Ets1�/�) chimeric mice. See Figure S3 for Ly49A, Ly49G2, and Ly49E staining onmNK cells frommixed

BM chimeras.

Immunity

ETS1 Targets in NK Cells

ETS1 DNA binding events in a CD4+ T cell line. This is a minimal

set because not all ETS1-dependent NK cell genes are

expressed in CD4+ T cells; therefore, ETS1 binding could not

be assessed at all NK cell targets. Nonetheless, 49% of ETS1-

dependent NK cell genes were bound by ETS1 in CD4+ T cells,

probably reflecting overlapping functions for ETS1 in these cell

types. Indeed, ETS1 regulates genes involved in T cell activation

in CD4+ T cells (Hollenhorst et al., 2007, 2009), and we identified

‘‘T cell receptor signaling’’ in addition to ‘‘NK cell cytotoxicity’’

as pathways associated with ETS1-dependent genes in

NK cells. Consistent with this common function, we found that

ETS1-regulated NK genes had ETS binding motifs nearly iden-

tical to an ETS1-specific motif reported in T cell studies (Hollen-

horst et al., 2007, 2009). However, although this site was not

associated with ELF1 binding in CD4+ T cells, it shows similari-

ties to ETS binding motifs reported for other ETS family proteins

(Hollenhorst et al., 2009; Wei et al., 2010). We speculate that

other ETS factors may occupy some of these sites in the

928 Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc.

absence of ETS1, providing an explanation for the only partial

decrease in many putative ETS1 target genes in Ets1�/�

mNK cells. In addition, some genes, such as Idb2, have an

ETS binding motif that can be bound by ELF1 and other ETS

family proteins (Hollenhorst et al., 2009), and we found ELF1

binding to this site in an NKP line. Therefore, ETS family proteins

probably play a more crucial role in NK cell development than

revealed by the Ets1�/� mouse.

An emerging question is why ETS1 induces some genes

specifically in NK cells but not in T cells and B cells, where it is

also expressed (Eyquem et al., 2004a, 2004b). The unique chro-

matin landscape present in each of these cells undoubtedly

plays an important role. However, there are also factors that

may influence ETS1 target gene selection such as ETS1 concen-

tration and DNA sequence affinity (Hollenhorst et al., 2007,

2009), posttranslational modification (Cowley and Graves,

2000; Lee et al., 2008), DNA methylation (Gaston and Fried,

1995; Yokomori et al., 1995), and cooperative interactions with

D

A

CS

S

FSC

Ets1+/+ Ets1-/-

0

25

50

75

100B

0

30

40

50

60

10

20

1 10 100

***

**

**

Ets1+/+

Ets1-/-

Be

my

zn

arG

+U

drB

+)

%(

IL-15 (ng/mL)

C 1 ng/ml

2.6

0.216

5.84

2.78 6.3

16.2 48.3

21

50

41.8

Ets1+/+

Ets1-/-

BrdU

Be

my

zn

arG

100 ng/ml 50 ng/ml 10 ng/ml 5 ng/ml

NS NS

)%(

evi

sn

op

ser

2-LI

DX5- DX5+

*

Figure 7. Ets1–/– mNK Cells Are Hyperre-

sponsive to IL-15

(A) Percent of single Ets1+/+ (filled) and Ets1�/�

(open) pro-NK and mNK cells giving rise to visible

colonies after 10 days in cultures supplemented

with IL-2.

(B) FSC versus SSC on typical mNK cell progeny

from (A).

(C) Flow cytometry analysis for GZMB and BrdU

incorporation in splenic mNK cells 24 hr after

initiation of culture in varying concentrations of

IL-15. The percent of Granzyme B+ BrdU+ cells is

indicated. One of three experiments is shown.

(D) Average ± SD of percent GZMB+ BrdU+ cells

from three independent experiments. *p < 0.05,

**p < 0.01, ***p < 0.001. The splenic mNK cells

were isolated from BM chimeras.

Immunity

ETS1 Targets in NK Cells

neighboring transcription factors (Cowley and Graves, 2000;

Pufall et al., 2005). In T cells, ETS1 and RUNX1 bind coopera-

tively at the Tcra enhancer, and in B cells, ETS1 is recruited to

the Cd79b promoter via association with PAX5 (Fitzsimmons

et al., 2009; Hollenhorst et al., 2009). Future studies analyzing

cis-regulatory elements at shared and NK cell-specific ETS1

targets could provide insight into the mechanisms of lineage-

specific gene expression by ETS proteins. Our study provides

a critical first step in this analysis by identifying potential shared

and NK cell-specific ETS1 target genes.

Multiple observations lead us to conclude that ETS1 limits

the NK cell response to cytokines. In addition to having an acti-

vated phenotype, Ets1�/�NK cells showed elevated Nfil3mRNA

and Nfil3 is regulated downstream of IL-15 and is sufficient to

rescue NK cell differentiation in Il15ra�/� NKPs cultured in vitro

(Gascoyne et al., 2009; Kamizono et al., 2009). Ly49G2 and

Ly49E were both highly expressed on Ets1�/� compared to

Ets1+/+mNK cells and their expression is upregulated by chronic

cytokine stimulation (Barao et al., 2011; Elpek et al., 2010; Fraser

Immunity 36, 921–9

et al., 2002). Moreover, in in vitro

cytokine-dependent cultures, Ets1�/�

NK cells cloned well with larger colony

sizes compared to Ets1+/+ NK cells and

both in vitro and in vivo, and under

competitive reconstitution conditions,

Ets1�/� NK cells had an activated pheno-

type. Most importantly, however, Ets1�/�

mNK cells incorporated more BrdU and

induced GZMB more rapidly than WT

mNK cells at all concentrations of IL-15.

The mechanism underlying the height-

ened activation in Ets1�/� NK cells prob-

ably involves the deregulation of multiple

genes encoding signaling proteins and

transcription factors (for example, Itk,

Jak1, Lck, Ppp1r3b, Ptpn3, Egr3, Tbx21,

Ikzf2, or Nfat1c). In addition, the

decreased expression and function of

activating NKRsmay change the ‘‘tuning’’

of the intracellular signaling milieu, result-

ing in an altered response to multiple cell

surface receptors (Joncker et al., 2009; Joncker and Raulet,

2008). Indeed, mice lacking NKp46 are hyperresponsive to

MCMV and the NK cell target YAC-1 and their increased respon-

siveness requires HELIOS (Narni-Mancinelli et al., 2012), which

is increased in Ets1�/� NK cells and therefore probably contrib-

utes to the hyperresponsive phenotype. However, in Ets1�/�

mNK cells the compounded defects in activating receptor

expression and degranulation probably limited NK cell-mediated

lysis. The hypothesis that ETS1 influences lymphocyte activation

potential is consistent with a previously reported role for ETS1 in

the B cell response to TLR9 (John et al., 2008; Wang et al., 2005).

Moreover, ETS1 influences cytokine responsiveness and activa-

tion in T lymphocytes (Clements et al., 2006; Grenningloh et al.,

2005; Higuchi et al., 2007; Moisan et al., 2007; Russell and Gar-

rett-Sinha, 2010), indicating that targets of ETS1 contribute to

the signaling milieu in adaptive lymphocytes. The barrier to

NK cell activation imposed by ETS1 may reflect involvement of

ETS1 targets in the unique mechanisms controlling NK cell

activation because Ets1�/� T cells fail to become activated after

32, June 29, 2012 ª2012 Elsevier Inc. 929

Immunity

ETS1 Targets in NK Cells

stimulation (Muthusamy et al., 1995). Importantly, although ETS1

deficiency phenocopies many aspects of chronic cytokine stim-

ulation, Ets1�/� mice do not develop leukemia as was observed

in IL-15 transgenic mice (Fehniger et al., 2001). Leukemogenesis

may be limited by the arrested differentiation that accompanies

ETS1 deficiency at the earliest stages of NK cell development.

EXPERIMENTAL PROCEDURES

Mice

C57BL/6 or 129/SvJ Ets1�/� mice (Barton et al., 1998) were housed at the

University of Chicago Animal Resources Center in accordance with the guide-

lines of the University of Chicago Institutional Animal Care andUseCommittee.

129/SvJ Rag2�/� mice were purchased from Jackson Labs.

Quantitative Real-Time PCR

RNA was purified with the RNeasy micro kit (QIAGEN), reverse transcribed

with SuperScriptIII (Invitrogen), and primed with random hexamers as

described (Boos et al., 2007). Expression is reported as DCT relative to Hprt

mRNA. qPCR primer sequences are available upon request.

Luciferase Reporter Assays

The 670 bp and 225 bp Idb2 promoter fragments were PCR amplified from

genomic DNA and cloned into pGL3. The 130 bp Idb2 fragment was digested

from pGL3-225-Idb2p with SacI and XhoI and cloned into pGL3. PTL cells

were transfected with DEAE-dextran with 8 mg of pGL3 constructs and

0.5 mg of pRL-CMV as an internal control (Kee and Murre, 1998). Lysates

were prepared 48 hr after transfection and assayed with the Dual-Glo Lucif-

erase kit (Promega).

Electrophoretic Mobility Shift Analysis

Nuclear extracts were prepared and EMSA performed as described (Kee and

Murre, 1998). The Idb2 EBS sequence was 50-GGTATTGGCTGCGAACGCG

GAAGAACC-30 and the Idb2 EBS mutant sequence was 50-GGTATTGGCT

GCGAACGCGGTAGAACC-30. Antibodies to ETS1, ELF1, and MEF1 were

purchased from Santa Cruz Biotechnology.

Cell Culture

Cells lines were maintained in Opti-MEM or RPMI-1640 supplemented with

10% FBS, 80 mM 2-mercaptoethanol, 100 units/ml penicillin, 100 mg/ml strep-

tomycin, and 29.2 mg/ml glutamine. Primary NKPs were grown on OP9

stromal cells (10,000 OP9 cells/well of a 96-well plate) supplemented with

IL-2 (1,000 IU/ml, NIH Reagents program), CD117 (1:1,000 dilution from

CHO-MGF cells), and FLT3 (10 ng/ml). Primary mNK cells and NK cell lines

were cultured in media supplemented with IL-2. The PTL line was generated

by H.-R. Rodewald by in vitro culture of fetal thymus-derived FcRgII or FcRgIII+

NK and T cell progenitors (Rodewald et al., 1992) and was adapted for growth

in Opti-MEM. The KY1, KY2, and NKCRq cell lines were provided by

W. Yokoyama and C. Roth (Caraux et al., 2006; Karlhofer et al., 1995).

IL-15 responsiveness was determined by culturing 1,500–3,000 flow cytom-

etry-sorted splenicmNK cells, isolated from chimericmice, inmultiple concen-

trations of recombinant mIL-15. At t = 24 hr, 1 mM BrdU was added for 45 min

prior to intracellular staining for BZMB and BrdU.

Flow Cytometry

Cells were stained with fluorochrome- or biotin-labeled antibodies for 20 min

on ice. The following antibodies conjugated to FITC, PE, PerCP-Cy5.5,

PerCP-ef710, PeCy7, APC, APC-ef780, Pacific Blue, or Brilliant Violet 421

were purchased from eBioscience, BD PharMingen, or Biolegend: CD45.2

(104), CD45.1 (A20), CD19 (1D3), B220 (Ra3-6B2), CD3ε (145-2C11), CD4

(RM4-5), CD8a (Ly-2), TCR-b (H57-597), TCR-gd (UC7-13D5), CD11b (M1/

70), Ter-119 (Ly-76), Gr-1 (RB6-8C5), IL-7Ra (A7R34), CD117 (2B8), Sca1

(E13-161.7), FLT3 (A2F10), CD122 (TM-b1), NK1.1 (PK136), CD49b (DX5),

CD94 (18d3), NKG2ACE (20d5), NKp46 (29A1.4), Klrg1 (CF1), CD69

(H1.2F3), Ly49D (4E5), Ly49H (3D10), Ly49G2 (ebio4D11), Ly49A (A1), LyEF

(CM4), Ly49CIFH (14B11), Ly-6d (49-H4), CD27 (LG.7F9), Klrg1 (2F1), BrdU

(Bu20a or PRB-1), IFN-g (XMG1.2), GZMB (NGZB or 16G6), and 2B4

930 Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc.

(ebio244F4). Propidium idodide was used to exclude dead cells. Cells were

analyzed on a FACS Canto, LSRII, or Fortessa or sorted with a FACS ARIAII

(Becton Dickenson).

Lineage cocktail for HSC, MPP, LMPP, CLPs: B220, CD3ε, CD4, CD8,

NK1.1, Ter119, CD11b, Ly-6G; for NK cells: CD19, CD3ε, CD4, CD8,

Ter119; and for CLP, pre-NKP, rNKP populations: CD19, CD11b, CD3ε,

Ly6d, and NK1.1.

NK Cell Activation

In vivo activation of NK cells was accomplished by an intravenous (i.v.)

injection of 1,000,000 IU IL-2 at t = 0 and t = 24 hr. At t = 48 hr, NK cells

were isolated by flow cytometry. Alternatively, 100 mg of poly(I:C) was injected

intraperitoneally (i.p.) followed by isolation of NK cells at t = 24 hr. For a-NK1.1,

a-NKG2D, or IgG stimulation of NK cells, mNK cells were cultured on antibody-

coated plates in 1,000 IU/ml IL-2 plus Golgi Plug (Becton Dickenson) for

5 hr. Fluorochrome-labeled a-CD107a (1D4B) or isotype control antibodies

(5 mg/ml) were added at t = 0. Cells were stained with DX5 and NKp46 prior

to flow cytometry analysis. Alternatively, cells were cultured with PMA

(100 ng/ml) and ionomycin (2 mM).

BM Chimeras

Chimeric mice were established by retrorbital injection of 2.5 3 106 bone

marrow cells from CD45.1 WT mice and 2.5 3 106 bone marrow cells from

CD45.2 Ets1+/+ or CD45.2 Ets1�/� mice into 8-week-old lethally irradiated

(1,000 rad) CD45.1 and CD45.2 or CD45.1 recipients. Recipients were

maintained on Bactrim and analyzed 8 weeks posttransplantation.

Microarray Analysis

cDNA prepared from 10,000 Lin�CD122+DX5+ cells was used to probe

Affymetrix MOE 430_2 arrays as previously described (Dias et al., 2008).

Raw array data were normalized with RMAexpress (http://rmaexpress.

bmbolstad.com/) and analyzed by dChip (http://www.biostat.harvard.edu/

complab/dchip/). Probe set annotation was obtained from Affymetrix.

MEME

Multiple Em for Motif Elicitation was used to identify repeated motifs in ETS1

ChIP-Seq sequences from the CD4+ T cell line Jurkat. MEME was run with

default setting, except that the minimum motif length was set to 8 and

maximum to 15. Only the motifs with the lowest E-value are reported.

Chromatin Immunoprecipitation

Primary mouse mNK cells were crosslinked in 1% formaldahyde and sheared

on a Branson sonicator. Protein-DNA complexes were immunoprecipitated

with polyclonal anti-ETS1 (C-20) or IgG (Santa Cruz). For each sample,

1,000,000 cell equivalents of chromatin were incubated with 5 mg of antibody.

Protein G-coupled magnetic beads were used to isolate immune complexes.

Crosslinks were reversed by heating at 65�C followed by proteinase K treat-

ment. DNA was purified with PCR spin columns (QIAGEN) and amplified

by qPCR with primers specific for the Tbx21, Cd122, or Idb2 EBS or irrelevant

genomic regions (SP5, Hbb, Ebf1). ChIP sequencing was described in Hollen-

horst et al. (2009).

ACCESSION NUMBERS

The microarray data are available in the Gene Expression Omnibus (GEO)

database (http://www.ncbi.nlm.nih.gov/gds) under the accession number

GSE37301.

SUPPLEMENTAL INFORMATION

Supplemental Information includes four figures and two tables and can be

found with this article online at doi:10.1016/j.immuni.2012.04.006.

ACKNOWLEDGMENTS

We thank E. Svensson and K. Barton for providing Ets1�/�mice. We are grate-

ful to V. Kumar for helpful discussions and comments on this manuscript. This

Immunity

ETS1 Targets in NK Cells

work was supported by the National Institutes of Health (CA099978 to B.L.K.,

CA099978-S to K.R., GM38663 to B.J.G., and CA42014 to the Huntsman

Cancer Institute for support of core facilities). B.L.K. is supported by a scholar

award from the Leukemia and Lymphoma Society. K.R. is supported by the

Interdisciplinary Training Program in Immunology (T32AI007090).

Received: October 26, 2011

Revised: March 2, 2012

Accepted: April 19, 2012

Published online: May 17, 2012

REFERENCES

Aliahmad, P., de la Torre, B., and Kaye, J. (2010). Shared dependence on the

DNA-binding factor TOX for the development of lymphoid tissue-inducer cell

and NK cell lineages. Nat. Immunol. 11, 945–952.

Bailey, T.L., Boden, M., Buske, F.A., Frith, M., Grant, C.E., Clementi, L., Ren,

J., Li, W.W., and Noble, W.S. (2009). MEME SUITE: tools for motif discovery

and searching. Nucleic Acids Res. 37 (Web Server issue), W202–W208.

Barao, I., Alvarez, M., Ames, E., Orr, M.T., Stefanski, H.E., Blazar, B.R., Lanier,

L.L., Anderson, S.K., Redelman, D., andMurphy, W.J. (2011). Mouse Ly49G2+

NK cells dominate early responses during both immune reconstitution and

activation independently of MHC. Blood 117, 7032–7041.

Barton, K., Muthusamy, N., Fischer, C., Ting, C.N., Walunas, T.L., Lanier, L.L.,

and Leiden, J.M. (1998). The Ets-1 transcription factor is required for the devel-

opment of natural killer cells in mice. Immunity 9, 555–563.

Boos, M.D., Yokota, Y., Eberl, G., and Kee, B.L. (2007). Mature natural killer

cell and lymphoid tissue-inducing cell development requires Id2-mediated

suppression of E protein activity. J. Exp. Med. 204, 1119–1130.

Caraux, A., Lu, Q., Fernandez, N., Riou, S., Di Santo, J.P., Raulet, D.H., Lemke,

G., and Roth, C. (2006). Natural killer cell differentiation driven by Tyro3

receptor tyrosine kinases. Nat. Immunol. 7, 747–754.

Carotta, S., Pang, S.H., Nutt, S.L., and Belz, G.T. (2011). Identification of the

earliest NK-cell precursor in the mouse BM. Blood 117, 5449–5452.

Chiossone, L., Chaix, J., Fuseri, N., Roth, C., Vivier, E., and Walzer, T. (2009).

Maturation of mouse NK cells is a 4-stage developmental program. Blood 113,

5488–5496.

Clements, J.L., John, S.A., and Garrett-Sinha, L.A. (2006). Impaired generation

of CD8+ thymocytes in Ets-1-deficient mice. J. Immunol. 177, 905–912.

Cowley, D.O., and Graves, B.J. (2000). Phosphorylation represses Ets-1 DNA

binding by reinforcing autoinhibition. Genes Dev. 14, 366–376.

Dias, S., Mansson, R., Gurbuxani, S., Sigvardsson, M., and Kee, B.L. (2008).

E2A proteins promote development of lymphoid-primed multipotent progeni-

tors. Immunity 29, 217–227.

Elpek, K.G., Rubinstein, M.P., Bellemare-Pelletier, A., Goldrath, A.W., and

Turley, S.J. (2010). Mature natural killer cells with phenotypic and functional

alterations accumulate upon sustained stimulation with IL-15/IL-15Ralpha

complexes. Proc. Natl. Acad. Sci. USA 107, 21647–21652.

Eyquem, S., Chemin, K., Fasseu, M., and Bories, J.C. (2004a). The Ets-1 tran-

scription factor is required for complete pre-T cell receptor function and allelic

exclusion at the T cell receptor beta locus. Proc. Natl. Acad. Sci. USA 101,

15712–15717.

Eyquem, S., Chemin, K., Fasseu, M., Chopin, M., Sigaux, F., Cumano, A., and

Bories, J.C. (2004b). The development of early and mature B cells is impaired

in mice deficient for the Ets-1 transcription factor. Eur. J. Immunol. 34,

3187–3196.

Fathman, J.W., Bhattacharya, D., Inlay, M.A., Seita, J., Karsunky, H., and

Weissman, I.L. (2011). Identification of the earliest natural killer cell-committed

progenitor in murine bone marrow. Blood 118, 5439–5447.

Fehniger, T.A., Suzuki, K., Ponnappan, A., VanDeusen, J.B., Cooper, M.A.,

Florea, S.M., Freud, A.G., Robinson, M.L., Durbin, J., and Caligiuri, M.A.

(2001). Fatal leukemia in interleukin 15 transgenic mice follows early expan-

sions in natural killer and memory phenotype CD8+ T cells. J. Exp. Med.

193, 219–231.

Fitzsimmons, D., Lukin, K., Lutz, R., Garvie, C.W.,Wolberger, C., and Hagman,

J. (2009). Highly cooperative recruitment of Ets-1 and release of autoinhibition

by Pax5. J. Mol. Biol. 392, 452–464.

Fraser, K.P., Gays, F., Robinson, J.H., van Beneden, K., Leclercq, G., Vance,

R.E., Raulet, D.H., and Brooks, C.G. (2002). NK cells developing in vitro from

fetal mouse progenitors express at least one member of the Ly49 family that

is acquired in a time-dependent and stochastic manner independently of

CD94 and NKG2. Eur. J. Immunol. 32, 868–878.

Gascoyne, D.M., Long, E., Veiga-Fernandes, H., de Boer, J., Williams, O.,

Seddon, B., Coles, M., Kioussis, D., and Brady, H.J. (2009). The basic leucine

zipper transcription factor E4BP4 is essential for natural killer cell develop-

ment. Nat. Immunol. 10, 1118–1124.

Gaston, K., and Fried, M. (1995). CpG methylation has differential effects on

the binding of YY1 and ETS proteins to the bi-directional promoter of the

Surf-1 and Surf-2 genes. Nucleic Acids Res. 23, 901–909.

Gordon, S.M., Chaix, J., Rupp, L.J., Wu, J., Madera, S., Sun, J.C., Lindsten, T.,

and Reiner, S.L. (2012). The transcription factors T-bet and Eomes control key

checkpoints of natural killer cell maturation. Immunity 36, 55–67.

Grenningloh, R., Kang, B.Y., and Ho, I.C. (2005). Ets-1, a functional cofactor of

T-bet, is essential for Th1 inflammatory responses. J. Exp. Med. 201, 615–626.

Grenningloh, R., Tai, T.S., Frahm, N., Hongo, T.C., Chicoine, A.T., Brander, C.,

Kaufmann, D.E., and Ho, I.C. (2011). Ets-1 maintains IL-7 receptor expression

in peripheral T cells. J. Immunol. 186, 969–976.

Held, W., Kunz, B., Lowin-Kropf, B., van de Wetering, M., and Clevers, H.

(1999). Clonal acquisition of the Ly49A NK cell receptor is dependent on the

trans-acting factor TCF-1. Immunity 11, 433–442.

Higuchi, T., Bartel, F.O., Masuya, M., Deguchi, T., Henderson, K.W., Li, R.,

Muise-Helmericks, R.C., Kern, M.J., Watson, D.K., and Spyropoulos, D.D.

(2007). Thymomegaly, microsplenia, and defective homeostatic proliferation

of peripheral lymphocytes in p51-Ets1 isoform-specific null mice. Mol. Cell.

Biol. 27, 3353–3366.

Hollenhorst, P.C., Shah, A.A., Hopkins, C., and Graves, B.J. (2007). Genome-

wide analyses reveal properties of redundant and specific promoter occu-

pancy within the ETS gene family. Genes Dev. 21, 1882–1894.

Hollenhorst, P.C., Chandler, K.J., Poulsen, R.L., Johnson, W.E., Speck, N.A.,

and Graves, B.J. (2009). DNA specificity determinants associate with distinct

transcription factor functions. PLoS Genet. 5, e1000778.

Huang, W., Sherman, B.T., and Lempicki, R.A. (2009). Systematic and integra-

tive analysis of large gene lists using DAVID bioinformatics resources. Nat.

Protoc. 4, 44–57.

Intlekofer, A.M., Takemoto, N., Wherry, E.J., Longworth, S.A., Northrup, J.T.,

Palanivel, V.R., Mullen, A.C., Gasink, C.R., Kaech, S.M., Miller, J.D., et al.

(2005). Effector and memory CD8+ T cell fate coupled by T-bet and eomeso-

dermin. Nat. Immunol. 6, 1236–1244.

John, S.A., Clements, J.L., Russell, L.M., and Garrett-Sinha, L.A. (2008). Ets-1

regulates plasma cell differentiation by interfering with the activity of the tran-

scription factor Blimp-1. J. Biol. Chem. 283, 951–962.

Joncker, N.T., and Raulet, D.H. (2008). Regulation of NK cell responsiveness to

achieve self-tolerance and maximal responses to diseased target cells.

Immunol. Rev. 224, 85–97.

Joncker, N.T., Fernandez, N.C., Treiner, E., Vivier, E., and Raulet, D.H. (2009).

NK cell responsiveness is tuned commensurate with the number of inhibitory

receptors for self-MHC class I: the rheostat model. J. Immunol. 182,

4572–4580.

Kallies, A., Carotta, S., Huntington, N.D., Bernard, N.J., Tarlinton, D.M., Smyth,

M.J., and Nutt, S.L. (2011). A role for Blimp1 in the transcriptional network

controlling natural killer cell maturation. Blood 117, 1869–1879.

Kamizono, S., Duncan, G.S., Seidel, M.G., Morimoto, A., Hamada, K.,

Grosveld, G., Akashi, K., Lind, E.F., Haight, J.P., Ohashi, P.S., et al. (2009).

Nfil3/E4bp4 is required for the development and maturation of NK cells in vivo.

J. Exp. Med. 206, 2977–2986.

Karlhofer, F.M., Orihuela, M.M., and Yokoyama, W.M. (1995). Ly-49-indepen-

dent natural killer (NK) cell specificity revealed by NK cell clones derived from

p53-deficient mice. J. Exp. Med. 181, 1785–1795.

Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc. 931

Immunity

ETS1 Targets in NK Cells

Kee, B.L., and Murre, C. (1998). Induction of early B cell factor (EBF) and

multiple B lineage genes by the basic helix-loop-helix transcription factor

E12. J. Exp. Med. 188, 699–713.

Kim, S., Iizuka, K., Kang, H.S., Dokun, A., French, A.R., Greco, S., and

Yokoyama, W.M. (2002). In vivo developmental stages in murine natural killer

cell maturation. Nat. Immunol. 3, 523–528.

Lacorazza, H.D., Miyazaki, Y., Di Cristofano, A., Deblasio, A., Hedvat, C.,

Zhang, J., Cordon-Cardo, C., Mao, S., Pandolfi, P.P., and Nimer, S.D.

(2002). The ETS protein MEF plays a critical role in perforin gene expression

and the development of natural killer and NK-T cells. Immunity 17, 437–449.

Lanier, L.L. (2005). NK cell recognition. Annu. Rev. Immunol. 23, 225–274.

Lee, G.M., Pufall, M.A., Meeker, C.A., Kang, H.S., Graves, B.J., and McIntosh,

L.P. (2008). The affinity of Ets-1 for DNA is modulated by phosphorylation

through transient interactions of an unstructured region. J. Mol. Biol. 382,

1014–1030.

Moisan, J., Grenningloh, R., Bettelli, E., Oukka, M., and Ho, I.C. (2007). Ets-1 is

a negative regulator of Th17 differentiation. J. Exp. Med. 204, 2825–2835.

Muthusamy, N., Barton, K., and Leiden, J.M. (1995). Defective activation

and survival of T cells lacking the Ets-1 transcription factor. Nature 377,

639–642.

Narni-Mancinelli, E., Jaeger, B.N., Bernat, C., Fenis, A., Kung, S., De Gassart,

A., Mahmood, S., Gut, M., Heath, S.C., Estelle, J., et al. (2012). Tuning of

natural killer cell reactivity by NKp46 and Helios calibrates T cell responses.

Science 335, 344–348.

Nishimori, H., Sasaki, Y., Yoshida, K., Irifune, H., Zembutsu, H., Tanaka, T.,

Aoyama, T., Hosaka, T., Kawaguchi, S., Wada, T., et al. (2002). The Id2 gene

is a novel target of transcriptional activation by EWS-ETS fusion proteins in

Ewing family tumors. Oncogene 21, 8302–8309.

Nutt, S.L., and Kee, B.L. (2007). The transcriptional regulation of B cell lineage

commitment. Immunity 26, 715–725.

Pufall, M.A., Lee, G.M., Nelson, M.L., Kang, H.S., Velyvis, A., Kay, L.E.,

McIntosh, L.P., and Graves, B.J. (2005). Variable control of Ets-1 DNA binding

by multiple phosphates in an unstructured region. Science 309, 142–145.

Rodewald, H.R., Moingeon, P., Lucich, J.L., Dosiou, C., Lopez, P., and

Reinherz, E.L. (1992). A population of early fetal thymocytes expressing

932 Immunity 36, 921–932, June 29, 2012 ª2012 Elsevier Inc.

Fc gamma RII/III contains precursors of T lymphocytes and natural killer cells.

Cell 69, 139–150.

Rosmaraki, E.E., Douagi, I., Roth, C., Colucci, F., Cumano, A., and Di Santo,

J.P. (2001). Identification of committed NK cell progenitors in adult murine

bone marrow. Eur. J. Immunol. 31, 1900–1909.

Russell, L., and Garrett-Sinha, L.A. (2010). Transcription factor Ets-1 in cyto-

kine and chemokine gene regulation. Cytokine 51, 217–226.

Seth, A., and Papas, T.S. (1990). The c-ets-1 proto-oncogene has oncogenic

activity and is positively autoregulated. Oncogene 5, 1761–1767.

Stewart, C.A., Walzer, T., Robbins, S.H., Malissen, B., Vivier, E., and Prinz, I.

(2007). Germ-line and rearranged Tcrd transcription distinguish bona fide NK

cells and NK-like gammadelta T cells. Eur. J. Immunol. 37, 1442–1452.

Vivier, E., Raulet, D.H., Moretta, A., Caligiuri, M.A., Zitvogel, L., Lanier, L.L.,

Yokoyama, W.M., and Ugolini, S. (2011). Innate or adaptive immunity? The

example of natural killer cells. Science 331, 44–49.

Vosshenrich, C.A., Ranson, T., Samson, S.I., Corcuff, E., Colucci, F.,

Rosmaraki, E.E., and Di Santo, J.P. (2005). Roles for common cytokine

receptor gamma-chain-dependent cytokines in the generation, differentiation,

and maturation of NK cell precursors and peripheral NK cells in vivo.

J. Immunol. 174, 1213–1221.

Wang, D., John, S.A., Clements, J.L., Percy, D.H., Barton, K.P., and Garrett-

Sinha, L.A. (2005). Ets-1 deficiency leads to altered B cell differentiation,

hyperresponsiveness to TLR9 and autoimmune disease. Int. Immunol. 17,

1179–1191.

Wei, G.H., Badis, G., Berger, M.F., Kivioja, T., Palin, K., Enge, M., Bonke, M.,

Jolma, A., Varjosalo, M., Gehrke, A.R., et al. (2010). Genome-wide analysis of

ETS-family DNA-binding in vitro and in vivo. EMBO J. 29, 2147–2160.

Yokomori, N., Kobayashi, R., Moore, R., Sueyoshi, T., and Negishi, M. (1995).

A DNA methylation site in the male-specific P450 (Cyp 2d-9) promoter and

binding of the heteromeric transcription factor GABP. Mol. Cell. Biol. 15,

5355–5362.

Yokota, Y., Mansouri, A., Mori, S., Sugawara, S., Adachi, S., Nishikawa, S.,

and Gruss, P. (1999). Development of peripheral lymphoid organs and natural

killer cells depends on the helix-loop-helix inhibitor Id2. Nature 397, 702–706.

Yokoyama, W.M., and Kim, S. (2006). Licensing of natural killer cells by self-

major histocompatibility complex class I. Immunol. Rev. 214, 143–154.