From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

HEMATOPOIESIS AND STEM CELLS

Estrogen signaling promotes dendritic cell differentiation by increasing expression of the IRF4

Esther Carreras,1 Sean Turner,1 Mark Barton Frank,1,2 Nicholas Knowlton,1,2 Jeanette Osban,1,2 Michael Centola,1 Chae Gyu Park,3 Amie Simmons,4 Jose´Alberola-Ila,4 and Susan Kovats1

1Arthritis & Immunology Program, and 2Microarray Research Facility, Oklahoma Medical Research Foundation, Oklahoma City; 3Laboratory of Cellular Physiology and Immunology, The Rockefeller University, New York, NY; and 4Immunobiology & Cancer Program, Oklahoma Medical Research Foundation, Oklahoma City

During inflammation, elevated granulo- tors (MPs). We now have identified inter- the estradiol/ER␣-dependent DC popula- cyte macrophage–colony-stimulating fac- feron regulatory factor 4 (IRF4), a tran- tion, indicating that an elevated amount tor (GM-CSF) directs the development of scription factor induced by GM-CSF and of IRF4 substitutes for ER␣ signal- new dendritic cells (DCs). This pathway is critical for CD11b؉ DC development in ing. Thus at an early stage in the MP influenced by environmental factors, and vivo, as a target of ER␣ signaling during response to GM-CSF, ER␣ signaling in- we previously showed that physiologic this process. In MPs, ER␣ potentiates and duces an elevated amount of IRF4, which levels of estradiol, acting through estro- sustains GM-CSF induction of IRF4. Fur- leads to a developmental program under- .gen receptor alpha (ER␣), promote the thermore, retroviral delivery of the Irf4 lying CD11b؉ DC differentiation. (Blood (GM-CSF–mediated differentiation of a cDNA to undifferentiated ER␣؊/؊ bone 2010;115:238-246 CD11b؉ DC subset from myeloid progeni- marrow cells restored the development of Introduction

Dendritic cell (DC) subsets in vivo are distinguished based on several transcription factors, including Irf4, to be crucial for location, surface markers, function, and migratory capacity, and development of specific DC subsets found in lymphoid organs.3 whether they are present in steady-state conditions or develop as a For example, mice deficient in relB, PU.1, or interferon regulatory consequence of inflammation.1 In non–lymphoid tissues such as factor 4 (IRF4) lack splenic CD11bhi CD8␣Ϫ CD4ϩ DCs, whereas dermis and lung, 2 major populations of CD11cϩ MHCIIϩ DCs mice deficient in IRF8 lack splenic CD8␣ϩ CD4Ϫ CD11bϪ DCs have been identified as CD11bhi CD103Ϫ and CD11blo CD103ϩ and plasmacytoid DCs. IRF4 also is necessary for the development ϩ Ϫ Langerin or .2,3 These interstitial DCs derive from monocytes or of MHC class IIϩ DCs during GM-CSF–mediated differentiation blood-borne precursors that seed tissue, while a distinct population from murine bone marrow (BM) or human monocytes in vitro.8-10 of epidermal Langerinϩ DCs (LCs) are replenished slowly from Less is known about transcription factor requirements for tissue local self-renewing precursors. The that mediate the DCs, and factors directing the development of recently defined development of interstitial DCs during homeostasis are not defined, interstitial CD11bϩ or CD103ϩ Langerinϩ DCs have not been and may include Flt3 or micro amounts of granulocyte reported. Analyses of IRF4Ϫ/Ϫ mice did not include assessments of macrophage–colony-stimulating factor (GM-CSF) or M-CSF tissue DCs.8,9 present in non–lymphoid tissue.3 During inflammation, DC differ- Transcription factor dosage is often important in cell fate entiation from Gr-1hi monocytes is largely mediated by elevated specification. Relevant to our study, quantitatively different levels amounts of GM-CSF.4 This GM-CSF–driven pathway is elevated of IRF4 protein alter B-cell differentiation pathways.11 IRF4 acts in infection and autoimmune disease and leads to the appearance of alone or dimerizes with PU.1 or Spi-B to mediate transcription.12 de novo DC populations in lymphoid organs and tissues.1,5 In Quantitative differences in PU.1 in hematopoietic progenitors also culture models, GM-CSF mediates CD11bϩ DC differentiation lead to alternate myeloid or lymphoid fates.13,14 Thus, IRF4 and from monocytes and myeloid progenitors (MPs), and a subset of PU.1 act as dosage-sensitive regulators to instruct distinct cell the CD11bint DCs express Langerin.6,7 However, it remains unclear differentiation programs, and therefore, factors (eg, hormones or how these DCs relate to Langerinϩ epidermal LCs or populations other transcription factors15,16) that regulate IRF4 expression are of CD11bϩ or Langerinϩ interstitial DCs. likely to alter IRF4-mediated developmental pathways. The lineage fate of hematopoietic progenitor cells is determined Environmental stimuli can influence DC differentiation, and we by ordered expression of transcription factors and receptor tyrosine previously have defined a role for (ER) signaling kinases such as Flt3, yet the molecular steps involved in the in DC development.17 ERs are ligand-dependent transcription development of specific DC subtypes from common precursors, factors that regulate expression by forming complexes with including myeloid and lymphoid progenitors, pro-DCs, and mono- chromatin-modifying coregulators.18 Immune cells and their pro- cytes, are not completely defined. Genetic analyses have identified genitors express ERs, suggesting that ER-mediated responses to

Submitted August 4, 2009; accepted October 7, 2009. Prepublished online as payment. Therefore, and solely to indicate this fact, this article is hereby Blood First Edition paper, October 30, 2009; DOI 10.1182/blood-2009-08- marked ‘‘advertisement’’ in accordance with 18 USC section 1734. 236935.

The publication costs of this article were defrayed in part by page charge © 2010 by The American Society of Hematology

238 BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2 From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2 ER SIGNALING INCREASES IRF4 EXPRESSION 239 circulating 17-␤-estradiol (E2) regulate cellular differentiation and driven DC differentiation cultures, or DC populations in BM were analyzed function. Indeed, elevated E2 levels decrease lymphopoiesis in directly ex vivo by flow cytometry. vivo by depleting primitive hematopoietic progenitors.19 E2 and ER␣ promote murine and human DC differentiation mediated by Microarrays 20-22 GM-CSF. We demonstrated that E2 specifically promotes the MPs were incubated with 1nM E2 or vehicle for variable time periods ϩ GM-CSF–mediated differentiation of murine CD11c CD11bint between 0 and 72 hours. Biotinylated amplified RNA was produced from Ϫ ϩ Ϫ Ly6C Langerin or DCs, a population that may represent in vivo 175 ng of total RNA per sample using an Illumina TotalPrep RNA populations of tissue or inflammatory DCs.7 Estradiol acting via Amplification Kit (Ambion). Amplified RNA was hybridized overnight at ER␣, but not ER␤, was required for the development of this 58°C to Sentrix Mouse-6 Expression BeadChips Version 1.1 (Illumina). CD11bint Ly6CϪ DC population from highly purified MPs (LinϪ These arrays contain 50-mer oligonucleotides coupled to beads that are c-kithi sca-1Ϫ IL-7R␣Ϫ flt3ϩ),17 but the molecular mechanism of mounted on glass slides. Each bead has an approximately 20- to 30-fold ER␣ action was unknown. redundancy. Microarrays were washed under high stringency, labeled with streptavidin-Cy3, and scanned using an Illumina BeadStation 500 scanner. We now report that E2/ER␣ signaling increases Irf4 mRNA and Ϫ To gain statistical power, response over time was protein expression in differentiating MPs and CD11c precursors, parameterized through a “mixed model”25 according to the following int Ϫ and promotes the development of the CD11c Ly6C DC subset equation: Normalized Fluorescent Units ϭ␣*Time ϩ␤1*Trx ϩ expressing the highest levels of IRF4. Enforced expression of Irf4 ␤2*Trx*Time, where ␣, ␤1, ␤2 are beta coefficients, Trx is Estrogen and Ϫ Ϫ cDNA in differentiating ER␣ / BM cells restored the develop- No Estrogen, and Time is in hours. The Trx*Time interaction P values were ment of the E2/ER␣-dependent CD11bint Ly6CϪ DC population, adjusted with a False Discovery Rate (FDR).26 An FDR of 10% was indicating that sufficiently high levels of IRF4 alleviate the considered statistically significant. A first order autoregressive covariance requirement for ER␣ signaling during GM-CSF–mediated DC matrix was used to account for the correlation structure inherent in the differentiation. Thus, an elevated amount of cellular IRF4 expres- longitudinal measurements of gene expression. All mixed model analysis sion, which can be induced by E2/ER␣ signaling, is required for was performed in the SAS system Version 9.1.3. CD11bϩ and Langerinϩ DC differentiation mediated by GM-CSF. qPCR

RNA and cDNA were generated and quantitative real-time reverse transcriptase– polymerase chain reaction (qPCR) was done as described.17 Relative Methods expression of was determined using the ⌬⌬Ct method with normaliza- tion to Gapdh expression, which does not change with E2 stimulation. Mice Specific primer sequences were: sense Gapdh 5Ј AGGTCGGTGTGAACG- Female C57BL/6 mice (CD45.2ϩ) and congenic CD45.1ϩ C57BL/6 mice GATTTG 3Ј, antisense Gapdh 5Ј TGTAGACCATGTAGTTGAGGTCA 3Ј; (8 to 12 weeks old) were purchased from the NCI Animal Production sense Irf4 5Ј GCCCAACAAGCTAGAAAG 3Ј, antisense Irf4 5Ј TCTCT- Program. ER␣-deficient (B6.129-Esr1tm1KskN10) mice from JAX were bred GAGGGTCTGGAAACT 3Ј; sense Irf8 5Ј GTTTACCGAATTGTC- at Oklahoma Medical Research Foundation. The Oklahoma Medical CCCGAG 3Ј, antisense Irf8 5Ј CTCCTCTTGGTCATACCCATGTA 3Ј; Research Foundation Institutional Animal Care and Use Committee Sense Pu.1 5Ј ATGTTACAGGCGTGCAAAATGG 3Ј, antisense Pu.1 5Ј approved the studies. TGATCGCTATGGCTTTCTCCA 3Ј. Id2 primers (sequences proprietary) were obtained from QIAGEN. GM-CSF–mediated DC differentiation Retrovirus-mediated gene delivery The GM-CSF–driven culture model for DC differentiation, in which E2 17 The Irf4 cDNA was cloned from murine DCs and inserted in front of the levels or ER signaling are manipulated, was used as described. Cultures 27 Ϫ Egfp gene in the retroviral vector MiG. To generate virus, MiG was were initiated from total BM cells, Lin BM cells, or purified MPs. transfected with pCL/Eco plasmid into 293T cells. LinϪ BM cells from Standard (hormone-replete) culture medium was RPMI, 10% fetal bovine ER␣Ϫ/Ϫ mice were isolated and cultured for 16 hours in stem cell factor serum (FBS). Steroid hormone-deficient culture medium was phenol red (50 ng/mL), IL-3 (20 ng/mL), and IL-6 (50 ng/mL). For viral transduction, free RPMI, 10% charcoal dextran stripped FBS (Omega). E2 (Sigma- 1.5 ϫ 105 LinϪ BM cells were mixed with 2 mL of retroviral supernatant Aldrich) was dissolved in EtOH (vehicle), or was in water-soluble form as plus 10 ␮g/mL polybrene in 12-well plates and centrifuged for 1 hour. cyclodextrin-encapsulated E2 with cyclodextrin as vehicle, and added to a Transduced LinϪ cells were cultured in GM-CSF in E2-deficient medium final concentration of 0.1nM to 1nM. ICI 182,780 was dissolved in DMSO for 5 days. (vehicle) and added to a final concentration of 100nM.

Statistical analyses Flow cytometry and myeloid progenitor isolation Statistical significance (P Ͻ .05) of differences in IRF4 mean fluorescence 17 Cells were prepared for flow cytometry as described. The goat anti-IRF4 intensity (MFI) values was determined using unpaired t tests in Prism antiserum M-17 (Santa Cruz Biotechnology) was detected with donkey software. Significant differences between ER␣ϩ and ER␣Ϫ/Ϫ DCs in mixed anti–goat Ig-Cy3, and isotype control goat IgG, or anti-IRF4 Ab preblocked BM chimeras were determined using paired t tests. with cognate peptide, controlled for the specificity of binding.8 The anti-Langerin mAb L31 was biotinylated and detected with streptavidin- allophycocyanin.23 Samples were run on an LSRII instrument (Becton Dickinson Biosciences) and analyzed with FlowJo software Version 8.8.5 Results (TreeStar). LinϪ cells and MPs were isolated as described.17 An early and brief exposure to estradiol significantly increases DC differentiation BM chimeric mice ϩ ϩ hi ϩ Mixed ER␣ϩ/ER␣Ϫ/Ϫ BM chimeric mice were generated by transferring Two major subsets of CD11c MHCII DCs (CD11b Ly6C and int Ϫ ϩ or Ϫ equal numbers of ER␣ϩCD45.1ϩ and ER␣Ϫ/ϪCD45.2ϩ BM cells into CD11b Ly6C Langerin ) differentiate from BM precursors 24 7 ␣ lethally irradiated [CD45.1xCD45.2]F1 recipients as described. On days 21 in GM-CSF–driven cultures. We previously showed that ER and to 24 after reconstitution, BM was isolated and placed into GM-CSF– physiologic amounts of E2 (0.1nM-1nM) were required for the From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

240 CARRERAS et al BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2

Figure 1. A 6-hour E2 exposure to myeloid progenitors yields a significant increase in DCs after 7 days. (A) MPs were stimulated with GM-CSF and 0.1nM E2 in standard medium at time 0. At the indicated times (0-96 hours), ER signaling was blocked by ICI 182,780 (100nM), such that E2 responsiveness was limited to the hour shown on the x axis. All cells were analyzed by flow cytometry on day 7. Shown are the average and range of percentages (left panel) and numbers (right panel) of CD11cϩ MHCIIϩ DCs in duplicate cultures. (B) Shown are the relative percentages of 2 DC subsets (CD11bhi Ly6Cϩ and CD11bint Ly6CϪ) present in cultures harvested on day 7 after E2 exposure for 0, 6, 48, or 168 hours. (C) Shown is the staining of anti-Langerin mAb L31 in DCs in the absence or presence of E2. Langerinϩ DCs are within the CD11bint Ly6CϪ population. Data are representative of 2 independent experiments.

GM-CSF–mediated differentiation of the CD11cϩ CD11bint Ly6CϪ vehicle or E2 (1nM) for variable times between 6 hours and subset from MPs.17 To determine the temporal window of E2 72 hours. This experiment identified Irf4 as an E2-induced exposure needed to augment DC differentiation, we incubated MPs gene. We focused on IRF4 because it is critical for develop- (LinϪ c-kithi sca-1Ϫ IL-7R␣Ϫ flt3ϩ) in standard (hormone-replete) ment of CD11cϩ CD11bϩ splenic DCs in vivo and in GM-CSF– medium supplemented with E2 (0.1nM) for various periods of time driven cultures. (6 hours to 96 hours) before blockade of ER signaling by addition To confirm E2 induction of Irf4 mRNA, independently isolated of ICI 182,780, an ER antagonist that induces ER degradation.28 MPs were incubated with GM-CSF and 1nM E2 (or vehicle) in Cells were cultured for 7 days, and the percentage and numbers of hormone-deficient medium for 3 to 48 hours before analysis of Irf4 ϩ ϩ CD11c MHCII DCs were quantified by flow cytometry (Figure expression by qPCR. E2 induced a marked (ϳ 3- to 4-fold over the ϩ 1A). A 6-hour E2 exposure was sufficient to augment CD11c DC vehicle control) increase in Irf4 mRNA by 3 hours, which contin- development approximately 2-fold with a marked increase in the ued to increase throughout the 48-hour time point (to 6- to 7-fold int Ϫ CD11b Ly6C population (Figure 1B). Within 48 hours of E2 over the vehicle control; Figure 2A). In contrast, in the absence exposure, the percentages of total differentiated CD11cϩ MHCIIϩ of E2, the GM-CSF–induced increase in Irf4 mRNA was DCs and the CD11bint Ly6CϪ subset were comparable with that of a significantly less and peaked at 6 hours. Due to these distinct 7-day (168-hour) E2 exposure. Within the CD11bint Ly6CϪ popula- kinetics, the difference in Irf4 mRNA levels in the absence or tion, E2 exposure also increases the percentage of Langerinϩ DCs presence of E2 at the early time points was 2-fold at 6 hours, (Figure 1C). Thus, within 6 hours, E2/ER␣ signaling in MPs biases with greater differences at 3 hours and 9 to 48 hours (Figure the developmental program initiated by GM-CSF toward develop- ϩ Ϫ 2B). These data show that upon stimulation with GM-CSF, ER␣ ment of the CD11c CD11bint Ly6C DC subset. signaling induces greater and more sustained amounts of Irf4 mRNA in MPs within 3 hours. Estradiol increases Irf4 mRNA levels in myeloid progenitors We investigated whether E2 exposure altered MP expression of To identify genes induced by E2/ER␣ signaling, we performed a other transcription factors involved in DC differentiation. Unlike microarray experiment in which MPs were incubated with Irf4, unstimulated MPs contained detectable levels of Irf8, Pu.1, From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2 ER SIGNALING INCREASES IRF4 EXPRESSION 241

Figure 2. E2 exposure markedly increases the amount of Irf4 mRNA present in GM-CSF–stimulated myeloid pro- genitors. MPs were stimulated with GM-CSF and 1nM E2 or vehicle in hormone-deficient medium. At the indicated time points, the amount of gene-specific mRNA was quantified using qPCR of the sample in triplicate. The relative expres- sion of each mRNA was calculated using the ⌬⌬Ct method. (A) Irf4 mRNA was not present at time 0 and was increased significantly in the presence of E2. (B) The fold increase in Irf4 mRNA due to the presence of E2 at early time points was reproducible. Error bars represent the range in 2 indepen- dent experiments. (C) Irf8, (D) Id2, and (E) Pu.1 mRNA were present at time 0, as determined by the Ct values, and the relative amounts did not differ in the presence or absence of E2. Data are representative of 2 to 4 independent experi- ments.

and Id2 mRNA. The mRNA levels of these factors did not change Estradiol increases IRF4 protein levels in differentiating significantly (Ͻ 2-fold) within 24 hours of GM-CSF stimulation, myeloid progenitors nor in the presence of E2 (Figure 2C-E). Amounts of these 3 mRNA decreased at 48 hours, but this did not depend on the Within 24 hours, GM-CSF–stimulated MPs exposed to E2 showed a presence of E2. Therefore, ER␣ signaling rapidly and specifically 2-fold increase in IRF4 protein, relative to vehicle-treated MPs (Figure increases Irf4 mRNA upon GM-CSF stimulation. 3A). An E2-induced increase in IRF4 protein was also detected

Figure 3. E2 exposure increases the amount of IRF4 protein in differentiating MPs 2-fold by 24 hours. MPs were stimu- lated with GM-CSF in hormone-deficient medium plus 1nM E2 or vehicle. (A) Shown is the binding of anti-IRF4 Ab to differenti- ating MPs in the absence (dotted line) or presence (solid line) of E2 at 24 hours. The shaded histogram indicates the binding of anti-IRF4 Ab preincubated with blocking peptide. The mean fluorescence intensity (MFI; with value for anti-IRF4 Ab bound to blocking peptide subtracted) of anti-IRF4 binding at 0 and 24 hours is plotted. Data are representative of 2 experiments. (B) MPs were stimulated with GM-CSF in the presence of E2 or E2 plus ICI 182,780 (100nM). The amount of intracellular IRF4 protein at 48 and 96 hours was assessed using flow cytometry. Shown are the MFI values (mean Ϯ SD) of anti-IRF4 binding (with value for anti-IRF4 Ab bound to blocking peptide sub- tracted) of cells in triplicate cultures. **P Ͻ .01, n ϭ 3. From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

242 CARRERAS et al BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2

Figure 4. E2/ER␣ signaling promotes development of DCs that harbor increased amounts of IRF4 protein. (A) LinϪ BM cells were incubated with 1nM E2 or vehicle in hormone-deficient medium. On day 7, IRF4 protein expression in CD11cϪ and CD11cϩ cells was determined. Shown are: binding of anti-IRF4 Ab to cells from vehicle-treated cultures (dotted line); binding of anti-IRF4 Ab to cells from E2-treated cultures (solid line); binding of isotype control IgG on cells from the E2-treated cultures (shaded gray). MFI values for anti-IRF4 binding with or without E2 are indicated (with value for isotype control IgG subtracted). Data are representative of 3 experiments. (B-E) MPs were incubated in standard medium with ICI 182,780 (100nM) or vehicle. On day 7, the amount of IRF4 protein expression in CD11cϪ and CD11cϩ cells was determined. (B) CD11c expression on all cells; the percentages of CD11cϪ and CD11cϩ cells are indicated. (C) Anti-IRF4 binding on CD11cϪ and CD11cϩ cells in the presence of ICI 182,780 (dotted line) or vehicle (solid line). The binding of isotype control IgG on cells in vehicle-treated cultures is indicated (shaded gray). MFI values for anti-IRF4 binding with or without ICI 182,780 are indicated (with value for isotype control IgG subtracted). (D) Within CD11cϩ cells, CD11bint DCs express higher levels of IRF4 than CD11bhi DCs. The percentage of DCs within each gate is indicated. (E) MFI values for anti-IRF4 binding with or without ICI 182,780 are indicated for each DC subset. Data are representative of 2 experiments. in undifferentiated CD11cϪ precursors at 48 and 96 hours Ly6Cϩ DCs) expressed 2-fold more IRF4 protein than ER␣Ϫ/Ϫ DCs (Figure 3B). (composed primarily of CD11bhi Ly6Cϩ DCs; Figure 5A-D). One pos- After 7 days of E2 exposure in cultures initiated from lineage- sibility was that DC precursors in ER␣Ϫ/Ϫ mice are defective due to the negative (LinϪ) BM cells in hormone-deficient medium, CD11cϪ absence of factors supplied by estrogen-responsive BM stromal cells in precursors and fully differentiated CD11cϩ cells contained 2-fold vivo.29 To determine whether the effect of ER␣ signaling on IRF4 more IRF4 protein than vehicle-treated cells (Figure 4A). expression was intrinsic to hematopoietic DC precursors, we generated ICI 182,780 inhibited the increase in IRF4 expression in CD11cϪ ER␣ϩ/ER␣Ϫ/Ϫ mixed BM chimeric mice, and BM cells from these precursors that occurred in hormone-replete medium (Figure chimeric mice were placed into GM-CSF–driven cultures. After 7 days, 4B-E). Within the CD11cϩ cell fraction, the CD11bint Ly6CϪ DC ER␣ϩ cells expressed higher amounts of IRF4 and had developed into subset, whose appearance is inhibited by ICI 182,780, expressed both CD11bhi and CD11bint DC subsets, whereas ER␣Ϫ/Ϫ cells ex- 2-fold more IRF4 than the CD11bhi Ly6Cϩ DC subset (Figure pressed lower amounts of IRF4 and had developed primarily into 4D-E). ER signaling also increased expression of IRF4 in the CD11bhi DCs (Figure 5E,F). These data show that ER␣ signaling acts in CD11bhi Ly6Cϩ DC subset (Figure 4E). These data show that when a cell-autonomous manner in DC precursors to increase expression of fully differentiated, the E2/ER␣-dependent CD11bint Ly6CϪ DC IRF4. Thus, E2/ER␣ signaling in MPs or CD11cϪ DC precursors leads subset normally expresses 2-fold higher levels of IRF4 than the to a reproducible 2-fold increase in IRF4 protein by 24 hours, which is CD11bhi Ly6Cϩ DC subset, suggesting that higher levels of IRF4 maintained throughout the 7-day culture period and in fully protein are required for their development and maintenance. differentiated CD11cϩ CD11bint Ly6CϪ cells.

ER␣ signaling in hematopoietic cells increases IRF4 Enforced expression of IRF4 rescues the development of the expression in a cell-autonomous manner ER␣-dependent DC population from ER␣؊/؊ BM precursors Comparison of differentiating ER␣ϩ and ER␣Ϫ/Ϫ BM cells showed that To determine whether elevated amounts of IRF4 protein were on day 7, ER␣ϩ DCs (composed of CD11bint Ly6CϪ and CD11bhi sufficient to induce differentiation of the CD11bint Ly6CϪ DC From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2 ER SIGNALING INCREASES IRF4 EXPRESSION 243

Figure 5. E2/ER␣ signaling acts in a cell-autonomous manner to increase expression of IRF4. (A-D) WT or ER␣Ϫ/Ϫ BM cells were incubated with GM-CSF in standard medium for 7 days. (A) The percentage of CD11cϩ CD11bϩ DCs was determined. (B) WT CD11cϩ cells harbor both CD11bhi Ly6Cϩ and CD11bint Ly6CϪ subsets, whereas ER␣Ϫ/Ϫ CD11cϩ cells primarily harbor the CD11bhi Ly6Cϩ subset. (C) Binding of anti-IRF4 Ab to WT (solid line) or ER␣Ϫ/Ϫ (dotted line) cells. The binding of isotype control IgG on WT cells is indicated (shaded gray). MFI values for anti-IRF4 binding to cells are indicated (with value for isotype control IgG subtracted). (D) The average fold change in IRF4 expression (anti-IRF4 Ab MFI) due to ER␣ in CD11cϪ and CD11cϩ is plotted; error bars represent the standard error in 5 independent experiments. (E-F) BM from ER␣ϩ/ER␣Ϫ/Ϫ BM chimeric mice was incubated with GM-CSF in standard medium for 7 days. (E) The percentage of CD11cϩ DCs in the WT/CD45.1 and ER␣Ϫ/Ϫ/CD45.2 cell fractions was determined. (F) The bar graph shows MFI values of anti-IRF4 Ab binding to CD11cϪ and CD11cϩ cells, averaged (mean Ϯ SEM) from 4 BM chimeric mice generated from the same BM transfer. Identical results were obtained upon analyses of 4 mice from a second BM chimera experiment. ***P Ͻ .001; *P Ͻ .05, n ϭ 4. subset in the absence of ER␣ signaling, we expressed the 6C,E) was not reproducible. While the majority of DCs was Irf4 cDNA in LinϪ ER␣Ϫ/Ϫ BM cells (Figure 6). LinϪ ER␣Ϫ/Ϫ CD11bhi Ly6Cϩ in the MiG-GFP–transduced cultures (Figure BM cells were transduced with MiG-GFP or MiG-IRF4-GFP 6C-F) and in the GFPϪ fraction of the MiG-IRF4-GFP– retroviruses and subsequently stimulated with GM-CSF. After transduced cultures (Figure 6I-J), the majority of DCs in the 5 days, the phenotypes of cells in GFPϩ and GFPϪ fractions GFPϩ fraction of the MiG-IRF4-GFP cultures was CD11bint were determined by flow cytometry. Enforced expression of Ly6CϪ (Figure 6K-L). The CD11bint Ly6CϪ population con- the Irf4 cDNA led to a 5-fold increase in the amount of tained both Langerinϩ and LangerinϪ DCs (not shown). intracellular IRF4 protein in GFPϩ cells, relative to GFPϪ cells Thus, increased expression of the Irf4 cDNA in differentiating in MIG-IRF4-GFP–transduced cultures (Figure 6H) or to cells in ER␣Ϫ/Ϫ LinϪ cells restores the development of the E2/ER␣- the MIG-GFP–transduced cultures (Figure 6B). This was as- dependent CD11bint Ly6CϪ DC population, indicating that sociated with a marked increase in the percentage of CD11cϩ sufficiently high levels of IRF4 substitute for the requirement cells (Figure 6I-K). The difference in percentage of CD11cϩ for E2/ER␣ signaling during GM-CSF–mediated DC cells in GFPϪ and GFPϩ MiG-GFP–transduced cultures (Figure differentiation. From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

244 CARRERAS et al BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2

Figure 6. Enforced IRF4 expression directs the development of the ER␣/E2-dependent CD11bint Ly6C؊ DC subset in the absence of ER␣ signaling. LinϪ ER␣Ϫ/Ϫ BM cells were transduced with MiG-GFP (A-F) or MiG-IRF4-GFP (G-L). Transduced cells were incubated in GM-CSF for 5 days before assessment of DC subsets by flow cytometry. IRF4 expression was assessed in GFPϪ (dotted line) and GFPϩ (solid line) fractions in (A-B) MiG-GFP– or (G-H) MiG-IRF4-GFP– transduced cells. The MFI of anti-IRF4 binding to GFPϩ and GFPϪ cells is indicated in panels B and H. The shaded histogram indicates the binding of anti-IRF4 Ab precincubated with blocking peptide. In MiG-GFP–transduced cell cultures, CD11cϩ DCs in the GFPϪ (C) and GFPϩ (E) cell fractions are primarily CD11bhi Ly6Cϩ (panels D and F gated on CD11cϩ cells shown in panels C and E). The percentage of cells with the bar or box gates is indicated. In MiG-IRF4-GFP–transduced cultures, CD11cϩ DCs in the GFPϪ cell fraction (I) are also primarily CD11bhi Ly6Cϩ (J, gated on CD11cϩ cells shown in panel I). In contrast, the percentage of CD11cϩ DCs in the GFPϩ fraction (K) is significantly increased, and the majority of DCs is CD11bint Ly6CϪ (L, gated on CD11cϩ cells shown in panel K). Data are representative of 4 independent experiments.

,ER␣ expression promotes CD11b؉ DC development during preparation). Thus, during radiation-induced inflammation in vivo inflammation in vivo ER␣ signaling in response to physiologic levels of endogenous estrogens promotes CD11bϩ DC development from short-term ␣ Our results in the GM-CSF–driven culture model suggest that ER repopulating BM progenitors. signaling promotes new DC differentiation during inflammation in vivo. To determine the relative efficiency of DC differentiation from ER␣ϩ and ER␣Ϫ/Ϫ BM precursors in vivo, we analyzed the CD11cϩ CD11bϩ DC subset present in BM of mixed ER␣ϩ/ Discussion ER␣Ϫ/Ϫ BM chimeric mice at an early time point (ϳ day 23) after reconstitution when myeloid cells, but few lymphocytes, have We have shown that an elevated amount of IRF4 protein, developed. In this model, DC development occurs in the postradia- induced by ER␣ signaling, promotes the GM-CSF–mediated tion inflammatory environment, which is characterized by elevated differentiation of CD11bint Ly6CϪ DCs from MPs. There is a serum IL-6 and high levels of costimulatory molecules on newly precedent for high and low amounts of IRF4 protein directing differentiated DCs, and normal levels of serum E2 (E.C., S. Bajan˜a, distinct transcriptional programs in differentiating cells. Dif- H. Agrawal, S.K., manuscript in preparation). In the BM of these ferentiating B cells expressed graded amounts of IRF4 protein, chimeric mice, newly differentiated CD11bϩ DCs were derived and high or low amounts of IRF4 specified plasma cell or preferentially from ER␣ϩ donor BM cells (ϳ 2:1 ratio of wild type germinal center gene expression programs, respectively.11 PU.1 [WT] to ER␣Ϫ/Ϫ DCs; Figure 7). Similar results were obtained pairs with IRF4 to regulate genes and, like IRF4, is required upon analyses of splenic and lymph node DC populations in these for the development of splenic CD11bϩ DCs.3,12 Quantita- chimeric mice (E.C., S. Bajan˜a, H. Agrawal, S.K., manuscript in tively different amounts (ϳ 4-fold) of PU.1 in hematopoietic From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2 ER SIGNALING INCREASES IRF4 EXPRESSION 245

In the GM-CSF–driven cultures, E2/ER␣ signaling promotes the development of CD11bint DCs, some of which are Lange- rinϩ. These could be the correlates of either epidermal Lange- rinϩ DCs or populations of CD11bϩ and CD103ϩ interstitial DCs found in the dermis and lung. CD103 is expressed on a significant portion of both the CD11bhi Ly6Cϩ and CD11bint Ly6CϪ DC subsets in our cultures (data not shown), so we could not use differential CD103 expression to correlate them with Figure 7. ER␣ expression promotes DC development during inflammation in vivo. Using flow cytometry, conventional CD11cϩ CD11bϩ DCs (cDCs) were interstitial DC populations. A role for IRF4 in the development identified in BM of ER␣ϩ/ER␣Ϫ/Ϫ BM chimeric mice on days 21 to 24 after reconstitu- of these tissue DC populations was not studied in published tion, a postradiation inflammatory environment. After gating on this DC population, reports of DC development in IRF4Ϫ/Ϫ mice.8,9 Our data now the percentage of cells bearing CD45.1 (WT ER␣ϩ) or CD45.2 (ER␣Ϫ/Ϫ) was determined. Shown are the (A) relative percentage and (B) number of cDCs derived link IRF4 to the development of one or more of these tissue DC from WT and ER␣Ϫ/Ϫ donor BM in individual mice (n ϭ 13), with the mean and SEM subsets in vivo. values indicated by the bars. Significant differences in the percentage and numbers The higher amount of IRF4 protein in E2-dependent CD11bint Ϫ Ϫ of WT and ER␣ / DCs were calculated using paired t tests. The total number of BM Ly6CϪ DCs may have functional implications. Fully developed cells in individual mice varied between 4 ϫ 106 and 1 ϫ 107. human DCs, CD11bϩ murine DCs, and macrophages retain expression of IRF4, and IRF4 negatively regulates Toll-like progenitors also specify distinct myeloid and lymphoid cell receptor signaling.39,40 Thus, IRF4hi DCs that develop in the ␣ fates, in some cases because the ratio of PU.1 to other context of E2/ER signaling may have a greater capacity to transcription factors is altered.14,30 This supports the signifi- regulate the magnitude of immune responses or promote cance of a 4- to 7-fold difference in IRF4 levels in MPs within tolerance. 3 to 9 hours of E2 exposure, and suggests that the ER␣/E2- During inflammation and autoimmune disease, elevated induced changes in IRF4 expression could lead to altered GM-CSF directs the development of new DCs from monocytes interactions with other transcription factors and/or to distinct or other proliferating precursors that infiltrate tissues and 5 transcriptional programs. secondary lymphoid organs. Increased estradiol synthesis also 41,42 IRF4 acts alone by binding ISRE or is recruited to composite has been linked to inflammation. Our data show that one IRF4/PU.1 binding sites by phosphorylated PU.1.12 Our data show physiologic factor that regulates de novo inflammatory DC ␣ that in MPs, Pu.1 mRNA changed little in response to GM-CSF or differentiation is E2/ER signaling, which leads to increased E2 stimulation over the 24-hour time period in which Irf4 mRNA is expression of the IRF4 transcription factor that is critical to ϩ significantly increased. Therefore, it is possible that increased CD11b DC development. amounts of IRF4 protein alter the recruitment of IRF4/PU.1 complexes to target genes, as demonstrated for IRF8/PU.1 com- plexes.31 Autoregulation of the Irf4 gene also could be amplified by ER␣ signaling because IRF4/PU.1 complexes bind to the Irf4 Acknowledgments 10 promoter. Alternately, elevated levels of IRF4 might allow IRF4 We thank J. Bass and D. Hamilton in the Oklahoma Medical to act independently of PU.1 to regulate genes that promote ϩ Research Foundation flow cytometry core facility, Drs H. Agrawal CD11b DC differentiation. and S. Bajan˜a for assistance with BM chimeras, and our Oklahoma ␣ ER binds directly to DNA at consensus estrogen response Medical Research Foundation colleagues for helpful discussions. 32 elements (EREs) proximal to genes. Algorithms within the ECR This work was supported by the Oklahoma Center for the Browser (www.dcode.org) at the National Center for Biotechnol- Advancement of Science & Technology grants HR06-157 (S.K.) ogy Information (NCBI) identified 4 conserved EREs proximal to and AR081-006 (M.C.); and National Institutes of Health grants ␣ the murine Irf4 gene. ER also may complex with other factors AI079616 (S.K.), RR020143 (E.C., M.C., S.K.), RR015577 (S.K., ␣ 18 such as Sp-1, NF-Y, FoxA1/HNF3 , or AP-1, and binding sites M.C.), and RR16478 (M.B.F.). for these factors are also present proximal to the Irf4 gene. Unbiased genome-wide chromatin binding (ChIP) and sequencing approaches to identify ER␣ binding sites in E2-stimulated cells revealed that many ER␣ binding sites and EREs are more than 5 kb Authorship away from target genes.33 Intriguingly, ER␣ bound to a region that is approximately 12 kb proximal to the start site of the human IRF4 Contribution: E.C. designed and performed research, analyzed gene.34 This region contains a consensus ERE site, which is and interpreted data, and wrote the manuscript; S.T. performed conserved in the same location 5Ј to the murine Irf4 gene. This research and analyzed and interpreted data; M.B.F. designed information is consistent with the hypothesis that E2-bound ER␣ research and analyzed and interpreted data; N.K. performed participates in a chromatin-modifying complex that directly regu- statistical analysis; J.O. and A.S. performed research; M.C. lates the Irf4 gene. analyzed and interpreted data; C.G.P. contributed vital new Alternately, ER␣ signaling may modulate expression or activity reagent; J.A.-I. designed research and analyzed and interpreted of an intermediate molecule such as nuclear factor–␬B (NF-␬B). data; and S.K. designed research, analyzed and interpreted data, Indeed, GM-CSF stimulated IRF4 expression in differentiating and wrote the manuscript. human monocytes by activating NF-␬B,10 and the NF-␬B subunit Conflict-of-interest disclosure: The authors declare no compet- Rel induced IRF4 expression in lymphocytes.35 RelB is also ing financial interests. important for the development of IRF4-dependent CD11bϩ splenic Correspondence: Susan Kovats, Arthritis & Immunology Pro- DCs.36,37 ER interactions with NF-␬B in other cell types have been gram, Oklahoma Medical Research Foundation, 825 NE 13th St, reported.38 Oklahoma City, OK 73104; e-mail: [email protected]. From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

246 CARRERAS et al BLOOD, 14 JANUARY 2010 ⅐ VOLUME 115, NUMBER 2

References

1. Shortman K, Naik SH. Steady-state and inflam- production by controlling the activity of IRF-4 zation and degradation of estrogen receptor- matory dendritic-cell development. Nat Rev Im- transcription factor. Immunity. 2008;29(6):899- alpha. J Biol Chem. 2006;281(14):9607-9615. munol. 2007;7(1):19-30. 911. 29. Yokota T, Oritani K, Garrett KP, et al. Soluble 2. Merad M, Ginhoux F, Collin M. Origin, homeosta- 16. Gauzzi MC, Purificato C, Conti L, Adorini L, frizzled-related protein 1 is estrogen inducible in sis and function of Langerhans cells and other Belardelli F, Gessani S. IRF-4 expression in the bone marrow stromal cells and suppresses the langerin-expressing dendritic cells. Nat Rev Im- human myeloid lineage: up-regulation during earliest events in lymphopoiesis. J Immunol. munol. 2008;8(12):935-947. dendritic cell differentiation and inhibition by 2008;181(9):6061-6072. 3. Merad M, Manz MG. Dendritic cell homeostasis. 1alpha,25-dihydroxyvitamin D3. J Leukoc Biol. 30. Nutt SL, Metcalf D, D’Amico A, Polli M, Wu L. Dy- Blood. 2009;113(15):3418-3427. 2005;77(6):944-947. namic regulation of PU. 1 expression in multipo- 17. Carreras E, Turner S, Paharkova-Vatchkova V, tent hematopoietic progenitors. J Exp Med. 2005; 4. Geissmann F, Auffray C, Palframan R, et al. Mao A, Dascher C, Kovats S. Estradiol acts di- 201(2):221-231. Blood monocytes: distinct subsets, how they re- rectly on bone marrow myeloid progenitors to dif- late to dendritic cells, and their possible roles in 31. Tamura T, Thotakura P, Tanaka TS, Ko MS, Ozato ferentially regulate GM-CSF or Flt3 ligand-mediated the regulation of T-cell responses. Immunol Cell K. Identification of target genes and a unique cis dendritic cell differentiation. J Immunol. 2008;180(2): Biol. 2008;86(5):398-408. element regulated by IRF-8 in developing macro- 727-738. phages. Blood. 2005;106(6):1938-1947. 5. Hamilton JA. Colony-stimulating factors in inflam- 18. Heldring N, Pike A, Andersson S, et al. Estrogen mation and autoimmunity. Nat Rev Immunol. 32. O’Lone R, Frith MC, Karlsson EK, Hansen U. receptors: how do they signal and what are their 2008;8(7):533-544. Genomic targets of nuclear estrogen receptors. targets. Physiol Rev. 2007;87(3):905-931. Mol Endocrinol. 2004;18(8):1859-1875. 6. Takahara K, Omatsu Y, Yashima Y, et al. Identifi- 19. Medina KL, Garrett KP, Thompson LF, Rossi MI, cation and expression of mouse Langerin 33. Kininis M, Kraus WL. A global view of transcrip- Payne KJ, Kincade PW. Identification of very (CD207) in dendritic cells. Int Immunol. 2002; tional regulation by nuclear receptors: gene ex- early lymphoid precursors in bone marrow and 14(5):433-444. pression, factor localization, and DNA sequence their regulation by estrogen. Nat Immunol. 2001; analysis. Nucl Recept Signal. 2008;6:e005. 7. Mao A, Paharkova-Vatchkova V, Hardy J, Miller 2(8):718-724. 34. Carroll JS, Meyer CA, Song J, et al. Genome- MM, Kovats S. Estrogen selectively promotes the 20. Paharkova-Vatchkova V, Maldonado R, Kovats S. wide analysis of estrogen receptor binding sites. differentiation of dendritic cells with characteris- Estrogen preferentially promotes the differentia- Nat Genet. 2006;38(11):1289-1297. tics of Langerhans cells. J Immunol. 2005;175(8): tion of CD11cϩ CD11b(intermediate) dendritic 5146-5151. 35. Grumont RJ, Gerondakis S. Rel induces inter- cells from bone marrow precursors. J Immunol. feron regulatory factor 4 (IRF-4) expression in 8. Tamura T, Tailor P, Yamaoka K, et al. IFN regula- 2004;172(3):1426-1436. lymphocytes: modulation of interferon-regulated tory factor-4 and -8 govern dendritic cell subset 21. Douin-Echinard V, Laffont S, Seillet C, et al. Es- gene expression by /nuclear factor kappaB. development and their functional diversity. J Im- trogen receptor alpha, but not beta, is required for J Exp Med. 2000;191(8):1281-1292. munol. 2005;174(5):2573-2581. optimal dendritic cell differentiation and [cor- 36. Wu L, D’Amico A, Winkel KD, Suter M, Lo D, 9. Suzuki S, Honma K, Matsuyama T, et al. Critical rected] CD40-induced production. J Im- Shortman K. RelB is essential for the develop- roles of interferon regulatory factor 4 in munol. 2008;180(6):3661-3669. ment of myeloid-related CD8alpha- dendritic cells CD11bhighCD8alpha- dendritic cell development. 22. Komi J, Lassila O. Nonsteroidal anti-estrogens but not of lymphoid-related CD8alphaϩ dendritic Proc Natl Acad Sci U S A. 2004;101(24):8981- inhibit the functional differentiation of human cells. Immunity. 1998;9(6):839-847. 8986. monocyte-derived dendritic cells. Blood. 2000; 37. Burkly L, Hession C, Ogata L, et al. Expression of 10. Lehtonen A, Veckman V, Nikula T, et al. Differen- 95(9):2875-2882. relB is required for the development of thymic tial expression of IFN regulatory factor 4 gene in 23. Cheong C, Idoyaga J, Do Y, et al. Production of medulla and dendritic cells. Nature. 1995;373(6514): human monocyte-derived dendritic cells and monoclonal antibodies that recognize the extra- 531-536. macrophages. J Immunol. 2005;175(10):6570- cellular domain of mouse langerin/CD207. J Im- 6579. 38. Kalaitzidis D, Gilmore TD. Transcription factor munol Methods. 2007;324(1-2):48-62. cross-talk: the estrogen receptor and NF-kappaB. 11. Sciammas R, Shaffer AL, Schatz JH, Zhao H, 24. Vremec D, Pooley J, Hochrein H, Wu L, Shortman K. Trends Endocrinol Metab. 2005;16(2):46-52. Staudt LM, Singh H. Graded expression of inter- CD4 and CD8 expression by dendritic cell sub- 39. Negishi H, Ohba Y, Yanai H, et al. Negative regu- feron regulatory factor-4 coordinates isotype types in mouse thymus and spleen. J Immunol. lation of Toll-like-receptor signaling by IRF-4. switching with plasma cell differentiation. Immu- 2000;164(6):2978-2986. Proc Natl Acad Sci U S A. 2005;102(44):15989- nity. 2006;25(2):225-236. 25. Littell R, Milliken G, Stroup W, Wolfinger R, 15994. 12. Marecki S, Fenton MJ. The role of IRF-4 in tran- Schabenberger O. SAS for Mixed Models. 2006. 40. Honma K, Udono H, Kohno T, et al. Interferon scriptional regulation. J Interferon Cytokine Res. Cary, NC: SAS Institute Inc. regulatory factor 4 negatively regulates the pro- 2002;22(1):121-133. 26. Benjamini Y, Hochberg Y. Controlling the false duction of proinflammatory cytokines by macro- 13. Dakic A, Wu L, Nutt SL. Is PU. 1 a dosage-sensitive discovery rate: a practical and powerful approach phages in response to LPS. Proc Natl Acad Sci regulator of haemopoietic lineage commitment to multiple testing. J R Stat Soc Ser B. 1995;57: USA.2005;102(44):16001-16006. and leukaemogenesis? Trends Immunol. 2007; 289-300. 41. Cutolo M, Villaggio B, Seriolo B, et al. Synovial 28(3):108-114. 27. Herna´ndez-Hoyos G, Anderson MK, Wang C, fluid estrogens in rheumatoid arthritis. Autoim- 14. Laslo P, Spooner CJ, Warmflash A, et al. Multilin- Rothenberg EV, Alberola-Ila J. GATA-3 expres- mun Rev. 2004;3(3):193-198. eage transcriptional priming and determination of sion is controlled by TCR signals and regulates 42. Salama SA, Kamel MW, Diaz-Arrastia CR, et al. alternate hematopoietic cell fates. Cell. 2006; CD4/CD8 differentiation. Immunity. 2003;19(1): Effect of tumor necrosis factor-alpha on estrogen 126(4):755-766. 83-94. metabolism and endometrial cells: potential phys- 15. Chen Q, Yang W, Gupta S, et al. IRF-4-binding 28. Long X, Nephew KP. Fulvestrant (ICI 182780)- iological and pathological relevance. J Clin Endo- protein inhibits interleukin-17 and interleukin-21 dependent interacting mediate immobili- crinol Metab. 2009;94(1):285-293. From www.bloodjournal.org by guest on March 23, 2015. For personal use only.

2010 115: 238-246 doi:10.1182/blood-2009-08-236935 originally published online October 30, 2009

Estrogen receptor signaling promotes dendritic cell differentiation by increasing expression of the transcription factor IRF4

Esther Carreras, Sean Turner, Mark Barton Frank, Nicholas Knowlton, Jeanette Osban, Michael Centola, Chae Gyu Park, Amie Simmons, José Alberola-Ila and Susan Kovats

Updated information and services can be found at: http://www.bloodjournal.org/content/115/2/238.full.html Articles on similar topics can be found in the following Blood collections Hematopoiesis and Stem Cells (3302 articles) Immunobiology (5287 articles)

Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests

Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints

Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.