17β-Estradiol Alters the Activity of Conventional and IFN-Producing Killer Dendritic Cells

Animals

Adult (>60 days of age) female (n = 400; 5–10/group/replication/experiment) C57BL/6 mice were purchased from the National Cancer Institute (Bethesda, MD) and Esr1^−/− females on a C57BL/6 background (n = 5) were purchased from Taconic Farms. All animals were housed five per cage in a microisolator room and maintained on a constant light:dark 16:8 cycle. Food and sterile tap water were available ad lib. The Johns Hopkins Animal Care and Use Committee (protocol # MO04H532) approved all procedures described.

Ovariectomy and hormone replacement

Animals were assigned to remain intact or be bilaterally ovariectomized (ovx). Mice were bilaterally ovx under ketamine (80 mg/kg body mass)–xylazine (8 mg/kg body mass) anesthesia (Phoenix Pharmaceutical) and given 2 wk to recover from surgery. Intact females received sham surgeries (sham). After recovery, females assigned to have E2 replaced, were each s.c. implanted with a 10 mm length Silastic capsule (i.d.= 1.02 mm, o.d. = 2.16 mm) filled with either 2.5 mm or 5 mm (5 mm of E2 was used in all experiments, except where noted; ovx+E2) of E2 (Sigma-Aldrich) and sealed with Silastic adhesive to form a controlled-release capsule as described previously (32). The remaining ovx females as well as sham females were each implanted with an empty Silastic capsule of equal length. The 5-mm dose of E2 was based on a previous report that this amount is sufficient to maintain physiological E2 concentrations (i.e., 100–200 pg/ml) in female mice for 14 days (33). Estradiol-containing and control Silastic capsules were incubated in sterile saline (37°C) for 24 h before s.c. implantation in the midscapula region. All animals were processed 2 wk after implanting capsules. Serum E2 concentrations were assayed using a standard ether extraction and a rodent E2 EIA kit, optimized in our laboratory (Cayman Chemicals). Uterine horns were removed and weighed as an additional bioassay for E2 concentrations.

Generation of DCs from BM cells

BM cells were isolated from the femurs and tibiae of adult female C57BL/6 mice and cultured in Iscove’s modified Dulbecco’s medium (IMDM) supplemented with 20% culture supernatant from GM-CSF-producing J558 cells (14). Because phenol red is weakly estrogenic (34), we conducted preliminary experiments to compare the phenotype and activity of BMDCs that were incubated in standard culture media that contained phenol red IMDM (Invitrogen Life Technology) plus 10% filtered FBS (Sigma-Aldrich) with cells incubated in estrogen-deficient media that contained phenol red-free IMDM (Invitrogen Life Technology) supplemented with 10% charcoal-dextran-treated FBS (HyClone). As reported previously (14), standard culture media had a significant effect on BMDCs (data not shown); thus, all in vitro experiments were conducted using estrogen-deficient media only. Total BM cells were suspended at 1 × 10⁶ cells per ml in estrogen-deficient medium in 12-well plates. Estrogen-deficient medium contained phenol red-free IMDM, 10% charcoal-dextran-treated FCS, l-glutamine, penicillin/streptomycin, gentamicin, and 2-ME (Invitrogen Life Technologies and Sigma-Aldrich). On days 3 and 6, half of the culture medium was removed and replaced with fresh estrogen-deficient medium containing GM-CSF. Estradiol (Steraloids) and/or ICI 182,780 (Tocris) were added at the beginning of the culture and were replenished with the medium on days 3 and 6. DCs were harvested on day 9.

Cell isolation and flow cytometry

Spleens were dissected and splenocytes were isolated by mincing and digesting whole tissue in 2 mg/ml collagenase D (Roche). Tissues were homogenized, cells were filtered and washed several times in PBS supplemented with 2 mM EDTA and 2% FCS, and mononuclear cells were isolated by resuspending cells in ACK lysis buffer (Invitrogen Life Technologies). Nonspecific binding was blocked by staining with FcR. Cell samples were stained with CD11c FITC, incubated with anti-FITC microbeads, and processed through LS columns using the quadroMACS magnet (Miltenyi Biotec). Retained cells were collected and viable cells were counted using a hemacytometer and trypan blue exclusion. Cells were then stained for appropriate mouse DC, NK, and IKDC markers. Cells were stained with the following mAbs: anti-mouse CD49b-PE (clone DX5), NK1.1-APC (clone PK136), CD11b-PE (clone M1/70), H2D^b-PE (clone KH95), CD8α-PE (Ly-2; clone 53-6.7), CD45R-APC (B220; clone RA3–6B2), Ly-6G-PE (Gr1; clone RB6–8C5), CD40-PE (clone IC10), NKG2D-PE (clone C7), and CD122-PE (clone 5H4) (BD Biosciences; eBiosciences).

Cell activation and cytokine production

Cells were stimulated with 5 ng/ml rIL-12 (eBiosciences), 5 ng/ml rIL-18 (MBL), 1 μg/ml LPS (Sigma-Aldrich), 1 μg/ml poly I:C (Invivogen), 1 μg/ml CpG (Integrated DNA Technologies), or 1 μg/ml a control oligonucleotide (Integrated DNA Technologies) for 24 h. Cytokine concentrations were assayed by ELISA using the manufacturer’s protocols for IFN-γ, IL-12p40, IL-12p70 (BD Pharmingen), and IFN-α (PBL Biomedical Laboratories).

NK cell lysis

Viable CD11c⁺ cells were adjusted to 1 × 10⁶ cells/ml and incubated with ⁵¹Cr-labeled YAC-1 target cells (American Type Culture Collection) for 4 h at 37°C at E:T ratios ranging from 100:1 to 12.5:1. After the 4 h incubation, ⁵¹Cr released from target cells was counted using a γ counter and lysis was calculated as: ((measured ⁵¹Cr release – spontaneous ⁵¹Cr release)/(maximum ⁵¹Cr release – spontaneous ⁵¹Cr release)) × 100. Maximum release was determined based on acid-lysed target cells and spontaneous release was determined by incubating target cells in the absence of effector cells. The average percent cytotoxicity for each set of duplicates was used in statistical analyses.

Microscopy

CD11c⁺ splenocytes were isolated as described above and sorted into CD49b⁺NK1.1^high and CD49b⁺NK1.1^low populations on a FACS Aria (BD Biosciences). For microscopy, sorted cells were fixed with 2% glutaraldehyde for 2 h at 24°C followed by postfixation with 2% osmonium tetroxide for 2 h at 24°C. The cells were embedded in Epon 812, sectioned at 2 μm, stained with toluidin blue, and examined by light microscopy using a Nikon Eclipse E800 light microscope (Nikon). Images were acquired at 60× using a SPOT RT CCD imager and software (Diagnostic Instruments).

Statistical analyses

The effects of E2 on all dependent measures were analyzed using one- or two-way ANOVAs with treatment group and/or TLR ligand as the independent variable(s). Significant interactions were further analyzed using the Tukey method for pairwise multiple comparisons. Comparisons between Esr1^−/− and wild-type (WT) females were conducted using Student’s t tests. If the data sets violated the assumptions of a normal distribution, then nonparametric statistics were used. Mean differences were considered statistically significant if p < 0.05.

17β-estradiol enhances the differentiation of naive BM-derived DCs

In vitro exposure to physiological concentrations of E2 facilitates differentiation of BM precursor cells into functional CD11c⁺ DCs with increased levels of MHC class II (MHCII), CD80, and CD86 on their surface (14). Many cells, however, express CD11c, including cDCs, pDCs, IKDCs, and NK cells (35, 36). We confirmed and expanded previous findings by generating CD11c⁺ cells from BM precursors, incubating cells with varying concentrations of E2 and/or the ER antagonist, ICI 182,780, and assessing cells for surface markers characteristic of cDCs, pDCs, IKDCs, and NK cells. Exogenous administration of 0.1–10 nM of E2 (i.e., concentrations within the range circulating in adult female mice (37)) facilitated differentiation of BM precursor cells into functional CD11c⁺ DCs and increased the expression of MHCII (Fig. 1⇓A), CD11b, CD40, and CD86, but not NK1.1 or CD49b (data not shown). The effect of E2 was mediated by the estrogen receptor because blocking ERα and ERβ with 100 nM of ICI 182,780 significantly inhibited DC differentiation (p < 0.05; Fig. 1⇓B and data not shown). Neither E2 nor ICI 182,780 alone or in combination affected BMDC viability as tested by 7-AAD staining (data not shown). ICI 182,780 alone also did not significantly affect differentiation of BMDCs (p < 0.05; Fig. 1⇓B and data not shown). These data illustrate that the in vitro effects of E2 on BM-derived CD11c⁺ cells are limited to cDCs and are mediated by the ER.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

17β-estradiol (E2) facilitates differentiation of DCs from bone marrow (BM). Bone marrow cells were cultured in estrogen-deficient medium with GM-CSF for 8 days in the absence or presence of the indicated concentrations of E2. Nonadherent cells were harvested and assessed for surface markers characteristic of mature cDCs by FACS. Plots represent the expression of I-A^b on CD11c⁺ cells. Values represent the percentage of MHCII⁺ (top) and MHC II^high (bottom) of the total CD11c⁺ cells (A). Bone marrow cells were cultured in estrogen-deficient medium with GM-CSF, E2 (1 nM), and the indicated concentrations of the ER antagonist, ICI 182,780. On day 8, cells were stained for FACS analysis. Plots represent CD40 v. CD86 expression on CD11c⁺ cells. Values represent the percentage of CD40⁺CD86⁺ of the total CD11c⁺ cells (B). Results are representative of three independent experiments.

17β-estradiol is required for activation of BMDCs by TLR ligands

To determine the effect of E2 on the ability of DCs to initiate an innate immune response, BM-derived cells differentiated in the presence of either no E2 or 1 nM of E2 (i.e., the median dose from Experiment 1) and were exposed to the TLR ligands LPS, poly I:C, CpG, media alone, or a control ODN. Cells were assessed for surface markers characteristic of mature cDCs and supernatants were collected to measure cytokine concentrations. E2 enhanced the expression of MHCII, CD40, and CD86 on CD11c⁺ BMDCs as compared with cells not treated with E2, regardless of the TLR agonist used (p < 0.05; Figs. 2⇓, A and B). Strikingly, BM-derived DCs differentiated in the absence of E2 showed no activation in response to stimulation with any of the TLR ligands. Among E2-treated BMDCs, exposure to CpG and LPS, but not poly I:C or the control ODN, further enhanced the expression of MHCII, CD40, and CD86 as compared with exposure to media alone (p < 0.05, Fig. 2⇓B). Treatment of BMDCs with E2 increased the production of IL-12p40 in response to CpG and LPS, but not poly I:C or the control ODN (p < 0.05; Fig. 2⇓C). Exposure to E2 also facilitated production of IL-12p70 by BMDCs in response to LPS (E2: 3.10 ± 0.41 ng/ml vs media: 1.54 ± 0.01 ng/ml; p < 0.05), but not CpG, poly I:C, or the control ODN (data not shown). E2 treatment tended to increase the synthesis of IFN-α in response to CpG, but this difference did not reach statistical significance (p > 0.05; Fig. 2⇓C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

E2 enhances the activation of CD11c⁺ BMDCs by TLR ligands. BMDCs were cultured in estrogen-deficient medium for 8 days and stimulated with TLR agonists or control ODN for the last 24 h of culture. Nonadherent cells were harvested and stained for FACS analysis. Representative plots of CD11c v. MHC II and CD40 v. CD86 (gated on CD11c⁺ cells) are shown (A). Values represent the percentage of MHCII^high and CD40⁺CD86⁺ cells. Summary of the proportion of CD11c⁺ BMDCs (±SEM) that expressed MHCII^high, CD40, and CD86 following culture with (E2⁺) or without (E2⁻) 1 nM E2 and activation with the indicated TLR ligands (B). Mean production of IFN-α and IL-12p40 (±SEM) by BMDCs exposed to either 1 nM of E2 or media alone and activated with the designated TLR ligands (C). CD11c⁺ BMDCs were generated by incubating cells with GM-CSF and either 1 nM of E2 (E2⁺) or media alone (E2⁻). During the last 24 h of the 8-day incubation, cells were activated with LPS, poly I:C, CpG, control ODN, or media alone. On day 9, cells were assessed for surface markers characteristic of mature cDCs by FACS and supernatants were collected to measure cytokine concentrations by ELISA. Results are representative of three independent experiments. An asterisk (∗) indicates that E2 enhances responses relative to media alone, a dagger (†) indicates that among E2-treated BMDCs, TLR ligands enhanced responses relative to media alone, p < 0.05.

Previous studies illustrate that BMDCs can produce IFN-γ following stimulation with IL-12 and IL-18 (38, 39). In addition to being responsive to classic TLR ligands, we sought to determine whether differentiated DCs from BM precursors were responsive to IL-12 and IL-18 and could produce IFN-γ in the presence of E2. BMDCs were differentiated as described above, exposed to either E2 or media alone, and activated with rIL-12 and rIL-18. BMDCs stimulated with IL-12 and IL-18 exhibited no changes in the expression of MHCII, CD40, and CD86 or the synthesis of IFN-γ, regardless of E2 exposure (p > 0.05; data not shown).

17β-estradiol increases numbers of CD11c⁺CD49b⁺NK1.1^low cells in the spleen

The data presented in this study and by others (14) illustrate that E2 impacts the in vitro differentiation of BM-derived precursors into functional cDCs. The effect of E2 on the phenotype and function of DCs in peripheral lymphoid organs, however, has not been addressed. To test whether E2 alters the number, surface phenotype, or activity of splenic DCs, we conducted a series of in vivo studies using intact/sham females, ovx females, and ovx females with E2 replacement (ovx+E2). Serum levels of E2 were significantly different among the three groups of females, in which concentrations were elevated in ovx+E2 females and reduced in ovx females relative to sham female mice (p < 0.05; Fig. 3⇓A). Exogenous replacement of E2 resulted in circulating estradiol concentrations that were within the physiological range (100–200 pg/ml) for female mice in estrus (37). Uterine horn mass was measured as a bioassay and was consistent with E2 concentrations, in which ovx+E2 and sham females had heavier uterine horns than ovx female mice (p < 0.05; Fig. 3⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

E2 differentially affects subpopulations of CD11c⁺ cells in the spleen. Serum, uterine horns, and spleens were collected from intact (sham), ovariectomized (ovx), and ovx females that had E2 replaced exogenously (ovx+E2). Circulating E2 concentrations were measured by ELISA (±SEM) (A) and uterine horns were weighed (±SEM) (B). Whole splenocytes were isolated, enriched for CD11c⁺ cells, and stained for CD11b (C) or CD49b (D). FACS analysis was performed by gating on CD11c⁺ cells. Bars with different letters indicate significant differences across groups (n = 10/group), p < 0.05.

CD11c⁺ splenocytes from C57BL/6 mice are a heterogeneous population of cells comprised of multiple subsets of DCs, including cDCs, pDC, and IKDCs, all of which are classified based on their surface marker profiles (30, 36, 40). To examine the effects of E2 on the absolute numbers and percentages of DCs in the spleen, CD11c⁺ splenocytes were isolated and stained for MHCII, CD11b, B220, CD8α, CD49b, and the costimulatory molecules, CD40 and CD86. In contrast to the effects on BMDCs, E2 did not significantly affect the absolute number of CD11c⁺ cells isolated from the spleen or the percentage of CD11c⁺ cells determined to be cDCs based on the high expression of CD11c and MHCII (p > 0.05, data not shown). Neither increasing nor decreasing concentrations of E2 in vivo altered the percentage of CD11c⁺CD11b⁺ cDCs, the percentage of CD8α⁺ cDCs, the percentage of B220⁺ pDCs, or the expression of the costimulatory molecules CD40 and CD86 on any of the DC subsets (p > 0.05; Fig. 3⇑C and data not shown).

There is overlapping expression of surface markers among DCs, NK cells, and IKDCs that can make the precise characterization of these cell populations complicated. Both IKDCs and NK cells express CD11c, CD49b, and NK1.1; thus, we sought to examine whether these cells were responsive to E2. Removal of circulating E2 via ovariectomy decreased, whereas sustained exposure to E2 dramatically increased the percentage of CD11c⁺CD49b⁺ cells in the spleen as compared with the sham controls (p < 0.05; Fig. 3⇑D).

To establish whether E2 alters the number of IKDCs in the spleen, we enriched for CD11c⁺ cells and stained for both CD49b (i.e., integrin subunit α2) and NK1.1 (i.e., NK receptor P1C). By gating on live CD11c⁺ cells, we observed that sustained exposure to E2 reproducibly increased the number of CD49b⁺NK1.1^low cells in the spleen as compared with sham and ovx female mice (p < 0.05; Fig. 4⇓, A and C). In contrast, the absence of E2 in ovx females increased the proportion of splenic CD49b⁺NK1.1^high cells as compared with ovx+E2 and sham female mice (p > 0.05; Fig. 4⇓, B and C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

E2 increases numbers of CD11c⁺CD49b⁺NK1.1^low, but reduces numbers of CD11c⁺CD49b⁺NK1.1^high cells (±SEM) in the spleen. Whole splenocytes were isolated from intact (sham), ovariectomized (ovx), and ovx females that had E2 replaced at low (2.5 mm) and high (5 mm) doses, enriched for CD11c, and stained for both CD49b (i.e., integrin subunit α2) and NK1.1 (i.e., NK receptor P1C). Using multiparameter FACS analyses, we gated on live CD11c⁺ cells and represent data as the proportion of CD11c⁺ cells that expressed CD49b and NK1.1^low/high (A and B). Representative dot plots are included (C). Bars with different letters indicate significant differences across groups; bars that share letters are not significantly different from each other (n = 10/group), p < 0.05.

To further illustrate the dependence of these changes on biologically relevant concentrations of E2, we implanted ovx females with a lower dose of E2 (i.e., 2.5 mm of E2 in 10 mm capsules). Similar to ovx females implanted with a higher dose (i.e., 5 mm of E2 in 10 mm capsules), ovx females exposed to a low dose of E2 also had elevated numbers of splenic CD49b⁺NK1.1^low cells and lower numbers of CD49b⁺NK1.1^high cells as compared with ovx female mice (p < 0.05 in each case; Fig. 4⇑).

17β-estradiol reduces numbers of splenic IKDCs

IFN-producing killer DCs are cells with surface expression characteristics of both DCs and NK cells (36). To further characterize the cell populations influenced by E2 as cDCs, IKDCs, or cells that share both cDC and IKDC markers, we stained and evaluated CD11c⁺ splenocytes from sham, ovx, and ovx+E2 mice for surface levels of MHCII, B220, CD8α, CD11b, CD11c, CD40, CD86, Gr-1, CD49b, NK1.1, CD122, and NKG2D (36). The CD49b⁺NK1.1^high cells, which were reduced in number by sustained E2 exposure, expressed appropriate levels of all the markers reportedly expressed on IKDCs, including CD11c, NK1.1, CD49b, CD122, MHCII, and B220 (Fig. 5⇓A) (36). The CD49b⁺NK1.1^low population of CD11c⁺ cells, which was increased in number by exposure to E2, did not express the NK cell markers CD122 or NKG2D, but did express MHCII, CD11b, B220, CD8α, Gr-1, and the costimulatory molecules CD40 and CD86 (Fig. 5⇓A), all of which are markers used to categorize DC subsets (30, 40). The CD49b⁺NK1.1^high cells had the morphological appearance of IKDCs, including large distinct nuclei and a very small cytoplasm (Fig. 5⇓B) (36). Conversely, the CD49b⁺NK1.1^low population was comprised of two predominant phenotypes: 1) cells with distinct nuclei and a large flowing cytoplasm, characteristic of cDCs (41); and 2) cells with distinct nuclei and a very small cytoplasm, typical of pDCs and IKDCs (Fig. 5⇓C) (36). These data suggested that estrogen-responsive CD49b⁺NK1.1^low cells are DCs that share many of the classic features of DC subsets, including IKDCs.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

CD11c⁺CD49b⁺NK1.1^high cells have phenotypic and morphological characteristics of IKDCs. Whole splenocytes were isolated from the spleens of sham, ovx, and ovx+E2 female C57BL/6 mice, enriched for CD11c, and stained for the following surface markers: CD49b, CD122, NKG2D, Gr-1, MHCII, CD11b, B220, CD8α, and CD86 (A). Morphological analysis using thick section microscopy illustrates the appearance of NK1.1^high (i.e., IKDCs) (B) and NK1.1^low (i.e., cDCs and IKDCs) (C) cells.

17β-estradiol increases IFN-γ production by CD11c⁺ splenocytes

Although IKDCs and NK cells comprise only a small percentage of the CD11c⁺ cells in the spleen, they are both potent producers of IFN-γ (35, 36). Therefore, we hypothesized that the effects of E2 on CD49b⁺NK1.1⁺ populations may result in functional changes in IFN-γ production by CD11c⁺ cells. CD11c⁺ splenocytes were isolated from ovx, sham, and ovx+E2 females and stimulated with either rIL-12 and rIL-18 or media alone (42). When stimulated with rIL-12 and rIL-18, CD11c⁺ splenocytes produced elevated concentrations of IFN-γ in the presence of high E2, despite the concomitant decrease in the percentage of CD49b⁺NK1.1^high cells (p < 0.05; Fig. 6⇓A). These data suggest that the E2-induced enhancement of CD49b⁺NK1.1^low cells contribute to increased IFN-γ production.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Production of IFN-γ is enhanced by exposure to E2 in vivo. Whole splenocytes were isolated, enriched for CD11c⁺ cells, and stimulated with the indicated TLR ligands or rIL-12 and rIL-18 for 24 h. Supernatants were used to measure the synthesis of IFN-γ (A), IL-12p40 (B), and IFN-α (C) by ELISA. Bars (mean concentrations ± SEM) with different letters indicate significant differences across groups; bars that share letters are not significantly different from each other (n = 10/group), p < 0.05.

Because CD11c⁺ DCs can produce IL-12 and IFN-α, we assessed the effects of E2 on the production of these cytokines. Splenocytes were isolated from ovx, sham, and ovx+E2 females, enriched for CD11c, and stimulated with LPS, poly I:C, CpG, or media alone. Although CD11c⁺ cells produced elevated levels of IL-12p40 following stimulation with LPS (p < 0.05), manipulation of E2 did not affect IL-12p40 production (p > 0.05; Fig. 6⇑B) and IL12p70 concentrations were below the limits of detection in all treatment groups. To ensure that the lack of effect of E2 on IL-12 production was not the result of functional or phenotypic changes induced during the positive selection procedure, CD11c⁺ splenocytes were isolated using a negative selection protocol. Similarly, manipulation of E2 in vivo did not affect production of IL-12p40 in negatively selected CD11c⁺ cells (p > 0.05; data not shown). Although poly I:C induced higher production of IFN-α by CD11c⁺ cells than did CpG, there was no enhancing or inhibitory effect of E2 on IFN-α secretion (p > 0.05; Fig. 6⇑C). These data indicate that E2 does not alter the activity of CD11c⁺ cell populations that produce either IL-12 or IFN-α (30, 40).

CD11c⁺ cells do not exhibit cytolytic activity

To further characterize the functional effects of E2 on CD11c+ splenocytes and to confirm that these CD11c⁺ cells were not NK cells, lytic activity was assessed. In the present study, CD11c⁺ splenocytes did not engage in significant killing of YAC-1 cells, regardless of hormone treatment (data not shown).

The effects of E2 on mature CD11c⁺ cells are partially mediated by ERα

There are two subtypes of the ER, ERα and ERβ; ERα, however, is the receptor subtype that is required for most of the known estrogenic effects (43). Both receptor subtypes are present in the thymus and spleen as well as in circulating lymphocytes, DCs, NK cells, and macrophages (7, 8, 9, 10). ER-mediated transcription can be examined using mice deficient in either ERα (Esr1^−/−) or ERβ (Esr2^−/−) (43). To test whether the effects of E2 on CD11c⁺ splenocytes were mediated by signal transduction activity through ERα, splenocytes were isolated from Esr1^−/− and WT C57BL/6 female mice, enriched for CD11c, and stained for CD49b and NK1.1 markers. Esr1^−/− females had significantly smaller uterine horns than their WT female counterparts, illustrating that the biological effect of circulating E2 was diminished in these females (p < 0.05; Fig. 7⇓A). Despite the effectiveness of the ERα deficit on reproductive development, the absence of ERα did not significantly alter populations of CD11c⁺CD49b⁺NK1.1^low or CD11c⁺CD49b⁺NK1.1^high cells as compared with WT females (p > 0.05; Fig. 7⇓, B and C). There was, however, a tendency for reduced numbers of CD11c⁺CD49b⁺NK1.1^low cells in Esr1^−/− as compared with WT females (Fig. 7⇓). Synthesis of IFN-γ by CD11c⁺ cells also did not differ between Esr1^−/− and WT female mice (p > 0.05; data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Deletion of ERα (Esr1^−/−) does not significantly affect populations of NK1.1^low or NK1.1^high cells in the spleen. Uterine horns (±SEM) from Esr1^−/− and WT female mice were weighed (A). Whole splenocytes were isolated from the spleens of Esr1^−/− and WT female mice, enriched for CD11c, and the percentage of CD11c⁺ CD49b⁺ NK1.1^low (B) or CD11c⁺ CD49b⁺ NK1.1^high (C) cells were determined (±SEM). Bars with different letters indicate significant differences across groups (n = 5/group), p < 0.05.