| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Immunology |
Department of Biomedical Sciences,3 Ontario Veterinary College, University of Guelph, Guelph, Ontario N1G 2W1, Canada
Department of Host Defense,4 Research Institute for Microbial Diseases, Osaka University, Osaka 565-0871, Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
are induced. Uterine NK cell precursors appear to home from secondary lymphoid organs to decidualizing uteri and localize mesometrially to the central decidua basalis, the site of maternal arterial modification at Gestation Days (gd) 9.5–10. In mice, genetic absence of uNK cells results in absence of pregnancy-induced spiral artery modification. Administration of IFN- to uNK-negative pregnant females induces arterial modifications without fetal loss. In this study, we investigated the roles of cytokines, known in other tissues to differentiate and activate NK cells, in induction of IFN- production in normal mouse implantation sites. Fecundity evaluation, implantation site morphometry, and IFN- quantification in interleukin (IL)-12p400/0, IL-180/0, dual IL-12p400/0/IL-180/0 and congenic strains revealed the importance of both IL-12 and IL-18 in the induction of spiral artery modification and IFN- synthesis. Immediately after implantation, IL-18 was localized transiently to decidual cells, but by gd8, IL-18 was produced solely by uNK cells, suggesting that early uNK cells are activated by stroma and lymphocyte-derived signals maintain later uNK cell activation. Mesometrial tissue of C57Bl/6J mice was examined by reverse transcription polymerase chain reaction assay in virgin, early postimplantation, and midgestation females for expression of the heterodimeric cytokines IL-23 (composed of IL-12p40 and a novel 
chain), IL-27 (composed of two IL-12-related chains) and IL-27R. No expression was detected in virgin uteri. The four genes were induced by gd6, and uNK cells isolated from midgestation transcribed IL-23 and IL-27R. This study advances the understanding of uNK cell activation during normal pregnancy.
decidua, growth factors, immunology, pregnancy, uterus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
have been reported (V.A. Paffaro et al., personal communication) [4]. Peak uNK cell numbers [5] and concentrations of IFN- [4] are achieved at gd 10, then both decline. In mice genetically ablated in IFN-, IFN- receptor (R) , or the IFN- signalling molecule STAT-1, uNK cells are abundant, but spiral arteries (branches of the uterine arteries leading into the placenta) do not undergo pregnancy-induced modification [6]. In mice genetically depleted of NK and uNK cell lineages (tg
26, RAG-20/0/common cytokine chain gamma (c)0/0, or IL-150/0), similar failure in spiral artery modification is seen (V.A. Paffaro et al., personal communication) [6, 7]. Most of the IFN- in mesometrial tissue at gd 6–12 is uNK cell derived [4] and treatment of pregnant uNK-negative mice with IFN- induces spiral artery change [6]. These data indicate that gene regulation by IFN- is an important feature in the maternal environment immediately after implantation during normal gestation.
Interleukin (IL)-12, a heterodimeric cytokine composed of p40 and p35 chains, is the most important stimulus for upregulation of IFN- mRNA expression and IFN- production in different NK cell subsets [8]. IL-12 was initially recognized as an inducer of IFN- synthesis in resting human peripheral blood mononuclear cells in vitro [9]. It is now recognized that IL-12 induces IFN- production in both resting and activated human T and NK cells in vivo [10, 11]. Macrophages and dendritic cells are the major sources of IL-12 in most tissues [12]. Preliminary reports suggest that IL-12 mediated signals contribute to normal uNK cell functions [13, 14], but IL-12 has also been associated with adverse gestational outcomes [14, 15].
IL-18, originally called IFN- inducing factor, is a cytokine of 18 kDa synthesized by Kupffer cells and activated macrophages [16, 17]. Synergy has been observed between IL-12 and IL-18 in induction and release of IFN- [18]. Based on immunohistological studies, a cascade of IL-18-enhanced IL-12 induction of uNK cell production of IFN- has been proposed to involve proliferative endometrial stroma [19]. However, the relative importance of IL-12 and IL-18 has not been examined directly. Mice doubly deficient for IL-12p40 and IL-18 are viable and fertile [20], but their implantation sites have not been analyzed for IFN-, spiral artery pathology, or the presence of other cytokines associated with induction of IFN-.
Recently, IL-23 and IL-27 were reported to contribute to the proliferation and activation of IFN- production in NK cells and T cells [21–25]. IL-23 is an IL-6 family member composed of the IL-12p40 soluble subunit and an IL12-p35 related chain named p19. The p19 chain is inactive biologically without its partner [21, 22]. IL-23 signals through IL-12Rß1 and a novel IL-23R chain [23]. IL-27 is composed of two chains, p28 and EBI3, analogous to IL-12p40 and IL-12p35, respectively. Engagement of the IL-27R (TCCR) results in IFN- production [24, 25].
To address the roles of IL-12 and IL-18 in activation of uNK cells, histological analysis of implantation sites from IL-12p400/0 (IL-120/0), IL-180/0, and IL-120/0/180/0 mice was undertaken from gd 6 to gd 12. Mesometrial tissues were assessed by morphormetry for ratios of tissue regions, uNK cell numbers, and spiral artery modification and by ELISA for IFN- production. Expression of IL-18 protein was addressed in virgin and pregnant uteri from genetically normal mice by immunohistochemistry. Expression of IL-23, IL-27, and IL-27R message was assessed by reverse transcription polymerase chain reaction (RT-PCR) using RNA prepared from uteri of virgin and pregnant IL-120/0/180/0 and B6 mice and using RNA prepared from isolated uNK cells collected at gd 10. The results indicate that IL-12 and IL-18 have nonredundant roles in spiral artery modification and that IL-23 and IL-27 are induced in the uterus after conception.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mice genetically ablated for IL-12 and IL-18 on a C57Bl/6J (B6) background (IL-120/0/180/0) were shipped as heterozygote breeding pairs from Japan to Guelph and housed under barrier conditions (OMAFRA Isolation Unit, Guelph, ON, Canada). Intercrosses were used to establish four sublines: IL-12+/+/18+/+ (used as normal congenic controls and referred to as wildtype [wt] to differentiate them from B6), IL-120/0/18+/+ (IL-120/0), IL-12+/+/180/0 (IL-180/0), and IL-120/0/180/0. Genotypes of the mice were obtained by PCR assay using tail DNA. For IL-120/0 mice, primers a (CAC TTG CCA AAC TCC TGT GAG CTA TGA) and b (TTC TTG TGG AGC AGC AGA TGT GAG TGG) were used to detect the wt allele, and primers b and c (ATC GCC TTC TAT CGC CTT CTT GAC GAG) were used to detect the mutated allele. For IL-180/0 mice, primers d (TAA TGG GTG GTC TTC TCA TCT CTG TGT) and e (TTG CTG CAC CTA GAG GTA TGT ACT GAC) were used to detect the wt allele, and primers e and c were used to detect the mutated allele.
The PCR conditions were hot start at 94°C for 4 min, 35 cycles of 94°C for 30 sec, 67°C for 30 sec, and 74°C for 1 min, then 74°C for 7 min prior to cooling to 4°C.
Mice on a B6 background and genetically deficient in IFN- (IFN-0/0) were purchased from Jackson Laboratory (Bar Harbor, ME) and housed under barrier conditions. Genetically normal B6 mice were purchased from the same supplier but were housed under conventional conditions and were used as normal controls in some experiments. For timed matings, females were selected for estrus, paired overnight with males of the same genotype, and checked for copulation plugs the following morning. The morning of plug detection was named gd 0. Mice were killed by CO2 asphyxiation and subsequent cervical dislocation if the tissues were for cytokine or RNA analyses. For collection of samples for histology, cervical dislocation was omitted and perfusion of the body was undertaken using 4% paraformadehyde and 0.1 M sucrose in PBS, pH 7.4 (PFA). Fetal viability, judged by color and size of implantation sites, ranged from 83% to 92%. Only viable implantation sites were used in analyses. All procedures were conducted under Animal Utilization Protocols approved by the Animal Care Committee of the University of Guelph and compliant with Guidelines of the Canadian Council for Animal Care.
Histological Procedures
Uteri from perfused mice were transected into individual implantation sites that were immersion fixed in PFA for another 15 min (<gd 11) or 60 min (
gd 11) and then transferred to 70% ethanol and embedded in paraffin. Blocks were transversely serially sectioned at 7 µm. All sections were stained by standard protocols for haematoxylin and eosin (H&E), periodic acid-Schiff, or DBA lectin [26]. The latter two techniques stain cytoplasmic granules of uNK cells; DBA lectin also stains the uNK cell plasma membrane. Numbers of uNK cells per square millimeter of tissue were measured on 11 sections from the middle of each implantation site and averaged [4]. Sections were 49 µm apart to avoid duplicate uNK cell counting (cell diameter is 
40 µm). At least two different implantation sites from two or more dams were used for each mean presented. Spiral artery wall and lumen surface areas were measured on adjacent sections that were stained with H&E, and vessel area:lumen area ratios were calculated. Surface areas of placenta, decidual basalis (DB), and mesometrial lymphoid aggregate of pregnancy (MLAp), the mural site of uNK cells from gd 10 to gd 12, were also measured on these slides. Measurements were made using OPTIMAS TM image analysis software (version 6.5; Optimas Corp., Media Cybernetics, Silver Spring, MD).
Tissue Acquisition for IFN- Quantificaton
To quantify IFN-, mesometrial decidua (gd 6–8) or MLAp (gd 10, 12, and 14) were dissected and minced using razor blades in Petri dishes on ice. The dissected region does not contain fetal cells on the these gestation days. Mesometrial sides of nonpregnant virgin uteri were also collected. Minced samples were pooled by litter in 1.5-ml microcentrifuge tubes containing 100 µl RPMI 1640 medium with 10% fetal calf serum (FCS) and immediately homogenized using a micropestle (Fisher Scientific, Nepean, ON, Canada). Samples were centrifuged (800 x g, 4°C, 5 min). Supernatants were collected and stored at -20°C until analyzed for IFN-.
Quantification of IFN- by ELISA
Capture antibody (purified rat anti-mouse IFN- antibody; Pharmingen, Mississauga, ON, Canada) at a concentration of 1 µg/ml in coating buffer (0.1 M Na2HPO4) was added in a volume of 50 µl to each well of the ELISA plate (Dynex Technologies Inc., Chantilly, VA). Plates were sealed with plastic film, incubated overnight at 4°C, and then washed, blocked (10% FCS in PBS), and emptied. Samples and recombinant mouse IFN- standards (Sigma, St. Louis, MO) were added to triplicate wells (100 µl/well), and plates were incubated (4°C overnight). After washing, 50 µl of 0.5 µg/ml biotinylated detection antibody (XMG1.2 rat anti-mouse IFN- antibody; Phamingen) was added for 60 min at room temperature (RT), followed by 100 µl horseadish peroxidase avidin D (1:3000; Vector Laboratories, Burlingame, CA) for 30 min at RT. Substrate (100 µl; 2,2'-azino [3-ethylbenzthiazoline-6-sulfonic acid], 1 mg/ml; 0.003% H2O2; 0.1 M citric acid) was added to each well. After 60 min at RT, absorbance was measured at a wavelength of 405 nm. IFN- concentrations were determined using a SAS program against serially diluted recombinant mouse IFN- (Sigma). To calculate IU/implantation site, the total units of IFN- found in the samples was correlated with the final sample volume and the number of recorded implantation sites pooled to create the sample.
Immunohistochemistry for IL-18
To detect expression of IL-18 protein in the early stages of murine pregnancy, virgin uterus and sections of implantation sites from IL-12+/+/18+/+ mice were studied by immunohistochemistry at gd 3, 5, 6, 8, and 10. After deparaffinization and rehydration, antigen was retrieved by immersion of slides in sodium citrate buffer (0.01 M, pH 6.0, 95°C, 3 min). Peroxidases were inactivated by applying 3% H2O2 solution for 6 min, and blocking serum (1% BSA in PBS) was then applied for 20 min and removed before the primary anti-IL-18 antibody (1:50 optimal dilution from titration studies, goat polyclonal IgG; Santa Cruz Biotechnology, Santa Cruz, CA) was applied for an overnight incubation at 4°C. After washing, slides were incubated for 30 min with biotin-conjugated secondary antibody (rabbit anti-goat IgG; Jackson ImmunoResearch Laboratory, Warrington, PA) and then treated with Extravidin-peroxidase (1:100; Sigma) for 30 min. After washing with PBS, slides were incubated in fresh 3,3'-diaminobenzidine (Sigma) solution for 2 min, rinsed in tap water, counterstained with Harris haematoxylin for 10 sec, dried, and mounted with coverslips using Permount (Fisher Scientific). Normal goat serum at the same dilution was used for the negative control for the primary antibody.
Acquisition of Samples for RT-PCR Analysis of IL-23, IL-27, and IL-27R
Virgin uteri and implantation sites harvested from B6 and IL-120/0/180/0 mice were used for RNA isolation. The gd 4 and gd 5 samples were dissected decidual swellings that included conceptuses. The gd 6 samples were dissected DB, and the gd 10 samples were dissected into DB, MLAp, and placenta (largely fetal trophoblast). RNA was isolated from these tissues using the RNAEasy Mini kit (Qiagen, Mississauga, ON, Canada), followed by reverse transcription (First-strand cDNA Synthesis Kit; Amersham Pharmacia Biotechnology, Nutley, NJ) according to the manufacturer's instructions. The following primers were used for RT-PCR: IL-23, AAT AAT GTG CCC CGT ATC CA (forward) and CTG GAG GAG TTG GCT GAG TC (reverse); IL-27, AAC TCC ACC AGA TCC ACG TC (forward) and AGC GGA GTC GGT ACT TGA GA (reverse); IL-27ß, TGT TTC CCT GAC TTT CCA GG (forward) and GGG GCA GCT TCT TTT CTT CT (reverse); IL-27R, TGA AGC CAG ACA CAC CTC AG (forward) and CAC ACA AGG TCT TGG GTC CT (reverse); ß-actin, GCT ACA GCT TCA CCA CCA CA (forward) and ACA TCT GCT GGA AGG TGG AC (reverse).
The predicted sizes of correct products are 213 base pairs (bp) for IL-23, 284 bp for IL-27, 340 bp for IL-27ß, 407 bp for IL-27R, and 477 bp for mouse ß-actin. PCR conditions were hot start at 94°C for 6 min, 35 cycles of 94°C for 45 sec, 57°C for 45 sec, and 72°C for 45 sec, and then 72°C for 7 min prior to cooling to 4°C.
To obtain RNA from uNK cells, dissected MLAp and DB from PBS-perfused gd 6 B6 females were finely minced, briefly dissociated in 1000 IU DNase in 2% BSA, filtered through an 80-µm mesh, and mixed 30 min with DBA lectin-conjugated magnetic beads. A magnet was applied, and the retained cells were washed three times using PBS and detached from the beads using a rinse of 0.1 M N-acetyl-D-galactosamine (Sigma), the specific sugar ligand of DBA lectin [27]. Uterine NK cells, with a purity of >98%, were prepared on two separate occasions and then used as above for RNA isolation and RT-PCR analysis.
Statistical Analysis
Results were analyzed by ANOVA using the SAS statistical program (SAS Institute, Cary, NC). Results were considered significant at P < 0.05. All values are shown as mean ± SD.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Litter size at birth and sex ratios of offspring at weaning (3 wk age) were recorded over 18 mo of breeding for IL-120/0, IL-180/0, and IL-120/0/180/0 mice and their congenic genetically normal counterparts (Table 1). Only IL-120/0 mice differed significantly from the wt mice in numbers born (fewer). For females killed during gestation, fetal resorption was occasionally observed. Litter sizes ranged from 1 to 15. Resorptions were significantly higher in the IL-120/0 strain than in the other strains (Table 1).
|
Assessment of Implantation Sites in Four Substrains
Pregnancies were analyzed at gd 10 and gd 12 when uNK cell densities in DB and MLAp peak [5] and pregnancy-induced spiral artery modification is expected to have been completed [6, 7]. Implantation sites in the four strains had an overall normal appearance (Fig. 1), and morphometric measurements of the surface areas of the placentae, DB, and MLAp showed no differences between the strains (data not shown). Uterine NK cells were numerous within implantation sites of all mice, and at both gd 10 and gd 12, more uNK cells/mm2 were present within the MLAp than within the DB. No differences in uNK cell numbers were found among the four strains in either DB or MLAp at gd 10 or gd 12 (Fig. 2).
|
|
Triggering of decidual spiral artery dilation is largely mediated by IFN- derived from uNK cells in normal mice at gd 9.5–10. Compared with the normal congenic strain, spiral artery vessel area:lumen area ratios for the three gene-ablated strains were significantly increased at each time point (P < 0.05; Figs. 1 and 3). Vessel area:lumen area ratios for the double knockout (KO) strain were not different from those of either of the single KO strains. Vessel area:lumen area ratios were also measured on archived gd 12 specimens from IFN-0/0 mice [6]. These vessels were the most constricted in the study and were significantly different from those in all other strains (Fig. 3).
|
IFN- Quantification
The observation that spiral artery vessel area:lumen area ratios in IL-120/0, IL-180/0, and IL-120/0/180/0 pregnancies were larger than that in B6 mice suggested that IFN- concentrations may be reduced in implantations sites of the KO mice. Therefore, concentrations of mesometrial IFN- were measured in homogenates of freshly dissected mesometrial tissues. IFN- was absent from homogenates of virgin uteri but was detected at every gestation day examined (gd 6–14). IFN- concentration per implantation site rose from gd 6 to a peak at gd 10 and dropped at gd 12 and 14 in all strains except IL-120/0/180/0, in which the peak concentration was achieved at gd 12 and then declined (Fig. 4). The concentrations of IFN- at gd 6 and gd 8 did not differ among strains. At gd 10 and gd 12, the IFN- concentration of 3 U/implant for the double KO strain was lower than that for all other strains. At gd 14, DB from all of the mutant mice had less IFN- than did the congenic normal strain.
|
Immunohistochemistry for IL-18
IL-18 is produced by several types of cells. If IL-18 plays a role in uNK cell activation, IL-18 should be produced very early during decidualization and in regions of uNK cell precursor recruitment and proliferation. IL-18 protein was not detected in virgin or preimplantation (gd 3) uteri using immunohistochemistry (Fig. 5, A and B). Very strong IL-18 labeling occurred throughout the decidua at gd 5 (Fig. 5C). However, this signal was transient, and stromal cell reactivity declined noticibly by gd 6 (Fig. 5D) and was absent by gd 8 (Fig. 5E). From gd 6 to gd 14, lymphocytes in the mesometrial triangle (gd 6 and gd 8) and in the MLAp and DB (gd 10) were the only cells that expressed IL-18 (Fig. 5F). Using DBA lectin to stain adjacent sections, the reactive cells were confirmed to be uNK cells (Fig. 5, H–J).
|
RT-PCR for IL-23, IL-27, and IL-27R
To evaluate whether IL-23 might contribute to uNK cell activation, chain (p19) message was assessed in B6 virgin and pregnant uteri at gd 4, 5, and 6. The same tissues were used to evaluate transcription of IL-27 and its receptor. None of the four genes were transcribed in mesometrial tissue of the virgin uterus, but all were induced by gd 4, the earliest time point decidua is present in mouse implantation sites. Transcription was sustained at gd 5, when the entire decidual capsule was studied. At gd 6, the earliest time sufficient decidua is available to microdissect, message for IL-27 was no longer detectable (Fig. 6A).
|
To assess whether uNK cells themselves transcribe any of these genes, the RT-PCR assay was repeated using purified uNK cells. Tissue from gd 10 was studied because sufficient cells could not be obtained at earlier times by the magnetic bead separation procedure. Uterine NK cells were positive for IL-23 and IL-27R but negative for IL-27 and IL-27ß (Fig. 6B).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
in mesometrial tissues of the mouse uterus during normal pregnancy. Previous work had established that uNK cells localize to the mesometrial region during decidualization [3] and are the major but not exclusive cell type producing IFN- [4, 6, 28]. The mesometrial region is the vascular portal to the uterus, and within each implantation site branches of the maternal uterine arteries become modified as they dilate and elongate to support fetoplacental development [29]. Maternal vessels crossing the DB are highly coiled and can be referred to as spiral arteries. In contrast to spiral arteries in human implantation sites, those in mice at midgestation are not associated with intraluminal or intramural trophoblast, except in the proximal 150 µm, but appear to rely on intravascular, intramural, and perivascular uNK cells when remodeling [6, 7, 29, 30]. IFN- is not detectable by ELISA in uteri prior to decidualization [4] but is induced during this process. In the four strains of mice in the current study, 1–2 IU IFN- per implantation site was measured at gd 6 (Fig. 4). This finding is consistent with concentrations measured in three other mouse strains: random-bred CD1, BALB/cJ, and outbred scid/scid [4]. IFN- concentrations of 2–3 IU/implantation site were also concordant between the strains at gd 8. In the previous study, peak IFN- concentrations of 6–8 IU/implantation site were found on gd 10, followed by a decline [4]. In the present study, peak concentrations in the normal B6-related mice were 6 IU/implantation site at gd 10 followed by decline. For IL-120/0 or IL-180/0 mice, 4–5 IU/implantation site was found, but these concentrations were not significantly different from those in the wt strain. Only the concentration of 3 IU/implant for the double KO mice was significantly lower than that for the control. This finding suggests that induction (IL-12) and enhancement (IL-18) of IFN- production in the implantation site may be partially compensatory.
A gestational rise in IFN- occurs in the absence of both cytokines. In a previous study, Ashkar and Croy [4] showed that in the absence of uNK cells, some IFN- is induced at mouse implantation sites (<1 IU/site), but this level remained constant across gestation. Gains in IFN- in the IL-120/0/180/0 pregnancies suggest there are alternative pregnancy-sensitive mechanisms in murine uteri for IFN- induction. Demonstration of IL-23 (p19) and IL-27 and IL-27ß transcription offers potential alternative pathways for uNK cell activation.
IL-12 (p40 plus p35) is the major inducer of IFN- synthesis. IL-12 has been demonstrated in murine decidua during normal pregnancy by an antibody recognizing the heterodimer [14] but has also been associated with pathologies such as pregnancy loss and pre-eclampsia [14, 15, 31–33]. In contrast (Table 1), absence of IL-12 was the only condition reducing fecundity in our study, possibly because of a compensatory gain in levels of other cytokines. Because IL-23 uses the p35 chain of IL-12, studies of p35 were not conducted and its presence in uterine decidua was assumed from published information [14]. The unique IL-23 chain was induced in the uterus of B6 mice by gd 4 and sustained to gd 6, the last day investigated. In mice lacking both IL-12 and IL-18, expression of IL-23 (p19) was not detectable in virgin uteri, and its induction appeared similar in time course to that in B6 mice. Further analysis by real-time PCR would be needed to reveal whether there is compensatory elevation in IL-23 message due to the absence of IL-12 and IL-18. IL-23 maybe a uNK cell product, like IL-18 and IFN-. No other studies have addressed uterine p19. Wiekowski et al. [34] reported that most p19 KO mice die in infancy, but those surviving to adulthood were infertile. All p19 KO mice died by 90 days of age.
IL-27 does not share chains with IL-12 but has chains highly homologous to IL-12p40 and IL-12p35 [25]. These chains were not expressed in virgin uteri from B6 or IL-120/0/180/0 mice but were induced at gd 4 and were sustained. The IL-12p40-related chain of IL-27 was initially characterized in human placental trophoblast cells [35] as EBI3 (Epstein-Barr virus-induced gene 3). Protein expression was shown in syncytiotrophoblast and extravillous trophoblast in all trimesters and was particularly strong in intramural trophoblasts of the spiral arteries. Additional cells were reactive with the antibody, but their morphology suggested that they were not dendritic cells, although this cell population is positive in other tissues. Cells reactive for EBI3 coexpressed IL-12p35. Villous cytotrohoblasts were EBI3 negative, although EBI3 could be induced in cultured trophoblast cell lines. Devergne and her colleagues [35] further showed that EBI3 rose in the plasma of pregnant women from Week 9 of gestation to term but most EBI3 was not associated with IL-12p35. They postulated that EBI3 had a different partner (subsequently identified as IL-27), that EBI3+ cells interacted with the adjacent CD56+ uNK cells in their decidual sections, and that EBI3 was the most frequent peptide presented by soluble or membrane-bound human leukocyte anitigen G [35] and thus must play a role in induction of maternal tolerance to the fetus. Study of implantation sites in mice ablated for the IL-27R TCCR [24] could be used to validate and extend these hypotheses. Our studies indicate that IL-27 and its receptor are both induced during early decidualization. Absence of IL-27 from uNK cells at gd 10 accompanied by expression of IL-27R suggests that IL-27 may be produced by stromal cells that signal to receptor-bearing uNK cells. P28 is expressed at a low level in 28-wk human placenta, the only time point reported [25], but has not been previously reported in mice. Our studies suggest its expression is very transient, with a peak duration of 48 h and then a decline. Further work, such as real-time PCR combined with in situ hybridization, will be required to establish whether this interpretation is artifactual because of inclusion of antimesometrial decidua and conceptus tissues in our gd 4 and gd 5 dissections but their exclusion at gd 6.
Analysis of the spiral arteries in the mouse strains lacking IL-12 and IL-18 suggest modest IL-23- and IL-27-mediated effects. If either or both of these cytokines could functionally replace IL-12, normal pregnancy-induced arterial modification would be predicted in the IL-120/0 mouse, which was not observed. Given the induction of IFN- in the three KO mouse strains and its increasing concentrations to midgestation, failure to modify the spiral arteries was not anticipated. One explanation is that spiral artery modification requires >5 IU of IFN- per implantation site, which was achieved only in the control strain. Alternatively, all of the IFN- detected in the tissue homogenates may not be functionally available in vivo, and the genetic cytokine deficiencies may result in less bioavailability. The small decline in vessel area:lumen area ratios in the IL-12/IL-18 series of mutant mice compared with IFN-0/0 females suggests that IFN- may be functionally available at borderline levels but not at levels needed for robust modification.
Numbers of uNK cells in implantation sites of the KO mice were similar to those in controls. This finding confirms the presence of functional IFN-, which must be >1 IU/implant site to promote terminal differentiation and apoptosis in uNK cells [6]. Transplantation studies have shown that differentiation of uNK cells within implantation sites requires only precursor cells plus IL-15 (B.A. Croy, unpublished data) [36]. Activation of the cells requires additional signals [37]. Most NK cell signaling receptors are inhibitory, and their engagement prevents lytic function [38, 39]. In a viral model system [40], activation receptors were functionally dominant over the inhibitory receptors in vivo, but only when NK cells were found in a proinflammatory environment. Expression of IL-18 in >90% of gd 4 decidual stromal cells suggests that any uNK cell precursors reaching the uterus would be placed into a proinflammatory environment and become activated to proliferate and produce cytokines. This environment appeared to be short lived (48 h), which may limit precursor recruitment or activation, thereby preventing septic shock-like consequences and outcomes such as pre-eclampsia [41]. However, spiral artery modification is not accomplished at this time. Thus, activated uNK cells may assume an autocrine role in sustaining the IL-18-based proinflammatory state for IFN- production.
This series of experiments has clearly shown that signaling pathways established in the uterus during decidualization and normal pregnancy support induction of IFN- production by uNK cells, and a certain concentration of IFN- per implantation site is needed for induction of spiral artery modification.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence. FAX: 519 767 1450; acroy{at}uoguelph.ca
Received: 15 January 2003.
First decision: 4 February 2003.
Accepted: 28 February 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
contributes to the normalcy of murine pregnacy. Biol Reprod 1999 61:493-502This article has been cited by other articles:
|
|
 |
|
  J. L. Herington and B. M. Bany Effect of the Conceptus on Uterine Natural Killer Cell Numbers and Function in the Mouse Uterus During Decidualization Biol Reprod, April 1, 2007; 76(4): 579 - 588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J Muniz, M. M Joyce, J. D Taylor II, J. R Burghardt, R. C Burghardt, and G. A Johnson Glycosylation dependent cell adhesion molecule 1-like protein and L-selectin expression in sheep interplacentomal and placentomal endometrium. Reproduction, April 1, 2006; 131(4): 751 - 761. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Tayade, Y. Fang, G. P. Black, P. VA Jr, A. Erlebacher, and B. A. Croy Differential transcription of Eomes and T-bet during maturation of mouse uterine natural killer cells J. Leukoc. Biol., December 1, 2005; 78(6): 1347 - 1355. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Xie, H. He, M. Colonna, T. Seya, T. Takai, and B. A. Croy Pathways Participating in Activation of Mouse Uterine Natural Killer Cells During Pregnancy Biol Reprod, September 1, 2005; 73(3): 510 - 518. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Dosiou and L. C. Giudice Natural Killer Cells in Pregnancy and Recurrent Pregnancy Loss: Endocrine and Immunologic Perspectives Endocr. Rev., February 1, 2005; 26(1): 44 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. V. Villarino, E. Huang, and C. A. Hunter Understanding the Pro- and Anti-Inflammatory Properties of IL-27 J. Immunol., July 15, 2004; 173(2): 715 - 720. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS |
