1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Gene ablation studies in mice indicate that lymphotoxin (LT)α, LTβ and LTβR are essential for the genesis of lymph nodes (LN), normal structural development of peripheral lymphoid tissues and the differentiation of natural killer (NK) cells. LTβR binds to the heterotrimeric cytokines LTα1β2 and LIGHT. LTs also regulate stromal cell expression of lymphocyte homing chemokines. Uterine decidualization in normal (+/+) mice is accompanied by the appearance and maturation of large numbers of uterine NK (uNK) cells that differentiate from precursors mobilized to the uterus from secondary lymphoid tissues. uNK cells accumulate in a transient, lymphocyte-rich region known as the metrial gland or, more recently, the mesometrial lymphoid aggregrate of pregnancy (MLAp). To determine if LTs contribute to development of the MLAp, and to the differentiation and/or localization of uNK cells, a histological study was undertaken of implantation sites from LTα null, LTβR null and gestation day-matched, normal mice. Implantation sites from the gene-ablated mice contained abundant numbers of uNK cells that localized appropriately. This indicates that the stromally derived molecules supporting NK cell differentiation in the uterus differ from those used in secondary lymphoid organs.


C57Bl/6 J mice


decidua basalis


dolicholos biforus agglutinin


gestation day


haematoxylin and eosin


herpes virus entry mediator


intercellular adhesion molecule


lymph nodes




mesometrial lymphoid aggregrate of pregnancy


mesometrial triangle


natural killer


periodic acid-Schiff reagent


Peyer's patches






uterine NK cell


severe combined immunodeficiency


vascular cell adhesion molecule


vascular endothelial cell growth factor

V : L

vessel : lumen area ratios

L, wall area to lumen area ratio.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The immune system is segregated into primary and secondary lymphoid tissues. Many secondary lymphoid tissues are associated with mucosal tissue in the respiratory, gastrointestinal and urogenital tracts.1 Mice, ablated for genes of the tumour necrosis factor (TNF) family have provided important insights into the biogenesis of secondary lymphoid tissues, particularly LN and Peyer's patches (PP). Mice deleted in lymphotoxin (LT)α, a gene encoding a soluble homotrimer (LTα3), had severe impairment in LN differentiation and no PP development. LTα null mice also displayed abnormal histological structure of the spleen.1–3 LTα3 binds to TNF-R1, TNF-R2 and HVEM.4,5 Because ablation of TNF-R1 and TNF-R2 did not alter lymphoid tissue structure in mice,6,7 the effects of LTα gene deletion were attributed to the membrane (m) form of LT, LTα1β2. Studies involving neutralization or genetic ablation of mlT and its receptor LTβR supported this conclusion. LTβ null mice lacked PP and most LN, retaining only mesenteric and cervical LN development while LTβR null mice lacked all LN.1,8–10

From studies of PP, a three-step model has been advanced for the biogenesis of secondary lymphoid tissue. Step one is expression of vascular cell adhesion molecule (V-CAM)-1 and intercellular adhesion molecule (I-CAM)-1 by stromal cell clusters in the gut wall. These regions become infiltrated by interleukin (IL)-7+, CD4+, CD3 cells which act as inducers expressing LTα/β and subsequently recruiting T and B lymphocytes.11,12 Similar cells have been found in developing LN,13 where a role has been postulated for natural killer (NK) lineage cells in organ biogenesis.1 LTα null mice have a significant deficit in NK cells. The action of LT in promotion of NK cell development is independent of IL-1514,15 and of TNFR-114 and dependent on LTβR.15 Adoptive cell transfer studies and in vitro stromal cell-dependent lymphogenesis assays involving LTβR null, LTα null and congenic wild-type mice, have shown that signalling, via LTβR on marrow stromal cells acting through mlT, is a key step in very early NK cell development. This interaction preceded the action of IL-15 and was independent from the IRF-1/IL-15 pathway found in the bone marrow stromal environment.15

The uterus is a mucosal organ with secondary lymphoid development. In humans, uterine lymphoid aggregates are comprised of CD8+ T cells surrounding a core of B cells. These aggregates are found in the cycling uterus but disappear at menses and are not present during gestation.16 LTα is expressed in human pregnancy in both early and term endometrial stromal cells that, during gestation, are present as transformed decidual cells.17 LTα and LTβR mRNA are also transcribed by at least two types of human placental cells, fetal trophoblasts and macrophages.18 In rats and mice, lymphocytes cluster into aggregates during mid-pregnancy on the mesometrial side of the uterus at each implantation site and surround blood vessels that enter the uterus. By day 10·5 of gestation, the aggregates have developed into significant mural structures, known as mesometrial lymphoid aggregates of pregnancy (MLAp) or metrial glands.19 LTβ as well as TNF-R1 and TNF-R2 have been found in the pregnant, decidualized mouse uterus.20 The lymphocytes which fill the MLAp are NK lineage cells with several specialized features and they are referred to as uterine (u) NK cells or granulated metrial gland cells.18,21,22 A key known function of murine uterine natural killer (uNK) cells is the production of IFN-γ. uNK cell-derived IFN-γ regulates genes that initiate the elongation and dilation of the uteroplacental spiral arteries as they pass through the maternal decidua to the placenta. It also sustains uterine stromal cells that have been transformed into decidua, a tissue found only in species with haemochorial placentation.23 No information is available on the biogenesis of the MLAp and whether all procurser (pro)-NK cells or precurser (pre)-NK cells have the potential to become uNK cells. Thus, a histological study of implantation sites in pregnant LTα null and LTβR null females was undertaken to ask whether LTs play roles in the biogenesis of the MLAp and/or the differentiation and localization of uNK cells.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Animals, tissue dissection and processing

LTα null mice on the C57Bl/6J background and C57Bl/6J (B6) controls were purchased from the Jackson Laboratory, Bar Harbor, ME. LTβR null mice on the B6 background were shipped from Munich to Guelph. Genotypes of all strains were confirmed by polymerase chain reaction (PCR) analysis in Guelph. Mice were used between 8 and 12 weeks of age. Virgin females were selected for oestrus by vaginal examination and paired with males of the same genotype. Detection of a copulation plug the following morning was indicative of mating and the mouse was considered to be at gestation day (gd) 0·5. Mated females were killed using CO2 followed by cervical dislocation on gd 10·5, 12·5 and 14·5. At least three pregnant females were studied for each time-point in each genotype. All matings were conducted in Guelph, under approved animal utilization protocols.

A ventral abdominal skin incision was made on euthanized females and the skin was retracted to permit examination of sites where subcutaneous LN are normally found. Next, the abdominal cavity was opened and mesenteric LN were evaluated to confirm the genotypes, then the gravid uterus was dissected.

For isolation of RNA, uteri from B6 females at gd 10·5 were opened along the antimesometrial side and the fetus, amnion and yolk sack were removed and discarded. Each placenta with its overlying decidua basalis (DB) was retracted off the uterine wall, separated and placed into RNA lysis buffer in weighed tubes. The MLAp was then dissected from the uterine wall and placed into a separate weighed tube of lysis buffer. Despite using great care, these dissections are imprecise and decidual cells would contaminate (about 30% level37) in each of the other two tissues. The MLAp region would not contain any fetal cells. For histology, the gravid uterus was fixed in Bouin's fixative (VWR, Mississauga, ON) 12–16 hr and processed for routine paraffin embedding and histology.

Analysis of gene expression

RNA was isolated from the MLAp, DB and placenta of B6 mice using Qiagen Rneasy mini kits, according to the manufacturer's instructions. PCR was performed on 25 ng of first-strand cDNAs in a total volume of 15 µl containing 15 ng of each primer, 200 µm dNTPs, 1·5–3 mm MgCl2, 50 mm KCl, 10 mm Tris-HCl and 0·2 Units of Taq DNA Polymerase (Sigma, Oakville ON). The PCR conditions were 94° for 5 min, 35 cycles of 94° for 30 seconds, 57° for 1 min, 72° for 1 min, followed by a further 7-min extension at 72°. PCR products were then examined by electrophoresis through a 1·5% agarose gel with 1 × Tris-acetate-EDTA (TAE) buffer. Gels were stained with ethidium bromide, examined under ultraviolet illumination and photographed. The primers were:

  • LTα
  • 5
  • LTβ
  • 5
  • LTβR
  • 5

The predicted sized of the amplicons were LTα 229 base pairs (bp), LTβ 120 bp and LTβR 271 bp.

Histological and morphometric procedures

Seven micron serial sections were cut for at least three implantation sites from each female. Some sections were stained with either periodic acid Schiff's (PAS) reagent (Sigma) or biotinylated DBA lectin (Sigma, developed with streptavidin–peroxidase) which recognizes glycoproteins in the membranes and granules of uNK cells.24 Other sections were stained for general morphology using haematoxylin and eosin (H&E). All the serial sections were examined microscopically. The central section was identified for morphometric analysis, which was also applied to 10 other tissue sections from each implantation site using OPTIMAS™ software (Optimas Corporation, version 6·2). These sections were on either side of the central section and separated from each other by 42 µm, to avoid duplicate counting of uNK cells which reach ∼40 µm. Cross-sectional area measurements of MLAp, DB and fetal placenta were collected at 25 × magnification. Spiral arteries, cut in cross section, were measured at 100 ×, as described previously,23 to determine the vessel area and the lumen area and calculate vessel : lumen area ratios (V : L). Twenty measures of this ratio were made between the 11 tissue sections. On the same 11 sections from each implantation site, uNK cells were enumerated using an ocular grid at a magnification of 400 ×. Only uNK cells showing a nucleus were counted. One mm2 was scored for uNK cell content in each of two regions, the MLAp and the DB, of every tissue section enumerated. To estimate cell diameter or granule number, 30 cells were measured. Data from the 11 sections of each embryo were averaged and data from the nine or more implant sites/genotype at each time-point were pooled. Statistical comparisons were made by anova and probability of < 0·05 was considered significant.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

LTα, LTβ and LTβR expression in mid-gestation implantation sites from normal mice

Expression of LTα, LTβ and LTβR was demonstrated by reverse transcriptase-ploymerase chain reaction (RT-PCR) in all of the mesometrial regions found in normal B6 mice at gd 12·5 whether these tissues represented fetal (placenta) or maternal tissue [decidua basalis (DB) and mesometrial lymphoid aggregate of pregnancy (MLAp)] (Fig. 1).


Figure 1. PCR analysis of cDNA from various tissues of gd 10·5 pregnant B6 mouse uterus: 100 bp ladder, lane 1. Blank water controls in place of cDNA, lanes 2, 7, 12 and no samples applied in lanes 6 and 11. LTα (229 bp amplicon): DB, lane3; MLAp, lane 4; placenta, lane 5; LTβ (120 bp amplicon: DB, lane 8; MLAp, lane 9; placenta, lane 10; LTβR (271 bp amplicon): DB, lane 13; MLAp, lane 14; P = placenta, lane 15.

Download figure to PowerPoint

Microarchitecture of implantation sites in pregnant LTα null and LTβR null mice

Implantation sites from LTα null and LTβR null mice had only minor differences to implantation sites from normal mice (Fig. 2). These minor differences were similar in both gene-deleted strains. Ablation of the actions of LTα3 or of LTα1β2 did not block development of the MLAp. Lymphocyte aggregation occurred in the appropriate location at the base of each placenta and it separated the myometrium between the circular and longitudinal smooth muscle layers. uNK cells were abundant in the MLAp and DB of the two gene ablated strains and in C57Bl/6 (Fig. 3). DB was less compact in both gene-deleted strains and this appeared to be due to a delay in the transformation of fibroblasts into decidual cells. DB appeared to have an increase in extracellular matrix and, in some females, the decidual cells were separated widely by proteinaceous fluid (Fig. 4). By gd 14·5, these histopathological differences between implantation sites in the knockout and control animals were diminished. Morphometrically, the area occupied by the MLAp increased as gestation proceeded in all strains and there were no differences between the strains. The area occupied by DB declined in all strains between gd 10·5 and 14·5 but DB in the control mice at gd 14·5 had the only significantly smaller area. The area occupied by median sections of the fetal placenta increased as pregnancy progressed in LTα null and control mice and was the same in all three strains at gd 10·5 and 14·5. In LTβR null mice, the gd 12·5 placenta was unexpectedly large and occupied an area significantly different to that of the other two strains (Table 1).


Figure 2. Low power histology of median tissue sections from gd 10·5 implantation sites in LTβR null (a) and B6 mice (b) showing the general structural microanatomy of the mesometrial side of the uterus. MLAp is at the top of the image. The placenta (PL) and fetus extend beyond the bottom of the image. The MLAp is fully developed in the null mouse and the region is filled with uNK cells. DB in the null mouse contains areas having very few decidual cells, which retain fibroblast-like characteristics, compared to the control. Overall microarchitecture of the implantations sites in LTα null (not shown) and LTβR null mice was similar at gd 10·5, 12·5 and 14·5 and resembled that in B6 control tissues. H&E, bars represent 250 µm.

Download figure to PowerPoint


Figure 3. Photomicrograph showing > 20 uNK cells/image (representative cells are marked by arrowheads) in the MLAp of LTα null (a) and B6 (b) mice on gd 14·5. uNK cells were large (largest cells measured 42 µm) and heavily granulated (22 granules/cell) in the null mice which resembled uNK cells in the control mice (largest cells 38 µm and 26 granules/uNK cell). SM indicates the circular smooth muscle layer of the myometrium that is disrupted during normal differentiation of the MLAp. PAS. Bars represent 75 µm.

Download figure to PowerPoint


Figure 4. Photomicrographs of central regions within DB in LTβR null, LTα null and B6 mice. Panels are all orientated with MLAp above and the placenta below. (a, b) Impaired development of the DB in LTβR null females at gd 10·5 (a) and 12·5 (b). Arrows show the persistence of poorly decidualized fibroblast-like cells at gd 10·5 [high power insert in (a)] while the asterisk marks accumulations of extracellular material, interpreted as proteinaceous fluid. Arrowheads mark a mature uNK cell within the oedematous region. These features were not prominent at gd 14·5. (c, d) DB in LTα null females at gd 12·5 (c) and 14·5 (d). Increased amounts of extracellular matrix (*) and poor differentiated cells were present in DB of LTα null females on gd 10·5 (not shown) and 12·5 (c). This can be recognized by comparing (c) to (e), a micrograph of the DB in day 12·5 B6 mice. By gd 14·5, the microarchitecture in DB of LTα null (d) and LTβR null females (not shown) was more organized (compare c to d) and closely resembled that in DB of gd 14·5 B6 females (f). Decidual cells in the control mice are larger and rounder. PAS. Bars represent 40 µm.

Download figure to PowerPoint

Table 1.  Area morphometry of microdomains on the mesometrial side of the pregnant uterus
GenotypeGestation dayArea morphometry (mm2 ± SD)
MLApDecidua basalisPlacenta
  • *

    Significantly different (P < 0·05) from the same microdomain in the same genotype at gd 12·5 and 14·5.

  • Significantly different (P < 0·05) from the same microdomain of B6 at gd 10·5 and 12·5.

  • Significantly different (P < 0·05) from the same microdomain in LTα null and in B6 at the same gestational time-point.

LTα null10·50·92 ± 0·14*3·64 ± 0·581·80 ± 0·45
LTβR null10·50·68 ± 0·13*2·90 ± 0·391·67 ± 0·18
 B610·50·66 ± 0·16*3·75 ± 0·522·11 ± 0·35
LT α null12·51·49 ± 0·282·44 ± 0·322·80 ± 0·34
LTβR null12·51·15 ± 0·233·15 ± 0·165·86 ± 0·36
 B612·51·58 ± 0·242·39 ± 0·383·96 ± 0·94
LTα null14·52·53 ± 0·192·26 ± 0·464·38 ± 0·69
LTβR null14·53·17 ± 0·931·92 ± 0·513·85 ± 0·54
 B614·52·03 ± 0·251·52 ± 0·255·56 ± 0·15

Numbers of uNK cells in the MLAp in pregnant LTα null and LTβR null mice

uNK cells were numerous in both the MLAp and DB of pregnant B6 females on all three gd investigated. Peak numbers occurred in both microdomains at gd 10·5 and numbers declined at gd 12·5 and 14·5, consistent with published reports. In contrast, in both LTα null and LTβR null females uNK cells peaked in the DB on gd 10·5 but did not reach peak numbers in the MLAp until gd 12·5. Numerical decline in uNK cells within the MLAp of the knockout strains was thus delayed slightly but was initiated by gd 14·5 (Fig. 5). At all three time-points in B6 females, uNK cells in the MLAp were smaller and less heavily granulated (properties thought to reflect immaturity) than uNK cells in DB. The granularity and diameters of the uNK cells in LTα null and LTβR null mice were similar to those in gd-matched B6 mice (Fig. 3), suggesting no detectable delay in uNK cell maturation in the gene-deleted strains.


Figure 5. Numbers of uNK cells/mm2 in the MLAp and in the DB of LTα null, LTβR null and B6 mice in the second trimester of mouse gestation (gd 10·5, 12·5 and 14·5). *Value statistically different from that of the same tissue region on different days of gestation in the same strain of mouse.

Download figure to PowerPoint

Morphometric measurement of decidual spiral arteries from pregnant LTα null and LTβR null mice

Murine uNK cells have not been cultured successfully and the only known functional action of the cells that can be measured quantitatively is their effect on dilation of the uteroplacental arterial segment called the spiral artery, as it passes through maternal decidua. These arteries appear to have undergone normal wall thinning and dilation in both LTα null and LTβR null mice (Fig. 2). This was confirmed by wall area to lumen area ratio (W : L) measurements of the vessels. The W : L ratios in the gene-ablated mice were the same as the ratios calculated for gd-matched B6 mice (Fig. 6). This indicated that the uNK cells detected in the histological sections display normal function.


Figure 6. Vessel area : lumen area ratios in the decidual spiral arteries at gd 10·5 and 12·5 for LTα null, LTβR null and B6 mice. There were no statistical differences within strains or between the strains at either gestation day, indicating that the normal physiological changes of pregnancy have occurred in these arteries.

Download figure to PowerPoint


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

In most murine secondary lymphoid tissue, LTα3 and LTα1β2 play key roles in initiating tissue biogenesis or in organizing tissue structure. It does not appear that these cytokines have either action in the pregnant adult uterus, although they are expressed in maternal uterine tissues during pregnancy (Fig. 1).17,20 Initiation of the MLAp occurred and a histologically normal structure developed, with only a slight delay in timing compared to control mice. This outcome may, perhaps, have been predicted since LTα3 and LTα1β2 influence LN and PP development during the late fetal and early neonatal periods while MLAp development is a re-occurring pregnancy-associated event in adults. During PP development, expression of V-CAM-1 and I-CAM-1 in the intestinal mesenchyme is followed by seeding of an inductive precursor cell defined as IL-7+, CD4+, CD3. In normal mice, small, non-granulated uNK cells (immature) are detected first histologically at gd 5·5 as DBA-lectin reactive cells localized to DB [24, unpublished]. This gestational stage is prior to development of the MLAp and suggests that decidual stromal cells or endothelium have uniquely expressed properties to attract lymphoid precursors to the uterus. Interestingly, Kruse et al. have reported studies conducted prior to full differentiation of the MLAp (gd 8·5 by our counting), showing that endothelium in DB has unique expression of V-CAM-1 in the absence of MAd-CAM.25 We have shown recently, in transplantation studies using pregnant alymphoid recipient mice, that uterus does not contain self-renewing uNK cell precursors.26 Precursors capable of generating uNK cells were found in all tissue examined from T and B cell-deficient severe combined immunodeficiency (SCID) [bone marrow (BM), spleen, liver] or B6 (LN, thymus) donors. However, the richest sources of uNK precursors were spleens from pregnant donors or LN pools (excluding nodes draining the pelvis and abdomen). Peripheral lymphocyte homing to the uterus appeared to be initiated at peri-implantation (gd 3·5) and to continue into mid-gestation.24 Although uNK cell precursors are not yet phenotyped, they would not express surface CD3 because they are present in SCID mice. Thus, some aspects of lymphoid development in the intestinal tract may be conserved during the establishment of lymphoid regions in the pregnant uterus.

The robust differentiation of NK cells within the pregnant uteri of LTα null and LTβR null females was unexpected. Co-cultures of NK cell precursors and bone marrow stroma show that stromal cells and/or stromal cell products are required for NK cell differentiation.15,27–29 Membrane LT has been considered the key early functional molecule in the differentiation process and to be required prior to IL-15.15 In pregnant mice lacking NK plus T cells or all lymphocytes, a major stromal defect has been reported, although the pregnancies continue to term and normally sized litters (average seven pups) are born and weaned. Decidua developing in the absence of uNK cells is hypocellular but viable. By ultrastructure, the decidual cells appear to be arrested in their differentiation from fibroblasts, suggesting that there may be an important interdependence between uNK cells and endometrial stromal cells.30 The finding of uNK cells normal in number, appearance and function in LTα null and LTβR null mice indicates that decidualized uterine stromal cells must use molecules other than the known LTs, to support NK cell differentiation. Further, decidual tissue is unlikely to be uniform in its expression of these NK cell support molecules because uNK cells are not found in either the primary decidua or in secondary lateral or antimesometrial decidua.

Absence of LTα or LTβR had a mild delaying effect on development of the mesometrial decidua. It also delayed by 48 hr, appropriate cell numbers within the developing MLAp. Our experimental approach does not distinguish whether these are two independent outcomes or if the delays in decidual cellularity/maturation and matrix development retard the influx of lymphocytes into the developing MLAp. This effect was minor because there were no significant differences in gd-matched surface areas of DB or MLAp between the null and control strains or in gd-matched uNK cells, which were found similar in maturity as judged by cell diameter and quantification of cytoplasmic granules. Further, by gd 14·5, the uNK cell population had begun its expected numerical decline.

One explanation for the absence of a role for LT in development of uNK cells is that uNK cells are a separate NK cell subset with a distinctive early lymphohematopoietic precursor. This seems unlikely because uNK cell differentiation appears to be highly dependent on IL-15. The lineage is absent in mice deleted for the IL-15Rβ, IL-15Rγ or IL-15 [21,22, unpublished]. IL-15, however, may be regulated differently in uterine compared to lymphoid organ stromal cells. IRF-1 is not detected by Northern analysis in the MLAp of normal mice, yet implantation sites in IRF-1 null mice contain uNK cells and IL-15 mRNA is expressed at that same level as in B6 mice (31 and A. Ashkar et al. manuscript in preparation). Progesterone, the dominant steroid hormone of pregnancy, is reported to regulate IL-15 transcription in human endometrium in vitro.32 These data, when combined with demonstration of full differentation of uNK cells in pregnant LTα null and LTβR null females, suggest that the progesterone-dependent transformation of endometrial fibroblasts to decidual cells creates a unique stromal regulatory environment within the uterus to promote specialized development of the uNK cell subset. Recent studies suggest that the functions of this NK cell subset relate to the initiation and promotion of maternal angiogenesis in implantation sites through the production and highly localized delivery of IFN-γ and vascular endothelial cell growth factor (VEGF) in mice.23,33 CD56 bright human uNK cells produce the endothelial cell mitogen NKG2;34 angiopoietin-2, a ligand for endothelial cell receptors; the lymphatic endothelial growth factor, VEGF-C and placental growth factor,35 a molecule used to release active VEGF by displacing it from an inactive, receptor bound conformation.36 Thus, it is possible that NK cells found in the pregnant uterus have highly specialized physiological functions not required in other sites, which are promoted uniquely by stromal cells transformed into decidua and independent from LT.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We thank the staff of the University of Guelph's OMAFRA Isolation Unit for their outstanding husbandry and concern for the mutant mice and Jennifer Lewis for her assistance with genotyping of the strains. These studies were supported by Awards from the Natural Sciences and Engineering Council, Canada, Ontario Ministry of Food, Agriculture and Rural Affairs, Studienstiftung des Deutschen Volkes and Deutsche Forschungsgemeinschaft (DFG).


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Fu Y-X, Chaplin DD. Development and maturation of secondary lymphoid tissues. Annu Rev Immunol 1999; 17 : 399433.
  • 2
    De Toni P, Goellner J, Ruddle NH et al. Abnormal development of peripheral lymphoid organs in mice deficient in lymphotoxin. Science 1994; 264 : 7037.
  • 3
    Bands TA, Rouse BT, Derley MK et al. Lymphotoxin-α-deficient mice. Effects on secondary lymphoid organ development and humoral immune responsiveness. J Immunol 1995; 155 : 168593.
  • 4
    Ware CR, VanArsdale TL, Crowe PD, Browning JL. The ligands and receptors of the lymphotoxin system. Curr Top Microbiol Immunol 1995; 198 : 175218.
  • 5
    Mauri DN, Ebner R, Montgomery RI et al. LIGHT, a new member of the TNF superfamily, and lymphotoxin alpha are ligands for herpesvirus entry mediator. Immunity 1998; 8 : 2130.
  • 6
    Pfeffer K, Matsuyama T, Kundig TM et al. Mice deficient for the 55 kd tumor necrosis factor receptor are resistant to endotoxic shock, yet succumb to L. monocytogenes infection. Cell 1993; 73 : 5767.
  • 7
    Erickson SL, De Sauvage FJ, Kikly K et al. Decreased sensitivity to tumour-necrosis factor but normal T-cell development in TNF receptor-2-deficient mice. Nature 1994; 372 : 5603.
  • 8
    Rennert PD, Browning JL, Medius R, Mackay R, Hochman PS. Surface lymphotoxin α/β complex is required for the development of peripheral lymphoid organs. J Exp Med 1996; 184 : 19992006.
  • 9
    Koni PA, Sacca R, Lawton P, Browning JL, Ruddle NH, Flavell RA. Distinct roles in lymphoid organogenesis for lymphotoxins α and β revealed in lymphotoxin β-deficient mice. Immunity 1997; 6 : 491500.
  • 10
    Futterer A, Mink K, Luz A, Kosco-Vilbois MH, Pfeffer K. The lymphotoxin β receptor controls organogenesis and affinity maturation in peripheral lymphoid tissues. Immunity 1998; 9 : 5970.
  • 11
    Adachi S, Yoshida H, Kataoka H, Nishikawa S. Three distinctive steps in Peyer's patch formation of murine embryo. Int Immunol 1997; 9 : 50714.
  • 12
    Honda K, Nakano H, Yoshida H et al. Molecular basis for hematopoietic/mesenchymal interaction during initiation of Peyer's Patch organogenesis. J Exp Med 2001; 193 : 521629.
  • 13
    Mebius RE, Rennert P, Weissman IL. Developing lymph nodes collect CD4+CD3LTβ+ cells that can differentiate into APC, NK cells and follicular cells but not T or B cells. Immunity 1997; 7 : 493504.
  • 14
    Iizuka K, Chaplin DD, Wang YI, Wu Q, Pegg LE, Yokoyama WM, Fu YW. Requirement for membrane lymphotoxin in natural killer cell development. Proc Natl Acad Sci USA 1999; 96 : 633640.
  • 15
    Wu Q, Sun Y, Wang J, Lin X et al. Signal via lymphotoxin-βR on bone marrow stromal cells is required for an early checkpoint of NK cell development. J Immunol 2000; 166 : 16849.
  • 16
    Yeaman GR, Collins JE, Fanger MW, Wira DR, Ledyard P. CD8+ T cells in human uterine endometrial lymphoid aggregates: evidence for accumulation of cells by trafficking. Immunology 2001; 102 : 43440.
  • 17
    Vince G, Shorter S, Starkey P, Humphreys J, Clover L, Wilkins T, Sargent I, Redman C. Localization of tumour necrosis factor production in cells at the maternal/fetal interface in human pregnancy. Clin Exp Immunol 1992; 88 : 17480.
  • 18
    Phillips TA, Ni J, Hunt JS. Death-inducing tumour necrosis factor (TNF) superfamily ligands and receptors are transcribed in human placentae, cytotrophoblasts, placental macrophages and placental cell lines. Placenta 2001; 22 : 66372.
  • 19
    Peel S. Granulated metrial gland cells. Adv Anat Embryol Cell Biol 1989; 115 : 1115.
  • 20
    Hunt JS, Rasmussen CA. The tumor necrosis factor gene family in pregnancy. Trophoblast Res 1998; 11 : 13947.
  • 21
    Guimond M-J, Luross JA, Wang B, Terhorst C, Danial S, Croy BA. Absence of natural killer cells during murine pregnancy is associated with reproductive compromise in Tgɛ26 mice. Biol Reprod 1997; 56 : 16979.
  • 22
    Guimond M-J, Wang B, Croy BA. Engraftment of NK cells reverses the reproductive deficits in pregnant NK cell and T cell deficient tgɛ26mice. J Exp Med 1998; 187 : 21723.
  • 23
    Ashkar AA, Di Santo JP, Croy BA. Interferon-γ contributes to initiation of uterine vascular modification, decidual integrity and uterine NK cell maturation during normal murine pregnancy. J Exp Med 2000; 192 : 25970.
  • 24
    Paffaro VA Jr, Haraguchi CM, Fonseca PM, Jacobucci GB, Joazeiro PP, Yamada AT. Glycoconjugates containing n-acetyl-galactosamine expressed by mouse uterine natural killer used as selective cell marker. Placenta 1999; 20 : A51.
  • 25
    Kruse A, Merchant MJ, Hallmann R, Butcher EC. Evidence of specialized leukocyte–vascular homing interactions at the maternal/fetal interface. Eur J Immunol 1999; 29 : 111626.
  • 26
    Chantakru S, Miller C, Roach LE, Kuziel WA, Maeda N, Wang W-C, Evans SS, Croy BA. Analysis of contributions from lymphocyte self-renewal and trafficking to the uterine natural killer cell population of early pregnancy. J Immunol 2002; 168 : 228.
  • 27
    Roth C, Carlyle JR, Takizawa H, Raulet DH. Clonal acquisition of inhibitory Ly-49 receptors on developing NK cells is successively restricted and regulated by stromal class I MHC. Immunity 2000; 13 : 14353.
  • 28
    Lu J, Patrene KD, Appasamy PM, Herberman RB, Boggs SS. Characterization of the stage in natural killer cell development in 14.5-day mouse fetal liver using adult bone marrow stroma. Exp Hematol 1999; 27 : 104656.
  • 29
    Williams NS, Klem J, Puzanov IJ, Sivakumar PV, Bennett M, Kumar V. Differentiation of NK1.1+, Ly49+ NK cells from flt3+ multipotent marrow progenitor cells. J Immunol 1999; 163 : 264856.
  • 30
    Greenwood JD, Minhas K, Di Santo JP, Makita M, Kiso Y, Croy BA. Ultrastructural studies of implantation sites from mice deficient in uterine natural killer cells. Placenta 2000; 21 : 693702.
  • 31
    Ashkar AA, Croy BA. Interferon regulatory factor-1 has important functions in placental growth and modification of decidual arterioles. Placenta 1999; 20 : A11.
  • 32
    Okada H, Nakajima T, Sanezumi M, Ikuta A, Yasuda K, Kanzaki H. Progesterone enhances interleukin-15 production in human endometrial stromal cells in vitro. J Clin Endocrinol Metab 2000; 85 : 476570.
  • 33
    Wang C, Umesaki N, Nakamura H, Tanaka T, Nakatani K, Sakaguchi I, Ogita S, Kaneda K. Expression of vascular endothelial growth factor by granulated metrial gland cells in pregnant murine uteri. Cell Tissue Res 2000; 300 : 28593.
  • 34
    Langer N, Beach D, Lindenbaum ES. Novel hyperactive mitogen to endothelial cells: human decidual NKG5. Am J Reprod Immunol 1999; 42 : 26372.
  • 35
    Li XF, Charnock-Jones S, Zhang E et al. Angiogenic growth factor messenger ribonucleic acids in uterine natural killer cells. J Clin Endocrinol Metabol 2001; 86 : 182334.
  • 36
    Carmeliet P, Moons L, Luttun A et al. Synergism between vascular endothelial growth factor and placental growth factor contributes to angiogenesis and plasma extravasation in pathological conditions. Nature Med 2001; 7 : 57583.
  • 37
    Rossant J, Croy BA. Genetic identification of tissue of origin of cellular populations within the mouse placenta. J Embryol Exp Morph 1985; 86 : 17789.