Leukemia inhibitory factor plays a major role in the uterus and in its absence embryos fail to implant. Our knowledge of the targets for LIF and the consequences of its absence is still very incomplete. In this study, we have examined the ultrastructure of the potential implantation site in LIF-null MF1 female mice compared to that of wild type animals. We also compared expression of proteins associated with implantation in luminal epithelium and stroma. Luminal epithelial cells (LE) of null animals failed to develop apical pinopods, had increased glycocalyx, and retained a columnar shape during the peri-implantation period. Stromal cells of LIF-null animals showed no evidence of decidual giant cell formation even by day 6 of pregnancy. A number of proteins normally expressed in decidualizing stroma did not increase in abundance in the LIF-null animals including desmin, tenascin, Cox-2, bone morphogenetic protein (BMP)-2 and -7, and Hoxa-10. In wild type animals, the IL-6 family member Oncostatin M (OSM) was found to be transiently expressed in the luminal epithelium on late day 4 and then in the stroma at the attachment site on days 5-6 of pregnancy, with a similar but not identical pattern to that of Cox-2. In the LIF-null animals, no OSM protein was detected in either LE or stroma adjacent to the embryo, indicating that expression requires uterine LIF in addition to a blastocyst signal. Fucosylated epitopes: the H-type-1 antigen and those recognized by lectins from Ulex europaeus-1 and Tetragonolobus purpureus were enhanced on apical LE on day 4 of pregnancy. H-type-1 antigen remained higher on day 5, and was not reduced even by day 6 in contrast to wild type uterus. These data point to a profound disturbance of normal luminal epithelial and stromal differentiation during early pregnancy in LIF-nulls. On this background, we also obtained less than a Mendelian ratio of null offspring suggesting developmental failure.

Decidualisation of uterine stromal cells is a prerequisite for implantation of the embryo in mice. Here we have used an in vitro culture system in which stromal cells decidualise as indicated by a number of markers, including an increase in alkaline phosphatase (ALP) activity. The latter was used as a quantitative marker of decidualisation in the presence of low (2%) fetal calf serum. Prostaglandin E(2) (PGE(2)), which is known to induce decidualisation, increased ALP activity, and this effect was blocked in a dose-dependent manner by indomethacin. Leukemia inhibitory factor (LIF) was then examined, but it had no effect on PGE(2) secretion. However, LIF suppressed ALP activity in a dose-dependent manner in the presence of 2% serum, while an inhibitor of LIF that competes for binding to its receptor reversed the effect of LIF and increased ALP activity above the control level. In serum-free cultures, stromal cells differentiated rapidly, and no differences were observed between LIF-treated and untreated cultures. Stromal cells produce LIF during in vitro culture, and this peaked at 48 h. Freshly collected stromal cells from both day-2 and -4 pregnant mice expressed mRNA for the LIF receptor, and the transcript level was higher in cells isolated on day 4. However, no differences were observed in the relative levels of transcripts in cells from day 2 and day 4 after culture, nor were there differences between the LIF-treated cultures and controls. Therefore, in this study, we have shown that LIF suppresses decidualisation of murine uterine stromal cells in the presence of serum, this is not due to the regulation of PGE(2) secretion by stromal cells.