The aim of this investigation was to establish a homologous culture system for study of the transcriptional mechanisms underlying endometrial expression of the pregnancy-associated genes encoding antileukoproteinase (ALP), an elastase/cathepsin G protease inhibitor, and uteroferrin (Uf), a transplacental iron transport protein. Glandular epithelial (GE), Luminal epithelial (LE), and stromal (ST) cells were isolated from pig endometrium at Day 12 of pregnancy by differential enzymatic digestion and sieve filtration. The three cell populations differed with respect to their morphology in culture and with respect to their expression of ALP and Uf. Expression of the ALP gene was much higher in GE than in LE cells and was undetectable in ST cells. Similarly, GE had the highest expression of the Uf gene, and expression in ST was lower but distinct. Western blot analysis of conditioned media (72 h) from GE, LE, and ST, using antiporcine Uf antiserum, detected significant levels of secreted Uf only in GE. The steroid hormone responsiveness of GE cells was monitored by changes in steady-state levels of ALP mRNA after 24-h exposure to estradiol 17 beta (E2; 10 nM) and/or progesterone (P; 10 nM). Glandular epithelial cells treated with E2, P, and E2 + P had increased (p < 0.05) ALP mRNA levels relative to those in control cultures. Glandular epithelial cells were transiently transfected with reporter constructs containing the 5'-flanking genomic regions of each gene. For ALP, the 1266-nucleotide (nt) region of the ALP 5'-flanking genomic DNA, and progressive 5' deletions within this region, were coupled to a luciferase reporter gene (LUCE). The most proximal 119-bp fragment (-119ALP LUCE), which contains the TATAA box (-21 to -26 nt) and a GC-rich sequence (-66 to -74 nt), was sufficient to confer transcriptional activity to the reporter vector. Progressively longer 5'-genomic fragments had promoter activities higher than or similar to those of the 119-nt fragment. Estrogen had no effect on the transcriptional activities of any of the ALP constructs. Uteroferrin 5'-flanking and promoter DNA constructs containing the chloramphenicol acetyl transferase (CAT) reporter gene also exhibited transcriptional activity in GE cells. The presence of multiple interacting cis-regulatory sequences within this region was demonstrated by increased promoter activity, relative to that of the smallest construct (-182 UFCAT-E; basal activity), with the inclusion of sequences between -182 and -484 nf, and drastic reduction to basal activity with the inclusion of sequences between -484 and -831 nt. In summary, primary cultures of GE from early-pregnant porcine endometrium express ALP and Uf, are steroid hormone-responsive, and support the transcriptional activity of endometrial-associated gene promoter and regulatory sequences. The use of primary GE cells thus provides a convenient in vitro system for further study of the endocrine, paracrine, and autocrine factors regulating endometrial gene expression during pregnancy.

The endogenous factors that underlie the transient induction of the gene encoding spermidine/spermine N1-acetyltransferase (SSAT), the rate-limiting enzyme in cellular polyamine catabolism, in pig uterine endometrium during periimplantation are not known. The present study examined a number of peptide growth factors and regulatory molecules that are present within the uterine environment at early pregnancy, coincident with maximal SSAT gene expression, for their ability to manifest endogenous SSAT gene-inducing activity. Basal SSAT expression in luminal epithelial cells was higher (p < 0. 01) than that for glandular epithelial (GE) or stromal (ST) cells. Recombinant human insulin-like growth factor-I (IGF-I; 50 ng/ml) had no effect on steady-state SSAT mRNA levels, but it increased mitogenesis in all three cell types. In contrast, IGF-I caused a marked induction (p < 0.01) of SSAT mRNA levels in the human endometrial carcinoma cell line Hec-1-A. Uterine explants incubated with interleukin-6, transforming growth factor alpha, epidermal growth factor (each at 1, 10, and 100 ng/ml), retinoic acid and retinol (each at 0.01, 0.1, and 1 microM), and estradiol-17beta (10 nM) had SSAT mRNA levels similar to controls. By contrast, leukemia inhibitory factor (LIF; at 10 and 100 ng/ml) caused a modest, but significant (p < 0.05), increase in SSAT mRNA levels over those of untreated explants. This effect of LIF, however, did not approach the level of induction observed in GE or ST cells after addition of medium conditioned by Day 12 or 17 porcine conceptuses and in endometrial explants supplemented with medium conditioned by Day 21 porcine conceptuses or a continuous cell line (Jag-1) derived from Day 14 porcine trophoblast. We suggest that transient induction of endometrial SSAT gene expression at implantation is mediated by the functional interactions of specific conceptus-derived regulatory factors, distinct from estrogen, with endometrial-derived factor(s) such as LIF. These complex interactions are probably requisite for the transient, yet dramatic, induction of SSAT gene expression and may be critical for successful implantation.

The present study examined the trans-activation potential of basic transcription element-binding protein (BTEB), a recently identified member of the Sp family of GC box-binding transcription factors, on the expression of the gene encoding the pregnancy-associated, epithelial-specific, and progesterone (P)-induced porcine uterine endometrial secretory protein, uteroferrin (UF). Endometrial expression of BTEB, P receptor (PR), and UF genes was analyzed by RT-PCR as a function of pregnancy stage and cell type and was correlated with the levels of endometrial BTEB that were quantified by Western blot and/or electrophoretic mobility shift assay. PR, BTEB, and UF messenger RNAs (mRNAs) were present in early (day 12) and mid(day 60) pregnancy pig endometrium, although expression levels varied for each mRNA (UF, day 12 << day 60; PR and BTEB, day 12 = day 60). Within the endometrium, glandular epithelial (GE) cells manifested higher amounts of UF mRNA than stromal fibroblastic cells, whereas both cell types had comparable amounts of BTEB and PR mRNAs. Expression of BTEB, however, was limited to endometrial GE cells. A BTEB expression vector (pcDNA-3BTEB) was used to examine the effect of increased BTEB protein on UF gene expression and promoter activity in primary cultures of pig endometrial GE cells. Cells transiently transfected with pcDNA-3BTEB had 2-fold higher UF mRNA levels than those transfected with the empty expression vector (pcDNA-3). Further, cells cotransfected with a UF promoter-luciferase (-1935UF-Luc) reporter gene and the BTEB expression vector had 2-fold higher Luc activity than those cotransfected with reporter gene and pcDNA-3. This effect of BTEB was not observed in transfected endometrial stromal fibroblastic cells, but was apparent in the human endometrial epithelial carcinoma cell lines ECC-1 and Hec-1-A, which exhibit low levels of BTEB protein and low or undetectable PR mRNA levels, respectively. The respective contributions of BTEB and PR to the modulation of UF promoter activity were examined by cotransfection of Hec-1-A and ECC-1 cells with expression plasmids for BTEB and PR and one of two UF promoter constructs (-831UF-Luc or -1935UF-Luc) in the absence or presence of P. The increase in UF promoter activity with BTEB was mimicked by PR in a P-dependent manner in both cell lines. The combined effect of PR/P and BTEB appeared additive in Hec-1-A cells and was synergistic in ECC-1 cells. These results highlight the cell context dependence of the trans-activation potential of BTEB and suggest its unique role, in concert with PR, in directing the temporal expression of endometrial epithelial genes of pregnancy.