Pre-B-cell colony enhancing factor (PBEF) was first isolated from an activated peripheral blood lymphocyte cDNA library and was found to be involved in the maturation of B-cell precursors. It was subsequently identified as one of the genes upregulated by distending the human fetal membranes in vitro. Here we report on the genomic organization of this gene, which is composed of 11 exons and 10 introns, spanning 34·7 kb of genomic DNA. Neither the gene nor the protein has any homology with other cytokines in any currently available database. The use of two promoters (proximal and distal) may result in differential, tissue specific expression of the PBEF transcripts. The 5 -flanking region lacks the classical sequence motif that would place it with the hematopoietic cytokines; however, it has several putative regulatory elements, suggesting that this gene may be chemically and mechanically responsive to inducers of transcription.The three PBEF mRNA transcripts were observed in both normal and infected human fetal membranes but were significantly upregulated (P<0·05) in severe infection. The PBEF protein was immunolocalized, in both normal and infected tissues, to both the normal fetal cells of the amnion and chorion and the maternal decidua of the membranes, and to the invading neutrophils. These stained strongly and were likely to contribute to the increased expression in infection. The amniotic epithelial cell line (WISH cells) has been used as a model to study PBEF gene modulation. Lipopolysaccharide, interleukin (IL)-1 , tumour necrosis factor (TNF) and IL-6 all significantly increased the expression of PBEF in 4 h of treatment. The addition of dexamethasone to IL-1 and TNF significantly reduced the response of PBEF to these cytokines. IL-8 treatment failed to alter PBEF gene expression. Thus PBEF is a cytokine expressed in the normal fetal membranes and upregulated when they are infected. It is likely to have a central role in the mechanism of infection-induced preterm birth.

Decidual and placental relaxins have been proposed as autocrine/ paracrine hormones in the remodeling of collagen in the amnion and chorion in the last weeks of pregnancy. The matrix metalloproteinase-1 (MMP-1) is a key enzyme in the degradation of the interstitial collagens which predominate in the fetal membranes. Distribution of the MMP-1 gene and of the MMP-1 protein was shown by in situ hybridization and immunolocalization, respectively, in amnion, chorion, and decidua collected from patients before the onset of spontaneous labor. The distribution of MMP-1 in the chorionic cytotrophoblast and decidua coincided with that of the human relaxin receptor, detected by tissue section autoradiography in tissues collected at the same stage of pregnancy. Fetal membrane explants were used to study the effect of exogenous human relaxin H2. These responded by a dose-dependent increase in expression of the MMP-1 gene, in its secreted protein, and in its enzyme activity in the medium. A similar dose-dependent increase in the tissue plasminogen activator (tPA) gene and protein upon exposure of the explants to relaxin H2 suggested a coordinated cascade system, resulting in increases in secreted activities of MMP-1, MMP-3 (stromelysin), and MMP-9 (gelatinase B). There was no effect on the genes or proteins for MMP-2 (gelatinase A) or tissue inhibitor of metalloproteinase-1 (TIMP-1), showing the specificity of the response. This coordinated regulation by relaxin H2 of tPA, MMP-1, MMP-3, and MMP-9 would result in more complete degradation of the fetal membrane extracellular matrix components.

The two human relaxin genes termed H1 and H2 are expressed in the choriodecidua and placenta and have been proposed to act via specific receptors as local modulators of collagenolysis in the fetal membranes. Such receptors have been inferred, but not demonstrated, from studies of the effect of adding exogenous relaxin to these tissues. Thus conditions were optimized for the binding of 32P-labelled human relaxin H2 to membrane-enriched particulate fractions of human fetal membranes, amnion and chorion, with adhering decidua. The membrane protein concentration was optimal at 250 micrograms, when incubated at 27 degrees C for 60 min, at pH 7.5 with Mn2+ and Mg2+ ion concentrations of 2.0 mM. Incubation of membrane particulate fractions with increasing amounts of labelled relaxin H2 suggested the presence of a single class of binding sites with an affinity constant (Ka) of 2.15 nM. The binding was primarily to the chorion and decidua with very little to the amnion layer. The competition for binding of the 32P-labelled human relaxin H2 with unlabelled relaxin H2 gave an IC50 of 28 pM, while an IC50 of 60 pM and 280 pM was obtained for relaxin H1 and porcine relaxin respectively. In contrast, unlabelled guinea-pig relaxin inhibited this binding by only 10% even at a 1000-fold greater concentration than H2, and human recombinant insulin failed to inhibit even at a million-fold concentration of unlabelled relaxin H2. Relaxins H2 and H1 can readily displace the binding of either 32P-labelled human relaxins H1 or H2 and gave very similar displacement curves.(ABSTRACT TRUNCATED AT 250 WORDS)

No abstract