Search

The endometrium is the only human tissue to undergo cyclic breakdown and regeneration. This physiological alternation renders it an advantageous system for studying tissue remodelling. Our previous observations indicate that menstrual endometrial breakdown is initiated by matrix metalloproteinases (MMPs), which are controlled overall by ovarian steroids but are also locally regulated by cytokines. We have therefore compared the effect of several endometrial cytokines on the gene expression of eight MMPs and their tissue inhibitors (TIMP)-1, -2 and -3, in primary cultures of human endometrial fibroblasts. Three categories of gene expression were identified: (a) MMP-13, -15 and -16 mRNAs were not detected despite stimulation by various cytokines; (b) MMP-2 and -14 as well as TIMP-1, -2 and -3 mRNAs were constitutively expressed but not markedly affected by the six cytokines tested; (c) mRNAs for MMP-1, -9 and -11 were selectively induced by specific cytokines: insulin-like growth factor-II, epidermal growth factor (EGF), platelet derived growth factor (PDGF)-BB and interleukin (IL)-6 stimulated MMP-11 expression; MMP-1 was induced by EGF, PDGF-BB, tumour necrosis factor (TNF)alpha and IL-1alpha, which also exerted additive effects. In contrast with MMP-1 and MMP-11 gene expression, which was sustained for 48 h, MMP-9 mRNA was quickly induced by TNFalpha, but disappeared within 12 h despite continuing stimulation. These results show that several cytokines are able to induce the selective expression of MMPs in cultured human endometrial fibroblasts and are thus good candidates for involvement in local triggering of menstrual tissue breakdown.

To quantify the kinetics of the plasma membrane flow into lysosomes, we covalently labelled at 4 degrees C the pericellular membrane of rat fibroblasts and followed label redistribution to the lysosomal membrane using purified lysosomal preparations. The polypeptides were, either labelled with 125I by the lactoperoxidase procedure, or conjugated to [3H]peroxidase using bisdiazobenzidine as a bifunctional reagent. Both labels were initially bound to plasma membrane, as indicated by their equilibrium density in sucrose or Percoll gradients and their displacement by digitonin, as well as by electron microscopy. Upon cell incubation at 37 degrees C, both covalent labels were lost from cells with diphasic kinetics: a minor component (35% of cell-associated labels) was rapidly released (half-life less than 1 h), and most label (65%) was released slowly (half-life was 20 h for incorporated 125I and 27 h for 3H). Immediately after labelling up to 30 h after incubation at 37 degrees C, the patterns of 125I-polypeptides quantified by autoradiography after SDS-PAGE were indistinguishable, indicating no preferential turnover for the major plasma membrane polypeptides. The redistribution of both labels to lysosomes was next quantified by cell fractionation. At equilibrium (between 6 and 25 h of cell incubation) 2-4% of cell-associated 125I label was recovered with the purified lysosomal membranes. By contrast, when 3H-labelled cells were incubated for 16 h, most of the label codistributed with lysosomes. However, only 6% of cell-associated 3H was bound to lysosomal membrane. These results indicate that in cultured rat fibroblasts, a minor fraction of plasma membrane polypeptides becomes associated with the lysosomal membrane and is constantly equilibrated by membrane traffic.

Rat polymeric IgA (pIgA) and galactosylated bovine serum albumin (GalBSA), once injected to rats, are avidly taken up by hepatocytes via receptor-mediated endocytosis. Of injected pIgA, 64% was transferred undigested into bile within 3 h, with a peak at 30-45 min. GalBSA was essentially digested in lysosomes. By electron microscopy using ligand-peroxidase conjugates, both ligands were internalized through coated pits/coated vesicles into similar electron-lucent vesicles and tubules. Subsequently, pIgA remained mostly associated with small vesicles clustering around and fusing with bile canaliculi, while GalBSA was predominantly found in large, heterogeneous endocytic structures and in lysosomes. By subcellular fractionation, they were associated at 3 min after injection with structures that similarly sedimented in the P fraction (250 000 - 3 X 10(6) X g X min) and equilibrated at densities of about 1.13 g/ml in sucrose gradients. At 10 min and 20 min, pIgA distribution remained mostly in the P fraction at the same equilibrium density. A minor component of the pIgA distribution was found at the density of lysosomes, but contrary to lysosomal enzymes, its distribution was not affected by Triton WR 1339. In contrast to pIgA, GalBSA was progressively recovered in the L fraction (33 000 - 250 000 X g X min) with organelles equilibrating around 1.11 g/ml, and, by 20-45 min, was found in the ML fraction (10 000 - 250 000 X g X min), around 1.20 g/ml, i.e. in lysosomes. Chloroquine did not reduce the efficiency but delayed the secretion of pIgA into bile. Similarly, it did not affect the uptake of GalBSA but apparently delayed GalBSA transfer along successive populations of host organelles. The low density, GalBSA-containing structures were devoid of proteolytic activity. Anti-secretory components IgG and F(ab')2 were selectively excreted into bile, partially or totally as compounds of lower molecular mass. These antibody fragments probably result from a disulfide reduction activity along the pIgA pathway. In conclusion, our data (a) strongly suggest that pIgA and GalBSA are sorted between 3 min and 10 min after injection in non-lysosomal acidic organelles, (b) identify two successive and physically distinct endosomal populations containing GalBSA, and (c) provide the first evidence for a disulfide reduction activity along the transcytotic pathway of rat hepatocytes.

In order to label the vesicles involved in transcellular transfer (transcytosis) through hepatocytes, polymeric IgA (pIgA) was conjugated to horseradish peroxidase (HRP) and injected into rats. The endosomes containing this ligand at 10 or 20 min after injection were isolated by the diaminobenzidine-induced density-shift procedure and their content in various marker enzymes was measured. The endosomes carrying pIgA-HRP 10 min after injection contained only traces of 5'-nucleotidase and low amounts of alkaline phosphodiesterase I. The estimated marker enzyme content is similar to that observed for the particles containing galactosylated bovine serum albumin conjugated to HRP, a ligand degraded in lysosomes. However, 20 min after injection, the transcytotic endosomes showed a marked enrichment in 5'-nucleotidase and especially in alkaline phosphodiesterase I. The results confirm the heterogeneity of rat liver endosomes and substantiate the concept of distinct endosomal compartments.

We have quantified, in cultured rat fibroblasts, the association to the lysosomal membrane of two classical plasma membrane markers, 5'-nucleotidase and alkaline phosphodiesterase I. To isolate highly purified lysosomal preparations, lysosomes were loaded with horseradish peroxidase (2-h cell uptake, 16-h chase) and isolated by isopycnic centrifugation in linear Percoll gradients, followed by a 3,3'-diaminobenzidine-induced density shift in sucrose gradients. Purified lysosomal preparations contained up to 50% of N-acetyl-beta-glucosaminidase of the homogenate. This lysosomal enzyme was enriched 33-fold in the most purified preparations. In the electron microscope, these preparations appeared to be highly purified and only contained organelles filled with diaminobenzidine reaction products. Analysis of purified preparations indicates that 0.5-0.8% of 5'-nucleotidase, but as much as 10.9-14.3% of alkaline phosphodiesterase I activities of the homogenate, are associated with lysosomes. After freezing-thawing, these activities remained essentially membrane-associated. The larger value obtained for alkaline phosphodiesterase I could not be ascribed to other lysosomal enzymes, as no such activity was detected at acidic pH. These two plasma membrane markers are thus unevenly distributed in the lysosomal compartment.