Unsupervised learning in detection of gene transfer.

The tree representation as a model for organismal evolution has been in use since before Darwin. However, with the recent unprecedented access to biomolecular data, it has been discovered that, especially in the microbial world, individual genes making up the genome of an organism give rise to different and sometimes conflicting evolutionary tree topologies. This discovery calls into question the notion of a single evolutionary tree for an organism and gives rise to the notion of an evolutionary consensus tree based on the evolutionary patterns of the majority of genes in a genome embedded in a network of gene histories. Here, we discuss an approach to the analysis of genomic data of multiple genomes using bipartition spectral analysis and unsupervised learning. An interesting observation is that genes within genomes that have evolutionary tree topologies, which are in substantial conflict with the evolutionary consensus tree of an organism, point to possible horizontal gene transfer events which often delineate significant evolutionary events.

Estrogen replacement decreases the set point of parathyroid hormone stimulation by calcium in normal postmenopausal women.

Estrogens decrease serum total and ionized calcium (Ca) concentrations in postmenopausal women with or without primary hyperparathyroidism, but cause little or no increase in serum PTH suggesting a modification of the relationship between the two. In order to define this relationship, we studied the effect of conjugated estrogens on total and ionized serum Ca and serum PTH concentrations in five normal postmenopausal women, before and after 3, 11, and 23 weeks of therapy. Dynamic tests of parathyroid gland function, based on 2-h iv infusions of CaCl2 and NaEDTA, were performed at each time. Total and ionized serum Ca and carboxylterminal PTH were measured every 15 min during the infusions, and parathyroid function was evaluated by a nonlinear 4-parameter mathematical model. Estrogen therapy caused decreases in serum total [2.36 +/- 0.04 (SD) mmol/L, baseline vs. 2.19 +/- 0.05 mmol/L, 23 weeks, P less than 0.005) and ionized calcium (1.27 +/- 0.01 mmol/L, baseline vs. 1.21 +/- 0.02 mmol/L, 23 weeks, P less than 0.005]; the decreases were evident at 3 weeks and persisted for the duration of the study. Serum PTH concentrations did not change (8.94 +/- 1.84 pmol/L, baseline vs. 8.98 +/- 2.38 pmol/L, 23 weeks). Three parameters of the parathyroid function, the maximal response to hypocalcemic stimulation, the nonsuppressible fraction of circulating PTH, and the slope of PTH on calcium at the set point were not affected by estrogen treatment. The fourth parameter, the set point of PTH stimulation by serum total calcium (2.16 +/- 0.04 mmol/L, baseline vs. 1.97 +/- 0.07 mmol/L, 23 weeks, P less than 0.0166) or by serum ionized Ca (1.19 +/- 0.04 mmol/L, baseline vs. 1.12 +/- 0.03 mmol/L, 23 weeks, P less than 0.01), was decreased by estrogen treatment. This was evident at the earliest time point studied and persisted thereafter. The decrease in ionized Ca set point only explained 40% of the decrease in total calcium set point, the remaining 60% being related to hemodilution of plasma protein during therapy. We conclude that estrogen replacement can influence parathyroid function in postmenopausal women by resetting the set point of PTH stimulation by ionized Ca. This in turn could contribute to the estrogen-induced changes in their Ca balance.

Novel cAMP PDE III inhibitors: 1,6-naphthyridin-2(1H)-ones.

Two series of medorinone (3) analogs were prepared by modifications at C(2) and C(5). The C(2)-series was prepared from 2-chloro-5-methyl-1,6-naphthyridine (4) by replacement of the chloro group with various nucleophiles. The C(5)-series was prepared from 5-acyl-6-[2-(dimethylamino)-ethenyl]-2(1H)-pyridinone (11), 5-bromo-1,6-naphthyridin-2(1H)-one (17), and 1,3-diketones 19 and 27. 1,6-Naphthyridin-2(1H)-ones are novel inhibitors of cAMP PDE III. Modification of the carbonyl group of 3 or N-methylation at N(1) resulted in a dramatic loss of enzyme activity. Absence of the C(5)-methyl group of medorinone (3) or its shift to C(3) or C(7) also resulted in reduced activity. Substitution at C(3) also diminished activity. However, substitution at C(5) by a wide variety of substituents led to improvement of enzyme activity and several C(5)-substituted analogs were more potent than milrinone.

Novel and potent adenosine 3',5'-cyclic phosphate phosphodiesterase III inhibitors: thiazolo[4,5-b][1,6]naphthyridin-2-ones.

The transformation of 3-bromo-1,6-naphthyridin-2(1H)-ones 8 to thiazolo[4,5-b][1,6]naphthyridin-2(1H)-ones 12 resulted in a 2-9-fold increase in cAMP phosphodiesterase (PDE) III inhibitory potency. Unlike the secondary binding sites on the cAMP PDE III isozyme which interact with the methyl group of milrinone (2) and CI-930 (4), the site which interacts with the 5-substituents of 1,6-naphthyridin-2(1H)-ones and the 8-substituents of thiazolo[4,5-b][1,6]naphthyridin-2(1H)-ones 12 is able to accommodate a diverse group of substituents which have different steric and electronic requirements.

Vascular endothelial growth factor and signaling in the prostate: more than angiogenesis.

In cloning tyrosine kinase genes in dog prostate cells, a fragment of the vascular endothelial growth factor (VEGF) receptor 1 or Flt-1 was sequenced. To test for a functional protein, Flt-1 antibodies were used to probe immunoprecipitated tyrosine phosphorylated proteins. Western blotting revealed a major 170-180 kDa band and a few bands below 116 kDa in dog prostate and human prostatic carcinoma PC-3 cells, with higher levels in PC-3. Similar results were obtained with human placental membranes used as a source of Flt-1. That the major Flt-1 tyrosine phosphorylated protein was likely VEGF-R1 and part of VEGF signaling pathways was shown by enhanced level of only this protein when PC-3 cells were exposed to VEGF. Accordingly specific cell surface receptor complexes, displaced by VEGF but not EGF and compatible with Flt-1 in size, were revealed by chemical cross-linking after 125I-VEGF binding. Similarly to the prostatic neuroproduct, gastrin-releasing peptide/bombesin, VEGF directly triggered the tyrosine phosphorylation of focal adhesion kinase and stimulated PC-3 cell motility. The titration of prostate tissue sections with VEGF-A antibodies revealed a confined staining in chromogranin A and/or serotonin positive neuroendocrine (NE) cells, including in primary tumors and lymph node metastases. Given that NE differentiation is associated with advanced disease, that NE cells are a significant source of VEGF in prostatic tumors, and that VEGF directly act on prostate cancer cells in vitro, VEGF-A may be more than angiogenic in prostate cancer and hence favor progression by affecting tumor cells.

Induction of heat shock protein 70 by herbimycin A and cyclopentenone prostaglandins in smooth muscle cells.

This study characterizes Hsp70 induction in human smooth muscle cells (SMC) by herbimycin A and cyclopentenone prostaglandins. The magnitude of Hsp70 induction by cyclopentenone prostaglandins was 8- to 10-fold higher than induction by herbimycin A. Hsp70 induction by delta12PGJ2 was first observed at 10 microM, rose to 4000-5000 ng/mL within one log unit and a maximum response was not observed; concentrations of delta12PGJ2 higher than 30 microM were toxic to the cells. A maximum response with herbimycin A (500 ng/mL) was reached at 0.05 microM and maintained to 1 microM without toxicity. Both, delta12PGJ2 and herbimycin A, were inhibited by dithiothreitol (DTT, 100 microM) at lower concentrations and became less sensitive to inhibition at higher concentrations. Hsp70 induction after incubation of SMC with delta12PGJ2 followed by addition of herbimycin A was significantly higher than Hsp70 induction after incubation with herbimycin A followed by addition of delta12PGJ2. When cells were incubated with [3H]-PGJ2, followed by protein denaturation, substantial radioactivity remained protein-bound suggesting that the prostaglandin must be covalently bound. Covalent binding was largely insensitive to DTT. Maximal Hsp70 induction was observed after 5 minutes of exposure of the cells to herbimycin A followed by a 20 hour recovery period in agent-free medium. Cells required 3-4 hours of exposure to delta12PGJ2 followed by a 20 hour recovery period in order to see high Hsp70 induction. Binding of the heat shock factor (HSF) to the heat shock element (HSE) in the presence of herbimycin A or delta12PGJ2, and the effects of DTT, mirrored the results of Hsp70 induction. The results suggest that probable differences between the 2 agents are at the level of the signal transduction prior to HSF activation.

In vitro and in vivo evaluation of a once-daily controlled-release pseudoephedrine product.

The functionality of a once-daily, osmotic dosage form--gastrointestinal therapeutic system (pseudoephedrine HCl) or GITS (PeHCl)--was studied in vitro and in vivo. The in vitro release profiles were close to identical from pH 1 to 7.5 and between USP apparatus 2 and 7, independent of paddle speeds from 50 to 200 rpm; GITS also released drug at the normal rate in aqueous media after incubation in bile salts or fatty media. Both strengths of GITS (PeHCl)--240 and 120 mg--were then compared with a commercially available pseudoephedrine solution given every 6 hours and a timed-release 12-hour pseudoephedrine capsule given every 12 hours in a randomized 4-way crossover study in 24 healthy men. All four formulations were equivalent in total drug absorbed. Both GITS treatments had AUCinf values equivalent to those of PeHCl solution and capsules, and Cmax values equivalent to PeHCl capsules. Cmax for GITS and capsule treatments were each significantly lower than for solution, but the differences were small (14-17%). A one-to-one correlation was shown between rate of absorption and in vitro release profiles for the GITS products, indicating that drug release from GITS controls absorption. Insensitivity to conditions of in vivo release accounts for the close in vitro/in vivo correlation of release rates. In a second randomized crossover trial (12 men), the effect of a high-fat breakfast on GITS performance was evaluated. Mean pseudoephedrine concentrations in plasma were close to identical with or without the breakfast, and the treatments were bioequivalent.(ABSTRACT TRUNCATED AT 250 WORDS)

The effects of serotonin antagonists in a behavioral despair procedure in mice.

Six serotonin antagonists (pizotifen, mianserin, cyproheptadine, ketanserin, trazodone and methysergide) were tested in mice in a behavioral despair procedure. The behavioral despair procedure detects most antidepressant compounds. Pizotifen, mianserin and cyproheptadine were found to be active and the others were inactive. The serotonin binding potency at serotonin1 or serotonin2 receptors, or the ratio of potency at these receptors did not correlate with activity in the behavioral despair procedure. However, the serotonin antagonists that were active in the behavioral despair procedure were all found to be potent antagonists at histamine1 receptors. It is suggested that the activity of some serotonin antagonists in the behavioral despair procedure is best explained by their antihistaminergic potency.

Differential pharmacologic sensitivity of cyclic nucleotide phosphodiesterase isozymes isolated from cardiac muscle, arterial and airway smooth muscle.

Phosphodiesterase isozymes were isolated by diethylaminoethyl ether (DEAE) column chromatography from cardiac muscle (canine, guinea pig), vascular (canine and guinea pig aortic) and airway (canine tracheal) smooth muscle. All peak I phosphodiesterases had a low apparent Km (0.29-0.49 microM) for guanosine 3':5' cyclic monophosphate (cGMP) and all peak III phosphodiesterases had a low apparent Km (0.35-0.58 microM) for adenosine 3':5' cyclic monophosphate (cAMP); trachealis peak III also had a high Km for cAMP (32 microM). The potency and selectivity for inhibition of peak I or peak III phosphodiesterase by theophylline and papaverine, the peak I selective inhibitor M + B 22948, and the peak III selective inhibitors amrinone, milrinone, imazodan, CI-930 and piroximone were approximately equal when isozymes isolated from aortic smooth muscle were compared to isozymes isolated from cardiac muscle of both species. Rolipram was relatively potent as a peak III phosphodiesterase inhibitor in canine cardiac muscle, but was impotent in the other cardiovascular peak IIIs. In tracheal smooth muscle, the cardiovascular selective peak III phosphodiesterase selective inhibitors were substantially less potent while rolipram was more potent as a peak III inhibitor. In summary, these studies show that while cardiac and vascular smooth muscle phosphodiesterase isozymes are pharmacologically similar, there is pharmacological and substrate heterogeneity of peak III phosphodiesterase in aortic vs. trachea smooth muscle within the same species.

In vivo cell interactions with calcium phosphate bioceramics.

Biointegration, resorption process, and solubility in physiological environments of calcium phosphate materials are scarcely described by ultrastructural studies. In vivo cells interactions with calcium phosphate materials are scarcely described by ultrastructural studies. In vivo cells interactions with calcium phosphate biomaterials are mediated by different proteins from physiological fluid, and in order to observe at the ultrastructural level the cell colonization, the resorption, process and the biointegration, we used in these experiments calcium phosphate materials precoated with fibronectin or not precoated. Two kinds of well determined materials were used for this study, Beta-tricalcium phosphate (B-TCP) and hydroxyapatite (HAP). The implants were soaked in human fibronectin diluted solution and were implanted in the connective tissue of rabbit abdomen. Our results showed that the fibroblasts and macrophagous++ cells interaction with the calcium phosphate crystal (B-TCP and HAP) was more important in the experiments with a fibronectin bilayer. In the presence of fibronectin at the grains surface of the material, cystic cavities' or fibrous encapsulation was suppressed and cells with fibers were in close contact with the material. The presence of fibronectin immediately after implantation seemed to increase the adhesion and the cell colonization. Fibronectin creates an organic interface between crystals and cells, and can promote interactions from cells and biomaterials.

Cardiovascular risk screening for women.

Cardiovascular disease is the leading cause of death of both men and women in Canada and the United States. The medical and societal emphasis on the occurrence of cardiovascular disease in men has resulted in an inclination to minimize its existence and severity in women. The purpose of this article is to assist clinical nurse specialists in cardiovascular risk-screening of women by providing a review of cardiovascular risk factors specific to women. Current knowledge about lipids, hypertension, diabetes, smoking, menopause, obesity, sedentary lifestyle, stress, and multiple roles are discussed. The clinical presentation for women and the clinical implications are presented. Lastly, implications for future research are described.

Molecular and histochemical characterisation of two distinct poplar Melampsora leaf rust pathosystems.

In this study, we compared interactions of two Melampsora foliar rust species with poplar, which resulted in either limited or abundant pathogen proliferation. In the pathosystem exhibiting limited pathogen growth, a defence response was observed after invasion of poplar leaf tissues by the biotroph, with late and clear production of reactive oxygen species (ROS) and other products. Characterisation of the histological, biochemical and transcriptional events occurring in both pathosystems showed striking similarity with components of plant defence reactions observed during qualitative resistance. Key components associated with development of an active defence response, such as up-regulation of pathogenesis-related (PR) genes, were observed during infection. Moreover, the time course and strength of gene induction appear to be critical determinants for the outcome of the tree-pathogen interaction. This work provides basic biochemical characterisation and expression data for the study of so-called partial resistance in the poplar-rust pathosystem, which is also applicable to other plant-pathogen interactions resulting in quantitative disease resistance.

Texture analysis of X-ray radiographs of iliac bone is correlated with bone micro-CT.

INTRODUCTION: Alteration of bone trabecular architecture is a predictor of fracture risk in osteoporosis. Until now, microarchitecture can only be measured on a bone biopsy, thus limiting microarchitecture analysis in routine clinical practice for osteoporosis. Texture analysis on X-ray images has been advocated to be a suitable means to assess two-dimensional (2-D) microarchitecture in the research field. But little is known about the relationships between three-dimensional (3-D) architecture and texture analysis, particularly in clinical practice. The purposes of the study were: (1) to explore the relationship between 3-D histomorphometric parameters and 2-D texture analysis, and (2) to see if cortical assessment may influence results. METHODS: In this study, the anterosuperior part of the iliac bone was removed from 24 cadavers. Large samples were prepared and comprised of the crest and a strip of bone approximately 3 cm wide and 5 cm long. These large specimens were used in order to preserve bone architecture; they also corresponded to the location used by histomorphometrists for the diagnosis of metabolic bone diseases on iliac crest biopsies. Bone samples were examined with a microcomputed tomograph for 3-D microarchitecture [BV/TV, C.BV/C.TV, Tb.P(f), structure model index (SMI), Tb.Th, Tb.N, Tb.Sp]. Texture analysis was done by several methods (skeletonization, run lengths, fractal techniques) from X-ray projection images. No correlation was found between bone mass parameters (BV/TV and C.BV/C.TV, which take into account both cortical and trabecular bone) and texture parameters. RESULTS: However, when specific descriptors of trabecular bone microarchitecture were used, several relationships with texture parameters were found [(Tb.N)/BOUND, r=0.628;/VGLN, r=0.596;/Fractal D, r=0.569]. CONCLUSION: When multiple correlations were used, the correlation coefficients were markedly improved with trabecular characteristics. X-ray texture analysis seemed to be a suitable approach for 2-D bone microarchitecture assessment. Furthermore, there is a good correlation between texture analysis of X-ray radiographs and 3-D bone microarchitecture assessed by microcomputed tomography.

Direct demonstration of a humorally-mediated inhibition of renal phosphate transport in the Hyp mouse.

The murine Hyp model reproduces the characteristics of human X-linked hypophosphatemia (XLH), an inherited disease causing renal loss of phosphate (Pi), severe rickets and osteomalacia. A current hypothesis considers that a humoral factor may be responsible for the renal Pi loss, although in vitro experiments with renal cell models have failed to demonstrate the presence of such a factor in XLH or in the Hyp mouse model. To test this hypothesis directly, we prepared primary mouse proximal tubule cell cultures (MPTC), expressing normal features of proximal tubule cells. These cells possess high alkaline phosphatase activity, and respond to human parathyroid hormone fragment 1-34 (PTH) with a four- to sixfold increase in cAMP production but do not respond to either arginine vasopressin (AVP) or to salmon calcitonin (sCT). They also show sodium-dependent phosphate, glucose and amino acid uptake. The presence of 10% Hyp mouse serum in HAMF12/DMEM media (1 mM Pi) for the last 48 hours of culture of MPTC reduced Pi uptake (0.1 mM 32P-Pi in the presence of 140 mM NaCl) by 45.7 +/- 3.9% (P < 0.01) as compared to normal mouse serum. This effect of Hyp mouse serum was dose-dependent between 5 to 20% (final concentration) in culture media for the last 48 hours of culture (P < 0.01 by analysis of variance). This effect of Hyp mouse serum was also time-dependent, with a lag time of at least 12 hours. Indeed, no significant inhibition of Pi uptake could be detected with incubations less than 12 hours in the presence of 10% Hyp mouse serum, whereas a maximal effect was obtained after 24 hours of incubation and remained unchanged after 36 and 48 hours. The inhibition of phosphate uptake by Hyp mouse serum was specific, since neither sodium-dependent glucose nor alpha-aminobutyric acid uptake was modified under these conditions. MPTC cells showed a very nice adaptation to Pi concentration in the media; low Pi (0.4 mM final concentration in the presence of 10% serum) stimulated Pi uptake, whereas high Pi concentration (3 mM) reduced Pi uptake by these cells as compared to regular HAMF12/DMEM media containing 1 mM Pi. Normal and Hyp mouse serum both inhibited Pi uptake by MPTC following adaptation in low or normal Pi media, however, Hyp mouse serum always showed a stronger inhibition than normal serum. In contrast, adaptation of MPTC in high Pi media resulted in no inhibition of phosphate uptake either in the presence of normal or Hyp mouse serum. We next questioned whether conditioned media from confluent Hyp mouse primary osteoblast-like cell cultures could affect Pi uptake by MPTC. These osteoblast-like cells expressed high alkaline phosphatase and produced the bone specific protein, osteocalcin. When MPTC were treated for 48 hours with Hyp mouse bone cell media conditioned for the last 48 hours of cultures, Pi uptake was specifically inhibited by 30.5 +/- 4.1% (P < 0.025) as compared to normal mouse bone cell-conditioned media. This effect of primary Hyp mouse bone cell-conditioned media is specific for these cells since it was not observed with CHO cell-conditioned media, nor with either mouse fibroblast (NCTC), normal mouse Kupffer cell- or Hyp mouse Kupffer cell-conditioned media. This effect also persisted through a number of passages of Hyp mouse bone cells, since conditioned-media from cells at their third passage still resulted in a 32 +/- 9.4% inhibition (P < 0.02). These results are the first to show an effect of Hyp mouse serum on Pi uptake by primary renal cell cultures in vitro. This effect is dose- and time-dependent, requiring 24 hours for maximum response, and is blocked in Pi rich media. These results also suggest that a specific intrinsic cellular defect, present in Hyp mouse osteoblasts, is responsible for the release of and/or the modification of a factor that can reach the circulation and which inhibits renal phosphate reabsorption. The molecular nature of this factor and its mode of action remains to be determined.

Endoglin expression on human microvascular endothelial cells association with betaglycan and formation of higher order complexes with TGF-beta signalling receptors.

Transforming growth factor-beta (TGF-beta) plays an important role in angiogenesis and vascular function. Endoglin, a transmembrane TGF-beta binding protein, is highly expressed on vascular endothelial cells and is the target gene for the hereditary haemorrhagic telangiectasia type I (HHT1), a dominantly inherited vascular disorder. The specific function of endoglin responsible for HHT1 is believed to involve alterations in TGF-beta responses. The initial interactions on the cell surface between endoglin and TGF-beta receptors may be an important mechanism by which endoglin modulates TGF-beta signalling, and thereby responses. Here it is shown that on human microvascular endothelial cells, endoglin is co-expressed and is associated with betaglycan, a TGF-beta accessory receptor with which endoglin shares limited amino acid homology. This complex formation may occur in either a ligand-dependent or a ligand-independent manner. In addition, the occurrence of three higher order complexes containing endoglin, type II and/or type I TGF-beta receptors, on these cells is demonstrated. Our findings suggest that endoglin may modify TGF-beta signalling by interacting with both betaglycan and the TGF-beta signalling receptors at physiological receptor concentrations and ratios.

Distinct profiles of phosphodiesterase isozymes in cultured cells derived from nonpigmented and pigmented ocular ciliary epithelium.

Alterations in either cyclic AMP (cAMP) or cyclic GMP (cGMP) may modulate the production of aqueous humor by the ciliary epithelium of the eye, thereby affecting intraocular pressure. We have found distinct profiles of phosphodiesterase (PDE) isozyme activity in cultured cells derived from bovine pigmented ciliary epithelium (PE) and cells derived from human nonpigmented ciliary epithelium (NPE), as well as corresponding differences in the effects of selective PDE inhibitors on the accumulation of cAMP and cGMP. In NPE cells, but not in PE cells, the major peak of PDE activity was stimulated by Ca++/calmodulin-stimulated (PDE I), and hydrolyzed both cAMP and cGMP. In contrast, PE cells contained a cGMP-specific PDE V not found in NPE cells. Rolipram, a selective inhibitor of PDE IV, was more potent and effective than the selective PDE III inhibitor CI-930 at potentiating intracellular cAMP accumulation in both cell types. Zaprinast, a selective inhibitor of PDE V, potentiated cGMP accumulation in PE but not in NPE cells. The results suggest that selective PDE inhibitors may modulate aqueous humor production by pigmented and nonpigmented ciliary epithelium, the two cell types may have different functional roles, and selective modulation of their functions may be possible. Furthermore, there may be distinct roles for intracellular calcium in regulating cGMP and cAMP in pigmented vs. nonpigmented ciliary epithelial cells.

Preconditioning of human smooth muscle cells via cyclopentenone prostaglandins protects against toxic effects of oxidized low-density lipoprotein.

Human vascular smooth muscle cells (SMC) exhibit upregulation of inducible heat shock protein 70 (Hsp70), upon exposure to oxidized low-density lipoproteins (LDL(ox)). The presence of Hsp70 is thought to protect the cell against the toxic effects of the modified lipoprotein. In order to test this hypothesis, Hsp70 in SMC was upregulated by exposure to Delta(12) prostaglandin J(2) (Delta(12)PGJ(2)) before cells were exposed to LDL(ox). Hsp70 levels were measured after exposure to Delta(12)PGJ(2) and before exposure to LDL(ox). Cell protection was monitored after LDL(ox) exposure by determination of cell toxicity measured by cell lactate dehydrogenase (LDH) release into the medium. Cells treated with Delta(12)PGJ(2) exhibited a 23-fold increase in Hsp70 levels and 56% lower LDH activity release after exposure to LDL(ox) when compared to cells that were not pretreated with Delta(12)PGJ(2). In addition, cells pretreated with prostaglandins that did not induce Hsp70 did not exhibit increased tolerance against the toxic effects of LDL(ox). The results support a protective role for Hsp70 against the toxic effects of LDL(ox) and hint at the potential for the use of small molecules for the prevention of deleterious effects of LDL(ox) through heat shock protein upregulation.

Characterization of endoglin on mouse uterine stromal cells.

During the oestrous cycle and early pregnancy, the uterus undergoes a variety of morphological and physiological modifications involving uterine cell proliferation and differentiation as well as extensive tissue remodelling. Transforming growth factor beta (TGF-beta) has powerful effects on these events and thus is thought to have a critical role in uterine physiology. Endoglin is a transmembrane glycoprotein that binds TGF-beta 1 and -beta 3 and interacts with TGF-beta signalling receptors to modulate many effects of this growth factor in different types of cell. Studies in mice revealed the highest concentrations of endoglin in the reproductive tract, notably on stromal cells of cyclic and pregnant uteri. The objective of the present study was to investigate the role of endoglin expressed on uterine stromal cells in binding TGF-beta and in the cellular responses induced by this growth factor. Highly purified populations of uterine stromal cells were isolated by cell affinity to the monoclonal antibody MJ7/18, which is specific to mouse endoglin. Affinity labelling of these cells with 125I-labelled TGF-beta followed by immunoprecipitation with endoglin-specific polyclonal 1256:4b antiserum indicated that endoglin expressed at the surface of uterine stromal cells binds TGF-beta 1 and interacts with TGF-beta signalling receptors. Treatment of uterine stromal cells with different concentrations of TGF-beta 1 induced a biphasic proliferative response and addition of MJ7/18 as well as neutralizing TGF-beta antibodies showed endoglin to be a modulator of TGF-beta-induced stromal cell proliferation. Given the importance of TGF-beta in the regulation of uterine physiology, these results indicate a role for endoglin during uterine tissue remodelling and decidualization.