1. alpha(2)-Macroglobulin is known to bind and inhibit a number of serine proteinases. We show that it binds thiol and carboxyl proteinases, and there is now reason to believe that alpha(2)-macroglobulin can bind essentially all proteinases. 2. Radiochemically labelled trypsin, chymotrypsin, cathepsin B1 and papain are bound by alpha(2)-macroglobulin in an approximately equimolar ratio. Equimolar binding was confirmed for trypsin by activesite titration. 3. Pretreatment of alpha(2)-macroglobulin with a saturating amount of one proteinase prevented the subsequent binding of another. We conclude that each molecule of alpha(2)-macroglobulin is able to react with one molecule of proteinase only. 4. alpha(2)-Macroglobulin did not react with exopeptidases, non-proteolytic hydrolases or inactive forms of endopeptidases. 5. The literature on binding and inhibition of proteinases by alpha(2)-macroglobulin is reviewed, and from consideration of this and our own work several general characteristics of the interaction can be discerned. 6. A model is proposed for the molecular mechanism of the interaction of alpha(2)-macroglobulin with proteinases. It is suggested that the enzyme cleaves a peptide bond in a sensitive region of the macroglobulin, and that this results in a conformational change in the alpha(2)-macroglobulin molecule that traps the enzyme irreversibly. Access of substrates to the active site of the enzyme becomes sterically hindered, causing inhibition that is most pronounced with large substrate molecules. 7. The possible physiological importance of the unique binding characteristics of alpha(2)-macroglobulin is discussed.

Objectives Generalised maternal endothelial cell dysfunction appears to be an underlying problem in pre‐eclampsia presumed to be caused, directly or indirectly, by one or more circulating factors derived from the placenta. Recently it has been suggested that tumour necrosis factor (TNF) may play an important role in pre‐eclampsia and contribute to endothelial activation. This study was designed to investigate this proposal. Design Plasma TNF‐α, IL‐6 and both forms of soluble TNF receptors (p55 and p75 TNF‐R) have been measured by ELISA in 31 pre‐eclamptic patients and 31 pregnant controls matched for age, parity and gestational age. Results Levels of IL‐6, TNF‐α and soluble TNF‐R (p55 and p75) were significantly higher in pre‐eclamptic patients, compared with age and gestation matched controls with a wide variation in levels between pre‐eclamptic individuals. There was a correlation between levels of IL‐6 and TNF or TNF‐R and between TNF and TNF‐R levels. However, when the pre‐eclamptic patients were subdivided on the basis of the severity of their disease, the median values of plasma concentrations of IL‐6, TNF‐α and TNF‐R were all higher in the group with lower platelet counts. Conclusions These new findings are consistent with the concept that the maternal syndrome of pre‐eclampsia is associated with endothelial dysfunction and provide evidence that at least part of this dysfunction could arise from excessive release of TNF‐α into the circulation.

Because of the widespread dispersion of mononuclear phagocytes throughout the body, there is little quantitative information on their total mass, relative numbers in different sites, or mobilization and redistribution in normal individuals or disease states. It is difficult to obtain such information by direct isolation of macrophages (M401 from all sites where they are known to occur and from tissues in which they are deeply embedded. The pool size of M4~ and their precursors in different compartments has been estimated by van Furth and his colleagues (1) by single-cell analysis after [SH]thymidine labeling, but this method favors enumeration of cells that turn over relatively rapidly, rather than more quiescent resident cells. In principle, antibodies that are specific for M4~ could be used to quantitate the content of antigen (Ag) in tissues directly, without cell separation. F4/80, a rat monoclonal antibody (Ab) directed against a plasma membrane glycoprotein of apparent Mr 160,000, is a specific and sensitive marker for mature mouse Mq~ after isolation (2) and in situ (3). In this study we have assayed F4/80 Ag content in various tissues of normal adult mice by adapting an absorption immunoassay developed by Williams and his colleagues (4) to measure lymphoid differentiation Ag in tissue lysates. We estimated M4~ number by calibrating F4/80 Ag content in a murine M4~ tumor-derived cell line, J774.2. Our results are in good agreement with immunohistochemical findings, and establish that relatively large amounts of F4/80 Ag are found not only in hemopoietic and lymphoid tissues, but also in other sites such as the gastrointestinal tract and normal kidney. This experimental approach provides a basis for further studies on the response of the mononuclear phagocyte system to inflammation and infection. Materials and MethodsAnimals. Tissues, organs, and peritoneal cells were obtained from 8-10-wk-old male mice of the C57BL/6 strain weighing 23 _ 2 g. Swiss mice, Pathology Oxford (PO), were used to prepare M~ target cells for binding assays. This work was supported in part by the Wellcome Trust and the Medical Research Council, UK. Address correspondence and reprint requests to S. Gordon. Preparation of M$ Target Cells. PO mice aged 8-10 wk were injected with 1.0 ml of thioglycollate broth intraperitoneally and peritoneal cells (TPC) harvested 4-6 d later. Cells were cultivated at 37°C in 5% CO~ in Dulbecco's modified Eagle's medium (DME) with 10% fetal calf serum (FCS) in 96-well plates (Sterilin Ltd., Middlesex, United Kingdom [UK]), at a density of 1 x 105 M4~ per well, for 2 d. They were then washed free of nonadherent cells, fixed with 0.125% vol/vol glutaraldehyde, and excess glutaraldehyde removed by washing and further incubation with 10% FCS in phosphatebuffered saline (PBS). Plates were stored in PBS with 10 mM sodium azide at 4°C for up to 12 wk. Preparation of Tissue and Cell Extracts.Organs: Animals were freshly killed by ether and organs removed into ice-cold extraction buffer (PBS containing 3 mM io...

Objectives To determine if placental syncytiotrophoblast microvillous (STBM) membranes contain factors which could cause the maternal endothelial cell disturbance thought to be central to the pathophysiology of the maternal syndrome of pre‐eclampsia. Design STMB membranes isolated from pre‐eclamptic or normal placentae were added to cultures of endothelial cells and their effect on the proliferation (measured by 3H‐thymidine incorporation), viability (measured by 51Cr release) and growth as a monolayer of these cells was determined. Membranes prepared from red blood cells, and nonendothelial adherent and nonadherent cell lines were used as specificity controls. Subjects STBM membranes were isolated from the placentae of primigravid women, 10 having caesarean sections for breech presentations and 10 for pre‐eclampsia. Results STBM membranes from the placentae of normal and pre‐eclamptic women suppressed endothelial cell proliferation to a similar extent and disrupted the cell monolayer to form a honeycomb‐like pattern. This change in morphology was seen before significant endothelial cell death occurred. Red blood cell membranes had no effect on either endothelial cell proliferation, viability or monolayer integrity. Endothelial cells from human umbilical arteries and bovine adrenal capillaries were similarly suppressed, but comparable concentrations of STBM membranes had no effect on nonendothelial cell lines. Conclusions Syncytiotrophoblast microvillous membranes specifically interfered with endothelial cell growth in vitro. Our results demonstrate that there are trophoblast products which could cause the maternal syndrome of pre‐eclampsia through endothelial cell damage.

No abstract