Marc Vasse

Frequency of protein Z deficiency in patients with ischaemic stroke

Summary Prothrombotic phenotype has been described in protein- Z deficient mice, but the thrombotic risk associated with protein Z deficiency in human beings is unknown. We saw a protein Z plasma concentration deficiency of about 20% in 169 patients, from two hospitals, who had ischaemic stroke, whereas the frequency in 88 controls was about 5%. We saw no increase in the frequency of protein Z deficiency in 56 patients with venous thrombophilia. However, why protein Z deficiency was only observed in arterial thrombosis remains unknown.

Oncostatin M induces angiogenesis in vitro and in vivo.

Neovascularization of the atherosclerotic plaque is responsible for its weakening and consequently for the complications of vascular disease. Macrophages are a source of growth factors that can modulate angiogenesis. In this study, we analyzed the effect of oncostatin M (OSM) on angiogenesis, as it could be involved in the development of atherosclerosis. The effect of OSM was compared with those of leukemia inhibitory factor (LIF) and interleukin-6 (IL-6). On human dermal microvasculature endothelial cells (HMEC-1s), OSM (22.5 to 112.5 pmol/L) induced a dose-dependent increase in cell proliferation greater than that induced by the classic angiogenic factors vascular endothelial growth factor (VEGF; 543 pmol/L) and basic fibroblast growth factor (bFGF; 1.1 nmol/L). LIF (19 to 475 pmol/L) induced only a 30% increase in cell proliferation, and IL-6 had no effect. Furthermore, in a modified Boyden-chamber model, OSM, LIF, and IL-6 were chemoattractant for HMEC-1s. In a tridimensional gel of fibrin, OSM increased tube formation and tube length, which were already noticeable by day 3. LIF and IL-6 induced a weaker effect that was only obvious by day 10. The angiogenic effect of OSM was also demonstrated in vivo in a rabbit corneal model: OSM was more potent than LIF, the length of the neovessels being longer with OSM than with LIF, whereas IL-6 was without effect. We tested factors that could be involved in the proliferative effect of OSM on HMEC-1s. OSM induced only a slight increase in the urokinase receptor and a 60% increase in VEGF secretion, whereas it does not modify IL-8 secretion or bFGF levels. The effect of OSM seems to depend on endothelial cell origin and cell species: OSM (up to 112.5 pmol/L) did not induce human umbilical vein endothelial cell proliferation and even had a small inhibitory effect (17%) on calf pulmonary artery endothelial cells. In conclusion, OSM induces an angiogenic effect on capillary endothelial cells, which could be, at least in part, implicated in pathological processes such as atherosclerosis or tumor growth.

Down-regulation of fibrinogen biosynthesis by IL-4, IL-10 and IL-13

High levels of fibrinogen are recognized as an important vascular risk factor; however, it is not known if the increase of plasma fibrinogen is directly responsible for this risk, or is only a marker of vascular inflammation. To support this second hypothesis, Oncostatin M (OSM) is a potent stimulator of fibrinogen biosynthesis and induces smooth muscle cell proliferation. In the same way, we analysed whether interleukin‐4 (IL‐4), interleukin‐10 (IL‐10) or interleukin‐13 (IL‐13), which protect vessel walls from monocytes injuries leading to atherosclerosis, could influence fibrinogen biosynthesis. The two levels of regulation of fibrinogen biosynthesis were tested: firstly, the direct effect of these cytokines on fibrinogen production by the hepatoma cell line Hep G2, and secondly their effect on the secretion of hepatocyte stimulating factor (HSF) activity in the supernatant of lipopolysaccharide (LPS)‐activated monocytes. IL‐4 and IL‐13 added to Hep G2 cells down‐regulated both the increase of fibrinogen secretion induced by IL‐6 and fibrinogen mRNA levels, this effect being more pronounced when Hep G2 were preincubated with the two cytokines before IL‐6 addition. The effect of IL‐10 was evidenced only on mRNA expression. IL‐10 and IL‐13 dose‐dependently decrease HSF activity secreted by LPS‐activated monocytes, whereas IL‐4 had no effect. However, the three cytokines decreased HSF activity when monocytes were incubated with the cytokines before LPS activation. The effects of these cytokines on HSF activity are related to variations of IL‐6 and OSM secretion. Our data strengthen the hypothesis that the fibrinogen level is a marker of vascular disease, since cytokines which have a protective vascular effect down‐regulate fibrinogen production.

Resistance to activated protein c: Evaluation of three functional assays

Resistance to Activated Protein C (APC) was evaluated using 3 different methods: two of them were based on the prolongation of the Activated Partial Thromboplastin Time (APTT) using 2 different APTT reagents in the presence of APC, whereas the third method was based on the prolongation of prothrombin time when APC is added. The three methods were significantly correlated. APTT-based assays were sensitive to factor XII deficiency, whereas thromboplastin-based assay was sensitive to factor VII deficiency (< 0.5 UI/ml), which surestimates the response to APC. In contrast, an increase in factor VIII (F. VIII) level is associated with a decreased response to APC, when APTT-based assays are used, whereas thromboplastin-based assay is unmodified. During pregnancy, a decreased response to APC is observed, which is not only due to the increase in F. VIII, since thromboplastin-based assay is also modified. In Protein S (PS) immuno-depleted plasma, the low response to APC is corrected by addition of free PS: the thromboplastin-based assay was the most sensitive one to PS deficiency. However, in patients with congenital PS deficiency, there was no correlation between APC-resistance and free PS level. In patients with lupus anticoagulant, discrepancies were observed between the 3 methods, but with a high frequency of low response to APC. For the 3 assays, there was a good differentiation and correlation between normal and pathological results, the thromboplastin-based assay being perhaps the most discriminating. However, 3 unrelated thrombophilic patients showed normal results using thromboplastin-based assay, although they were APC-resistant using APTT-based assays. For 2 patients, this discrepancy can be explained by high levels of F. VIII. For the last patient, an abnormal F. VIII, resistant to APC can be suspected.

Cyclooxygenase‐2 activity is necessary for the angiogenic properties of oncostatin M

Macrophages play a major role in angiogenesis. We recently reported that oncostatin M (OSM), a cytokine of the interleukin (IL)‐6 family secreted by macrophages, has a potent angiogenic activity on human microvascular endothelial cells (HMEC‐1), but has no effect on macrovascular cells (human umbilical vein endothelial cells (HUVECs)). In this work, we show that in HMEC‐1, OSM (0.5–2.5 ng/ml), leukemia inhibitory factor (LIF) (25 ng/ml), bFGF (25 ng/ml) and IL‐1β (5 ng/ml) induced production of cyclooxygenase (COX)‐2. In contrast, in HUVECs, neither OSM nor LIF induced COX‐2 mRNA, suggesting that COX‐2 might be implicated in the angiogenic activity of OSM. This was confirmed by the inhibiting effect on OSM‐induced HMEC‐1 proliferation of specific COX‐2 inhibitors. In vivo studies confirmed this findings. We conclude that induction of COX‐2 by OSM is necessary for its angiogenic activity, but is not sufficient since IL‐1β, which also induces COX‐2 in HMEC‐1, has only a poor proliferative effect.

Protein Z, a protein seeking a pathology

Protein Z (PZ) is a vitamin K-dependent factor identified in human plasma in 1984 characterized by an homology with other vitamin K-dependent factors (factor VII, IX, X, protein C). In contrast to these factors, PZ does not possess any enzymatic activity but is involved as a cofactor in the down-regulation of coagulation by forming a complex with the protein Z-dependent protease inhibitor (ZPI). ZPI inhibits the activated factor X (FXa) on phospholipid surface. In mice, the disruption of PZ gene is asymptomatic, but the association with the factor V Leiden mutation leads to a quasi complete mortality during the neonatal period with microvascular thrombosis. In humans, PZ is characterized by an unusual wide distribution in plasma, and a major decrease induced by warfarin. Isolated PZ deficiency does not seem to constitute a risk for venous thrombosis, but a severe PZ deficiency could increase the risk of well recognized venous thrombotic risk factors such as factor V Leiden, G20210A mutation or hyperhomocysteinemia. Unexpectedly, a relationship between PZ deficiency and ischemic arterial diseases such as stroke, acute coronary syndromes or peripheral arterial disease was described but not confirmed by all studies. PZ deficiency could be also a risk factor for early fetal losses, and increases the arterial risk in antiphospholipid syndrome. This review analyzes the different studies so far published and discusses the various results obtained in order to understand whether or not protein Z deficiency could be considered as an arterial ischemic risk factor.

Endothelial metalloprotease-disintegrin protein (ADAM) is implicated in angiogenesis in vitro

Recently two metalloproteinase, disintegrin, cysteine proteins (MDCs), also called ADAMs were identified on endothelial cells. However the role of these ADAMs are not defined on these cells. In order to elucidate whether ADAMs associated with endothelial cells could be involved in angiogenesis, we have tested the effect of an inhibitor of ADAM (GL 129471) in models of angiogenesis in vitro. Our results showed that GL 129471 inhibited endothelial cell migration and adhesion and increased the number of cells in the G2/M phase leading to an inhibition of cell proliferation. The effects of GL 129471 are not mimicked by the endogenous matrix metalloproteinase inhibitor TIMP-2. These data suggest that ADAMs may play important role in angiogenesis and could provide a new target for inhibition of angiogenesis in cancers.

Cooperation between monocytes and breast cancer cells promotes factors involved in cancer aggressiveness

In breast cancers, clinical symptoms of inflammation localised around the tumour at the time of diagnosis have been considered to have poor prognosis significance. In this study, the biological mechanisms responsible for the deleterious action of monocytes in cancer were investigated. The incubation of the breast-cancer-derived MDA-MB231 cells with monocytes resulted in an increase in factors involved in cell invasion (i.e. both cancer cells and monocytes-associated urokinase and Tissue Factor, and PAI-1 and MMP-9 secretion). Moreover, the functions of monocytes were also modified. Incubation of monocytes with MDA-MB231 cancer cells resulted in a downregulation in the secretion of the antiproliferative cytokine Oncostatin M, while the apoptotic factor TNF alpha was dramatically increased. However, MDA-MB231 cancer cells have been shown to be resistant towards the apoptotic action of TNF alpha. These findings demonstrate that incubation of MDA-MB231 cancer cells with monocytes induced a crosstalk, which resulted in an increased expression of factors involved in cancer cell invasiveness and in a modification of monocytes function against cancer cells, while inflammatory effects were increased.

Increased levels of tissue factor activity and procoagulant phospholipids during treatment of children with acute lymphoblastic leukaemia

The use of l‐asparaginase (l‐ASP) in paediatric patients with acute lymphoblastic leukaemia (ALL) is associated with thrombotic complications. We evaluated the activities of tissue factor (TFa), thrombomodulin (TMa) and procoagulant phospholipids (PPL) in 26 consecutive children with ALL (25 B‐ALL and one T‐ALL) treated by the French Acute Lymphoblastic Leukemia group (FRALLE)‐2000 protocol. Samples were obtained at diagnosis, after glucocorticoid (GC) therapy, during the induction phase with l‐ASP, vincristine (VCR) and adriamycin (ADR), during the re‐induction and within the week after treatment. Plasma levels of TFa, TMa and PPL increased gradually and significantly during the different phases of the treatment, with higher levels observed during the induction period, and decreased after treatment discontinuation. In vitro studies showed that the different drugs used for ALL treatment could induce a weak expression of TF and procoagulant activity (PCA) on normal and leukaemia blood cells, while a marked effect was observed on endothelial cells. In conclusion, these data indicate that, in addition to the well‐identified increased in coagulation factors and inhibitor deficiencies, the injury of the endothelium could lead to the release of TF and PPL and could contribute to the hypercoagulability of children treated for ALL.

Human endothelial cells synthesize protein Z, but not the protein Z dependent inhibitor

Protein Z (PZ) is a vitamin K-dependent protein isolated from human plasma, and acts as a cofactor for a serpin, called protein Z-dependent protease inhibitor (ZPI). A prothrombotic phenotype has been reported in PZ deficient mice, and PZ deficiencies have been observed in patients with arterial thrombotic events. PZ was immunologically detected in the endothelium of atherosclerotic arteries, suggesting that endothelial cells could be involved in the production of PZ. In this study we analyzed the synthesis and release of PZ and ZPI by human umbilical vein endothelial cells (HUVEC), representative of the macrovasculature, and by HMEC-1, a microvascular endothelial cell line. PZ was quantified by a specific ELISA in the supernatant and in the lysates of both cellular types. Western blotting of the supernatants showed the presence of a band of 62 kDa, identical to PZ synthesized by the hepatoma cell line HepG2. mRNA of PZ was also detected in each cellular type. PZ biosynthesis was unaffected by inflammatory cytokines in HUVEC, whereas a slight decrease of mRNA and PZ antigen (53.5 +/- 14.5% of protein synthesis as compared to the control, p < 0.01) and a modest increase (126 +/- 8.5% as compared to the control, p < 0.05) were induced respectively byTumor Necrosis Factor (TNF)-alpha (25 ng/ml) and oncostatin M (5 ng/ml) in HMEC-1. Immunological studies showed the presence of PZ near the nucleus and a possible expression of PZ at the membrane. In addition, PZ was present in the endothelial cells of both normal arterial and venous vessel sections. In contrast, neither ZPI nor its mRNA was detected in endothelial cells.