Capillary and arteriolar pericytes attract innate leukocytes exiting through venules and 'instruct' them with pattern-recognition and motility programs.

Coordinated navigation within tissues is essential for cells of the innate immune system to reach the sites of inflammatory processes, but the signals involved are incompletely understood. Here we demonstrate that NG2(+) pericytes controlled the pattern and efficacy of the interstitial migration of leukocytes in vivo. In response to inflammatory mediators, pericytes upregulated expression of the adhesion molecule ICAM-1 and released the chemoattractant MIF. Arteriolar and capillary pericytes attracted and interacted with myeloid leukocytes after extravasating from postcapillary venules, 'instructing' them with pattern-recognition and motility programs. Inhibition of MIF neutralized the migratory cues provided to myeloid leukocytes by NG2(+) pericytes. Hence, our results identify a previously unknown role for NG2(+) pericytes as an active component of innate immune responses, which supports the immunosurveillance and effector function of extravasated neutrophils and macrophages.

Distinct Pathogenesis of Pancreatic Cancer Microvesicle-Associated Venous Thrombosis Identifies New Antithrombotic Targets In Vivo.

OBJECTIVE: Cancer patients are at high risk of developing deep venous thrombosis (DVT) and venous thromboembolism, a leading cause of mortality in this population. However, it is largely unclear how malignant tumors drive the prothrombotic cascade culminating in DVT. APPROACH AND RESULTS: Here, we addressed the pathophysiology of malignant DVT compared with nonmalignant DVT and focused on the role of tumor microvesicles as potential targets to prevent cancer-associated DVT. We show that microvesicles released by pancreatic adenocarcinoma cells (pancreatic tumor-derived microvesicles [pcMV]) boost thrombus formation in a model of flow restriction of the mouse vena cava. This depends on the synergistic activation of coagulation by pcMV and host tissue factor. Unlike nonmalignant DVT, which is initiated and propagated by innate immune cells, thrombosis triggered by pcMV was largely independent of myeloid leukocytes or platelets. Instead, we identified externalization of the phospholipid phosphatidylethanolamine as a major mechanism controlling the prothrombotic activity of pcMV. Disrupting phosphatidylethanolamine-dependent activation of factor X suppressed pcMV-induced DVT without causing changes in hemostasis. CONCLUSIONS: Together, we show here that the pathophysiology of pcMV-associated experimental DVT differs markedly from innate immune cell-promoted nonmalignant DVT and is therefore amenable to distinct antithrombotic strategies. Targeting phosphatidylethanolamine on tumor microvesicles could be a new strategy for prevention of cancer-associated DVT without causing bleeding complications.

Disulfide HMGB1 derived from platelets coordinates venous thrombosis in mice.

Deep venous thrombosis (DVT) is one of the most common cardiovascular diseases, but its pathophysiology remains incompletely understood. Although sterile inflammation has recently been shown to boost coagulation during DVT, the underlying molecular mechanisms are not fully resolved, which could potentially identify new anti-inflammatory approaches to prophylaxis and therapy of DVT. Using a mouse model of venous thrombosis induced by flow reduction in the vena cava inferior, we identified blood-derived high-mobility group box 1 protein (HMGB1), a prototypical mediator of sterile inflammation, to be a master regulator of the prothrombotic cascade involving platelets and myeloid leukocytes fostering occlusive DVT formation. Transfer of platelets into Hmgb1-/- chimeras showed that this cell type is the major source of HMGB1, exposing reduced HMGB1 on their surface upon activation thereby enhancing the recruitment of monocytes. Activated leukocytes in turn support oxidation of HMGB1 unleashing its prothrombotic activity and promoting platelet aggregation. This potentiates the amount of HMGB1 and further nurtures the accumulation and activation of monocytes through receptor for advanced glycation end products (RAGE) and Toll-like receptor 2, leading to local delivery of monocyte-derived tissue factor and cytokines. Moreover, disulfide HMGB1 facilitates formation of prothrombotic neutrophil extracellular traps (NETs) mediated by RAGE, exposing additional HMGB1 on their extracellular DNA strands. Eventually, a vicious circle of coagulation and inflammation is set in motion leading to obstructive DVT formation. Therefore, platelet-derived disulfide HMGB1 is a central mediator of the sterile inflammatory process in venous thrombosis and could be an attractive target for an anti-inflammatory approach for DVT prophylaxis.

Differential Effects of Erythropoietin Administration and Overexpression on Venous Thrombosis in Mice.

BACKGROUND: Deep vein thrombosis (DVT) is a common condition associated with significant mortality due to pulmonary embolism. Despite advanced prevention and anticoagulation therapy, the incidence of venous thromboembolism remains unchanged. Individuals with elevated hematocrit and/or excessively high erythropoietin (EPO) serum levels are particularly susceptible to DVT formation. We investigated the influence of short-term EPO administration compared to chronic EPO overproduction on DVT development. Additionally, we examined the role of the spleen in this context and assessed its impact on thrombus composition. METHODS: We induced ligation of the caudal vena cava (VCC) in EPO-overproducing Tg(EPO) mice as well as wildtype mice treated with EPO for two weeks, both with and without splenectomy. The effect on platelet circulation time was evaluated through FACS analysis, and thrombus composition was analyzed using immunohistology. RESULTS: We present evidence for an elevated thrombogenic phenotype resulting from chronic EPO overproduction, achieved by combining an EPO-overexpressing mouse model with experimental DVT induction. This increased thrombotic state is largely independent of traditional contributors to DVT, such as neutrophils and platelets. Notably, the pronounced prothrombotic effect of red blood cells (RBCs) only manifests during chronic EPO overproduction and is not influenced by splenic RBC clearance, as demonstrated by splenectomy. In contrast, short-term EPO treatment does not induce thrombogenesis in mice. Consequently, our findings support the existence of a differential thrombogenic effect between chronic enhanced erythropoiesis and exogenous EPO administration. CONCLUSION: Chronic EPO overproduction significantly increases the risk of DVT, while short-term EPO treatment does not. These findings underscore the importance of considering EPO-related factors in DVT risk assessment and potential therapeutic strategies.

Protein disulfide isomerase acts as an injury response signal that enhances fibrin generation via tissue factor activation.

The activation of initiator protein tissue factor (TF) is likely to be a crucial step in the blood coagulation process, which leads to fibrin formation. The stimuli responsible for inducing TF activation are largely undefined. Here we show that the oxidoreductase protein disulfide isomerase (PDI) directly promotes TF-dependent fibrin production during thrombus formation in vivo. After endothelial denudation of mouse carotid arteries, PDI was released at the injury site from adherent platelets and disrupted vessel wall cells. Inhibition of PDI decreased TF-triggered fibrin formation in different in vivo murine models of thrombus formation, as determined by intravital fluorescence microscopy. PDI infusion increased - and, under conditions of decreased platelet adhesion, PDI inhibition reduced - fibrin generation at the injury site, indicating that PDI can directly initiate blood coagulation. In vitro, human platelet-secreted PDI contributed to the activation of cryptic TF on microvesicles (microparticles). Mass spectrometry analyses indicated that part of the extracellular cysteine 209 of TF was constitutively glutathionylated. Mixed disulfide formation contributed to maintaining TF in a state of low functionality. We propose that reduced PDI activates TF by isomerization of a mixed disulfide and a free thiol to an intramolecular disulfide. Our findings suggest that disulfide isomerases can act as injury response signals that trigger the activation of fibrin formation following vessel injury.

Chemokine fractalkine mediates leukocyte recruitment to inflammatory endothelial cells in flowing whole blood: a critical role for P-selectin expressed on activated platelets.

BACKGROUND: The membrane-bound chemokine fractalkine (CX3CL1) is expressed on various cell types such as activated endothelial cells and has been implicated in the inflammatory process of atherosclerosis. The aim of the present study was to dissect the role of fractalkine in leukocyte recruitment to inflamed endothelium under arterial shear forces. METHODS AND RESULTS: With the use of immunofluorescence and laminar flow assays, the present study shows that human umbilical vein endothelial cells stimulated with tumor necrosis factor-alpha and interferon-gamma abundantly express CX3CL1 and promote substantial leukocyte accumulation under arterial flow conditions. In the presence of high shear, firm adhesion of leukocytes to inflamed endothelial cells is reduced by approximately 40% by a function-blocking anti-fractalkine antibody or by an antibody directed against the fractalkine receptor (CX3CR1). With the use of intravital video-fluorescence microscopy we demonstrate that inhibition of fractalkine signaling attenuates leukocyte adhesion to the atherosclerotic carotid artery of apolipoprotein E-deficient mice, which suggests that the CX3CL1-CX3CR1 axis is critically involved in leukocyte adhesion to inflamed endothelial cells under high shear forces both in vitro and in vivo. Surprisingly, platelets were strictly required for fractalkine-induced leukocyte adhesion at high shear rates. Correspondingly, specific inhibition of platelet adhesion to inflamed endothelial cells also significantly reduced leukocyte accumulation. We show that both soluble and membrane-bound fractalkine induces platelet degranulation and subsequent surface expression of P-selectin, which thereby promotes direct platelet-leukocyte interaction. CONCLUSIONS: Fractalkine expressed by inflamed endothelial cells triggers P-selectin exposure on adherent platelets, which thereby initiates the local accumulation of leukocytes under arterial shear, an essential step in the development of atherosclerotic lesions.

Histological comparison of arterial thrombi in mice and men and the influence of Cl-amidine on thrombus formation.

AIMS: Medical treatment of arterial thrombosis is mainly directed against platelets and coagulation factors, and can lead to bleeding complications. Novel antithrombotic therapies targeting immune cells and neutrophil extracellular traps (NETs) are currently being investigated in animals. We addressed whether immune cell composition of arterial thrombi induced in mouse models of thrombosis resemble those of human patients with acute myocardial infarction (AMI). METHODS AND RESULTS: In a prospective cohort study of patients suffering from AMI, 81 human arterial thrombi were harvested during percutaneous coronary intervention and subjected to detailed histological analysis. In mice, arterial thrombi were induced using two distinct experimental models, ferric chloride (FeCl3) and wire injury of the carotid artery. We found that murine arterial thrombi induced by FeCl3 were highly concordant with human coronary thrombi regarding their immune cell composition, with neutrophils being the most abundant cell type, as well as the presence of NETs and coagulation factors. Pharmacological treatment of mice with the protein arginine deiminase (PAD)-inhibitor Cl-amidine abrogated NET formation, reduced arterial thrombosis and limited injury in a model of myocardial infarction. CONCLUSIONS: Neutrophils are a hallmark of arterial thrombi in patients suffering from acute myocardial infarction and in mouse models of arterial thrombosis. Inhibition of PAD could represent an interesting strategy for the treatment of arterial thrombosis to reduce neutrophil-associated tissue damage and improve functional outcome.

Retroinfusion of embryonic endothelial progenitor cells attenuates ischemia-reperfusion injury in pigs: role of phosphatidylinositol 3-kinase/AKT kinase.

BACKGROUND: Adult endothelial progenitor cells (EPCs) reduce myocardial infarct size and improve postischemic myocardial function. We have recently shown that clonal embryonic EPCs (eEPCs), derived from 7.5-day-old mice, home specifically to hypoxic areas in tumor metastasis mouse models but spare normal organs and do not form carcinomas. Here, we assessed the potential of eEPCs to limit organ dysfunction after ischemia and reperfusion in a preclinical pig model. METHODS AND RESULTS: Pigs were subjected to ischemia (60-minute left anterior descending [LAD] artery occlusion) and reperfusion (7 days). At the end of ischemia, we applied medium with or without 5 x 10(6) eEPCs by either pressure-regulated retroinfusion or intravenous transfusion. One hour after reperfusion, 99Tc-labeled eEPCs engrafted to a 6-fold higher extent in the ischemic myocardium after retroinfusion than after intravenous application. Regional myocardial function (subendocardial segment shortening [SES] at 150/min, given in percent of nonischemic circumflex region) and infarct size (TTC viability and Methylene-blue exclusion) were determined 24 hours and 7 days later. Compared with medium-treated animals, retroinfusion of eEPCs decreased infarct size (35+/-4% versus 51+/-6%) and improved regional myocardial reserve of the apical LAD region (SES 31+/-4% versus 6+/-8%), whereas intravenous application displayed a less pronounced effect (infarct size 44+/-4%; SES 12+/-3%). Retroinfusion of an equal amount of neonatal coronary endothelial cells (rat) did not affect infarct size (49+/-5%) nor regional myocardial reserve (16+/-7%). The eEPC-dependent effect was detected at 24 hours of reperfusion (infarct size 34+/-7% versus 58+/-6%) and was sensitive to Wortmannin coapplication (50+/-5%). CONCLUSIONS: Our findings show that eEPCs reduce ischemia-reperfusion injury in a preclinical pig model. The rapid effect (as early as 24 hours) indicates a role for enzyme-mediated cardioprotection, which involves, at least in part, the phosphatidylinositol 3-kinase/AKT pathway.

Selective retroinfusion of GSH and cariporide attenuates myocardial ischemia-reperfusion injury in a preclinical pig model.

OBJECTIVE: Reperfusion after ischemia may contribute to loss of myocardial function and increase in infarct size. Scavenging of reactive oxygen species (ROS) by glutathione (GSH) and inhibition of the sodium-proton-exchanger by cariporide are both capable of reducing myocardial reperfusion injury. We tested the efficacy of both agents applied regionally into the myocardium immediately before reperfusion. METHODS: Neonatal rat cardiomyocytes (NRCMs) were exposed to either hypoxia (H, 8 h)/reoxygenation (R, 1 h) or H2O2 (300 microM) in the presence or absence of GSH (10 mg/ml). In pigs (n=5 per group), percutaneous LAD occlusion was performed for 60 min. Application of GSH (250 mg/kg) and/or cariporide (1 mg/kg) was achieved by pressure-regulated retroinfusion of the anterior cardiac vein draining the ischemic area starting 5 min before reopening of the occluded LAD. Seven days later, subendocardial segment shortening (SES) was analyzed by sonomicrometry. Infarct size was determined by methylene-blue staining of the non-ischemic area and tetrazolium red staining of the viable myocardium in the area at risk (AAR). RESULTS: NRCM incubated with GSH (10 mg/ml) survived H/R or H2O2 (0.3 mM) to a larger extent than untreated cells. In pigs, infarct size of untreated hearts (51 +/- 6% of the AAR) was not significantly altered by GSH or cariporide retroinfusion alone (41 +/- 3% and 42 +/- 6%). In contrast, combined retroinfusion of cariporide and GSH significantly reduced infarct size (29 +/- 3%). SES of the infarcted area was improved only after cariporide/GSH retroinfusion as compared to untreated hearts. Additional systemic application of CD18-antibody IB4 (1.5 mg/kg) did not alter infarct size or SES in comparison to GSH/cariporide retroinfusion alone. CONCLUSION: Timely application of GSH scavenging ROS and cariporide targeting ion imbalance provides cardioprotection to the postischemic heart, which is superior to either treatment alone. The lack of an effect of additional IB4 treatment may indicate that GSH/cariporide retroinfusion itself affects leukocyte-dependent reperfusion injury.

Deep vein thrombus formation induced by flow reduction in mice is determined by venous side branches.

BACKGROUND: Interaction between vascular wall abnormalities, inflammatory leukocytes, platelets, coagulation factors and hemorheology in the pathogenesis of deep vein thrombosis (DVT) is incompletely understood, requiring well defined animal models of human disease. METHODS AND RESULTS: We subjected male C57BL/6 mice to ligation of the inferior vena cava (IVC) as a flow reduction model to induce DVT. Thrombus size and weight were analyzed macroscopically and sonographically by B-mode, pulse wave (pw) Doppler and power Doppler imaging (PDI) using high frequency ultrasound. Thrombus size varied substantially between individual procedures and mice, irrespective of the flow reduction achieved by the ligature. Interestingly, PDI accurately predicted thrombus size in a very robust fashion (r2 = 0.9734, p < 0.0001). Distance of the insertion of side branches from the ligature significantly determines thrombus weight (r2 = 0.5597, p < 0.0001) and length (r2 = 0.5441, p < 0.0001) in the IVC, regardless of the flow measured by pw-Doppler with distances <1.5 mm drastically impairing thrombus formation. Occlusion of side branches prior to ligation of IVC did not increase thrombus size, probably due to patent side branches inaccessible to surgery. CONCLUSION: Venous side branches influence thrombus size in experimental DVT and might therefore prevent thrombus formation. This renders vessel anatomy and hemorheology important determinants in mouse models of DVT, which should be controlled for.

Fractalkine is expressed in early and advanced atherosclerotic lesions and supports monocyte recruitment via CX3CR1.

Fractalkine (CX3CL1, FKN) is expressed in the inflamed vascular wall and absence of FKN reduces atherogenesis. Whether FKN is expressed throughout all stages of atherosclerotic disease and whether it directly contributes to monocyte recruitment to atherosclerotic lesions is not known. We collected human atherosclerotic plaque material and blood samples from patients with carotid artery disease undergoing endarterectomy. Plaques were analyzed by immunohistochemistry and qPCR. We found that FKN is expressed at all stages of atherosclerotic lesion formation, and that the number of FKN-expressing cells positively correlates with the number of CX3CR1-positive cells in human carotid artery plaques. In the circulation, soluble FKN levels are significantly elevated in the presence of high-grade (sub-occlusive) stenosis. To determine the role of the FKN-CX3CR1 axis for monocyte adhesion in vivo we then performed intravital videofluorescence microscopy of the carotid artery in ApoE(-/-) mice. Notably, FKN-CX3CR1 interactions are critical for recruitment of circulating monocytes to the injured atherosclerotic vascular wall. Thus, this chemokine dyad could represent an attractive target for anti-atherosclerotic strategies.

A novel role of sphingosine 1-phosphate receptor S1pr1 in mouse thrombopoiesis.

Millions of platelets are produced each hour by bone marrow (BM) megakaryocytes (MKs). MKs extend transendothelial proplatelet (PP) extensions into BM sinusoids and shed new platelets into the blood. The mechanisms that control platelet generation remain incompletely understood. Using conditional mutants and intravital multiphoton microscopy, we show here that the lipid mediator sphingosine 1-phosphate (S1P) serves as a critical directional cue guiding the elongation of megakaryocytic PP extensions from the interstitium into BM sinusoids and triggering the subsequent shedding of PPs into the blood. Correspondingly, mice lacking the S1P receptor S1pr1 develop severe thrombocytopenia caused by both formation of aberrant extravascular PPs and defective intravascular PP shedding. In contrast, activation of S1pr1 signaling leads to the prompt release of new platelets into the circulating blood. Collectively, our findings uncover a novel function of the S1P-S1pr1 axis as master regulator of efficient thrombopoiesis and might raise new therapeutic options for patients with thrombocytopenia.

Monocytes, neutrophils, and platelets cooperate to initiate and propagate venous thrombosis in mice in vivo.

Deep vein thrombosis (DVT) is a major cause of cardiovascular death. The sequence of events that promote DVT remains obscure, largely as a result of the lack of an appropriate rodent model. We describe a novel mouse model of DVT which reproduces a frequent trigger and resembles the time course, histological features, and clinical presentation of DVT in humans. We demonstrate by intravital two-photon and epifluorescence microscopy that blood monocytes and neutrophils crawling along and adhering to the venous endothelium provide the initiating stimulus for DVT development. Using conditional mutants and bone marrow chimeras, we show that intravascular activation of the extrinsic pathway of coagulation via tissue factor (TF) derived from myeloid leukocytes causes the extensive intraluminal fibrin formation characteristic of DVT. We demonstrate that thrombus-resident neutrophils are indispensable for subsequent DVT propagation by binding factor XII (FXII) and by supporting its activation through the release of neutrophil extracellular traps (NETs). Correspondingly, neutropenia, genetic ablation of FXII, or disintegration of NETs each confers protection against DVT amplification. Platelets associate with innate immune cells via glycoprotein Ibα and contribute to DVT progression by promoting leukocyte recruitment and stimulating neutrophil-dependent coagulation. Hence, we identified a cross talk between monocytes, neutrophils, and platelets responsible for the initiation and amplification of DVT and for inducing its unique clinical features.

IRAG mediates NO/cGMP-dependent inhibition of platelet aggregation and thrombus formation.

Defective regulation of platelet activation/aggregation is a predominant cause for arterial thrombosis, the major complication of atherosclerosis triggering myocardial infarction and stroke. A central regulatory pathway conveying inhibition of platelet activation/aggregation is nitric oxide (NO)/cyclic GMP (cGMP) signaling by cGMP-dependent protein kinase I (cGKI). However, the regulatory cascade downstream of cGKI mediating platelet inhibition is still unclear. Here, we show that the inositol-1,4,5-trisphosphate receptor-associated cGMP kinase substrate (IRAG) is abundantly expressed in platelets and assembled in a macrocomplex together with cGKIbeta and the inositol-1,4,5-trisphosphate receptor type I (InsP3RI). cGKI phosphorylates IRAG at Ser664 and Ser677 in intact platelets. Targeted deletion of the IRAG-InsP3RI interaction in IRAGDelta12/Delta12 mutant mice leads to a loss of NO/cGMP-dependent inhibition of fibrinogen-receptor activation and platelet aggregation. Intracellular calcium transients were not affected by DEA/NO or cGMP in mutant platelets. Furthermore, intravital microscopy shows that NO fails to prevent arterial thrombosis of the injured carotid artery in IRAGDelta12/Delta12 mutants. These findings reveal that interaction between IRAG and InsP3RI has a central role in NO/cGMP-dependent inhibition of platelet aggregation and in vivo thrombosis.

Endothelial nitric oxide synthase overexpression provides a functionally relevant angiogenic switch in hibernating pig myocardium.

OBJECTIVES: We investigated whether retroinfusion of liposomal endothelial nitric oxide synthase (eNOS) S1177D complementary deoxyribonucleic acid (cDNA) would affect neovascularization and function of the ischemic myocardium. BACKGROUND: Recently, we demonstrated the feasibility of liposomal eNOS cDNA transfection via retroinfusion in a model of acute myocardial ischemia/reperfusion. In the present study, we used this approach to target a phosphomimetic eNOS construct (eNOS S1177D) into chronic ischemic myocardium in a pig model of hibernation. METHODS: Pigs (n = 6/group) were subjected to percutaneous implantation of a reduction stent graft into the left anterior descending artery (LAD), inducing total occlusion within 28 days. At day 28, retroinfusion of saline solution containing liposomal green fluorescent protein or eNOS S1177D cDNA (1.5 mg/animal, 2 x 10 min) was performed. Furthermore, L-nitroarginine-methylester (L-NAME) was applied orally from day 28, where indicated. At day 28 and day 49, fluorescent microspheres were injected into the left atrium for perfusion analysis. Regional functional reserve (at atrial pacing 140/min) was assessed at day 49 by subendocardial segment shortening (SES) (sonomicrometry, percent of ramus circumflexus region). RESULTS: The eNOS S1177D overexpression increased endothelial cell proliferation as well as capillary and collateral growth at day 49. Concomitantly, eNOS S1177D overexpression enhanced regional myocardial perfusion from 62 +/- 4% (control) to 77 +/- 3% of circumflex coronary artery-perfused myocardium, unless L-NAME was co-applied (69 +/- 5%). Similarly, eNOS S1177D cDNA improved functional reserve of the LAD (33 +/- 5% vs. 7 +/- 3% of circumflex coronary artery-perfused myocardium), except for L-NAME coapplication (13 +/- 6%). CONCLUSIONS: Retroinfusion of eNOS S1177D cDNA induces neovascularization via endothelial cell proliferation and collateral growth. The resulting gain of perfusion enables an improved functional reserve of the hibernating myocardium.

Percutaneous approach to a stent-based ventricle to coronary vein bypass (venous VPASS): comparison to catheter-based selective pressure-regulated retro-infusion of the coronary vein.

AIMS: Percutaneous stent-based ventricle-to-coronary vein bypass (venous VPASS) is a new approach to chronic venous arterialization as a treatment modality in an otherwise no option patient with coronary artery disease. In this study, the efficacy of venous VPASS was compared with catheter-based selective pressure-regulated retro-infusion of arterial blood during acute ischaemia. METHODS AND RESULTS: In seven pigs, venous VPASS was established using a percutaneous ultrasound-guided puncture from the anterior cardiac vein to the left ventricle, with subsequent implantation of an ePTFE-covered stent graft. During left anterior descending artery (LAD) occlusion, coronary venous pressure in the distal anterior cardiac vein increased to 55+/-4 mmHg under conditions of venous VPASS compared with 78+/-5 mmHg during selective pressure-regulated retro-infusion. Significant preservation of regional myocardial function was observed during venous VPASS (67+/-6% baseline) and during selective retro-infusion (83+/-4%) compared with control LAD occlusion (0.4+/-2%). CONCLUSION: Percutaneous implantation of a PTFE covered stent (venous VPASS) was feasible and associated with significant preservation of regional myocardial function during acute ischaemia in pigs at reasonable levels of mean coronary venous pressure to avoid tissue damage during chronic application.