ALBERT

Description: Drug targeting to inflammatory brain pathologies such as stroke and traumatic brain injury remains an elusive goal. Using a mouse model of acute brain inflammation induced by local tumor necrosis factor alpha (TNFα), we found that uptake of intravenously injected antibody to vascular cell adhesion molecule 1 (anti-VCAM) in the inflamed brain is 〉10-fold greater than antibodies to transferrin receptor-1 and intercellular adhesion molecule 1 (TfR-1 and ICAM-1). Furthermore, uptake of anti-VCAM/liposomes exceeded that of anti-TfR and anti-ICAM counterparts by ∼27- and ∼8-fold, respectively, achieving brain/blood ratio 〉300-fold higher than that of immunoglobulin G/liposomes. Single-photon emission computed tomography imaging affirmed specific anti-VCAM/liposome targeting to inflamed brain in mice. Intravital microscopy via cranial window and flow cytometry showed that in the inflamed brain anti-VCAM/liposomes bind to endothelium, not to leukocytes. Anti-VCAM/LNP selectively accumulated in the inflamed brain, providing de novo expression of proteins encoded by cargo messenger RNA (mRNA). Anti-VCAM/LNP-mRNA mediated expression of thrombomodulin (a natural endothelial inhibitor of thrombosis, inflammation, and vascular leakage) and alleviated TNFα-induced brain edema. Thus VCAM-directed nanocarriers provide a platform for cerebrovascular targeting to inflamed brain, with the goal of normalizing the integrity of the blood–brain barrier, thus benefiting numerous brain pathologies.

Description: Sepsis induces a procoagulant state, which in its most extreme form can result in disseminated intravascular coagulation (DIC), a syndrome of widespread microvascular thrombosis, ischemia, and organ dysfunction. Suppression of the protein C (PC) pathway has long been thought to be an important part of the pathophysiology, leading to development of a number of candidate therapeutics, including soluble human thrombomodulin (shTM, or ART123), currently being evaluated in a phase III clinical trial. While conferring less risk of bleeding toxicity than previous pharmacologic interventions into the protein C pathway, shTM fails to reproduce the spatial localization of its endogenous counterpart, preventing optimal interaction with PC bound to its endothelial receptor. To test the importance of localization to the endothelial membrane, we compared the antithrombotic activity of shTM to that of hTM/R6.5, a fusion protein therapeutic that anchors recombinant human TM to endothelial ICAM-1. In modified thrombin generation assays, in which platelet-poor plasma was exposed to cytokine-activated human endothelial cells, cell bound hTM/R6.5 was more potent than shTM mixed into the plasma. A similar result was found in a human whole blood microfluidic assay of sepsis-induced microvascular thrombosis, in which endothelial bound hTM/R6.5 more effectively inhibited fibrin deposition than its soluble counterpart. To investigate its mechanism of action, the fusion protein was tested in the setting of moderate PC deficiency, a common occurrence in patients with DIC. While PC depletion alone had no effect on fibrin deposition in this model, it significantly reduced the efficacy of hTM/R6.5, which was restored by supplementation using plasma-derived PC concentrate. These results indicate that endothelial bound TM acts not only as a direct inhibitor of thrombin, but derives antithrombotic activity from the activation of PC, such that it may be insufficient to reverse the coagulopathy of patients with significant deficiency in plasma PC. Likewise, the data match the clinical experience with PC supplementation, which appears to be of nominal benefit in sepsis or DIC, despite repeated demonstration of efficacy in patients with severe congenital PC deficiency, who have normal levels of endothelial TM and EPCR. The synergistic effect observed with simultaneous replacement of plasma PC and endothelial TM, however, suggest that this combination may represent a novel, translational therapeutic strategy. Figure (a) Thrombin generation assay using fluorescent thrombin substrate. Thrombin was generated following addition of platelet poor plasma to static monolayers of tissue factor expressing, TNF-α activated HUVEC. Treatment of cells with hTM/R6.5, followed by washing to remove non-specifically bound fusion protein, was more effective than addition of shTM to the plasma. Inset shows AUC (area under curve) analysis: * = p 〈 0.05 vs. TNF only, # = p 〈 0.05 for hTM/R6.5 vs. shTM at each concentration. (b) Fibrin generation (measured using AF647-conjugated fibrin antibody) in whole blood microfluidic assay. 3D-confluent endothelial monolayers were grown within flow chambers and activated with TNF-α prior to the infusion of whole blood at 5 dynes/cm2. hTM/R6.5 shows more prolonged inhibition of coagulation than shTM. Inset shows fluorescence microscopy images at 30 min after infusion of whole blood and AUC analysis: * = p 〈 0.05 vs. TNF only. (c) Fibrin deposition in control vs. PC deficient blood. PC supplementation with plasma-derived concentrate restores the anti-thrombotic activity of hTM/R6.5. Inset shows fluorescence microscopy images for each condition at 7.5 min after infusion of whole blood. AUC analysis is also shown: ** = p 〈 0.05 vs. all other conditions. Figure. (a) Thrombin generation assay using fluorescent thrombin substrate. Thrombin was generated following addition of platelet poor plasma to static monolayers of tissue factor expressing, TNF-α activated HUVEC. Treatment of cells with hTM/R6.5, followed by washing to remove non-specifically bound fusion protein, was more effective than addition of shTM to the plasma. Inset shows AUC (area under curve) analysis: * = p 〈 0.05 vs. TNF only, # = p 〈 0.05 for hTM/R6.5 vs. shTM at each concentration. (b) Fibrin generation (measured using AF647-conjugated fibrin antibody) in whole blood microfluidic assay. 3D-confluent endothelial monolayers were grown within flow chambers and activated with TNF-α prior to the infusion of whole blood at 5 dynes/cm2. hTM/R6.5 shows more prolonged inhibition of coagulation than shTM. Inset shows fluorescence microscopy images at 30 min after infusion of whole blood and AUC analysis: * = p 〈 0.05 vs. TNF only. (c) Fibrin deposition in control vs. PC deficient blood. PC supplementation with plasma-derived concentrate restores the anti-thrombotic activity of hTM/R6.5. Inset shows fluorescence microscopy images for each condition at 7.5 min after infusion of whole blood. AUC analysis is also shown: ** = p 〈 0.05 vs. all other conditions. Disclosures No relevant conflicts of interest to declare.

Description: Introduction: We previously demonstrated prolonged circulation and improved efficacy of erythrocyte (red blood cell, RBC) coupled thrombomodulin (TM) fusion proteins as thromboprophylactic and anti-inflammatory agents. To further this therapeutic platform, we produced humanized analogues and investigated their efficacy in models where RBC-coupling may provide not only a pharmacokinetic advantage, but also a pharmacodynamic advantage by local delivery to RBC membranes. Previous fusions were constructed with mouse TM fused to single-chain variable fragments (scFv) of a rat-derived anti-glycophorin A antibody. Therefore, to produce clinically translatable therapeutic fusion proteins, modification of both cargo and targeting moiety are necessary for use in humans to render them non-immunogenic and capable of binding to human RBCs. We aimed to use human-like antibodies against high-prevalence antigens, fuse these antibodies to the extracellular domain of TM, confirm their binding and enzymatic activity, and demonstrate their efficacy in whole-blood models of human vascular pathology. Methods: An IgG Fab phage display library was prepared from a Rhesus macaque immunized with human RBCs. By panning on intact RBCs, Fab/phage specific for human RBCs were identified. We selected two clones from this library, one against a high-prevalence Band 3 antigen (Wright B (Wrb), Diego blood group, 〉106 copies/RBC) and one against a high-prevalence RhCE antigen (Rh17, Rhesus blood group, ~105 copies/RBC). The variable chain sequences of these antibodies were cloned into a scFv construct and fused to the extracellular domain of human TM. Fusion proteins were produced in S2 cells using a metallothionein promoter expression system. The soluble extracellular domain of human TM was similarly cloned and produced as a control. Binding of the fusions to RBCs was measured by indirect agglutination and by ELISA with immobilize erythrocyte ghosts. Activity of the TM fusions was measured by colorimetric cleavage of an APC substrate. Adverse effects of fusions on RBCs were investigated in a model of osmotic stress in hypotonic saline as well as mechanical integrity against agitation with glass beads. Assays of endothelial protection were performed on human umbilical vein endothelial cells (HUVEC) activated lipopolysaccharide (LPS), TNF-α, heme, and HMGB1. Culture supernatants were analyzed by ELISA for IL-6, IL-8, and vWF. An endothelialized, whole-blood microfluidic platform was developed using the Bio-Flux (Fluxion) microfluidic system. Models of vascular pathology included TNF-α activation of the endothelialized channels as well as localized, light-induced injury with hematoporphyrin. Results: We successfully produced fusion proteins of human TM and human-like scFv antibody derivatives capable of specifically binding to human RBCs (Figure 1, a). The fusions demonstrated affinities suitable for translation, and enabled comparison of different levels of surface loading. The scFv/TM fusions did not induce direct RBC agglutination, and did not have adverse effects on RBC integrity under osmotic and mechanical stress. The fusions maintained similar enzymatic activity to their soluble TM counterparts (Figure 1, b) and remained active when bound to RBCs (Figure 1, c). The TM fusions also demonstrated efficacy in protection of HUVECs against activation by inflammatory mediators such as LPS and thrombin (Figure 1, d-f), both as soluble proteins and when bound to RBCs. In a whole-blood endothelialized microfluidic system, the fusions reduced fibrin deposition and channel occlusion in activated and injured endothelium. Figure 1 - (a) hTM-Wrb binds specifically to human RBC with detectable binding at

Description: Abstract 3351 The thrombomodulin-protein C pathway has important antithrombotic and anti-inflammatory roles in a variety of disease models and human illnesses, including severe sepsis, acute lung injury, and focal ischemia & reperfusion. We recently reported that anchoring recombinant thrombomodulin (TM) to PECAM-1 on the luminal surface of the vascular endothelium is protective in a mouse model of inflammatory lung injury. To this point, it has been unclear whether targeted thrombomodulin is able to partner with the endothelial protein C receptor (EPCR) in the same way as its endogenous counterpart. Here we demonstrate that anchoring TM to endothelial ICAM-1, rather than PECAM-1, results in approximately 10-fold greater activation of protein C. Furthermore, blocking protein C binding to EPCR results in marked reduction of protein C activation when TM is targeted to ICAM-1, whereas protein C activation is largely independent of EPCR when PECAM-1 is used as the target ligand. Consistent with this in vitro observation, anti-ICAM/TM fusion protein provides greater in vivo protection in a mouse model of inflammatory lung injury than its PECAM-targeted analogue, more potently blocking expression of pro-inflammatory cytokines and more effectively stabilizing endothelial barrier function. Since ICAM, endogenous TM, and EPCR are all thought to localize to microdomains on the apical surface of endothelial cells, we hypothesize that ICAM-targeting more effectively mimics the natural configuration and brings TM within sufficient proximity to allow access to its membrane co-factor. These observations could have profound implications for the therapeutic efficacy of a whole series of endothelial-targeted proto-drugs and their potential for translational success. Disclosures: No relevant conflicts of interest to declare.

Description: Delivery of bio-therapeutics by red blood cells (RBCs) can greatly enhance pharmacokinetics and pharmacodynamics of the appended or loaded agents, and may even potentiate induction of immunologic tolerance. Our group and others have successfully used fusion proteins, antibodies, and peptides to couple therapeutics to murine, but not human, RBCs. It is known that extracellular ligands have the potential to induce marked, epitope-dependent changes in red cell physiology, including changes in deformability, phosphatidyl-serine (PS) exposure, and reactive oxygen species (ROS) production, particularly for agents targeted to glycophorin A and Band 3, two highly-expressed membrane proteins. To produce clinically translatable strategies for human RBCs, it is critical to identify optimal red cell target epitopes, understand their effects on red cell physiology, and create humanized or human-like ligands to minimize immunogenicity. We constructed single chain antibodies (scFv) against antigenic determinants on Band 3 protein (Wrb) and RHCE protein (Rh17/Hr0) on human erythrocytes using phage display libraries prepared from immunized cynamolgous macaques (Macacafascicularis). Both these antigens are present on essentially 100% of the human population. Unfused scFvs were produced in E.coli while fusions of scFv with the extracellular domain of human thrombomodulin (TM-scFv) were produced in Drosophila S2 cells. Binding of recombinant proteins to human RBCs was measured by radioimmunoassay and flow cytometry. Generation of activated protein (APC) by RBCs loaded with TM-scFv fusions was measured by colorimetric assay. RBCs pre-incubated with varying concentrations of anti-Band3 and anti-RHCE fusions were assessed for osmotic resistance and mechanical integrity by exposure to hypo-osmolar medium and rotation in the presence of glass beads, respectively. PS exposure was measured by annexin V binding, and ROS generation was measured by dihydrorhodamine-associated fluorescence. Effects on RBC rheology were measured by flowing through microfluidic channels under controlled shear rates. Efficacy of TM-scFv fusions in diseased micro-vessels was assessed using a TNF-alpha stimulated, endothelialized microfluidic model. Single-chain antibody fragments and TM fusion proteins targeted to conserved epitopes on Band 3 protein and RHCE protein bound to human, but not murine or porcine, RBCs with high specificity and affinity (~50 nM), and in numbers consistent with the expected level of target expression (105 and 106 copies/RBC for RHCE and Band3, respectively). Coating RBCs with proteins targeted to Band 3 lessened RBC hypo-osmolar hemolysis (20% reduction) but increased hemolysis (2-fold) under mechanical stress, changes compatible with decreased red cell deformability. Proteins targeted to RHCE did not induce significant changes in hemolysis of RBCs under either osmotic or mechanical stress. Targeting neither Band 3 nor RHCE induced significant exposure of PS or production of ROS. Target-dependent effects on RBC rheology were observed under varying shear stresses in a microfluidic system. Fusion proteins of TM targeted to both epitopes demonstrated dose- and surface-copy-number-dependent generation of APC in the presence of PC and thrombin. Both TM-scFv fusion proteins were efficacious in a microfluidic model of disseminated intravascular coagulation using whole human blood by demonstrating near complete abrogation of fibrin generation in response to endothelial activation with TNF-alpha. In summary, we designed human RBC-specific non-human primate single chain antibody fragments capable of fusion to therapeutic cargoes. The TM-scFv fusions maintained therapeutic activity when bound to human RBCs and showed effective thromboprophylaxis in a whole-blood model of vasculitic injury. These antibodies and fusion proteins bound to erythroid-specific epitopes, and demonstrated target-dependent effects on several aspects of red cell physiology. The non-human primate origin of the antibodies should minimize their potential immunogenicity and the findings provide a platform to translate red cell targeted drug delivery into the clinical realm. Disclosures No relevant conflicts of interest to declare.

Description: * These authors contributed equally to this work The microvasculature plays a key role in the pathogenesis of sepsis, ARDS, multiorgan dysfunction, and a variety of other human diseases characterized by a pro-coagulant, pro-adhesive endothelial phenotype. The complex interactions that occur at the interface of blood and activated endothelium are difficult to resolve in animal models and challenging to recreate in cell culture systems. While individual processes - e.g., leukocyte adhesion and transmigration - have been extensively studied, development of a well-characterized model integrating the full range of pathogenic processes (coagulation, barrier dysfunction, innate immune system activation, etc.) remains an important unmet goal, which could both elucidate mechanisms of disease and aid in the design and testing of putative therapeutics. To this end, we sought to model an inflamed vascular segment using a Fluxion Bioflux system. 3-dimensional confluent endothelial cell (EC) monolayers were established within fibronectin-coated laminar flow chambers. ECs were flow adapted, activated with TNFa, and then perfused with whole blood (WB) at a variety of shear stresses. Real time fluorescence microscopy allowed continuous monitoring of fibrin deposition and leukocyte and platelet adhesion. The multi-channel format, which allows simultaneous testing of multiple conditions with replicates, proved to be a critical asset, given substantial day-to-day variability. Both fibrin deposition and adhesive events showed dependence on dose (1 vs. 10 ng/mL) and duration (4 vs. 6hr) of TNF activation. Confocal microscopy revealed TNF-dependent, increases in EC expression of ICAM-1, VCAM-1, and tissue factor (TF), as well as suppression of endothelial thrombomodulin (TM). Activation of the coagulation system was completely abrogated by treatment of the EC monolayer with a TF-inhibiting antibody, suggesting a primary role for the extrinsic pathway. Hirudin also limited fibrin deposition when added to whole blood prior to perfusion, although "breakthrough clotting" was seen in some channels. Finally, the role of endothelial TM was investigated in several ways, including by the use of a blocking antibody, which prevents thrombin binding. Treatment of ECs with this antibody markedly increased fibrin deposition, whereas TM/R6.5 scFv, a novel targeted fusion protein therapeutic, which anchors recombinant TM to endothelial ICAM-1, inhibited fibrin deposition upon subsequent infusion of WB. Neither soluble TM (sTM) nor anti-ICAM-1 R6.5 scFv alone had any effect on coagulation when infused in this setting (i.e., prior to WB) and even addition of a large excess of sTM to whole blood was less effective in reducing TNF-dependent fibrin deposition than pre-treatment with the ICAM-targeted TM fusion protein, indicating potential importance of precision drug delivery on the microscale. In summary, the described microfluidic, "endothelialized", whole blood model of an inflamed microvessel may prove useful in interrogating specific aspects of a variety of vascular pathologies and in devising and improving therapeutic interventions. Figure 1. (a) Fibrin deposition and (b) leukocyte and platelet adhesion upon whole blood perfusion of endothelialized microchannels, pre-activated with 1 vs. 10ng/mL of TNF for 6 hours. (c) Role of endothelial TM in TNF-dependent microvessel thrombosis. TM blockade exacerbates fibrin deposition whereas ICAM-targeted TM fusion protein effectively eliminates coagulation. Figure 1. (a) Fibrin deposition and (b) leukocyte and platelet adhesion upon whole blood perfusion of endothelialized microchannels, pre-activated with 1 vs. 10ng/mL of TNF for 6 hours. (c) Role of endothelial TM in TNF-dependent microvessel thrombosis. TM blockade exacerbates fibrin deposition whereas ICAM-targeted TM fusion protein effectively eliminates coagulation. Disclosures No relevant conflicts of interest to declare.

Description: Despite continued achievements in antithrombotic pharmacotherapy, difficulties remain in managing patients at high risk for both thrombosis and hemorrhage. Utility of antithrombotic agents (ATAs) in these settings is restricted by inadequate pharmacokinetics and narrow therapeutic indices. Use of advanced drug delivery systems (ADDSs) may help to circumvent these problems. Various nanocarriers, affinity ligands, and polymer coatings provide ADDSs that have the potential to help optimize ATA pharmacokinetics, target drug delivery to sites of thrombosis, and sense pathologic changes in the vascular microenvironment, such as altered hemodynamic forces, expression of inflammatory markers, and structural differences between mature hemostatic and growing pathological clots. Delivery of ATAs using biomimetic synthetic carriers, host blood cells, and recombinant fusion proteins that are activated preferentially at sites of thrombus development has shown promising outcomes in preclinical models. Further development and translation of ADDSs that spare hemostatic fibrin clots hold promise for extending the utility of ATAs in the management of acute thrombotic disorders through rapid, transient, and targeted thromboprophylaxis. If the potential benefit of this technology is to be realized, a systematic and concerted effort is required to develop clinical trials and translate the use of ADDSs to the clinical arena.