Harvey Jacobs

Design of nonthrombogenic polymer surfaces for blood-contacting medical devices.

Although significant progress has been made in the design of blood-compatible polymers in the past decades, there is no ideal polymer surface which is comparable with a natural endothelial surface in preventing surface-induced thrombosis and maintaining hemostasis. This is due to the complex pattern of protein and cellular interactions with foreign surfaces, which still demands defining a proven hypothesis to develop non-thrombogenic surfaces. Synthesis of new polymers with optimal mechanical properties and the in vitro and in vivo characterization of these surfaces will require many more years of work. In this article, the surface modification of existing medical polymers for the improvement of blood compatibility is introduced. Surface immobilizing of heparin onto polyurethane, coatings of a polyurethane-poly(ethylene oxide)-heparin graft copolymer, and a coating of thermosensitive polymers on polyurethane will be discussed. All modified surfaces demonstrated superior blood compatibility both in vitro and in vivo. The biological response of these designed systems in vitro, ex vivo and in vivo should provide state-of-the-art materials for the specific application of controlling thrombosis and solving biocompatibility problems.

Heparin surface immobilization through hydrophilic spacers: Thrombin and antithrombin iii binding kinetics

The immobilization of heparin onto polymeric surfaces using hydrophilic spacer groups has been effective in curtailing surface induced thrombus formation. In this study, the effect of hydrophilic spacers (PEO) on the binding kinetics of immobilized heparin with antithrombin III (ATIII) and thrombin was investigated. Monodispersed, low molecular weight heparin was fractionated on an ATIII affinity column to isolate high-ATIII affinity heparin. This high-ATIII affinity fraction was immobilized onto a styrene/p-amino styrene random copolymer surface using hydrophilic poly(ethylene oxide) (PEO) spacer groups. Styrene/p-amino styrene random copolymer was chosen as the model surface to provide quantitative and reproducible surface concentrations of available amine groups, grafted PEO spacers, and immobilized heparin. The polymer substrate was coated onto glass beads, tolylene diisocyanate modified PEO was covalently coupled to the surface, followed by heparin immobilization. The bioactivity of immobilized heparin was 16.2%, relative to free heparin, and a 1:1 binding ratio between heparin and PEO was achieved. The binding of ATIII and thrombin to control surfaces (no heparin), soluble heparin, heparin immobilized directly onto the surface, and heparin immobilized via spacer groups, were compared. Soluble heparin bound both thrombin and ATIII, while heparin immobilized directly onto the surface bound only thrombin. Spacer-immobilized heparin bound both ATIII and thrombin, although to a lesser extent than soluble heparin. Thus, the enhanced bioactivity of spacer-immobilized heparin, compared to direct-immobilization, may be attributed to the retention of ATIII binding.

Stabilization of insulin by alkylmaltosides. A. Spectroscopic evaluation

Abstract The aggregational behavior of native bovine insulin has been studied by circular dichroism and quasi elastic laser light scattering. The influence of a homologous series of alkylmaltosides on the self-association and aggregation of insulin was investigated. Significant stabilization of insulin was observed for the alkylmaltosides. Circular dichroism revealed that dodecylmaltoside was the most promising compound from a stabilizer stand point. Using dodecylmaltoside as the model compound, it was found that micelles were formed in the concentration range of the experiments and micell formation appeared to be important for the stabilization of insulin. Moreover, the results may suggest that insulin partitions into the micelles in an equimolar fashion and that insulin in this ‘mixed micellar’ form is relatively stable.

Binding kinetics of thrombin and antithrombin III with immobilized heparin using a spacer

The immobilization of heparin onto polymeric surfaces using a hydrophilic spacer was effective in curtailing surface induced thrombus formation. In this study, the binding kinetics of immobilized heparin with antithrombin III (ATIII) and thrombin were investigated. Low molecular weight heparin (molecular weight, 6,000 daltons) was fractionated on an ATIII affinity column, and it was immobilized onto a styrene/p-amino styrene random co-polymer surface via hydrophilic spacer groups. This polymer substrate was coated onto glass beads (diameter range, 0.088-0.105 mm). PEO (molecular weight 3,400), modified by tolylene diisocyanate, was covalently coupled as a spacer group, followed by heparin. The bioactivity of immobilized heparin was approximately 16.2%, relative to free heparin, and nearly 1:1 binding between heparin and PEO was calculated. The binding constants of immobilized heparin and ATIII, and immobilized heparin and thrombin, were 0.958 x 10(7) M-1 and 1.76 x 10(8) M-1, respectively. The immobilized heparin bound with both ATIII and thrombin, and the binding mechanism was similar to that of free heparin.

Effect of fibronectin on the binding of antithrombin III to immobilized heparin.

An objective of this research is to verify the mechanism of anticoagulant activity of surface-immobilized heparin in the presence of plasma proteins. The competition and binding interaction between immobilized heparin and antithrombin III (ATIII)/thrombin have been described in vitro. However, the strong ionic character of heparin leads to its specific and nonspecific binding with many other plasma proteins. Most notably, fibronectin contains six active binding sites for heparin which may interfere with the subsequent binding of heparin with ATIII or thrombin. Heparin was covalently immobilized through polyethylene oxide (PEO) hydrophilic spacer groups onto a model surface synthesized by random copolymerization of styrene and p-aminostyrene. The binding interaction of immobilized heparin with ATIII was then determined in the presence of different fibronectin concentrations. The binding interaction was studied by first binding immobilized heparin with ATIII, followed by the introduction of fibronectin; heparin binding with fibronectin, followed by incubation with ATIII, and simultaneous incubation of surface immobilized heparin with ATIII and fibronectin. The extent of ATIII binding to heparin in each experiment was assayed using a chromogenic substrate for ATIII, S-2238. The results of this study demonstrate that the displacement of ATIII from immobilized heparin was proportional to the fibronectin concentration, and was reversible. Furthermore, the binding sequence did not play a role in the final concentration of ATIII bound to immobilized heparin.

Anti-GAD monoclonal antibody delays the onset of diabetes mellitus in NOD mice.

Insulin Dependent Diabetes Mellitus (IDDM type I) is the result of autoimmune destruction of insulin producing pancreatic β-cells by the cellular immune system, specifically, autoreactive T cells. Disease progression is evident by multiple autoantibodies responding to self-antigens in a cascade mechanism, wherein the first self-antigen induces the activation of the immune system, leading to the destruction of β-cells and consequently, exposure of other antigens. Glutamic Acid Decarboxylase (GAD) is recognized in the literature as a primary autoantigen involved in the cascade. We questioned the immunological involvement of this autoantigen in the overall progression of the disease, specifically if antigen recognition by the cellular immune system (T cells) is necessary for organ specific autoimmunity and cellular toxicity. We tested this hypothesis by isolating, purifying and injecting monoclonal antibodies against GAD (anti-GAD Ab; 0.1 mg or 0.3 mg) into non-obese diabetic (NOD) mice on a weekly basis. We suggest that the anti-GAD Ab will bind to the GAD antigen, or perhaps bind to the epitope presented in association with APC-MHC and prevent T cell recognition, thereby delaying disease onset. Our results demonstrate a delay in the onset of diabetes and a decrease in the severity of insulitis in our test animals, when compared to controls. The mechanism of action of the anti-GAD Ab may be associated with a passive protection mechanism, as evidenced by the fact that splenocytes transferred from anti-GAD Ab treated mice did not prevent or delay diabetes in syngeneic irradiated NOD mice. The mechanism of diabetes prevention by administration of anti-GAD antibody could be associated with an interference in recognition of GAD by T cells, and continuing research will be perform to investigate this hypothesis.

Heparin immobilization by surface amplification

A method to increase the amount and improve the bioactivity of heparin (HEP) immobilized on a polymer surface was developed. The surface of polyurethane-urea (PU) coated glass beads was first modified with diisocyanates, followed by surface grafting of polyfunctional polymers (PFP), including: poly(vinyl alcohol), poly(ethyleneimine), and poly(allylamine). The functional groups of the surface grafted PFP (-OH, -NH, or -NH2) were modified with diisocyanates (TDI) to amplify the surface concentration of isocyanate groups, alpha, omega-diamino-terminated polyethylene oxide (PEO; molecular weight, 4,000 daltons) was then coupled to the surface grafted PFP, and the free amino groups were derivatized with TDI. Finally, HEP was coupled to the amplified surface through free -NCO groups of PU-PFP-PEO. The surfaces were quantified during each step of the procedures for -NCO groups and HEP. All grafted surfaces showed a four to eightfold increase in -NCO content and a twofold increase in immobilized HEP content compared with HEP immobilized directly onto the PU surface. The HEP bioactivity tests (including activated partial thromboplastin time, thrombin times, and factor Xa) demonstrated an increased bioactivity of HEP when immobilized through PFP-PEO compared with PFP and PU alone.

Stabilization of insulin by alkylmaltosides. B. Oral absorption in vivo in rats

Enteral absorption of insulin is hampered by instability and self-association, degradation of insulin by digestive enzymes and by low macromolecular permeability. Reduction of the influence of these factors through protein stabilization should hypothetically result in increased absorption due to a higher concentration gradient of intact insulin across the intestinal mucosal barrier. Insulin in a stabilized form was shown to be absorbed after duodenal administration in normoglycemic and in diabetic rats. A homologous series of alkylmaltosides were found to stabilize insulin in solution (Hovgaard et al., 1996). For dodecylmaltoside, only minimal aggregation was observed over extended periods (60 days) under agitating conditions. In comparison, regular insulin aggregated and lost complete biological activity after 8 days. In an intraduodenal rat model, blood glucose levels were depressed to 70% of initial values and serum insulin concentrations reached 250 μU/ml. The bioavailability of stabilized dodecylmaltoside insulin was found to be 0.5-1% based on area under the curve (AUC) determination for plasma insulin levels and decreased AUC (dAUC) for blood glucose level depression.

PGE1—Heparin conjugate releasing polymers

Abstract Previous work in our laboratories has shown prostaglandin E 1 (PGE 1 ) releasing polymers reduce platelet adhesion and aggregation. Heparin released from polyurethane maintained bioactivity as measured by APTT, TT and Factor Xa assays. In in vivo animal studies using a dog model, heparin releasing polyurethane catheters showed small fibrin formation compared to the untreated polyurethane catheter. However, platelet adhesion and aggregation on both the heparin releasing system and untreated catheter were similar, which indicates significant thrombus was still present on the surfaces. A covalently bonded conjugate of commercial grade heparin and (PGE 1 ) was synthesized for use as a controlled releasing system for blood contacting sur faces in order to improve the blood compatibility of polymer surfaces. The compound was synthesized using a modified mixed carbonic anhydride method of amide bond formation between the carboxylic acid moiety of PGE 2 and a primary amine group on heparin. Quantitation of coupling was measured by spectroscopically monitoring the PGB 1 conjugate. Bioactivity tests on the conjugates (APTT and platelet aggregation) confirmed that the antithrombic activity of heparin was maintained. However, PGE 1 bioactivity, as measured by ADP induced platelet aggregation, had decreased, but was still active. Rabbit arteriovenous (A-V) shunt experiments revealed that heparin—PGE 1 releasing polyurethane demonstrated the prevention of both fibrin formation and platelet interactions. This approach can be utilized for the design of nonthrombogenic medical devices in contact with blood.

Self-Regulated Insulin Delivery-Artificial Pancreas

AbstractA chemical mediated artificial pancreas has been designed that will deliver glycosylated insulin (G-Ins) in response to glucose concentrations. This system is based on the competitive binding between G-Ins and glucose to a saccharide binding substrate, Concanavalin A (Con-A). Con-A has was evaluated as a soluble monomeric protein, enclosed within microcapsules, formed into microspheres, and synthesized as a soluble high molecular weight oligomer. These forms were designed to maintain the binding characteristics of Con-A and G-Ins and to prevent leakage or permeation of Con-A across the device membrane.